OPERATIONS RESEARCH gael bal SOLUTIONS ee oil age , mT DP: 49 vod BE / eee [alco elie a oe JK SHARMAOPERATIONS RESEARCH PROBLEMS AND SOLUTIONS Third Edition JK SHARMA Formerly, Professor Faculty of Management Studies University of Delhi Delhi MACMILLANFA © Macmillan Publishers India Led, 2002, 2004, 2009 ‘All rights reserved, No part of this publication may be reproduced or transmitted, in any form or ty any means, without permission. Any pperson who doc: any unauthorised actin relation to this publication ‘maybe lable to criminal prosecution and civil claims for damages. Firs published, 2002 Second edition, 2004 ‘Reprinted, 2005, 2006, 2007 twice), 2008 Thirdedtion, 2009 _ Reprined, 2009 2010 MACMILLAN PUBLISHERS INDIA LTD Delhi Bangalore Chennai Kolkaia Mumbsi Ahmedabad Bhopsl Chandigarh Coimbatore Cuttack Guwahati Hubli Hyderabad Jaipur Lucknow Madurai Nagpur Patna Pune Thirwvananthapuram Visakhapatnam ‘Companies and representatives throughout the world ISBN 10:0230-63659-4 ISBN 13:978-0230-63659-0 Published by Rajiv Beri for Macmitlan Publishers India Lid 2/10, Ansari Road, Daryaganj, New Delhi 110 002 Printed by S.M Yogan at Macinillan India Press, Chennai ~ 600 041 This took ismean for tuctional ad leaing pups. The ath) ofthe ook haha taken ll reasonable ae tenure tat the content of he Doak dona ose any eising coyigs or ober inelctal propery rigs of ay penn nny manner whatsoever. Inthe eventhe ano) basave ‘been umbleto ack ay Source and any copyeight bs been nave ininged lease oy he publ in wring for conecive acon.Preface to the Third Edition Preface to the First Edition Brief Contents ‘Chapters Solved ‘Self Practice Pages Examples Problems 1 Operations Research: An Introduction — = 110 2. Linear Programming: Model Formulation 6 % 11-8 3, Linear Programming: Graphical Solution Method 2 Js 6591 4. Linear Programming: Simplex Method 2 B 9471 5. Revised Simplex Method 5 Jo IBID 6. Dual Simplex Method 5 16 193203 2. Bounded Variables LP Problems 3 a 205218 &_ Duality in Linear Programming B 2 219-239 9. Sensitivity Analysis ary 2 241-270 10, Parametric Linear Programming 4 8 2285, 11, Integer Linear Programming 2 a IAS 12 Goal Programming 6 a U3 13, Transportation Problem 7 » 329-373 14, Assignment Problem 1s 2 378407 15, Decision Theory and Decision Tree Analysis, 0 6 409457 16, Theory of Games 2 B 459494 17, Project Management: PERT and CPM 35 ® 495569 18 Inventory Control a Pv STI-638 19, Queuing Theory a 2 639685 20. Replacement and M: 2 35 (67-122 21, Markov Chains n ” 23146 22. Simulation B 2 147183 2B, Sequencing Problem 15 19 785810 24, Information Theory 8 B 811-818 25, Dynamic Programming m » 819-856 26, Non-Linear Programming Methods B a 857888 Appendix 889-894Operations Research: An Introduction emetoues The Historical Development Significance of Operations Research Features of Operations Research Approach. Definitions of Operations Research ‘Scientific Method in Operations Research Models and Modelling in Operations Research Advantages of Models ‘Methods for Solving Operations Research Models ‘Methodology of Operations Research ‘Advantages of Operations Research Study Foatures of Oparations Research Solution Applications and Scope of Operations Research Review Questions THE HISTORICAL DEVELOPMENT Itis generally agreed that Operations Research (OR)came into existence as a discipline during World War Il. However, 1 particular modeland technique of OR ean be traced back much carlier. The tem operations rescarch was coined as result ofresearch on military operations during this war Since the war involved strategic and tactical problems which ‘were so complicated, that to expecting adequate solutions fiom individuals or specialists in a single discipline was unrealistic. Therefore, groups ofindividuals who collectively were considered specialists in mathematics, economics, statistics and probability theory, engineering, behavioural and physical science were formed as special units within the armed forces to deal with strategic and tactical problems of various military operations. Such groups were first formed in England and the United States. One of the groups in England came to be known as Biackett’s Circus, This group, under the leadership of Prof. P M S Blackett was attached to the Radar Operational Research unit and assigned the problem of analyzing the coordination of radar equipment at gan sites, Following the success of this group, such mixed-team approach was also adopted in other allied nations. After the war ended, scientists who had been active in the military OR groups made their efforts to apply the ‘operations research approach to civilian problems, related to business, industry, research and development, etc. A key person in the post-war development of OR was George B Dantzig. In 1947, he developed linear programming and its solution method known as simpler method. Besides linear programming, many other tools of OR such as statistical quality control, dynamic programming, queuing theory and inventory theory ‘were well-developed by end of the 1950.image not availableimage not availableimage not availablees | subject to restrictions on resources in the context of the problem under consideration and non-negstivity of variables. Static Models Static models present a system at some specitied time and do not account for changes over time. For example, an inventory model can be developed and solved to determine an economic order quantity for the next period assuming that the demand in planning period would remain the same as that for today. Dynamic Models Ina dynamic model, time is considered as one of the variables and admit the impact of changes generated by time in the selection of the optimal courses of action. Thus, a sequence of interrelated decisions are made toachieve the optimal course of action o optimize the given objective, Dynamic programming isan example of a dynamic model Deterministic Models 1fall the parameters, constants and functional relationships are assumed to be known with certainty when the decision is made, then the mode is said to be deterministic. Thus, in such a case, the outcome associated with a particular course of action is known. That is, fora specific set of input values, there is ‘a uniquely determined output which represents the solution of the model under conditions of certainty. The results of the models assume single value, Linear programming modelsare examples of deterministic models. Probabilistic (Stochastic) Models Models in which at least one parameter or decision variable is a random variable are called probabilistic (or stochastic) models. Since at least one decision variable israndom, therefore, an independent variable which isthe function of dependent variable(s) will also be random. This ‘means consequences or payoff due to certain changes in the independent variable cannot be predicted with certainty. However, it is possible to predict a pattern of values of both the variables by their probability distribution Insurance against risk of fire, accidents, sickness, ete, are examples where the pattem of events is studied in the form of a probability distribution. Analytical Models ‘These models havea specitic mathematical structure and thus can be solved by known analytical or mathematical techniques. Any optimization model (which requires maximization or minimization of ‘an objective function) is an analytical model. Simulation Models These models also have a mathematical structure but are not solved by applying ‘mathematical techniques to get a solution. Instead, a simulation model is essentially a computer assisted ‘experimentationon amathematical structure of real-life problemin order to describe and evaluateits behaviour under certain assumptions over a period of time. Simulation models are more flexible than mathematical ones and therefore, can be used to represent a ‘complex system which otherwise cannot be represented mathematically. These models do net provide general Solutions like those of mathematical models. ‘ADVANTAGES OF MODELS Models in general, are used as an aid for analyzing complex problems. However, amodel can also serve other purposes as given below: 1. A model provides economy in representation of the realities of the system. That is, models help the decision-maker to visualize asystem so that he can understand system’s structure or operation in a better ‘way. For example, itis easier to represent a factory layout on paper than to construct it It is cheaper to try ‘out modifications of such systems by rearrangement on paper. ‘The problem can be viewed in its entirety, with all the components being considered simultaneously. 3, Models serve asaidsto transmit ideas and visualization among people in the organization. For example, a process chart can help the management to communicate about a better work method to shopfloor workers. 4. A model allowsus to analyse and experiment in a complex situation to a degree that would be impossible in the actual system and its environment. For example, the experimental firing of INSAT satellitemay cost millions of rupess and require years of preparation, 5. Models simplify the investigation considerably and provide a powerful and flexibletool for predicting the future state of the process or system,image not availableimage not availableimage not availableProduction Management (a) Facilities Planning + Location and size of warehouse, distribution centres and retail outlets (>) Manufacturing ‘+ Aggregate production planning, assembly line, blending, purchasing, inventory control ‘+ Employment, raining, layoffs and quality contro! (©) Maintenance and Project Scheduling, ‘+ Maintenance policies and preventive maintenance ‘© Maintenance crew size and scheduling Personnel Management + Manpower planning, wage/salary administration + Skillsand wages balancing Techniques and General Management + Decision support systems and MIS; forecasting + Organizational design and control Government + Economical planning, natural resources, social planning, energy Operations Research: An Introduction | of Logistics, layout, engineering design Transportation, planning, scheduling Project scheduling and allocation of resources, Scheduling of training programmes Project management, strategic planning. Urban and housing problems Military, police, pollution control, etc. REVIEW QUESTIONS |, Briefly ace the history of operations research. What happened 10 ‘the development of operations research after World War II? 2. Is operations research a discipline or a profession or a set of teciniques ora philosophy ora new name for old things? Discuss, 3. What are the situations (areas) where OR techniques will be applicable? 4. Discussthe following: (@) OR is an interdisciplivary approach, (b) Scientific methodin OR, (©) ORis more than ¢ quantitative analysis ofthe problem. ‘5. What are the essential characteristics of operations research? Mention different phases in an operations research study. Point out its limitations, if any. ICA, Now. 1992} 6. (@) Outline the broad features of the judgement phase and the research phase of the scientific method in OR. Discuss in detail any one of these phases. (b) State the phases of OR study and their importance in solving problems. 7. Define operations research. Explain critically the limitations of | various definitions as you understand them. 8. Give any three definitions of operators research and explain them, [Meerut Unis. 1PM, 1991; MSe (OR), 1990), ‘onseapz 9. Discuss the role and scope of quantitative methods for scientific decision-making in abusiness envionment, [Dethi Uni, MBA, 1995) 10, Operations research isan aid forthe executive inmaking hs decisions by providing him the needed quanvitative information, based on scientifie method analysis. Discuss, 11, What were the significant characterises of OR applications during ‘World War II? What caused the discipline of OR to take on these ‘characteristics during that period? 12, Comment on the following statements (2) OR isthe art of winning war without actually fighting it. (b) OR isthe art of finding bad answers where worse exist (©) OR replaces management by personality [Dethi Univ, MBA (HCA), 1990} 13. Discuss the significance and scope of operators researchin modern ‘management, [Delhi Uni, MBA (HCA), 1995, 97, 2001) 14, Quantitative techniques complement the experience and judgement ‘fan executive in decision-making, They do not and cannot replace it. Discuss, 15, Decision-makers are quick to claim that quantitative analysis talks tothem in ajargon that does not sound like English, List fourterms that might not be understood by a manager. Then explain in non: technical terms what each term means.image not availableimage not availableimage not availableLinear Programming: Model Formulation 1. Linear programming rests all relationships among decision variablesas linear. However, generally, nether the objective functions nor the constraints in real-life situations concerning business and industrial problems are linearly related to the variables. 2 While solving an LP model, there is no guarantee that we will get integer valued solutions. For example, in finding out how many men and machines would be required to perform a particular job, a non-integer valued solution will be meaningless. Rounding off the solution to the nearest integer will not yield an optimal solution. In such eases, integer programming is used to ensure integer value to the decision variables. 3. Linear programming madel doesnot take into consideration the effectof time and uncertainty. Thus, the LP ‘mode! should be defined in sucha way that any change due to internal as well as external factors can be incorporated 4, Sometimes large-scale problemscan be solved with linear programming techniques even when assistance of computer is available. For it, the main problem can be fragmented into several small problems and solving each one separately. ‘5. Parameters appearing inthe model are assumed to be constant but in real-life situations, theyare frequently neither known nor constant. 6 Itdeals with only single objective, whereas in real-life situations we may come across conflicting multi- objective problems. In such cases, instead of the LP model, a goal programming model is used to get satisfactory values of these objectives. GENERAL MATHEMATICAL MODEL OF LINEAR PROGRAMMING PROBLEM ‘The general tinear programming problem with » decision variables and m constraints can be stated in the following form: Optimize (Max or Min) Z=¢, x, Fey +... +643, subject to the li sar constraints, ay + aint to + aiyhy Gm 2dD, Oy, + Gy9Xq +... + ag, BA 2DO, Opt Fr + Oy X3 + os + Oem Sy 752) Bq and poy nt20 where c;'s are coefficients of decision variable 3, representing their per unit contribution to the value of objective function. The a,’sare called the echnological coefficients ot input-output coefficients, and represent the amount of resources Consumed per unit of variable x,. In the given constraints, the a,”s can be positive, negative or zero. The bs represent the total availability ofthe ith resource. The term resource is used in a very ‘general sense to include any numerical value associated with the right hand side of a constraint. It is assumed that b, > 0 for all i. However, ifany 6, <0, then both sides of constraint can be multiplied by ~ 1 to make +,> Oand reverse the inequality of the constraint. GUIDELINES ON LINEAR PROGRAMMING MODEL FORMULATION ‘Step 1: Define Decision Variables (@)_ Express exch constraintin words, For this firstsce whether the constraintis ofthe form, > (at least as large 3 as), or ofthe form < (nomore than) or = (exactly equal to). (b) Then express the objective function in words. (©) Step 1(a) and (b) should then allow you to verbally identify the decision variables. If there are several decision alternatives available, then to identify the decision variables you have to ask ‘yourself the question: What decisions must be made in order to opiimize the objective function? Having accomplished Step I(a) through (c), decide symbolic notation for the decision variables and specify their units of measurement. Such specification and units of messurement would help in interpreting the final solution of the LP problem.image not availableimage not availableimage not availablei ei a has available in each period 2,000 hours of machine time and 1,400 hours of assembly time. The production of ‘each C, requires 3 hours of machine time and 2 hours of assembly time, whereas the production of each C, requires hours of machine time and 3 hours of assembly time, Formulate this problem as an LP model so as to ‘maximize the total profit to the company. LP Model Formulation The data of the problem is summarized as follows: Resources/Constrainis Components , Total Availabilty Budget (Rs.) to/unit 440/unit Rs. 4000 Machine time 3 has/unit 2hysiunit 2,000 hours, Assembly time 2 hrs/unit 3 hisiunit 1,400 hours Seling price Rs, 30 Rs, 70 Cost (wages + materia Rs. 10 Rs. 40 Decision variables Let x, andx, The LP mode! Maximize (total profit)Z= Selling price ~Cost price =(30-10)x, +(70-40)x, = 20x, +30x, subject to the constraints number of units of components C, and C, tobe produced, respe (i) The total budget available (ii) Productiontime 10x, +40x, $ 4,000 3ry+ ry S 2,000 and ¥,20 2e,+ 3x, < 1.400 Example 2.7 A company has two grades of inspectors I and 2, who are io be assigned fora quality control inspection. It is required that at least 2,000 pieces be inspected per 8-hour day. Grade | inspector can check pieces atthe rate of 40 per hour, with an accuracy of 97 percent. Grade 2inspector checks atthe rate of 30 pieces perhour with an accuracy of 95 per cent. The wage rate of aGrade | inspector is Rs. Sper hour while that of Grade 2 inspector is Rs. 4 per hour. An error made by aninspector costs Rs. 3 to the company. There are only nine Grade | inspectors and eleven Grade 2 inspectors available in the company. The company wishes to assign work to the available inspectors $0 a8 to ‘minimize the total cost of the inspection. Formulate this problem as an LP model so as to minimize daily inspection cost. [Delhi Univ., MBA, 2000; NIIFT (Mohali), 2000] LP Model Formulation The data of the problem is summarized as follows Inspector Grade 1 Grade 2 Number of inspectors ° " Rate of checking 40 piecesshe 30 piecestir 1-0.97-0.03 1-098 005 Cost of inaccuracy in checking Rs. 3/piece Rs. Sipiece ‘Wage rate/hour Rs.5 Rs4 Duration of inspection = 8 hrs per day ‘otal pieces which must be inspected = Decision variables Let x, and x, = number of Grades | and 2 inspectors to be assigned for inspection, respectively. The LP model Hourly cost of each of Grades 1 and 2 inspectors can be computed as follows’ InspectorGrade 1: Rs. (5+3 40 0.03)=Rs. 860 InspectorGrade2:Rs.(4+3 «30% 0.05)=Rs. 850image not availableimage not availableimage not availableimage not availableimage not availableimage not availableimage not availableaS Hence, napkins required constraints are written as: 160, x= 120 and x, =60, X447,=90, x5ty= 110, 4475 +2) (Gi) Total Number of Soiled Napkins Constraints Since number of soiled napkins on the jth day is equal to ‘number used on jth day + number left on (j~ I)th day. Hence, Wyte ty, 160; ya tzyty.= 120+ Y; y, 25 +1; = 60+, yg tZytry OV, Yet At y=NO+VG Yet Ze+vg=1ODH I; yy +2, +¥;= 1204 vy But, given that y=. and z, =74~2,=2,=2, = 0. Hence, these constrains are rewritten as follows: 20+ 9,3 py ty 60412; 20+% 100, x, +y,+2)= 120. Example2.18 _Considera company that must produce two products aver a production period of three months. ‘The company can pay for materials and labour from two sources: Companty Funds and Borrowed Funds. The firm faces three decisions: () How many units should it produce of product 1? How many units should it produce of product 2? (ii) How much money should it borrow to support the production of the two products? In making these decisions the firm wishes to maximize the profit contribution subject to the conditions, stated below: (@ Since the company’s products are enjoying a seller’s market. itcan sell as many units as itcan produce. The company would, therefore, ike to produce as many units as possible subject o production capacity and financial constraints. The capacity constraints together with cost and price data, ae given in the tablebelow: (@ The evailable company funds dur- SellingPrice Cast of Required Hours per, ing the production period will be Producr —(Rsunit) Production Untein Depa. Re. 3 lakh. isin Gy ae (A bank will give loans upto Rs. 2 lakhs per production period at an 4 a ” = spe A interest rate of 20 per cent peran- ut ‘ a ‘num provided the company’s acid Available hoursper production «500400200 (quick) test ratio is at least 1 to 1 __Peviod of taree months ‘hile the loan is oustanding. A sim- plified acid test ratio is given by ‘Surplus cash on hand after production + Accounts receivable ‘Bank borrowings + Interest accrued thereon (iv) Also make sure that the needed funds are made available for meeting the production costs. Formulate this problem as the linear programming problem. ICA, Now, 1992] LP Model Formulation Decision variables Let x, andx, = number of units of products 1 and 2 to be produce, and x, = amount (in rupees) to be borrowed. The LP model Maximize (netprofit) Z = [Profit from selling products ~ Cost of borrowed funds} = [14-195 01-3) - 26 subject to the constraints. (0 Production capacity of depit. A : 05x, +03x, £500, B : 0.3x,+0.4r, $400, © : 02x, +0.1x, $200, oniseaysimage not availableimage not availableimage not availablecr oper a oman QE Mackine Product A Rowe Product B Route Product Route Machine Hours Se eee 1 2 3 1 a 1 2 3 Lathes 0.5 07 03 = os 06 os 03 200 Drills 0S 03 02 oa 03 07 04 a 250 Grinders 0.6 04 06 07 os 04 03 = 300 LP Modol Formulation Decision variables Letx, = number of units of product i (i = route (j= 1, 2and 3), respectively Band C) to be manufactured through The LP model (@ Maximize (total sales revenue) Z, = 20(x4, + X4p + X45) + 15C¢) + Xp) + 25% ey + keg +X). ‘subjects to the constraints ( Machine hours available for each machine, Lathes © 05x 4g) +07xp)+03x4;+O.Sxp + 0.6r¢ +0.5t¢p* 03k < 200, Drills: OSty + 03x44 0245+ Ong) +0.3%p + 0.7¥¢y +Obicy +O.Ley $250, Grinders + 06r4, +04 + 06% 43 +0.2p, +0.5tp9 4 0.4, +0.3%¢, $ 300. and x, 2 O forall iandj (b) The fixed order for product A, B and C is 250 units, 200 units and 150 units respectively. Thus on each ‘machine, number of units ofeach produced must be more than or equal . the fixed order size. That is, Lathes : x4, Hi, +X 2280, Drills: xp +%p,2200, Grinders: ¥¢, +¥¢y-+¥42 150. Maximize (total sales revenue) Z, = 250 «20+ 15(X4) +X 4. +X 4,250) +200 * 15+ 10(¥p, +Xp9~ 200) + 150% 25 +20 (xe +.X¢9 +X cq - 150). . subject tothe constraints in (a) and (b) slong with non-negativity conditions (©Since sales constraint on product C is: X¢y + cq *¢q $ 200, Maximize Z>, subject to the constraints in (a), (6) and current sales constraint on product C. © Letx,),x49 and.x,, = oumber of units of product A manufactured during overtime through routes 1, 2 and 3 respectively. Then the objective of maximizing sales revenue will be replaced by profit function as production during overtime is less profitable than regular production, Example 2.23 A person running a warehouse purchases and sells indentical items. The warehouse can ‘accommodate 1,000 such items. Each month, the person can sell any quantity he has in stock. Each month, he ccan buy asmutch as te likes to have instock for delivery atthe end of the month, subjectto a maximum of 1,000 items. The forecast of purchase and sale prices forthe next six Monh,i = tS ‘months is given below: Purnhase price (Rs)e,: 12-4 IT 8D Hat present hehasa stock of sate price (Rs,) 3, BS 16 200 items, what should be his policy? Formulate this problem as an LP model LP Model Formulation Decisionvariables Let x, andy, = number of an item j(= 1, 2, FE 1,2, 146), respectively. Maximize (otal profit) Z = Total seles revenue ~ Total purchase cost = (139) + Sy) + 16, +203, + 2Iy +239) (12x, + be, + 175+ 1984 +20,+ 2x4) 6) purchased and sold in a monthimage not availableimage not availableimage not availablea a a Re Example2.27 PQR Feed Company markets two mixes for cattle. The firstmix, Fetilex, requiresat least twice ‘as much wheat as barley. The second mix, Multiplex, requires at least twice as much barley as Wheat. Wheat ‘costs Rs. 1.50 per ke and 1,000 kg are available this month. Barley costs Rs. 1.25 per kg and 1,200 kg are aveilable. Fertilex sells for Rs. 1.80 perkg upto 99 kgand each additional kgover 99 sells for Rs. 1.65. Multiplex sells at Rs. 1.70 per kg upto 99kg and each additional kg over 99 sels for Rs, 1.55. Bharat farms will buy any and all amounts of both mixes of PQR Feed Company. Formulate this problem as an LP model. LP Model Formulation The data given in the problem can be summarized as follows Ingredienss Feed Mixes Wheat Barley __ Selling Price (Rs.) per kg Fertilex 2 1 1.80 upto 99 kg 1.65 above9 kg, Multiplex 1 2 1.70 upto 99 kg 1.55 above99 kg, Costikg (Rs) 1.50 125 Available quantity (kg) 1,000 200 Decision variable Letx, andx,= quantities (in kg) f feedmixes—Fertilex and Multiplex to be marketed by the company respectively. The LP model Maximize (monthly profit) Z= Total sales revenue ~ Total cost 99 « 1.80 + 1.65 (x, -99) +99 * 1.70 + 1.55(x-99)} = [iso( ced }otas(Sae2a)} =(1.65x, +0415 99+ 1.551, +0.15*99) — Lx + Ln, +1254 +254) =0.23x, +0.22,+29.7 subject to the constraints 2,4! @ Avsilabity ofwheat: $5) +5% $1,000 or 2xy+x, < 3000, 142 (Availabilty ofbarley: 3741 +5%2 $1,200. or x, +2x) S 36,000, and xy20 Examples on Marketing Example 228 An advertising company wishes to plan an advertising campaign in three different media television, radio and a magazine. The purpose of the advertising is to reach as meny potential customers as possible. Following are the results of a market study: Television Prime Day PrimeTime Radio Magazine (@) (Rs) @s) @), Cost of an advertising unit 40,000 75,000 30,000 1,000 ‘Number of potential customers reached per unit. 4,00,000-9,00,000 -$,00,00 -—_2,00,000 ‘Number of women customers reached per unit 3,00,000 _4,00,000 _2,00,000 _1,00,000image not availableimage not availableimage not availableLinear Programming: Model Formulation () Themaximum rupee amount tobe invested in alternative Fis Rs. 250,000. i No more than Rs. 5,00,000 should be invested in alternatives A and B combined. Gi) Total weighted risk should notbe greater than 0.10, where -.- (Amount invested in alternative j) (Risk of alternative j) Tos Seishet Rie Total amount invested in all the alternatives (iv) Forthe sake of diversity, at least 100 shares of each stock should be purchased. (v) Atleast 10 per cent of the total investment should be in alternatives A and B combined. (va) Dividends for the year should be at east Rs. 10,000. Rupees retum per share of stock is defined as price per share in one year hence less current price pershare plus dividend per share. Ifthe objective isto maximize total rupee return, formulate the linear programming ‘model for determining the optimal numberof shares to be purchased in each of the shares under consideration. ‘You may assume that the time horizon for the investment is one year. LP Model Formulation Decision variable Letx,.x,.x¢ x, andxp = number of shares to be purchased ineach of the six alterns- tive investment proposals. The LP Model Maximize (total rupee return) Z = (80 1.80~80+4)x, (100 * 1.07-100+450)x9 + (160 x 1.1 160+ 7.50). + (120 »1.12—120+5.50)3,, + (150% 1,09 -150+5.75) r_+ (200% 1.15~200+0.00) x5. = 14x, +1155 423Sx¢+ 19.9% p+ 19.25x p+ 30Kp, subject to the constraints © 200%, $2,50,000 or xp-$ 1,250. G) 80,-+100xy < $00,000 oF O.8t,+y < 5,000 }) 80, x 0.05 +100, x0,03+ 160-*0.10+ 120%, «020+ 150r, «0.06 +200,» 0.08 $0.10 (80x, + 10039 1603+ 120x5 + 150x,+200x,) OF —4x4—Trg + Oxc Ip ~ 6rp — Arp SO oF dr, + Ixy ~ 12ty4 Org dry 20. (iv) 80x, + 100xy > 0.1 (80x, + 1005+ 160;¢* 1209 * 150g +2005) OF 724+ 90% _~ 165¢~ 12xp— 15xp- 20x 2 O. (0) 4xg+ 48g +7Sxct S.5xp + 5:75xp + Ory > 10,000, 80, + 100x9+ 1604+ 120;p + 150 +2001, <25,00,000 and X40 ¥pe ey Ips py Zp 20. Examples on Finance ‘Example 2.33. An engineering company is planning to diversity its operations during the year 1998-99. The ‘company has allocated capital expenditure budget equal to Rs. §.15 crore inthe year 1998 and Rs, 650 crorein the year 1999. The company has five investment projects under consideration. The estimated net returns at the present value and expected cash expenditures on each project in the tWo years are given in the table: ‘Assume that the retum from a particular project would be in direct proportion to the investment in i, s0 that, for example, i in a — project, say A, 20 percent(of 120in 1998 and Year 1998 Year 1999 (0f320in 1999) is invested, then theresulting 35. 5g. 9 Project Estimated Net Returns (Cash Expenditure (in 000 Rs) (in ‘000 Rs.) 2i0 0 x netretum in it would be 20 percent (of240). ff cap aay oy This ‘assumption also. implies thet a at a individuality of the project should be 150 250 340 ‘ignored. Formulate this capital budgeting 182 ua 474 problem asan LP model to maximize the net return.image not availableimage not availableimage not availableLinear Programming: Model Formulation | 44 LP Model Formulation ‘The data of the problem is summarized as follows: Consirains Cows Hens Crop Extra Hours Total Paddy Bajra Jowar Sept-May June-Aug_—_‘”*llabliy ‘Man-hours Sept-May 1000 40.20 23 1 = 3,300 June-Aug 50 od so 38 40 1 4:000 Land i 1 1 tos - 100 Cow 1 a - = - 32 Heas - 1 =- = - = - 4,000 'Netannual eash income(Rs.) 3,500 200—*1,200 800850 2 3 Decision variables Let x, and.x, =number of dairy cows and laying hens, respectively. Xp X4 and x, =average of peddy crop, bajra crop and jowar crop, respectively. 1X “extra man-hours utilized in Sept-May. xy =extra man-hours utilized in June-Aug, The LP model Maximize (net cash income) Z=3,500x, + 200, 1,200 + 80x, +850x, +24, + 3x, subject to the constraints @ Man-hours: 100x, + 0.6x, +40x, +20r,+25x,+x,=3,500 (Sept-May duration) 50x, +044, + 50K, +35x,+40x,+x,=4,000 (June-Aug duration) (i) Landavailability: 1.5x, +x,+x4+x55100 Gil) Livestock: $32 dairy cows); x, and xp IpXyXyXy%y%720 4,000 (laying hens) Example 2.38 A certain farming organization ‘operates three farms of comparableproductivity.The Farm Usable Acreage ‘Water Available ‘output of esch farm is limited both by the usable (incubic feet acreage and by the amount of water available for 1 400 1,500 irrigation, The data for the upcoming season is as 2 600 2.000 shown in the table: 3 300 900 ‘The organization is considering planting crops ‘hich differ primarily in their expected profit per acre and in their consumption of water. Furthermore, the total acreage that can be devoted to each of the erops is limited by the amount of appropriate harvesting equipment available. Crop ‘Maximum Water Consumption Expected Profit Acreage __(imeubicfeet) __ per Acre(Rs.)_ A 3 B 4 c 3 In order to maintain a uniform workload among the farms, itis the policy ofthe organization that the percentage of the usable acreage planted be the same at each farm. However, any combination of the crops may be grown at any of the farms. The organization wishes to know how much of each crop should be planted at the respective farms inorder to maximize expected profit Formulate this problem as an LP model in order tomaximize expected total profit onionsimage not availableimage not availableimage not availableimage not availableimage not availableimage not availableimage not available1. A wine maker has a stock of three different wines wi Consider a small plant which makes two types of automobile yarts,say A and B. It buys castings thatare machined, bored and polished. The capacity of machining is 25 per hour for and 24 per hour forB, capacity of boring is 28 per hour for A and 35 pec tour for B.and the capacity of polishing is 35 perhour for A and 25 per hour for B. Castings for part A cost Rs. 2 and sell for Rs, 5 eachand those for part B cox Rs. 3 and sel for Rs, Geach. The three machines have running costs of Rs. 20, Rs. 14 and Rs, 17.50 per hour. Assuming tha any combination of parts A and B ean be sold. Formulate this problem as an LP model to etermine the product mix which maximizes profit . On | Octeber, a company received a contact to supply 6,000 nits of a specialized preduct. The tems of contact require that 1,000 units be shipped in October; 3,000 units in November and 2,000 units in December. The company can manufacture 1,590 ints per month on regular ‘ime and 750 units per month in ‘overtime, The manufacturing cost per item preduced during regular time is Rs. 3 and the cost per item produced during overtime is Rs, 5. The monthly storage costs Re. 1. Formulate this problem 48 an LP model o a5 to minimize ‘otal costs {Dethi Unis, MBA, 1999) the following characteristics Wine Proofs Acid(%) Specific Stock Gravity (Gallons) x 7 032 170 20 B 33 0.20 1.08 34 c 32 0.30 1.04 2 ‘A gocd dry table wine should be between 30 and 31 degree proof, it should contain at least 0.25 percent acid and should have a specific gravity of atleast 1.05. The wine maker wishes to blend the three types of wine to produce as large aquantity as possible of a satisfictory dry table wine. However, his siock of wine A must be completely used in the blend because further storage ‘would cause it to deteriorate. What quantities of wines B and C should be used in the blend. Formulate this problem as an model, ‘The XYZ Company assembles and markets two types of transistor radios, A and B. Presently 200 radios of each type are manufactured per week. You are advised to formulate the production schedule which will maximize the profits in the light of the following information: Tune Total ManHHows Average Selling Component of Assembly Man-Minute Price Coz per Time per of Inspection per Radio ——‘Redio’ and Corree- acho (Rs) tion Time (Rs) per Radio A 200 12 10400 B 160 6 35280 ‘The company employs 100 assemblers who are paid Rs. 10 per thous actually worked and who will work up to a maximum of 48 hours per week. The inspectors, who are presently four, have 10. Linear Programming: Model Formulation agreed toa plan whereby they average 40 hours of work per week ‘each, However, the four inspectors have certain other administrative duties which have been found to take up an average of Shours per week between them, The inspectors are each paid a fixed wage of Rs, 800 per week Each radio of either type requires one speaker, the type being the same for each radio. However, the company can obtain a ‘maximum supply of 600 in any one week. Their cost has been included in the components cost given for each radi in the table above. Only speakers actually used need be paid for. The only other cost incurred by the company are fixed overheads of Rs, 20,000 per week, ABC Fonds Company is developing a low-calorie high-protein dietsupplement called Hi-Pro. The specifications for Hi-Pro have ‘been established by a panel ofmedical exper's, These specifications along with the calorie, protein and vitamin content of three basic foods, are given inthe following table: Nutritional Units of Nutritional Elements Basic Elements (per 100 gm Serving of Basic Foods Foods) HiPro ——————— Specifications 1 2 3 Calories 350 250 200300 Proteins 250 300 150.200, ViuminA 100 150 78 100 ViuminC 75 128 150 100 Cost per serving(Rs) 1.50 2.00 1.20 ‘What quantities of foods 1,2, and 3 should be used? Focmulate this problem as an LP model to minimize cost of serving. [Delhi Univ., MBA (HCA), 1995, 2000) ‘Omega leather goods company manufactures two types of leather soccer balls X and Y. Each type of ball requires work by both types of employees: semi-skilledand skilled. Basically, the semi- skilled employees employ machines in the manufacture of the balls, while the skilled employees hand-sew the balls. The available time (per week) for each type of employeeand the time requirement for each type of ball are given below: Memufacturing Time Type of Time Available Requirement (hr) Bepkjes: = Reretemet nach ives “BallX any ee. Semi-skilled 2 3 80 Shiled a é 130 ‘The cost of an hour of semi-skilled labour is Rs, 5.50 and that of ‘an hour of skilled labour is Rs. 8.50. To meet weekly demand requirements, at least 15 balls of type X and at least 10 balls of, type Y must be manufictured. Formulate thie problem as an LP ‘model so as to minimize cost of production. ‘A pharmaceutical company has developed a new pill to be taken by smokers that will nauseate them if they emoke. This ‘new pill is a combination of four ingredients thet are costly and in limited supply, The available supply and costs are as follows:image not availableimage not availableimage not available20 Activity Anticipated Return on Investment (%) A, 10 Ay 2 Ay “ Ay 6 ‘The investor wishes to know how much to invest in each activity to maximize the total return on the investment. Formulate this problem as an LP model [Delhi Univ, MB4, 1990) ‘The board of directors of «company has given approval forthe construction of anew plan, The plant will require an investment ‘ofRs. $0 lakh. The required funds will come from the sale of @ ‘proposed bond issue and loans from two financial corporations. For the company, it will net be possible o sell more than Rs. 20, lakh in bonds a! the proposed rate of 12 per cent. Financial corporation A will Ioan up to Rs. 30 lakh atan interest rate of oper cent but insist that the amount of bond debt plus the mount owned to financial corporation B be no more than twice the amount owedto financial coporation A. Financial corporation B will loan an amount up to the amount leaned by financial corporation A but at an interest rat of 18 per cent. Formulate this problemas an LP model to determine the amount of funds to obvain from each source in manner that minimizes total annual interest charges. (Delhi Univ, MBA, 1989] ‘An investor wishes to diversify his portfolio and make due allowance for tona-temm potentiates, butat the same time wishes to maximize his current dividend income. He has considered ‘vious sccuritics in which he might invest, and has classified them into four types: ‘Type A: Relatively high element of risk, with commensurately high dividend and considerable growth potential ‘Type B: Speculative stock with considerable risk, bigh dividends, but less growth potential than type A. ‘Type C: Stock with little risk, considerable growth potential, ‘but relatively low dividend income. at present. ‘Type D: Stock with ite risk, not much growth potential, and fairly igh dividends. Because of the element of risk, the investor wishes to restrict purchases of types A and B to not more than 30 per cent of his investment. To enhance prospects for long-term growth of his investments, hae wishes to have at last 4 per cent of his total outlay in types A and C, Within these restrictions, he wshesto maximize his current dividend income. Total investment is Rs. 1,00,000. Dividend returns onthe four types offnvestments are A: percent, B: 7 per zat, C:3 pe cen, D-5 per vent Formulae this problemas an LP ‘medel to suggest the total investment to be allocated The most recent audited summarized balance sheet of Shop Financial Service is given below: The company intends to enhance its investment inthe lease portfolio by another Rs. 1,000 lakh. For this purpose, it would like to raise mix of debt and equity in such a way that the overall cost of raising additional funds is minimized. The following constraints apply to the way the funds ean be mobilized: (i) Total deb divided by net owned funds, cant exceed 10. (i) Amount borrowed from financial insttuiions cannot exceed 25 percent of the net worth Linear Programming: Model Formulation Balance Sheet as on 31 March 1994 Liabilities (Rs. in lakh) Assets (Rs, in lakh) Equity Share Capital 65 Fixed Assets Reserves & Surplus 110. Assets on Lease ‘Term Lean from IFCL 80 (Original Cost: Public Deposits 150 Rs. $50lakts) 375 Bank Borrowings 147 Other Fixed Assets 50 Other Current Liabilities 50. Investments (on wholly owned subsidiaries) 20 Curren Assets: Stock on Hire 80 Receivables 30 Other Current Assets 35 Miscellaneous Expenditure (not written off) LR Total 602 Total 602 (ii) Maximum amoun’ af bank borrowings eanaot exceed three times the net owned funds. (iv) The company would like to keep the total public deposit limited to 40 per cent of the total debt. The post-tax costs of the different sources of finance are as follows: Equity TermLoas Public Bank Deposits Borrowings 25% 85% ™ 10% Formulate this problem as an LP model to minimize cost of funds raised. Note: (a) Total Debt = Term Loans from Financial Institutions + Public Deposits + Bank Borrowings Equity Share Capital + Reserves and Surplus (©) Net Owned Funds = Net Worth ~ Miscellaneous Expenditures (CA, May 1994) (b) Net Wort Miscellaneous Problems 29, The security and traffic force must satisfy the staffing requirements as shown below on the eve of Republic Dey «celebration, Oicers work S-hour shits starting at each of the 4-hour intervals as shown below. How many officers should report, for duty atthe beginning of each time period in order to minimize ‘he total number of officers needed to satisfy the requirements? Tine Number of Officers 0:01-4:00 4:01- 8:00 8:01-12:00 1 12-01-1600 16:01-2000 20:01-24:00 Formulate this problem as aa LP model se as to determine the minimum number of officers required on duty at beginning of ‘each time period.image not availableimage not availableimage not availableimage not availableimage not availableimage not availableimage not availablerage 30. a ee (i) Addl. equity cepital+ Adal. term loan+ Adal public deposits + Adal. bank borrowings = 1,000 (since the company wans to enhance the investment by Rs. 1,000 lakh) or x, =x, +x, +x4= 1,000 and 3y549.45,2420- 29. x,=number of officers who statin shift (/= 1, 2,3, 6) Min Z= x1 +20) 4.25 4444x546 subjectto i) x, +x, 275 Gi) x +85 215; fii) x3 42,275 Gv) ay #5212; ©) 2545529; wi) xetm2 5 and 4, 20, forall -xyand x) ~ number of class A and B trucks tobe despatched, respectively Min Z = 3x, + 4r, 31 32. 34, 35. subject to (i) 15x, +1023 $3,000 (i) 41495 44598 (policy of retaining aclass of truck with every two class B truck in reserve) and aye, 20. yp % pnd 5c = Weight (in kg) of commodities A, B and Co be accommodated inthe direction {i= 1, 2 3 — forward, centre and after), respectively. Max Z= 60 (x4 +23 4454) +80(%19 +29 t%9) +50 o HXac tHe) subject to (i) Ay + Xa + yg 54,000; Gi) yp ap 1 5p $4000; Gil) ayy ayy + Sy $4,000; G8) 44 #9 * Xe $ 2,000 0) Myo + an + Hye $4000; (i) 44 + yp + Hye S 1,500 (i) x,y + Ory + 2Sxy¢ $ 1,00,000 ; (Wil) 60%, + 50g + 25%0 $ 1,35,000 x4 + Dry + 28x, $30,000 and Jugs Nps Sie 20 forall -xp.py£5.%4and xg = percentage of the total fund that should be invested in all given five schemes, respectively. Max Z= 6x, +13x, +1033 +20x4 +25x5 subject to (i) 18x, +34) +5x5 46x, 410%, 218; Gi) x) 43x: 42m +Su 4 <4 (ii) x5 2025; Gv) stay tay tae and 4) 20 forall) x5, 45 = number of scooters of type A, B and C, respectively. Max Z = 0.7 (6,000 ~5,000)x, + 0.8 (8,500 — 7,000)x + 0.6 (10,500 - 8,500), subject to (@) $000, +2 % 7,000: < 1,00,000; Gi) 2 * 7,000, + 8,500x, = 100,000 and ayer 20 {Xp =mumber of station wagons, minibuses and large buses, respectively to be purchased, Mex Z (total profit) = 15,000r, + 35,000:, + 45,000, subject to (i) x +2, + $30 (Availability of drivers); (i) +(3/5)x,+ (1/2)x;$80 (Maintenance capacity) (Gi) 2,45,000., + 3,50,000x, +5,00,000x, < 50,00,000 (Budget) and Ay ¥y 520, Nove: 1.SW. (9/5) M.B., Because (5/3) 5.0. = 1 MCB. and 1.1 = (3/5) MB, because (5/3) SIV. MD. ‘The data ofthe problem can be summarized as given below: Quary A Quarry B Qua ‘Specifications Material X 20% 0% 10% 30%, Material Y 60%, 30% 40% <40% Material Z 20% 30% 50% between 30% and 402% Costtonne (Rs.) 10 2 15image not availableimage not availableimage not availableimage not availableimage not availableimage not availableimage not availableame OF ee WIG CIEDD NB Fig. 35. Graphical Solution of LP Problem Extreme Point Coordinates (Objective Function Value yx) 2= 1x, +15 ° (0,0) 100) + 150)= 0 A (13,0) 10(13) + 15() B 10) 10(8) + 15(10) ~ 230 c +611) 10(6) + 15(11) = 225 D (0.5) 10(0) + 15(8) = 75 ‘The maximum value of the objective function Z = 230 occurs at extreme point B (8, 10). Hence, optimal solution to the given LP problem is: x, = 8, x, = 10 and Max Z =230. 1 Example 3.4 Use graphical method to solve the following LP problem @-3ry+ 456 Gil) x + 2q <4 No lower bound constraint forx,. [Karn. Univ, BE (Mech), 1995] Solution space a Fig.3.6 Graphical Solution of LP Problemimage not availableimage not availableimage not availableThe contribution per person from each type of the tour packages sold is as follows: Packages Price Hotel Coste Meal and Other NetProfit Offered (Rs) (Rs) Expenses (Rs) (Rs) a @ o @=()-12+@1 Deluxe 10,000 3,000) 4,750 * 2250 Standard 7,000 2,200 2,500 2300 Economy 6,500 1,900 2,200 2.400 (a) The linear programming model ofthe given problemis: Maximize Z=2,250x,+2,300x, + 2.400%, ~2,00,000 subject to the constraints @ —1— 20.10 or 9%,-x)-x,20 Rim ty ) —2— 2035 or ~1x,+13x)-7x,20 Roe tay ames 2 2030 or 3x31, 414,23 © Trae SE Sas — 2 <0,70 or 7x, +3y,—74,<0 htm tay @) x $60; (V) x, 4+,2120; (Wi) tay tx, =200 and x,xy.%520. (b) Since x, +x, +x, =200, x, =200—x, —x, therefore the new LP model is as follows: Maximize Z= 2,250 x, + 2,300 x +2400 (200—x,—x3)-2,00,000=— 150%, 100, +2,80,000 subject to the constraints ® 9,-x,—200-x,-x)20 or *,220; Gi) ~3x, ~3x,+7200-x,-¥,)20 or x,+x,5140 and) 7x, +3x)-7(200-x,-x,) $0 or xy $140 Gi) x50; (¥) xy 4x2 120 and ¥,x,20. (© The coordinates of extreme points of the feasible solution space or region shown in Fig. 3.10 are: AGO, 70), B(60, 70).C(60, 80). D(20, 120) and E(20, 100). The value of objective function ateach of these points is as follows: “75, +13x2-7200-x,-x)=0 or 24270 wl © ® @ Oxu-20 @x-00 On QOa-n me @atn=u0 @xta=n0 100 (20, 100) | o 20 4 60 80 100 120 10% Fig.3.10 Graphical Solution of LP Problemimage not availableimage not availableimage not available° @. 0) =1@ + A @, 0) -1@) + B (4, 2) -1@) + c @4) =1@) + D (, 108) =1@ + 20103) ‘The minimum value of the objective function Z = 2 occursat the extreme point A (2, 0). Hence, the optimal solutionto the given LP problem is:x, = 2, x)= 0 and Min Z=- 2. Example3.12 GJ.BreweriesLtd. have twobotting plants one located at G" and the other at “I”. Each Drink Plant at plant products three drinks, whisky, beer and brandy ev named A, Band C respectively. The number of the § —____“_*_ bottles produced per day are shown in the table: Whisky 1,500 1,500 ‘A market survey indicates that during the month Beer 3.000 1.000 Of July, there will be a demand of 20,000 bottles of __ Brandy 2,000 $000 whisky, 40,000 bottles of beer and 44,000 bottles of ‘brandy. The operating cost per day for plants at G and J are 600 and 400 monetary units. For how many days each plant be run in July so as to minimize the production cost, while still meeting the market demand? Solve ‘graphically. [Pune Univ., MBA, 2000] Solution Let us define the following decision variables: x, and x, = number of days of work at plant G and J, respectively. ‘Then the LP model of the given problem can be expressed as: Minimize Z= 600, + 400x, subject to the constraints @ 3xj+3x,240; i) 3x,+ 240; Gil) 2x, +54, 244 and pk 20. ‘The feasible solution space depicted in Fig. 3.14 is unbounded on the upper side. The coordinates of ‘extreme points of the feasible solution space bounded from below are: A(22,0),B(12, 4) and C(0,40). O atin @ at a0 @ @ x45 -44 oi 2% 30 Fig. 3.14 Graphical Solution of LP Problem ‘The value of objective function at each of the extreme points is shown below:image not availableimage not availableimage not availableSince minimum value of objective function occur at the extreme point A(100, 100), the solution to the given problem is: x, = 100, x)= 100 and Min Z=Rs. 60,000, That is, the firm should bay 100 quintals of scrap each from supplicr A and B to minimize the total cost of purchase. Datm-3 Ones Oxtnce Oue32 Example 3.16 Use the graphical method to solve the following LP problem. 5 Maximize Z=7x, +3x, subject to the constraints 4 @ x,42R23 Gi) x 4y,54; Gil) x, 552; (iy) 532 and Ay%2 0. Solution Ploton. graph each constraint by fist treating itas (0,372) «linear equation. Then use the inequality condition of each i constraint to mark the feasible region as shown in Fig. 3.18. ‘The coordirates of the extreme points of the solution space are: A(S2, 1/4), BES/2, 3/2), and C(0, 3/2}. The value of the oT 2 3 4 5 5 objective function at each of these extreme points is 8 Fig. 31g Graphical Soluion of LP Problem Exireme Point Coordinates (Objective Function Value Gm) Z= n+ 3t A (572,14) 14512) + 3(0/4) = 734 B (872,322) 1482) + 33/2) =22, c (0,32) (0) + 3(3/2) ‘The maximum value of the objective function Z~ 22 occurs at the extreme point B(S/2, 3/2). Hence, the ‘optimal solution tothe given LP problemis: x, = 5/2, x,= 3/2 and Max Z=22. ‘Example3.17 Use the graphical method to solve the following LP problem. Minimize Z=20r,++ 1d, subject to the constraints @ +2540; i) Axj+ 4230; ii) 4x, +34, 260 and xy2,20. ig) Solution Plot ona graph each constraint by firs treating them as. linear equation. Then use the inequality condition of each constraint to mark the feasible solution space as shown in Fig. 3.19. Q@ x+25=40 @n+730 3s Os +35, 2 (40,0) os ws 2% 2 3 35 0 8 5 Fig. 3.19 Graphical Solution of LP Problemimage not availableimage not availableimage not availableECs Operations Research: Problems and Solutions 3 irae a Thenthe LP modelofthe given problem canbe written QO sy +4q=200 @ 31, +55.= 150 as: 50 @ 5, +42, = 100 @ 8 +45.=90 Maximize (total profit) 2=300x, + 400%, subject to the constraints @ Input (a) 5x,+4x, $200; (b)3x +5xyS 150 Ouput a) 5x, +4x,2 100; (b) 8x, #42280 and. x,,4)20. For solving this LP problem graphically the inequality condition of each constraint to mark the feasible region is as shown in Fig. 3.23. The coordinates of extreme points of the feasible region are: A(20, 0), B(40, 0), C(400/13, 150/13), DO, 30) and E(0, 25). The value of the objective function at ° 8 oe O each of these extreme points is as follows: Fig. 323. Graphical Solution of LP Problem Exirene Gooratnates Objective Function Value Point Gym) Z= 300, +400", A 20,0) 300(20) + 400(0) = 6,000 B (40,0) 300440) + 400(0) = 12,000 c 400/13, 150/13) '300(400/13) +400(150/13) = 1,80,000/13 D (0,30) 300(0) + 400(80) = 12,000 E (0,25) 300(0) + 40025) = 10,000 ‘The maximum value of the objective function occurs at the extreme point (400/13, 150/13). Hence, the manager of the oil refinery should produce, x, = 400/13 units under process 1 and x, = 150/13 units under process 2in order to achieve the maximum profit of Rs, 1,80,000/13. Example3.22 A manufacturer producos two different models: X and Y, of the same product. Model X makes ‘contribution of Rs. 50 per unit and model Y, Rs, 30 per unit towards total profit. Raw materials r; and ry are required for production. Atleast 18 kg of, and 12 kgof'r, must be used daily. Also at most 34 hours of labour are tobe utilized. A quantity of? kg of ris nceded for model X and I kgofr for model Y. Forcach of X and Y, | kg of ris required. It takes 3 hours to manufacture model X and 2hours tomanufacture model Y. How many units of each model should be produced to maximize the profit? Solution Letus define the following decision variables: x, and x, = number of units of model X and Y to be 25 produced, respectively. ‘Then the LP model of the given problem can be written as 20 Maximize (total profit) 2= Str, +30x, subject to the constraints © @ 2x,+ 218; (b)x,+ 14212 Rawmaterial) (i) 3x,+2%y534 (Labourhour) and xy,x,20. ‘The problem solution is depicted graphically in Fig. 3.24. The feasible region is shaded according to the inequality condition of each constraint. os is mH Fig. 3.24 Graphical Solution of LP Problemimage not availableimage not availableimage not availableOperations Research: Problems and Solutions Example3.26 Use graphical method to solve the following LP problem: MaximizeZ=3x, +2x, subject to the constraints OxH-Hzh Gi) yty23 and xy) 20. Solution The constraints are plotied on a graph as usual as shown *2 in Fig. 3.28. The solution space is shaded and is bounded from below. It is noted here that the shaded convex region (solution space) is 3 unbounded from above. The two comers of the solution space are, ‘A(0,3) nd B(2, 1). The value of the objective function at these points is: Z(A)=6 and Z(B)=8, ‘Since the givenLP problem is of maximization, there exist a number of points in the shaded region for which the value of the objective | fnetion is more than 8. For example, the point (2,3) lies inthe region and the function value at this point is 12 which is more than 8. Thus, both the variables x, and x, can he made arbitrarily large and the value of Z also inereases. Hence, the problem has an unbounded solution. Example3.27 Use graphical method to solve the following LP problem Maximize Z=5x, +4x, subject to the constraints O y-2 21 Gi) 4428, 23 and x2 20 (PT Univ, BE, 2001] Oa-n-! O4+y-3 ‘Unbound soliton lA,3) 1 2 3 4 3.28 Unbounded Solution Solution Constraints are plotted on a graph as usual as shown in Fig. 3.29. The solution space is shown shaded and is bounded from below. This shaded convex region (solution space) is unbounded from above. The two extreme points ofthe solution space are, A(0, 3/2) and B(2, 1/2). The value of objective function at these points is 2(A) ~ 6 and 2(B) ~ 12.Since the given LP problem is of maximization, theseexists a number of points in the solution space where the value of objective function is much more than 12. Hence, the unique value of Z cannot be found as it occurs at infinity only. The problem, therefore, has an unbounded solution. Fig. 329 Unbounded Solutionimage not availableimage not availableimage not availableFEIN coxrsons researc: rotioms and Soins 3. A manufacturing firm produces two machine parts P, and P, | 9. Accompany machines and drills two castings X and Y. The time ‘using milling and grinding machines. The differentmachining times required to machine and drill one casting including machine set up required for each part, the machining times available on different time iss follows: mutes dle ion ech mache pares penn =a, ince ee Manfacerng Tie Maron Te 5 5 ; Meche Regret ava 4 3 pe tet Nt) a AP ‘ec aro aes ad ee ling washes, Te working cratic hans resist end veine Vr Lathe 10 5 2,500 costs for both castings are Rs. 120 per unit while total fixed costs Millingmachine 4 10 2,000 amount to Rs. 1,000 per week. The selling price of casting X is Grinding machine 1 1s 450 Rs. 300 per unit and that of Y is Rs. 360 per unit. There are no pouperint(is) so 100 Tniaots ent mer X end Vast ea The = ; pong aeons earner erat Demite Remade ofpe oF, aat,tovemmeand | 6 eeice a iear popumeig tee fr te roe ad pee prc to macxatn Oe (Gi) solve the problem graphically. [Dethi Univ, MBA, 1999) 4. A furniture manufacturer makes two products: chairs and tables. 40, company possesses two manufacturing plants, each of which Processing ofthese productsis done on twomachines A apd B.A chair requires 2 hours on machine A and 6 hours on machine B.A table requires 5 hours on machine A and no time on mackine B. ‘There are 16 hours pe day available on mackine A and30 hours on machine B. Profit gained by the manufacturer from a chair anda table is Rs. 2 and Rs. 10, respectively. What should be the daily production of each of the two products? 5. A.company produces two types ofeather belts, say Aand B. Belt ‘A isof superior qulity and B is inferior. Profit onthe twoare 40 and 30 prise per belt, respectively. Each belt of type A requires, twice as much time as required by a belt of type B. Ifall the belts were of type B, the company could produce 1,000 belts per day. But the supply of leather is sufficient only for 800 belts per day. Belt A requires fency buckle and only 400 of them are available per day. For belt B only 700 buckles are available per day. How should the compary manufacture the wo types of belts in order tohave amaximum overall profit? 6. Oldhenscan be bought at Bs 2 each and young onesat Rs. Seach. Theold ens lay three exgs per week—eich egg worth 30 puse. A hhen costs Re. I per week to be fed. If there are only Rs. 80 for purchasing the hers, how many of each kind of hens should be ‘bought inorder to havea maximum profit per week, assuming that itis not possible to house more than 20 hens ata time? {Meera Univ, M Se (Math), 1990) 7. A person requires 10, 12 and 12 units of chemicals A, B and C, respectively for his garden. A liquid product coatains 5,2 and 1 units of A, Band C respectively perjar. A dey product contains 1, 2 and 4 units of A, B and C per carton. If the liquid product for Rs.3 perjar and the dry product sells for Rs. 2 per carton, how ‘many of each should be purchased in orderto minimize cost and ‘meet the requirement? 8 Consider a small plant which makes two types of automobile parts, say A and B. Itbuys castings that are machined, bored and polished. The capacity of machining is 25 per hour for A and 40 ‘per hour for B, capacity of boring is 28 per hour for A and35 per ‘hour for B, andthe capacity of polishing is 35 per hour for A and 25 per hour for B. Casting for part A costs Rs. each and for part Bitcosts Rs. 3 each. They sell for Rs. $ and Rs. 6, respectively. ‘The three machines have running costs of Rs. 20, Rs. {4 and Rs. 17.50 per hour, Assuming that any combination of parts A ‘and B can be sold, what product-mix maximizes profit? ean produce three products: X, Y and Z from a common raw material. However, the proportions in which the products are produced are diferentin each plant and soare the plant's operating costs per hour. Following are the data on production perhourand costs together with current orders in hand for each product. Plant Products Operating Cost XY @_perHour(Rs.) A 2 4 3 9 B 4 4 2 10 Ordersonhand S024 60 ‘You are requiredto use the graphic method to find out the number ofproduction hours needed to fulfil the orders on hand at minimum, cost Acompany manufactures wo products Aand B. Product A yields ‘ contribution of RS. 30 per unit and product B RS, 40 per unit towards fixed costs. I is estimated that sales of product A for the ‘coming month will not exceed 20 units, Sales of product B have not been estimated but the company does have @ contract t0 supply at least 10 units toa regular customer. Machine hours available for the coming month are 100 and products A and B take 4 hours each, respectively, to produce Labour hours available are 180 and products A and 3 take4 hours and 6 hours of labour, respectively. Materials available are restricted to 40 units of two products while each uses one unit ‘material perunit. The company wishes tomaximize contribution. Using the graphic method, find the optimum product-mix. A company produces two types of presentation goods A and B that require gold and silver. Each unit of type A requires 3 g of silver and | gof gold whileB requires | gof silver and 2g of gold. ‘The company can produce g o silver and 8 g of gold. feachunit of type A brings a profit of Rs. 40 and that of type B Rs. 50, determine the numberof units ofeach type that should be produced tomaximize the profit. AA firm makes two types of furniture: chairs and tatles. The contribution to profit by each product as calculated by the accounting depactment is Rs. 20 per chair and Rs. 30 per table, Both products are to be processed on three machines M;, Mand.14, 15, availablein hours per week on each machine areas follows: ‘My, The time required in hours by each product and total time | | | ‘Machine Chair ‘Table dvailable Tine (brs) My 3 3 36 M; 5 2 50 M; 2 6 60 How should the manufacturer schedule his production inorder to maximize profit? In a certain manufacturing firm, the production consisis of a machining process which takes raw materials and converts them into (unassembled) parts. These parts are then sent to one of the two divisions for assembly into the final product—Division | for product A and Division? for product B. Product A requires 40 units of raw material and 10 hours of machine processing time, ProductB requires 80 units of raw material and 4 hours of machine processing time. During the period, $00 units of raw material and 80 hours of machine processing time are available “The capabilitics ofthe two assembly divisions during the period are 6 and 9 units, respectively. The profit contribution per unit to profit and overhead (fixed costs) is Rs. 200 for each unit of product A and Rs. 120 for each unit of product B. With this information, formulate this problem as a lineur programming model and determine the optimal level of output for the two ‘products using the graphic method ‘A company manufacturing television sets and radios has four major departments: chassis, cabinet, assembly and final testing Montily capacities areas follows: Linear Programming: Grahicel Souton Method. FRE 6. A umber company cuts raw timber—osk and pine logs—into ‘wooden boards. Two steps are required to produce boards fiom logs. The first step involves removing the bark fiom the logs, ‘Two hours are required to remove the bark from 1,000 foot of ‘oak logs and three hours per 1,000 feet of pine logs. After the logs have been ebarked, they must be cut into boards. It takes 2.4 hours per 1,000 feet of oak logs to cut into boards and 1.2 hours per 1,000 feet of pine logs. The bark removing machines ‘cam operate up 10 60 bours per Week, while the cutting machines are limited to48 hours per week. The company can buy a maximum, ‘of 18,000 feet of raw oak logs and 12,000 feet of raw pine logs each week, The profit per 1,000 feet of processed logs is Rs. 1,800 and Rs. 1,200 for oak pine logs, respectively, Solve the problem to determine how many fet of each type of log, should be processed each week in erderto maximize profit. 17. A-company manufactures two products A and B, Product A. ‘makes contribution of Rs 30 per unit and product B, Rs. 40 per unit, toward fixed cos, Its estimated that sales of product A for the coming month will nt exceed 20, Sales of product Bhavenot ‘boon estimated but the company does have a contract to supply at least 10 units tos regular customer. ‘Machine hours available for the coming month are 100 and products A and B take 4 hours and 2 hours, respeciively to produce. Labour hours available are 180 and products A and B take hoursand 6 hours of labour, respectively. Materials available are restricted to 40 units and thetwo products use one unit each of | ‘material, Formulate this problem as an LP model end solve it to determine wits of products A snd B to be produced 18. A cold drinks company has two botling plants, located at two different places. Each plant produces three different drinks A, B ad C. The capacities of two plants, in number of battles per day are as follows: Department Television Capacity Ratio Capacity | ProductA Product Product C_ Chassis 1,500 or 4,500 } Patt 200) op: 2,000 Ga: 1000 «© 8000 Pact 1,000 1,000 6,000 Assembly 2.000 = or 4,000 ‘Amarket survey indicates that during any particular month there ‘retin aco 80 ‘will bea demand of 24,000 bots ofA: 16,000 botles of Band Thecontribution of television is RS. 150 each and the contribution of radio is Rs. 250 each. Assuming that the company ean sell ‘any quanity of ether product, determine the optimal combination of cutput 48,000 bottles of C. The operating costs, per day, of running plants | and If are respectively 600 monctary units and 400 ‘monetary units. How many days should the company run each plant during the month so that the production cost is minimized ‘hile sill meeting the matket demand? Use Graphic method. [Delhi Univ, MFC, 1996] HINTS AND ANSWERS Altemative solutions Unbounded solution No feasible solution No feasible solution 6and Min Z (Gi) Anfeasible solution (vii) Unbounded solution () y=3.% (9). Redundant constraints are second and thind and MinZ = 19 (ii) Unbounded solution (si) Infeasible solution x) x, =20/19, x, = 45/19 and Max 2 and Min Z (i) x= 6.8;fa Operations Research: Problems and Solutions 3. x, andax, = number of pieces of P, and P, to be manufsctured, respectively. Max Z = 50%, + 1005; subject to () 10s, +Sx,$2,500; Gi) de, + 10r,€2,000; ii) x, + LS, < 450 and x20 ‘Ans: x, = 1875,x,~ 125 and Max Z = RS21,875. 4, x and.x, = numberof chairs and tables produced, respectively Max Z = 2e)+ 10x, subject to () 2x, +5", 516: Gi) 6x, < 30 and x20 a 0,x)= 32 an Max Z=Rs.32. 5. x, and x, = numberof bets of type A and B produced, respectively. Max Z.= 0.4, + 0.30% subject to () x, +4 <800; i) 2x, +4; <1,000; Gi) x, <400; iv) x, < 700 and xy. 20 a 200, x; = 600 and Max = Rs. 260 6. x, and.x, ~ number of old hens and young hens bought, respectively Max Z = Monthly gain ~Monthly cost = 0.33, + 5x3)~ (x, +) subject to (i) 2x, +5x, $80; i) x, +3, $20 and sty 2 0 ‘Ans: xy =0,3,~ 16 and Max Z = Rs. 8. 7. x, andxy ~ number of units ofiquid and dry produced, respectively Min Z= 3x, + 2x subject to (i) Sej+x)210; (il) 2 +2212; Gil) x, +4%212 and xy y2 0 Anst x= yx; Sand Min Z ~ 3. 8. x, and.x, = number of parts manufactured of automobile A and B, respectively. Max Z= {5—2+20/25+14/284175/35} x, + (6~3+20/ 40+ 14/35+175/25}.x5 = 4.80x, + 4.60%, Hy 2 < Ayes yh, sojetio) Sask WAHT Sk iy + esl md 9. yan seumbert uns ofcasing Xan Yt e manure reapecively Max 2 (604 +366) (1205 +1203) 1,0 = 1805 2405-1000 aij) ey #245240; |G) 24 +853 40 ad 520 Amst, =10,1,=20 an Mix 2 5400. 1b, syanday~ numberof duction hou for pln A and Bede io complet the ones, especie Min Z= 9x, + 10x; subject to (i) 2x) 44%, 250; Gi) 4x, +3x,224 Gil) 3x +2e,2 60 and 4.420 Ans: x, = 35/2 and.x, = 15/4 and Min Z = Rs. 195. U1. x, and.x, = number of units of product A and B, respectively Max 2= 30s, + 40x) subject (0 (i) 2054210; (i) ax, +45 100; (HI) 4x, +6 S180, (Iv) x, +4 S40 and apy20Linear Programming: Graphical Solution Method 12, x and x,= number of units of A and B to be produced, respectively, Max Z = 40x, + 50%, subject © (i) 35,459: Gi) x) H2QSB and ayn 20 ‘Ams: x,=2,x)=3 and Max Z= Rs 230. 13. x, and.x,= number of chairs and tables to be made, respectively, Max Z = 20x, + 30%, subject (2) 3x, 43% $36; (4) Sy #25550; (ii) 2x, + Oxy $60 and ayy 20 Ans: x, =3,x,=9 and Max Z=Re 330. 14, x, and x, = number of units to be produced of product A and B, respectively Max Z = 20x, + 120%, subject to () 40x, + 80x, $800, Gi) 10x, +458; Gi) <6; GW) SO and x20 ‘Ams x, ~5,x)77.Sand Max Z~ Rs. 1900. 15. Letx, and x, = number of units of television and radio tubes produced, respectively. Max (total profit) Z = 150x, + 250%, subject to (i) xy/1,500+ 254,500 1, Gi) x/1,000++x,8,000 > 1 (Gi) x/2,000+x52,0002 1, Civ) xy/3.000+ 9,000 1 and spay 20 16, Let.x, and x number of feet of rw oak lops and raw pine logs to be processed each week, respectively Max 2 = 1.800%, + 1,200, subject 0 (i) 24,434 56; (i) 24x, +12yS4% (ii) x, $18; (W) SID and med 17. Letx, and x, ~ numberof units of product A and B to be produced, respectively. Max (total profit) Z = 30s, + 40x, subject 19 () x4$20; Gi) x2 10; Gil) 4x, +2sy $100; Gv) 4x, +6r, $180; (¥) x, 4) S40 and xpty20 18. Minimize Z; = 600, + 400x, subject 19 (2) 3,000, + 1,000, 24,000; Gi) 1,000, + 1,000%, 2 16,000; Git) 2,000, +6,000x, 2 48,000 and xty20Linear Programming: Simplex Method Chapter Outline © Introduction © Simplex Algorithm (Minimizaton Case) ‘© Siandard Form of an LP Problem © Some Complications and Their Resolution © Definitions * Solved Examples ‘* Reduction of Feasible Solution to a Basic © Self Practice Problems. Feasible Solution © Simplex Algorthm (Maximization Case) Hints and Answers INTRODUCTION The concept of simplex method is similar to the graphical method. In graphical method, extreme points of the feasible solution space are examined (0 search for optimal solution at one of them. For LP problems with several variables, we may not be able to graph the feasible region, but the optimal solution will still lie at an ‘extreme point of the many-sided, multidimensional figure (called an n-dimensional polyhedron) that represents the ares of feasible solutions. The simplex method examines the extreme points in a systematic manner, repeating the steps of thealgorith until an optimal solution is reached. Itis for this reason that iis also called the Herative method. Since the number of extreme points (comers or vertices) of feasible solution space are finite, the method assures an improvernent in the value of objective function as we move fiom one iteration (extreme point) to ‘another and achieve optimal solution in a finite number of steps and also indicates when an unbounded solution is reached. ‘STANDARD FORM OF AN LP PROBLEM ‘The use of the simplex method to solve an LP problem requires that the problem be converted into its standard form. The standard form of the LP problem should have the following characteristis: (i) Allthe constraints should be expressed as equations by adding slack or surplus and/orartifiial variables. (i) The right-hand side of each constraint should be made non-negative; if it is not, this should be done bby multiplying both sides of the resulting consiraint by —1. i) The objective function should be of the maximization type.ETTey Operations Research: Problems and Solutions The standard form ofthe LP problem is expressed as Optimize (Max or Min) Z= ec, x,+¢)%)+... 6,3, +05, +05, +. +05, subject tothe linear constraints yx, + ait eH aig%y +5, A,X, + Oy X +. Hay,%, +53 Gg FF gy Ht + Gragg t Sp and Kya ovmegaienrsag. 20 ‘The standard form of LP problem can be expressed in the compact form as follows: Optiize Maxon miny2~ $ «3, + $05, (Obectv isin) subject to the constraints +2s-+4m (Constraints) Pei +5; and 520. foralliandj (Non-negativity conditions) In matrix notations the standard form is expressed as: Optimize (Max or Min)Z= ex +0s subject to the constraints Axt , and x,s20 where €= (c,Cy,-. 4, isthe row vector, X = (yyy... sty) B= (bj By. by, and $= (5455p. ySyy) ARE column vectors, and A's the m x n matrix of coefficients of variables XX... %, in the constraints, Remarks The constrained optimization (maximization or minimization) problem, may have 1. (@ no feasible solution, i.e. there may not exist values x; (j=, 2,.. .. 1 that satisfy every constraint. () aunique optimum feasible solution. (©) morethan one optimum feasible solution, i.e. altemative optimum feasible solution. (@ a feasible soluion for which the objective function is unbounded, 1. the value of the objective function can be made as large as possible in a maximization problem or as small as possible in a minimization problem by selecting an appropriate feasible solution. 2. Any minimization LP problem can be converted into an equivalent maximization problem by changing the sign of g's in the objective function. That i, Minimize i 957 Maximize 3 (5 dy 3, Any constraint expressed by equality (=) sign may be replaced by two weak inequalities. For example, 1%) +4;24) +... + @),x, =b, is equivalent to following two simultaneous constraints, ary tayaty te Fa yghy Sy amd ayy age ts «+ a44,25, 4, Threetypes of additional variables, namely (i) stack variables (s) Gi) surplus variables (— s), and (i) artificial variables (4) are added in the given LP problem to convert it into the standard form forthe following reasons: (These variables allow us to convert inequalities into equalities, thereby converting the given LP prob- lem into a form that is amenable to algebraic solution. (b) These variables permit us to make a more comprehensive economic interpretation of a final solution, (©). Help us to get an initial feasible solution represented by the columns ofthe identity matrix. The summary of the extra variables to be added in the wo astandard form is, given in Table 4.1 en LP problem io convertUns Pari Si aoe ES ‘Table 4.1 Summary of Extra Variablesto be Added Types of Constraint Era Variable Coefficient of Extra Variables Presence of Extra Variables in Needed inthe Objective Function the inital Solution Mix Max Z Min + Less than o& A slack variable 0 0 Ye equal to (3) is added + Greater than or A surplus variable ° 0 No equal w @) Is subtracted, and an artificial variable -M M Yes is added + Bqual to (~) Only an artificial -M M Yes variable is added DEFINITIONS Basicsolution Given asystem of m simultaneous linear equationsin n(> m) unknowns, Ax=B, where A is an m xn matrix and rank (A)=m. Let B be any m x m non-singular submatrix of A obtained by reordering m linearly independent columns of A. Then, a solution obtained by setting nm variables not associated with the columns of B, equal to zero, and solving the resulting system is called a asic solutionto the given system of equations. ‘The m variables, which may be all different from zero, ate called basic variables. The m x m non-singular submatrix B is called a basis matrix and the columns of B as basis vectors If Bis the basis sub-matrix, then the besic solution to the system of equations will be x, = Bb, Basic feasible solution A basic solutionto the system Ax b is called basic feasible if ¥ =D. Degenerate solution A basic solution to the system Ax =) is called degenerate if one or more of the basic variables vanish. Associated cost vector Let x, be a basic feasible solution to the LP problem, Maximize Z= ex subject to the constraints Ax=b, andx20. Then the vector ¢y ~ (cps Cpps «+» Cg)» Where ep are components of¢ associated with the basic variables, is called the cost vector associated with the basic feasible solution Xp. -REDUCTION OF FEASIBLE SOLUTION TO A BASIC FEASIBLE SOLUTION Consider a system of m simultaneous linear equations in (> m) unknowns: a,x +8 x)... +a, x, =D where A= (a), a3...) isan m real matrix and b is an m x 1 column vector of real numbers. Suppose there exists a feasible solution to the above system of equations. Then this feasible solution can be reduced to a basic feasible solution as discussed below: Step 1 Irtecolumn vectors of A are linearly dependent, then express one of the column vestora, Aas a linear combination of the remaining column vectors of A as: a, = 1,2, +...+ Aa, where 2,’s are scalars, Evaluate the values of the scalars Step 2 Compute the ratios (x)/Aj;A)>0; j= 1,2,...,m} and choose the minimum value. Let it be x, 2, Step 3 Reduce to zero the value of the variable corresponding to the minimum ratio, x, 2, Step 4 The values of new variables are given by: -(E) assoly ‘Operations Research: Problems and Solutions SIMPLEX ALGORITHM (Maximization Case) ‘The steps of the simplex algorithm to obtain an optimal solution (ift exists) to a linear programming problem are 2s follows: Step 1 Formulation of the mathematical model (@ Formulate the mathematical model of the given linear programming problem. (ii) If the objective function is of minimization, then convert it into onc of maxi following relationship ization by using the Minimize Z = -Maximize Z*, where Z* =~ Z. (iii) Check whether all the 6, (i= 1,2, ...,.m) values are positive. If any one of them is negative, then ‘multiply the corresponding constraint by ~ 1 in order to make b, > 0.1n doing so, remember to change a type constraint to a® type constraint, and vice versa. (iv). Express the mathematical model of the given LP problem in the standard form by adding adcitional variables to the left side of each constraint and assign a zero cost coefficient to these in the objective function. (9) Replace each unrestricted variable with the difference of two non-negative variables; replace each non-positive variable with a new non-negative variable whose value is the negative of the original variable. Step2 Set up the initial solution Write down the coefficients ofall the variables inthe LP modelin the tabular form, as shown in Table 4.2, to get an initial basic feasible solution [xy =B™ b] After having set up the initial simplex table, locate the identity matrix and column variables involved in it. ‘This matrix contains all zeros except positive elements 1's on the diagonal. This identity matrix is always a square matrix and its size is determined by the number of constraints. ‘The identity matrix so obtained is also called a basis matrix (because basic feasible solution is represented by B= I}. Assign the values of the constants (b's) to the column variables in the identity matrix (because ap=B'b=1b=b). The variables corresponding to the columns of the identity matrix are called hasic variables andthe remaining ones are non-basic variables. In general, if an LP model has m variables and m (< n) constraints, then m variables would be basic and n —m variables non-basic. That is, simplex algorithm works with basic feasible solutions (pfs), which are algebraic versions of exireme points. A bfs is 2 feasible solution obained by choosing one basic variable for each quality constraint and remaining ones are non-basic and have zero value. However, in certain cases some basic variables may also have zero values. This situation is called degeneracy and is discussed later. ‘Table4.2 Initial Simplex Table 6° 1 & & 0 0 .. 0 Coeffcientof Variables Valueof Basie Variables Basic Variables inBasis Variables | or Bh . r) B bE xp) ea 5 a 42 ay 1 oO o &m 8 4 % ot o Son Sy Om Ma a) i Z=Eey ty, o o 0 0 e- oo 0 ‘The first row in Table 4.2 indicates the coeificients cof variables in the objective function which remain the same in successive simplex tables. These values represent the cost or profit per unit to the objective function of each of the variables and are used to determine the variable to be entered into the basis matrix B.vn ari Spb The second row provides the major column headings for the simplex table. Column ‘ey’ lists the coefficients of the current basic variables in the objective function. These valves are used to calculate value of Z when one unit of any variable is brought into the solution, Column headed by xy represents the current values of the corresponding variables in the basis. The identity matrix (or basis matrix) represents the coefficients of slack variables which have been added to the constraints. Each column of the identity matrix also represents a basie variable to be listed in column B. Numbers jin the columns under each variable are also called substitution rates or exchange coefficients ‘because these represent the rate at which resource i(/= 1, 2,... , m) is consumed by each unit of an activity FURAN 2-09) The values z; represent the amount by which the value of objective function Z would be decreased (or increased) if one unit of given variable is added to the new solution. Each of the values in the c-z; row ‘represents the net amount of increase (or decrease) in the objective function that would occur when one unit of variable represented by the column head is intraduced into the solution. That is: 2 (net effect) = o, incoming unit profit/cost) —z, (outgoing total profit/cost) Step3 Test for optimality Calculate the cj —z; value forall non-basie variables, To obtain the value of , multiple each element under Variables Column Head (columns, a of the coefficient matrix) with the Corresponding elements under Coefficients of Basic Variables Column (¢q-column). Examine the values of ¢,~ There may arise three cases. {@ fall ¢)-2,<0, thon the basic feasible solution is optimal i) Ifatleast one column ofthe coefficients matrix (i.e. a, ) for which c,~z,> Oandall elements arenegative (i.e. ay <0), then there exists an unbounded solution to the given problem. (ii) fat least one c,~z >0 and each ofthese has at least one positive element (e.g) for some tow, then it indieates that an improvement in the value of objective function Z is possible. ‘Step 4 Solect the variable to enter the basis. If casc (iii) of Step 3 holds, then select a variable that has the largest c,~z, value to enter into the new solution, That is, g7 7,7 Max (6-2); G30) The column to be entered is called the Key or pivot column. Obviously, such variable indicates the largest per unit improvement in the current solution. Obviously, such a variable indicates the largest per unit improvemeat in the current solution. Step 5: Test for fea (variable to leave the basis) After identifying the variable to become basic variable, the variable to be removed from the existing set of basic variables is determined. For this, each number in xg-column (i¢. 6 values) is divided by the corresponding (but positive) number in the key column and we select the row for which this ratio, [(constant column) (key column)] is non-negative and minimum. This ratio is called the replacement (exchange) ratio. That is, we =i und 44 ‘These ratios indicate, for example, the number of units of. variable (product) that can be made by trading away all of the current level of the respective basic variables. In other words, this exist criterion limits the number of tnits of incoming variable that can be obsained from the exchange. It may be noted here that division by negative or zero clement in key column is not permitted. ‘The row selected in this manneris called the key or pivot row and represents the variable which will leave the solution. The clement that lies atthe intersection of the key row and key column of the simplex table is called key or pivot element. Step 6 Finding the new solution (@ Ifthe key element is 1, then the row remains the same in the new simplex table. (ii) If the key clement is other than 1, then divide each element in the key row (including elements in Xp-column )by the key element, to find the new valves for that row.FEY cverstions Research: Problems and Sclutons (ii) The new values of the elements in the remaining rows for the new simplex table can be obtained by performing elementary row operations on allrows so that all elements except the key element in thekey ‘column are zero. In other words, for each row other than the key row, we use the formula: (ees is ig oe above or all (srr sumer "| Number i ew row «| the new row, that is row drow key 3 ° aia replaced in Step 6 (ii) ‘The new entries in eg (coefficient of basic variables) and x, (value of basic variables) columns are updated in the new simplex table of the current solution. Step7 Repeatthe procedure Go toStep 3 and repeat the procedure until all entries inthe c,-z,row are either negative or zero SIMPLEX ALGORITHM (Minimization Case) In certain cases as discussed below, it is difficult to obtain an initial basic feasible solution. Such cases arise. Case 1 Whenthe constraints are of the Stype, i.e.) ay x) Sb, ,x,2 Obut some right-hand side constants [i.e. <0}. In this case after adding the ron-negative slack variable s, (i= 1,2, ... m), the initial 30 obtained will be 5; ~ ~ b; for some i, It is not the feasible solution because it violates the non- negativity conditions of slack variables (i.s,20). Case 2 When the constraints are of the 2 type, ie. 2 ay x; 2, 3 2 0. In this case to convert the inequalities into equation form, adding surplus (negative slack) variables, iat Letting x)= 0 (/=1,2,.. 7m), we getan initialsolutions, = B, ors, =~ b, Itisalsonota feasiblesolution as it violates the non-negativity conditions of surplus variables (i.e. s, > 0). In this case, we add artificial variables, 4, (i=, 2, .. m) to get an initial basic feasible solution. The resulting system of equations then becomes: Say ays +4, =by where x96 p4,20, 7 1,2,..9m ‘and has m equations and (n +m +m) variables (i.e. m decision variables, m artificial variables and m surplus variables). An initial basic feasible solution of the new system can be obtained by equating (n+ 2m — mm) = (rt + m) variables equal to zero. Thus, the new solution to the given LP problem is: 4,=6,((= 1,2, .m), whieh does not constitute a solution to the original system of equations because the two systems of equations are not equivalent. Thus to get back to the original problem, ertificial variables must be dropped out of the optimal solution. There are two methods for eliminating these variables from the solution, (@ Two-Phase Method (i) Big-M Method or Method of Penalties The simplex method both forthe minimization and the maximization LP problem may be summarized inthe form ofaflow chart Remark Artificial variables have no meaning in « physical sense and are only used as a tool for generating ‘an initial LP problem solution. Before the final simplex solution is reached, all artificial variables must be dropped out from the solution mix. This is done by assigning appropriate coefficients to these variables in the objective function. These variables are added to those constraints with equality (=) and greater than or equal to () sign.Two-Phase Method In the first phase of this method the sum of the artificial variables is minimized subject tothe given constraints to get a basic feasible solution of the LP problem. The second phase minimizes the original objective function starting with the basic feasible solution obtained at the end of the first phase. Since the solution of the LP problem is completed in two phases, this is called the Tixo-phase Method. ~ Conver LP model into standard form by adding either slack variables, surplus variables andlor anificial variables + Decide coefficient of these variables inthe objective function] T ‘Setup intial simplex table to obtain initial solution Minimization Maximization Do" Do’ 9-4 No 9-4 ‘negative values» ———+| This solution is optimal positive value exist? exist? Yes ‘Select Key colunin with Select key column largest negative ¢,~ 3; value largest positive ¢ ~, value + Select key row with Min {iyi /¢y3 j=} + Halla, 5 0, then current solution is unbounded and stop the procedure I entity Key element at the intersecion of key row and key column q Uplate te entries inthe Simplex table by (a) firstobtaining key row values, and (©) apply elementary row operaions Fig. 4.1. Flow Chart of Simplex Method Advantages 1. No assumptions on the original system of constraints are made, ie, the system may be redundant. inconsistent or not solvable in non-negative numbers. 2. Itis easy to obtain an initial basic feasible solution for Phase |. 3. The basic feasible solution (iit exists) obtained atthe end of phase Tis used to start Phase I. Steps of the algorithm: Phase | Step 4 (a) [fall the constraints in the given LP problem are (<) type, then phase II can be directly used to solve the problem. Otherwise, the necessary number of surplus and artificial variables are added to convert constraints into equality constraints. (b) Ifthe given LP problem is of minimization, then convert ito the maxim mn type by the usual method. onesStep 2 Assign zero coefficient 1 each of the decision variables (x,) andto the surplus variables; and assign ~ 1 coefficient to each of the artificial variables. This yields te following auxiliary LP problem, Maximize Z"= $ (1) 4, ii subject to the constraints Say x) +4, int and 34,20 Step3 Apply the simplex algorithmto solve this auxiliary LP problem. The following three cases may arise at optimality. (@ Max Z* = Oandat least one artificial variable is present inthe basis with positive value. Then no feasible solution exists forthe original LP problem. (b) Max Z* = 0 and no artificial variable is present in the basis. Then the basis consists of only decision variables (x's) and hence we may move to Phase II to obtain an optimal basic feasible solution on the original LP problem. (©) Mex Z*=Oandat least one artificial variable is present inthe basis atzero value. Then a feasible solution to the above LP problem is also a feasible solution to the original LP problem. Now in order to arrive at ‘the basic feasible solution we may proceed directly to Phase Il or else eliminate the artificial basic variable and then proceed to Phase IL ‘Once an artificial variable has left the basis it has served its purpose and can, therefore, be removed from the simplex table. An artificial variable is never considered for re-entry into the basis. Remark The LP problem defined above is also called an auciliary problem. The value of the objective function inthis problem is bounded from below by zero because the objective function represents the sum of artificial variables with negative unit coefficients. Thus, the solution to this problem can be obtained in a finite number of steps. Phasell Assign actualcoefficients to the variables in the objective function and zero tothe artificial variables ‘hich appear at zero value inthe basis at the end of phase |, that i, the last simplex table of phase I can be ‘used as the initial simplex table for phase Il. Then apply the usual simplex algorithm to the modified simplex table to get the optimal solution to the original problem. Artificial variables which do not appear in the basis ‘may be removed. The Big-M Method ‘The Big-M Method is another method of removing artificial variables from the basis. In this method, we assign coefficients to artificial variables, undesirable from the objective function point of view. If objective function Z is to be minimized, then a very large positive price (called penalty) is assigned to each artificial vatiable, Simitasly, ifZ is tbe maximized, then a very large negative price (also called penalty) is assigned to each of these variables. The penalty will be designated by ~ M fora maximization problem and + M for a ‘minimization problem, where M > 0. ‘Step 1 Express the LP problem in the standard form by adding surplus variables and artifical variables. ‘Assign a zero coefficient to surplus variables and a very large positive number + M (minimization case) and —M (maximization case) to artificial variable in the objective function. Step 2 Calculate the values of c)~z, and examine these values. (@) fall ¢)~z;2 0, then the current basic feasible solution is optimal (©) Iffora column, k, c,~z,is most negative and all entries in this column are negative, then the problem has an unbounded optimal solutionLinear Programming: Simplex Method (©) fone or more ¢,-z; <0, then select the variableto enter into the basic with the largest negative ¢,~z, value. That is, e727 Min {g- 356 cj 2,50} 5 ‘The column to be emtered is called key or pivor column. Stop 3 Determine the key row and key element in the same manneras discussed in the simplex algorithm of the maximization case. Remarks Atany iteration of the simplex algorithm, any one of the following cases may arise. 1, If.at least one artificial variable is present in the basis with zero value and the coefficient of M in each 6-4)/~ 1,2. +) values ismon-negative then the given LP problem has no solution, Thatis, the current basis feasible solution is degenerate. 2. [fat least one artificial variable is present in the basis with positive value and the coefficient of M in each 6-2) U= 1.2, 1) values is non-negative, then the given LP problem has no optimum basic feasible solution. SOME COMPLICATIONS AND THEIR RESOLUTION Unrestricted variables Usually in an LP problem, itis assumed that the variables x, should have non- negative values. In many practical situations, however, one or more ofthe variables, x, can have either posi- tive, negative or ze1o value, Variables which ean assume positive, negetive or zero value are called unre- stricted variables, Since the use of the simplex method requires that all the decision variables must have non- ‘negative value at each iteration, in order to convert an LP problem involving unrestricted variables into an equivalent problem having only restricted variables, we have to express cach of unrestricted variables as the difference of two non-negative variables. Suppose variable x, be unrestricted in sign, We define two new variables say x and x7, such that xe x—eT5 xf .a720 when x/ 2.x” thenx, 20 and when x/ apt % sa pa fet 23 phy x 2 Soe] parent eh * Here it may be nated that all these solutions are non-degenerate,Linear Programming: Simplex Mathod BEE Example 4.2 Computeall the basic feasible solutions to be the system of linear equations: (@ 44,4244 57, Gi) -x,+4ry+2Q= 14 Solution Setting x, =0, the resulting square matrix of coefficients of x, and x, is 2) [2] ‘This matrix is singular because |A] = 0. Thus, it cannot be a basis because the columns are not linearly independent and in fact, neither are rows. As a consequence, we now set r, =0 and obtain a resulting square matrix of coefficients of x, and x,. aot Ala 2 This matrix can bea basis because | # 0 and so e=-tyat{? “HP 7] 1/9) _7 0 =0, x4=639=7 ae OL1 a}Las] 963] ~Learop sr SOO Here it may be noted that x,, although not a non-besic variable, still has a zero value in this solution. Finally, setting Owe get a coefficient matrix of x, and.x;, which may serve as a basis: 42 1 4 ap 1 [4-2 7177 _[ 0 me eae Thus, xB ATE], allsal2aeloa] | eae fie e202 26318 =72. and again a basic variable, x, has a zero value. A Example43 _Computeall the basic feasible solutions to the LP problem: Maximize Z= 21, +3x, 44%,+ 7%, subject to the constraints @ 2x 43x)~ x, 44x, 1y-2k, + 6r,— 74, and yoke ayy G20 [Meerut Univ, MSc (Maths), 1994] Solution Since m-+x=6 and m=2, the maximum number of basic feasible solutions to the given LP problem are *C,=4, (i) Putting x, x,~Oin the constraints, we get 2x, +3x,=8 and x Solving these equations for x, and x, by using matrix inversion method, we get a basic feasible solution to the given LP problem: Basic: x, =1,1)=2; Non-basic: x, =x,=0 Similarly the other three basic feasible solutions are (ii) Basie = 119; Non-basie : x) Non-basie: x, (iv) Basic :1,=44/17,x, =45/17; Non-basie: x, The values of objective function Z at each of these solutions is given by @ Z=2(1)+3Q)+4O)+70)=7 Gi) Z=2 22/9) +3 (0)+ 4 (0)+7(7/9) =93/9 (ii) Z=2()+34916)+4(716)+ T0)= 1972 Civ). Z=2(O)+ 340+ 444/17) +7 (ASIT)= 14S ‘The maximum value of objective function Z ~ 144/5 occurs at the basic feasible solution, x) ~ x) 24744117, 4 =45/17,FREY coeratons Resear: Example4.4 Compute all the basic solutions to the following LP problem: Maximize Z =x, + 3%, +31, subject to the constraints @ x,+2r+3xy84 and (ii) 2x +3x,+5x5,=7 Also find which of the basic solution are (a) basic feasible, (b) non-degenerate basic feasible, and (c) optimal basic feasible. Solution Since there are m + n= 3 variable and m ~ 2 constraints in this LP problem, the maximum number of basic feasible solutions to given LP problem are °C, = 3. These are obtained by setting any one (n= 1) of the nonbasic variable to zero. ‘The solutions in which all basic variables are 2 0 will be basic feasible solutions and solutions in which all tasic variables are> O will be non-degenerate basic feasible solutions that optimizes the objective fanetion will be the optimal basic feasible solution. Putting.x,=Ointhe constrains, we get x,*2xy=4, and 2x, +3x,=7. Solving theseequations forx, and ty, we geta basic feasible solution to the LP problem, where Non-basie variable: 1y=0 and Max Z=5. Basie variables: x=2,1) ‘The summary of various basic feasible solutions is shown below: ‘Summary of Basic Solutions Basic Non-basic Valuesofthe Basie Valueofthe ‘Inthe Zsthe Solution Is the Solution Variables Variable Variables Given by the Objective Solution _-‘Non-degenerate Feasible and ‘Constraint Equations Function Feasible? (are all basie Optimal? (are all 207) variables > 0?) yx 4 xy +2y 2x, + 3x) =7 5 Yes Yes Yeo yr a 4 Yes Yes Yes = No No No ‘The first two solutions are non-degenerate basic feasible solutions. The solution: x, Max Z.= Sis the optimal: Example4.5 _A firm manufactures four different machine parts M,, M,, M, and M, made of copper and zine. ‘The amount of copper and zinc required for each machine part, their exact availability and the profit earned from ‘one unit of each machine part are as follows: MM, My My ExactAvailabiliy ti) he) tg) (hg) Copper 5 4 2 1 100 Zine 2 3 8 1 15 Profit (Rs) 12 8 4 10 How many of each part be manufactured to maximize profit? Forthis LP problem find (a) basic solutions (b) basic feasible solution (©) non-degenerate basic feasible solutions (@) optimal basic feasible solution. [Punjab Univ., BE(E), 1996}image not availableExample 4.6 Show,using matrix-vector notation, that the following system of linear equations has degener- ate solutions: 2x; +4y—15=2, and 3x, +2x 444-3. Solution The givensystem of equations can be written in the matrix formas: Ax=b, where ay. fe -t Since A isthe matrix of order 2» 3, there can be*C, = 3 submatrices i He 1] and R ‘| eachof L the order 2 x2, Also each ofthese submatrices is non-singular, any one of them can be taken as the besis matrix B. Thus to obtain all possible basic solutions, putting variables equal to zero not associated with the columns of B. Putting xy 0, ie. the variable not associated with the columns of B, we get b ae)-0) . CIE TC TEE] meen Thus basi feasible solutionso obtained is: x, = 1, x. ~0. Non-basic variable: x, =0. Similarly, considering the other two submatrices, we get the following basic feasible solutions: B,1,=-1/3; Nombasic: x,=0 1 = 133-05 seat ‘Thesecond solution: feasible solutions are: (i), 1/3 is not feasible. Thus the remaining two degenerate basic Example 4.7 Isthe solution: x, =1, =x,= a basic solution of the equations: yt 2g tay te, ‘and x, +2x,+0Sx,+x5-2? Solution The givensystem of equations can be written in the matrix formas follows: awd 12 1 1 0)/7 - ep th 2 os o af]? 12 ‘Since matrix A is of order 2 5, there can be °C, = 10 submatrices of the order, 2 x 2 as shown below: 12] ft 1] fv] fi oy y2 17 f2 t] f2 oyfa ppt 07 po [ 3} [i al (i th [i ih [; wb [; i} G ih les i} las t] a0 [0 ‘| Putting x, = x, =x, = 0 in the constraints, we get x, = 2 and x, = 2 as one of the basic feasible solution. ‘Similarly, the other basic solutions can be obtained as follows: Non-basie: x, =x, Non-basic: x;=2; Non-basic: x; Non-basic: Non-basic: Non-basie:Linear Programming: Simplex Method (vii) Basi Non-basie: 1) =1)=*5=0 (vill) Basie: x,= Non-basic: 1, =1)="4=0 Since none of these solutions corresponds to the solution: x) ~ 1, x) ~ 1/3, 5 = x4 ~A5 ~ 0, the given solution is not basic. 5 Example 48. Letx, =2,x,=4 and.x, = | be a feasible solution to the system of equations: (9 2xj— xy #2ej= 2, and (il) xy +4ry=18 Reduce the given feasibie solution toa basic feasible solution {Delhi Univ, B Se (Maths), 1995] Solution Writing the system of equations in matrix notation: x a 2) or ax=b 14 ole 18 2-1 F 2 : where A=(a, asaye| i a of mil and x=] x5 |. %5, The above system can also be expressed a8: ay xy + ay Xp +s xy =D. Since it s given that.x, = 2, 4) =4 and X47 Lwehave 2a, +4a,+a,=b. Since rank (A) =2, only two out of three column vectors a,, a, and a, are linearly independent. Assuming that these vectors are linearly dependent, expressing one of themas a linear combination of the remaining two ay Aya tga, 2 2]. far Al -ali [al al or 2=22)-AyandO=A, +4, where A and A, are scalars not all zero. On solving these two equations, we get, combination: 3/9 and 2, = -2/9, Substituting values of A, and 2, inthe linear 8 a=Ajay tiga where A, ~8/9,2,=~29 and, ‘To reduce the number of positive variables, the vector to be removed following result: a 2 oa, >0} sm } Since x,/A, =9/4 corresponds toa vector ait should be removed to obtain a new solution with the twonon- negative variables, The values of the new variables are given by: ” a9 "219 *=1 a dotg tle? oe -=1, This gives 1 4 Bie Sie SIS Sie eln SS whe Prey‘The new solution (0, 9/2, 13/4) so obtained is also feasible solution. As the two column vectors a, and a, associated with non-zero variables x, and x, are linearly independent. Therefore, the required basic feasible solution x) = 9/2 and x, = 13/4. This result can be verified by substituting the values of x, x, and.x, inthe equation, a x, +2,.x,+; x Example4.9 Show that the feasible solution: x, = 1,x,=0, x,= I; Z= 6 to the system of equations: ( xptxytx5=2, and (il) x, xy +x,=2 with Maximum Z= 2x, +3x, + 4x, isnot basic. Solution Writing the system of equations in matrix notation: 4 1 o4d 2 ab ron af? |b} & % 14 2 zi where A=(ayapa)=| >| 3 and a=] 4 “The above system can also be expressed as: ay x, +8, x) +5 ¥)= x5 Lwehave a, +a,=b. Since rank (A) =2, only two out of three column vectors a,, a and a, are linearly independent. Assuming that these vectors are linearly dependent. Then expressing one of them asa linear combination of the remaining two as: Since it is given that x, = I,x,=0 and a, = Aya, tay, [Jaffe] tessa tana where Ay, 2p are scalars notall zero. On solving these two equations, we get A, = | and 2, = 0. Substituting values of A, and A, in the linear combination: +i, oF a=a, (a) Taking a, and A,=~ 1. To reduce the number of positive variables, the vector to be removed is chosen in accordance with the result } = x x ae: se =Miny—4 2, >0p=Min y=, <2. 3) = Min) i, of > | nf Ay ag fT Since x/2,= 1 corresponds to vector a, itshould be removed to obtain anew solution with not more than, ‘ovo non-negative variables. The values of the new variables arc given by: x x This gives es zh =1-10) x a =0-10) xh sao is IE) Then we shall have 2, A, = Oand A, = 1, and in this case = Min fea, ~of- 1 (corresponds to a5)Linear Programming: Simplex Method Thus, the vector a, can be removed to obtain anew solution with not more than two non-negative variables. ‘The values of the new variables are given by: 0-1(0)=0 and The new solutions (0, 0, 2) and (2, 0, 0) so obtained have two linearly independent column vectors, and hence both of these solutions are basic feasible. However, these two solutions are different from the given (2,0, 1) solution. Thus, the given solution is net basic. Example 4.10 Computeall basic feasible solutions to the LP problem, Maximize Z=3x,+2x, +, subject to the constraints (@ 3x, 42x) +2x5 444-8, (il) 3xy He Ha, +¥=7, and ky ky AyXs2O Obtain the new basic feasible solution other than the initial solution and verify that the corresponding new value of Z is improved. Solution The given system of equations can be written in matrix form as follows: where The y's for each column a, of A but notin Bare reel DEEL] lo afta} 13} Ly: 2] _ Piz 3 [2p Lo} eb FES] cS 4 3,-0f3] Since ¢4~(3, 2, 1,0, 0) and xy~(0, 0,0, 8,7), we get Z = eg xy =0. To obiain an improved basic feasible solution, we have to determine one of the column vector in B which can be replaced by a column vector in A but not in B. It is determined as follows: x Xp x <8 = Min|*®.y, >0 -min 30. 22} Fy y, Yun" Ya crrimage not availableLinear Programming: Simplex Method ‘Table4.4- Initial Solution wos a a a Tin Rao Bais Values Da besa c i « 2 @ © 1 0 o| @ ‘ 2 o fo F $s 0 1 0 | a2 a Z we {3 2 4 0 o 1 | aso Zo oo 0 0 0 © 33 4 8 ot ~1,2,3,the current solution isnot optimal. Thus, variable x is chosen toenter into the basis as it corresponds to the largest net contribution in the x,-column where all elements are positive. ‘The variable leaving the basis is determined by dividing the valves in the xp (constant) column by their corresponding elements in the key column as shown in Table 44, Since the minimum ratio isin row 1, the variable s, should leave the solution. Heration 1 Since the key element enclosed incircle in Table 4.4is not 1, divide all elements of the key row by 3 to obtain new values ofthe elements in this row. The new values of the elements in the remaining rows for the rnew table can be obtained by performing the following elementary row operations on all rows so that all elements exceptthe key element I in the key column are zero. R, (new) ~> R, (old) +3 (Key element); R, (new) > R, (old) -2R, (new); R, (new) > R, (old) -2R, (new) ‘The new solution is shown in Table 4.5, Table4.5 Improved Solution oF 3 5 4 0 G 6 % lariablesin Solution | = ~~~ ~SCOS*«SS:C*«dSC« Rao Basis Values xp B bx) 5 » 83 wm 1 0 13 0 0 = 0 a was }-43 0 =@ -23 1 0 | aasys 6 5 293 so 0 4-30 1 | evya Z= 408 5 Cr gy [-u 0 4-53 ‘The improved basic feasible solution can beread from Table 4S as: x ‘The improved value ofthe objective function is Z= 40/3. s shown in Table 4.5,c,~2,>0, the current solution i not optimal Iteration 2 Repeat steps 3 10 5. Applying following row operations to enter variable x, into the basis and to drive out s, from the basis. R, (new) = Ry (old)5 (key element); R, (new) — Ry (old) -4R, (new) ‘The new solution is shown in Table 4.6. cred‘Table4.6 Improved Solution 6? 35 4 0 ot % Variabiesim Solin | + ~=*S~*~«S,S*~C<~*“S!C“C*‘“ SSC Ra Basis Values alt, Bo bx) 5 4 % 2% 1 0 1 0 0| @By@a) 4 5 tis |-ans 0 1 -2ns uso | ° a sas |Gims 0 0 ans 4s 1 | aonsyavisy 2=256/5 3 [ows 5 0 ims 45 Oo 3 [ums 0 4-15-45 oo Neration 3 Thevariablex, enters the basisand leavesthe basis. For this, we shall apply the following row operations. Ry (new) > R, (old) x(15/41); Ry (new) > Ry (old)+ (15/41) Ry (new): R, (new) R, (old) ~(2/3) R, (new) ‘The new solution is shown in Table 4.7. Table4.7 Optimal Solution 97 3 5 4 o o ° ep Variablesin Solution |, 2 % * 5 5 Basis Vals B bea) 3 % 50/41 0 0 14l SAL — 10041 5 exiat ° ° 1-041 sat ava 3 x ovat 1 ° wal 121 sia Z- 76541 3 3 s 4 aay Dama = o o 0-441 24 11/41 In Table 4.7, all cj ~ +; <0 for non-basic variables. Therefore, the optimal solution is reached with, x, =89/41,x= 50/41, x,=62/41 and the optimal value of Z = 765/41. Example 4.13 Three grades of coal A, B and C contain phosphorus and ash as impurities. Ina particular industrial process, fuel up to 100 ton (maximum)is required which should contain ash not more than 3 per cent and phosphorus not more than 0.03 per cent. It is desired to maximize the profit while satisfying these condi- tions. There is an unlimited supply of each grade. The percentage of impurities and the profits of grades are given below: Coal Phosphorus Ash Proftsin 6) (6) Rupees per Ton A 002 30 1200 B 004 20 1500 c 003 50 1400 Find the proportions in which the three grades be used. | Solution Let usdefine the following decision variables: | xy; and x; = amount (in tons) of three grades of cool 4, Band C to be used, respectively. Then the LP model for the problemis written as: | MaximizeZ (total profit) = 12x, + 15% + 14x,Linear Programming: Simplex Method [EET subject to the constraints ( Phospiorus content must not exceed 0.03 per cent 0.02%, + 0.04 + 003%; $ 0.03 (x) 4x3 +45) de, +4, +34, $ 3ey tay tx) or (@_Ash content must not exceed 3 per cent. 3x) #24) + 5, 3, tay tay) or (Gi) Total quantity of fuel required is not more than 100 ton. xy tay tay < 100, +150. yt 2y $0, and yey ty 20 Introducing slack variables s,, 5 and s, the LP problem can be expressed in its standard form as follows: Maximize (total profit) Z = 12x, + 1Sr, + 14x,+0s, +0s, +05, subject to the constraints © stay te — 0) (i)—ay +2xy 45-0, My Xp Apts Hy 53 20. An initial basic feasible solution is obtained by setting x, =0, x, 3+ and 100 0, =0 inthe constraints. Thus initial solution shown in Table 4.8 is: 5,=0,s)=0, ,= 100 and Max Z Table48- InitialSolution y> oP os m8 Varablsin Seaton | OCS Rl ‘Basis Values | sais cp B bx, L 7 0 7 © 0 yo 0 | 05 oe 0 oy 3 ° | 5 100 rot oo 1 | z= a0 5 °° 0 oO a gi, | 2 os wo ‘The initial basic feasible solution shown in Table 4.8 is not optimal. Since c, 14 in x,-columnis the largest positive value, replace basic variables, into the basis by non-basie variablexy. For this apply following row operations Ry(new) > R, (old)+ I(key element); Ry(new) —> R, (old) ~R, (new) to get the new solution shown in Table 4. Ry (new) —> R, (old) +R, (new): Table4.9 Improved Solution oP Bw Tarablesin Sawin | =~ OSC*S*~«S*S YR ‘asie Value aly ‘s 3 is % 0 rr ° 0 % o 4 6 ot = 0 i wo | @ 0 1 10 1 | 1002-20 z-0 “5 oS no 4s ot Solution shown in Table 49s not optimal as ¢,—= =27 inx column isthe langest positive value, Ths, center variable x, toreplace variable s, into the basis. For this apply following operations:R; (new) > Rj (old) + 2(key element); R, (new) — R, (old)+R, (new); Ry (new) > R, (old) Ry (new) to get the new solution shown in Table 4.10. ‘Table4.10 improved Solution g> ks Veriblasin Sotaion [Sy |i Rao Basis Xpiky a 8 5 m 2 2 0 12 | 10 ° ” o @ vw 1 wl ws 2 A 10 0s sos | z= 1350 3 Rs m2 m 0 2m 0 320-2 ‘Once again solution shown in Table 4.10 is not optimal because c; ~ 7, positive. Replace s, variabl 1/2 value in x,-column is still to the basis by-x, variable using necessary row operations. The new solution so ‘cbtained is shown in Table 4.11. ‘Table4.11 Optimal Solution 4 Ro “ o 4 o Variablesin Solution % % % yn Basis Values os B a) 15 i) 40 ° 1 os 25 4 4 20 0 0 vs us 2 4 40 1 0 os 2s = 1360 2 15 4 8S SOBs 5 0 0 0-85 Vs — 6815 Since all the ¢)-z)$0in Table 4. 11, the current solution is optimal. Hence, it is desirable to use x, = 40 tons, = 40 tons and.xj~ 30 tons of grade A B and C coal, espectively to maximize ttl profit to Rs. 1360. Example4.14 A.company makes two kinds of leather belts. Belt A is a high quality belt and belt Bis of lower ‘quality. The respective profits are Rs. 4 and Rs. 3 per bel. The production of each of ype A requires twice as much time as abet type B, and ifall belts were oftypeB, the company could make 1000 per day. The supply ofleather is sufficient for only 800 bel per day (both A and B combined). BetA requires a fancy buckle and only 400 per day are available. There are only 700 buckles a day available for belt B. ‘What should be the daily production of each type of belt? Formulate this problem as an LP model and solve it tysimpler metnod, Solution Letus define the following decision variables: 1 and x, = numberof belts of type A and B, respectively manufactured each doy. ‘Then the LP model for the problem is written as: Maximize (total profit) Z= 4x, +3x subject to the constraints ( 2x, +2,5 1000 (Time availability), (i) xy+x,< 800 (Supply of leather) Ot. 1 (Buckles availability) and p20< Linear Programming: Simplex Nethod JEP) Introducing slack variables s,, 83,5, ands, to convert given LP model into its standard form as follows: Maximize Z= 4x, +31, + (s, +05 +055 +0s, subject to the constraints (9 2xj ty +5,=1,000, Gi) xy +xy+5=800, (i) x, 455-400, (iv) x) 434-700 and xp. pSpSp.%5.5,20 ‘An initial feasible solution is obtained by setting x, =.x,= 0 in the constraints. Thus, the initial solution shown in Table 4.12is: s, = 1,000, = 800, s,=400, s,=700 andMax Z=0. Table 4.12 Initial Solution 6? 43 0 0 0 0 Varies Soluion ] =O | Rao inBaxis Values 4h n B br_) ° ” 1000 > 1 1 0 0 0 ° 5 a0 1 1 0 4 0 0 ° 5 so | @ 0 0 0 14 0| 4001-400 ° se 100 oto o_o _ 1 fede zo oo o 0 0 0 y | 4 3 0 o o 0 In Table 4.12, since ¢) ~=, ~4is the largest positive number, we apply the following row operations to replace basic variables s, into the basis by non-basic variable x. R, (new) + R, (old) I (Keyelement); Ry (new) > R, (old)—2 Ry new); R, (new) + R, (old)~R, (new) ‘The new solution is shown in Table 4.13. ‘Table 4.13 Improved Solution 47 43 0 0 0 0 VariallesSolwion | = % 5 ~~), | MnRaio inBais, Values eh & BR berg ° 200 o0 @ 10-20 | 2001 =200 ° 8 400 o Tr oo 4 1 0 4 y 400 rr ° 5 70 oo oo __t_ | 7001-700 = 1600 0. 0 0 ww o 3 9 o -« 0 The solution shown in Table 4.13 is not optimal because <2 ~ => Oin x -column, Thus, egain applying the following row operations to get anew solution by entering variable x into the basis and removing variable s, from the basis, R, (new) +R, (old)+I (keyelementy, Ry (new) —> R, (old) —R, (new); Ry (new) > Ry (Old) ~ R, (new) ‘The new solution is shown in Table 4.14. oniseayTable4.14 Improved Solution 6 43 0 0 o 0 Prof Vriabes Sotuion | ~<% ~*% ~~ ~«Sy]~—«MinRato perUnit—inBasis Values Xplsy & B bax) 3 % 200 0 7 70-20 = % 200 0 0-1 4 @_— | 20071 = 200» 4 400 1 0 Cr) 1 0 | 00/1 =400 ° 55 500 oo -1 o 21 | so0a=250 2= 2200 % 4 3 a 0 5-4 oo -3 0 ° The solution shown in Table 4.14 is not optimal because ¢,—z,> 0 in s,-column. Thus, again applying the following row opcrationsto get a new solution by entering variable, s, nto the basis and removing variable 5 from the basis, Ry (new) —>R, (old) + | (key element) R, (new) > R, (old) +2R, (new) R, (new) -¥R, (old)-R, (new); R, (new) R, (old) -2R, (new) ‘The new improved solution is shown in Table 4.15. Table4.15 Optimal Solution oo 4 3 o 0 0 o Variables Solution x hs Sy in Basis Values fo B bax) 3 ” 600 0 1-1 2 0 0 0 5 200 0 oo -1 1 1 0 4 x 200 1 ° 1 -1 0 ° ° 4 100 o ° 1 20 1 7=2600 zy 4 3 7 2 0 ° 4-4 o o -1 2 0 0 Since all cj ~z;£ O corresponding to non-basic variables columns, the current basic feasible solution is also the optimal solution. Thus, the company should manufacture, x, = 200 belts of type 4 and.x, = 600belts of type Btoobiain the maximum profit of Rs. 2,600. Example4.15 A pharmaceutical company has 100kg of 4, 180 kg of B and 120 kg of C ingredients available permonth. Company can use these materials to make three basic pharmaceutical products namely 5-10-5, 5-5- 10 and 20-5-10, where the numbers in each case represent the percentage of weight of A,B and C, respectively {in each ofthe products. The cost of these raw materials is shown in table, Selling pricesof these productsare Rs. 40.5, Rs. 43 and 45 per kg, respectively. There isa capacity restriction ofthe company for ingredient Cost per kg (Rs.) product 5-10-5, so that company cannot produce more than 30 kg_§ << permonth. Determine how much of each of the products company A be) should produce in order to maximize its monthly profit. z & Solution Let the P,, P, and P3 be the three products to be inertingredients 20 ‘manufactured. Then the data of the problem can be summarized as follows:Prosuct Product Ingredients 4 zB C Inert P, 5% 10% 5% 80% P; 5% % 10% 80% P 20% m% 10% 65% Cos perkg Rs.) 80 20 50 20% Cost of P, = 5% * 80+ 10% x20+5% x 50+ 80% x20 =4+2+2.50+ 16=Rs. 24.50 perkg Cost of P, = 5% x 80+5% x20+ 10% x 50+80% *20 =4+1+5+16=Rs.26perke Cost of P, = 20% x 80+5% x 20+ 10% x 50+ 65% x 20= 16+ 1+5+13=Rs.35 perks, Let x, x5 and.x be the quantity (inkg) ofP, , P,and P, respectively to be manufactured. Then the LP problem can be formulated as: ‘Maximize (netprofit) = (Selling price ~ Cost price) = (Quantity of product) = (40.50 24.50) x, + (43-26) x + (4535) x,= 16x, + 17x,+ 10, subject to the constraints 1 1 1 pit apt tg%s £100 or + xy xy < 2,000 1 1 ToT tap t7g7 S180 or 2yt mt xy $3,600 dentin tay st00 or apt? 52400 $30 and Xjtpy20. Introducing slack variables s,, and s, toconvert the given LP model into its standard form as follows: Maximize Z= 16r, +17x, + 10x, +05, +05, +0s,+0s, subject to the constraints (x, 42x, +4x;+5,=2,000, (i) 2x, +x +x, +5,=3,600, (ii) x, #244 + 2xy55°2,400, Gv) x, +5430 and x ,.43.%5 51.5305. 5420. An initial basic feasible solution is obtained by setting x, =. 0. Thus, the initial solution shown in Table 4.16 is: s, = 2,000, s, = 3,600,s,=2,400, s,=30 and MaxZ=0. Table 4.16 Initial Solution s> 6 7 + 09 0 0 Oo Variables Solin | x = 5) 5-5) 5 | MinRatlo imbasis —Yaues 5% we 2 bese) 0 4 zoo | 1 1 4 1 0 0 0 | 20001-2000 0 5 3000 | 2 1 1 0 1 0 0 | 3¢00/~3.600 o a zaoo | 1 @ 2 9 0 1 0 | 2¥002-1200 a & 50 1 0 @ 0 0 o 1 = Zo o 0 0 0 0 0 0 gy [16 7 wo 0 0 0 0 Since ¢, ~z, = 17 in:x,-colurnn is the largest positive value, we apply the following row operations to get a new improved solution by entering variable x ino the basis and removing variable sy from the basis.Ry (new) > R, (old) +2 (key element); Ry (new) -> R, (old) ~ Ry (new); Ry (new) Rj (old)-R, (new) ‘Thenew solution is shown in Table 4.17. Table4.17 Improved Solution 6 7 0 0 0 0 © aH S 8 8 Min Ratio yh, B ° 0 3 10-12 0 | s00|172)= 1,400 0 32 0 0 0 1 V2 0 | 2,400/3/2y= 1,600 0” 1 1 0 0 12 6) 1200172)= 2.400 ° @ 0 0 0 o 0 1] smi=30 Z= 20400 5 |e 7 WW 0 0 Wao oy [20 -7 0 0 -1 0 The solution shown in Table 4.17 is not optimal because c, ~2,>Oinx-column, Thus, again applying the following row operations to get a new improved solution by entering variable x, into the basis and removing the variable s, from the basis: Ry (new) > R, (old) + 1 (Key elementy, —__R, (new) + R, (old) ~(1/2) R, (new) Ry (new) Ry (old) ~ (3/2) Ry (new); R, (new) + R, (old)—(1/2) R, (new) ‘Thenew solutions shown in Table 4.18, Table4.18 Optimal Solution o> 6 7 0 0 0 0 0 Variables Solution HOR OH OS in Basis Values ag 2 bay) 10 4 785 00 8 10 a 2355 0 0 0 oO 1 = -32 7 x L185 oO 1 1 0 0 wm -1 16 x 30 1 0 0 o 0 o 4 = 20625 5 ‘6 7 7 0 0 WR 18D a4 oo -7 09 0-172 ~192 Since all ¢j—z < 0 corresponding to non-basic variables columns, the current is also an optimal solution. ‘Thus, the company must manufacture, x, =30 kg of P,, x, = 1,185 kg of P, and x,= 0 kg of P, to obtain the ‘maximum net profitofRs. 20,625. Example 4.16 A furniture manufacturer produces and sells desks, chairs and bookshelves. They have no difficulty in selling their items. However, limited availability of machine time, labour and floor space restrict production, Data on usage of resources, supplies and profits on items are given belo Desk Bookself Supply Machine hrs per unit 8 5 1,000 brs Labour brs per uni 5 3 1650 brs Floor space sq. ft per unit 9 9 1,260 9. Contribution Rs. per unit 270 205 (@) Formulate this problem asan LP problem. (b) Using x, x,,x, a8 the respective decision variables, simplex algorithm was initiated for the above prob- em. At one stage, the following table was reached.6 m0 dS o 0 ¢e Variables in Solution x) 4 % s Ss ty Basis Values z (x, 270 x 125 fl 12 si8 ie 0 0 0 25 o 13 a 0 0 5 135 0 32 28-98 0 1 2= 33,190 9-4 ° 9 254 135940 0 From the above table, what is the current solution? Is it optimum? (©) Ifthe curreat solution above is not optimum, carry out the simplex algorithm unit an optimum solution is reached. State the optimum solution, (@) The Personne! Manager of Good Furnishings claims that by recruiting additional labour force, the profits ‘can be increased, Assuming no other changes, is the claim valid? Explain with reasons. Solution Let us define the following decision variables: 1x1»; and x, ~ mumber of desks, chairs and bookshelves to be manufactured, respectively. (a) The LP mode! of the given problem is written as: Maximize Z=270x, + 144x, +2251, subject to the constraints () Bx, +44, +5x,$1,000, (il) Sx, +3x,+3x,$650, (ii) Oe, +6x,+ 9x45 1,260 and x.% 420, (b) From the given table we read the current solution as: x: No. of desks = 125, s,: slack labour hours=25 hrs, and s,: slack floor space = 135 sq.ft Revenue = 125 270=Rs.33,750 (©) This solution isnot optimal because c, ~7,=225/4 is the largest positive vaiue corresponding to a non-basic variable. Improved solutions are shown in Tables 4.19 and 4.20. ‘Table4.19 Improved Solution o> mi sot @ Wwiaiiesn Solan | CMa Basis Values gl, 7 beep m0 y ps | 12 sem 00 | 300 9 4 p | 0 Ww -1s -s8 1 0 = ° 5 ps | o 32 <8 0 1 40 Z= 3350 09 254-9 00 Table 4.20 Optimal Solution 6° 20 20 Wenham Swim [oA Bass Fates B bx) 270 i 100 1 28 ° % x0 59 zs 7 “0 oe 5 0-16(@ The solution in Table 4.20 shows that labour force is slack ass, = 130 in the basis. Hence, addition of further labour force will not change, the optimum solution. The Personnel Manager's claim, therefore, is not valid. Example4.17 A farmer has 1,000 acres of land on which he can grow corn, wheat or soyabeans. Each acre of com costs RS, 100 for preparation, requires 7 man-days of work and yield a profit of Rs. 30. An acre of wheat cost Rs. 120 to prepare, requires 10 man-days of work and yields a profit of Rs. 40. An acre of soyabeans cost Rs. 70 to prepare, requires 8 man-days of work and yields a profit of Rs. 20. Ifthe farmer has Rs. 1,00,000 for preperation and can count on 8,000 mar-days of work, how many acres should be allocated to each crop to maximize profit? [Jammu Univ., MBA., 1996} Solution Let us define the following decision variables: 1», and.x, = acreage of land on which farmer can grow corn, wheat and soyabeans, respectively. ‘Then the LP model based on the problem data is written as: Maximize Z=30r,-+40r, +20r, subject to the constraints (@ 10x, + 12c5+7x, $10,000 (Budget available), Gi) 7x, + 10r, + 8x, $8,000 (Man days of work) (@ 3, +x)+%, $1,000 (Land available); and xy.%7320. Introducing non-negative slack variables s,, s, and s, to convert the LP problem into its standard form as follows: Maximize Z = 30x, + 40x, + 20x, +05, + 0s, +05, subject to the constraints (9) 10x, + 12, +7, +5, = 10,000, (i) 74, + 10x, + Bx, +. and xy XXyS 89520. ‘An initial basic feasiblesolution: x, =x). in Table 421. Proceeding in the same manner as discussed in previous examples to improve the initial solution as shown Tables 4.22and4.23. $000, Gil) x+y Hy + = 0, s, = 10,000, s,~ 8,000, 5, 1,000 and Max Z~ 0 isshown Table 4.21 Initial Solution 0 9 mM 09 oo Profit. Variables 8 5H & % | MinRatio perUnit in Basis ph % 5 ° my mo 2 7 1 © 0 | 10012 0 h 7 8 0 10 | socnt0 -» ° 45 1 i 10 o 1 | toon Z=0 oo 68 Oo oo 0 0 ww 0 oo ‘Table4.22 Improved Solution 67 0 0 mM 0 0 oO Profit Variables Solution H 8% 5 55 | Minkaiio perUnit — imBasis Values alk & B bx) 0 7 200 “36101 a0 0 | «00016 > 0 4 £00 Bio = vto So | 800007 0 4 200 2100 mot‘ | 20008 Z= 32,000 z zo10 0 ~«AOSC« O74 -2 0-4 0Linear Programming: Simplex Method ‘Table4.23 Optimal Solution 97 30 4 0 o 0 Prot. Variables. Solution % ay QO perunit in Basis Values a z boxy) 30 -%g 250 1 0 2616 1016-1216 40 % 2s 0 1 316-7116 1046 0 5 125 0 ote -¥16 woot Z= 32300 z 30 ~OSCGONSSC‘CS*«CSC 9-4 ° Ce In Table 4.23, since all ¢ ~z)$ 0 the optimum feasible solution is arived at. Hence, farmer must use x, =250, x4 625 and x, = O acreage of land to grow corn, wheat and soyabeans, respectively to achieve a maximum profit of. 32,500. Example 4.18 By means of some sharp bargaining wit the union and subsequent reduction of union “make- ‘work’ restrictions in his former contract, a small paper towel manufacturer has created some spare capacity in ‘each of his three main production departments: Cutting, Folding and Packaging. For the purpose of identifica: tion, three different sizes of paper towel are called products A, B and C, Owing to itssmall size, the company can sellin the market, all that it can produce at a constant price. Management is inclined to be conservative and does not wish to expand production facilities at this time, although they do wish to utilise fully the present spare capacity. ‘The paper towelling is received fiom another manufacturer in large rolls. These rolls are subsequently cut, folded and packaged in three sizes. The pertinent manufacturing and profit information for each size of paper towel is summarized as below: Department Size Constraint for Time 4 B c Period per Week Cutting 107 50 20 ms Folding 54 100 40 2210 Packaging 07 10 20 MS Profitcontribution per unit (in Rs.) 100 150 200 (@) Formulate this problem as an LP model, (b) Find the optimum soh Solution Letus define the following decision variables: {Xy-%, and x, = number of units of paper towel of sizes A, Band C tobe produced, respectively. ‘Then the LP model based on the problem data is written as Maximize Z= 10x, +151) +20, subject to the constraints (10.7%, + 5x, +2x,£2705 (Cutting), i) Sx, + 10x, +4r,<2210 (Folding), (i) 07x, +x+2x,5445 (Packaging) and yp Xyp520. Introducing non-negative slack variables s,, 3 and ss to convert the LP problem into its standard form as: Maximize Z= 10x, + 151, +20x, +05, + 05, +055 subject to the constraints (i) 107x,+5x,+2n+5 0. using simplex method =2105, (ii) 54x, +10x, +41, +5, 210, (il) Ox, +x, +2x,+ and ).%5,43.545 427image not availablenear Programming: Simplex Method (b) In Table 4.27, since all c)~z)$0, the optimum solution obtained is: x, = 200, x» = 65,5 = 120andMax Z= ‘5375, Hence, the present spare capacity is utilized fully when x, = 200 units of towel A, x= 65 units of towel Band.x,~ 120 units of towel C are manufactured. Example 4.19 A firm assembles and sells two different types of out-board motor 4 and B, using four re- sources. The production process can be described as follows: Resources Capacity per Month ‘Motor unit shop resource + 400 type A units or 250 type B units or any linear combination of the two. Type A gear and drive shop resource : 175 type A units ‘Type B gear and drive shop resource : 225 type B units Final assembly resource + 200 type A units oF 350 type B units or any linear combination of, the two. ‘Type A units bring in a profit of Rs. 90 each and type B units, Rs. 60 each, What should be the optimum product-mix? Formulate the problem as LP mocel and solve by simplex method. Solution Let us define the following decision variables: x, and x = number of units of type A and B motors to be assembled, respectively. ‘The LP model based on the problem data is writen as: Maximize (ital profit) Z = 90x, +60x, subject to the constraints @ Sx,+8x)$2000, i) S175, il) 4225, (v) Try +4x $1400 and 4,220. Introducing non-negative slack variables s,s, and s, to convert the linear LP into its standard form. Maximize 2=90x, + 60x, + Os, +053 +0s; + Os subject to the constraints ® Sx, 48x, +5,=2000, Gi) 4y4+5,=175, Gil) x4. and ,.X3S4. 53.55.5420. Aninitial basic feasible solution: x, == 0, , = 2000, s3= 175, ,=225, s,= 1400 and Max Z=0is shown in Table 4.28. Proceeding in the same manner as discussed in previous examples to improve the initial solution. ‘An optimal solution is shown in Table 4.31 400 128, iv) Tr, +4e, + sy ‘Table 4.28 Initial Solution “7 0 0 0 0 6 0 Variables Soluion | 8) SOS Mn Rao inBais, Values pik A B Hex ° 5 2000 5 8 1 0 0 @ | 20005 ° Ss 175 ®@ o 0 1 0 0 isn 0 55 225 0 1 0 60 1 0 — 0 54 1400 7 4 0 oo t_ | 4007 Zo oo oo oo o 0 o o o o ort)0 % 1175 . «8 = 00 | 1ase %” x 175 1 0 0 1 0 0 - ° a ms > 1 0 o 1 oO | 250 ° ti 178, 0 @ 0 20 1 | 17845 = 15,750 a a a 0 @ o -» o 0 Table430- Improved Solution oF 9 60 0 a 0 oO Variables Sotion | «4% ~~OSC*yS*C«wyS«dY’sCMin Rao inBasis Values xp fa B bex,) 0 4, 775 >. 0 1 @ 0 2 | 1759 = %0 x 175 1 0 0 1 0 o | asa ° 45 514 Cr © a 154 Ce ee ee S = 18375 % oo 0 8 0 5 o 0 Oo ws 0 «Ss ‘Table431 Optimal Solution 6 0 oO o o o Profit Variables Solution xy 4S 5 8 5% perunit ——inBasis Values & B boxy) 0 % 7199 0 o wi 4 o 9 ” x 8009 1 0 -9 0 0 29 o 5 299 ° 0 76 0 1 sn6 o S 17809 0 1 B60 0-586 2= 19,666.66 a a ar) os 6%) ° 0-3 0 0-383 In Teble 4.31, since all ¢,- z, $0, the optimum solution is arrived at. Hence, the firm must assembled x, = 800/9 units of motor A and.x, ~ 1750/9 units of motor B to earn a maximum profitof Rs. 19,666.66. Example 4.20 An advertising agency wishes to reach two types of audiences. Customers with monthly incomes greater than Rs. 15,000 (target audience A) and customers with monthly incomes of less than Rs. 15,000 (target audience B). The total advertising budget is Rs. ,00,000. One programme of TV advertising costs Rs. 50,000, one programme of radio advertising costs Rs. 20,000. For contract reasons, atleast 3 programmes ‘ought to be on TV and the number of radio programmes must be limited o 5. Survey indicates that a single TV programme reaches 4,50,000 customers intarget audience A and 50,000 in target audience B. One radio programme reaches 20,000 in target audience A and 80,000 intarget audience B. Determine the media mix to maximize the total reach.Linear Programming: Simplex Method Solution Let us define the following decision variables -x, and x, = number of advertising programmes to be released on TV and radio, respectively. ‘Then the LP model based on the problem data is writen as: Maximize (total each) Z =(4,50,000 + 50,000) x, +(20,000+ 80,000) x, subject to the constraints (9) 50,000 x, +20,000 x,<2,00,000 or Sx, +23, and xp.x)20. Introducing non-negative slack variables s,s) and sto convert the LP problem into its standard formas: Maximize (otal reach) Z.=5,00,000 x, +1,00,000 x, +0s, +05; 0, subject to the constraints () Sr,+2x,45;=20, i) -m+8) and %,X3,5,.5))5320. Aninitial basic feasible solution: x, ~x9~ 0,5) ~ 20,53 ~~ 3,s3~S and Mex ZO is shown in Table 432. Proceeding in the same manner as discussed in previous examples to improve the initial solution. An optimal solution is shown in Table 4.34 ,00,000.x, +1,00,000r, 20, (ii) x, 3 or -yS-3, Gi) n<5 Gi) x +55=5, ‘Table432 Initial Solution 67 500.900 1.0000 0 0 o Variables Solution x 2 an 5 | Mnkaio inBass Values alt op B exp) 0 m 20 3 2 H 0 0 Ds 0 “3 @ o o 1 0 | 0 5 5 0 1 ° 0 1 - Z=0 z 0 0 ° 0 0 4-4 | 500900 1,00,00 0 ° ° Table433 Improved Solution G2 500900 1000 6 ° o Variables Solution % % i % | Mm Raio inBasis Values _! fs 8 ex) 0 i 3 0 2 T 0 [55-15 5,00,000 x 3 1 ° © o = ° 85 3 ° 1 6 1 - =15,00,000 S000 o 500.0000 © -190,000 a 5.00.00 0 Table 434 OptimalSolution 97 5.00000 100,000 0 o ° Yarlabies Solution x % a % 5 inBasis Values os B Bx) 0 Sy 1 ° ws 2s 1 0 500,000 4 4 1 28 us ° ° 0 5 5 ° 1 o 0 1 Z= 20,00,000 7 $00,000 200,000 100,000 0 0 0 ~1,00,000 -1,00,00 0 °image not availableimage not availableimage not availableLinear Programming: Simplex Method ‘subject to the constraints (3x, +xy+4x,+5,=600, Xs Ayy Hyp Spy A AQzO. (i) 2x, +4x,+2s, 0, (ii) 2x, +34, 43x, +A, = 540, and ‘The initial basic feasible solution is obtained by setting x, = x, = s,= 0. The initial solution: s, = 600, ‘A,=480, A, = 540, Max Z =-1,020 M so obtain is shown in Table 41 Table4.41 Initial Solution oo 203 4 0 oO —M Prof ‘Variables Solution yo) 5 5 2 A ‘Min Ratio per Une —nBasls Values nlp a B ber) 0 sy 600 3 4 4 0 0 of 600 -M 4 480 2 @ 2 0 + 1. 0} 4804» M A 540 2 3 3 0 oo 1 | 5408 7100 M —iN_ 7M 5M 0.~~=«M MOM 2e4M 347M 445M 0 -M 00 In Table 4.41, cz, value in x,-column is largest postive, entering variable x, to replace basic variable A, into the basis. For this, apply the following row operations Ry(new)— R,(0ld)* 1/4 (Key element); R, (new) + R,(old)-R,inew); Ry(new)— R,(old)—3Ry(aew) to get the now solution as shown in Table 4.42, Table4.42_ Improved Solution 7 2 3 4 0 0 -M Variables Solution x A] Minato inBasis Values xp! &e B 0 5) 2 Ol 1 v0 | 9607 3 % moo o 414 0 240 A wo oo 120 180M z 32-M2 3 32-3M2 0 -3/4~3M4 12+M2 0 52+3M2_0 443M 0 InTable 4.42 c,—z, value in.x,-column is largest positive, enter variable x, to replace basic variable A; into the basis. For this, apply Following row operations: Ry(new)— R,(old) * 2/3 (key clement); Ry(new) > R,(old) —(1/2) (new). to get the new solution as shown in Table 4.43. ‘Table4.43 Optimal Sotution Ry(aew) > R, (old) ~(7/2)R, (new), 6 2 3 4 6 0 Variables Solution * x % 5 5 fn Basis Values a in) 5 60 48 ° ° 1 32 = 0 13 1 0 o 2 4 120 v3 o ' ° 2 78 3 4 ° 2 3 “13 0 0 0 -12In Table 4.43, all ¢ ~z, R,(old) +1 (key element); R,(new) — R,(old) ~(1/3) R; (new) 10 get the new solution as shown in Table 4.45, Table4.45 Improved Solution 6 300400 o Variables Soltion x Xo % % s | Makaio inBasis Values Salty & B (=) 0 4 3000 ° 0 1 H 0 | 3000 ° a 4500 ° o @ 0 1 | 2m 300 % o 1 o 2 0 0 = 400 x 0 o 1 10 0 = z=0 300 400-1000 0 ° 0 40 0 0 Again, the c,—z,~400 value in.xy-column is largest positive, entering variable xy toreplace basic variable sp into the basis. Apply following row operations: R, (new) R, (old) = (3/5) (Key element); R, (new) > R, (old) —R, (new); R, (new) R, (old) +2 R, (new): R,(new) > R, (old) + R, (new) ‘to get the new solution shown in Table 4.46. ‘Table446 Optimal Solution 9? 300 400 00 0 0 Variables Solution xy Xo xy 8 inBasis Values oe B be xy) 6 5 300 ° ° 0 1 ys -600 dy 2700 ° 0 1 0 3/5 300 4 5400 1 0 0 0 os 400 xo 2700 ° 1 ° 0 us Z= 10,80,000 300 400 — 600) 0 ° 0 0 0 5400 unit, 9 =2,700 units, x3. 5, = 300 units, and Max Z= Rs. 10.8 lakhs. (b) Since, =300 units of resource | remain unused, it does not resirict the achievement of the target. Also 7 ~ 6, implies that resource 2 is fully used. It has opportunity cost of Rs. 240/hr and restricts the achievement of target. (© If the production capacity of resource 2 is inereased by As, hours, then the maximum value of As, is restricted by the fact that the new value of, should remain 2 0, i. 300-(3/5)As,20 or =(3/5)As,2~300 or As; < 500, ‘This additional production capacity of 500 hours can be used as: Incremental profit = Rs.(500 x 240)=Rs. 1.2 lakhs Total profit =Rs.(10.8+ 1.2)=Rs. 12 lakhs © Since all ¢,~z, $n Table 4.46, the optimal solutionis: onties10With production capacity of 4,500 + 500 =5,000 hours, the new values of variables x, x, and xq become: 2y =A + B/S) x 500=2,700+ 300=3,000 units, 5400+ (615) x 500=6,000 units, ty 700+ (3/5) x 500= 3,000 units, ‘Max Z = 300 * 6,000 + 400 * 3,000-600 « 3,000=Rs. 12 lakhs. Example 4.25 A company has two manufacturing plants, each of which can produce three products X, Y and Z from a common raw material, However, the proportions in which the products are produced are different in each plant and so are the plant's operation costs per hour. Data on production per hour and costs are given below, iogether with current orders in hand for each product. Operating Cost Products per Hour # Yy Zz Bs, Plant A 3 4 3 9 Phin 4 2 0 Orders on hand 0 0 You are required to use the simplex method to find the number of production hours needed to fulfil the orders in hand at Solution Let us define the following decision variables: .x, and x = number of operating hours used in plant A and B, respectively. ‘Then the LP model based on problem data is written as Minimize (‘otal cost) Z = 9x, + 10x, subject to the constraints (@ 2x, +4r,250 (ProductX), (ii) 4x, +3x,224(ProductY), (ii) 3x, +2e,260 (Product Z) and p20. Introducing non-negative surplus variables 5,,5>,ands, and artificial variables Ay, A, and A, to convert the LP problem into its standard formas follows: Minimize (jotal cost) Z=9x, + 10x, +0, +05,++0s, + MA, +MA, +MA, subject to the constraints (i) xy #2x-5,+A,=25, Hi) ALY F3K-—SQHAQ=M, and Ay Xp Sp Sy Sy Ay Ay Ay 20. An initial basic feasible solution: x, =x, =0, Ay =25, Ay =24, A, = 60 and Min Z = 109M is shown in Table 447 ) 3x, + 2x; 85+, = 60 Product Z) Table 4.47 Initial Solution st 9 0 Veriables Solon | 4 ~~ | Min Rao inBests Vales lh & Bo bless) M 435 ee M 4 m4 | @ 3 0 4 0 1 0 | 4804 M4 | 3 2 0 0 4 0 0 1| sms Z= 10981 7 i 1M -M -M_-M MMM 9-8 W-T MoM M 0 0 0 Proceeding in the same manner as discussed earlier to improve the initial solution, Improved solutions are shown in Table 4.48 t0 4.50,Linear Progremming: Simplex Method [KY] Table4.48 Improved Solution oo oo 0 0 o Mm mM Tavis Sotuion [A | Maio inBasis Vales en g 2 bes) u a 1 ° a 7°) 78 9 a ‘ 1 340 o of — M 4, 2 0 34 0 of Tava oa ve OM is oM Meo M00 ‘Table4.49 Improved Solution Py om 0 0 0 Mm Weise Coin Say a ng ya inBxt Tatu vl e 3 bsp M 4; 5 0 rr rn er 9 # 20 1 om 0 6 Bw 9 | 3% 0 A 5 o wo 1 wo Pariyey) 9 6eawa M03 eMB 0 44m3 Mw 0 3B Table) Optimal Solution 9 . mo oo Taahies Solin TO ; mbar ahr _#@ 3 mn y 0 1 3H 9 4 1 o 2 6 ia Q HI asi Q om ts 7a 108 q °C os 0 0 30 2 In Table 4.0, since all ¢,~=,>0, an optimum solution isarrived ai, '¢ x, =17.5 hours of plant A and =3.75 hours of plant B are needed to fulfil the orders on hand at minimum cost of Rs. 195 Example 4.26 A small jewellery manufecturing company employs a person whois highly skilled gem cutter, land it wishes to use this person at least 6 hours per day for this purpose. On the other and, the polishing facilities can be used in any amount upto 8 hours per day. The company specializes in three kinds of semipre- cious tones P Q R Cutting (hrs) 2 1 1 Polishing (hrs) 1 1 2 Costperstone (Rs) 30 M10 P,Qand R. Relevant cutting, polishing, and cost requirements are lised in the table, How many gemstones of ‘each type should be processed each day 1o minimize the cost of the finished stones? What is the minimum cost?image not availablesy «2: Example 4.27 A bicycle manufacturer makes two model Sports cycle and a Racer. In order to make & sports model, 6 man hours are needed, while a racing model requires 10 man hours. The manufacturer can employ no more than 15 men and these men work & hours per day for 6 days each week. The cost of material amount to Rs. 500 per cycle and the manufacturer's weekly quota of such materialsmay not exceed Rs. 40,000. ‘The firm has a contract to supply atleast 30 ‘sports models” and 20 *racers’ per week. How many cycles of each type should be made in order to obtain the maximum possible profit, ifthe profit on each sports cycleis Rs. 100 and on each racing model is Rs. 300? Solution Letus define the following decision variables: x, and x, = number of Sports and Racing cycles to be manufactured, respectively. ‘Then the LP model based on the problem data is written as: Maximize (total profit) Z= 100%, + 300, subject to the constraints () 64, + 10x, $ 15% 8 » 6 (= 720) Man hours, f)500x, + 500%, $40,000 (Material cost), Gi) x, 230; 220 (Supply contract), and %4,5)20. Introducing non-negative slack, surplus and artificil variables to convert the LP mode! into its standard form as: ‘Maximize (total profit) = 100x, + 300x, + Os, +0, +053 +0s,-MA)—MAy subject to the constraints 6x, + 10x, +s, =720, (@ 5x, +50x, +s,=4000, (i) x-53+4,=30, (o) x,-5,4 4, =20, and x4.25) 53,555 p Ay Ay20. An initial basic feasible solution: s =54=0 asshown in Table 4.54 is obtained by setting x, =x, =, ‘Table 4.54 Initial Solution 62m 0 Veriabis Sohaion |, Ma ‘mBas aes ee cp B (b(=x,) om | 6 160 06 0 | raoi0 oy 400 | 505 | 00050 wo 4] tO Ot -w 4 ow | 0 Mo oo ot ot fms 750m 3 pcan w0+M 304M 00 MM 0 Apply necessary row operationsas usual and proceed to improve current solution tll an optimal solution shown in Table 4.57 is obtained. Table455 Improved Solution oo wom 0 Teli soatin | ~OC mh | Minato ‘nBisis Vas a 4 P besp 4 mo [0 10 0 wo) S06 0 oo | 50 to 50 | 00150 -m 4 wo | @ 0 0 0 4 0 1} 30 300 5 wo | v1 0 0 o «+ of- F000 7300 30 cos 0 Ms 0image not availableimage not availableimage not availableLineer Programming: Simplex Method. ‘Table 4.63 Optimal Solution 97 600 300 o 0 Variables Solution % In Basis Values s B ) 00 x 10083 1 0 300 % 403 ° L = 80,0003 7% 300 G73 0 700340073 In Table 4.63, all ¢,-:, 20 andalso artificial variables have been reduced to zero, an optimum solution has been arrived at with x, = 106/3 batches of hardcover books, x, = 40/3 batches of paperback books, at a total minimum cost of Z=Rs, 80,000/3. Example 430 Anadvertising agency wishes o reach two typesof audiences: Customers with annual income greater than Rs, 15,000 (target audience A ) and customers with annual income less than RS. 15,000 (target audience B). The total advertising budget is Rs. 2,00,000. One programme of TV advertising costs Rs. 50,000; one programme on radioadvertising costs Rs. 20,000. For contract reasons, at leastthree programmes ought to be on TV, and the number of radio programmes must be limited to five, Surveys indicate that a single TV programme reaches 4,50,000 customers in target audience A and 50,000 in target audience B. One radio programme reaches 20,000 in target audience A and 80,000 in target audience B. Determine the media mix to ‘maximize the total reach, Solution Letus definethe following decision variables: x and x» = number of insertions in TV and radio, respectively. ‘Then the LP problem can be formulated as follows: Maximize total reach) Z = (4,50,000+ 50,000) x, +(20,000++80,000) x, 00,000 x, + 1,00,000.x, = Sx, +x, subject to the constraints (i) 50,000 x, +20.000.x, < 2,00,000 or Sx, +2x,<20(Adwt budget}; Gi) x23 (Adv. on TV); (ii) x, $5 (Advt. on Radio) and aym20. Introducing slack, surplus andior artificial variables inthe inequalities ofthe constraints, the standard form ofthe LP problem becomes: Maximize Z= Se, +x) +0s, +05, +08; subject to the constraints (@ Se,42e+ 3. Gi) +45 and Xy.Xy84,53,5yAj20- ‘An initial basic feasible solution is obtained by setting x, = x, = 530. This initial solution s, =20, A, 45S ond Max Z ~~ 3Mis shown in Table 4.64 ‘Table 4.64 Initial Solution G7 5 4 0 0 0 n Variables Solution x SA | Minato inBasis Values pit, 8 dense) ° 3 2 1 0 0 0 =4 -M A o 0 -1 oot 0 53 l oo 1 0 = 7M oo wo -M 1 o -w oo oryThe ¢; ~z, value in x,-column of Table 4.64 isthe largest positive value, enter variable x, to replace basic variable A, into the basis. For this apply following row operations: Ry (new) +R, (old) +1 (Key element), Ry (new) > R, (old) —5 R, (new) to get the new solution as shown in Table 4.65, ‘Table4.65 Improved Soludon = 5 1 0 ° ° a Variable Solution % nO | Minato inBasis Values alts Bo Bmx) ° 1 3 ° z 1 ® 0 | wi1> 5 z 3 1 ° oo of — ° 55 5 o 1 0 o if Z=15 3 ° 0 30 ° 1 0 50 The c, ~ 2, value in s,-column is the largest positive, enter variable s, to replace basie variable s, into the basis. For this apply following row operations: R, (new) +R, (old) +5 (Key element); Ry (new)—> R, (old) +R, (new) to get the new solution as shown in Table 4.66. Table4.66 Optimal Solution “> 5 1 0 0 0 a Variables Solution yO 5 OH InBasie Values, 5 Bexs) - ° 1 oa us 1 0 5 x 4 128 uo 0 ° 5 s o 1 oo 1 5 2 1 ° 0 0-1-1 0 0 In Table 4.66, since all the numbers cz, 0 the total reach of target audience cannot be increased further. Hence, the optimal solutionis: x, = 4 insertions in TV and x,=0 in radio with Max (total audience) Z=20,00,000. Example 4.31 Use the penalty (Big-M) method to solve the following LP problem. Minimize Z=5x, +3x, subject to the constraints @ AWt4ysl2, Gi) 2x,+25 and xy,r,20. 0, Gil) Sx, +2210 Solution Introducing slack variable s,, surplus variable s, and artificial variables A, and A, in the constraints of the given LP problem. The siandard form of the LP problem is siated as follows: Minimize Z= 5x, +3r,+0s, +0s,+MA,+MA, subject to the constraints (@ 2x, 44x, +5,=12, (i) 25, +24, 4A, =10 ii) Sx, 42st A, Sy Ay A 20. and xp,Linear Propramming: Simplex Method An initial basic feasible solutions, = 12,4, 10, A, =10 and Min Z=20M shown in Table 4.67 is obtained by leting x) =x" 330. ‘Table467 Initial Solution 6? 5 30 0 OM OM » Vriables Solution | x, m8 % A Ay | MinRatn inBass Values yin, ’ de xp i 2 2 71-0 © 0 | wae A, 10 2 2 0 -0 1 0 | W2-=5 AL 10 ® 2 09 -1 0 1 | 5-2 ™ 4 0 MM M g-3 | S-7M 3-4M 0-0 In Table4.67, sinoe the value c, -z,=5—7M isthe smallest value, therefore enter nor-basic variable x, toreplace ‘variable A, inthe basis for this apply the following row operations. The new solution is shown in Teble 4.68. R, (new) —>R, (old) + 5 (Key elements); Ry (new) —> R, (old)— 2R, (new); Ry (new) > Ry (old) -22, (new). ‘Table 4.68 9 3 3 o oom Variables Solution 1 s) ry Min Ratio inBasis Values yl B ex) 0 5 8 RCD 1-250 | S165) =4016- M A, 6 0 16s 0-25 1 | 6465) =6 3 x 2 1 1s 0 -¥s 6 | 2108) Z=10+6M 5 @MS)+2 0 @MS)-1 0 CoMs)+1 0G 2Ms)+1 0 In Table 4.68, since the value c, ~, is smallest, introduce variable x» into the basis and removes, from the basis by applying the following elementary row operations. The new solution is shown in Table 4.69, Ry (new) ~> R, (old) * 5/16 (keyelements), Ry (new) ~> Ry (old) ~ (6/5) Ry (new) Ry (new) > Ry (old) ~ 2/5) R, (new) Table 469 q> 3 2 ° oom Hirwbks Sout | my 4 a Am | Mra inBasis Values Nils B b=) 3 % 52 0 i 316 is 0 | 20 M A 3 ° ° a8 1] 2s 5 fi 1 1 ° -e tao | Z=2523M 2, 5 3 -3MRTSNG MAT OM gy [0 0 aMB-5/16 — M4478 0erry Operations Research: Problems and Solutions In Table 4.69, the value c,—z4is smallest, introduce non-basie variable s, ‘A, For this apply the following row operations: Ry (now) —9 Ry (old) 4 (key elements) ; Ry (new) > R, (old) ~ (1/8) R, (new) Ry (new) — Ry (old)+ (1/4) Ry (new). ‘The new solution is shown in Table 4.70 ‘Table470 Optimal Solution 0 the basis to replace basic variable 6? 5 3 0 0 oa Variables Soliton a 2 In Basis Values B 3 % 1 0 1 1 0 0 % 2 0 0 -32 1 3 x 4 t o 2 0 2-23 3 3 =I 0 0 0 Il 0 In Table 4.7, it is observed that all ¢ ~ > 0. Thus, an optimal solution has arived at with value of variables as Wand Min Z=23. Example 432 Usethe penalty (Big-M) method tosolve the following LP problem, Maximize Z=2r,+-x,+34, subject to the constraints () xytayt 2B (i) 2m, yt 4ey=12 and xyxyx,20 (Bharthidasan Univ. B Se (Maths) 1995: Dethi Uni, B Se (Maths), 1996) Solution introducing sick variables, and arificil variable A, in the constraints ofthe given LP problem, The standard form of the problem is stated as flows: Minimize Z=2x,-+x,+355+0s,-MA, subject to the constraints 0 xtx42y +5 and at xy ty A, 20 2,430,441 +A,=12 The initial basic feasible solution is given in Table 4.71. ‘Table 4.71 Initial Solution 47 2 1 3 0 = a a a inBais Vales mh BK 0 3 H T @® 1 c 2 3 + 0 1 “MMM 2s2M 143M 3s4M 0 Introduce variable x, into the basis and remove variable s, from the basis to geta solution by applying the following row operations. Ry (new) +R, (old)+2 (key element); Ry (new) > R, (old) —4R, (new). ‘The new solution is shown in Table 4.72.image not availableimage not availableimage not availableimage not availableop Variables in Basis B 1 % 3 x Z-473 133 Linear Programming: Simplex Method [EES % B 193. In Table 4.82, all ¢ ~z)<0,an optimal solution has arrived at with the values ofthe variables as: x{ = 11/3 or X{=2/' = 11/3-0=11/3;x,= 143 and Max Z= 47/3. Example 4.35. Solve the following LP problem by using thetwo-phase simplex method. Minimize Z=x, +x, subject to the constraints ( 2x, +4r,24, and x,,3520. Solution Adding surplus variables s, and s, and artificial variables A, and A,, becomes: Maximize 2*~~x, x) subject to the constraints (9) 2x + bxy-5, +A, =4, and Xp ppp Ay AgzO where Z*=—Z. Phase! Minimize Z'=-~A,-A, subject to the constraints (0 2x +4ty-5, +A, and XX Spr Ay Az2O. ‘The initial solution isshown in Table 4.83. i) x +7, (at Iy-st A, @ x,47%27 [Bharthidasan Univ. B Sc (Maths), 1994) ytAy=7 ‘This phase starts by considering the following auxiliary LP problem: standard form problem Table483 Ini 6? o o o + ce Tariables Solution » 8 8 8 A A in Basie Values B dex) =i Ay 4 2 tl o 1 ° 1 Ay 1 1 @ oo «1 ° L wai 8 H 1 a 3 a o Applying following row operations to enter non-basic variable x, the basis as shown in Table 4.84. Ry (new) Rj (old) x (1/7) (ley element); oniseatt the basis and removing variable A, from R, (new) ~ R, (old) ~ Ry (new)° % 1 in 1 0 - ° Zens y -B7 o 1-7 -l 3 Br 0 “1 wn o Apply the following row operations to enter non-basic variable sin the basis and removing variable A, from the basis as shown in Table 4.85. Itmay be noted that if variable x, is chosen to enter into the basis, then itwill lead to an infeasible solution. R, (new) —>R, (old) * 7 (key element); Ry (new) Ry (old) + (17) R, (new) Table 4.85 6° 0 o o ort cp Variables Solstion x % x % A a in Basis Values B b(=xp) 0 = 2 B ° -7 7 7-1 0 aS 4 2 1 ° 1 0 % 0 0 0 0 ° 9-4 ° ° 0 0 -1 + Since all ¢ ~z <0 comespond to non-basic variables, the optimal solution: x, = 0, x) = 4, $= 0,$)=21, A, =0,A)~O.with 2* = 0 is reached. However, this solution may or may not be the basic feasible solution 1o the original LP problem. Thus, we have to move to Phase Il to get an optimal solution to the original LP problem. Phasell The modified simplex table obtained from Table 4.85 is represented in Table 4.86. ‘Table 4.86 ot tt a a ‘mbar Yalu mi 2 bea) = a | a a a 2 24 Pt to 2 zm aa i a ote Introduce variable x, into the basis and remove variable s, from the basis by applying the following row operations: Ry (new) R, (old) « (1/3) (ey clement); Ry (new) — R, (old) ~2R, (new) The improved solution so obtained i given Table 4.87. Since in Table 4.87, all¢.~2,<0, the currentsolution is optimal. Thus, the optimal basic feasible solution to the given LP problem is: x, = 21/13, x,= 10/13 and Max Z*=~31/13 or MinZ=31/13.ep Variables in Basis ‘Table487 Optimal Solution tinesr Programmning! sinpioxWeined ERI mis 213 3 5 1 1 ° Solution i % Values v 0-73 3 Il 13 1 613 0 0-613 Example 4.36 Solve the following LP problem by using two-phase simplex method. Minimize 2 = x, -2t,-3x, subject to the constraints () 2x, 44473472, and x)Xy¥320. Minimize Z*=—A,—A, subject to the constraints () -2x,* xy 43x34 Ay and xpxy xy Ay A,20. Phase! Minimize Z* Ay Ay subject to the constraints () 2x4 Ky+3y+A, i) 2x, #32) 4 4x,= 1 M3 [Meerut Unix, MSe (Maths), 1991) Solution After converting the objective function into maximization form and adding artificial variables A, and ‘A, in the constraints of the given LP problem, the problem becomes (i) 2a +3 Fey AQT This phase starts by considering the following auxiliary LP problem: Gi) 25, t3my tar tAj=1 andy) Xy XA, AQZO. ‘The initial solution is presented in Table 4.88. ‘Table4.88 Initial Solution 97 0 a a 2 -1 a Variables in Soluiion x * % Ay A; Basis Values B beers) =1 Ay 2 = 1 1 0 =1 Ay 1 2 3 0 1 Ut=-3 z 0 -4 - 1 “1 ory o 4 ° o Introduce non-besic variable x, into the basis to remove 4, from the basis by applying the following row operations. R, (new) > R, (old) + 4; ‘The improved solution so obtaine 0,1,=0,x,= 1/4, 4, =S/4and 4, =| an R, (new) > R, (old)—3R, (new) dis given in Table 4.89. Since al, ¢, with Z* +, $0. the optimal solutions is: 5/4, But at thesame time, the value of Z* < Oand the artificial variable A, appears in the basis with positive value 5/4. Hence, the given original LP problem does not possess any feasible solution.FE cvmrstone Resco: Problems and Solutions ‘Table 4.89 Optimal but not Feasible Solution 5 0 6 0 = ey Variablesin Solution x x % A Basis Yalues B AL ~~ 94 0 1 5 4 12 34 1 6 7 Ey 0-1 -72_-94 0 0 Example4.37 Solve the following LP problem Maximize Z=3x, +9x, subject to the constraints ( x44) <8, (i) x 42Q <4 and X,.1,20. Solution Adding slack variables s, and s, to the constraints, the LP problem becomes: Maximize Z=3x, +9x,+0s, + Os, subject to the constraints @ 44,45, (i) x) 42045) and xy. ty 5 5920. The initial basic feasible solution in given in Table 4.90. Table4.90 Initial Solution 6 3 9 o 0 ‘Variables ‘Solution % 5 s; | MinRetio in basis ‘Values pl B bex, ° * 8 1 4 1 0 wa=2 0 8 4 1 2 0 1 42-2 Z=0 0 0 0 0 3 9 ° ° In Table 4.90, c, ~2,=9is the largest positive value, entering non-basic variable x, into the basis. However, both variabless, and s,are eligible to leave the basis as the minimum ratio is same. ie. , so there isa tie among the ratio in row s, and s. This is an indication of the existence of degeneracy. To obtain the unique key row, apply the following procedure for resolving degeneracy. (@ Write the coefficients ofthe slack variables as follows as shown in Table 4.90. Row | Column 51 2 a [1 0 » | 0 1 Gi) Dividing the coefficients by the corresponding element of the key column, we obiain Row Column 5 % VE W4=0 % a2=0 © 2=1/2Linear Programming: Simplex Method )) Comparing the ratios of Step (i) from left o right columnwise, the minimum ratio occurs for the second, row. Therefore, the variable s, is selected to leave the basis. The new solution is shown in Table 4.91. ‘Table491 Optimal Solution cha 67 3 ’ 0 o Variables Solution x % % % inBasis Values B 0 a =1 0 T ° 4 2 v2 1 ° 12 Z=18 on 9 ° 2 6 0 In Table 4.91, all ¢,~z,<0. Therefore, an optimal solution is arrived at, The optimal basic feasible olutionis: x, =0,x)=2and Max Z=18, Example 4.38 Solvethe followingLP problem Maxitnize Z = 6x, +4, subject to the constraints (i) 2x, +3x,530, and xy,%,20. (i) 3x,+2xy 24, Gi x + Is the solution unique? If not, give 2 different solutions [Jodhpur Univ., M Se (Maths), 1993] Solution Adding slack variables s,s surplus variable s, and artificial variable A, in the constraint set, the LP problem becomes Maximize Z= 6x,+ 4x, + ry + Os) +08;- MA, subject to the constraints ( 2xy+3x+5,=30, Gi) 3x, +2e,+5,=24, and xy.Xq.5y, 5 5A, 20. ‘The optimal solution for this LP problem is presented in Table 4.92, ‘Table 4.92 Optimal Solution Gil) yx, -85 +A, ey ‘ + 0 0 0 we Variables Saution rs infant Values sie 2 7 7 asa i 5 Seo wat “ 2 1 mo ia 0 | ma 5 6 on) 20 0 0 0 3 0 ‘The optimal solution shown in Table 4.92 is: x, =8,x)=0 and Max Z = 48, But as c, and.x, is notin 2 the basis. Thus, an alternative optimal solution can also be obtained by entering non-basic variable.x, into the ‘basis and remove s, from the basis. The new solution is shown in Table 4.93,‘Table 4.93 Alternative Solution 6 4 a 0 o es Variables * % s es i in Basis B 4 % 0 1 35 = 215 0 0 5 0 0 ws vs 1 6 a 1 0 -us 3S 0 = 48 6 4 0 2 0 0 ° 0 =2 0 ‘The optimal solution shown in Table 4.93 is: x, solution. ‘The infinite number of solutions which can be obtained for this LP problem are as shown in Table 4.94. 12/5, x,=42/5 and Max Z = 48 which is an alternative optimal, Table 4.94 Infinite Solution ‘Solution Values General Solution if 2 * 8 15 = 84 (SKI) a ° aus yO + (DISH =D) 5 4 0 3, =i + 01-2) 53 5 305 55 =5A+ BY5)U-2) For each arbitrary value of 2, the value of objective function will emain same. Example4.39Solvethe followingLP problem Maximize Z=3r, +5x, subject to the constraints @ 41-2456, fi) 1,510, Gi) x21 and xj,%,20. Solution Adding slack variables s,,s,, surplus variables and artificial variable A, in the constraint set, the LP problem becomes Maximize Z=3r, +5x, + Os, +05, +0s,—MA, subject to the constraints @ y-2y+5,=6, GD y= and psy Spy A, 20. ‘The initial solution to this LP problem is shown in Table 4.95, ‘Table4.95 Initial Solution 0, @ H-5,4A 62 3 5 oo 0 op Variables Solution ry » 5 2 ‘Min Ratio inBasis Values Xl B Bex) 0 4 6 1 =2 1 0 0 0 = 0 % 10 1 0 o 4 o 0 — 4 1 0 1 oo -1 1 13 0 “mM 0)0OU0)COUM OM 3 stm 0 0 MoSince c,~z, in x,-column is largest positive value, entering non-basic variable x, inthe basisand A, leaves the basis. The new solution is shown in Table 4.96. Table4.96 Unbounded Solution 9? 3 5 0 0 oO -M Variables Solution * % rn % % A in Basis Values B ix, 0 s 3 T 0 1 0 2 2 0 sy 10 l 0 0 1 0 0 5 » 1 ° 1 ° o 1 1 Z-5 5 ° ° os 3 a3 3 0 0 0 5 -w-s In Table 4.96, c,~z, ins,-column is largest positive value, entering non-basic variables, in thebasis. But the ‘coefficients in the s,-column are all negative or zero, This indicates that the variable s cannot be entered into the basis. However, the value of s, canbe increased infinitely without removing any one of the basic variables ‘rom the basis. Further, since is associated with x, in the third constraint, x, will also be increased infinitely, because it can be expressed as: x, Example 440_ Solve the following LP problem Maximize 2= 6x, +41, subject to the constraints © xx 55, and 4,520. Solution By adding slack, surplus and artificial variables, the problem becomes Maximize Z = 6x;+ 4x, +0s, +0s,-MA, subject to the constraints Oxtyty-s 6 and yp 5,5, 20. ‘The initial solution to this LP problems shown in Table 4.97. ‘Table 4.97 Initial Solution Gi) 28 xy tAy=8 + 5,—A\. Hence, the solution to the given LP problem is unbounded. 6 4 oo “ ‘Variahtes mA) Minato in Basis pay a T o 1 a) si AL 0 c o-oo wi 5 7 om 0 ° 24M 0 wo‘Since cy -2, value in the x,-column is largest positive value, entering non-basic variable x, leaves the basis. The new solution is shown in Table 4.98, ‘Table4.98 _Infeasible Optimal Solution 97 6 4 a a oe Variables Solution * % * " A in Basis Values B bexp 4 % 3 1 1 T 0 D -M 4 3 =1 ° “1 =1 1 2=20-3M 3 eM 4 43M OM OOM ory 2M 0 -4-M o Since all ¢ ~z,< 0, the solution shown in Table 4.98 is optimal. But this solution isnot feasible forthe given LP problem since it has x, = 0 and x = S (recall that in the second constraint x, 2 8). The presence of artificial variable A, =3 in the optimal solution indicates that the optimal solution does not satisfy second constraint (x,28). Hence, the solution shown in Table 4.98 is infeasible eptimal solution. Example 4.41 _ Use the two-phase simplex method to solve following LP problem Maximize Z=5x,—4x, +31, subject to the constraints (@ 2x, +x) 6x,=20, (i) 6x, +5x)+ 10x, $76, (i) 8x, 3x, +6x,<50 and pexgnts 20. (Punjab Unix, BE (Elect,), 1994) Solution Adding slack variables s,s. and artificial variable 4, in the constraints, standard form of given LP model becomes: Minimize Z” = Ox, + x, + 0x, +0s,+0s,+A, subject to the constraints ( 2x, +e, ~6r, +A, =20, Gi) 6x, + 5x, +10r,+ and x),X3,%555,5p,A, 20. i) 8x, 3x, +6, +8, Phase | The basic feasible solution to the auxiliary LP problem is obtained by setting variables '5=0. The basic feasible solution: 4, =20, , = 76, , = 50s shown in Table 4.99. Table 4.99 Initial Solution 6° 0 oo o o 2 Gp ‘Variables ‘Solution | % = 5% Ay | Minka inBasis alt, 8 rr 2 1-6 0 0 1 0 ° n 6 sm 1 0 0 | 380 55 ® 3 60 10 254 > 7-20 5 2 1-6 0 oO 1 6-4 | 2 -1 6 0 0 0 In Teble 4.99, c,—z, <0 in,-column, enter non-basic variables, toreplace basic variables into the basis. For this apply following row operations: Ry(new)— R,(old) * (1/8) (key clement); (new) FR (old)-2 R,(new); Ry{new) > R, (old) -6R(new) to get the new solution as shown in Table 4.100.Ue rig Si i ES ‘Table4.100 Improved Solution 5° 0 o 0 9 0 1 oe Tariables Solution | SH A) Min Rano inBasis Vales walt: B 1 4 0.) 2 0 14 307 > ° 4 oe ets | sas 0 x 1 a8 soo | 508 5 0 m™ S20 a ons om 2 0 In Table 4.100.c, —2, <0 in, ~column, enter non-basic variable x, to replace basic variable A, into the basis. Applying following row operations R, (new) > R (old) 4/7 (key element), Ry(new) > Ry(okd)+(3/8)R (new) to get the new solution as shown in Table 4.101 ‘Table4.101 Optimal Solution R, (new) > R(old) ~(29/4)R (new), ad 0 o 0 0 0 Variables Solution * % % 4 Sy in Basis Values B Bex») 0 a 3077 0 i 0 7 0 1 su 0 0 1 27 0 x 397 1 0 0 ma z=0 % 0 0 0 0 0 67% ° o ° o 0 In Table 4.101, all c)—z; values are non-negative and hence the current solution is optimal. Also, since ze and no artifizial Variable appears in the basis, this solution is also the basic feasible solution Phase ll The initial simplex table for this phase is the optimal simplex Table 4.101 of Phase I, as shown in Table 4.102 ‘Table 4.102 Improved Solution 67 5 “4 i o ° o Taraites — Soluton | % 5 aw in Basis Values B ex) ~4 % 307 7 a 7 ° a 227 ° oe 27 5 ¥ 387 1 0-07 oo uns Za 1557 5 3 a oF 0 ia ° 0-97 oti Since all c;—z, values are either negative or zero nor-basic variables columns, the solution shown in Table 4,102 is an optimal basic feasible solution. The optimal solution is: x, = 55/7, x, = 30/7, x, = 0, and Max 2: $57.Example 4.42. Usetwo-phase simplex method o solve following LP problem Maximize Z=3x, +2r,+2x, subject to the constraints Osx tI t4yS7, CD) —4r,+7H+5xj2-2, (il) 3x, 4+4x,—6x,22977 and Xp Xyxj20. [Punjab Univ, BE (E & C) 2006, BE (17) 2004) Solution Since RHS of constraint 2 is negative, multiplying it by - 1 on both sides and expressed it as: 4r,— Try - Sx, $2. Phase! Introducing sleck, surplus and artificial variables in the constraints, the standard form of LP problem becomes: Minimize Z°= Or, + 0x, + 0x, +05, +05, +05,+A, subject to the constraints @ Sx +Ieyt4xyt5,=7, (i) 4e)—Txy-Sxy+5)=2, (iil) 3x, +4e~6r,—8,+4,=2917 and pty Ayes Spy A120. Setting variables x, =x, =x, =s, = 0, the basic feasible solution shown in Table 4.103 to the auxiliary LP problem is: 5, , Ay = 29/7 Table 4.103 Initial Solution o 0 0 0 9 © Variables "4% 5% Ss Ay] MinRatio in Basis aly & B 0 Hi 7 5 @® # 1 » @ of i> 0 5 2 4-7 -5 0 1 o of - 1 4 3004-6 0 0 -1 1] 298 72297 3 4 6 O 8 -1 Jv 3-4 6 0 ot Since c, ~z,=~4 is the largest negative value underx,-column in Table 4. 03, replacing basic variable s, with the non-basic variable x, into the basis. For this, applying following row operations: R, (new) 98 (old) = 7(key element); Ry(new)—> Ry(old)+7R, (new), Rylnew) > R(old) -4R, (new) to get an improved solution as shown in Table 4.104 ‘Table 4.104 An Improved Solution go 0 et Taiabls Solan || a ‘nBass Talus, ne 2 bem 0 S 1 m1 aan] o = ° 9 0 4 4 to of 4 1 4 nm | @ 0 #7 nm 0 4 1] 13 Zin 5 a gy [an 0 a a ot Solution shown in Table 4.104 is not optimal and there isa tie between the s) and A, -Tows forthe key row. The A,-row is selected as the key row because preference is given to artificial variable, Thus (1/7) is the key clement, Replacing basic variable A, in the basis with the non-basic variable x,. Applying suitable row opera- tions to get the new solution as shown in Table 4.105.Table 4.105 Optimal Solution g oo 0 09 0 0 a ie ean om imBash Values a Heap 7 & mi I 0 i 0 o 0 © 7 1 8B 0 # 1 oe Zo a i} 0 0 } In Toble 4.105, all ¢,~z, values are 2ero under non-basic variable columns. Thus, the current solution is optimal. Also Min 2’ = 0 and no artificial variable appears in the basis, and hence this solution is the basic feasible solution tothe original problem. Phase Il Using the solution shown in Table 4.105 as the staring solution for Phase II and carrying out computations to get optimal solution as shown in Table 4.106 ‘Table 4.106 Initial Solution ae 3 2? @ 6 0 cy Variables Solution |» yA | Mina inBasis hh B 2 2g a ° Te 3 © 3 | war 0 s ° ° .@ 7 1 6 o> 3 i 1 1 en 5 0-7 = Z=a57 3 ?-% 68 ee ee In Table 4.106, c,~z,~ 92 is the largest positive value, replacing basic variables, with non-basic variable 4% into the basis. For this apply necessary row operations as usual. The new solution is shown in Table 4.107 Table 4.107 OptimalSolution 9? 32 2 o o o oe Variables Solution | 2% % ss inBasis Values B em) 2 Ps 7 0 1 0 9a ~aa/s2i — 41/521 2 % o ne) Ws2i 63/521 3 x 1 ee ee ee Z-357 >a 2 aywsai salsa _asisaa 0-0 0275219521 - 651521 Since all c)~ z, values are either negative or zero in Table 4.107, the current solution isthe optimal basic feasible solution: X= 1,1) =2/7,x,= OandMax Z=25/7,Pr 1 A television company has three major departments for the ‘manufacture of its two models, A and B. The monthly capacities are givenas follows: ‘Per Unit Time Requirement Gar) Hours Available this Month ModelA _ MocklB Department 1 4.0 20 1,600) Department I 2.5 10 1,200 Department Ill 4.5 1s 1,600 ‘The marginal profit per unit of model A is Rs. 400 and that of model B is Rs. 100. Assuming that the company can sell any quantity of either product due to favourable rarket conditions, determine the optimum output for both the models, the highest possible profit for this month and the slack time in the three departments. . A manufacturer of leather belts makes three types of belts A, B and C which are provessed on three machines My, M and My. Belt A requires 2 hours on machine M, and 3 hours onmachine M, and 2 hours on machine M,, Belt B requires 3 hours on machine M,,2 hours on machine M, and 2 hours on machine M, and Belt C requires 5 hours on machine M, and 4 hours on machine M, ‘There are 8 hours of time per day available on machine M,, 10 hours of time per day available on machine M, and 15 hours of time per day available on machine M,.The profit gained from belt AsRs, 3.00 per unit, fiom Belt Bis Rs. 5.00 per unit, from belt is Rs. 4,00 per unit. What should be the daily production of each, type of belt so thatthe profit is maximum? [Meerut Univ M Se (Maths), 1998) ‘A company produces thee products A, B and C. These products require tree ores 0, 0, and O. The maximum quantities ofthe ores 0}, O, and 0, available are 22 tonnes, 14 tonnes and 14 tonnes, respectively. For one tonne of each of these products, the ‘ore requirements are 4 a c 0 3 = 3 0, 1 2 3 0; 3 2 3 Profit per tonne 1 4 5 (Rs. in thousand) “Thecompany makes a profitof Rs. 1,000, 4.000 and 5,000 on each tonne ofthe products A.B and C, respectively. How many tonnes of each product should the company produce to maximize its profits. ‘A manufacturing firm has discontinued production of a certain "unprofitable product lie. This has created corsiderable excess production capacity. Managsmeat is considering to devote this excess capacity to one or more of three products call them product 1, 2and 3. The avilable capacity on the machines which might limit output is summarized inthe following table: Available Time (in Machine-hours per Week) Milling Machine 250 Lathe 130. Grinder 0 ‘The numberof machine-hours required foreach unit ofthe respective product is as follows: ‘Machine Type Productivity (2 Machine-hours per Unit) Product — Product Product 1 2 3 filling Machine 8 2 3 Lathe 4 3 0 Grinder 2 1 ‘The profit per unit would be Rs. 20, Rs. 6 and Rs. 8, respectively for product 1, 2 and 3. Find how much of each produet the firm sbould produce in order to maximize profit. ‘The annual handmade furniture show and sale occurs next month and the school of vocational studies s planning to make furniture for thesale. There arethree wood-working clarses—I year, year and II] year, atthe school and they have decided to make styles of chairs —A, Band. Each chair must receive work ineach class and the time in hours required for exch chair in each clas is: Chair Tear Year M1 ¥ear A Zz 4 3 B 3 3 2 c 2 1 4 Dring the next month there will be 120 hours available to the I year class, 160 hours to the II year cass, and 100 hours to the IIL {year cls to produce the chars. The teacher ofthe wood-working, lasses fels that a maximum of 40 chairs can be sold a the show. The teacher has determined thatthe profit from each type of chair will be: A, Rs. 40; B, Rs.35 and C, Rs. 30. How many chairs of each type shouldbe made in order tomaximize profits atthe show and sale? Me Jain, the marketing manager of ABC Typewriter Co. is tying, to decide how to allocate his salesmen to the company’s three primary markets, Market Is an urban area andthe salesman can, sell, on the average 40 typewriters « week. Selesmen inthe other two markets, II and III can sell, on the average, 36 and 25, typewriters per week, respectively. For the coming week, three of the salesmen will be on vacation, leaving only 12 men available for duty. Also because of lack of company care, a maximum of $ salesmen can be allocated to market ares 1. The selling expenses, per week fo salesmen in each area are Rs 800per week For area I,10. Rs. 700 per week for area I, and Rs, $00 per week for ares IL. The budget for the next week is Rs. 7,500, The profi: margin per typewrier is Rs. 150. Determine how maay salesmen should be assigned to each area in order to maximize profits? An animal feed company must produce 200 kg of a mixture | consisting ofingredientx, and x, daly. x, costs Rs. 3per kgand.x, ost Rs. 8 per kg Not moe thn 80 kg o can be used and at | least 60 kg ofxy must be used. Find how much ofeach ingredent_| shouldbe used if the company wants to trinimize costs. | | | | | | ‘A diet ito contain atleast 20 ounces of protein and 15 ounces of| ceuohydrate. There ae thtoe foods A, B and C available in the ‘marke, costing Rs, 2, Re. ! and RS. 3 per unit, respectively. Each LunitofA contains ources of protein and 4 ounces of carbchydrate; cach unit of B contains 3 ounces of protein and 2 ounces of carbohydrate; each unit of € contains 4 ounces of protein and 2 ‘ounces of carbohydrate, How many units ofeach food should the dict contain so that the cost per unit et is minimum? | ‘A person requires 10, 12 and 12 units of chemicals A, B and C, | respectively for his garden, A liquid product contains 5,2 and | uit of A. B and C, respectively per jer. A dry product contains 1. 2.and 4units of A, B and C per unit. Ifthe liquid preduct sells for [s.3 per jarand the dry product for Rs, 2 per eaton, how many of each should be purehased in order to minimize the cost and meet the requirement? ‘A scrapmetal dealer as recived an oder from @ customer fer at least 2,000 kg of scrap metal. The customer requires that at last 1,000 kg of the shipment of metal be of high quality copper that ccan be melied down and used to produee copper tubing Furthermore the customer will rot accept delivery ofthe order if near Programming: Simplex Method ETS} ‘manufacturing of components for assembling and for packing. A dozen units of model A require one hour of manafacturing, two ‘hours for assemblingand one hour for packing. The corresponding figures fora dozen units of mode! Bare 2, | and 2, and for a dozen ‘its of Care 3,5 and 1, whilea dozen units of medel D require 1 hour of packing only. During the forthcoming week, the company sll beable to make available 15 hours of manufacturing, 20 hours ‘of assembling and 10hours of packing time. Determine the optimal ‘rodution schedule for the company. [Madras Univ., BE (Civil, 1991; IAS, 1995] A transport company is considering the purckase of new vehicles for transportation between the Delhi Airport and hotels in the city. ‘There are three vehicles under consideration: Station wagons, ‘minibuses and lange bases. The purchase price wouldbe Rs. 1,45,000 for each station wagon, Rs. 2,50,000 for the minibus and Rs. 4,00,000 forthe large bus. The board of directors has authorized 2 ‘maximum amount of Rs. 50,00,000 for these purchases. Because of the heavy air travel, the new vehicles would be utilized at ‘maximum capacity, reganiless of the type of vehicles purchased ‘The expected net annual profit would be Rs. 15,009 forthe station wagon, RS.33,000 forthe minibus and Rs. 45,000 for the large bus. ‘The company has hired 30 new drivers forthe new vehicles. They ‘are qualified drivers for allthree types of vehicles. The maintenance department has the eapacity to handle an additional 80 station ‘wagons. A minibus is equivalent to 1.67 station wagons and each large bus is equivalent to 2 station wagons in terms of their use of the maintenance department. Determine the number af each type ‘of vehicle wo purchase in order to maximize profit. [Delhi Univ, MBA, 1998) and D which have profit contributions of Rs. 8, Rs. 1S and Rs. 25 ‘on models A,B and C, respectively and a loss ofRe. 1 on mode! D. Each type of radio requires a certain amoust of time for the is first made in the carpentry shop and then furnished, waxed and polished in the finishing shop. Man-hours required in each are: it contains more than 175 kg metal that he deems unfit for | 'S- The Omega Data Processing Company performs thres types of outsell sae. fac woul tat carton ca cacoaaive acanaat oF stvities: Payroll, account receivables, and inventories. The profit ‘impurities and cannot be melted down and refined profitably. an, eck es rasp - ilies? oe i office The dealercan purchase terap metal fom twodiffrent suppliers printing fora standard job are shown inthe following table: {in unlimited quantities with following percentages (by weight of Tob Profi/Standord Time Requirement (in) high quality copper and unfit scrap. fo Peck Gi BE pA Spr end at oom J0b'RS) Kespunch Conputtion Printing — Sep Sirti: B Payroll 215 1,200 20 100 Copper 25% 75% Arereceivable 125 140015 0 Unfit scrap 5% 10% Inventory 225 800 35 80 ‘The cost per kg of metal purchased from supplier A and supplier Omega guarantees overnight completion of the job. Any job Bare Re. | and Rs.4,respecively. Determine te optimal quantities | _sthedvled during the day canbe completed during the dey or night ‘of metal to be purchased for the dealer from each of the two Any job scheduled during the night, however, must be completed supplies [Gujarat Unix, MBA, 1990) ‘during the night. The capacities for both day ard night are shown in . A markcting manager wishes W allocate his annual advertising | the following table: budget of Rs. 2,00,000 to two media vehicles A and B.The unit | costofamessage inmedia AisRS, ,000andthatofBisRs. 1.500, | _Careior Min) _Kerpench Computation _ Print “Media A i¢amonthly magazine snd not more than one insertion is Day 4200 150 400 desired in one issue, whereas at least five messages should appear | Night 9.200 250 650 inmedis B. The expected audience forunit messagesin media Ais tS 40,000 and that of media B is 53,000. Develop an LP model and | Determine the mixture of standard jobs that should be accepted solve it for maximizing the total effective audience. | daring the day and night. A sanssor radio company manufactures fou males A B,C | 16, A fumiture company can produce four types of cis. Each chat‘Carpentry stop 4 9 7 0 Finishing shop 1 1 3 40 Contribution per chair(Rs) 12-20-1840 ‘otal number of man-hours available per month in carpentry and Finishing shops are 6,000 an 4,000, respectively. Assuming abundant supply ofraw material ard demand for fished products, determine the number of each type of chairs to be produced for profit maximization 17. farmer has a 100 acre farm. He can sell ll the tomatoes, lettuce ‘or radishes which he produces. The price hecan obtain is Re. per kilogram for tomatoes, Re. 075 a head for lettuce and Rs 2 per Kilogram for radishes. Te average yield per sere is2,000 kilograms ‘of tomatoes, 3,000 heads of Tetuce and 1,000 kilograms of radishes. Fertilizer is available at Re. 0.50 per kilogram and the mount required per acre is 100 kilogram each for tomatoes and lettuce and 50 kilogram for radishes. Labour required for sowing cultivating and harvesting per acre is 5 man-days for tomatoes and radishes and 6 man-days for lettuce. A total of 400 man-days of labour are available a Rs. 20 per man-day. Determine crop mix 0 as to maximize the farmer's total profit. {Delhi Uni, MBA, 2000] 18. A manufacturer has two products P, and Pa, both of which are prodaced in two steps by machines M, and M. The process time per hundred forthe progucts on the machines are M, My Contribution (per 100 unis) 0 pf os 2 s Available (hrs) “The manufactureris ina market upswing and can sell as much as hhe can produccof both produc. Fermulate the mathematical model and determine product mix, using simplex method. 19. A company sells two different products A and B. The selling price and incremental cost information are as follows: 22, Solve the following LP problems using the simplex method. (i) Max Z= 3x, + 2x, Product A Product B Selling Price Rs, 60 Rs. 40 Variable Cost Rs. 30 Re 10 The two products are produced ina common production proe- ess and are sold in two different markets. The production process ‘ha aeapacity of 30,000 men hours. takes three hour to produce ‘unit of A and one hour to produce a unit of B. The market has ‘been surveyed ard the company official feel thatthe maximum ‘numberof units of A that cen be sold is 600 and maximum forB is 1,200 units. Subject to these limitations two products canbe sold in any combination. Formulate and solve this problem simplex method. ‘A product is manufactured by blending thee different raw materi- ‘als. The finished product should meetcerain quality requirements. Given the folowing data wia is your recommendation with regard to quantity for raw materials tobe blended, which will met the ‘quality requirements with minimum cost. Quality ‘Coniributionto Minimum Characteristies Quality Quaiity 4 pc Retirements 1 3.00 «0 10 2 $1 2 15 (Cost of raw materials perunit(inRs) 2S LICWA, June 1990) A firm has an advertising budget of Rs. 7,20,000. It wishes to allocate this budget to wo media: magazines and television, sothat the total exposureis maximized. Each pageof mgacine advertising is estimated to result in 60,000 exposures, whereas each spot on television is estimated to result in 1,20,000 exposures. Each page ‘of magazine advertising costs Rs 9,000 ani each spot on television ‘costs Rs. 12,000. An additional condition that the firm bas speci- fied is that atleast two pages of magazine advertising be usedand at Jeast3 spots ontelevision. Determine the optimal medis-mix for this firm, i) Max Z= $x, + 3x subject to xtys4 subject to ats 2 aja $2 Sx, 125 $10 and yxy 20 By, +8, <12 [Madurai Univ, MCom, 1997; 14S (Main), 1992 and x4.% 20 Calicut Unt, B Tech, 1990) (ii) Max Z=3x, +244 5x, (iv) Max 2= x, ~3x,+ 2, subject to x,t 2p + $430 subject to By - yt ys 7 art 2k 5 460 -2y +4 S12 xt dy, < 420 4x, + 3x, + By S10 and Ayipye 0 and XpXh 2 0 [Karnataka Univ, BE (Mech), 1992; IAS, 1993; Dibrugarh Uni, M Se (Stet, 1904)(vy) Max Z: bx, + Sx +945 + Hx, subjectto x, + xp + mt mys IS Te, + Sy 3x4 2x4 S120 3x, + Sep + 10x, + 15x, < 100 phat 0 3+ 2y+ 4 <3 + xt 2452 papery 20 fix) Max Z= 107s, +) +2 subjectto Mr, + gy ~ Gry t+ Ong <7 lex, + 05 + 6x, <5 a Hy <10 and Rpt ayry = 0 [Meerut Univ, MSe (Maths), 1992; (x) Max Z: subject to at x S10 ys 2 at dys 0 and 2 0 Linear Programming: Simpiox Nethod EE (i) Max Z=x, +3, +3; subject to dey + Sey + 3x, € 18 Hox, + 7x + x, $12 and SpAyHy2 0 (ii) Max 2 2r, + 41, 4 3x5 ty sbjetto x,t 3y+ oS 4 ryt oy <3 tay cd and Kyla tye 20 (6) Max Z= 41, + 10x, subject to 2yt as 50 2x, + Sx, < 100 2x, +3m $90 and aye 0 (Gai) Max 2= 4 +4 +345 +5, subject 04x) 61 — Sx, #4xy 2 -20 2x thet my < 10 Br —3x, +3, +244 $20 and 20 ‘has an unbound solution, 23. Use the two phase method to salve the following LP problems: (Max Z= 3x, subject t0 d+ my 22 xt xy s2 aed ani sum 20 (ii). Min 2= 7.5, ~3ry subject to ary 23 Hoyt ye? and rakes 20 (0) Min Z= 55, +86, subject to 3 +2y 23 stn 24 mapas and mm 20 (il) Min Z = x, ~21— 3x, subject to = 2x, + x + 2x =2 2x) + By + 2x = 1 ond spp 20 (iv) Min Z= 35, 5 subject to y+ x22 at 3m, $2 a4 and x20 [Kanpur Unis, B Se (Stat), 1997; Bharthidasan Unis, B Se (Maths), 1995) (i) Min Z= 31, +2, subject to dt ye 2 Bx, $4x, 2 12 end yxy 0 [Kanpur Unin, B Sc (Sta, 1997: Bharthidasa Univ. B Se (Maths), 1995) 24, Use penalty (Big-M) method to solve the following LP problems: (@ Min Z= 3x, —x; subject to dat x 22 ayt3y $3 x24 xyz 20 (ii) Min Z= 25, +, subject to Bxt 3 4x +3x, 26 x +2 $4 xem 20Gil) Max Z= 3x, +24) + 3x, subject to mt gt 4s Bx + 4 + 25 58 and Ste 20 [Madurai Uni, B Se (Maths), 1998] (©) MinZ=x, 3x, +28, subject to ay gt igs 7 2+ 4+ S12 = 4r, + 3, + Bry S10 and Xp 2 0 [AS (Main), 1993) (iv) Max Z= x, +4, subjectio§ xt x+y tay ed ay t 2g t ay tay 4 xt2gtn <4 [MS Triunvelli Univ, B Se (Maths), 1995] (vi) Min Z= 5x, + 2x, + 10e, subject 10 = £10 xt 210 and ptph 2 0 25. Solve the following LP problems and remove the complication (if any) () MaxZ= 2x, +3x,+ 101, subject to xt 2y=2 gt aol and XH 20 Gi) MaxZ= 5, +32, subect to nt ms Sr, + 2x, < 10 3x, + 8S 2 and phe 0 (©) Max z= 8, subject to inn RS 2x, + 3x, S 6 and 2% unrestricted. (i) Max 2 = Sx, -2x,+ 3% subject 0 2x, + 2-422 By 453 yt dy ss and Xptpy 20 (Gv) Max 2 = 22x, + 30x, + 25x, subject 10 2x + 2n+ S100 2x + Rt 3 2 100 xj + 2x + 2x $100 and pty 2 0 26. Solve the following LP problems wo show that these have alternative optimal solutions. @ Max Z= 6c, +32, subject to mt us 8 3 + xy S18 ns3 apn? 0 2435-24 stint ay =15 d+ xt Sx 220 at dy t yt my 210 and Synths 2 0 Gi) Min 2=2e, + 8x, 27, Solve the following LP problems to show that these have an unbounded solution subject to * <5 2x, - 3x $6 and p20 Gil) Max Z= 107x, +, +21, subject © Mx +a — Gry + Sry = 7 Mr, 405-6 9 <5 wo ho SO and Fy kp tury 20 subject io Sy t 210 2x, + 2 214 xt 4 212 and xm 20 (i) Max Z=3x,+ 6, subject 0 Bayt Ay B12 “tt ys 4 and X20 (iv) Max = subject 10 2y- 52 <4 and Aye 20i a Linear Programming: Simplex Method BES 28. Solve the following LP problems to show that these have no feasible solution, () Max Z= 2x, +32 (i) Max Z= ar, +454 445+ 5xy subject to x24 subject to dr, + Gx, = Sey + dx, 2-20 x + $6 Bry y+ dre ys 10 $2 Bx - yt drys xy S10 20 a, - 3x, > By + 2y S20 sympa 2 0 Gil), Max Z= x, + 3x, (iv) Max Z= 3x, +25 subject to ao 1 subject to dnt xy $2 3x; 3 3x, + 4x, 212 and ° and apap 20 [Meerut Unin., M Se (Maths), 1990) HINTS AND ANSWERS 1.x, and x= units of model A and B to be manufactured, respectively Max Z= 400, +400, subject () 4x,+ 21, $1,600, Gi) 55/2 +x, 5 1,200, Ans. x, = 385.5, 2) = Oand Max 7 = 1 42,2292 2. xox; and x, units of type A, B and C belt to be manufactured, respectively Max Z= 3x, + 5:3 + 4r, subjectto @) 24,435 S8 i) 2y +5110, (il) 3xj 424; FAS 15, and xy, Ans. x, = 80/41, x, =S0/41 x, = 64/81 and Max 2 = 775/41 3. jx, and x,— quantity of product A, B end C to be produced, respectively Max Z=x, #41, +51, subjectto () 3x, +38, 522; Gi) xy +2x, +3514 ii) 3x, +255 14, and xy.x5 Ans. ¥; ~ 0,3 =7,%;~ Oand Max 2 = Rs, 26,000 Sxy/2+ 3e/2$ 1,600, and x,,1,20 4. x,.%5 and x, = numberof unit of product 1, 2 and 310 be produced per week. respectively Max Z= 201, + 6x, +88 subjectto () 8, +2xy +345 S250; Gi) 4x, +34, S150 i) 2c, +H,S50, and x,y .2520 Ams. x; = 0,35" 50, ~ SO and Max Z = 700 5, xyyx, and.x,~ acreage of com, wheat and soysbean, respectively Max Z= 30, +408, +20, subjectto ()10r, + 12x, + 7x, $10,000; (i) 7x, + 10x, * 8x4$ 8,000. Gil) x, +x, +24 $1,000, end xy, x5, 20 Ans. x; =250,x,= 625, x5 = 0 and Max Z=Rs. 32,500. 6. xj.x, and x,~ number of units of chair of styles A, B and C respectively. Max Z= 405, + 35x, + 30%, subjectto () 2x; + 3x5 +2x,$120; (i) 4r, +3x, +x, $160 Gi) 3x, +2r, 44x, $100; (iW) x) Hx, Hx 40 and x4.%5.% 20 ANS, x, = 20, x = 20,3 =0 and Max Z'= Rs. 1,500 7. jx, and x) salesman assigned to arca, 1,2 and 3, respectively. Max Z= 40 x 150, +36 * 150, +25 > 150: ~ 800s, + 700x, + 500x,) subjectto () xy 4x, +x,$12; (i) x, $5; (ii) 80Gr, +7001, + 500K, 7,500 and x5, 23,520 8, x and y= number of kx of ingredients x, 1, respectively Min (total cost) Z = 3x + 8y subjectto (i) x+¥=200; Gi) x80; Gil) y260; and x.y20 Ans. x 80,y= 120 and Min Z= Rs. 1,200 cruseeht29.x. and.x, = number of units of food A, B and C, respectively which a diet must contain. in (total cost) Z~ 2x, 4.ap +83 subject to (i) 2x, +2r,+ 4x 220; (il) 4x, +2xq +2215, and x,.¥.2520 10, .x, and.x,= number of units of liquid and dry product produced, respectively. Min (otal cost) Z=3x, + 2x, sujet to (i) Sxj+x,210; (ii) 2x,+25212; ii) x, + 44,2125 and xy,2,20 Ans, x,~ 1,x)= SandMinZ=13 11, x and.xy = numberof scrap (in ke) purchased from supplier A and B, respectively Min total cost) Z= x, + 4x, subject to (f) 2.25%, +0.75%2 1,000; (i) 0.05 x, + 0.102175; Gili) x,+%22,000; ard xy.x20 Ans, x, ~ 2500, x5~ 500 and Min 2~ 4,500 12. x, and x,= number of insertions messeges for media A and B, respectively. (total effective audience) Z = 40,000, + 55,000, subject to (i) 1,000 x, + 1,500.x,2,00,000; (i) x, $125 Ans, x, = 12, x= 16/3 and Max Z=77,3333.33, 13. x, and.r,~ number of radio and television programmes, respectively. Max (total audience) Z = (4,$0,000 + $0,000) x, + (20,000 + 80,000) x, subject to (i) 50,000x, + 20.000x, < 2,00,000; Gi) x, <3; (il) x, $2, and y,x,20 Ans, x,=4,x, = Vand Max Z-=4 * 5,00,000 14, x,y), nd 3, = unit of model A, B, Cand D to be produce respectively. Max (total income) Z= 8x, + 15. + 25%5~x4 subject to () x, +2431, 15; i) 2x, +x, 451, =20; (i) x, +2ey +x, +xy= 10: and y.xy.44420 Ans, x = 5/2.) = 5/255 =5/2,x4=0 and Max Z= 120. 15, xj,x, and.x, = number of station wagons, minibuses and large buses to be purchased, respectively Max Z= 15,000, + 35,000x, + 45,000, subject to i) x) +4; +xy £30; Gi) 1,45,000x, + 2,50,000s, + 400,000, < $0,00,000 Gi and 21.%9.%420 16. xy represents ith job and jth activity Max Z=275 (3, +43) + 1254) # X92) #225 (ry) tx) subject 0 (9) 1,200 Gy +352) + 1,400 (x + 459) + 800 Oy) 7x39) $13,400 (i) 20 Ge, + x43) + 1S Geqy +32) +35 yy +59) $400 Gi) 100 (x44 + 512) + 60 Gy, + 249) + 80H) +.%52) $1050, (iv) 1.2005; + 1400r3» + 800%» $9,200 (0) 20,3 + 1Styp + 35x59 $ 250 (0) 100s) + 60%, + 80x,, $ 650 and x20 forall 17. xy. xyand x, = chairtype 1, 2,3 and d to be produced, respectively Max Z= [2x + 20r, + 18x, + 40%, subject to i) Ax, + 9x, + Tay + 10, $6,000; (i) x, # 2x, +3x5 + 40x, < 4,000 sand 45%,%5,%420. Ans. x, = 4,000, x 18, Let, x, and x, = number of units of tomatoes, lettuce ard radishes to be produced, respectively. Max (profit) Z Selling price Ferilize: cost— Lebour cost 1 «2,000 — 0.50 » 100-20 5).x, + (0.75 x 3,000- 0.80 = 100-20 x 6) 4+ (2 * 1,000~ 0.50 = $0 ~20 5) xy =1,850.x, + 2,080.x, + 1875.1, subject to () x, +x, +x, $100; (ii) Sx, + 6% + Sr, $400; and x),%5,1520. i) x25, and x,x,20 xy + 167i + 055) £80; 200/3 and Max Z= Rs. 56,00031. a4 24, 25. M. Linear Progamming: Simplex Method 1x, and x = number of units of product P, and P,, respectively. Max Z= 10x, + St, subject to (i) 4x, +515 100; i) 5x) + 2x 580; and xy,4320. Ans. x, = 200/17, x,= 190/17 and Max Z = 2,900/17, 2 and x, = number of urits of product A and B, respectively Max Z = (60 ~30)x, + (40 ~10)x) subject to (i) 3x, +x, $30,000; (il) x, $6,000; (ii) x, € 12,000, and 14,x,20 Ans. :, = 6,000; x3 = 12,000 and Max Z = Rs. 5,40,000. Ay, Xpand. = units of raw material A, B and C, respectively Min Z = 2r, + 5x3 + 3x, subject to (i) 3x, +x,210: ii) Sry +x, +2x52 15 Gil) xy +2428 and x), xy.25 20. Ans. 5, = 1849, %, = 0,1, = Oand Min Z = 3688, 41 and x, = numberof pazes of magazine and television, respectively, Max Z = 60,000x, + 1,20,000c, subject to 1) 8,000, + 12,000x, $7.20,000; ii) x, 22; (il) x23; and xy.x,20. ‘i 47 Vand Max Z= 11 (ii) x, =2.3,= and Max Z= (iv) Max Z*=—x, +31, ~2x, where 2* 5.x) = and Z* nd Max. by3 (coefficient of x.) and Max Z=6 (i) Alle~s< Out Z2=—5/4 (<0) and actificial variable A, = 5/4 appears in the basis with postive value. Thus the given LP problem has no feasible solution (ii) 5, ~ 5/4, ~ 0,35 ~ Oand Min Z~ 75/8 ) -»= 0 and Max wi) 2,2 0,butZ=—4 (<0) and at so feasible solution. Sand Max Z= 40 appears in the basis with a positive value. Thus the given LP problem has fal variable Ay (ii) x, =3/5.x," 6/5 and Min Z= 12/5 appears in the basis with zero value. Thus an optimal solution to the given LP problem exists, (Gv) Intoduce anificel vaiable onl in the third constraint. x ~ 0, x5 = 0,25 = 0, x5 = O and Max Z~ 4. ©) and Min Z == 11 (i) x)= 0,25 = 10,x,= Oand Min Z= 20 (@) x, °0,22 Lay = O-and Max i) Degeneracy occurs atte initial stage. One ofthe variable eligible to leave the basis isartifcial variable, therefore, thereis no need of resolving degeneracy. Remove the artificial varabe from the basis. iy 0 @ Gi) Gi) (iv) x, ~ 10003, x, ~ 50/3, 5, = $003 and Max 2 1650 = 615 and Mex (Atihe current solution: x, = 5, =9and Max Z = 15, itmay be observed that ¢)~z) = 3/2 but all elements inthe second column are negative, Solution is unbounded, Gi) Atthe current solution: x,» 4,s, = 4and Max Z: negative, Solution is unbounded, (Gil) Atsecond best solution, ¢)~ 25 ~ 113/3 but all elements in the third column are negative. Solution is unbounded, x 5, = Sand Max Z = 0 iv), Optimal solution: x, =4, x,= 6 and Max Z~= 12. ince inthe initial simplex table all the elements are negative inthe second column, the feasible solution is unbounded but the optimal solution is bounded, and Max Z=4~ 2M: Infeasible solution. 24, itmay be observed that c, ~ =, 15 but all elements in the second column are 4 ai Co) (i iv) and Max Z=9—4M; infeasible solution. Infeasible solution,Copyrighted materialomen © Introduction © Review Questions ‘© Computational Procedure for Standard Form | * Self Practice Problems ‘+ Solved Examples © Hints and Anowers ‘® Altemative Approach to Revised Simplex Method INTRODUCTION “The revised simplex method is wed for reducing computational time of simplex method by arranging data ina smaller array. The revised simplex method, which proceeds through the same steps as simplex method but keeps all important data in a smaller array. ‘In matrix notations, the LP problem in its standard form by adding slack variables in the set of constraints ‘an be expressed as: Maximize Z = ex subject to the const Ax=b,and x20 If objective function is considered as one of the constraints equations, then the LP problem in its standard form I is written as: x20 «fee ‘Making Z= ex as the part of constraints implies that all the column vectors will have (m +1) components instead of m. Consequently, the notation a," is used due to (m + 1) components, Hence, we write aj) = F-¢,aJ. Similarly, new (m+ 1) component vector b” becomes b= (0, b]. (COMPUTATIONAL PROCEDURE FOR STANDARD FORM | For the initial basis matrix in revised simplex method, the columns a, of coefficients matrix A which form the initial identity matrix | are used. Since in the simplex method we always start with an initial basis (identity)‘matrix B of order m, therefore, forthe revised simplex method the inverse ofthis inital basis matrix can be written as: B, =(B4?, Bi? 8328071 ‘where the superscript | indicates that, is of order (m1) whose first column is the unit vector corresponding t0Z (= Bg). Thus we may also write a? oP. we o] 0 a) Further, if columns of matrix A form an initial basis matrix of order m which corresponds to the slack or surplus variables, then cy,=0 (i= 1,2, .. ., m). Thus Eqn. (1) reduces to the form. ‘This implies that the inverse of initial basis matrix B, ‘The initial basic solution is then given by 1 0 Ir97_fo ef JIL) This solution is obviously feasible because last m-components of it are non-negative, and the first component Zan have any sign. ‘After obtaining a basic feasible solution and the inverse (= 1p, ;) ofthe initial basis matrix, Bj!, we need to improve the solution by using the revised simplex method. For which we first calculate cj —z, for each column a) not in the basis B, by taking scslar product of the first row of B;' with each aj!’ as explained catlier. The vector a{” to enter the basis is determined by the criterion. 5 -4)> 0}, forall. 6 #~ Max (gj) 5 (af) ofthe inal bass matic inverse By’ =I, ; will never be removed from the basis at any iteration. The vector to be removed from the basis is determined by the criterion 2 = Min {= va od. for alli Tie Ya ‘m) are the components of vector yj.), where 0 wy [Fee ve Ne Since we start with an identity matixB,, the new inverse ofB, denoted by 8; !isobtained by multiplying the basis matrix inverse B;' at the previous iteration by an elementary matrix I”1. Here I~" is the inverse of an identity matrix I with rth column replaced y,. Since the variable Zis always desired in the basis, the first column | where yy (i= 1,2,"| Revieed Simplex Method Remarks |. _Ifthere isa tie in the selection of the key column, then choose the column from left to right (i.e. smallest index, ), 2. Atie in selecting the outgoing vector can be broken by any of the methods discussed earlier. Steps of the Algorithm Step 1 Express the given problem in the standard revised simplex form by considering the objective func- tion as one of the constraints, and adding the slack and surplus variables, ifneeded, to inequalities to convert them into equalities Step 2 Obiain initial basic feasible solution. That is, obtain an intial besis matrix B=, and then find By and B;' bto form the initial revised simplex table as shown in Table 5.1. ‘Table$.1 Initial Revised Simplex Table Variables Solation Basis Imerse. Bi inBais. Vas * : = B bead) | Bf a 8 Pu Z) 52) Sim) Z 0 1 0 0 0 | ak 4 ° Hi D : 0 vu & ° 0 1 0 Ms ie tn ° ° 0 ° 1 ia Step 3° Select a variable to enter into the basis (key column). That is, for each non-basic variable, caleulate ¢,~ by using the formula - 1 at) 4-3 = 9-0 ay where B;! af? represents the product of the first row of By" and successive columns of A not in By! '$, then the current basic solution is optimal. Otherwise go to Step 4. z; are positive, then variable to enter into the basis may be selected by using the 6-3) = Max {c; 6-370} Step 4 Select a variable to leave the basis (key row). That is, calculate y!! = Bj! af!) = al!); &=1) where af = ¢,, a]. Ifall yy $0, the optimal solution is unbounded. But if at least one yy > 0, then variable to be removed from the basis is determined by calculating the ratio x Aor = win 2, Ya a Je) Vik ‘That is, the vector Bi is selected to leave the basis and go to Step 5. Ifthe minimum ratio is not unique, i.e. the ratio is same for more than one row, then resulting basic feasible solution will be degenerate, To avoid cycling to occur, the usual method of resolving the degeneracy is applied. ‘Step 5 Update the current solution table by introducing a non-basic variable x, (= #4” ) into the basis and. removing basic vaiable x, (~ B'? from the basis, Repeat Steps 3 to $ until an optimal solution is obtained or there is an indication for an unbounded solution. a PreExampleS.1 Use the revised simplex method to solve the following LP problem. Maximize Z=2e) +2 subject to the constraints (3x 44x56, (il) 6+ 283 and 14,2320. [Dethi Unix, B Sc (Maths), 1993; Meerut, M Se (Maths), 1998] Solution Introducing slack variables s, and s, to consirainis to convert them to equations and considering the objective function as one of the constraints. Then the given LP problem can be rewritten as: (2-2 ==, Gi) 34+ 4x +5,=5, (ili) Ox, +xy+ y= 3 and xy X45 8)20. ‘The new system of constraints equations can be expressed in the matrix forms as follows: af) a af af af Pz 1] -2 -1 0 0)]x| fo) | =/6 a} 3. * 1 -e]fz]_fo ee eee 2, The basis matrix B, of order (2+ 1)=3 can be expr x L o 0 (68, 8. 1=[0 [10 o|o 1 2,0] 10 Jewe-[h S] 0 aa ‘Key Column Variables notin the Basis ‘mnase Values g ” 0 y? a? a B blexy’) © ) 4) (en) Zz 0 0 ana of =I * 6 ° 1 ° 3 4 b 3 0 0 1 6 1 Iteration 1a) To select the key column, ic. a non-basic variable to enter into the basis, we compute 4-24 ~ Max {(c; —=))> 057 =1,2} ~Max {-Crirstrow of By") (Columas a rot in basis, Bi )} ee =Max {-(1,0,0)] 3. 4] t= Max {-2,~1)} =2 (corresponds 10 ¢— 2)) 6 1 ‘Thus, variable x, (= a”) is selected to enter into the basis, for k= 1.Revised Simplex Method [Eg (0) To select the key row, ic. abasic variable to leave the basis given the entering non-basic variable x), we compute y(? for k= 1, as follows: 2 0 yf) =Byla(?=a =| 3} fork=1; and xf) ~B;' b~b=]6 ‘ 3 ‘The minimum ratio to seleet the basic variable to leeve the basis is computed as follows pS ain {0 Ju >} inf two. a : Jun Yar} x ~Min{S, 3} -2 (somspont 0 2 6) 6 Jn ‘Thus, the variable (= ()) for r = 2 is selected to leave the basis. Table 5.2 is again reproduced with the new entries in column y/ and rhinimum ratioas shown in Table 5.3. Table 5.3 Variales Solution Basis Inverse, y? inBaiis Values a a 0, 0 " (k=1) Min Ratio B (=x) a m 2) en) or) xP z ° 1 ° © =2 — * 6 ° 1 © 3 8 % 3 0 ° 1 6 36> 1 (© The intial basic feasible solution shown in Table 53 i now updated by replacing variable withthe variable x; inthe basis. For this we apply the following row operations in the same way asin the simplex method. o oD wm Ro 0 0 0 -2 R> 6 1 a 3 Ro 3 0 1 © Rs (new) Rj (ld) +6 they element), Ry (new) —¥R, (ld) +22; (new) Ry (new) — Rp (old) - 32; (new) Remember While determining the entries of thenew table foran impoved solution, it should be remembered «hat column p> will never enange, Taus, entries in x", y"” , 8," columns will be changed due to the above mentioned row operations. The improved solution is: ‘itowe in Table 5.4.TableS.4 Improved Solution Basis Ivers. Bi! MinRatlo Variables not nthe Basis at o » | Oo oO » Bo Br yf) | xp al a (2) sy) (x) ©) Z 7 7 ° | | 2 @ 2 » #2 |o 1 -2 | @] 2 o 4 2 yom o ° ws | 6 =z! 1 t The column vectors not in the basis and new basis matrix as shown in Table 5.4 are: =I ° 10 wB 4] and af =|0)andB;'=|0 1 172 1 1 oo 6 Iteration 2 Repeat Steps 3 to 5 to get the new improved solution. (a) To select the column vector comesponding to a non-basic variable to enter into the basis in Table 5.4, we compute 62, = Max ((¢)-2)) > 0;/=2,4) = Max {- (First row of B;') (Columns a)” not in basis)} -1 0 =Max {(-1,0,13)| 4 0] > =Max { © (corresponds to e222) 1d Thus, the variable x, (vector a!) is selected to enter into the basis, for k= 2. (b) In order to find the variable (vector B\")) corresponding to the basic variable to leave the basis, we compute y(” for k= 2 as follows. 10 1B]f-1] [-24) y® =B;'a=|0 1-12]! 4]=| 72 | The values y$" are shown in Table 54. ‘The minimum ratio for a predetermined value of k (= 2) is given by se els 50} ain (22.2) 92 12) _[% 4|_9 x aa 2 te ot 2 22 io (22.21 6 t + [somsipns *x] is selected to leave the basis as shown in Table 5.4, (©) The solution shown in Table 5.4 is now updated by replacing variable s, with the variable. into the basis For this applying the following row operations in the same way as in Iteration I: ‘Thus, the variable s (vector B")) for: Rovied Sempox Mood ERY o o 7 om 2 By By if no ° B 28 Ro 7 1 =n n Ro 2 0 16 Ry (new) —> Ry (old) * (2/7) (Key element); Ry (new) —> Ry (old) + (2/3) Ry (new) Ry (new) —> Ry (old) ~ (1/6) Ry (new) ‘The improved solution is shown in Table 5.5 ‘Table 5.5, Variables Seton Basis ere, Varies ot inthe Rass abe Vata: i) ay a ay “ B bx) fy 0 a au (2) ex) x) ex) es) z A 7 wai 521 ° ° si 97 ° 37 7 ° 7 4 a 0 -12l 421 1 ° The variables (column vectors) not in the basis as shown in Table 5.5 are: 0 0 1 421 5/21 sy= all = [1], 5,= al =/0] and By! =l0 wm -n 0 1 0-221 421 Iteration 3 Repeat Steps 3 0 5 to get thenew improved solution. (@) Toselect the variable (vector) corresponding to non-basie variables to enter into the basis in Table 5.5, we compute, eg—24 = Max {(¢) -2))> 4} =Max {— (First row of B, ') (Colurnns af" notin the basis)} 0 0 = Max }~(1, 4/24, 921)}1 0|} =Max (-(421, 5/21)} o1 Since all ¢,~zj< 0 in Table 5.5, the current solution is optimal, Thus, the optimal solution is: ,~ 9/7 and Max Z-= 1317. Remark Once the revised simplex algorithm for solving an LP problem is fully understood there is no need. of giving details about the steps of the algorithm. This is illustrated in the following two examples. Example 5.2 Use the revised simplex method o solve the following LP problem Maximize Z=3x, + Sx, subject to the constraints @s4, fi) ns6, Gil) 3x) + 217518 and 1,220. Solution Expressing the given LP problem inthe standard form I of the revised simplex method as follows: @ 2-28, 5x fi) x+4=4 aa +52 =6. (wy) 3x +2 +s; and Hp 557520.Now representing the new system of constraint equations in the matrix form as follows: af) 9 af” a a pz x] fo x] _|4 | [6 | Ls 3 0) 9, M29, 4 2 2g where, af) =p0), a=, a =p? and al = The basis matrix B, of order (3 + 1) = 4 can be expressed as: gM gi go 0 ~ [Bt 8\” By 8S]- cgB ee 0 0 Then |-s2- o 1 o|-[pl?. pS, B?] a-10, 01 o1 ‘The initial basic feasible solution is shown in Table 5.6. ‘TableS.6 Initial Solution Basis Inverse, Variables not inthe Basis oy or a o 0 o Bo 8 8; Ye 4 % 2) es) ex) EX) Zz 0 f 0 0 [ann 3 =5 4 4 0 1 0 1 0 2 6 ° ° ° © 1 53 18 0 ° 1 3 2 Iteration 1 Toselect a non-basic variable (i.e. key column) to enter into the basis, we compute oy-2p = Max (G2) 70; j= 1,2) =Max {- (First row of By!) (Columns 2}! notin the basis )} -3 -8 =Max}—(1, 0,0, 9)] 4 | f=Max {-(-3,~5)} =5 (corresponds to eo 20) 302 = 2. To select key row, ie. basic variable to leave the basis, we ‘Thus, vector a$) (=.x3) enters the basis, for ‘compute -5 °l tork=2 and x) =By'b=b= ip Sp Sp e=e= o 4 6 8.‘The minimum ratio to select the basic variable to leave the basis is computedas follows: Yat = Min { Thus, the variable s» (vector Bl) for = x a “Br = Min}=B, y,, >0, Ya jm =6 (corresponds to is selected to leave the basis. x eo Xas M2 In In ) Table 5.6 is again reproduced with the new entries in column y‘"” and minimum ratio as shown in Table 5.7 ‘Table5.7 Variables Solution Basis Inverse, By! inBais Values Bex) | BP 8” 8, 0 | Min Ratio © ex a=) | iy? z 0 1 ° 3 ai + ° 1 @ % 6 ° o @ | si-65 43 18 ° ° 2 | 1wa=9 t ‘The initial basic feasible solution shown in Table 5.7 is now updated by introducing variable x2 into the basis and removing sp from the basis. o a Bp o o yy R> 0 0 Xp B, Ro 4 1 R> 6 0 R> 18 0 For this applying the following row operations. Rs (new) —> Rs (old) + I (Keyelement); Rj (new) Ry old)— R, (new) The improved solution is shown in Table 5.8. 0 0 1 0 5 0 @ 2 Ry (new) > R, (old) + 5R3 (new); ‘Table5.8 Variables Solution Basis Inverse, B-* y> Variables notin the Basis oP ewe ” ©Z) ©s) (x) (82) sy ‘ 0 0 -2 1 | @ 3 °‘The variables not in the basis and new basis matrix as shown in Table 5.8 are given below. ° 10 so o o 1 00 1 oo 10 0 00 -2 1 Iteration 2 Again select the column vector corresponding to a non-basic variable to enter into the basis, we compute 64-24 = Max {(6,-2)>0;j=1,4} =Max {— (First row of B) (Columns a” notin the basis )} -3 0 0 = Max}—Cl, 0.5.99] 4 | f =Max {-(-3,5)} =3 (comesponding toc, —2;) 30 Thus, the variable x, (vector a{!) ) enter the basis for k= 1. To select the basic variable to leave the basis, we compute 10 5 O7[-3) f-3 o1 00 1 =plall= =| hae yP=Bi a =o 9 1 off 0/7 of Fm? oo -2 la) ls ‘The values of y(? are shown in Teble 5.8. The variable to be removed from the basis is computed for predetermined value of & (~ 1) as follows: Sar In {4.6 8) 6 ceponding to XB tieess}-$! spending oy) ‘Thus, the variable s, (vector 8) ) is selected to leave the basis as shown in Table 5.8. ‘The solution shown in Table 5.8 is now updated by introducing variable x, into the basis and removing variables s3 from the basis. o a a o Xp Bi By yy Ro 30 0 s 3 ho 4 1 ° ° Ro 6 ° 1 1 A> ° © For this apply the following row operations. Rg (new) >, (old) +3 (Key element); Ry (new) Ry (old) — Rg (new); Ry (new) 8, (old) +32, (new) ‘The improved solution is shown in Table 5.9.Revised Silex Mettod BEEY ‘Table 5.9 Variables ‘Solution Basis Inverse, Variables not in the Basis InBosi Yas ah o Bo byex") " 2 3 ¢ Ess) Zz 36 7 ° 3 7 0 0 a + ° 1 TT) ° 0 2 6 0 0 1 0 1 0 Xi 2 0 1. -» 0 1 ‘The variables not in the basis and the new basis matrix as shown in Table $.9 are: a 1 10 3 1 | 8) a _f-1] gp fo 1 28-18 4 ays 0 0 0 1 0 -U3. V3. Oo 0 -23 v3. eration 3 The procedure illustrated in iterations 1 and 2 is repeated to updated the solution as shown in Teble 5.9. Firstto select the variable to enter into the basis, we compute a.-4 = Max {(c, ~ =,)>03/=3,4} = Max {- (First row of By") (Columns a( not in basis)} 00 = Max -03,0]) 4 = Max {-(3, 1}. oo. Since all cj ~ z; < 0, the current solution shown in Table 5.9 is optimal. Thus, the optimal solution is: y= 2,2 = 6, and Max Z = 36, Example $.3 _Usethe revised simplex method to solve the following LP problems. Maximize 11x) 43%, subject to the constraints (9) 3x) +26) +25 $3, (ii) Wey Hy + 2m 52 and xpnHZ0. Solution Expressing the given LP problem in the standard form I ofthe revised simplex method as follows: () Za, —x2~ 343-0, (i) 3p +e HHT 5)=3, Gi) 2K Hy H2xy r= 2 and Xp 9X 7, Xyp 5,57 20. ‘The initial basis matrix B, isgiven by 1 oo 1B,” 8,” ,p0'1=|0 [10 ojo i The initial basic feasible solution: s, = 3,s, = 2andZ~0 is shown in Table 5.10.‘Table 5.10 Initial Solution Variables Solution Basis Inverse, B;' inBasis—Yaies rr | Bb tex) : é 2) eH) ) z ° 1 ° eZ =I 4 =3 n 3 0 7 ° 0 3 2 7 h o 0 1 1 2 1 2 lteration 1 To select a non-basic variable out of, x;, x3 and z3 to enter into the basis, we compute ena Max ((9-4)> 05) 3} = Max {- (First row of By! ) (Columns a not in the basis)} =Max}—(1,0,0)] 3 2 1} =Max {1, 1,3) =3 (corresponds to; -£3) Tah ati itt’ ste cari tho ail 2 Now wade aie variable to leave the basis, we compute © 1 0 yf-3] [-3 ¥3 a=l0 1 olf t=] th: k=3 o o ifL2} 2 1 0 O70] fo and x9=B)b=|0 1 0f/3]=|3 oo aja} [2 The values of y$) and x§) are shown in Table 5.11. Table 5.1 Variables Solution Basis Inverse, By! ‘Mn Raiio in Basis Values 7 7 : ns a” Bo p” BY ” 7 fs) a) 5) es (e=3) xp/y.. Zz 0 1 0 0 = 3 0 1 0 1 ° ° 1 2 , ‘The variable to be removed from the basis is computed for the predetermined value of k=3 as follows: fee Win {84.74 >o} = tin Tr Ya = Min f. 3} = 1 (corresponds to 282. 12 Jn Thats, -=2 and therefore basic variable sis tobe removed from the basis as shown in Table 5.11.Revised Simplex Method [EEE Apply the following elementary row operations to update solution shown in Teble 5.11 Ry (new) —9R; (old) +2 (Key element); Rj (new) —» R; (old) +3R3 (new) Ry (new) —¥ Rp (Old) ~ Ry (ew); ‘The improved solution is shown in Table 5.12. “Table 812 Variables Solution Basis Inverse Variable notin the Basis Base, rae a) a yo a aw a Box?) | Bo i 5 : : G@2) Gs) Gx) | &=3) (x) Gx Gx) Z 3 1 ° 2 =a =I ° a 2 ° 1 =12 3 2 ° 5 1 ° ° 12 2 1 t =r -1 0 = 2| and al? =| 0 2. 1 1 Meration 2 Again repeat Steps 3 to 5 to get the new improved solution, if possible. 4 = Max {-(First row of BY”) (Columns {"” notin the basis )} -1 -1 0 = Max }—(1, 0, 32)) 3 2 0|}=Max {-2,-1/2,-322) 2 11 Since alc) ~z; <0, the current basic feasible solution is optimal as shown in Table 5.12. Thus, the required ‘optimal solutionts: x, = 0, x)= 0,x5= land Max Z=3. Example 54 Use the revised simplex methodio solve the following LP problem Maximize 26x, —2xy+ 3x5 subject to the constraints @ 2x) +x +2ey82, (i) ey t4ysd and X1,243320 [Dethi Univ., B Sc (Maths), 1993; Madras Univ, BE, 1999] Solution Expressing the LP problem in the standard form I by addingnon-negative black variables as follows: () 2-6) +2) 3%; (il) xj-xy +2 +92, (ill) x, HAY +s, and ry tas.s1 0. ‘The initial basis matrix B, is given by 1 0 0 » gt? gi 1B’ By .By’)=|0 | FO) =1, o]0 L ‘The initial basic feasible solution: s, =2,s=Oand Z=Ois shown in Table 5.13. ontseay.13‘Table $.13 Initial Solution Variables Solution Basis Inverse, By! Variables not in the Basis intoss Yalus, amy oo Bo obex!) ' 2 eH es ce 6a) Zz 0 1 0 0 [a-4 -6 2 3 % 2 0 1 0 2 2 a 2 & 4 o 0 1 1 1 0 4 Iteration 1 To select the key column, i.e. a non-basic varieble out of, to enter into the basis, we compute 4-24 = Max {( 6)-2))>0; 1, 2,3} = Max {- (Firstrow of B, ' )(Columns a{? not in the besis)) -6 2 -3 =Max}-(1, 0,0) 2-1 2/}=Max{6, -2, 3} =6(comesponds to¢,—z)) too 4 ‘Thus introducing vector a, (=.x,) into the basis for k= compute . To select a basic variable to leave the basis, we 1 0 Olf-6 6 eget al ol ax yl) =By o 1 off 2[=/ 2k kaa oot 1 L 1 0 ojfo 0 ait x2=Bo=]o 1 olf2|=|2 o 0 14 4 ‘The vauesof y is shown in Taste 3.14 Table5.14 Variables Solution Basis bvere, Min Ratio inBas = Valus 5 ; 5 , ‘, a a a ”” 7 oem) | ez) es) ) den) xy? z ° 1 ° @ = = i 2 ° 1 @ =I a 4 ° o 1 ans ; ‘The minimum ratio column of Table 5.14 shows that the basic variable s, (= B{”) should leave the basis. Apply the following elementary row operations to update the solution as shown in Table 5.15. Ry (new) —> Ry (Old) +2 (Key elementy, Ry (new) — Ry (old) + GR; (new); R; (new) = R; (old) — R; (new)‘Revised Simplex Method EEA Table 5.15 Variables Solution Basis Inverse, By! Variables not in the Basis ita Te : wa Sd ©Z) Es) (k=2) (en) Gn) Zz 3 1 3 0 ot 0 2 -3 ti 3 o -2 1 | @> 0 6 4 Iteration 2 Repeating Steps 3 to 5 to get the improved solution, if possible. Toselect a non-basic variable to enter into the basis, we compute 4 ~24 = Max {( 6-2) > 0:5 =2, 3,5} = Max {- (First row of B;' ) (Columns a? not in the basis)} 0 2 -3 Max -(I, 3,0)] 1-1 2|} = Max {- (3, -1, 3} =1 (corresponds to ¢) 2) 0 0 4 Thus introducing vector a, (=x) into the basis for k= 2. To select a basic variable to leave the basis, we compute 3 of2] ft vw olf-1]=|-12 -2 1 0 o} | 2 oyfo) fp -r 1 of]2 [| o sila} | vw o ‘The values of y$? are shown in Table 5.15. Applying the following elementary row operations to update the solution as shown in Table 5.16, O gt Y. “By a) = 10 = BY y= Bie aloe = Ry (new) — R; (old) * 2 (hey element); Ry (new) > R, (old) —R, (new) Ry (aew) > Ry (old) + (1/2) R (new). ‘Table 5.16 Variables Solution Basis Inverse, B;' Variables not in the Basis ea a aor yest | Oe Lk 2) Ca) Cm) ea) Gx) a3) oe fo ot eration 3 To select a non-basic variable to enter into the basis in Table 5.16, we compute ¢y—74 = Max ft Girst row of BY?) (Columns a{” notin the basis)} oo 3 =Max}—(1, 2, 3)}1 0 2|}=Max {-(2,3,13)) o1 4 2,-3,- 13)err Operations Research: Problems and Solutions, Since all cj ~z;<0, the current basic feasible solution shown in Table 5.16 is optimal. Thus, the required ‘optimal solution is: x, =4,.x)=6 and Max Z= 12. While solving an LP problem using usual simplex method, at each iteration, the inverse ofthe basis matrix is obtained. However, in revised simplex method, the inverse of the basis matrix is obtained from the original constraint equations provided current basic variables are known in any simplex table. Example SS _Userevised simplex method to solve the following LP problem Maximize 22x, +21) +3x5-my subject to the constraints, (x, 42x,+3y=15, (ii) 2xytxyt5xy=20, Gl) x +2ey ty +4,=10 and yp TpXy X42. Solution Expressing the LP problem in its standard form as follows: Maximize Z=x, + 2c, +3x,—x,-MA,—MA, subject to the constraints Ox, +2ey+3x+4,=15, (ii) 2xjtay+Sxy44,=20, ill) x, +2ey+xy 404-10 and Aya tardysdg20 Leta, a, ay, ,,a, and a, be the column vectors corresponding to the variables xx, that xg Tespectively, 0 15 =|1 | anav =| 20 0 10 “Thecolumn vectors ay, asandag form the ntal basis matrix Bof order 3, ie 0 1! B oo 4 3 o 0. ‘The inital basi feasblesolution: 4, = 15,4,=20,andx,= 10s shown in Table $.17. ‘TableS.17 InlualSotution Variables Solution Basis Inverse, B* Key ‘Minimum te Baek Yaar ao B b(=x_) xy A) GA) ‘A 8 0 1 o | ® | ws-35 Mw » oO 1 1 | 20-20 4 zi 10 1 oo 3 103 =3.6 valuesas follows: o 1 oO CM-M-D}0 9 1]=C1,-M,—y 10 0. i =(1,-M,—)) jo 1Revised Simplex Method | 189 | 2 =2-C1,-M.-M)] 1] =443a0 2 x 3-(-1,-M,-M)| 5|=6+6M 1 Since c;~ 2, value is largest positive, variable xis selected to enter into the basis. To select ¢ variable to leave the basis, computing key column values and minimum ratio as shown in Table 5.17, © 1 ovf3y fs Keycolumn,a(?=Blaj=/0 0 1)/8)=|1 1 0 oli} 3 ‘The minimum ratio column shows that the basic variable 4, should be replaced with non-basic variablex, in the basis. For this, applying following row operations: Ry (new) Ry(old) +5 (key element) —_—_R (new) FR, (old)— Ry (new) Ry (new) > Ry(old) -3R, (new) The improved solution is shown in Table 5.18. =a-% ena “Ys ‘TableS.18 Improved Solution & Hariabies Solution Basis Imerse, 8 Key Minirum inBass values. [aS Clem Ratio B dex) 3 x 3 3S) = 15 -M Ay 7 17915) = 859 1 % 3 34/8) = 1877 > Iteration 2 y=cBt= ey 8y e Yay isthe largest positive value, variablex, is selectedto enter into the basis. The key column then 0 us O72] [us ta=|0 -us 1{]1| =|95 1-35 off2 Pivot column, a!‘These values along with minimum ratios are shown in Table 5.18, Hence basic variable x, is selected to leave the basis, For this, applying following row operations. R, (new) -> R; (old) (5/7) (Rey element); R, (new) +B (old) -(9/5) R, (new) Rj (new) — R, (old)—(V/5) R, (new) ‘The new solution is shown in Teble 5.19. ‘Table5.19 Improved Solution Cy Variables Solution inBasie Values B b(t) 3 x 187 -M A san 2 4 1 eration 3 y=egB"=(3, -M, 2 4724 = ea YMG Since c, ~z, is the largest positive value, variable x, is selected to enter into the basis. The pivot coluran then becomes “V7 27 Oo) fo" a =B'a=|-97 47 allo|=|1 s7 -37 ollt} |o. ‘These values along with minimum ratios are shown in Table 5.19. The basic variable A, is selected to leave the basis, ‘ableS.20 Improved Solution co Variables Solution Basis verse, Pivot | Moin ae on a; a] coum | io 2 tem | ey 6 68 3 . wf moa o | @ | 4 S an | 97 47 1 «7 oan 2 n ist | -37 sn o | -w es -1 mo nteration 4 y=epB'=-@,-,|-97 47 1 -[$ 4, >] m7 -37 0. 16 TOTRevised Simplex Method ERY Thec,—z5 and c,—7, values are not considered because A, and A, are artificial variables. Since ¢, —z, is positive, variable x; is eligible to enter into the basis and key column as shown in Table 5.20 becomes. -vI 27 oft) [37 a, =B4a,=|-97 47 1/]2/=| 67 sa 37 olla] [17 ‘The pivot column along with minimum ratios are shown in Table 5.20. To enter variable x into the basis, applying following row operatior Ry(new)— R, (old) *7/3 (key element), R, (new) > R, (old) + (1/7) R, (new). ‘The new solution is shown in Table 5.21. R, (new) > Ryfold)—(6'7) R, (new) ‘TableS21 Optimum Solution cy Variables Solation Basis Inverse, B inBasis values 7 a Bo bexp) eye 1 a 6 23 0 1 *y sor o 1 2 » 207 -13 0 “13 23 0 eration 5 y 14-1,2] -1 0 1]=@, 0-1 2B -1B 0 3 e5-¥65=3-@ 0, -D/5 1 Since c,—z, =—2 (negative), optimum solutions arrived at as shown in Table 5.21, i, 167 and Max Z= 6 +2(21/7) + 3(0)~(56/7)= 287 REVIEW QUESTIONS 21 =6,2)= 217,25, 1. Formulate a near programming problem intheformof the revised | 3. Give a brief outline forthe standard form I of the revised simples simplex method, method. 2. Develop an algorithm forsolvinga linear programming problemby | 4. Whatis the difference between simplex method andrevised simples revised simplex method. method? When and where the two should be applied? SELF PRACTICE PROBLEMS Use the revised simplex methodto lve he fllowing LP problems: | 3. Max 2= ny +2e5 + 5x 1. Max Z=xy +283 subjest to (i) x, #23445 < 430; Gi) 3x +2x5 < 460 subjectto (x; +32 $ 3, Gi) xy +28 $ 5, (il) 3xy+xy <6 Git) 3, +4 < 420 and xyz 2 0. and Maem 2 0. (Gaudhati Unis, MCA, 1992] | 4 MO 2 or TF ie a subject to (@) 2x +3m~ 1 +4xy S 2 Max Zon | (i) ~25, 4 2ey— Sarg 5 38 subjectto (i) 3x) +3%, <6, Gi) +4 <4 apt ayia loo and pp 20. and Xt ty ty 2 0.5, MinZ=—4x; +, +2x 8. Max Z=2x, +x, +25, subject to (i) 2x, 3x) +2%<12; (ji) —5x, +21 +324 subject 10 (i) dr, +3x) + 8x42 12; (ii) dx, +x, + Id <8 i) 3x,—3x =-1 (ii) 4x, —2y + 385 $8 and Sythe 0. and Hips 20. [Panjab Unix, BE (1), 2004) | 9. Max Z=~25, 4x, x, 6 Min Z=2r, +45 sibjectio 5, +25-% $5 GLI —H +2502 subject © (i) 3x ty Gi) 4x, #35326 Gi) - 42m, 42x, 21 Gi) x, +283 and Roxy 20. and int 20. 10. Max 230. + 23x, +291, 1. Min Z=x, +x, subject 10) 6r, + 5x; +3x,$26; (ii) 4x, + 2x, +5, $7 subject to (i) x, 42x) 2% Gil) ax #2 6 and XX 20. and Spy 20, 2. x= 27, a= 3S and Max Z= 11/7 x1 =0, x= 100, x)= 230 and Mex Z= 1,350. 4. x)= 71/4, xy=1, 15 = 2972, 14=4 and Max Z=445/4, 6 8 24, = 59/17, 2 = 36017, x5 = 97/17 and Min Z=— 6/17 5, x= 0S and Min Z= 12/5 07, y= 22/7 and Min. ay = 32, y= 2,45 = 0 and Max Z=5 4,50, 4.50, 4, = 1 and Max 2 10. x, =0, x,=7/2,4,=0, and Max 2= 161/2 07 Copyrighted materialDual Simplex Method | SS i ! ! ou * Introduction ‘© Review Questions * Dual-Simplex Algorithm ‘© Self Practice Problems * Solved Examples © Hints and Answers INTRODUCTION ‘The simplex method is an algorithm that always deals with a basic feasible solution and the algorithm is, terminated as soon as an optimal solution is achieved. That is, the procedure should be stopped when all ¢,~ =) 0 for maximization problem and c) ~ > 0 for minimization problem. However, ifone or more solution ‘values (i.e. are negative and optimality condition c;-z both for maximization and minimization is satisfied, then current optimal solution may not be feasible (because ¢)—z)= c)-¢pB™! a, is completely independent of the vector b). In such cases, it is possible to find a starting basic, but not feasible solution that is dual feasible, ive, allc)—2)$ 0 fora miximization problem. In all such cases a variant of the simplex method called the dual- ‘simplex method would be used. In the dual simplex method we always attempt to retain optimality while ‘ringing the primal back to feasibility (i.e.xp,2 0 forall’). _DUAL-SIMPLEX ALGORITHM Step 4 Convert the given LP problem into the standard form by adding slack, surplus andartifiial varinbles and obtain an initial basic feasible solution. Display this solution in the initial dual-simplex table. Step 2 If all solution values are positive (ie. x) 2 0 for all /), then there is no need of applying a dual- simplex method because improved solution can be obtained by simplex method itself. Otherwise goto Step 3 Step 3 If there exists a row, say r, for which solution value is negative (i.c.xg, <0) and all elements in row rand column j are positive (i.e. y,, 2 0 for all j), then current solution is infeasible; hence go to Step 4. Step 4 (i) Select a basic variable associated with the row (called key row) having the largest negative solution value, ie, = Min {5X51 0}i _ Determine the minimum ratios only for those columns having a negative element in row r. Thenselect a variable for entering into the basis associated with the columa for which 6= _ sin Jk | ay The element (i. yx ) at the intersection of key row and key column is called key element. The improved solution can then be obtained by making y,_ as 1 and. other element of the key column zero. Here, it may be noted that key element is always positive. ‘Step 5 Repeat Steps 2 to-4 until cither an optimal solution is reached or there exists no feasible solution. ‘A flow chart of solution procedure of dual-simplex method is shown in Fig. 6.1. ‘Transform the LP model into standard form Set up initial simplex table [compte ante = va] g-3s0 Yes all vo. # Select key row with lagest negative solution values Ge. xpj<0 forall) Select the key column with Min ((¢)~2)y: yy <0} ratio, where yy are the negative elements inthe key row ‘+ Determine an element (called key clement atthe intersection of key row and key column. '* Develop the next improved solution with the usual method. Fig.6.1 Flow Chartof Dual-Simplex Method Example 6.1 Use the dual simplex methodto solve the LP problem: MaximizeZ=-3x,-2%, subject to the constraints @ x tx21, @ x+ys% GD x +2m210, Gi) yS3 and xyH20,‘Dual Simplex Method [EEE Solution In order to apply the dual-simplex method, make all the constraints of the type < by multiplying by ~ 1 and then add slack variables in the constraints of the given LP problem. Thus the problem becomes 3x, ~2eq +05, +05) + Os, subject to the constraints @ —x- Hts) Gill) —x, -28 +5; and X19 Sp 830555420. An initial basic solution (infeasible) is obtained by setting ry =x =0 as shown in Table 6.1. This givesthe solution values as: s; = ~1, s;= 7,83 =~ 10, s4~3 and Max Z =0. Maximize Gi) txts (iv) +54 Since all ¢)~z 5 Oandall solution values arenot non-negative, an optimal but infeasible solution has been obtained. Now in order to obtain a feasible solution, we select a basic variable to leave the basis and a non- basic variable to enter into the basis as follows: Key row (variable to leave the basis) = Min {xp;: ¥p)<0} ‘Thetis, the basic variable s; leaves the basis. Key column (variable to enter the basis) = Min { ‘That is, variable x, enters the basis ‘Table61 Initial Solution 47 “3 > 0 ° ° ° ca larigbles Solution inBasis Values a B b(= x5) 0 7 1 1 1 ¢ 0 0 ° 2 7 1 1 0 1 ° ° ° 5 =10 - @ o « 1 ° o i 3 o 7 0 ° ° 1 Z=0 ° oo o 0 ° en) ‘ ° ° Mteration 1 ‘The new solution is obtained after introducing x, into the basis and dropping s; from the basis as shown in Table 62. Table 62 97 4 0 ° ° ° ce Wertbies ————Sotaton in Basis Values x on © @ om w B ) 0 1 4 a ° -m 0 ° 3 2 2m 0 1 2 0 - 4 s m 1 0 o -2 0 ° Ss 2 Gado 6 1 o 2 0 0 10 4 20 0 0 aBETH Operations Research: Problems and Solutions : ee ‘Table 6.2 shows that the solution is still infeasible (because s, eration 2. 2) but optimal, Thus we proceed to Heration 2 Since s4 (= xp, )=—2is the only row with negative solution value, variable sg leaves the basis. ‘Also in key row ya; is the only element having negative value, thus ,-colurmn is the key column and variable x enters the basis. The new solution is shown in Table 6.3. Table 6.3 > -3 -2 o 0 6 0 oe Variables ‘Solution in Basis Values * % 4 % % 4 B (2x5) 0 4 6 0 0 1 0 -1 + ° 8 D 0 0 ° 1 1 1 -2 * 3 ° 1 ° ° ° 1 -3 Xi 4 1 0 0 0 “1-2 Za-18 3 2 0 0 3 4 6-4 0 0 0 0 -3 0-4 Asshown in Table 63. all ¢)-z)S0.as well as all solution values are also positive. therefore, current solution is the optimal solution, Hence, the optimal basic feasible solution to the given LP problem is: x; = 4,1) =3 and Max Z=~18. Example6.2 Use the dual simplex method to solve the LP problem Maximize Z=-2x,—x5 subject to the constraints Ox tyes, Gi) xy-2xy H4gns and st. 20. [Roorkee Univ, BE, 1990; Banasthali Univ., M Sc (Maths), 1993] Solution Since the objective function of LP problem is of maximization, all the constraints should be of < type. Thus convert the constraints to the < type by multiplying both sides by - I and rewriting the LP problem. Maximize Z=—2x, +0r.—x subject to the constraints () =x) =m tay $-5, (ii) -x + 2x) -44yS-8 and 14n%520. Convert this problem into the standard form by adding slack variables s, and sin the constraints. The given problem can then be written as: 2x, +Ox,— 45+ 05, +063 Maximize. subject to the constraints @ -y-m ty +5) and, Ay XyXy S572 0. An initial solution to this LP problem is showa in Table 6.4, x 42x, —4x5+ 5Table6.4 Initial Solution 47 30 to 0 ce Variables Soliton inBasis Values < *» & 3 o a 3 7 1 1 ° ° es -8 -1 2 @ 0 1 z=0 ° ° 0 ° 0 3 oH ° 0 In Table 6.4, the initial solution is infeasible because s,; =— S and s, = — 8, but it is optimal, since all G-#)$0. Thus, we need to apply the dual simplex method to get both feasible as well as an optimal solution. eration 1 n order to obiain a feasible solution we select a basic variable to leave the basis and a non-basis. variable to enter into the basis as Follows: Key row (variable to leave the basis) = Min (xp): xp, <0} = Min {— 5,~ 8} =—8 (corresponds to.xp9) ‘That is, basic variable s, in row 2 leaves the basis ac pea a = inf Ly, <}- van {=2,=1} = (1/4) (corresponds to Column 3) ‘Thats, non-basis variable-x enters the basis. The new solution is obtained after entering r3 into the basis and removing 5 from the basis is shown in Table 6.5. “Table6s —~ 2 9 -- 0 o a Vaiabis Soon J inBasis Vals ee ee ee z tm) ° 7 7 “4S 71a “1 5 2 “GD 1 ou 5 a oa a 4-1 o 0-14 t Table 6.5 shows that the solution is still infeasible (because s second iteration. —1) but optimal. So we proceed to the eration 2. Since variable s, = ~ 7s the only variable having negative value, select variable s, to leave the basis. The variable to enter into the basis is determined as: Key column (variable enter the basis) = Min { 4 4 = Min {= 12 | = (corresponds to Column 2) A | =1 (conespo umn 2) ‘Thats, the variable x, enters into the basis. The new solutjon is shown in Table 6.6.ce Variables in Basis x 2 % 5 2 B " sn 1 ° = -12 “1 5 an 0 1 <1 =m =” 0 1 1 12 - 0 0 “1-2 In Table 6.6, all ~ 2 <0 and all solution values are also positive, therefore the current solution x = 0, 2) = 14,x;=9 and Max Z= 9 is the optimal basic feasible solution, Example 6.3 Use dual simplex method to solve the LP problem Minimize Z= 3x) +x subject to the constraints x tm21, i) ey + 3x22 and 7320. UAS(Marhs), 1990} Solution Conver the given minimization problem into maximization problem, and convert the constraints of 2 type into < type by multiplying them by ~ 1. Thus, the given LP problem becomes Maximize Z” =-3x)-x) ‘subject to the constraints () =x -5- and Ay 20 (ii) -2x) -3x) $2 follows: Maximize Z* =~ 32, —1 + Os, + Os subject to the constraints @ -x-2+5)= and HpkaspH20. An initial solution ofthis LP problem is shown in Table 6.7. ‘Table6.7 Initial Solution (i) -2x,-3xy +5) a “2st ° ° ce Variables inBasis 4 n * bs B 0 n 7 7 ° ° 8 2 @_o 1 3 ° ° 0 ° 6-4 ee o ° In Table 6.7, the initial solution is infeasible because s, =— 1, and s)=— 2 but it is optimal, since all 640. Thus, applying dual simplex method to get both feasible as well as optimal solution.Hteration 1 In order to obtaina feasible solution we select a basic variable to leave the basis and anon-basic variable to enter into the basis as follows: Key row (variable to leave the basis) = Min {rpy:q;<0} = Min {-1,-2} = (ato nee (-3-} ‘The new solution obtained after entering non-basic variable, basic variable, x» into the basis and removing basic variable, 32 fiom the basisis shown in Table 6.8. 2 (corresponds to 59) Key column (variable to enter the basis) — Mi 1 3 (Corresponds tox-tolumn) Table 6.8 4? = 2 0 0 or Variables Solon inBasis, Values n Boos 2 Bens) ° A 13 mw OU CD -I a 23 2s 10 8 Z=-28 % =23 0 1B 73 ° 13 ‘Table 6.8 shows that solution is still infeasible (because s, iteration, Iteration 2 Variable s, = — 1/3 is the only basic variable having negative value; it should therefore be removed from the basis. Variable to enter into the basis is selected as: 1/3) but optimal. So we proceed to second Jia -78 -13) Key row (variable to enter the basis) = Min 35 <0}= Min J 5 Yo = 1 (corresponds to s,-column) ‘That i, the non-basie variable 32 enters into the basis. The new solution is shown in Table 6.9. Table 6.9 - 97 - J 0 o Variables Solution inBasis Valves 4 & 5 8 Bb, 0 a 1 1 0 3 1 “1 s 1 1 1-1 oO 1 1 1 1 0 -2 0 1 0 InTable 6.9, since all ¢)~zj<0 and all solution values are also positive the current solution: x; =0,x= 1 and Max Z= |, is the optimal basic feasible solution.Example 64 Use dual simplex method to solvethe following LP problem Minimize Z= 10x + 63+ 2x5 subject to the constraints @ =x txytx2 1, Gi) 3x, +xy— 4522 and 19.8520. [MS Tirunvelli Univ., B Se (Maths), 1996] Solution To apply dual simplex method, coavert < type inequalities into > type as usual and Min Z Max Z*, where Z* =~ Z, The standard form of LP as follows: Maximize Z*=— 10x 6x ~2xy+ 05; +05, subject to the constraints @ y-m-wyt9, and Hy p¥H 515220. ‘The initial basic solution (infeasible): s, =—1,s) =—2 and Min Z=0 is obtained by setting x, =x, =1, = 0as shown in Table 6.10. Gil) -3,, wntayte= ‘Table 6.10 Initial Solution 67 “0 <3 oo cs ‘Variables Solion inBass Values 4 ” 5 5 8 2 oan) ° a 71 1d a) o 2 3 Q = o 4 z=0 : 0 0 ° oO <0 6 2 oo 2 is the largest negative solution value, the basic variable s, is selected to leave the basis. The non-basic variable to enter the basis in place of sis selected as: Iteration 1 Since sy GE5. Key column= Min}; y,, <0} = Min M5 Thus non-basic variable x, enters the basis as shown in Table 6.1). ‘Table 6.11 g> “0 7 o ° cy Variables Soltion inBasis ‘Values y * % si 4 B be 0 PA 3B 0 “2B T 1B 10 x 23 1 13 v3 © v3 Z=-208 ° 103103 o 103 -10 88-1683 0-103 Iteration 2 Since s, =— 5/3, the solution shown in Table. 1 is still infeasible. Thus, is selected to leave the basis. The non-besic variable to replace s} in the basis is selected as: g 0 ne mead ‘Thus non-basic variable x, enters the basis as shown in Table 6.12. Key column= Min‘Table6.12 47 -10 0 0 ce Variables Solution inBasis Values x m a a 2 B 6 2 sia 0 1 2-34 Wh -10 x 4 1 0 -12 v4 14 Z=-203 0 0 2 2 0 0 -4 2-4 In Table 6.12, all ¢ ~z)$0 and all xp,2 0. Thus, the current solution is the optimal basic feasible solution: 2) V4, x2 = 5/4 and Min 2= 10. Example 6.5 _Usedual simplex method to solve the following LP problem Minimize Z~2x, + 2x)+4%y subject to the constraints @ 2x +3x45x22, Gi) 3x + 4-75 $3, iil) + 4ry-Gy SS and 2X94 20. [Punjab Univ, BE: (17), 2004] Solution To apply dual simplex method, convert S type inequalities into 2 type as usual and Min Z =Max Z*: where Z* =~Z. The standard form of LP problemis as follow: Maximize Z* =~ 2x, ~2x)~483 +05, + 0x;+ Oxy subject to the constraints © ~2x)-3x- 5x34 and Hy ae45,5).S965520. Aninitial basic solution: s, =- 2, in Table 6.13. 2, (i) 3x txt Tey 57=3, (hil) x) Fart Oy +5=5 83 = 5 and Max Z= 0 is obtained by putting x; =x)=13 = Oas shown ‘Table6.13 Initial Solution a 2-4 09 0 6 ce Varabes in Basis ee ee ee ee 2 0 2 @D -s 1 ° o 0 8 3 3 ro o7 0 1 0 5 5 1 46 0 oo Zao % ° ° a -, | 2 -2 -§ 0 0 . Since all cj~zj $0 and all solution values are not positive, the solution shown in Table 6.13 is optimal but infeasible, Horation 1 Since s, ~~ 2 is the only negative solution value, the basic variable sis selected to leave the basis. The non-basic variable to enter the basis in place of s, is selected as follows: S74 =2)-3'° =5 + o Thus, the non-basic variable.) iseligible to enter into the basis as shown in Table 6.13. To enters, variable into the basis, applying necessary row operations. The new solution is shown in Table 6.14, Yn onseaps‘Table6.14 Optimal Solution Ce <2 -2 0-4 o 0 o co —-ariables Solution inBasis x a ny 4 nos B a 23 2B 1 SB 1B ° n 2B 1B 0 163 us 1 0 3 1B -53 0-23 438 1 % 4B «-2 108 0 0 ey -23 0-23-23 ° InTable 6.14, all ¢,-z;<0 and all solution values are positive, the current solution is optimal solution. The ‘Show that the value of the objective function of the dual for any feasible solution is never less than the value ofthe objective fune- tion ofthe primal corresponding to any feasible solution. Use dual simplex method to solve the following LP problems Min pect to () 201 42222, ad wea 20. Max 2= 2x, +28) subject to () x) +2521, and se 0. Min 24,4 24g + Sey subject to () xj —3y #24, Gi) y-m22 and e320 Min 2= 39,4205 +5 +454 subject to (i) 2x, + 4x, + Sxy+.xg2 10, Gi) Bx ~ y+ 7xy-2yS2 (ii) Sxy+2xy + x5 + 6x42 15 nd amnnnt 20 [Meerut Univ,, M Sc (Maths), 1995) ate Gi) -x-m21 Gi) Aytxyzt (@ +m 42q88 Min Z= 21, + subject to ()) 3x; +x 23, (ily 24 #2523 and a2 0. @ 4x +326 (Meerut Univ., M Se (Math), 1994, ‘Madras Unix, BE (Mech), 1999) 3. Show with the help of an example how when one solves an LP protlem by simplex method going through infeasible but better than optimal solution, ne directly goes through infeasible beter than optimal solution of:the dual LP problem. How is this fat utilized in the solution of the dual? 6. Min Z= x, +2x) subject to () 2ry + p24, and (@ x42Q57 saa 0. [Meerut Univ, M Se (Maths), 1999) 7. Min Z= 6x, + 2xp+ 3x) + Seq subject to (i) Sx + 6x, —3xy + 4xq2 12 22 *S4j— 6B 10 Gil) 2+ 5+ 14+ 428 ita ty%420 [Meerut Univ, M Se (Maths, 1990) and 8, MinZ=3xy + Sx) dry subject to @) ~ 21, ~ xy +5422 GD 3x, +2x + 445 216 and yt 20- 9. Max Z=— 4x; ~ 64) ~ 18x subject to @) xy +3423, G@) mtIqn 2S and sytnts2 0, 10, Min Z=2x; + 3x subject to (2) 2xj—x9-423, (Wy +22 and sytn2 0, 11. Min 2= 80x, + 60x, + 802, subject 9 ()) x) #2m+3y24, Gi) 242% +324, sity 20, (i) 2m #3483 Ge) aty+ xp $3526 and Copyrighted materialMin Z= 2x, +3425 subjectto (i) 4x, + 6c) +3x328, Gi) xj 9+ 4yS-3 (i) -2y)-3yy ty S—4 and smn sy 20. Max Z= 3x) —2 subjectto () y+ x21, Gi) 2x, 434,22 and aap? 0. Min Z= 105; + 642 + 2x5 subject to () —2yxyay2 1. and Gi) 3x, tga >2 1a 520. Dual Simplex Method Max Z =~ 6x) ~ Tx, ~ 325 —Sxy subject to (i) 2) +5xy 425 +4428 (i) xy #45 x42 10 Gi) Sy +6, Bey + Ay 2 12 Sty 20, [Meerut Univ., M Se(Maths), 1999, 99} and Max Z=-2x)-2x3 subject 0 (i) a) 4-25, ond ) x2 tages sayy 20. [Karnataka Univ, BE(Mech), 1996] HINTS AND ANSWERS. ‘Write Min 2~ Max (- Z) and convert all the constaints ofthe Stypes.xy = 1,4) ~ Cand Min Z=~1 xy = V3.2 = Vand Min Z= 43 3 1-32 2. = O-and Min Z= 10 x, = 65723, xy © 1yy = 20723 and Min Z = 215/23 5. xy = 3/5.) = 6/5, and Min Z= 12/5 ‘and Min 7. xy = 16/13, x)= 6113, x= 8/13 and Min Z= 2280/13 x)=)" 0, = 4 and Min Z=16 9. x) =0, 8° 3.x Land Min Z x) 5/3,43~ 0,45 ~ 13 and Min Z~ 138 H, 1 = 16/13, 3, ~ 6113, 33 ~ 8/13 and Min 2 ~ 2280/13 2)" 0,1 = 4/3, x5 = Cand Min Z = 4/3, 13. x) = 0.x = Tand MinZ =~ 1 44, x) = 5/4,x)=0 and Min x) = 30/L Ly = 16/11, x4 = Oand Max Z=~258/11Coli ramesC ‘© Introduction ‘© Sof Practice Problems. ‘+ Simplex Method for Bounded Variables = Hints and Answers © Solved Examples INTRODUCTION Inaddition to the constrains in any LP problem, i the value of some or ll varcblesis restricted with lower and upper limits, then in such cases standard form of LP problem appears as: Optimize (Max or Min) 2= $63, sibjecito Be contin. = ay x js where /, and 1, denote the lower and upper bounds for variable s, respectively. ‘The inequality constraints /, 0 for the non-basic variable columns, and this column has atleast one entry positive (ie. yy >0) for some row i, then it indicates that an improvement in the value of objective function, Z is possible. Step4 IfCase(b) of Step 3 holds, select anon-basic variable to enter into the new solution according to the following criterion: ible solution. Ifany of the basic variables is at a positive lower bound, ogy = Max fe) — 3, 6 -#) >0) Step 5 Aficridentifyingthe non-basic variable which will enter into the basis matrix B, the basic variable to be removed fiom basis, Bis calculated. For this calculate the quantities: 8, = Min fe ed ML, and 8 = Min (0,, ®,u,} where u, is the upper bound for the variable x, in the current basic feasible solution. Obviously, ifall y,,>0, then 6. (a) 1f@=6,,then basic variable x, (column vector a,) is removed from the basis and is replaced by non-basic variable, say x, (column vector a ), as usual by applying row operations. (b) 1f@=6,, then basic variable x, (column vector a, ) is removed and replaced with a non-basic variable x, (column vector a, ). But at this stage value of basic variable x, = Xp, is not at upper bound. This must be substituted out by using the relationship (pe) ~ pa) Mires OS (pe), Sm, where (xpy ); denotes the value of variables x,. ‘The value of non-basic variable x, is given at its upper bound value while the remaining non-basic les are put at zero value by using the relationship: x,=u,—x/; O0Sx/ Su, If @ = u,,the variable x, is given its upper bound value while the remaining non-basic variables are put at zero value by the relationship: x,~ u,~x/;0 £ 6 we mw we ww 67 3.2 09 0 0 0 029 ‘Variables Soluion inBasis, —“Vawes | x, xy py 8g | my Bo bp) ° 5 3 @® 3 1 0 0 0 of w3=0 o 5 4 Fo-2 0 1 0 00] wend 0 85 20 2 1 6 0 1 6 0x20 0 % 30 130 0 0k Jo 30 ° %5 6 “11 0 ot fo Zo 5 oo o 0 0 0 0 32 6 o o o 2 Since, ~, ~3 is largest positive, variable x, is eligible to enter into the basis. As none ofthe basic variables 5, t0 5g are at their upper bound, thus for deciding about the variable to leave the basis, we compute o.= Mn 9¢} mn 2.4.2.2) aot : rary - 2B )} = 2x8 = nr sa <0 py 722 Ceoresponds toss) ‘a Thus © = Min {8,,0,1¢,) = Min (3, ,8) =3 (corresponds 0, ) ‘The non-basic variable sis eligible to leave the basis and therefore p= 1 becomes the key element.Introduce x, into the basis and remove s, from the basis by applying row operations in the same manneras . 9 3 0 0 0 0 on -3 0 0 oo 0 5, variable to leave the basis, we compute un {5 mo} =Min { T We } = 1 (corresponds to x,) unt 8-3 = 5 open [2 oy cee B25 2S apt Therefore 6 = Min {6,, 6, v9} = Min (1, 5/3, 6} = 1 (corresponds tos.) Hence s, will leave the basis and y,, = 1 becomes the key element. Introduce.x, into the basis and remove ‘5 fom the basis as usual. The improved solution is shown in Table 7.3. 3 Table 73 47 owe 6> 302 0 0 0 0 oO @% Variables Solution inBais Values | x, mp | my B bx») 3 3 6 1 « @ 3 0 0 oO 2 " 1 > 1 Fr 1 0 0 0 o 53 7 > 0 $ -7 1 0 0 o % 2 9 © 5 6 0 1 0 0 55 u oo = 2 0 0 ftw Z=20 > 2 -8 0 0 Oo 5 0 ¢ 8 0 0 oO 2, isthe largest positive, therefore variable s, e- win {2 > 0M 2.22} = 2 (eomespond tos) "Vee eligible w enter into the basis. We compute 8-6 6-1 =(-2)'-GD'-CD win =m.» $ 6 © ww 2 67 320 0 0 0 0 Variables Sobution inBasis Vales «| ox 8p tet | Bo bx) 0 5 3 “2 0 1 32 0 0 2 y -2 -2 1 0 -12 0 0 ° 85 2 m2 0 0 Wm 1 0 o % 36 2 0 0 3 oO 1 o 55 8 -! 0 09 mio 0 ; 1 2 0 -1 0 0 oO o 4 0 0 1 0 0 0 Since c, 2, = 4s largest positive, variable x, is eligible to enter into the basis. Also the upper bound for variable. is 8; therefore we update the value of basic variables by using relationship and data of Table 7.4 3p1 =5) =p ~My =-3-C12)8 =I Sag = ¥27 hy ~Yayt =— 2-128 =2 yy 8h P51) = 22 (GIB HD Jaa fa “Yat = 36 (5/2)8 = 16 ps fs Ys = 8—(-1/2)8 = 12 ‘Also one of the non-basic variables x, at its upper bound can be brought at zero level by using the substitution y= my— x] = 8-4 08 af <8 ‘The data of Table 7.4 cannow be updated by substituting new values of basic variables as well asnon-basic variables as shown in Table 7.5. Table7.5 we v0 we oo 320 0 0 0 0 cp Vriables Soition inBess oes | Xm B b(=x_) ° A 1 rr er er er) 2 = 2 2 o 12 0 0 0 ° z -2 0 0 @i 0 o ° i w |-s2 0 0 wm 0 1 0 ° 5 2 moo 0 mio 0 4 Zama 12. 0-1 0 0 0 = 4 0 0 1 0 0 0 Since ¢,~z is the only positive value, s will enter into the basis. For deciding about the variable to leave the basis, we computePat Operations Research: Problems and Solutions Sy emin{2. 16 2. 7 oy mn 2, 1 of mtn 26 2 nn 20024 4 (corresponds t0 3) 0,=Minj@—*# y< Oh =Min f=) | 8 Gorespocids tox) t Ya 32)'=(-12), 2 ‘Therefore @ = Min (6,,0,,144}=Min (4, 8,0} = 4 (corresponds tos, ) Hence, the basic variables, will leavethe basisand y,4~1/2becomes the key element. Introduces into the basis and remove s, from the basis as usual. The improved solution is shown in Table 7.6. Table76 a 6> 3 2 0 09 0 0 0 cy Veriables — Soiation ies Wier | om a oye | Bo bsg) ° mr 7 a 2 = 4 QDs o o 1: o of 62 ° 5 4 “5 0 0 1 2 0 | #2 ° % 0 5 0 0 9 -3 1 0 | wt ° a 0 20 0 0 1 0 | wnt Zane4 a rr a =2 aes 1 0 0 0 -2 0 0 Since e, ~7 isthe ony postive value, variable xj willenter the bass. To decide which variable will eave the basis, we compute 12 |= 2 canegentson ts} 1 (corresponis tox, ) Mia f: Vy 5) Therefore © = Min {0,, @,, 4) = Min {12/5, 1, 8} =1 (corresponds to x, ) Hence, the basic variable x, will leave the basis and y,, =~2 becomes the key element. Introduce 2 into the basis and remove x, from the basis. The new solution is shown in Table 7.7. ‘Table7.7 49 8 6 www 9? 3 2 0 0 0 0 0 Variables Solution inBasis Values Hoo yy "Xy Bo b-x,) 0 4 -7 0 -m2 1 0 - oO 0 “3 ay -2 | - 0 0 -2 0 0 0 Sy -6 0 -2 9 1 -2 0 0 o 5 2 o 82 0 oO - 1 0 0 55 16 o 2 0 oO mio 4 Z=30 30 2 0 0 8 0 0 o wi io 0 -1 o 0Bounded Variables LP Problems Since the upper bound for variable x, is 6, we update the value of basic variables by using the following relationships and data of Table 7.7. Xp =, = 8h — €72)6=14 Bag =X] Xb9 Yan th - €12)6= 1 3 = 52 = Xf3 yoy - €52)6= 9 ey ps “Yaa - 626=7 ps = 85 = X55 —J50 Mp - 26-7 The non-basic variable x, at its upper bound can be brought at zero level by using the substitution Xp = t= 35 =6- x5, 0Sx5 56 The data of Table 7.7 can now be updated by substituting new values of basic variables and non-basic variables as shown in Table 7.8. Table 7.8 “> s 6 © © @ 47 —_— 0 6 0 0 6 en Variables Solution inBasis Ho my Om 8s B 0 5 o om 4 o 12 0 0 -3 x 1 12 0 6 -1 0 0 o as 0 s2 0 1-12 0 0 54 0-32 0 0 -12 1 0 0 85 0 32 0 9 mo 1 3-2 0 0 m0 0-12 0 0 -32 0 In Table 7.8, all ~z)<0, an optimal solution is arrived at with values of variablesas:x',=1orx, = 8-1= 7x; = Wy—*',=6-0= Gand Max Z=33, Example 7.2. Solve the following LP problem Maximize Z= 3x, +51) +2%y subject to the constraints 0) xy#2x,+25,514, and (ii) OSx,$4; 2Sx,$5; OSx,53. Solution The variable x, has a positive lower bound, therefore taking 1, =x, ~2 or x) = x')+2. Thea the fourth constraint ofa given LP problem can be written as 0x", $3 and new LP problem will become Maximize Z= 3x, +5(x!,+2)+2ey= 3x, + Sx, +25 +10 subject to the constraints © x +245 +2)+2,514 or Gi) 2x +4044 #2434523 or and Osx,S4; OSx5<3; OSx,53 (i) 24, 4x, +34, £23 xy +2x5 +255 $10 2x, +4x5+3y,515 Introducing non-negative slack variables s, and s, to convert inequality constraints to equations. Then standard form of LP problem becomes PoMaximize Z= 3x, +5x}+2x,+ 10+ Os, +05, subject to constraints © Hy t2x5 4 Deets = 10, i) 2x, 44x34 3x44 xy=15 and Aye Kho pSpo 5) ZO. ‘The initial basic feasible solution to this problem is:xg,=5;= 10, xg)"s)~ 15. Also for the basic variables ‘5, and s, no upper bounds are specified, itis therefore assumed that both of these have an upper bound at oo, The initial basic feasible solution can be read from the initial simplex Table 7.9, ‘Table 7.9 Initial Solution "> 4 3 3 @ q7 3 5 2 0 0 cp ‘Variables Solution inBosis Values Se ee ee a B 5(=x5) 0 my 10 1 2 2 1 0 | #-W== o % 1s 2 @ 3 9 1 See Z=10 oo o 8 0 oy 3005 2 0 0 Since ¢)~2,~ Sis largest positive, variable x willenter into the basis. As none of the basic variables s, and ‘5, are at their upper bound, thus for deciding about the variable to leave the basis, we compute xpi 1s 0, = Min} 8, = (corresponds tos) P Da 4 ©, =, because all entries in Column 2 are positive, ie. 7; >0 forall i Therefore @ = Min {6,, 0,4) = Min {15/4, , 3} =3 (corresponds tox}. Hence, the non-basic variable x; is substituted at its upper bound and remains non-basie. The non-basic variable =r at its upper bound can now be pat at zero value by using the substitution: xp sujoxfa3-xf; Osx7 <3 The value of other basic variables are updated by using Fat 81> Pita %p2 = 27 tho Yaa The data in Table 7.9 cannow be updated by putting new value of basic variables and non-basic variable x3 as shown in Table 7.10, Table 7.10 "> 4 3 3 # 97 3 “5 2 0 0 tp Variables Solution ‘in Basis Values x os xy 1 3x Fai B bla) 0 nh + 1-2 2 1 0 | «due 0 hy 3 @ -4 3 0 bo | 2-3-0 Z=15+10 A oo 0 0 Oo 5 3-5 2 oo In Table 7.10, c; ~2, = 3s the largest positive, therefore variable x, will enter into the basis. For deciding about the variable fo leave the basis, we computeBounded Vaiables LP Probione PARE = tn, >of 0, forall Therefore 8 = Min {@,, 8), ;} = Min {3,<2, 4) =3 (corresponds to 8,).. Hence, the basic variable s, will leave the basis and yy, =2 becomes the key element, Introduce 4 into the basis and remove s, from the basis as usual. The improved solution is shown in Teble 7.11 3 (comesponds to s,) ‘Table 7.11 “> ‘ oo = a 3 > 0 6 @ Yes — Sots aes ee | www | B (=p) 0 i sa ce nT a a a 3 5 32 1 @ wo w)3-sa-30 75190 36S =592 o 1 = 0 =a In Table 7.11, c, -z_= 1 is the only positive value, variable x3 will enter into the basis. To decide which variable will leave the basis, we compute 6, ==, because all entries in Column 2 are cither negative or zero, i.e. yy $0, forall 7 4 52 e=min| a J <9) =f, 2} -§ (corresponds tox,) Therefore @ =Min {0,,,, u,} = Min {= 5/4, 3} = 5/4 (corresponds to 8,). Hence, variable x, will leave the bass. To put x, atitsupper bound, substitute, improved solution is shown in Table 7.12. 4-x{ in Table 7.11. The ‘Table 7.12 “> 4 3 3 ~ ~ o> 3 a 2 0 0 & Variables Solution in Basis Values ca ag % s s 5 OG x5) ° sy 52. 0 0 12 1 m12 -5 x -v4 -12 1 -v4 0 -"4 Z=25 +154 % 52S 15/4 0 34 1154 57% V2 1 -14 0 ~54 The value of non-basic variable x, at its upper bound 4 can be put at zero level by substituting -x{; OSais4 4 The value of other basic variables are updated by using the relationship: 1 A xg yy = 52-0%4=52 qo =49 = xhp —Yoit, = -3M4—C 1D) x 4= 514 ‘The data in Table 7,12 can now be updated by putting new values of basic variables and non-basie variable -x{ as shown in Table 7.13.Eg] versions Resear: Probie and Sottions ‘Table 7.13 42> 4 3 3 ° - 9? 2 0 0 Variables Solution in Basis Values x s % 4 * B Jp) 4 sn 0 12 1 -12 x si 1 34 va % -5 40 3 5 0 =m 0-804 {In Table 7.13, all¢)~z)$0, an optimal solution is arrived at with values of variables: x 745 XY = 5/4003 x5 *5/4 0F3— (x2) = $/4 or xy ~ 15/4 and Max Z~ 123/4. Example7.3 Solve the following LP problem: MaximizeZ=2x, +x, subject to the constraints @ x #2eS10; @ x 4x86 Gil) t=) 5! (iv) ,-28, and 0Sy 53, 05x82 Solution Introducing non-negative slack variables 1, 63, standard form as stated below: MaximizeZ=2x, +1, +05, +05,+05+0s, subject to the constraints @ x, 42445, Gi) xy += ind s, to convert given LP problem into its 0; Gi) tat, dy aay taged and pep Spry 5420. ‘The initial basic feasible solution to this problem is: s, = 10, 6,5, =2 ands, = 1. This solution is shown in Table 7.14. ‘Table 7.14 Initial Solution 49 a 2 - ° ° 52 2 ’ o o 0 ° &y_ Variablesin Solution % a % 55 4H Basis Values B dex) ° iy 0 7 2 H ° ° ¢ 0 b 6 1 1 0 1 0 0 0 55 2 1 -1 ° 0 1 0 o 4 \ @ 2 0 0 0 1 Zo z 0 oo 0 0 ° 5 2 1 0 0 0 4 As shownin Table 7.14, enter variable x, into the basis. To decide the variable to leave the basis, we compute 6, = Min ny 16 621 BL yy > 0) = Min {8 4} = 1 (corresponds to s,) Ya 1 - since all elements in x,-colurmn are positive, and. 1, =3Bounded Variables LP Problems ‘Therefore, 8 = Min {8, 85, u,} = 0, (= 1. corresponds to s,). Hence, variable s, leaves the basis. The new solution is shown in Table 7.15. ‘Table 7.15 "> 3 2 « we ~ “> 2 1 0 0 D 0 ce ‘Variablesin Solution HOB 5 3 % se im Basis Values B bx) 0 % 9 0 4 1 ° ° 1 | --9-—= 0 5 5 0 3 0 1 o +1 0 fs 1 0 0 0 | 2 s l 1 0 ° ° 1 2-2 5 2 ° ° 0 2 0 5 0 0 0-2 In Table7.15,c,~2; is the only positive value, the variable.x, will enterinto the basis. To decide the variable to leave the basis, we compute Xu 951 BL, yj > 0} ~ Min{>,3, 1) — 1 (comesponds to» {22 Yas | 34} (wonesponds 10.) 6, = = in| and m3 2 =y_<0f -(-2) Therefore, 8 = Min {0,, 0,1} =8, oF 0, Selecting arbitrarily 0, so that variables, leaves the basis. The new solution is shown in Table 7.16. Table 7.16 a 302 mm “- 2 1 9 o 0 2 ¢y ‘Fariablesin Solution x % s 8 83 S| okey Basis Values 2 besy ° s 5 oo «oO - 3 ° 5 2 o 0 0 4 32 1 4 1 o 1 o 0 boot 2 x 3 a ee 2 QDI 3-0 a7 5 2190 = 3 oo oo oS In Table7.16,c,-z.=3 isthe only positive value, variables, will enterintothe basis. To decide the variable toleave the basis, we compute Xai ar ua § 3} Min} 8, y,4 >0}= Min }>. t fe var | ora in | Yin <0 ay Therefore, 8 = Min {0,, 0,145} = 0 1 (corresponds to s,) |-0 (corresponds tox) corresponds 10 x). Por‘Thus, selecting variable x, leave the basis. The new solution is shown in Table 7.17. Table 7.17 3 2 ew 2 1 0 0 0 0 Ho 0 3 0 1 0 2 0 ° 2 ° ° 1 1 ° 1 -1 1 0 0-1 ° ° -1 0 0 0-2 1 Ze-2 % =I 1 0 0-1 0 3 ° ° ° 1 ° In Table 7.17, the variable.x is eligible to enter into the basis. Also the upper bound for variable x, is 3, therefore, we update the value of tasic variables in Table 7.17 by using the following relationships 1 = I-G)3=S tps = py Ym =8-O)3=2 %pp = Xbs Py =-2-C1)3 5 %py = Abs —Yant =—3-(1) 3=0 The non-basic variable x, at its upper bound can be brought at zero level by using the substitution, xytmy-xf=3-x( 08x, $3, The data of Table 7.17 can now be updated by substituting new values of basic variables and non-tasic variables as shown in Table 7.18. Table718 4 302 @ we me 67 2 1 0 0 6 0 “Variables ‘Solin [| om = 5 ~~ % | 4%, mBois Yates B b=xs) ° nh 5 37 0 1 0 2 0 ° 5 2 2 0 0 1 1 0 1 4 1 1 1 0 0 @& o ° 5 ° 1 oo od Z=1 a 1 1 o 0 -1 Oo gy | -3 9 0 oo 10 In Table 7.18,c, ~25~1 makes the variable s, eligible to enter into the basis. To decide the variable to leave ‘the basis, we compute es ais es in|, va>0h = inf is (corresponds to) 6, = Minj“i“*8, 5, <9 4 { ag 78 | =E Therefore, 0 — Min (0,,0,, us) ~0, (~ 1, corresponds to x,). ‘Thus, non-basic variable x, will leave the basis. The new solution is shown in Table 7.19. (corresponds to x,)wander one ESET ‘Table 7.19 4 3 2 o © ~ go 2 07 0 0 a 0 ep Variables Soluton |X; XS SOS]SSCSSCS inBais Values B 4 0 5 7 | -1 2 1 0 0 0 ° 8 3 1 ° 1 ° ° 0 & 1 “1 0 0 ' 0 0 54 -2 -2 0 0 0 L 2-6 0 0 0 0 0 0 1 0 o 0 0 In Table 7.19, the variable x, is eligible to enter into the basis. Also the value of non-basic variable x, at its upper bound 2 can be put at zero level by using the following relationships 7-Q)2 3-2 D2 =1 yy = thy ¥gt=-2-CD2— =: Also the non-basic variable at its upper bound can be brought at zero value using the substitution =3-x:05x453. ‘The data of Table7.19 cannow be updated with new values of basic variables as well as non-basic variables as shown in Table 7.20. Xin Yaka gp = bY =-1 ioe ‘Table 7.20 "> 32 mm mw go>? lt “ Toriaties Soon | x XS SSCS in Basis ‘Values B Hex) o a 3 a ° % 1 ee ee en) ° a 1 fr 0 ss 2 12 oot Z=8 . 0 0 0 0 10 0 oo 2and Max Z= SELF PRACTICE PROBLEMS Solve the folowing LP problems 2. Max Z= Bn -ty +55 +7 1 Max Z=4 +34, subjest to @) 2x, +3m—m +4n,< 40 subjecto (@) x txytms10 Gi) x, 25,50 (i) ~ 2x, +24, # 54,0 $35 Gi) 25,-¥,5 10 (i) 5, Fxg 2ey + Buys 100 ant Osx SHOSx 54 520. and 522, a lay dong 2. Max Z= 4x, +415 + 3x, [4 Max Z=45, 11055495 + Hy subjectto (i) -1, +2, 434,515 (ii) y+ ys 4 | subject to (i) 2x, + 2p +2, #25455 Gi) Ay FH—MyS6 GW) X-_tAASIO | (i) 48, + 80, + 160s, + 240545257 and 02x $8 05x 54; Ox, 54 and Oem j-1,2,345, MinZ=—2x,~ 45-2, 7. Max Z= x, +3438, subject to (i) 2x, +x, +255 10 subject to ()) x +x, +4812 ata sd @ -x+H85 and 0sx,S4; 0SnS6 15x54 (i) mp +2y<8 6 Min Z=x, + 2x) + 3x3—4 and OSx,53; 051,56; OS, 54. subject to (i) x, x2 +x; ~2r, $6 . 8 Max Z=2x, + 3x —2r5 Gi) 5, Paymay tags 8 subject to () x, +x, +y)S8 ii) ey +xy-y22 () 2542-223 and Sx $3; 1S $4 05x, 510; 281,55, and OSx,S4; 25x56, 1522 "2003, = 0,35 = 108 and Max Z= 10 2. x= 1715, x," 165, x) =4—x,=Oorx, = 4.and Max Z= 192/5 3. xf = 63/4 or? —x, = 63/4 orx, = 71/4; x5 y x4 *23 or 3-1," 28/2 rx, = 292; x = or 4—x,= 0 orx,=4 and Max Z= 387/4, 4, x)= 9/16, x5=0orl-x)=Oorx=1; xy=15/15, 1X] =O orx,=O.and Max Z= 331/16. 5. x72) ~6,~ 8 and Min Z = 28, jor 1-x,=O0rx, Copyrighted materialChapter Outline ‘© Introduction '* Review Questions © Economic Analysis of Dual = Self Praciice Problems ‘© Rules for Constructing the Dual irom Primal © Hints and Answers © Solved Examples BREROBUETION Each LP problem (either maximization or minimization) stated in ts original form is associated with another unique linear programming problem (called dual linear programming problem or in short dual), tased on the same data. In general, itis immaterial which ofthe two problems is called primal or dual, since the dual of the (greater than or equal to) type constraints and vice versa, Thus, the inequality sign is reversed in all the constraints except the non-negativity conditions. 3. Each constraint in the primal corresponds toa dual variable in the dual and vice versa, Thus, given a primal problem with m constraints and m variables, there exists a dual problem with m variables and » constraints. 4, Theright hand side constants bb, .b, of the primal becomes the coefficients of the dual variables y,, Vy- «Jy in the dual objective function Z,. Also the coefficients, ¢, ... ¢, of the primal variables x;, Xs Xp in the objective function become the right hand side constants inthe dual 5. The matrix of the coefficients of variables in the constraints of dual is the transpose of the matrix of coefficients of variables in the constraints of primal and vice versa. Thetis, coefficients of the primal variables x,,¥,,.. .%, in the constraints of primal LP problem are the coefficients of dual variables in first, second, ... nth, constraints for the dual problem, respectively. 6. Ifthe ith dual variable is unrestricted in sign, then jth primal constraint is = (equality) type and vice versa. SOLVED EXAMPLES 1 ee Example 8.1 Write the dualto the following LP problem Maximize Z= x, —x, + 3x, subject to the constraints @ xptx tx S10, (i) 2x)- Ox —x, $2, (i) 2x, -2xy-3x, 56 and Appts = 0.‘Duality in Linear Programming Solution In the given LP problem there are m = 3 constraints and n = 3 variables. Thus, there must be dual variables and n= 3 constraints. Further, the coefficients of the primal variables, c, = 1, =1, cy = 3 become right hand side constants of the dual. The right hand side constants b, = 10, b,~2, bs ~6 become the coefficients in the dual objective function. Finally, the dual must have a minimizing objective function with all 2 type constraints. Ify,, 72 and y, are dual variables corresponding to three primal constraints in the given order, the resultant dual is Minimize Z, = 10», +2y, + 6y; subjectto the constraints @ y,+ 29424215 Wy, ~Oy)-2y,2-1, Gi) and I ae520. ey 239523 Example 8.2 Writethe dual ofthe following LP problem Minimize Z, = 3x, ~ 2x, +44, subject to the constraints 3x, +54, 444,27; WH) 6x4 x4 3,2 4 i) Tx, -2x —¥y $10; (iv) xj 2x) +4423; (¥) Ax, + Ixy 2x22 and xp xpey20. Solution Since the objective function of the given LP problem is of minimization, the direction of each inequality has to be changed to 2 type by multiplying on both sides by 1. The standard primal LP problem so. obtained is Minimize Z, = 3x, ~ 2x) + 4x, subjectto the constraints © 3x, 45x, 44r,27; @ br, +x 43x,24: ii) Fx, +2x, 4242-10; (i) x,-24, + 5x23; (¥) 4x, 7x)-2y22 and XppXypty20. Ify,,9’p-¥3.4and ysare dual variables corresponding tothe five primal constraintsin the given order, the dual of this primal LP problem is stated as: Maximize Z,= 7); + 4y3 ~ 10); + 394+ 2y'5 subject to the constraints 0 By, + 6p Bry hyg hy S3s Gil) 49, + 3y +99 + Sy4- 25 S4 and PpYeVp lp ¥s20- Example 8.3 Obtain the dual problem of the following primal LP problem: 24 D) 591+ yy+2y,—2y4+ Ivy 5-25 Minimize Z=2, subject to the constraints 0 2x, 44x 5160; (i 4)-=30; i) x, 210 and xy20. Solution Transform all< type constraints to > type constraints by multiplying the constraint on both sides by =L. Also write = type constraint equivalent to two constraints of the type 2 and <. Then the given primal problem can be written as: Minimize Z,=.x) + 2%) subjectto the constraint © ~2x,~4r,2 160; Gi) x, -¥,230: (ii) x;-2, $30 or -x, +, 2-30 (iv) x2 10 0. and yk Pa8 Operations Research: Problems and Solutions: Let y;,.¥p,)'s and y, be the dual variables corresponding to the four constraints in the given order, then the dual of the given primal problem can be formulated as follows: Maximize Z, ~~ 1603, + 30y ~ 30h, + 10), subject tothe constraints (i) —2y, ty -¥y tg 81 (ii) -4y,-y,+y3S2 and MpVarV5rV42 0. Let y=y2—Ys (7393's 20). Then the above dual problem reduces to the form Maximize Z,~ — 1603, +30) + 1034 subject tothe constraints (@) -2,+y+ 7481 (ii) -4y,-ys2 and Dio¥qz 0; y being unrestricted in sign, “To apply rule 6, note that the second constraint in the primal is equality, therefore, the corresponding second dual variable y (= y,~y;) should be unrestricted in sign. Example8.4 Obtain the dual problem of the following Minimize Z, =x, -3x~2r, subject to the constraints () 3x)—xy +28 $7 i) 2, -4x,2 12 (ili) — 4x, + 3x, + 8x and xyo%32 0; 2 unrestricted in sign. al LP problem: Solution Since x, is an unrestricted variable, therefore, it can be expressed as the difference of two non- negative variables, .xy= 5 — af, x4, 2f 20. Then the given LP problem can be written as: 1-3 2 (35 — 38) Minimize Z,. subject to the constraints () 3xj—x,+2(xh AF) ST or ~3y, +-A xh X8)2-7 (ii) 2x, -4r, 212 (ill) — 4x, +33, +8 (x5 — xf) 10 and xyty 8 20. Let y,, 2 and yy be the dual variables corresponding to three primal constraints in the given order. As the given problem is of minimization, all constraints can be converted to 2 type by multiplying both sides by -1. Since the third constraint of the primal is an equation, the third dual variable y, will be unrestricted in sign. Now the dual of the given primal can be formulated as follows: Maximize Z, =—7y, + 12y) + 10) subject to the constraints () = +2y,-4<51 i) 9, — 49, +39, 5-3 = 2y, + By, $-2 and ¥in¥%5 20; y unrestricted in sign. Example 8.5 Obtain the dual of the following primal LP problem Maximize Z, = x, ~ 2x, + 3x subject to the constraints () ~25, +) +34 and ry 20. Solution Since both the primal constraints are equality type, corresponding dual variables y, and y,, will be unrestricted in sign. Following the rules of duality formulation, the dual of the given primal LP problem is Minimize Z, = 2v, +3 subject tothe constraints (@) =, +221 Gi) y, +39, 2-2 (iii) 3y, + 4y, 23 and ¥p>) unrestricted in sign, (i) 2x, + 3x4 45,‘Duality in Linear Programming [PES Example 8.6 Write thedualof the following primal LP problem Maximize Z= 31, +1) +245 subject to the constraints () 2x) -ay + 3x5 Fx, and ayy 82 O and x5, x, unrestricted in sign, (i) xy $x) 357%, Solution Here we may apply the following rules of forming a dual of the given primal LP problem, (i) Thex, and x, variables in the primal are unrestricted in sign. Therefore, the third and fourth constraints inthe dual shal be equalities. (i) The given primal problem is of maximization. Therefore, the firsttwo constraints in the dual will be> ‘type constraints (iii), Since both the constraints in the primal are equalities, the corresponding dual variablesy, and y, will beunrestricted in si Ify, and y, are dual variables corresponding to the two primal constraints in the given order, the dual of the given primal ean be written as: Minimize Z,=y, + 3y, subject to the constraints ( %+y,23 (i) -y, 4,21 Gi) 3y,-,=2 i) Hy and ‘Vy ¥2 unrestricted in sign. Example8.7 The optimal solution simplex table for the primal LP problem. Maximize Z= 31, + 4x, +25 subject to the time constraints (i) x, + 2x, + 3x5 90 (Operation 1) (ii) 2x, + + x, £60 (Operation 2) Gil) 3x, + x44 2x5 $80 Operation 3) and xp ZO isgiven below: 6 3 4 1 o oo Profit per Variables Solution nO 5 Oy Unit — in Basis. Values oe B xn) 4 % 0 0 1 106 4603 * y 10 1 0-18-1300 Boo 0 lig 0 0 0 8%) 8H 10 Z=190 4-4 00-286 -106 23 (a) Find the solution, maximum profit, idle capacity and the loss of total contribution of every one unit reduced from the right hand side of the constraints. (b). Write the dual of tho given problem and give the initial simplex table. Solution To illustrate the interpretation of the dual variables, let us consider this example in a profit ‘maximization production context with constraints on time (resource) input. ‘The primal LP problem in this example is concemed with maximizing the profit contribution from three products say A, B and C, while dual LP problem will be concerned with evaluating the time used in the three operations to produce three products. ‘The productive capacity of the three operations isthe valuable resource to the firm; the production manager wonders whether it would be possible to place a monetary value on its worth. If yes, then how much? His problem can be solved along the following lines.i nl nd sa aa (a) The optimal solution of the primal LP problem can be read from the given table as: x17 10,x)=40,x,=D and Max (profit) Z=3 « 10+4% 40+0x 10=Rs. 190 Since the value of slack variable, s, in the optimal solution is 10, the idle capacity of operation 3 is 10 units. ‘The absolute value of numbers in the ¢,- =, row of final simplex table under the slack variables column representthe value of dual variables, .c, marginal value (shadow price) ofthe constraint resources. Ifa resource (such as time constraint for operation 3 in this example) is not completely used, i.e. there is slack (s5~ 10), its marginal value is zero. A value (neglecting negative sign) in this row under columns 5,3) and 35 represents time. These valves can be interpreted as the amount of increase in total profit which can be achieved when one additional hour in the department heading that column is available. ‘The optimal value of dual variable y, represents per unit price (worth or marginal price) forthe first resource (ie. time available in operation 1), viz., Rs. 10/6 or Rs. 1.66. This means that marginal contribution of operation [to the total profit is Rs. 1.56. In other words, ifone productive unit is removed from the righthand side of the first constraint, the total contribution would reduce by Rs. 10/6, ie. Rs. 1.66. However, ifone productive unit is removed up to the extent of idle capacity from the right hand side of the third constraint, the total contribution would not be affected. (b) Lety,,.7) and, be the dual variables representing marginal value of oneunit ofthe resource corresponding to the constraints 1, 2 and 3, respectively of the primal, that i, the firm pays per unit price of y,,)_ and y, for the time available in the three operations. The firm, however, wants to make sure thet it makes as much profit 4s it would ifit remained in business for itself. Then the dual ofthe given problem is stated below: Minimize Z=90y, + 60), + 80) subject to the constraints @ y, 4 2y, +3523, i) yt ye and Vpn 20. 24 i) 3y,+ 9) +24ZL Example 8.8 XYZ Company has three departments—Assembly, Painting and Packing, and can make three types of almirahs. An almirah of type I requires one hour of assembly, 40 minutes of painting and20 minutes of packing time, respectively. Similarly, an almirah of type II needs 80 minutes, 20 minutes and one hour, respectively. The almirah of type IIL requires 40 minutes each of assembly, painting and packing time. The total time available at assembly, painting and packing departments is 600 hours, 400 hours and 800 hours, respectively. Determine the number of each type of almirahs that should be produced in order to maximize the profit. The unit profit for types I, Il and Ill is Rs. 40, 80 and 60, respectively, ‘Suppose that the manager of XYZ Company is thinking of renting the production capacities of the three departments to another almirah manufacturer—ABC Company. ABC Company is interested in minimizing the rental charges. On the other hand, the XYZ Company would like to know the worth of production hours to them, in each ofthe departments to determine the rental rates. (a) Formulate this problem as an LP problem and solve it to determine the number of each type of almirahs that should be produced by the XYZ Company in order to maximize its profit. (b) For LP problem in (a) formulate its dual and interpret your results. [CA, May 1990; Delhi Univ, MBA, 1999) Mathematical formulation ‘The production data given in the problem may be summarized as below: Tipe of Almira — | __Nunber f Hows Required er Unt] rftper Un Aeenbly Paling rang | ofalmeah ts) 1 1 8 1B ° u a 13 i %” ff 2 6 nn 5 [wa | Let.x, x, and x, be the number of units ofthe three types of almirahs, respectively to be produced. Then the given problem can be represented as an LP model as:Duality in Linoar Programming Maximize Z=40x, + 80x, + 60x, subject to the constraints O +43 +23p; $600, Gi) (3) y+) +23) x5 £800 and ets 20. If ABC Company pays per hour rental price of y,,’, and y, forassembly, painting and packing departments, respectively, then company XYZ must get a total rent equal to: 600y, + 400y, + 800y;, Since company ABC wants 'o pay toial rent per hour as low as possible, its objective becomes: Min (otal rental charges) Z, = 600», + 400s, + 800y, After knowing this rental value, the objective of ABC Company becomes to know what minimum offer shall be acceptable to it. The company, however, wants to make sure that it makes as much profitas ican. The ‘given data show that an almirah oftype I requires 1 assembly hour, 2/3 painting hour and 1/3 packing hour. The time used for making almirah I would be eaivalent to the rent for y, + (2/3) y3 + (1/3)... f ABC Company ‘used that time to make almireh I it would ear Rs. 49 in contribution to profit, and so it will not rent out the time unless, + (2/3), +(1/3) y3240. Similarly, the company will work for almirabs Il and 1. Since the total rent of all the departments should be greater than or equal to the profit from one unit of the ‘almirah, the dual objective functioa along with constraints which determines for ABC the value of the productive resources can be written as: Minimize (total rent) Z, = 600y, +400), +800), subject to the constraints Cy + 2B)y_ + 19) ¥52 40, Gi) 43) y+ (3) y9 +y5280 Gi) Cy, +2D)y, + (3) y,>60 and YoPnya20- Here the dual objective function, i. minimum total rent acceptable by ABC Company is equal to the primal objective function, ie. maximum profit which could be earned by XYZ Company from ts resources. The dual variables »,, », and y, represent the rental rate of various departments corresponding to the slack or unused capacity (called slack variables s,s, and sy, respectively) of these departments and are called the shadow prices ot marginal profitability of various departments. i) (2/3) x, + (13) xy+ (23) x $400 Example 8.9 _A firm manufactures two products A and B on machine | and IL as shown below: Machine Product Avellabie Hours a B 1 30 20 300 " 5 10 ho Profit per unit (RS) 6 8 ‘The total time available is 300 hours and 110 hours on machines I and I, respectively. Products A and B contribute Rs. 6and Rs. 8 per unit, respectively. Determine the optimum product mix. Writethe dual ofthis LP problem and give its economic interpretation. Mathematical formulation The primal and the dual LP problems of the given problem are Primal Problem Dual Problem x, and.x = number of units of A and B to 1y, and y, = cost of one hour on machines I and Il, bbe produced, respectively. respectively. Max Z = 6x, + 8x, Min Z, = 300); + 1103 subjectto 30x; + 20x, $300 subject to 30, + Sy) 26 Sx, + 10x, < 10 20y, + 10, 28 and x2 0 and Yada 20 peedSolution of the primal problem The optimal solution of the primal problem is given in Table 8.2. Table8.2 Optimal Solution of Primal Problem 69 6 8 0 0 4 Variables Solution x) x 5 s, " 1 2 y 2 im Bosis Values zs is B 6 x 4 1 0 v0 10 8 4 9 0 1 -110 320 6 0 orto Table 8.2 indicates that the optimal solution is toproduce: x, =4 unitsof product A; x, Band Z,=total maximum profit, Rs.96 ‘units of product Solution ofthe dual problem The optimal olution of the dual problem can be obtained by applying the Big-M method. The optimal solution is shown in Table 8.3. Table&3 Optimal Solution of Dual Problem b> 300 10 o 0 D Variables Solution yy ve Sa Sg in Basis Values 5 Ja 300 v1 110 v ° “120 (WAD 10 Ye eo 0 1 vio __-3720 96 b-, 0 0 4 9 The optimal solution as given in Table 8.3 is: Rs. 1/10 perhour on machine I; v,=Rs. 6/10 per hour on machine I and Z,= total minimum cost, Rs. 96 The values ofy, and y, indicate that each unit of the two resources (available machine hours on machines and 11) contributes Rs. 1/10 and Rs. 6/10 per hour, respectively to the total profit of Rs. 96. Comparison of the solutions For interpreting the optimal solution of the primal (or dual), its solution values ean beread directly from the optimal simplex table of the dual (or primal). The method can be summarized in the following steps. 1, The slack variables in the primal correspond to the dual basic variables in the optimal solution and vice versa. For example, s,, is the slack variable of the fist primal constraint, it comesponds to the first dual variable »,. Likewise s,, corresponds to the second dual variable y,. Moreover. s,, and s,, are the dual surplus Variables and they correspond tox, and x,, respectively of the primal problem. The correspondence between primal and dual variables is summarized in Table 8.4. mete Mains |" 514) supe varies ae Sd “2p V2 2. The valve in the ¢, ~z, row under columns of the slack/surplus variables ignoring negative sign give directly the optimal values of the dual/primal basic variables. For example, c; ~z,~ 1/10 and ¢, —z,= 6/10Duality in Linear Programming FEZ values in Table 8.2 under s,, and s,, correspond to solution values of, and y’ in Table 8.3 and vice versa. ‘The primal-dual relationship for these problems which are feasible can be summarized in Table 8.5. Tables. Primal Dual + Values ofthe basic variables + (6-5) valuesiin the non-basic ‘surplus variable columns + (G-=)valucs in the non-bssiec | + Value of the basic variables slack variable columns 3. The values ofthe objective function are the same for primal and dual problems. 4, The dual variable y; represents the marginal value of a unit of resource i. For example, y, the marginal value of time on machine 1, equals Rs. 1/10 and y,, the marginal value of time on machine I, equals Rs. 6/10, as shown in Table 8 3, Example 8.10 Food A contains20 units of vitamin X and 40 units of vitamin Y per gram. Food B contains 30 units each of vitamin X and Y. The daily minimum human requizements of vitamin X and Y are 900 units and 1,200 units, respectively. How many grams of each type of food should be consumed soas to minimize the cost iffood A costs Re. 0.60 per gramand food B costs Re. 0.80 per gram? Mathematical formulation The primal and the dual LP problems of the given problemare: Primal Problem Dual Problem Xj» % = number of units (gm) of food A and B, Yyp = perunit price of vitamins Xand Y, respec- respectively to be purchased tively to be fixed for selling Max Z = 0.60x, + 0.80x, Min Z, = 900); + 1200, subject to 30x, + 20x, $300 subject to 30y, + Sy Sx, + 10x, £10 and > 0 and Solution of the primal problem Expressing the primal problem in the standard form by adding surplus variables s, and s, and artifical variables A, and A,, we have Minimize Z, = 0.60%, + 0.801, + 0s, + 0s) + MA, +MA, subject to the constraints (i) 20x, + 30r,-s, +A, =900 (ii) 40x, +302, and Xp Sy gy Ayy AQ 20. ‘The solution to this problem is shown in Tables 8.5 t0 8.7. Table8.6 Initial Solution 670m om rants Sommon [mm] inbasis—Ylues tal B bes) " 5 a " Ay 1200 wo 0 1 oo Z= 200m Sy | 0080 NOM — 60M _— 60M& Variables ho OH A, | MinRatio in Basis Xgl B M Ay °. @ -1 in 1 | 3001520 0.60 x 1 v4 0-140 | 30314) 40 ‘T= 300M+ 18 0 035-15 M O01S-M2 0 ‘Table8.8 Optimal Solution of Primal Problem go> 060 080 o 0 ‘Variables, Solution x h a 4 inBasis Values B boxy oR = 20 0 1-18 130 0.60 a 15 1 o v2 -120 0 0 B 1B Since all c,—z,2 Oin Table 8.6, the optimal soluti 5, 4,= 20 and Min Z,=Rs. 25 Solution of the dual problem Expressing the dual problem in the siandard form by adding slack variables 5 and s, we bave Maximize Z, = 900y, + 1200y, + 0s, + Os subject to the constraints (8) 20y, + 40), +5,=0.60 Gi) 30y, +30y, +5) and WFar $5420 ‘The solution to this problem is shown in the Tables 8.9to 8.11: Table 8.9 Initial Solution o> 90 1420 0 o ‘Variables Soluion | oS 2 | MinRaio inBais Vales Yala B oe ° 7 » @ 10 | sow=sa ° 4 0» ° 0 z=0 90012000 0 “Table8.10 Improved Solution 9 1420 0 ° p‘Reribies ” 4 4 | MinRaio in Basis Jo! B 1.200 % 12 1 1400 | 0015K1/2)=0.03 ° o @® 0 a4 1 | o3sms-o00 z-18 g- [30 0 30 0Table8.11 Optimal Solution of Dual Problem 6900 1.200 0 0 Variables Solaion | % 5 in Basis Values B bey) 1.200 v2 0.003 o 1 v2 ~ 150 900 v 0023, 1 ov os Z=25 5 o 0 35-20 a err 2 Sinceall ¢,~z, $0 in Table 89, the current solution is optimal solution an and MaxZ, “Rs. 25 given by: »,=023, y,=0.003 Comparison of the solution The absolute values ofc —2, in Table 89 showing the optimal solution of dual problem under columns s, ands, (slack variables) are 55 and 20. These values are the marginal worth or accounting price or shadow price of one unit of food A and B, respectively. This information of worth of each {ood will help the decision-maker to decide whether additional units of food should be purchased or not and at what price Example 8.11 The XYZ company has the option of producing two products during period of slack activity. For the next period, production has been scheduled so that the milling machine is free for 10 hours and skilled labour will have & hours of time available. Product Machine Time Skilled Latour Time Profit Contribution Per Unit Per Unit Per Unit A 4 2 5 B 2 2 3 Solve the primal and dual problem by the simplex method and bring out the fat that optimum solution of one ean be obtained from the other. Alsoexplain inthe cortextof the example what you understand by shadow prices or dual prices or original value of resources Solution The primal and dual LP problems of the given problem Primal Problem | Dual Problem x), = number of units of product A andB tobe | _y,,.7)= cost of one hour on machine and skilled produced, respectively. | labour, respectively. Maximize Z,=5¢,+3¥, | Minimize Z,= 10y, +89, subject to constraints (i) 4x,+2x,<10 | subject to constraints () 4y,+2y,25 (@ 2x, 42x58 G 2, +2y,23 and rym20, | ana 3yed220. Solution to primal problem The optimal solution tothe standard form of primal LP problems shown in Table 8.13. Max Z,= 5x, +3x, +05, +0s, subject to the constraints ( 4x, +2x)+5,=10 @ 2,42 +5, 0. and xi RRETable 8.12 Initial Solution 5? 5 3 0 o Cost ariales Solution |x, 4 * s ] MinRatio peruntt —nBasis’ Vales ah & B ery) 0 i 10 @® 2 1 0 ° 8 2 2 ° 1 = 5 3 0 0 ‘Table8.13 Optimal Solution “> 5 3 0 0 Cot ‘Variables Solaion | % An % per Unit inBasis Values on B boxy) 5 % 1 1 ° 12 2 3 * 3 ° 1 -12 1 z14 0 0 “1-1” ‘Table 12 indicates thatthe optimal solution is to produce, x, = 1 unit of product A;.x)~3 units of product B and Max (total profit) Z,=Rs. 14. Solution to the dual problem The optimal solution ofthe dual LP problem canbe obtained by applying, the Big-M method. The optimal solution is shown in Table 8.14. Min Z,= 10), +8), +NA, + MA, subject to the constraints © 4442-9 44,=5 GD 242%, and Ya Fn Sy Sod yydy20. ‘Table8.14 Optimal Solution ee) 8 o o Cost Variables Solution | y, 2 4 2 perUnit — inBasis Values on B 10 1 f ° “2A 8 % Ww o 1 w= Z=14 6-3 0 0 1 3 ‘The optimal solution in Table 8.14 i cost, and Min (total cost) Z, = Rs. 14 Re. 1 per hour cost on machine, and y, = Rs. 1/2per hourlabour Comparison of the Solutions (i) The cbject function value of both the problems is the same. (i) The numerical value of each of the variables in the optimal solution of the primal LP problem is equal tothe value ofits corresponding variable in the dual contsined in the ¢,~ z, row. Thus, in the primal LP problem: x, = 1 and whereas in the dual LP problem: s, =1 and s, = 3. (ii) The numerical value of each ofthe variables in the optimal solution ofthe dual LP problem is equal to the value of its corresponding variable in the primal, as contained in the ¢j~ =, row (only absolute values). However, artificial variables are not considered because they do not correspond to any variable in the primal, and are introduced only fora specific purpose.Duality in Linear Programming Fxg) Example 8.12 A company wishes to getat least 160 million ‘audience exposures’ the numberof times one of the advertisements is seen or heard by a person. Because of the nature of the product the company wants at least 60 million month and atleast 80 million of the exposures to involve persons between 18 and 40 years of age. The relevant information pertaining to the two adverising media under consideration—magazine and television is given below: Magazine Television ‘+ Cost per advertisement (Rs. thousand) ” 20 + Audience per advertisement (million) 4 0 + Audience per advertisement with monthly income ‘over RS. 10,000 (million) 3 0 ‘+ Audience (per advertisement) in the age group 18-40 million) 8 0 ‘The company wishes to determine the number of advertisements to be released each in magazine and television s0 asto keep the advertisement expenditure tothe minimum. Formulate this problem asa LP problem. What willbe the minimum expenditure and its allocation among the two media? Write “dual” ofthis problem. Solve the ‘dual’ problem to find answer to the problem, Solution The primal and dual LP problems of the given problem are: Primal Problem Duat Problem yy = number of advertisements in magazine 1.33 = shadow price (or worth) of one unit of and television, respectively. advertisement over audience characte- tistics, respectively. Minimize Z,= 40x, +200x, Maximize 2, = 160), + @y, + 80, subject to dv, + 40x, > 160 subjectto 4), + 3y,+8y, $40 3x, + 10x, 2 60 40y,, + 10y, + 10y, $ 200 8x, + 10,280 and Vip Van¥520. and xpXp20, ay Solutionto the primal problem The optimal solution tothe standard form of primal LP problem is showa in Table 8.16. MinZ,=40r, +200x,+0s, +05, +08, +MA, +MA,+ MA, subject to the constraints (@ 4x, +40x,-5,+4,=160 (ii) 3x, +10, A;=60 (ili) 8x, +10x~s,+4, and Zp Sp Spry Ay Aged 20. Table 8.15 Initial Solut 97 0 m 0 0 oO M M M Cost Variables Solution | SOSSCiSSCiCSSC Ag perunit in Basis Values os B v4 wf 4 @ 1 0 0 1 0 0 uw o4 | 3 0 0 1 0 0 1 wo 4 | 8 wm 60 0 4 0 0 4 7= mow ys |” 2 ow MM 0 0 0 ISM ~ 60MFEB] omeratons Rescaren: Problems and Solutions Table 8.16 Optimal Solution 9? ‘Cost! Variables Solution perusit —inBasis’ Vales 4 B bexp) 200 % 3 ° 55 30 40 a 10 000 In Table 8.16, since all ¢~z)2 0, the current solution is an optimal solution andis given by: MaxZ,,=1000, @ 2m 0 0 0 x * 5 2 5 0 1 -20s60 ° o 58 I ° 18 0 3210 0,x,=3and Solution to the dual problem The optimal solution o the standard form of dual LP problemis shown in Table 8.18. Max Z,= 1603), +603, +80), +05, +0s, subject to the constraints @ 4x, +39, +8), +5, =40 @ 40y, +10y, + 10y, +5 and PPI S520 ‘The initial solution and the optimum solution of this dual problem is shown in Table 8.17 and 8.18 respectively, Table 8.17 Initial Solution “eo 10 Cost Verabes i a S| Min Ratio pont inBasis min a 3 ° 7 3810] aaani0 ° E 1 1 01 | 2000-5 ° oy | 0 8 0 ~~0 ‘Table 8.18 Optimal Solution 62m mo Cost Variables Solution n dos 5) 3 perm —mBasis, Vales oa B by) 0 % 10 ° 1 +” 12 1m 160 * sn 10-98 im _380 1000 zs .. ww 0 3 In Table 8.18, since all c,—z, <0, the optimum solution of the dual problem is obtained and is given by: 5/2, ¥)= 10 and Max Z, Comparison of solution Since c,~ z, values (sbsolute) in s;-column and s,-column in Table 8.18 are 04—24= 10 and c,—c5=3, value of primal decision veriables is: x, (0, x, =3 and Min Z, = Max Z, = 1000. ‘The shadow prices (optimum values of the dual variables), are the coefficients of the slack variables in the optimum solution of the primal objective function with sign changedDuality in Linear Programming PE¥] ‘Duel variable noo oN ‘Slack variable noo SS ‘Shadow price 2 row coefficient s2 Example 8.13 XYZ manufacturing company operates a three-shift system at one of its plants. In a certain, section of the plant, the number of operators required on each of the three shifts is as follows: Ship Number of Operators Day (6am.t02p.m) 30 Afternoon (2 p.m. to 10 pam) a Night (10 pm. to 6 pm.) 10 ‘The company pays its operators at the basic rate of Rs. 10 per hour for those working on the day shif. For the aftemoon and night shifts, the rates are one and a half times the basic rateand twice the basic rate respectively. In agreement with each operator at the commencement of his employment, he is allocated to one of three schemes 4, B or C. These are as follows: A: Work (on average) one night shif, one afternoon shift, and two day shifts in every four shifts. B : Work (one average) equal number of day and aftemoon shifts. Cs Work day shifts only In schemes A and B, it is necessary to work strictly alternating sequences of specified shifts, as long as the correct proportion of shifts is worked in the long run (a) Formulate a linear programming model to obtain the required number of operators at minimum cost, (b) By solving the dual of the problem, determine how many operators must be employed under each of the three schemes. Does this result in over-provision of operators on any one of the three shifts? Solution ‘The primal and dual LP problems of the given problem are: Primal Problem ‘number of operators employed under scheme 4, B and C respectively yy Xpand x, Yoana, Dual Problem adow price (or worth) per unit of resources—operators in three shifts respectively. Minimize Z,= 20% 72, +15 Maximize Z,=10y,+24r, +50), i subject to the constraints = (i (12y,+ (12942572, Gi) _y,<10 Fat and Ypoton 520. subject to the constraints (i) (4)x2 10, (ii) (/4)n, + (1/2)xQ2 24 Gi) 2px, + (12)x, +5250 and papery 20. Solution to the dual problem The optimal solution to he standard form of the dual LP problem is shown inTable 8.20: Maximize Z, = 10), +24y, +50y, +05, +08, +0s, subject to the constraints © Wy, +0d)y, + 0/2), +5, =55/4, Gi) 95 +5)=10 MMos yS8yH20. @ (2, +0 Ay, +5, and onsea}ssEER cverstions Resear: Protiems and Selutens Table 819 Initial Solution 9 0 4 3 0 0 8 Cont Variables Solution nO 8 | MinRario peranit —inBasis Values es & Bo bey 0 % 35/4 “ow wt oo | ss2 0 4 292 o Mm wo 1 o| 2s 0 4 10 oo oo 1] ws Z,=0 0 24 5 0 oO 0 Table 8.20 OptimalSotution 6 0 M4 9 0 oe Cost ‘Variables Solution | >, nO j eH peranit in Basis. Values a 7 be. 0 a 20 1 ° ° 4 2 1 4 dy 1s ° 1 o o i 30 4 10 ° 0 1 0 6 1 Z,=0 ° 0 0-4 286 In Table 820, since all c,-z,<0,optimumsolution to the primal LP problem is obtained. Thus from Table 820, the optimum solution of primal LP problem is read as: . 1070. , =40, x, = 28, x5= 16 and Min (total cost) Z,= REVIEW QUESTIONS 1. Define the dual ofa inear programming problem. State the functional properties of duality. 2. Explain the primal-dual relationship. 3. What is duality? What isthe significance of dual variablesin a LP 4 3. ‘State the general ules for formulating a dual LP problem from its primal, How the concept of duality can be useful in managerial decision- making? ‘model? 6, ‘State and prove the relationship between the feasible solutions of LP problem and is dual ¥ SELF PRACTICE PROBLEMS Write the dual ofthe following primal LP problems 5. Min Z,= 2x, + 3x, +4r, 1. Max Z,=2x, + 5x; + 6%, subject to (i) Sx, +65—x,$3 (il) 2x +x, h4yS4 subjectto (i) 2x, +3x; + 5x,22 (ii) x $454 6x, <5 (i) 3x + y+ 7-3 (ii) S04 3y51 fiw) ~3x, 34, +e, <6 and ‘yt? 0; is unrestricted and xk H201 6 MinZ= xt 43 2, Min Z,= 7x, +32, + 8x, subjectto () xj—3y 44-5 Gi) m-2x, $3 subject to () By +254 sy E3 Gi) xj tO ADA 24-224 Gi) 4x, QF S21 iv) x, +5my 42527 and 0,2; is unrestricted. and Apt H 20. [Meerut Uni, MSe (Maths), 1994) 3. Max Z,~2xj +3xy 45, 1. One unit of product A contibutes Rs. 7 and requires’ units of raw subject to () 4x, #3y+a=6 Gi and ayxy 20. As Man Z;= Bayt 35m subject to ()) 285+ 3x, +,=1 Gi tity 3 peppy? O- xy tet Sra 4 ‘material and 2 hours of labour. One unit of product B contributes Rs, Sand requires one unitof raw material and one hour of labour. Availability of raw material at present is 48 units and there are 40 hours of labour. (@) Formulate this problem as a linear programming problem. (b) Write its dual,10, (©) Solve the dual by the simplex method and find the optimal product mix and the shadow prices of the raw material and labour A company makes three products: X,Y and Z out of three raw | ‘materials A, B and C. The raw material requirements are given velow: | Duality n Linear Programming [Ext 12. A manafacturing firm has discontinued production of a certain ‘unprofitable product line and this has created considerable excess production capacity. The management is considering to devote this ‘excess capacity to produce one or more of three products 1, and 3. The available excess capacity onthe machines which might limit ‘output, i summarized in the following table: Raw Materials Number of Units of Raw Material Required “Machine ype “Available Excess Capacity ta Produce One Unit of Product (Machine Hours per Week) x y Zz Milling machine 250 a 7 zi ; Lathe 150 ‘ i i i Grinder 50 c 2 5 1 “The number of machine-hours foreach unitof the respective product ‘The unit profit conribusion ofthe proguets: X, Y andZ. is Rs. 40, leevenbetow, 25 an 50, respectively. The number of uit ofrawmatra available ———=siSe ESS are 36, 60 and 45, respectively. Canesly Resa een acane boere U8) (a) Determine the product mix that will maximize the total prof Machine Tipe __Prochct! Product? Product 3 (©) Using the final simplex table, write the solution to the dual Vice 1 #¢. 3 problem and give its economic interpretation ae i $ 5 ‘Thre focd preduets are available at costs ofRs. 10, Rs. 36 and Rs, Gainder 3 ° i 24 per unit, respectively. They eontain 1,000, 4,000 and 2,000 caloris por unit, respectively and 200,900 and 500 protein wits por unit, respectively. It is required to find the minimum-cost diet containing at least 20,000 calories and 3,000 units of protein Formulate andsolve the given problem as an LP problem. Write the ) How many runita;of each, product should be produced to and C cost Rs. 6.00 and Rs. 7.50 per litre, respectively. aaticaize reverse! Anco Rs 60 nS TSO especie (©) Wetted roblem Raw Marra 19. Aperaneonsues two type fod A snd Bevery wo obsin a be Sanit prasing, [unis aTearbolydatesand9 unis of ats which ish clymininumrequiemens. igo fOodAcanalns2, and Combustion poi, °F _—~200~—~«180~—~—«280 T units of protein, carbohydrates and fats, respectively, Gama content % * 31 Tigoffood 8 contains and unison carohydatsand pau A Ss a ff fa epotely. Foo A cost R80 perk, while B cats Ras 4 even S| arg Determine how mary kgot each od shuld by diy to The procurement manager wishes 0 minimize the cost of raw __- minimize is costo feod an til meet the minum equrem=ns rates pr ie of product. Use linear programming to find the Formulate on LP problem mathematically. Wee it ual and Opti proprtons of ach raw material o ase ina lite ofthe gave the dl by he mpon meth sed proce. Also wrt the dal ofthe given problem and give {[Gijart Uni MBA, 1996) Westie eT acre nes 20. Afi prodaces thecal. ¥ 2a tol cont of. 4, Ra 3, tf aa aang aa peta i HG A ‘and Rs. 6 per item respectively. Total number of X and Z item = clay . ae 1 Se —e produced should be at least 2 and numberof Y and Z together be at indicate thatthe minimum dail adlt dosage for effective treatment Teast 5, The firm wantsto minimize the cos. Formulate this problem i 10 mig of drug X, Gig of drag S and 8 img of drug EL. To ‘as an LP problem. Write its dual. Solve the dual by the simplex substance ar ready nvallatefr preparing pls fr disibutin | Sea? Dablem. Wes onl Sove the du! hy te impor tocoldsufeers Bothaibstncesconainalthre ofthe required | eNO Ca¥ thugs Esch ni ofsabanee A conan 6g, Imgand mot 9, em drugs K, S andH, respectively and each unit of substance B contains 2 mg, 3 mg and 2 mg of the same drugs. Substance A costs RS. 3 per unit and substance B casts Rs. S per unt (@) Find the least-cost combination of the two substances that will yield a pill designed w contain the minimum daily recommended adult dosage. inaseortments of wo sizes I and I, The size I contains 20 bisouits of type A, 50 of type B and 10 of type C. The size Il contains 10 biscuits of type-4, 80 of type Band 60 of type C. A buyer intends to buy at least 120 biscuits of type 4, 740 of type B and 240 of type C. Determine the least number of packets be should buy. Write the dual LP problem and interpret your answer.22, Consider the following product mix problem: Let x, denote | ‘number of units of Product | to be produced daily and x, the | ‘number of units of Product 2 to be produced daily. ‘The production of Product 1 requires one hour of processing time in department Dj, Production of 1 unit of Product 2 requires 2 hours of processing time in department D, and one hour in, department D,, The number of hours available in department Dy are 32 hours and in department D,, 8 hours. The contribution of one unit of Product 1 is Rs. 200 and of Product 2 is, Rs, 300, ‘The solution tothis LP model is given below: Duality in Linear Programming FPxyg 420 30 0 Basie Solution] x) 8) Variables Values 200 y 2 ro2 1 Oo 0 S 8 o 1 0 1 Z= 6400 0 [-100 20006400 Given the dual of the primal model Obtain the opiimum solution to the dual LP model from the above ‘able, Interpret the dual variables. HINTS AND ANSWERS Min Z, = 39,44), +354 Oy subject io) Sy; ~ 2y,+ 5p) ~ 37422 (ii) 6y,+ y~ Sy, ~ 39425 | - y+ 4s +30 + 79426 and Heda ro%2o | MinZ, = 67,+ 4s, | subject o (49+ y22 Gi) 39, 420,23 | Gi) 821 | . Max Z,=2y,+ 395 $y | subjectio () 24 43y~ yS2H) 34+ 3y-4yS3 | Gi) 5),479,-6y=4 | ad yn) @and yy umesrted | Prinal tod x; = mimber of its of posts A snd B, respectively to be prodced Max 2, =x + 54 subject'o Gxt S48 i) 24, Hy S40 an sume 0 Aas 5 = 0,340, Max 2, Rs. 200 Prinal sf ad x,= un ofthe products X, Y and 2, respectively tbe produced Max Z,= 405, +254, +505, abject @ Ht yt 4596 (24+ Qtayseo Gi) 45m + S45 and 2154520. Ars 5 20,5 "0,3, = 5; Mas Z, = 1,050 2. Max Z,=3yj +4 +75 + 704 subject t0 (i) 80, +39, +405 + 947 Gi) 2y,+63* yy+5%4 3 Gi) y, +4, + 5), +245 8 and Yio daz 0. 4. Min Z,=9, +30 subject to (i) 2y,+9,2 3 Gi) —y+y,21 Gi) 3y,-y,23 Gv) yy ty 2-1 and -yjuip tnrestricted in ign. 6. Max Z,=~59) 3974 subject (i) —y)-95$1 (i) 3y42y,42y ST (iii) ~4y, “yy S1 and tps 20 and y is unrestricted. Dual {y) andy = worth of one unit of raw material and labour, respectively. Min Z,~48y, +40), subjectto (@) 3yj+2,27 i) yy + 2S and Yiesn2d Ans, y,=0,y, = $ and Min Z,= Rs, 200 Dual ys¥3 and y,= worth (or shadow price) per unit ofraw materials A, Band C, respectively. Min Z, = 367, + 60), + 4575 subject io (2) yj +29 429;240 Gi) y+ yy 594225 Gi) 9 +4, + 95250 and ary 20. Ans. y; ©0,', © 0,3, © 10: Min Z, = 1,05010. Primal 21,3, and x, ~ units ofthe products P, Q and R to be produced, respectively Max Z=3x, + 5x, 4%, subject to (i) 2x, +3x,<8 (ii) Sq +4x 515 (i) 5x,+2s,+2x,<10 and. ApApaye OL 11. (@) Primal 2x; and x2 ~ units of fresh food I and TI, to be bought, respectively Min Z=3x, + 2x, subject 9) 7x, +24, 230 Gi) 25, +842 16 and Aye 0. (i) 5x, +4220 12, Product | = Ounit; Product 2 = $0 units; Product and Max profit = Rs. 700, 13, Primal xptp tyandy, = $0 units antes of coffee seeds, respectively which the company buys from each plantation. Min = 35, 425, +123 Ley subject toi) (dx, +0.2x, +0.6r, + 08x, 2 03 i) 06, + 0.8, + 04x, + 02,2 0.7 and Nyy O. 14, Primal ‘xy-4pand.x5= number of valves of the types A, B and C. respectively to be produced. Max Z,= 18x, + 13.54, + 10, subject 9 () Sry + 3x 215 $1,200 Gi) I, + 2h 4y, < 500 (i) 4x, + 10x, + 5x, $1,300 and Hyp 20. Ans, x, = 22,4000217, 2 = 5.500162, x5 = 0 and Max 2, Rs, 8125/31 17. Primal Problem ie) = number of units of product A and B respectively to be produced Max 2, 51, * 3%, subject to) 4xy 42x $10 Gi) Ax, +2, <8 18, Primal Problem -Xy+49)43 = number oF units of product P, Qand R to be produced, respectively. Max Z,= 3x, + 5x +4x, subject 19) 2x) +3588 i) Sx, 44x, $15 and My yy 20- Ans. x)= 4441, = SIAL, val (iy 5x, 42%) 4 25,510) Dual Min Z=8y, + 10), + 15) subject © @) 2), +5923 (i) 245 25 Gi) +29 44,24 and Viv Var¥s 20. Duat _yy'zand yj wort per unit of vitamins A, Band C, respectively to the body. Max 230), +20, 16), subjectto 1H, +5y,+27553 il) 2, +49 +852 and Pp InIs20. () The optimal solution to the dual problem is: », = 20/52, = 0,73 = #52 and Max Z= Rs. 14 Dual and y, = worth of coffee seeds A and B, respectively Max Z, = 03y, + 0.7), subject 10 () O4y, +06y 53 Gi) 0.2y,+ 08y, 52 Gi) 0.6), + 04y512 Gv) 0.8, + 0.29, 51.5 and 220 Duat 33g and ys = work of Chambers 1,2 and 3 production cana respectively, Min Z,= 1.2003, + $00y, + 13009, subject to @) Sy tr 4215 3y, +29, + 1y,2 135 Gi) 29, +4y,+ 59,210 and Pan I520, Ans. y, = 0,9, = ABI, 9 =~ 649102. Dual Problem Yas Ys = worth (orshadow price) per unit of resource operators in teee shifts respectively. Min Z,= loy, +8) subjecto @) Ay +2228 Gi) 242,23 and Vyedg2 0. Ans, y, = 1,y,= 1/2 and Min Z, = 14 Dual Problem Jip Yan Ys = Worth, per unit of material m,, my and ms respectively. Min Z,= 8), +10), + 15%, subject to (i) 24, +323 Gil) 3y, + 2y, F4y, 24 and YoPn Wz. (i 22+ 5952520, 21 Primal Problem xy)" number of units of food A and B to be consumed, respectively 80x, + 43, ( 2xj+x,28 Gi) xj +3429) and pe 0. Ans. x, ~ xy" 6 and Min 2, 6525 Primal Problem Max 2, subject t0 (ii) 6x +212 | 2X). = mumber of units of articles X, ¥ and Z respectively. Max Z,= 4e, +32, + 61; subject to (i) x, +x422 Wi) tyes and ape as 20. Ans. = 0%, = 3% ‘with Min Z Primal Problem | yay" numberof astormens of size Land I respectively. | Manda | subject to (@) 20%, 10x2 120i) 50x, +80:,2740 | Git) 10x, + 60x, 2 240 | aod ayn20 | (Avs. = 2,4)" 8 and Min Z,= 10 1. Prima! Problem | Max Z, = 200r, + 300% sbjec to) 425832 i) Oxy ty £8 ad synee ‘The optimum selution to the dual LP problem is: y; ~ 200, "3 ~ 0s, Note that, y = marginal inerease in profit for an adi . Dally in Linear Programming ERY Dual Problem y= worth per unit of proteins, carbohydrates and fats respectively. Min Z, =8y, +125, + 9» subject to (i) 2y, +65 +); $ 8.50 Gil) yy +y5 43S 4 and My Par ¥520 Ans. y, =31/8, yz = 1/8,y =O and Max Z, = 6512; Dual Problem ‘worth per unit of resources to be used, respectively 4 + SY Ons4 Gi) ty, 6 Bide Min Z, subject to (i) yS3 Dual Problem YipYy's= Worth per unit of biscuits of type A, B and C, respectively, Min Z, ~ 120), + 14099 + 240 subject to) 20y, +50y, + 10) $1 (i) 10y, + 80y, + 609, <1 and Pp In ¥520 Dual Problem Min Z, =32y,+ 87, © 9 #072200 PaO subject to Gi) 29, +92 B 300 and ~0s,~ 100, 2° ~ Rs 600 ion I hour of capacity. Since y, = Rs. 200, profit becomes Rs. 6400+ 200= Rs, 6600, Also y) = marginal increase i proft given an addition hour of cepacity in D. Since y) =, there would be no change SL pti ty beentorm peter Bindicates that there isan ex {cess copacity of 8 hours in D,. Thus, in dual solution we have y, ~ 0. ‘unit profit ceived from manufacturing and selling ! unit of Product 1Sensitivity Analysis ‘Addition cf a New Variable (Column) ‘Addition of a New Constraint (Row) Solved Examples Review Questions Self Practice Problome Hints and Answers Sensitivity Analysis Discrete Change in Objective Function Coefficient (o) © Discrete Change in Availability of Resources (b) * Discrete Change in Coefficients (a,) IWTRODUCTION, = Inan LP model, the input data (also known as parameters) such es: (i) profit (cost) contribution (c) per unit of decision variables, (i) availability of resources (6), and (ii) consumption of resources per unit of decision ‘variables (a,), are assumed constant and known with certainty during s planning period. However, in real-world situations, these input parameter value may change over a petiod of ime duc to dynamic nature of the business environment: Such changes in any of these parameters may raise doubt on the validity ofthe optimal solution ofthe given LP model. Thus, a decision-maker in such situations would like to know how sersitive the ‘optimal solution is wo the changes in the original parameters value. ‘Sensitivity analysis and parametric tinear programming are the two teckniques that evaluete the relationship ‘between the optimal solution and changes in the LP model parameters. Sensitivity analysis considered the ‘effects of variations in the input coefficiens (also called parameters) when these coefficients are changed one ata time, whereas parametric analysis considered the effects of simultaneous changes in data, where the coefficients change as a function of one parameter. However, while sensitivity analysis provides the sensitive ranges (both lower and upper limits) within which the LP model parameters can vary without changing the optimality ofthe current optimal solution; parametric linear programming provides information to such changes outside the sensitive range, as well asto changes in ‘more than one parameter at atime,FLem Operations Hesearch: Problems and Solutions = SENSITIVITY ANALYSIS ‘Sensitivity analysis isthe study of sensitivity of the optimal solution of an LP problem due to discrete variations (changes) ints parameters. The degree of sensitivity of the solution due to these variations can range from no change at all to a substantial change in the optimal solution of the given LP problem. Thus, in sensitivity analysis, we determine the range over which the LP model parameters can change without affecting the current optimal solution. For this, instead of resolving the entire problem as anew problem with new parameters, we ‘may consider the original optimal solution as an initial solution for the purpose of knowing the ranges, both lower and upper, within which a parameter may assume a value. The process of studying the sensitivity of the optimal solution of an LP problem is also called post- optimality analysis because itis done after an optimal solution, assuming a given set of parameters, has been obtained forthe mode! Different categorics of parameter changes in the original LP model discussed in this chapter include: @ Profit (o cost) per unit (c) associated with both basic and non-basic decision variables (coefficients in the objective function). (i) Availability of resources (right-hand side of constants, b). Gi) Consumption of resources per unit of decision variables x, (coefficients of decision variables on the left-hand side of constraints, a) (iv) Addition of a new variable to the existing list of variables in LP problem. (¥) Addition of « new constraint to the original LP problem constrainis DISCRETE CHANGE IN OBJECTIVE FUNCTION COEFFICIENT (c) Consider the LP problem: Maximize Z=ex subjectto — Ax=bandx>0 Let the coefficient cy ofa variable. is changed from ¢;t0 ¢, + Acy. Since the condition of optimality requires that¢;~z,$0 forall jin case of maximization LP problem, the effect of such a change will beseen inthe c,—z) row of the optimal simpletable. Two cases may arise (@ cis the coefficient ofa non-basic variable x, G_cyisthe coefficient of a basic variable-xp In case of (i)to retain the optimality of current optimal solution we must have 4-4 = (Gt Ae) — 450.12. ot Ay Sz {In case of (i) to retain optimality of the current optimal solution we must have G4) Gay, +6)-250 ‘Thus the range of variation Ac, which satisfies the optimality condition can be determined by solving the following system of linear inequalities: Min! F be ney = Mae] dy <0 vy >O DISCRETE CHANGE IN AVAILABILITY OF RESOURCES (0) ‘The range of variation Ad, can also be obiained by solving the system of linear inequalities (condition of feasibility ofthe current optimal solution), x4=B- b> 0. where = matrix of coefficients corresponding to slack variables in the optimal simplex table amount of change in the resource & basic variables appearing in B-column of simplex table.sensviy Ania BEY ‘The range of Ad, so that the current optimal solution remain feasible is given by =n =m Max 4h, s Min| 221 ies} ‘ ital Remark If someentries in the ‘x,’ column of the simplex table are negative, then the dual simplex method can be used to get an optimal solution to the new problem by maintaining feasibility. DISCRETE CHANGE IN COEFFICIENTS (a,) ‘Suppose that the elements of coefficient matrix A are changed. Then two cases arise: Change in acoefficient, when variable is a basic variable, and i Changein a coefficient, when variable is a non-basic variable, Case 1 When a non-basic column a, € B changed to aj, the only effect of such change will be on the optimality condition. Thus te solution will remain optimal, if C- 2 =q— ty Bay <0 otherwise the simplex method is continued, after column k of the simplex table is updated, by introducing the ‘on-base variable x; into the basis. However, the range for the discrete change Aa, in the coefficient of non-basic variable x, in the constraint, ‘ican be determined by solving following linear inequalities sinc S| cy sof | egB,>0| 7 3B) <0 Here B, isthe ih column B™'. If¢y B, then Aa, is unrestricted in sign. Alternative Method The change in the coefficients (a,’s) values associated with non-basic variables in the ‘optimal simplex table can be analysed by forming a corresponding dual constraint from the original set of ‘constraints: & a, y,2¢,3 forx,non-basie variable. =I ‘The values of dual variable y;’s can be oblained from the optimal simplex table. The reason behind this dual ‘constraint formulation is that an activity is considered as fully undertaken provided all the marginal values (shadow prices) ofits resources become equal to its per unit contribution to total profit. Case 2 Suppose a basic variable column a, € B is changed to aj. Then conditions to mai feasibility and optimality of the current optimal solution are: (@ Max eicctaes SA; < n both te PN Fee [Son Bp —*ap Bas > £20 [Say By ip Bur <0 (b) Max {++ _______.< Ag, ¢ Min } ———_+_4_____ “ ce aaErl : i= Wn =i =| ADDITION OF A NEW VARIABLE (COLUMN) Let anextra variable x, , , with coefficient c, . , be added in the system of original constraint, Ax = B, x > 0. ‘Thus, it creates an extra column a,,,, in the matrix A of coefficients. To see the impact of this addition on the current optimal solution, we compute and Cy 41 Zn = Ene 1 SB Yn |FEES Operations Research: Problems and Solutions S ‘Two cases of maximization LP model may arise: LIfe, 12) 41 $0, then xp =O, and hence current solution remains optimal 2 Uf¢y4 1 ~Z,-+1 > 0 then the current optimal solution can be improved by introducing a new column a,,. into the basis to find the new optimal solution. ADDITION OF A NEW CONSTRAINT (ROW) Suppose that a new constraint Bet iT tent. 2e tte b is added to the system of original constraints Ax=b, x>0,where b, is positive, zero or negative. Then, two cases may arise: 1. The optimal solution (xg) of the original problem satisfies the new constraint, If it is $0, the solution remains feasible as well as optimal, because the new constraint either reduces or leaves the feasible region of the given LP problem unchanged. 2 The optimal solution (xp) of the original problem does not satisfy the new constraint, Then the optimal solution to the modified LP problemis re-obtained, Let B be the basis matrix forthe original problem andB, be the basis matrix for the new problem with m + 1 constraints. That is, matrix B, of order (m+ 1) is given by BO Birla 1 where the second column of B, corresponds to slack, surplus or artificial vaviables added to the new constraint and = (ay 4 15 @p 41,25 “+ %y¢ |) 188 FOW vector containing the coefficients in the new constraint and corresponds to variables in the optimal basis. In order to prove that the new solution pte *e|s is a basic feasible solution to the new LP problem, we shall compute the inverse of B, using partitioned methods as given below: ee[Se Te see[Ss hhh] where b* is the resource vector in the new LP problem. If x}, 2 0, the current solution remains basic feasible for the new LP problem, otherwise apply dual simplex method to obtain the basic feasible solution. x, in on Ba S Doe slack variable ‘SOLVED EXAMPLES Example 9.1 Soive the following LP problem: Maximize Z=3x, +5x, subject to the constraints © 3x, +2x,<18; x54; Gi) <6 and p20. (@ Determine an optimal solution te the LP problem. (b) Discuss the change in con the optimality of the optimal basic feasible solution. Solution (a) The optimal solution of the given LP problem is shown in Table 9.1. From Table 9. 1, the optimal solution can be read as: x, = 2, 5 =6 and Mex Z= 36. (0) Here c= (cy, ¢q,€y,Cy.€), and the cost coefficients associated with basic variables x,,.x,, and s, are 57 (1+ €2:€4) = @, 5,0). The changes in c, can be classified as under:Sonsivty Anavsis BEY (0 Changes in the Coefficients ¢ (ue. ¢5 and cs) of Non-basic Variables 5, ands Let us add Ac, and Ac; to the objective function coefficients c, and cs. Then the new objective function coefficients will become c= c,+ Ac, and 5 =¢,+ Ac, Since ¢,=0 and c,=0, c= Ae, and cf = Ac, Thus, the new values of cy ~z, and c, ~z, will become Ac, - 1 and Ac. —3, respectively. Now in order to maintain optimality we must have ‘Ac-1SVand Acs~3S0(maximizationcase) or Ac,S1and Acc $3 Gi) Change in the coefficients cj (i.e.cy, ¢ and cy) of basic variables xy, and sy ‘The value of additional increment Ac, k= 1,2, 4in the coefficients c,, c, and c, which satisfy the optimality condition can be determined by solving the following system of linear inequalities [refer to (1). Hi e-% re ES Table9.1 Optimal Solution Min gn 3 5 6 o o “pMariables ‘Solution % 4 4% % %5 inBasis Values B Bex) 3 % 2 1 ° 18 a 23 ° s o ° ° 2a H 43 5 4 6 0 1 0 a i Z=36 0 o = @ 3 Fork=1 (ie. basicvariable,, inrow 1), we have 3 = i 2Aqz Max); j7=3,5 or 9/22Ac,2~ nonf st teansranl lh goss oe setae Here it may be noted that yy, = (7,3 = 1/3, 45 =~ 2/3) are only for those columns corresponding to which ‘variables are not inthe optimal basis (i.e. non-basie variables). The current optimal solution will not change as Jong as 1s (343) 2e,2G-3) or [2e,20 Fork=3 (ie. basic variable x inrow 3), we have we fabsaasl Hence, the current solution will not change aslong as: (5 +9) 2c, 2(5~3) oree2 c 22. Example 9.2 Discuss the effect of discrete change in the resources values to the LP problem discussed in Example 95. Solution: From Table 9.1, we have V3 0 2/3 B=(,5).42); X=(2,0,6); b= (18,4,6) and B'=(s,,53.8,)=]-2/3 1 4/3 oo 1 ‘The range of variation in 5, (i = 1, 2, 3) can be obtained by solving the system of linear inequalities representing condition of feasibility of the current optimal solution, i. y= Br! b > 0.eos sy % re 7S (@) Variation in by: Let Ab, be an addition in the resource b,. Then by making use of the condition that the current solution remain fensibl as ong as 13 0 -2/3][18+4] fo -2/3 1° 473} 4 |>|o| fie. B'b>o] oo als | fe 2 Fas+any4@- (6120 or ~65 Ab, 3 2 4 ~F (8+ ab) +4()+5 (620 oF Ab,s0 Since there is no upper bound, the optimal solution will not change as long as ~65Ab,S0 of (I8-6) 0 gives 13 0 -2/3]f 18 ] [o -2/3 1 4/3|| 4 elo 00 ile+aa} Lo 4 ays0ce-2¢6vany 20 or ays 4 = 3 (18)+14)+5 (6 +Aby) 20 or 4,20 0(18)+04)+1(0+ Ab.) 20 or Ab,2-6 From these three inequalities, we have ~65Ab,<3 or (6~6) o 1 2 = 18 13 Z=16 4 6 8 108 2B 6 0 6 = 103 = The optimal solution is: x;~ 1.x) ~2 and Max Z~ Rs. 16. (b) Effect of Changein the Coefficient ey of Non-basic Variable x, for Product C: Inoptimal simplex Table 9.2, the variable x, is non-basic and its coefficient c, =2 is not listed in the ¢, column of the table. This means a further decrease in its profit contribution c, (= Rs. 2) perunitwill have no affect on the current optimal product ‘mix. But, if is increased beyond a certain value, the product may become profitable toproduce. Hence, there isonlyan upperlimit onc, for which the current optimal productmix willbe affected Asie know, a change Ac; in c, will cause a change in 2, and c xy. The change in c, results in the modified simplex Table 9. Table93 Table witha iy Changein, the column corresponding to gt 4 6 2h dey o o cq Varlables Solution 7 ¥i % % 4 inBasis. Values B bay) 4 x 1 1 0 “1 43 -18 6 % 2 o 1 2 13 13 z-16 + 6 8 103) 23 4 0 dey 6 103, 23 For an optimal solution shown in Table 9.3 to remain unchanged, we must have Ac, — 6 0 or ‘Acy $6. Recalling that c, = 2 + Ac,, of Ac, = ¢ ~ 2, we substitute this emount in above inequality c, 26 ore, £8. This implies that as long as the profit contribution per unit of product C is less than Rs. 8 (ic. change should not be more then Rs. 8) it is not profitable to produce it and the current optimal solution will remain unchanged. (0) If the Value of eis Increased from Rs. 2 to Rs. 10, the New Value of =I 25 = (cy— yay) will be Thus, if coefficient of variable x, is increased from Rs. 3 to Rs. 10, the value of e, ~ z, will become positive as shown in Table 9.4. The variable x, (corresponding to product C) becomes eligible to enter into the basis Hence, the solution cannot remain optimal any more.Variables Solution -* 2 % ‘Min Ratio inBasis Values ay B bes) 4 % 1 1 0 -1 & 1B = 6 % 2 o 1 @ -~ 2 3 4 6 8 1s 2 ony oo 2-03-28 Applying the following row operations to enter variable x, into the new solution-mix Ry new) >R, (old) + 2 (Rey elementy; Ry (new) —»R, (old) + Ry (new) ‘The new optimal solution after entering non-basic variable x; into the solution is shown in Table 9.5. Table os 67 4 6 10 a a Variables — Solution x % 5 x inBasis Values B Bex) 4 x 2 1 2 0 6 16 10 5 1 ° 2 1 6 16 Z-18 4 7 10 3 1 6-3 0 -1 0 -3 1 Since all ¢ ~2, $0 in Table 9.5, the new optimal solution obtained is: x, = 2, x, MaxZ=Rs. I (@) Effect of Change in the Coefficient c, of Basic Variable x, for Product A: Inthe optimal simplex Table 9.2, the variable-x, appears in column B. This means a further decrease in its profit contribution, (= Rs.4) will make it les profitable to produce and therefore the current optimal product mix will be affected. Also, an increase in the value of c, beyond a limit will make the product A much more profitable and may force the decision-maker to decide to produce product A only. Thus, in either case, the current optimal product mix will be affected and hence, we need toknow both the lower as well as upper limit on the value of ¢, within which the optimal solution willnot be affected. Referring again to Table 9.2, the range of change in the value ofc, (andior inc, also) which does not affect the current cptimal product mix can be determined once again by calculating the values of cj ~ 3 values corresponding to non-basic variables x,,., and s, respectively. For this reproduce Table 9.2 once again with unknown value of ¢, as shown in Table9.6. ‘Table9.6 Table witha Ac, Change in c, and x = 1 and 97 4446) 6 2 0 o ep Variables. Solution x % Ss % inBasis ‘Values B bexp ara, 1 1 ° “1 48 V3 6 4 2 0 I <8 B Z=12+4 +e) 44k, 6 B-Aq — 103+Ae/3 23—Ae\/3 0 0 Ae-6 — ~103-Ac\/3_Ac\3-23Soneivty anaysis [ERB For the solution shown in Teble 9.6 to remain optimal we must haveall ¢/~. SO givesAc, 6 s ~F SO wives dey 2-F <0 gives Ac, $2 33 ‘Thus, the range of values within which c, may change without affecting the current optimal solution From these calculations, it may be concluded that in the case of competitive pressure or in the absence of | any competition, the decision-maker krows how much prices can be adjusted without changing the optimal product mix (©) Fffect of Change in the Coefficient of Basic Variable x, for Product A: \f the profit contribution ¢, from product A is taken as Rs. 9 instead of Rs. 4, then putting c, ~9 in Table 9.6, we have 54 4 Weqreg=2—¥ e172 = 10 These entries are shown in Table 9.7 Table 97 67 9 6 2 0 0 “p——Variables—‘Soluion Ho % Min Ratio inBasis Values xa 8 9 my 7 1 0-1 4-1 = 6 4 2 0 1 2-3 @ 2/3) 721 > 6 3 Ww -) oo 1-0 1 ‘The solution shown in Table 9.7 is no longer optimal because c,~z,> 0. Entering slack variable , into the solution by epplying the following row operations, Ry (new) Ry (old) *3; Ry (new), (old) *(1/3) Re(new) ‘Thenew solution is shown in Table 9.8, Table98, 6 9 6 2 0 0 ¢g ‘Variables. Solution x % % 3) 53 inBasis Values B b(=x_) 9 % 3 T T T T 0 0 & 6 0 3 6 -1 1 Za 9 9 9 ° 0 o 4a 7 9 o Since all ¢ ~z)<0, the solution is optimal: x, ~3,1)~0 and Max Z~Rs.27, (6) Effect of Change in Coefficient of Basic as well ax Non-basic Variables in the Objective Function: Reproduce Table 9.3 as shown in Table 99. Since the objective function isnow changedto Z= 3x, + 8, +Sx,the entries inthe ¢, column of Table 9,2 will be (3, 8) instead of 6, 4). To check the optimality ofthe solution as shown, in Table 9.3 with the new values of c,, we determine. ~ =, values once again as shown in Table 9.9 2s usual onset?py ‘Varables Solution A si i 5 % inBasis Values B bexy) 3 % 7 0 =1 eB 18 8 % 2 ° 1 2 “13 1B 3 3 8 3 4a 53 9-3 0 0 -8 -43 -53 Since all c, - z, $ 0 in Table 9.9, the optimal solution does not change. The optimal solution is: ,,=2and Max Z= 19. a Example9.4 Solverhe followingLP problem Maximize Z=4r, +6x,+2x, subject to the constraints xt Ht 4 s3 Gi) xp +4 tI <9 and XpXyX 20. (@)_ Discuss the effect of discrete change in the availability of resources from [3, 9]" to (9, 61". (b) Which resource should be increased (or decreased) to get best marginal increase in the value of the “objective function? Solution The given LP problem in its standard form can be stated as follows: Maximise 24x, + 61,+21 +0s, +05 ‘subject to the constraints O xt yt Ht (x, tart Tey t= 9 and yp Xyp5198g20. Applying the simplex method, the optimal solution so obtained is shown in Table 9.10. Table9.10 Optimal Solution 97 4 6 2 o 0 ‘Variables. ‘Solution % % 5 a iy inBasis ‘Values B exp x 1 1 o - “3 -13 6 2 2 ° 1 2 =" 1B Z=16 % 4 6 8 103 2B 6-4 ° 0 -6 103 -2 t (a) Ifnew values of right-hand side constants in the constraints are [ 9, 6 ]’, then the new values of the basic variables (xg=B™ b)shown in Table 9.10 will become [hE ele) «3 10,andx,=—1ae Since the value of x is negative, the optimal solution shown in Table 9. will become infeasible. To remove infeasibility we use the dual simplex method. Reproducing Table 9.10 with new values, as shown in Table 9.11. ‘Table9.11 “7 ‘ ‘ 2 5 0 ey Verabia —Souton _ A HR inBasis Values, Bb . % 10 1 . 1 WB 6 on “1 0 1 2 aD z= a ¢ . 0328 ° a ws -28 Since:x,=~I, the second row is the key row and x; is the outgoing variable. For identifying the key column find the following ratios -103 vin ya <9 = {22} -r0(emns Hence, column ‘s,” isthe key column and variable s, will enter into the basis. Revise the simplex table using following row operations: Ry (new) —> Ry (old) x3; Ry (new) > R, (old) (43) R, (new) ‘The new solution is shown in Table 9.12. ‘Table9.12 Optimal Solution 67 4 ‘ 2 ° 0 Variables Solution x x % sy fs inBasis ‘Values B 4 x 6 I 4 7 ° 1 0 sy) 3 0 -3 -6 1 I Z=24 % 4 16 28 ° 4 -% 0 -10 26 ° 4 As all c, ~ z, $ 0 and all 6, > 0, the solution given in Table 9.12 is optimal. The optimal solution is: 0,5 = 0 and Max Z= 24. In Table §.12, slack variable 5, i presentat positive level in thesolution mix. Thus, the sensitivity limits for the corresponding right-hand side, ie. first constraint, are determined as follows: Lowerlimit = 9-3=6; Upperlimit=<» (b) In Table 9.11, ¢,—24= 10/3 and c,~z,= 2/3 values (ignoring negative sign) corresponding tos, ands, ‘columns represent shadow prices of resource 1 and 2, respectively. Thus, increasing the amount of resources, 1 ard 2 will inerease the value of objective function Z by Rs. 10/3 and Rs. 2/3 respectively. Now in orderto know how much these resources may be increased so that each additional unit continues to increase the objective function by 10/3 and 2/3, respectively, til the optimal solution remains feasible, we proceed as follows, Let Ab, be an increase in first resource (RHS of first constraint) so that 4 Xp=B p20 x] _f #3 13 7[3+45,]_[i+aa)a0, or => L-a3 wll 9 J>[2-aaae, : Sonstivity Analysis Pryie. 1+(4/3)Ab,200rAb, 2-3/4 and 2-(1/3)Ab,200rAb, <3 From these inequalities, we have-3/4SAb)<3° or 3-(3/4) SB ‘Solution em Values bexp) ws ° 1-15 WS 1 ° ws 2s us 2s 2s 2/5 =29)5 -M+2I5 (@) fnew values of right hand side constants in the constraints are changed from [5,2]"to [7, 2]", then thenew values of the basic variables (xp =. B-'b) shown in Table 9.14 will becomeimage not availableimage not availableees Sensitity Anais EEE The variable x» in Table 9.17 isa basic variable, the solution will remain optimal long as non-basie variables remain non-positive, 3 1@,c,,0)|0)0 or <0 =,c,,0)| 1 ]<0 oF cy2-1, =I 0-(0,c,0)| 1]/<0 or 4 0 ¢,=-M-(,e,,0)| 0]<0 or M20 4 ‘The variable.x, in Table 9.17 isa non-basic variable, and to find its upper limit, we must have c; $0, ie. 2 €5-(,2,0)| 1)$0 ore,-2<00re,s2. value is obtained from largest value of == 1(cy2,0) |0|s0 or -1~ 3c, <0 or c,2- 1 2 = &y 2,0) 1 |s0 or -1+2c,-250 or es > 3 2 0-0 (cy, 2,0)| 1/50 or 0+ c,-250 oF cy s2 3. 5 aL Hence, limits for ¢, are: ~ 1/3 $c, § 3/2 ‘The variable s, in Table 9.17 is a basic variable and its limiting value is obtained from largest value of Gs 2e and Z as follows: 3 ~ (2,3) |0|s0 or -1~cg 1 <0 or es <0 or—1-(2+3e,) £0 or cg 2~ =. 2.¢5)fcc eee “1 1/50 or 0-@+4es)S0 oF e525, 4 % = 0-(0,2,¢5) 0 0]s0 or -M+cs<0 or cg M. -l % =-M-O,2¢9 Hence, limits for eg are:—1/2Seg or eg $~ 1/2. ‘The variable #5 in Teble 9.17 is non-basie variable and to find its lis iting value <0, we have -1 1]s0 or «250 oF cys 2. 4 Example9.7 Consider the following LP problem Maximize Z = 3x, + Sr, + 4x3, subject to the constraints @ 2x, +358, and yt 20. ‘The optimal solution is givenin Table9.18. 4 ~ (0,2, 0) (i) 2x + 553510, Gil) 3,42 +4xy S15 Table9.18 Optimal Solution os 3S 4 0 0 0 Varicbies — Soltion | x SSS inBasis Values: zB b exp) 5 ay sot o 4 O15) wat 101 4 xy 24 o 0 1-641 sist 4d 3 x 2941 10 0-24) -1241 15/41 308 4 45/4) 2841/41 o 0 0 ~45/4) —24/41 1141 (@), How much cand c, can be increased till the optimal solution given in Table 9.18 remains optimal? Also find the new value of the objective function if possible, (b) Find the range over which b, can be changed maintaining the feasibility of the solution [Delhi Univ. B Se (Math) 1995] Solution (a) i) Variation in the Coefficient c, of Basic Variable x, : The value of c, can only be changed 50 Jong its upper limiting value © (j=4, 5,6) remains non-positive. is/4i so 6 = (5, ey. 3) Jso -siai|s0 or [BSS aaa -2/41 8/41 sai fso or ~[ ~12/AL a-H]s0 ore2- 4 = 6,643) aI 10/41 wai Js0_ or -[ ssa 30 ~ Gey 3)Hence, limits fore, are: 5!4 $c, $23/2594,5/45 < Max Z.< 1230/41 (i) Variation in Coefficient c, of Non-basic Variable s,: Value of s, can only be increased so long as remains non-positive, ie 15/41 4-6, 4,3) 6/4150. or cy— 2/41 z 24 ‘Js 45 £) Sc, <45/41 and hence max Z= 765/41 (b) The new values of the current basic variables when value of resource b, changes from 10 to 10 +A 6), (say) ate given by: 138 10 Teal 2 Petal x {-rte=|-£ 5 4 lhoras elo a4 x 2 2 ft B JL 44) 20, 80 Bag, 9 30 or ane, aa ai a 48 50,5 60 Tr i 6 my, naa S30 or Abe al + 3 20 ot ab, 4 <2 12 From above inequalities, we have-25/4 < Ab, <89/12 or 15/4 1 ° Ss 10 1 0 3 ° IL 44 Zs 5 a 2 2 o . 2 0 -1 6 o3 ° > 2 — [Moora Unit, MSe (Maths), 1093) Solution (i) Variation in by: Let b, be 10 + A by. To find the range of variation in b,, we make use of the condition that the current optimal solution remains feasible so long as 5 1 0 -1fto+aH) fo a 0 0 ti] 6 alo). o 1 ail 4 0 Thisyiclds, 10+ 4 b,-420, iz. A b, 2-6 or 6A b,. Since there is no upper bound on A by, 6S. b, So or 10~65d,510+% oF 456, S(Gi) Variation in by: Let by be 6 + A by. Then we have 5 10 -If 10 0 my |=B'b=|0 0 1//6+46,|2}0}. 4 o 1 ail 4 0 This yields, 6 + & b, + 1620, ie. 4 b, $10. Since there isto lower bound to A by,—e S.A by $10 or--0+ 6S 4b 510+60r-- Aay,2—169 or a,,23-169, 1.0.4, 2-163 oF dyy> Br biea,2- 149 oF 44324-1940. 0,215; 19 4 OF ay32 1 Fy dys 2 <2 Shay, S@, ie. © S033 S00, Case2 a, isinthe basis B. Since x,, x, and s, are in the basis, therefore any discrete change in the coefficient a, to any of these basic variables (or vectors) may affect the feasibility as wellas the optimality ofthe original optimum basic feasible solution Xp. Consider the diserete changes in ay corresponding to x, ~ By. The range for the discrete change in the element a,, in order to maintain the feasibility ofthe solution is given by 16,8 5 1 7 126 is 3 rorba-toatr= Beh -3(-2) =F ses Ba in Bae Bx S-S(-3) 3 ‘The range for the diserete change in the element dq in order to maintain the feasibility of the solution is given by. 1619 -126/19 Max\a3a 32 | S42 oF sca, ae emg acento Similarly, the ranges for discrete changes in the elements aq, ayy, ete., can also obtained.ea Example 9.12 Find theeffect ofthe following changes in the optimal simplex Table 9.24 of the following LP problem. Maximize Z=45x, + 100x,+30x, +50x, subject to the constraint (Tey # 10x, + 4x, +9x421200, (i) 3x, + 40x, +15 +4800, and XypXpp yy F420. Table9.24 Optimal Solution “8m 0 ton Horas Soon | 3S i we a & uo 7a as wo os 010-130-150 Za2803 = | o30 10090 08 ts é 2900888 tas @ ‘x'-column in the problem changes from [7, 3]' to [7, 5]". (b) ‘xj-column changes from (7,3]" to [5,8]! Solution (a) The variable r, is a non-basic variable in the optimal solution shown in Table 9.24. The upper limit & the coefficient of x, is calculated as follows: B =e, tqa, =a -egh tay =C4—dy- where ¢, as and =¢98-'=(20, 10)| m | 5]: |-nso a7 J" [373 — 4g _[22 2] [7 15410) _ Thus 3 -5-[2.2][{]--(-4)- Since value of & is negative, the current optimal solution remains optimal for the new problem also. cg Ba, =¢,—2a, -4s-[2 2][;]-«5-(2 ¥)- 8 a3 Since value of @; is positive, the current optimal solution can be improved. Also ra [3 *TE)-[ ais ] so 275]L8]~ [2750. ‘Table9.25 Optimal Selution ©) Ge e\-ep m= Pe ee ee ee) Varabies Sohn | mH | Mn Rai imBass——“Vedues ae os B bx) oO wos | as 01 7a wis ans | 0003 wo oy 403 10 __~1n0 ~1s0_275 | 200027 = 280008 1300S HOTS S078 300-883-328Eee comtcions Research: Problems and Solutions. Introducing non-basic variable x, into the basis to replace basic variable x, using suitable row operations: In Table 9.25, all ~z $0, the solution is optimal with: x =2000/27,x,=0,x, = 460027 and Max Z= 86,000/9, ‘Table9.26 Optimal Solution 5? 43 10 30 o 0 9 Variables Solution | inBasis Values oe B beexy 30 a soo2r | 0 49 1 827-527 45 x 200027 109 0 27427 45 3500305959 6591009 0-503 0 1459 4599-1009 Example 9.13 Discuss the effect on optimality by adding anew variable to the following LP problem with column coefficients (3, 3, 3)” and coefficient, 5 in the objective function. MinimizeZ=3x, +8, ‘subject to the constraints (@) x42 =200, (ii) xy and xy. 20. Solution The given LP problem in its standard form can be expressed as: MinimizeZ=3x, + 8+ 0s, + 0s) +MA, + MA; subject to the constraints @ xy+x,+A,= 200, @ x,+5,=80, i) -s, +A, and XpXp 5p) Ay Ap 20. Applying the simplex method, the optimal solution is shown in Table 9.27. The optimal solution is: x, =80,1,= 120 and Min Z= 1.200. Table9.27 Optimal Solution : 97 3 8 0 0 Mw op Variables Solution x a % A, inBasis Values z be xy) 0 » 60 0 0 1 1 1 3 a 80 1 0 1 0 0 8 5 120 0 1 at 0 1 Z= 1,200 3 8 -s 0 8 9 0 ° 5 ° Mog To ses the changes in the optimal slutioa, we are given that c, = 5,column, a, =(3, 3,3). Thus, o see the change, we calculate =1o1 ys rays) fs ¢)-2,=¢~tg B! a,=5-(0,3,8)| 1 0 0|/3|=-4,anda;=B o olfs|=|3 -10 a3 o ais] Lo Since c, ~2, = ~4, is negative, the existing optimal solution can be improved. For this, we start with the optimal solution shown in Table 9.27 and adding entries corresponding to variable x, as follows.‘Table9.28 New Column with Variable x, Added 9? 308 0 0 MS & Variables Solution we Ay | MinRatio inBosis Values py Bo beay ° % o oo «a a 4 3 xi 80 1 0 1 0 0 8 x 120 o 1 + 0 4 2= 1200 30 8 5 0 8 ° 50 M8 ‘As shown in Table 9.28, variable x, must enter into the solution and s, should leave it. The new solution is shown in Table 9.29. Table 9.29 ga 8 o 0 Mos ee Variables Solution | x, 8A in Basis Values a esp) 4 20 « ° a 3 x 20 1 0 2-1-1 0 8 » 120 ° 1 “10 ro Z= 1,120 3 8-18 43 48S o 0 3,43 Mean 0 Since all ¢,~z,20 in Table 9.29, the solution is optimal with x, = 20, 20, x, = 20and Min Z= 1,120. Example9.14 Consider the following LP problem Maximize 2=3n, +5¢, subject to the constraints @ 3x,+2y<18, (i) x, +283 and 1,320. ‘Obtain on optimal solution ofthe given LP problem. (Suppose variable x, is added to the given LP problem. Then obtain an optimal solution tothe resulting LP problem. It is given that the coefficients of x, in the constraint of the problem are 1, 1 and 1, and its ‘coefficient in the objective function is 2. (©) Discuss the effect on the optimal basic feasible solution by adding a new constraint 2x, +x, 810 the given set of constraints Solution The optimal solution: x)= 2, x, = 6 with Mex Z= 36 to the given LP problem is shown in Table9.30. i) x $6 Table9.30 Optimal Solution o> 3 5 0 0 0 o Taichi * * * * 5 B 3 1 0 18 0 -28 ° o ° 0 -28 1 43 5 6 ° 1 0 0 1 Z=36 ° 0 - 0 =3B11: Operations Research: Problems and Solutions (a) After adding the new variable, say x, in the given problem, the new LP problem becomes: Maximize 23x, +51, +26 subject to the constraints By Fy trS18, GD xy txgS4, @ Ht4656 and Hyp Xy p20. . Given that the column vector associated with variable, x is ag =(1, 1, 1). With the help of Table 9.30, we compute v3 0 -23)f1] [-18 31 4p}li 53 Yor Bt ag a oo 1 jit 1 Since in the current optimal solution, ey =(3, 0, 5), we have -13 =c5—tp¥e=2-@,0,5)| 5/3 |=-2(<0) 1 As ¢q~Z¢S 0, the optimality of the current solution remains unaffected with the addition of x, (b) The optimal basic feasible solution given in Table 9.30 does not satisfy the additional constraint 2x, +x, $8. Thus, a new optimal solution is obtained by adding slack variable s, to this constraint. Then, itis ‘written together with the entries in Table 9.30, as shown in Table 9.31. % Table 9.31 go 3 5 ° 0 0 0 ys Variables ‘Solution | x, % 4 % % Ss in Besis Values 5 OC =xp) 3 * 2 1 ° 13 0 -28 0 0 fa ° 0 0-28 1 48, 0 5 x 6 0 1 0 0 1 0 0 5 8 Zz 1 0 0 ° 1 2=36 9-4 [0 oO 1 0 a 0 InTable 9.31, itmay be noted that the basis matrix B has been disturbed due to row 4, Thus, the coefficients in row 4 must besome zero. This ean be done by using the row operations: R, (new) —> R, (old) ~ 2R; ~ Ry. We get anew table as shown in Table 9.32. Table 9.32, eo} 5 8 Wks Soon [SSCS imnBasis Vacs B b= xy) 3 x > pro 0% 0 0 zy 0 fo o -2» 1 4 0 5 zg s jo 4 en) on) 0 4 > |o o0 GD o ww 1 736 Ce Since the solution given in Table 9.32 is optimal but not feasible, apply the dual simplex method to get an optimal basie feasible solution. Introduce s, into the basis and remove s, from the basis. The new solution is shown in Table 9.33.Table 9.33 go 3 5 Variables — ‘Solution | x, % in Basis Values B 3 1 r 0 0 a 2 ° 0 3 4 6 0 1 0 s 3 0 0 3 0 0 0 0 0 0 0 -12 12 1 1 -1 0 1 ° 0-12-32 0 =32 12 objective function. REVIEW QUESTIONS Discuss the ole of sensitivity analysis in linear programming Under what circumstances ist needed, and under what conditions do you think tis not recessary? ‘What do you understand by the term ‘sensitivity analy the effet of) variation of, (i variation of Band (i ofa new constraint. Discus he changes in the coefficients. forthe given LP problem: Max Z= ex, subject to Ax=b, 120 Given, thatthe problem: Max Z'= ex such that Ax = b,x2 0 has Discuss ) addition ‘an optimal solution, can one obiain a linear programming problem ‘which has an unbounded solution changing b alone. 5. Consider the LP problem: Max Z = ex; subject to Ax=b and x > O.where e.x"e E",b” e E" and A is m n coefficients matrix. Determine how muck the components of the cost vector € can be changed without affecting the optimal solution of the LP problem. Find the limits of variation of element a, so that optimal feasible solution of Ax = b, x2 0, Max Z= ex remains optimal feasible solution when (i) a € B, (ia, B. SELF PRACTICE PROBLEMS {Ina LP problem, Max Z = ex, subjectto Ae = b, x20, obtain the variation in coefficients c, which are permitted without changing the optimal solution. Find this for the following LP problem Max Z =35, + 3x) subjectio (x, tx <1; Gi) 2x tay st [Meerut Univ. B Se, 1990] . Discuss the effect of changing the requirement vector from {6 4 24}t0[6 2 12}entheoptimal olution ofthe following LPprotiem: Max Z=3x, +65 + suyject io) xy tx tH ZO Gi) HF Se, ty 524 and HX 20. | . Discuss the effeet of discrete changes in the parameter | +b, = 1,2, 3) forthe LP problem, Max Z=3x,+4y ty + Tey subjectto (i) 8x, +3y +45, + 57; (2x, +64, +25 4 98453; xy tae, +5, +2558 opto %420[ Meerut Univ, MSc (Maths), 1992] and xx, 20. On x tome 4 4. Giventhe LP problem Max Z~—xy +2y—a5 subject to (i) Ixy +4y—1y $10; (ii) —x, +44 +52 6 Gi) ty s4 and Xyetaets20. Determine the range for discrete changes inthe resource values +b, = Wand, = 6 of the LP model sos io maintain optimality of the current solution, [Meerut Univ,, MSc (Matis), 1998] 5. Given the LP problem Max 35, +3 subject to (2) 3x,+2x,S18, (i) 44$4; (ii) x56 and xp 20. Diseuss the effect on the optimality of the solution when the ‘objective function is changed to 3x, +x. 6, Inan LP problem, Max Z~ ex, subject to Ax= b; x2 0, obtain the variation in coefficients c, which are permitted without changing the optimal solution, Conduct sensitivity analysis on c/s for the following LP problem. Max Z= 3x, + $5 subject 10 and (@ x+y Sl; Gi) 24 tay st x20. [Meerut Univ, BSc, 1990]1 10. nL. 13. Given the LP problem Max Z=r, +2n +5 subject to () 3x, +a—ay S105 (il) -x $y ty 56; Gi) xytxyed and Xpky AZO. (8). Determine optimal solution to the problem, (©) Determine tne effect of ciscrete changes in those components ‘of the cost vector which corresponds to the basic variable Find the optimal solution tothe LP problem Max Z= 15x, + 45e5 subject ©) x, + 16r, $250, ii) x50 and xy. 20, IfMax Z=2c,%) j= 1,2 andc, iskeptfixedat 45, determine how ‘much cane, be changed without affecting the optimal solution of i) Sx, +24 = 162, the problem. [Bombay Univ. B Se (Maths), 1991) Given the LP problem MaxZ=—2, +2x)—x5 subject to (i) 3x, +ay—x4 S10, )—x +4 +2526 Gi) xteys4 and py hzO. Determine the ringe for discrete changesin the resource values 2b, =10 and by ~ 6 of the LP model 80.08 to maintain optimality of the current solution. Consider the following LP problem Max Z~ 4x, + 65 subject to (i) x,+2y<8, Gi) Gx, +4r, $24 and yep 20. (8) What is the optimal solution? (©) If the first constraint is altered as: x, + 3x, <8, does the optimal solution change? A stainless sect utensil manufacturer makes thre types of items. ‘The restrictions, profits and requirements are tabulated below: Variables inBasi & B 200 T 150 c 2~ 15,8007 whore sy and 5, arethe slack variables for unased word and labour, respectively. (@) Determine the sensitivity limits forthe available wood and laboar within which the present product mix will remain optimal. (b) Findthe new optimal solution when the available wocd is 90 board feet and labour is 100 hours. Refer to the data in Practice Problem 12. 12, 4 Utensil Type ion i Raw material requirement(kgperunit) = 63S, Welding and finishing time (hours perunit) 3 4 5) Profit per unit (Rs.) 301 4 inlessstel (raw materia) availability is 25 kg and weldingand finishing time available is 20 hours per day, then the optimum product mix problem is expressed as Max Z=3x, +x, + 4x; subject 10 Its 6, +3x, + Sx, $25 (raw material restriction) Bx, + 4x5 + Sy $20 (time restriction) and spats 20 where 5 (j= 1, 2,3) is the number of units of the th type of the item to be produced. Get the optimal simplex table and encircle the appropriate answer tothe following questions. (@)_ The second ype of utensil would change the current optimal, basis fits profit per unit is: © 1; = 1.02; 1.06; = 2; 2 3. (©) The simplex multiplier associated with the machine time restriction of20 hoursis (~3/5). Thus, the multiplier remains unchanged for the upper limit onthe machine time availability of 25; 27.5; 35; 42.5; 32.5 hours, ‘The inereasein the objective function for each unit availability cofmachine time higher than the upper limit indicated in part ()is. (show calculations). The profit ofthe third type of wensil is Rs. 4 per unit. The lower unit on its profitability such that the current basis is still optimal is 4;3; 2.5; 2;<2. ‘The following LP model applies for Shastri & Sons Wood Furniture ‘Compaay which makes tables (T), chairs (C) and book shelves (), along with other toms, © @ Max Z= 200+ 1505+ 150%y subject to 10x; + 3g + 10g = 100 (00d) Sx,+Sig+ Sty 60 (labout) and ptora2 0 The optimal simplex ble is shown below: 5> 200150 «150 o ‘Solution MeO Values exp oar? 1 1 i ~670 207 o 4 ° won rr a a 00-5 07-807 (2) Determine the sensitivity limits for unit profits within which the current eptimal solution will remain unchanged, What is the total profit wien each table yields a profit of Rs. 180 and each chair yields a profit of Rs. 100. A-company makes two products A and B. The production of both products requires processing time in wo departments | and Il The hhourly capacity of and I, unit profits for products: A and B and the processing timerequirementsin LandIlaregiveninthe following able: ©)Capacity B Hours) 1 1 2 32 0 o 1 8 Unit profit (Rs.) 200 300 ‘Now the company is considering the addition of a new product C to its line, Product C requires one hour each of departments I and TL. What must be product C's unit profit in order to profitably add it othe firm’s product line? 15, A company produces three products A, B and C, Each product, requires two raw materials: tee! and aluminium, The following LP. ‘model describes the company’s product mix problem. Max Z= 30x, + Oxy + 50% subject 64, + Brg Src 450 (Steel) Bx, + dg Se 300 (Aluminium) and Eeteke2 0 ‘The optimal production plan is given in the following table: 52 30 1 530 0 8 Unit Profit Variables Solution | xy Xe S584 in Basie Values a B 0 bs 30-1 0 1 -1 so te ws 45 1 0 US 3,000, 0-30 0 0 -10 ‘wheres, and. ae the slack variables for unused stel and aluminium, quantity, respectively. (@) Suppose an additional 300 tonnes of tee! may be procured at, ‘a cost of Rs, 100 per tonne, Should the company procure the adtitional steel? (©) Unit profit of product A is Rs. 30. How much should this [price be increased so that A is produced by the company? ‘A pig farmer is attempting to analyse his feeding operation. The ‘minimum daily requirement of the three nutritional elements for the pigs and the number of units of each of these nutritional clemen's in two feeds is given inthe following table; 16 Required Units of Nuriional ‘Minimum Nueritional Elements(inhg) Requirement Flement Food | Food2 A 20 30 200 B 40 28 350 c 30 45 430 Cost per kg (Rs) 5 3 (@) Formulate and solve this problemas an LP model (b) Assame thatthe farmer ean purchase a third feed ata cost of Rs. 2 per kg, which will provide 35 units of autrient A, 30 units of ‘nuientB and 50 units of nutrient C, Would thischange the optimal mix of feeds? Ifyes, how? 17. A company sells two different products A and B. The selling price tnd ineremental cost information is as follows: Product Product B Sellingprice(Rs.) 60 40 Incremental cost (Rs.) 30 10 Incremental profit (Rs.) 30 30 ‘The two products are produced ina common production process and ae sold in two different markets. The production process has capacity of 30,000 labour hours. I takes three hours to produce ‘unit of A and one hour to produce a unit of B. The market has been surveyed and company officials feel thatthe maximum number ‘of unis ofA that can be seld is 8,000; and that of Bis 12,000 units (2) Find the optimal product mix. (©) Suppose maximum number of units of A and B that can be sold is actually 9,000 urits and 13,000 unis, respectively instead of as given in the problem, what effect does this have ‘on the solution? What is the effect on the profit? What isthe shadow price for the constraint on the sales limit for both these products. ‘Suppose there are 31,000 labour hours available instead of 30,000 asin the base case, what effect does this have en the solution? What s the effect on he profit? What is the shadow rice for the constraint of the number of labour hours. ‘An organizationean produce a particular component forpassenget ‘ars, jeeps and trucks. The production of the component requires utilization of sheet metal working and painting facilities, the details ‘of which are given below. © Resource ‘Comtumption 1 Produce Availability (a Unit for Passanger Car Jeep Tak Shoot metal working o2shr thr OShr 12 hes Painting 03 br thr 2 hr 30s The profits that can be eamed by the three categories of ‘componens, ie. for passenger cars, jeeps and trucks are Rs. 600, Rs. 1,400, and Rs, 1.300, respectively. (2) Find the optimal product mix for that ganization, () What additional profit would be eamed by increasing the availability of: (i) Sheet metal working shop by an hour only. (Gi) Painting shop by an hour only. ‘What would be the effect on the profit camed if a leat one ‘component for jeep had to be produced? ‘A firm uses three machines in the manufacture of three products, Each unit of product I requires 3 hours on machine 1, 2 hours on ‘machine 2and I houron machine 3. Eachunit of product requires hours on machine I, 1 hour on machine? and 3 hours on machine 3. Each unit of product Ill requires 2 hours on machine 1, 2 hours ‘on machine 2, and 2 hours on machine 3. The contibution margin of the three products is Rs. 30, Rs. 40 and Rs. 35 per unit, respectively. Available for scheduling are 90 hours of machine | time, 54 hours of machine? time, and 93 hours of machine 3time (2) Whatis the optimal production schedule forthe frm? (6) Whatis the marginal value of an additional hour of time on machine? ©ea (©) What is the epportunity cost associsted with product 1? (@) Suppose thatthe contribution margin for product is increased ‘What interpretation should be given to this opportunity fom Rs 30 Rs.43, wouldthis change the optimal production cost? plin? Give reasons. 1, -S¢,S3+(13; 5-(12)Se 55+ oF 925,50 2, The cureat basic Feasible solution consisting of, x and remains feasible and optimal atthe new values:xy ~7,x,% 5 and.x, 0 with Min 2=%6, 3. 52h, 2-32/5; S4/5262-Sid; Aby2~ 126/19 4, 6 <10; -S2S 26 5. Initial optimal solution: x, = 2, x, = 6 and Max Z= 36. Where cy ~Z, =~ I and ¢~z,=—3. After hevinge, =3 and ¢,= of ¢,~ z= 1 and ¢,~ 2, = 2/3. Thus, introduce 5 inthe basis 1 get the new Solution: x, = 2, = 6 and Max Z= 12, 7. (@) y= 0,494,150 and MaxZ=8 (6) (1/2) Sc, 8, (15- 195/16) sc, < (15+ 195/16) 9. b,<10; ~52< 6,26 12, (a) 36x, $120; 50x, 5003 ()xp=5477; 4077, xe=307; 10077, 2= 15,3077, 23,000/7. 13, 15Sx7$50; 6Sx¢$20; —eSxpS20 17. (@) x4=6,000, p= 12,000, Max Z=Rs. 5,40,000 (©) As4 8,000 not binding because x, ~ 6,000in the solution. Thus, an increase in the limit on 4 will have no effect onthe solution oF profit. Shadow price i zero. For increase inthe limit of B, the now solution is,x, = 5,670 andx =13,000 and Max Z= Rs. 5,60,000.The shadow price is (5,60,000 ~ 5,40,000) = Rs. 20,000, (©) New solution is x, = 6330 and xy ~ 12,000 with Max 2 ~ Ra, $,50,000, The shadow price is (5,50,000 ~ 540,000) ~ Rs. 10,000. calculate new valuesParametric Linear Programming Cae tity Introduction © Solved Examples Variation in the Objective Function Coofficients © Rey Variation in the Availabilty ofResources(RHS —® Self Practice Problems values) * Hints and Answers 1w Questions INTRODUCTION Parametric linear programming is used to determine the effect of predetermined continuous variations in the input data (orparameters , b, and g,) ofthe LP problem on its optimal solution. The parametric analysis aims at finding various basic solutions which become optimal one after the other due to continuous variations in the LLP model parameters. Since LP model parameters change a linear funtion ofa single parameter, this technique is known as linear parametric programming. ‘The purpose of this analysis is also to reduce the computational time required to obtain the changes in the optimal solution. The following two types of variations will be discussed in this chapter: 1. Variation in objective function coefficients, c, 2. Variation in resources availability (Right-hand side values), b, Let be the unknown (positive or negative) scalar parameter with which coetficients in LP model vary. We start the analysis at optimal solution obtained at k= 0, Then using the optimality and feasibility conditions of the simplex method we determine the range of A for which the optimal solution at = 0 remains unchanged, Let 2slies berween 0 and 2. Itmeans 0 < is the range of A beyond which the current solution will become infeasible and/or non-optimal, Thus ~ 2, a new solution is determined which remains optimal and feasible in other interval, say 2, << 2. Again a new solution at A =2, is obtained. The process of determining the range of" is repeated til « stage is reached beyond which the solution either does not change or exist. VARIATION IN THE OBJECTIVE FUNCTION COEFFICIENTS Let us define the parametric linear programming model as follows: Maximize2 = 3 (6; + Ref). a)subject to the constraints E ay x =b; i=1,2,. in and 320; j=1,2, where Acj represents predetermined variation in the parameter c, and A> 0 is a scalar (positive or negative) ‘parameter. Now the aim is to determine such consecutive values of 2. from —° tose at which current optimal ‘basic feasible solution tends to change with a change in the coefficients c) Such consecutive values of aro called critical (range) values of 2. When A = 0, the parametric linear programming problem reduces to the original LP problem. Thus given LP problem is initially solved by using simplex method at A= 0. ‘As. changes from zcro to positive or negative valuc, the values of basic variables remains unchanged, while the relative costcoetficient c's chenge. Since changes in cost coefficient c only ffect the optimality of the current solution, therefore as A changes only c, ~z, values are affected. Hence, for the perturbed LP problem let us caleulite ~z values corresponding to nor-basc variable column inthe optimal simplex able s follows: 6 )- 2,0) =0,Q)- cg) B Cj RY) * MF - CRYD=(G— BAG —Zs y= CY For. solution tobe optimal for ll values of & emust have c, (A) —z;(A) $0 (maximization case) and c;(A)— =,(4)20 (minimization case). These inequalitss fora given solution ae also used for determining the range Ay ‘S$A.S;, within which the current solution remains optimal. (@ fA, =, then the problem has an optimal solution for all. >,. Gi) If A, is finite, then the range of A beyond A, is improved as: (Gq *AC)-Cptrep)yy y= Bla where ¢j -2/ > 0 for maximization and cj - 2} <0 for minimization. ‘When A= A,, an altemative solution exists and whend. > h,, the solution is improved using simplex method. Let us define the parametric linear programming model as follows: Maximize Z= $6.5 subject to the constraints a and x20, FEN where 6, A 6; is the predetermined variation in resource values (right-hand side values), where 42 Ois a scalar parameter. Now aim is to find the range (or critical values) of 2. so that the current optimal solution remains unchanged with a change in the right-hand side constants 8,, for all i, Let B and xg = B"' b be the optimal basis matrix and optimal basic feasible solution, respectively of the original LP problem when itis solved at 4=0. Ifbisreplaced by b+ Ab’, then optimality condition c)~z, wll not be affected. However, such changes will affect the value of the basic variablesand the new values are given by Xp) =B! (b+ Ab) =B' b+) B'b'=x,+ Any Now s long as Xp (A) 20, the current basis remains optimal. Thus this criterion can be used to determine the range of A within Which the solution remains optimal as follows ~btRb;, imLet=A,. Then for 2 [0, A), the current solution remains optimal and at this solution the value of the objective function is givenby Z (A) =2+2.2"= ey xy+Aty xj. AIA, the current basis x (right-hand side) is replaced by x, (A)= B'' (b+ Ab’) and.x,, is removed from the basis by the usual simplex method. The process of finding the new range [2., 2] of values ofA isrepeated over which the new basis is optimal. The process is terminated when xB“ b'20 forall. This also implies that the current basis is optimal forall values of”. greater than or equal to the last value ofA. Example 10.1 Consider the parametric linear programming problem Maximize Z~ (3—6h)x, + (2-2h).xy+ (5+ SA) xy subject to the constraints: @ x,+2x,+4,5430; (ii) 3x,+ 2,<460; (ii) 3x, +4r,<420 and Xk tz20. Perform the parametric analysis and identify all the critical values ofthe parameter A. [Bharathidasan Univ, B Se (Maths), 199; Meerut Uni, MSc (Maths) 1998) Solution ‘The given parametric LP problem can be written inits standard formas Maximize Z = (3-64) x, +(2—2A) x, +(5+5A)x, +0s, +05 +05, subject to the constraints (0) xt Wey tay $s, =430, (ii) 3x,+2cy+5,=460, Gill) xy+42,+5,=420 and X).X.15.5),5p5320. According tothe problem: c()=6/+A¢j =(3,2,5,0,0,0)+A(-6,~2,5,0,0,0). Solving the given LP problem with A = 0. The optimal solution at =0 isshown in Table 10.1. ‘Table 10.1 Optimal SolutionatA=0 97 3 2 3 ° o o Gp —‘Foriables Solution inBais Values’ |x, 4 5 4 % 8 B baa) 2 y 100 | 14 i 0 mus 0 3 x 230 32 Q 1 0 2 0 0 % 20 2 a 0-2 1 1 21,350 5 7 2 5 1 ° g-4 | -4 ° o 4 - o ‘The optimal solution is: x, =0, x5 = 100, x, =230 and Max In order to find the first critical (or range) value of A for which the solution shown in Table 10.1 remains optimal, we first find ¢f — 2/ values corresponding to nom-basic variables x, s, and s, columns as follows: -14 2-14 g- fey, =C60,0)--2,5,0] 32 0 If f= 1,45 20-2001 1 as . =C60Q-lD+5, (6,0, 0) - 8, -1, 3) = © 14, 1,-3) Fora maximization LP problem, the current solution will remain optimal provided all c,()~z,(2) <0. Since cj —24 0, the first critical value of is given by‘This means that for’, € (0, |], the solution given in Table 10.1 remains optimal. The objective function value in this interval is given by ZQ) =Z+Z’ (A) =cpip tA C5 Xp = 1350+ 9500 Now for values of 2 other than zero in the interval (0, 1], we compute ¢, (.) -z,(2) values for non-basic variables x,, s, ands, asshown in Table 10.2. 1 A)—z, O) =(ey— 2) +2 (ef — 2f) =- 414A 0 or 222/7 40) —24 09 = (ey 24) +A(ch— 24) =-1 +R $0 or ASI 05 (A)—25 (A) = (cy- 25) +A (ch — 28) =-2-3A <0 or A2-2/3 ‘The optimal solution for any value of A between -2/3 and 1 is given in Table 10.2. Table 10.2 ga 6 2 5 o o o 6° 3 2 5 o o o thy ‘Variables—‘Solution inBasis Values x s, B bex,) 2.2 z 100 | 14 1 . @® 0 ss 4 230 32 0 1 0 12 0 oo 5 20 2 0 0-2 1 1 20) = 1,350-+ 950% 9-4 =A ° oc . 2 © gj-4y | -4 o 0 1 -3 0 4 Q)-20) |-4-1a 0 0 -1+A 2-20 AUL=1,¢4 (X)—z, (A)=Oin the ‘s,’ column. But for > 1, ¢4 ()—z, (A)> Ofor the non-basic variables, and hence the solution in Table 10.2 no longer remains optimal. Enter variable s, in the basis to find new optimal solution. The new optimal solution shown in Table 10. Table 10.3 Optimal Solution at.21 go -6 -2 5 o o 0 3 2 3 0 oo ch ty ‘Kariables in Basis yO8 % ay 2 oy B n -12 2 0 . -12(0 5 2 0 1 0 nm 0 4 1 4 ‘ o 0 1 ~97 2 6 0-520 -272-2 © 0-52 0 4 Q)-20) | -18 0 6 0 “5 0 ‘The solution shown in Table 10.3 will be optimal ifallc, (2) ~z, (4) <0, j= 1,2,5.To check the optimality, we ‘compute c, (A)~z, (A) values forthe non-basic variables x,,x, and s, as follows:3 2 6-2) = (6-21) + Ae} = 2{) = 7 PASO Or €0)-20) =(- 2) +(e = 24S Oor A> |, which is true 3s % 5 (0) 25 (8) AsOorke-1 Therefore, for Az 1, the G (A)—z, (4) $0 forall non-basie variable columns and hence the solution in Table 10.3 is optimal: x, ~1,=0, x5 = 230 and Max Z=2,300. Fork -2/3 ,c,(2)—2, (A) value fornon-basic variable s, becomes positive and again solution shown in Table 104 no longer remains optimal. Entering variable sin the bass to find new optimal solution, The variable s; will replace basic variables, in the basis. The new optimal solution is shown in Table 10.4. (e525) + A(c§ - 2h) = — ‘Table10.4 Optimal Solution G7 5 0 o o 6 5 0 0 0 Variables Solution | x xy Ss. inBaris Values a B b (=p) 22 % 105 we ° ° oa 5 5 es 220 Rm 0 1 1 0-12 oo 3 20 200 0 2 1 1 310 +8902, 9-4 0 0 o 4s 0 2 6-3 -8 0 o 4 0 3 Solution shown in Table 10.4 will be optimal only when c,(&)~z, (i) $0 for/=1,4,6 ©, A)-z, A) =(c, - 21) +2 (ce, - 2) =0-8A $0 or 220 4 00) —24 2) = (Cy 24) +4 (Cy 7) =-S-—SASO oF 12-1 6 O) - 26) = (eg - 29) + (CG - 2%) = 243A 50 or AS23 ‘Thus for-1 $.$-2/3, the optimal solution ix, =0, x, 105,x, =220 and Max Z=1310+890 2. Hence Table 10.2 to 10.4 give family of optimal solutions for 2/3<1£ 1,12 1 and—1S2-2/. The critical value off are-2/3 and 1 Example 10.2 Consider the linear programming problem Maximize Z= 4x, + 6¢, +2x,, subject to the constrains Oytytys3, Gi) n+4yt+y<9 and py 520. The optimal solution to this LP problem isgiven in Table 10.5, ‘Table 10.5 Optimal Solution 9? 46 2 0 0 Cost Variables. Solution Oe 45 5, 8; porUnit — inBasis Values & B 4 x 1 1 0 “1 430° 1B 6 y 2 o 4 2 413 a8 0 0 6 03-28 Solve this LP problemif the variation cost vectoris:c/ 2,-2,2), Identify all critical values of the parameter A.Solution The given parametric LP problem can be written in its standard formas: MaximizeZ=(4+ 2A)x, + (6—24) x)+ 2+ 2A)x5 +08, + 05, subject to the constraints @ xytrytayts=3, Gin tt Tx,+y=9 and ApXpty5pSp20- Since LP problem is of maximization, thecurrent solution shown in Table 10.5 will remain optimal when all 5 A450. ‘The values of 2 other than zero, c) values become linear function of 2. Thus to caleulate these, an additional profit coefficient row c/ is added in the Table 10.5 and shown in Teble 10.6. ‘Table 10.6 Optimal Solution go? zo-2 2 o o 57 46 2 o o ee Variables’ = Soluion =| x, ty Hi inBasis Values a boxy) x 1 100 =I 3 x 2 oon 2 6-3) 6 0 6-103 -23 6@-4@ | 9 0 8-43 43 oI a ED For solving the given LP problem with A> 0, we find ¢)—z, values corresponding to non-basic variable.x, 1s, and 5, columns as follows: yan -1 43-13 ~9-4y-0.-2.90-9 [9 | “1313 “Tabln 10,6 repronent a bans fons parametric LP problem. This solution will be optimal if alle, (A) z, (4) $ 0 for j= 3,4, 5. Thus we con determine the range of 2 for which solution remains optimal as follow: 3 0) =23 (0) = (6 ~ 25) + (c=) =- 6+ SSO or AS3/4, 4 0) ~24 0) = (€y ~ 24) +h (C, ~ 24) = (10/3) ~ (10/32) $0 or A2— 40) —24 Q) = (Cy ~ 29) +h (Ch 29) =— CIB) + 4B)ASO of ASIN. “Thus, the solution in Table 10.6 is an optimal solution for the given parametric LP problem for all values of A between I and /2.and Max 2= 16-24. For A> 12, the (A) Zs () value corresponding to non-basic variable sy-zolumn becomes positive and solution no longer remains optimal. Entering variable s, in the basis o find new optimal solution. The new ‘optimal solution is shown in Teble 10.7. Table 10.7 Optimal Solution and Z(Q)=Z+12'= 16-22 forthe given G7 2 0 0 67 4 2 a 0 cy ep Variables Solution A % 5 s % inBasis Values zB bx) 20 4 x 3 1 1 1 0 oo L 6 o 3 6 1 = 2-4 ° Z@)=12+6 6@-3@) | 0-4 . 2 °3,4, Therange of within which The solationshown in Table 10.7 will be optimal ifallc,(.)~z,(A) $0 forj=: the solution remains optimal, we compute €(B) 25.0) = (6-9) +2 (Cy - 24) €3 (A) —25 A) = (€,— 4) +A (4-74) 4 (9) £4 A) = (64-29) +2 (Cy -24) = 4-200 or AZ-2 ‘Thus for all A2 1/2, the optimal solution is: x, =3, x,=0., 5,7 6 and Max Z= 12+ 6), ForA.<-1, ¢, (2) 2, (2) for the non-besic variable s, becomes positive andhence solution shown Table 10.6 ‘no longer remains optimal. Entering s, non-basic variable in the basis to replace basic variable x,. The new solution is shown in Table 10.8. - 440 or A212 250 Table 108 go 2-22 o 67 46 2 o ey ‘Variables ‘Solution | x % inBass Values B 6 oo sy aa a0 lt 2 6 » ous a us 5-4 520 HD 2Q)=272-9n2 4@-3@ | 32 0 ao 2 ‘The solution in Table 10.8 will be optimal ifcA)—z(2) $0 forj=1,3,5. Thus we compute 1 (A) 2, A) = (¢, - 4) +2 (C, - 24) = (512) + (SYA SO or ASI, 5 (8) = 25 A) = (C- 4) +A (C4) = (172) + (112) $0 0rd S171, 5 (A) ~25 (A) = (C5 ~ 29) +2 (C5 — 25) = - GB/2) + (IZA SO OFAS3. Forall 2-1, the optimal solution is:x, ~0,x,~9/4,x3~ 0,8) =3/4,s) =O and Max Z=27/2—(9/2)h. Hence the rangeof A for which solutions are optimal will be:-1<4.<1/2, 22 1/2 and AS—I Example 10.3 Consider the linear programming problem Minimize 2-—x-3x) subject to the constraints @ mtn s6 G4 +2 <6 and xp 20. ‘The optimal solution to this LP problem is given in Table 10.9. Table 10.9 Optimal Solution at 4=0 67 “1 =3 0 0 ce Variables Solution inBasis. Values x % s % B boxy) =1 x 2 1 ° 23 = -3 % 4 o 1 13 1B 1 3 38 - 28 0 0 3B 2B Solve this problem if the variation in the cost vector is c'=(2, 1, 0,0). Identify all the critical values of parameter.Minimize 2=(-1+22).x,+(-3+A)x,+0s, +05, subject to the constraints @ x4 HyH5,=6, i) -x,+2x,+5,=6 and Xpty5pH20. ‘When 2 = 0, the given parametric LP problem reduces to the ordinary LP problem whose optimal solution: x,=2,x,=d.and Max Z=~ 14s given in Table 10.9 In order to find the first critical value of 2 other than zero for which the solution shown in Table 10.10 is optimal, we first find cj — z/ values corresponding to the non-basic variables s, and s, as follows: j shy any aun] “EP j v3 = (0, 0) - (513, 13) = (5/3, 13) ‘Since LP problem isof minimization atthe optimal solution we must havec, (2) —z,(2)20 forall j. Also as ',<0, the first critical value of is given by ‘This means that for A, € [0,1], the solution given in Table 10.9 remains optimal. The objective function valuein this interval is given by Bh) -Z+2" aruthehag-- weap] 14+ 8a ‘Thus for values of A other than zero in the interval [0, |]. we compute the solution as shown in Table 10.10. Table 10.10 g- 2 1 0 0 57 -1 ot 0 cy ey Variables Solution inBasis Valuer x * 4 8 B bem) 2-1 x 2 Hl 0 2B =13 133 4 o 1 ws 1B Z0)=- 4% ° 0 38 2B o o -s0 13 38 BA 5s. Baha ° ° aoa" 33 ‘The optimal solution given in Table 10.10 will remain optimal ifal c, (A) ~2, (A) 2 Ofor 6,0)-2,0) = (6-4) 44-25) = GBM 4) 2408) = (Cy 24) +2 (Cy 244) = (2/3) + (113) ALA= 1,5 ®)—2, (A)=Oand c, (A) —z,(2)> 0. So entering variables, inthe basis. Thenew solutions shown in Table 10.11.Parametric Linear Programming Fag) Table 10.11 Optimal Solution at 2. go 2 : ° ° 67 1 -3 0 0 x ta ‘Fariables Solon inBasis Values 4 4 5 % Bo bes) oo 7 ° 7 1-3 4 \ 0 ° ° o ° o Since LP problem is of minimization, therfore for the solution to be optimal we must have 6) -2,(0.)2 0 forall. But c,(4)~z, (4) 20 for non-basic variable columns I and 4at 4= 1. Thus thecurrent solution: x, =0,.x,=3 and Min Z=— 6 is the optimal solution. However, we need to find the new critical value (of Ain the interval [1,A, ] over which the solution shown in Table 10. remains optimal. For this computing: 2, = Min (64-24) } -32 G4) <0f = This shows that fA € [1 3] the ¢,(4) ~z; (A) 0 for j= 1,4 and the solution shown in Table 10.11 is optimal, In the interval [1,3], the value of objective function is given by 943k 20y-qeehein-t0-a[ 2.08] AtA=3,c, (A)—z,(A)= 0,80 entering variable s, into the basis and remove x, from the basis to get a new solutionas shown in Table 10.12. ‘Table 10.12 Optimal Solution at}: ed 2 1 o o > 1 3 0 6 a ~—~—CNariables. Solution mBasis Values x m cf 8; B xy) 0 0 5 6 Hl 1 1 0 o ° 2 6 1 iz ° 1 “1 3 ° o 2, 2 1 0 6,0)-2 0) 5 ° ° Alllc,(.)~z2) 2 Oin Table 10.12. However, we find the interval 3, 4 in which this solution remains optimal as follows: dy = Mi van se] / 20, this solution will remain optimal for all values of A in the interval [3, =]. Butall ¢j ~7Example 10.4 Consider thelinear programming problem Maximize Z=3x,+2x, + 5x, subject to the constraints @ x, 42e,+x,5430+2, (8) 3x+2x,S460-4, il) x + 41 $420-44 and xy 38,20 Determine the critical value (range) of 2 for which the solution remains optimal basic feasible, Solution The given parametric LP problem can be written in its standard form as: Maximize Z= 3x, +2x, + Sx, +05, +0s,+0s, subject to the constraints @ xj42e,+x,+5,= 430+, @ 3x, +2x,+5y Gi) x, 44x, +5,=420-44 and ppp 5155395520. ‘The optimal solution when A ~0 is shown in Table 10.12. Table 10.13 Optimal Solution st 2-0 97 3 2 a o o o @ Variables Solution inBasis Values 4 ey ep 4 5 a vo 2 % ~04 1 ° wm 4 ° 5 4 32 0 1 o 2 0 0 55 2 0 ° -2 1 1 1350) 7 2 5 1 2 0 -4 0 ° -1 -2 0 In orderto find the range in which the solution shown in Table 10.13 is optimal, we first calculate B'b’ we -v4 olf 1) [32 xy ow olf-4}=| -2 -2 0 if-a] [-10 Fora fixed A, values of basic variables in Table 10.12 becomes.x,= 100+ (3/2)A, x3 =230~2- ands, =20-10A, ‘The optimal solution shown in Table 10,12 will remain optimal as longas 4 = 100+ (3/2)A 2 0 or AS - 2003, 7230-22 20 or ASLIS 5, =20-10 20 or As? Consequently the solution in Table 10.13 will remain optimal between ~ 200/3 and 2, i.e. - 200/3 $12. In particular for any A = [0, 2], the objective function value and the right-hand side values are given by 100 Ed ZQ) = ey Xp +h qh =(2,5,0)| 230] +2(2,5,0)} -2]=1,350-72 10 100 +302 XQ) = xp tA xp= 89 +2) -2|=]230-20 20-10 |.z vem nc Poor EEN Evidently for > 2, the new solution will be primal infeasible because-x5(=s,) will become negative. The solution atA = 2is shown in Table 10.14. Table 10.14 Optimal Solution at A=2 qe 3 2 s 0 0 « ce Verables Soliton inBass Yas | x, ‘, h oO4 z z 7% “a om -1A s * wo 1 ow 0 % 2 0 o @ 1 1 2=1336 z 7 2 3 1 2 -4 0 ott For .>2, the basic variable s, becomes negative. Consequently solution becomes infeasible. Therefore dual simplex method is applied to find the new optimal solution. Remove s; (because 1,5 <0) from the basis. Determine the ratio {¢)~z;) (3 ¥-<0} = 1/2(correspondsto s, ) and enters, into the basis. The new solution isshown in Table 10.15. Table 10.18 Optimal Solution at A>2 97 3 2 3 0 0 0 op Variables Solution inBasis ‘Values x % s 4 5 B vob 2 % 1 ys | a 1 0 0 om 3 xy 2 m3 | 32 0 1 0 v2 o ° Ss) 3-0 | 4 0 ° 1 “2-10 Z= 1,336 3 8 2 3 ° 32 5-4 zs 0 0 a 52-19 In orderto find the next critical value (range) of in the interval [2, Ap] in which the solution shown in Table 10.14 remains optimal, we first find 0 0) yf ay per x,=B'b’=|0 12 0||-4)=|-2 1-12 -1}[-4] | 5 ‘The solution shownin Teble 10.15 willremain optimal aslong basic variable x,,x, ands, remain non-negative, i. 1X2 = 105-220 oF AS 105; x;=230-2A20 or AS IIS 5, =-10+5420 oF AZ? ‘Thus the solution ie optimal for all values of in the range 2 Az 10s, For € [2, 105], the optimal objective function value and the right-hand side values are given by 105} -1 ZQ) = yxy +h eg xf = (2,5,0) |230| +2 [2,5,0]| -2 |= 1360-122, 10 5 10s] [-1] [10s~ x Xp) = xptAx =| 230|+2|-2|= |230-22 0 s} | +s onises)18Evidently, > 105, thenew solution will be primal infeasible because basic variable x, becomes negative. Hence, no optimal solution exists forall A2 105. For A-<—200/3, the basic variable in Table 10.14 becomes negative. Applying dual simplex method to dint solution for 4.< ~ 200/3. Entering non-basic variable s, into the basis to replace basic variable x,. The new ‘optimal solution is shown in Table 10.14. ‘Table10.16 Optimal Solution at A= 105 3 2 5 6 o 0 s) -6 40 | 1 -4 0 -2 4 0 8 x 1 30 | 4 201 1 0 0 ss 4 0 | 4 0 0 Oo 1 150 z, s 0 s $s 0 0 z “8 0 -5 0 0 ‘The critical values of 2, for which solution shown in Table 10.16 remains optimal are calculated as follows: 2 2 1 0 x |=B'H=| 1 0 0 % oo. Hence the basic solution in Table 10.15 will remain optimal provided 43 =430+A200r2-430; 5, =-400- 642 0 or S—2003 5, =420-4h200rAS 105 This implies that the solution is optimal in the range —430<1.<—200/3. For 4 <—430, the basic variable.x, in Table 10.16 becomes negative. As there is no negative entry in.x, row of ‘Table 10.15, the prime solution is infeasible, Hence, there exists no optimal solution to the problem forall A <— 430. Hence, Table 10.13, Table 10.15 and Table 10.16 gives range of A values, - 200/3 <2, 2A 105 and ~ 430 < i < ~ 200/3, respectively for which solution is optimal. Example10.5_ Consider te linear programming problem Maximize Z=4r, +6r,+2r, subject to the constraints O x +x +43, and y%y.¥,20. ‘The optimal solution to this LP problem isshown Table 10.17. Table 10.17 Optimal Solution ) ry Hae, + Ixy $9 9? 4 6 x 6 o Variables — Solution | x, 7 OO inbais —Yalws & B ber) ° q & 1 5 8 2=16 oo 1 be S 1 8eee Parametric Linear Programming FETE Solve the problem ifthe variation in right-hand side vectors: (3, ~ 3)". Perform complete parametric analysis and identity al critical values of parameter A. Solution The given parametric LP problem can be written in its standard form as: = Arr, + 6, +21, +05, + Os ‘subject to the constraints @ xtxytxyts=343A, (x td tI ty and ppt 5520. ‘The optimal solution when A = 0 is given in Table 10.17. For values of other than zero, the values ofright- hhand side constants change because of the variation in vector 8’. The new values are computed as follows: 4] epty-[ 43 “PTY S al 7 "| wal [-s}7[-2)° As A changes, values of basic variables x, and x, also change and solution shown in Table 10.17 remains ‘optimal provided value of basic variables remains non-negative That is, solution remeins optimal provided xy = 145020 or 22-15, xy =2-2A20 or ASI. ‘Thus, solution remains optimal inthe range—1/5 SA I andis given by: x, = 1+ 5A,x=2—2A,35 =0,and Max Z= 16+ 8h. For A > |, the basic variable x, becomes negative. Consequently solution becomes infeasible for the primal, but remains feasible for the dual because all c,~ z, <0. Apply dual simplex method to find the new optimal solution for A > 1. Evidently x, is the variable that leaves the basis. Determine the ratio {(6)~ =)! Yj Yq < 0} = 10 (Corresponding to 5,) and enters, into the basis. The new solution is shown in fabte 10.8, Maximize 3h 5 Table 10.18 OptimalSolution 6? 4 6 2 oo Variables Sotuion | xy x) 5 inBasis Values os B vos 4 38 14 7 0 1 0 4 6-6 | o -3 ~~ 1 -l 2-36-12, 4 6 2% 0 4 0-0 -2% 0-4 ‘The basic solution shown in Table 10.18 is: x, =9~3A, x, =0,.x=0,x4=—6+ 62,x,=0and Max Z= 36-121. ‘This solution will remain optimal provided: x, =9-3200r.&3 and.x,=—6+ 642 0orA22. Thatis, solution is optimal for all 124.23. For 4> 3, the basic variable x, becomes negative. As there isno negative coefficient in the x, row, the primal solution is infeasible. Hence there exists no optimal solution to the problem for all A>3. For 4 S -1/5, the basic variable x, in Table 10.17 becomes negative. Consequently, solution becomes infeasible forthe primal, but remains feasible for the dual, because allc)~z) = 0. Applying dual simplex method to find the new optimal solution for 4 <—1/5. Evidently x, is the variable that leaves the basis. The ratio {(6)-2) 302g <0} = {6,2}. Enters, into the basis. The new solution is shown in Table 10.19.Table 10.19 Optimal Solution 97 46 2 o 0 Variables Solution yom 5 inBasls Values oe B vie 0 f -18 -3 | -3 0 30-4 1 6 % 303 14 1 U 0 Z=18 +18 x 6 6 6 6 ° 6-4 20-46 0 ‘The basic solution shown in Table 10.18 is: x,=0,x,=3 +32,.x,=0,x4=0,x,=—: This solution will remain optimal provided +3A20 of A2-1 and xs=-3-15A20 or AS-IS. ForA-<-1, the basic variable.x in Table 10.19 becomes negative. As there is no negative coefficient in the 1% row, the primal solution is infeasible. Hence there exists no optimal solution to the problem forall A<— 1. Thus Tables 10.17, 10.18 and 10.19 give families of optimal solutions for ~ 1/5 $4< 1, 1SAS3 and-1<1-< U5 respectively. eye ee REVIEW QUESTIONS |. Replan the basic difference between sensitivity analysis and where 0152.5 8. ,by simple method, the new basis which ye parametric programming. the optimal solution, alo yields aminimum for at east one value 2. Ina linear programming problem Of. If A; SAS Apis the entire range of values of A. for which the A new basis yields a minimum, then show that, Min z= ex subject to Ax-band x20 discuss the effect of (@) discrete changes in the requirement vector, (b) diserete changes inthe cost vector. 3. Show that in solving a parametric linear programming: | Mii 2eGR093 | bjec to ‘Ax=bandx20 | SELF PRACTICE PROBLEMS. L. MinZ-As-y | Perform complet parame programming analysis. denify he subject to () x24, Gi) 2e+ys3 range of critical values ofthe parameter 1 and all optimal basic, a wmcashspso | feasible solutions. ‘where ais an arbitrary, and small scalar number but finite and Bis | > © ‘i ecbifvary sla po lac iunibr ict Beis. Paced cociplets x, +x $10-2 parametric programming analysis. A easeea 2 (9 MmZ=0— D5. oy Meza 85 ty subject to () xy +2x510 (H) 24+ xy S11 subject to (i) &, +31 + 4x, +x,S7—A Gi) x, -2753 Gi) m+ 6a ts ened eh (ii) x, + 4,4 525 +2ej 584) and x pekay By %q2 0. subjet to) 3xy-3ap—ay Fg S Perform a complete parametric programming analysis to find the (i) 2x,-2y+Hy-442 3 range ofl. for which the solution remains optimal basic feasible. and pty kya 2 0. 4. Max 2=3x, + 2x) + 51 (Weerut Univ, M Se (Mats), 1993] subject to () x, +2; +x55430-+ 100,(ii) 3x, + 2x5 460-2008 (ii) x, xy $420 + 400% and wd 20. Perform parametric analysis to determine the range of for which the solution remains optimal basic feasible. 5. Max 2=(6—Aix, +12—A)x, + 4A subject to () 3x, +42, +2582, Gi) x, 43x +2481 and Ry 420. Perform a complete parametric programming analysis ad identify all the critical values ofthe parameter A. Max Z= 3x, +45, subject to (i) 2x, #45212 and aym20. Determine the optimal sctuton for =0 and find the ange of Gi) x4 S6-% | | | Ree ee | 8, | | inate inal eooanee ERE] under which the solution remains optimal for A 0. Man Z=7x, + 41, * 615 + Sxy subject © (i) 2x, txyt e+ xy S10+R (ii) 3x, +x3- 24,4 3x S45—A and tye ty 420. Perform a complete parametic programming analysisand identify all the critical values ofthe parameter Max Z=(4 — 102i, + (8— 4A} subject to (i) x) #254 and yet 20 Gi) 2x, +S3-2 ‘Study the variations in the optimum solution with the parameter, where = << HINTS AND ANSWERS x= 815, y=— 15 and MinZ: . @) 4G“ S:00532 i.xg73:32SA83 yt 15h23. 1/5 + (8/5) A. Problem has one characteristic solution: ~ 2< 2 <3 and a multiple solution for = 3, (©) x-xy7 43-05 x4 Sand Max Z~5;-2<2<3. @) 4)90,x,"2,2As6 ) xen On R4sas7 372.7 54523 5.x, 3sAs9 9 10, 0sh22 8 12, 2sks 72 4) =.= 72,437 0,x,= 12, LU2SAS 35/2 No feasible solution when A > $9/2. osas3 6. x) 732x=92: -382837 -0,-sshs-5 No feasible solution when 2.> 3Creat) © Introduction © Branch and Bound Method © Types of Integer Programming Problems © Solved Examples ‘© Gomorys all Integer Cutting Plane Method © Review Questions © Gomory’s Mixed-inleger Cutting Piane © Self Practice Problems Method © Hints and Answers INTRODUCTION In linear programming, each of the decision variable as well as slack and/or surplus variable is allowed to take any real or fractional valve, However, there are certain practical problems in which the fractional value of the decision variables has no significance. For example, it does not make sense saying 1.5 men working on a project or 1.6 machines in a workshop. The integer solution to a problem can, however, be obtained by rounding off the ‘optimum value ofthe variables tothe ncarest integer value. This approach can be easy in terms of economy of effort, time and cost that might be required to derive an integer solution butthis solution may not satisfy allthe n constraints, Secondly, the value of the objective function so obtained may not be optimal value. All such difficulties ean be avoided if the given problem, where an integer solution is required, is solved by integer programming techniques. There are certain decision problems where decision variables make sense only ifthey have integer values in thesolution, Capital budgeting, construction scheduling, plant location and size, routing and shipping schedule, batch size, capacity expansion, fixed charge, ete, are few problems which demonstrate the areas of application of integer programming. TYPES OF INTEGER PROGRAMMING PROBLEMS Linear integer programming problems can be classified into three categories: (i) Pure (all) integer programming problems in which all decision variables are required to have integer values.(i) Mixed integer programming problems in which some, but not all, of the decision variables are required to have integer values. Gil) Zero-one integer programming problems in which all decision variables must have integer values of orl. ‘Various methods for solving an integer programming problem are summarized on the next page. Inthis chapter, we shall discuss two methods: (i) Gomory’s cutting plene method and (ii) Branch and Bound ‘method for solving integer programming problems. ‘The pure integer linear programming problem in its standard form can be stated as follows: Maximize Z=0, x, +0)4)+...+¢q%, subject to the constraints 4% $09) tee. Fy9 X=) yy X, +09 X) +... FA yy Sy =D a ee and pty ++ o%q2 0 and are integers. GOMORY'S ALL INTEGER CUTTING PLANE METHOD An iterative procedure for the solution of an all integer programming problem by Gomory’s cutting plane ‘method may be summarized as follows: Step 1 Initialization Formulate the standard integer LP problem. Solve it by simplex method ignor- ing integer requirement of variables. Step2 Testthe optimality (a) Examine the optimal solution. Ifall basic variables (i integer values, the integer optimal solution has been derived and the procedure should be te current optimal solution obtained in Step 1 is the optimal basic feasible solution to the integer linear Programming. (b) If one or more basic variables with integer requirements have fractional values, then go to Step 3. Step 3 Generate cutting plane Choose a row r corresponding to a variable x, which has largest fractional value f, and generate the cutting plane (a Gomory constraint) GSE fee) 8 hase Ey ier er where O| Current solution is integer sl She required ineger eae LP problem solution [v0 ~ Select the basie variable with largest fractional value «= Generate te cutting plane ‘+ Add the cutting plane tothe bottom of optimal simplex table ‘+ Find new optimal solution using cual simplex method Fig. 1 ‘Step 3. Generate cutting plane Choose a row r corresponding to a basic variable x, which has largest fractional vaiue f, and generate a cutting plane as explained eariier in the form Flow Chart for Solving ILP Problem snot 3 ayy (& ae ayy ht Efi Osh ety 3 dq 2 Oand x; non-integer 2 ay fp and x; integer ‘Step 4 Obtain the new solution Add the cutting plane generated in Step 3 to the bottom of the ‘optimal simplex table as obtained in Step 1. Find a new optimal solution by using dual simplex method and return to Step 2. The process is repeated until all restricted basic variables are integers, ‘This retod is suitable fo the integer programming problems involving ony to variable. The eave procedure is summarized as follows: Step 1 Obtain the optimal solution of the given problem ignoring integer restriction on the variables. If the solution to this LP problem (say LP-A) is infeasible or unbounded, then the solution to the given all-integer programming problem is also infeasible or unbounded, as the case may be.Examine the optimal feasible solution. If the answer satisfies the integer restrictions, then the optimal integer solution has been obtained. If one or more basic variables do not satisfy integer requirement, then g0 to Step 2 Let the optimal value of objective function of LP-A be Z, Step 2 Branching step (a) The value Z, will be considered as initial upper bound on objective function value for integer LP problem. Let it be denoted by Zy. The lower bound on integer LP problem can be obtained by rounding off to integer all values of the variables. (b) Let x; be the basic variable having largest fractional value. (© Branch (or partition) the LP-A into two new LP sub-problems based on integer value of x, that are immediately above and below its non-integer value. That i, partitioning is done by adding two mutually exclusive constraints 4 Sty] and x,26J+1 to the original LP problem. Here [x,] is the integer portion of the current non-integer value of the variable Ap Obviously it is done to exclude the non-integer value of the variable 24. The two new LP sub-problems are as follows: LP Subsproblem B LP Sub-problem C Max Z= 3 ox, Max Z= § 6x, a it subject to is subject to Sous x, S by) 2b) + 1 and 320 and 320 Step 3. Bound step Obtain optimal solution of sub-problems B and C. Let the optimal value of the objective function of LP-B be Zy and that of LP-C te Z,. The best integer solution value becomes the lower bound on the IP problem objective function value. (Initially this is the rounded off value.) Let the lower bound be denoted by Z,. Step4 Fathoming step Examing solution of both LP-B and LP-C, which might contain optimal point. Exclude a sub-problem from further consideration if (@) any of the sub-problems has an infeasible solution, itis seid to be fathomed by infeasibility, or (b) a sub-problem yields a solution that is feasible but not an integer, then return to Step 2; (©) a sub-problem yields a feasible integer solution, examine the value of the objective functions. If this value is equal to the upper bound, then an optimal solution has been reached. But if it is not equal to the upper bound but exceeds the lower bound, then this value is considered as new upper bound and return to Step 2. Finally, if itis less than the lower bound, terminate this branch. Step 5 Termination The procedure of branching and bounding continues until no further sub-problem remains to be examined. At this stage, the integer solution comesponding to the current lower bound is the optimal all-integer programming problem solution. Example 111 Solve the following integer programming problem using Gomory’s cuting plane algorithm. MaximizeZ: subject to the constraints @ 3x,42y55, (i) yS2 and X1.¥)20 and are integers. [Madras Univ., BE (Mech & Prod.), 1990] Solution 1. Obtain the optimal solution to the LP problem ignoring the integer restriction by the simplex ‘method, The optimal solution is shown in Table 11.1 itecp Variables 2 5 % in Basis B 1 x) 1 0 1B 1 » 2 1 0 1 67% 0 =18 =18 In Table 11.1, since alle, —z,$0, the optimal solution is: x, = 1/3, x,=2 and Max Z=7/2. 2. In the current optimal solution, all basic variables in the basis are not integers and the solution is not acceptable. Since both decision variables x, and.x, are assumed to take on integer value, a pure integer cutis, developed under the assumption that all the variables are integers, as explained in Step 3. 3. Since x, is the only basic variable whose value is a non-negative fraction, we shall consider the firstrow for generating the Gomory cut. Considering x, -equation as the source row in Table 11.1, we write ‘The factoring ofthe x,-row yields aern(0e2) (104) a= (ord) Notice that each of the non-integer coefficients is factored into integer and fractional parts in such a manner that the fractional par is strictly positive. Rearrange the equation so that all of the integer coefficients appear on the left-hand side. This gives toa got gang tm) Since x, and s, are non-negative, left-hand side must satisfy Talis! g tl St ZHZZ OF GTZ SM 1 1 or “Estat wheres, is the new non-negative (integer) slack variable By adding the Gomory cu atthe bottom of Table I. the new table o obtained is skown in Table 11.2. ‘Table 11.2 67 1 1 o 09 0 cp Voriables mR na inBasle z 1 % 1 oO ° 1 a ° 1 ° 0 0 ‘a 0 o GD 1B 1 2-72 ry "aa Ratio: (6-2), <0) | — = 1 1 =4. Apply the dual simplex method to find the new optimal solution. The key row and key column are marked in Table 11.2, The new solution is shown in Table 11.3. ‘Table 113 Optimal Integer Solution go 1 1 0 0 0 os Variables * * a ” in Basis B 1 0 1 1 ° 1 1 ° o 0 1 1-3 0 0 0 0-1 The solution given in Table 11.3 is: xy = 0) ~2,s,, = 1 and Max 22. This also satisfies the integer requirement. Example 11.2 Solve the following integer linear programming problem using the cutting plane algorithm. Maximize Z=2x, +208, ~ 10x, subject to the constraints @ 2x, +20, +4xys 15 (@ 6c, +20x,+4x%,=20 and 2), Xy%; Non-negative integers. Also show that it is not possible to obtain a feasible integer solution by using the method of simple rounding off Solution Adding slack variable s, in the first constraint and artificial variable in the seeond constraint, the problem is stated in the standard form as: Maximize Z= 2x, +20r,—10x,+0s,-MA, subject to the constraints © 2x, +20n, +424 @ 6, +20x, +45, +4, and Xp XpXy 5p» A, 2 Oandare integers. The optimal solution of the problem, ignoring the integer requirement using the simplex method (Big-M technique) is shown in Table 11.4, ‘Table11.4 Optimal Non-integer Solution 672 20 » Variables Solution |x, * 5 So in Basis Values 20 * se | 0 1 us 340 2 x sit 1 0 ° “14 Za15 3-5 | 0 0 —4 1 ‘The non-integer optimal solution shown in Table 11.4 is: x, = 5/4, x, = 5/8, x, = 0 and Max Z = 15. x= 0 and Max Z= 2. This sotution does not satisfy the rounding off the values of the variables. To obtain the integer-valued solution, we proceed to construct Gomory's constraint (fractional cut). Since the fractional part of the value of x, =(0 + 5/8) is more than that ofx, (= 1+ 1/4), the.x,-row is selected for constructing the fractional cut as given below:3 syfow Or, +245 5+ 3p The factoring of the x,-row yields : ange (oed) ye oe) aneed oe LyeZaad * Onre-arranging above equation, we obtain the Gomory’s fractional cut as: ~S3-gitinn gE (cut Adding this additional constraint at the bottom of optimal simplex Table 11 4, the new valves so obtained are shownin Table 11.5. ‘Table 11.5 97 2 2-10 ° o oe Variables ee % 5 5p, in Ronis 1 ‘ B 20 a 0 1 S30 0 2 x ° 0 -us 0 0 "i o 0 ws 1 Z=15 9% 0 o -4 1 0 Ratios (6 — Vr (<0) 7403 Remove the variables,, fromthe basis and enter variables, into the basis by applying the dual simplex method. ‘The new solution is shown in Table 11.6 ‘Table 11.6 6 2 2-10 6 o o Variables ‘Sohtion | x, % s “a inBasis B 20 a 0 1 0 -1 2 x 103. 1 0 28 6-103 0 s 2513 0 ° 8B 1 403 0 0 -4B 6 ‘The optimal solution shown in Table 11.6 is still non-integer. Therefore, one more fractional cut should be added. This time consider the x,-row for constructing the cut 2, 0, siminie tty = The factoring x,-row yields (1+0)5,+ (042) +(-442 67 (+4) or 2x, 425g, 24(Onre-arranging the terms in zbove equation, we obiain Gomory’s fractional cut as: 2 FF thyt 3 Cu) Adding this cut to the optimal simplex Table 11.6, the new table so obtained isshown in Table 11.7. ‘Table 11.7 9° 2 0 -0 0 o o 4 Variables Solution |x, x, x 4 5 a ‘6 trict ite i 9 % 1 1 cn B bex) 20 % 0 ° 1 oo =1 ° 2 x 103 1 o © ~1083 © 0 5, 253 o 0 88 1-408 o 0 ig -3 |o 0 G@ o -» 1 Z=203 67% 0 0-348 0 =4073 ° Ratio: (6-3) [— a 0 Remove the variables, from the basisand enter variable x, into the basis by applying the dul simplex method. ‘The new solution is shown in Table 11.8 Table 1.8 2 2 -10 0 0 o cp Variables xO a Se, Soy in Basis zB 20 % 0 1 0 0 1 0 2 x 1 0 0 0 -4 ° o s 0 0 0 1-16 4 -10 % 2 ° o 1 0 1-32 Z=1 6-3 0 0 0 0 -2-17 ‘The optimal solution shown in Table 11.8 is still non-integer because variable x, does not assume integer value. Thus, a third fractional cut will have to be constructed with the help of ¢,-row: The required Gomory’s fractional cut obtained: Adding this cut to the bottom of the optimal simplex Table 11.8, the new table so obtained is shown in Table 11.9, XyTOW Ixy + 5, = 1+ 0)x +0405, + (2+ “2a +5, (Cutty,‘Table 11.9 6 2 0 -D 0 0 9 96 gp Variables — Solution HOR 8 Sf Se in Basis Values ' “ ° 6 ex) 20 8 ° o 0. 0 o ft 0 0 2 ie 3 1 0 0 0 - 1 0 ° 5 1 o 0 0 1 -6 4 0 -10 5 7 oo 1 0 1-32 ° ° " “1 o 0 09 0 0 GA 1 9-3 0 D 0 0 20-17 0 Ratio: (= 3Myy {sO} — = Remove the variable 5, from the basis and enter variable s,, into the basis by applying the dual simplex ‘method. The new solution is shown in Table 11.10. ‘Table 11.10 g> 2 0-10 o 0 o 0 Variables Solution | x, 7 in Basis Values ' ° * B bay) 20 * ° ° 1 ° 0 1 oO 2 x 2 1 0 0 0-4 o 2 ° 5 3 0 0 0 1-16 0 8 10 2 0 0 1 0 1 @ 3 ° 1 0 o o o o 1-2 Z=-16 S 0 0 0 0 -2 0-34 Since all variables in Table 11.10 have assumed integer values and all c,~z, <0, the solution solution: x; ~2,x,—0,.r,—2andMaxZ=~ 16 Example 11.3 Solve the following integer linear programming problem Maximize 2= Se, +75, subject to the constraints @ -2y +3x, 56, ( 6x, +x, $30, and xs) 20 and integer. [Devalbagh Ed. Inst, M.Tech, 2003] Solution Expressing the integer LP problemin its standard form as follows Maximize Z= 51, +7) + 0s, + 05, subject to the constraints @ =, +356, Gi) Gr, +x, + 5-30, and yop Sp Sy 2 0 and integer ‘The optimal solution to the problem, ignoring integer requirement using simplex method is shown in Table AL‘The non-integer optimal solution shown in Table 11.11 is: x, = 21/5,x,=24/5, and Max. Variables in Basis ‘Table 11.11 OptimalNon-integer Solution 7 67 5 Solution % % Vues b(=x,) ws ° 1 2u5 1 « y 5 7 Z ° « ano 3737 3720 73/5. Since this solution does not satisfy the integer solution requirement, we proceed to construct Gomory's constraint. For this, first selecting x,-row which has the larger fractional part, 4/5. Each of non-integer coefficients are factored into integer and fractional components in such a manner that fractional parts are postive. Accordingly, xpTow is writenas 3 1 4 xytow 3 (1+0)x,+ (a) + ope a445, On rearranging this equation as discussed before, we obtain the Gomory’s fractional cut as: 3.14 4 301 __4 jot 25 or ~ Fone ts, =~ FC, Adding this additional constraint at he bottom of optimal simplex Table 11.11. The new values so obtained areshown in Table 11.12 Table 11.12 Optimal Non-integer Solution 5 7 0 o 0 Variables x yy 5 a in Basis B a ° 1 ano mo 0 x 1 0-120 3200 1 ° 0 -0 1 22185 3 7-370 = Roo o 0-370 -2920 0 Ratio: 2 - 37% rn Since basic variable ,, is negative, in Table 11.12, solution is optimal but infeasible. Applying dual simplex ‘method as usual o remove infeasibility. Enter non-basic variable s in the basis to replace basic variable ‘The new solution is shown in Table 11.13. ‘Table 11.13 Optimal Nop-integer Solution 5 Variables x a Sy 58) in Basi Gy B 1 2B ° 1 ° 0 1 5 x t D 0 6-16 o ° ° 1 13-106 71493 3 7 0 36 ~«OTI6 ° D 0 56 316image not availablehours, respectively in stitching one Zee-shirt, and 4 hours and 3 hours, respectively in shirt, How many shirts of both types should be stitched in orderto maximize daily profit? (@) Formulate and sotve this problem as an LP problem. (b) Ifthe optimal solution is not integer-valued, use Gomory technique to derive the optimal integer solution [Delhi Univ., MBA, 1995] Mathematical formulation Let x, and x, = number of Zee-shirts and Button-down shirts to be stitched daily, respectively. ‘Thea the LP model of the LP problem is stated as: Maximize Z= 3x, + 12x, subject to the constraints @ 2x, 44x, <7[Time with Tailor A], (ii) Sx, +3x, $15 [Time with Tailor B] and Xx, 20 andiinteger Solution (2) Adding stack variables 5, and s», the given LP problems stated into its standard form as: Maximize Z=3x, +12x,+0s, +08, subject to the constraints @ 2x, F4xy+5,=7, (ii) 5x, +32, +5)= 15 and yp Sys) 20 ‘The optimal solution of the LP problem, obtained by using the simplex method is given in Table 11.16. Table 11.16 Optimal Non-integer Solution ge 3 R o o oe Yeriables Solution |x * 1 inBasis Values B dex) 2 x m | 1 4 o 0 304 | 72 ° -3i4 1 6-3 | -3 0 =3 ° The non-integer optimal so 21, = 0, x)= 7/4 and Max Z=21. (6) To obtain the integer-valued solution, we proceed to construct Gomory’s fractional cut, with the help of prow as follows: c 1 1? (out ‘ont ig? xppw > pay tayt gang or [OFS] + +)x+ (042) s= (145 1 3 Therequired Gomory’s factional cutis: — 5 xj 48, F4p,-—Z (Cut) ‘Add this additional constraint to the bottom of the optimal simplex Table 11.16. The new table so obtained is shownin Table 11.17. ‘Table 11.17 67 3 2 o 0 o * % Ss 5 wo rr 0 0 m 0 -34 1 0 0 -14 0 1 2221 o-4 ° 3 ° ° Rats {i syy EO) | 6 — 2 = =aa You have either reached a page that is unavailable for viewing or reached your viewing limit for this book.aa You have either reached a page that is unavailable for viewing or reached your viewing limit for this book.Integer Linear Programming FeTag Table 11.22 5> - 1 3 0 9 oo ce Variables Solution 48 8 8 Fy inBasis, ‘Values B bexp 1 % 37 “718 0 3 4877 Cr ee ee 3 x 137 yu 0 61 Oo 0 5, -o7 =| -3 0 00-370 1 a Dd 22447 5» | 0 0-7 0-570 Ratio: (6) s¥g( 2b 8 0 0 0 90 Variables Solution AOS 8 8 Sy inBasis —Yalues B bea) He “12 83 5 x 1 8 1 -52 Since all ¢ ~z,$0 aswell as the variable x has assumed integer value, the optimal mixed integer solution is: 2) =0,2)= 8/7, x, | and Max Z=2917, Example 11.6 Solve the following mixed-integer programmingproblem Maximize Z: subject to the constraints @ 3x +2455; G42 and yo) 05x; an integer Solution After converting the given LP problem into its standard form, obtain an optimal solution ignoring the integer restriction on.x; by the simplex method. The optimal solutionis given in Table 11.24. ‘Table 11.24 Optimal Non-integer Solution 47 1 ’ o o os Variables Solution | » " % in Basis Kalues 5 bea) | H 18 1 0 1 = as 1 2 o 1 o 1 ° 0 1B Fy Bema ge | BIE all ¢ ~ 2; 0,the optimal non-integer solution is:x, = 1/3,,=2 and Max,In the current optimal solution the variable x, which is restricted to take integer value is not an integer, therefore, we generate Gomory eut considering x,-row as follows: 1 1 z a= ai + (fissi + figs) oF teme(tn -3) Since the coefficient ofs,ispositive, applying rule weget f3 = aL Sy 2 agi ay 0 Also the coefficientof is negative, applying rule, we get 2 ot _j18 fa ( Vs {att ‘Thus, Gomory’s mixed integer cut becomes: +8 a tn +5, = i 3 (Mixed integer cutt wheres, is Gomory’s slack variable, Inirodvcing this Gomory cut atthe botiom of Table 11.24, the new table soobtainedis shown in Table 11.25. Table 11.25 1 1 o o o oe Variables HOR 3 a Ap in Bosie B 1 x 1 nn) 0 1 % ° 1 0 1 0 ° Sey ° ° G&® -~ 1 ° eT) ° Ratle: (6 — 5) ©) = - 1 1 = Applying the dual simplex method, we obtain the revised solution as shown in Table 11.26. Table 11.26 67 1 1 ° o o os Variables ‘Solution % A in Basis Values B 1 x 0 1 0 oA 1 1 % 2 o 1 ° 1 ° ° s 1 ° o 1 133 Since all c - z) $0, therefore the required mixed integer optimal solution given in Table 11.26 is: 3170.3) 2,5,= Tand Max Z=2. Example 1.7 Solve the following mixed-integer programming problem Maximize 2=4r, + 6%, +2, subject to the constraints 0 4x-4% 55, (i) -x, +645, (ii) =x, +2, +3555 and py ay 2 05 4,25 integer.Integer Linear Programming [ET Solution The optimal solution to the mixed-integer programming integer restriction on.x, and.x, variables is given in Table 11.27 ‘Table 11.27 Optimal Non-integer Solution 9? 4 6 2 0 o 0 Variables Solution Ho 4 5 8 % inBasis Values & B bea) 4 % 52 1 ° 0 30s 0 6 x sia 0 l 0 10 us 0 o 53 2514 ° o ee) 0 LU 2-352 6 0 32 2 0 g-y ° 0-32 72 ° ‘The optimal basic feasible solution shown in Table 11.27 is: = 5/2, x, = Oandmax Z=35/2 Since variable x, restricted to take integer values is notan integer, we generate Gomory cut considering x,-row as follows. ‘1895;4(02)yaf¢el |g ere 10 5 Bo to yt 1 Batty 2d or - =-4 [aetorile ff =a, (2,0) ieghes iG 3 (tone 440) ‘Table 11.28 Optimal Non-integer Solution 5° 46 2 9 9 0 0 Variables Solution Hoe 8 5 8B InBasis Values & Bo be) 4 x 52 1 0 0 3% 1S 0 0 6 % 54 o 1 0 12 Is 0 0 2 % 2514 o 0 1 mM 0 1 0 oy a2 o 0 (Oo “3 0 1 7352 3, 2 2 2 2 6-4 60 0 09 2 2 Ratio: (Gg -2/ yO} - = - mm Apply the dual simplex method, to obtain the revised solution as shown in Table 11.29 ‘Table 11.29 Optimal Non-integer Solution so 4 62 Ot Yaiables— Sauion “Jy a 2 5 a inBasty Talus eben ar 2 [|i e© 0 0 0 © 4 | 6 og wi o|o 1 0 0 Geo w 2 # sie | 0 6 1 0 6 1 56 oF so | 08 0 1 80-08 7-608 | 2 8 8 -83aa You have either reached a page that is unavailable for viewing or reached your viewing limit for this book.aa You have either reached a page that is unavailable for viewing or reached your viewing limit for this book.aa You have either reached a page that is unavailable for viewing or reached your viewing limit for this book.aa You have either reached a page that is unavailable for viewing or reached your viewing limit for this book.aa You have either reached a page that is unavailable for viewing or reached your viewing limit for this book.aa You have either reached a page that is unavailable for viewing or reached your viewing limit for this book.aa You have either reached a page that is unavailable for viewing or reached your viewing limit for this book.aa You have either reached a page that is unavailable for viewing or reached your viewing limit for this book.{in department 11 and are sold on @ weekly basis. The weekly Production cannot exceed 25 refrigerators in department I and 35 ranges in department II because of the limited facilites in the departments. The company regularly employs a total of 50 workers inthe departments. A refrigerator requires twoman-weeks of labour and arange oneman-week. A refrigerator contributes aprofitof Rs. 300 and a range of Rs 200, Formulate ard solve this preblem as an integer LP problem to determine the unitsofrefrigeratorsand ranges that the company should produce to realize maximum profit? 5. The XYZ Company produces two types of tape recorders: A- reel-to-reel model and a cassette model, on two assembly lines. ‘The company must process each tape recorder on each assembly line and it has found thatthe following time is required: Assembly Line Reel-t-reel Cassette 6 hours 2 hours 2 3 hours 2 hours ‘The productionmanager says that line I will be available 40hours per week and line 2 only 30 hours per week. After these hours of ‘operation each line must be checked for repairs, The company realizes a profit of Rs. 300 on each reel-to-reel tape recorder and Rs, 120 oneacheassette recorder. Formulate and solve this problem as an integer LP problem to determine the numberof recorders of ‘each type to be produced each week in order to maximize profit ‘A company manufacturing metal products is planning to buy any of the following three types of lathe machines—manual, semi- ‘automatic and fully-automatic. The manual lathe machine costs Rs. 1,500, while the semi-automatic and fally-automa machines cost Rs. 4,000 and Rs. 6,000 respectively. Thecompany's budget for buying new machines is Rs. 1,20,000. The estimated ‘contribution towards profit from the manual, semi-automatic and fully-automatic lathe machines are: Rs. 16, Rs. 20 and Rs, 22, reapectively. The avsilable floor apoce allows forthe installation of ‘only 36 new lathe machines. The maintenance required on fully ‘automatic machines is low and the maintenance department can maintain $0 fully-automatic machines in a year. The maintenance ‘of asemi-cutomatic machine takes 20 per cent more time than that of a fully automatic machine and a manually-operated machine takes 50 per ceat more time for maintenance of a fully-automatic machine. Formulate and solve this problem as an integer LP problem to determine the optimal number of machines to be bought, A manufeeturing company produces two types of screws— metal and wooden. Each screw has to pass through slotting and threading machines. The maximum time that each machine ean be run is 150 hours per month. A batch of SO wooden screws requires 2 min on the thresding machine and 3 min on the slotting machine, Metal screws of the same batch size require 8 min on the threading machine and 2 min on the slotting machine. The profi contribution for exch batch of wooden and metal screws is Re. 1,00 and Rs. 2.50, respectively. Determine the optimal product mix for maximum profit contribution. AA dietician for a hospital is considering a new breakfast menu involving oranges and cereal. This breakfast mast meet the minimum requirements forthe Vitamins A and B. The number of milligrams of each ofthese vitamins conteined in a purchasing unit foreach of these foods is as follows: Yitanin Milligrams per Purchasing Minimum Unit of Food Requirement Oranges (doz) Cereal (bo) ie) A 1 2 20 B 3 2 0 ‘The costof the food ingredients iss. 15 per dozen for oranges and Rs. 12.50 per box for cereal, For dietary reasons at least one unit, ‘ofeach food type must be used in the menu plan, Formulate and solv this problemas an integer programming problem. 9. building contractor has just won a contact io build a munici- pal library building. His present labour workforce is inadequate to take up this work immediately as he has already got other jobs on hand and, therefore, has to decide immediately whether to hire one or more labourers on a full-time basis (cight hours a day each) or to allow overtime to one or more ofthe existing labour fore (five hours a day each), Extra labourers can be hired for Rs. 40 per day (for eight hours) while over-time costs Rs. 43 per day (for five hours pet day). The contractor wants to limit his extra payment to Rs. 400 per day and to use 10 more than twenty labourers (both fulktime and overtime) because of limited supervision. He estimates thet the new labour employed (om a full-time basis will generate Rs. 15 a day in profits, while overtime labour Rs. 20 a day. Formulate and solve tis problem as an integer liner programming problem and help the building contractor to decide the optimum labour force. 10. The ABC company requires an output of at least 200 units of f product per day and to accomplish this target it can buy ‘machines A or B or both. Machine A costs Rs. 20,000 while ‘machine B costs Rs. 15,000 and company has a budget of Rs. 2,00,000 for the stme, Machines A end B will produce 24 and 20 units, respectively of this product per day. However, ‘machine A will require a floor space of 12 square feet while ‘machine B will require 18 square feet and the company has a total floorspace of 180 square feet only. Determine the minireum number of machines that should be purchased. (A nor-integer solution for this problem will not be accepted.) [Delhi Univ, MBA, 1909] 11. A company produces two products A and B. Each unit of product ‘A raquires one hou ofengineeringservices and five hours ofmachine time. To produce one unit of product B, two hours of engineering ‘and 8 hours of machine time are needed. There are 100 hours of ‘engineering and 400 hours of machine time available. The cost of production is a non-linear function of the quantity produced as ‘given inthe following ible: Product A Product B Production Unit Cast ‘Production Unit Cast. (units) a) (unis) (s) 0-50 10 0-40 7 50-100 8 40-100 3 ‘The unitselling price of product A is Rs. 12 and of product B isRs. 14, The company would like a production plan which gives theaa You have either reached a page that is unavailable for viewing or reached your viewing limit for this book.aa You have either reached a page that is unavailable for viewing or reached your viewing limit for this book.aa You have either reached a page that is unavailable for viewing or reached your viewing limit for this book.aa You have either reached a page that is unavailable for viewing or reached your viewing limit for this book.aa You have either reached a page that is unavailable for viewing or reached your viewing limit for this book.aa You have either reached a page that is unavailable for viewing or reached your viewing limit for this book.aa You have either reached a page that is unavailable for viewing or reached your viewing limit for this book.aa You have either reached a page that is unavailable for viewing or reached your viewing limit for this book.aa You have either reached a page that is unavailable for viewing or reached your viewing limit for this book.aa You have either reached a page that is unavailable for viewing or reached your viewing limit for this book.aa You have either reached a page that is unavailable for viewing or reached your viewing limit for this book.Table 12.7 so 0 ° SP, 3, P, eg Variables Solution * * dj 45 PA dj inBasis Values B 0 5 0 I 1 1 0 0 o x 70 1 > ° ' ° ° Bd as 0 o -1 ' ' 0 P, G 10 0 D 0 ° It Py 0 7 = 1 0 - = », 0s - — 3 o - - gy 0 = 7 0 0 - _ P, o 7 = 1 ° = - It can be seen that the Table 12.7 presents an optimal solution, because all c) ‘or zero inthe P,-row. Thus, the second goal (P.,)is fully achieved. ‘The solution can be read from the solution basis: x, = 70,x,= 10, d; Values are either positive 36 and dy = 10. REVIEW QUESTIONS 1. Whatis goal programming? Clearly state its assumptions, Compare the differen and goal programming. 3._ Explain the following terms (Deviational variables (ii) Pre-empiive priority factors, Explain the differences in solving a linear programming versus goal programming problem by the simplex method, ‘5. Whatis goal programming? Why areall goal programming problems 2 /similarities between linece programming 1 inimization problems? Why does altering the goal priorities resalt in a different solution toa problem? Explain What is meant by “satisfying” and why is the term often used in ‘ccnjunction with goal programming? W nat are dev ional variables? How do they differ from docision ‘va ablesin traditional linear programming problems? Woat does it mean to rank goals in goal programming? How does, thi, affect tho problem's solution? SELF PRACTICE PROBLEMS 1, Anoffice eyuipmentmanufacturer produves two kinds of product chairs and lamps, Production of either a chair ora lamprequires | ‘our of production eapacity in he plant The plant has a maximum troduction eapacity of 10 hours per week. Because of the limited ‘ales capacity, the maximum number of chairs and lamps that can be sold are and 8 per week, respectively. The gross margin from the sale ofa chair is Rs. 80 and Rs. 40 for that of 2 lamp. The plant manager hasset the following goals aranged inorder ‘of importance: (He wants to avoid any underutiization of production capacity. (i) He wants to sell as many chairs and lamps as possible Since the gross margin from the sale of a chair is set at twice the amount of profit from a lamp, ie has twice as much desire to achieve the sales goal for chairs as for lamps. He wants to minimize overtime operation ofthe plant as ‘much as possible. Gin Formulate and solve this problem as a GP problem, so that the plant maragermakes a decision tht will achieve his goals asclosely, as possible [Delhi Unix, MBA, 1999] A production manager faces the problem of job allocation among, thre: of his teams. The processing rates of the three teams are 5,6 and 8 units per hour, respectively. The normal working hours for each team is 8 hours per day. The production manager has the following goals forthe next day in order of priority (i) The manager wants to avoid any underachievement of production level, which is set at 180 units of product. (ii) Any overtime operation of team 2 beyond two hours and te1m3 beyond three hours should be avoided, (ii). Minimize the sum of overtime. Formulate and solve this problem as goal programming problem. XYZ company preduces two products—record players and tape- recorders. The production of both products is done in wo separate ‘machine conties within the plant. Each record player requires twoEPa Operations Research: Problems and Solutions hours in machine centre A and onehourin machine centre B. Each tape reconer, onthe other hand, requires one hour inthe machine centie A and three hours in machine centre B. In addition, cach product requires some in-process inventory. The per-unit in- process inventory required is Rs. 50 for the record player and Rs, 30 forthe tpe recorder. The firm has normal monthly operation hours of 120 for both machine centres A and B. The estimated profit per unitis Rs. 100 for the record player and Rs. 75 for the tape recorder. According to the marketing department, the forecast ‘of sales for the record player andthe tape recorder are 50 and 80, respectively fer the coming month. ‘The president ofthe firm has established the following goals for production in the next month, in ordinal rank of importance: Limit the amount ted up in in-process inventory forthe rmonth to Rs. 4,500. (i) Achieve the sales goa! of 80 tape recorders for the month (Gi). Limit the overtime operation of machine centre A to 20 hours (Gv) Achieve th sale goal of $0 record players forthe month (©) Limit the sum of overtime operation for both machine centres. (i) Avoid any underutiization of regular operation hours of both machine centres. Formulate andsolve thisprobicm as goal programming problem. 4, ABC Fumitures produce three products, tables, desks and chair. The furmiure i produced in the central plant. The production of the desk requires 3 hours in the plant, the Table 2 hours and the chair only 1 hour. The regular plant capacity is 40 hours a week. ‘According tothe marketing department, the maximum number of desks. tables and chairs that can be sold weekly are 10. 10 and 12. respectively. The president ofthe firm has established the following goals according to their importance: {i) Avoid any underutiization of production capacity. (i) Meet the order of XYZ. Store for seven desks and five chairs (Gi) Avoid overtime operation ofthe plant beyond 10 hours (is) Achieve sales goal of 10 desks, 10 tables, and 12 chairs, ()_ Minimize overtime operation as much as possible Formulate andsolve thisproblem asa goal programming problem. [Dethi Univ, MBA, 1995, 1998] 5. ABC Computer Company produces three different types of computers—Epic, Galaxie and Utopia. The production of all computer is condvcted in a complex end modem assembly line The production ofan Epic requires S hours in the assembly line, a Galaxie require 8 hours and a Utopia requires 12 hours, Thenormal operation ours of assembly ine are 170 per month. The marketing and accounting departments have estimated that profits per unit, forthe three types of computers are Rs 1,00,000 forthe Epic Rs 144,000 for the Galaxie and Rs, 2,52,000 forthe Utopia. The ‘marketing department farther reports thatthe demandis such that, the firm can expect to sell all the computers it produces in the ‘month, The chairman ofthe company hss established the following goals listed according to theirimportance: (i) Avoid underutitization of capacity terms of regular hours ‘of operation ofthe assembly line. (i) Meet the demand of the north-eastern sales district for five Epics, five Galaxies, and eight Utopias (diferent ‘weights should be assigned acconting tothe profit ratios among the three types of computer). (i) Limit overime operation of the assembly line to 20 hous. (Gv) Mect theses goal for each ype af compute: Epic—10; Galaxie—12; and Utepia—10 (again assign weights according to te relative profit function for esch compute). (¥) Minimize the total overtime operation of the assembly line Formulate and solve this problem as a gcal programming modal through two iterations (thre tables) by the simplex method. ‘The manager ofthe only recont shop in a town bas a decision problem that involves mutiple goals. The record shop employs five full-time and four par-time salesmen, The normal working hous per month fr a full-time salesman are 160 hours and for & part-time salesman 80houss. According to periormance reco of| the salesmen, the average sales has been five records perhou for full-time salesmen andtwo records per hour for part-time salesmen. ‘The averagehourly woge rates are Rs.3 foc ullsime-salesmenand Rs.2 for parttime salesmen, Average profit from the sles ofa record is Rs. 50. In view of past record of sales, the manager feels thatthe sales goal forthe next month shouldbe 5,500 records. Since the shop isopensix éays a week, overtime is often required ofsalesmen not necessarily overtime butexta hows forthe part- time salesmen). The manager believes that a good employer- cmployes relationship isan essential facior of business success ‘Therefore, he fees thata stable employment level with occasional overtime requirement is a better practice than an unstable enploymentievelwithno overtime. However. he fees thatoverime ‘of more than 100 hours among the full-time salesmen should be avoided because ofthe declining sales effetivencas caused by fatigue “The managerhasset the following goals: (@) The first goal is toachieve a sales target of 5.500 records forthe next month (ii) The second goal isto limit the overtime of falltime salesmen to 100 hours. (ii) The third goals to provide job security to salesmen, The rmarager feels that fll utilisation of employses' regular working hours (no layoffs is an important factor fora goo! employer-employee relationship. Howerer, he is twice as concerned with the full utilization of full-time salesmen as with the fll lization of part-time salesmen, (iv) Tholast goa ist minimize the sum of overtime for both {ulltimeand partsime salesmen. The manager desires to assign differential Weights tothe minimization ofoverime according tothe net marginal profit ratio between the full- time and part-time salesmen. Formulate and solve the given problem as a goal programming problem. hospital administration is reviewing departmental requests prior tothe design of anew emergency room. At issue isthe number of ‘beds for each department. The current plans call for a 15,000 square fet fit. The hospital boardhas established the following sal in order of importance:Department No.of Costper Areaper Peak Beds Bed Bed—_Reguirement Requesied — (incl —(sq-fu)_—_ (Max. no equipment) ofpatienss (Rs.) atone time) A Ss 12600 474 3 B 20 5400 542 is c 20 8600438, Is (Avoid overspending of the budget Rs, 3,00,000, | (i) Avoid pln requiring more than 15,000 39 | (iii) Meet the peak requirement. | tiv) Meet the departmental requirement. | Formulate ar solve (yp to thee erations) the given problem 3s 0a programming problem, | MrX has inherited Rs. 10,0,000andseeksyouradviceconcerting ‘investing his money, You have determined that 10 per ceat can be | eamed na bank account and 14 per centbynvestingincerificates of deposit In real esta, you esimate an annual reum of 15 per eat, while onthe stock market you estimate that 26 per cent san beeamed annual. | MIX his established the fllowing goals in oder ofimportance: | (. Minmize risk theefore he wansto invest not mare than 35 per oat In anyone ype of avi. | | ) Must have Rs, 1,00,000 in the bank account to meet any emergency i) Maximum anual cash retur Formulate ad solve (up to thre iterations onl) thi problem asa goal programming medel that determines the amount of meney ‘which X should invest in each investment option ‘A company plans to schedule is annual advertising eampign. The total advertising budgets sctat Rs. 10,0000. Th firm ean purchase local radio spots at Rs. 2,000 per spot, local television spot at Rs. 12,000 per spot oF magazine aiversing at Rs 4.000 per insertion. The payofT rom each advertising medium sa function ofits audiones sie and audiencecharictersties Latthispayof be 7 5 32 Cost in huncreds of rapees of shipping one million litre from each plant to each distribution centre is given inthe following table: Distribution Centre Dy D; Ds Ds P 2 3 o 7 Plan Py 1 0 6 r P, 5 8 Is 9 Find initial basic feasible solution for given problem by using (a) North-west comer rule; (b) Least cost ‘method; and (¢) Vogel's approximation method if the object is to minimize the total transportation costimage not availableTransportation Problem (©) Fogel’s Approximation Method: The method is self-explanatory as shown in Table 13.4 Table 13.4 By D, Dy Ds Supply Row Penalty Pp © | © j 3 i i oe hat 7, ms * © |o 3 - 4 2 ‘The number of allocated cells in Table 13.4 re six, which is equal to the required number m +n —1 = 6), therefore, this solution is non-degenerate. The transportation cost associated with this solut 8.2 1+ 3X SH1x1+Sx6+15*3+9% 1) x 100=Rs. 10,200 Total cost It can be seen that the total transportation cost found by VAM is lower than the costs of transportation determined by the other two methods. Therefore, itis of advantage to use this method in order to reduce ‘computational time required o obtain optimum solution Example 13.4 Obtain an optimal solution tothe transportation problemby MODI method given in Example 13.1 >, D, Ds >, Sioply s, ]s 0 30 0 7 s, | 7 x 40 0 ° s, | 40 8 70 20 8 Demand | 5 ' 7 “ Fy Solution Applying Vogel’sapproximation method to obtain an initial basic feasible solution. This solution is again reproduced in Table 13.5 for ready reference. ‘Step 1 In Table 13.5, the number of occupied cells are m +n — 1 =3+4— 1 = 6, and initial solution is non- degenerate. Thus, an optimal solution can be obtained. The total transportation cost associated with this solution is Rs. 779, Step 2. Inorder to calculate the values of us (i= 1, 2, 3) and v/s ( arbitrarily assign v, 1,2,3,4)for each occupied cell, we to simplify calculations. Given ,=0, uy, u, and u, can be computed immediately by u, + v, for occupied cells as shown below’ or = 60 or 10-4, +0 or w=10‘Table 13.5 Initial Solution, VAM. Dy D: Ds Ds ‘Supply uy Given u,,1,, and 1,, value of v,, v, and ¥, can also be calculated as shown below: cuemty, or 19=10+y, or y= 9 eg=ytu, or 40=60+v, oF yy =-20 CT WytY, OF 8=2+1 oF ¥,=-12 Step 3 The opportunity cost for each of the occupied cell is determined by using the relation y= ¢,~ (4, + ¥)andis shown below. y= ey.—(u, +p) =30- (10-12) = B dys=¢y3~(u, +¥3)=50-(10-20)= © dyy= ey ~ (ty +¥,)=70-(60+ 9)=1 10 -(60~ 12) 4020+ 9)= 11 33 ~ €33~ (us + ¥3)~ 70-20-20) = 8 ‘Step4 According to the optimality criterion for cost minimizing transporation problem, the current solution is not optimal, since the opportunity costs of the unoccupied cells are not all zero or positive. The value of =~ 18 incell (S,, D,) is indicating that the total transportation cost can be reduced inthe multiple of 18 by shifiing an allocation to this cel. ‘Step 5 A closed loop (path) is traced along row S, to an occupied cell (S,, D,). A plus sign is placed in cell (SqyD,) and minus sign in cell (S,, D,). Now take aright-angle turn and locate an occupied cell in column Dy. An ‘occupied cell (S,, D,)exists atrow S,, and a plus sign is placed in this cell. ‘Continuing the process and complete the closed path. The occupied cel (S,, D;) must be bypassed otherwise it will violate the rules of constructing closed path. ‘Step 6 In order to maintain feasibility, examine the occupied cells with minus sign at the comers of closed loop. and select the one that has the smallestallocation. This determines the maximum number of units that can be shified along the closed path. The minus signs are in cells ($3, D,) and (S>, D,). The cell (Sy,D,)is selected because it has the smaller allocation, ie, 2, The valve of thi allocation is then added to cell (Sy, D,) and (S,, D,), which carry plus signs. The same value is subtracted from cells (S,, D,) and (Sy, D,) because they carry minus signs. Step 7 The revised solution is shown in Table 13.6. The total transportation cost associated with this solution as shown in Table 13.6 is: Rs. 743,Transportation Problem Step & Test the optimality ofthe revised solution once again in the same way as discussed in earlier steps. ‘The values of us, v's and djs are shown in Table 13.6. Since each of +s positive, the current basic feasible soltionis optimal with minimam total transportation cost of Rs. 743. ‘Table 13.6 Optimal Solution D, Ds Ds Ds Supply % 5 [9 30 ® 432 7 5 (no 30 nt +| @ 9 3, ao ® 70 20 = 10 +n| © +n| @ 8 Demand s $ 7 4 3 5 [yet yee ~= 10 Example 13.5 A company has factories at F,, F, and F, which supply to warehouses at W7,, W’, and W, ‘Weekly factory capacities are 200, 160 and 90 units, respectively. Weekly warehouse requirement are 180, 120 and 150 units, respectively. Unit shipping costs (in rupees) are as follows: Warehouse ", Sipps 2 2 200 Factory 8 18 160 38 i6 0 {__ Demand 180 12 150 450 Determine the optimal distribution for this company to minimize total shipping cost. Solution Initial basic feasible solution obtained by North-West Comer Ruleis given in Table 13.7. ‘Table 13.7 Initial Solution w, W, Sipply ] © © Demand 180 120 150 450 tt‘The initial solution has m + n—1=3 +3 -1=Sallocations. Therefore, itis a non-degenerate solution. The optimality test can, therefore, be performed. The total transportation cost associated with this solution is, Total cost = 16 « 180-+20x20+8 x 100+ 18% 60+ 16 90=Rs. 6,800 Determining the values of u's and v,'s as usual by assigning, = 0 arbitrarily. Given u, = 0, the values of others variables so obtained by using the equation, cj, =u, + v, for occupied cells are shown in Table 13.8, Table 13.8 W, , Wy, Sipply 4 +0 5 26 24 16 4-4 %0 14 +18 Demand 180 120 150 450 % yn ie % = 20 30 or 16= 0+ oF or 20= 04%, of or 8= +20 or or 1249, or or 16= u,+30 or u; =-14 The opportunity cost for cach of the unoccupied cells is determined by using the equation, cy (4+ v) as shown below: d, ‘The value of d,, =~ 18 in the cell (F,, M3) indicates that the total transportation cost can be reduced ina multiple of 18 by introducing this cell inthe new transportation schedule. To see how many units of the commodity could be allocated to this cell (route) we shall form a closed path as shown in Table 13.8. ‘The largest number of units of the commodity which should be allocated to the cell F,, IV) is 20 units because it doesnot violate the supply and demand restrictions (minimum allocation among the oecupied cells bearing negative sign at the comers of the loop). The new transportation schedule (solution) so obtained is shown in Table 13.9.aa You have either reached a page that is unavailable for viewing or reached your viewing limit for this book.Table 13.11 ” my, Wy z= © 1 90 Tl a zs i ‘The total transportation cost associated with this solution is Total cost= 16% 140+ 12 « 60+14 x40+8 x 120+16«90=Rs. 5,920 To test the optimality ofthe new solution shown in Table 13.11 again calculate the opportunity cost ofeach unoccupied cell in the same manner as discussed earlier. The calculations are shown in Table 13.12. ‘Table13.12 Optimal Solution ", ", Sipply 4 16 20 2 un 16 +10 200 % 4 8 18 yo +8] 100 yy 26 4 16 = 20 “6 +10 0 Demand 180 120 150 » ya-6 ayy =e2—(u,+¥y) oF 20-(16-6) dy ~ex-(uy t¥) or 18-(14-4) = 8 =cy-(uy+v,) or 26-(20+0) = 6 10 Since none of the unoccupied cells in Table 13.12 has a negative opportunity cost value, therefore, total ‘transportation cost cannot be reduced further. Thus, the solution shown in Table 13.12 is the optimal solution giving optimal trasportation schedule with a total cost of Rs. 5,920. ay yy = ey-(uj+¥,) or 24-(20-6) Example 13.6 The following table shows all the necessary information on the availability of supply to each ‘warehouse, the requirement ofeach market and unit transportation cost (in Rs) from each warehouseto each market. Market Q Supply 2 Warehouse 13 8 5aa You have either reached a page that is unavailable for viewing or reached your viewing limit for this book.aa You have either reached a page that is unavailable for viewing or reached your viewing limit for this book.Transportation Problem & ‘The cost of manufacturing the product at different production shops is, Shop eA B Variable cost 9: 4 16 Is Fixedcost. 7,000 4000 5000 Find the optimum quantity to be supplied from each shop to different warehouses at minimum total cost. [Delhi Univ, MBA, 1997; Nagpur, MBA, 1998] Solution In this case, fixed cost data is of no use, The transportation cost matrix will include the given ‘transportation cost plus the variable cost as shown in Table 13.17, ‘Table 13.17 I a Mt v v Supply 4 6+ 14-20 | ae is—is | aeis-ae | tensa | seis-19 100 B s+i6~21 | 6+ 16-22 | 7416-23 | 4416-20 | 8416-24 125 c sris-is | 4+is—i9 | 6+1s-21 | a+is—as | aris—19 175 Demand oo 80 85 105 0 400 ‘The optimal solution obtained by applying MODI method is shown in Table 13.18. Table 13:18 7 7 ww v ‘Sippy % 4 [2 uu 9 na 8 rr af a [at 20 m 6 4 wp 5 ¢ |i 8 9 aT) 2 " 75 105 70 ¥3=0 y=0 ‘The transportation cost associated with the solution is, Rs. 7,605, Example 13.8 A company has received a contract to supply gravel for three new construction projects located in towns 4, B and C. Construction engineershave estimated the required amounts of gravel which will be needed at these construction projects Plant location DOA B c Weekly requirement: 72 102 4a (Truckloads) The company has’ gravel pits located in owns X; Yand Z. The gravel required by the construction projects can be supplied by three pits. The amount of gravel which can be supplied by eack pit is as follows: Plant 4 x Yy z Amount available (truckloads): 16 82 1 oniseay2aa You have either reached a page that is unavailable for viewing or reached your viewing limit for this book.aa You have either reached a page that is unavailable for viewing or reached your viewing limit for this book.aa You have either reached a page that is unavailable for viewing or reached your viewing limit for this book.aa You have either reached a page that is unavailable for viewing or reached your viewing limit for this book.aa You have either reached a page that is unavailable for viewing or reached your viewing limit for this book.aa You have either reached a page that is unavailable for viewing or reached your viewing limit for this book.aa You have either reached a page that is unavailable for viewing or reached your viewing limit for this book.aa You have either reached a page that is unavailable for viewing or reached your viewing limit for this book.aa You have either reached a page that is unavailable for viewing or reached your viewing limit for this book.aa You have either reached a page that is unavailable for viewing or reached your viewing limit for this book.aa You have either reached a page that is unavailable for viewing or reached your viewing limit for this book.aa You have either reached a page that is unavailable for viewing or reached your viewing limit for this book.aa You have either reached a page that is unavailable for viewing or reached your viewing limit for this book.aa You have either reached a page that is unavailable for viewing or reached your viewing limit for this book.aa You have either reached a page that is unavailable for viewing or reached your viewing limit for this book.aa You have either reached a page that is unavailable for viewing or reached your viewing limit for this book.aa You have either reached a page that is unavailable for viewing or reached your viewing limit for this book.aa You have either reached a page that is unavailable for viewing or reached your viewing limit for this book.aa You have either reached a page that is unavailable for viewing or reached your viewing limit for this book.aa You have either reached a page that is unavailable for viewing or reached your viewing limit for this book.aa You have either reached a page that is unavailable for viewing or reached your viewing limit for this book.aa You have either reached a page that is unavailable for viewing or reached your viewing limit for this book.aa You have either reached a page that is unavailable for viewing or reached your viewing limit for this book.aa You have either reached a page that is unavailable for viewing or reached your viewing limit for this book.aa You have either reached a page that is unavailable for viewing or reached your viewing limit for this book.aa You have either reached a page that is unavailable for viewing or reached your viewing limit for this book.aa You have either reached a page that is unavailable for viewing or reached your viewing limit for this book.aa You have either reached a page that is unavailable for viewing or reached your viewing limit for this book.aa You have either reached a page that is unavailable for viewing or reached your viewing limit for this book.aa You have either reached a page that is unavailable for viewing or reached your viewing limit for this book.aa You have either reached a page that is unavailable for viewing or reached your viewing limit for this book.aa You have either reached a page that is unavailable for viewing or reached your viewing limit for this book.aa You have either reached a page that is unavailable for viewing or reached your viewing limit for this book.aa You have either reached a page that is unavailable for viewing or reached your viewing limit for this book.aa You have either reached a page that is unavailable for viewing or reached your viewing limit for this book.aa You have either reached a page that is unavailable for viewing or reached your viewing limit for this book.aa You have either reached a page that is unavailable for viewing or reached your viewing limit for this book.aa You have either reached a page that is unavailable for viewing or reached your viewing limit for this book.aa You have either reached a page that is unavailable for viewing or reached your viewing limit for this book.aa You have either reached a page that is unavailable for viewing or reached your viewing limit for this book.aa You have either reached a page that is unavailable for viewing or reached your viewing limit for this book.aa You have either reached a page that is unavailable for viewing or reached your viewing limit for this book.aa You have either reached a page that is unavailable for viewing or reached your viewing limit for this book.aa You have either reached a page that is unavailable for viewing or reached your viewing limit for this book.aa You have either reached a page that is unavailable for viewing or reached your viewing limit for this book.aa You have either reached a page that is unavailable for viewing or reached your viewing limit for this book.aa You have either reached a page that is unavailable for viewing or reached your viewing limit for this book.aa You have either reached a page that is unavailable for viewing or reached your viewing limit for this book.aa You have either reached a page that is unavailable for viewing or reached your viewing limit for this book.aa You have either reached a page that is unavailable for viewing or reached your viewing limit for this book.aa You have either reached a page that is unavailable for viewing or reached your viewing limit for this book.aa You have either reached a page that is unavailable for viewing or reached your viewing limit for this book.aa You have either reached a page that is unavailable for viewing or reached your viewing limit for this book.aa You have either reached a page that is unavailable for viewing or reached your viewing limit for this book.aa You have either reached a page that is unavailable for viewing or reached your viewing limit for this book.aa You have either reached a page that is unavailable for viewing or reached your viewing limit for this book.aa You have either reached a page that is unavailable for viewing or reached your viewing limit for this book.aa You have either reached a page that is unavailable for viewing or reached your viewing limit for this book.aa You have either reached a page that is unavailable for viewing or reached your viewing limit for this book.aa You have either reached a page that is unavailable for viewing or reached your viewing limit for this book.aa You have either reached a page that is unavailable for viewing or reached your viewing limit for this book.aa You have either reached a page that is unavailable for viewing or reached your viewing limit for this book.aa You have either reached a page that is unavailable for viewing or reached your viewing limit for this book.aa You have either reached a page that is unavailable for viewing or reached your viewing limit for this book.aa You have either reached a page that is unavailable for viewing or reached your viewing limit for this book.aa You have either reached a page that is unavailable for viewing or reached your viewing limit for this book.aa You have either reached a page that is unavailable for viewing or reached your viewing limit for this book.aa You have either reached a page that is unavailable for viewing or reached your viewing limit for this book.aa You have either reached a page that is unavailable for viewing or reached your viewing limit for this book.aa You have either reached a page that is unavailable for viewing or reached your viewing limit for this book.aa You have either reached a page that is unavailable for viewing or reached your viewing limit for this book.aa You have either reached a page that is unavailable for viewing or reached your viewing limit for this book.aa You have either reached a page that is unavailable for viewing or reached your viewing limit for this book.aa You have either reached a page that is unavailable for viewing or reached your viewing limit for this book.aa You have either reached a page that is unavailable for viewing or reached your viewing limit for this book.aa You have either reached a page that is unavailable for viewing or reached your viewing limit for this book.aa You have either reached a page that is unavailable for viewing or reached your viewing limit for this book.aa You have either reached a page that is unavailable for viewing or reached your viewing limit for this book.aa You have either reached a page that is unavailable for viewing or reached your viewing limit for this book.aa You have either reached a page that is unavailable for viewing or reached your viewing limit for this book.aa You have either reached a page that is unavailable for viewing or reached your viewing limit for this book.aa You have either reached a page that is unavailable for viewing or reached your viewing limit for this book.aa You have either reached a page that is unavailable for viewing or reached your viewing limit for this book.aa You have either reached a page that is unavailable for viewing or reached your viewing limit for this book.aa You have either reached a page that is unavailable for viewing or reached your viewing limit for this book.aa You have either reached a page that is unavailable for viewing or reached your viewing limit for this book.aa You have either reached a page that is unavailable for viewing or reached your viewing limit for this book.aa You have either reached a page that is unavailable for viewing or reached your viewing limit for this book.aa You have either reached a page that is unavailable for viewing or reached your viewing limit for this book.aa You have either reached a page that is unavailable for viewing or reached your viewing limit for this book.aa You have either reached a page that is unavailable for viewing or reached your viewing limit for this book.aa You have either reached a page that is unavailable for viewing or reached your viewing limit for this book.aa You have either reached a page that is unavailable for viewing or reached your viewing limit for this book.aa You have either reached a page that is unavailable for viewing or reached your viewing limit for this book.aa You have either reached a page that is unavailable for viewing or reached your viewing limit for this book.aa You have either reached a page that is unavailable for viewing or reached your viewing limit for this book.aa You have either reached a page that is unavailable for viewing or reached your viewing limit for this book.aa You have either reached a page that is unavailable for viewing or reached your viewing limit for this book.aa You have either reached a page that is unavailable for viewing or reached your viewing limit for this book.aa You have either reached a page that is unavailable for viewing or reached your viewing limit for this book.aa You have either reached a page that is unavailable for viewing or reached your viewing limit for this book.aa You have either reached a page that is unavailable for viewing or reached your viewing limit for this book.aa You have either reached a page that is unavailable for viewing or reached your viewing limit for this book.aa You have either reached a page that is unavailable for viewing or reached your viewing limit for this book.aa You have either reached a page that is unavailable for viewing or reached your viewing limit for this book.aa You have either reached a page that is unavailable for viewing or reached your viewing limit for this book.aa You have either reached a page that is unavailable for viewing or reached your viewing limit for this book.aa You have either reached a page that is unavailable for viewing or reached your viewing limit for this book.aa You have either reached a page that is unavailable for viewing or reached your viewing limit for this book.aa You have either reached a page that is unavailable for viewing or reached your viewing limit for this book.aa You have either reached a page that is unavailable for viewing or reached your viewing limit for this book.aa You have either reached a page that is unavailable for viewing or reached your viewing limit for this book.aa You have either reached a page that is unavailable for viewing or reached your viewing limit for this book.aa You have either reached a page that is unavailable for viewing or reached your viewing limit for this book.aa You have either reached a page that is unavailable for viewing or reached your viewing limit for this book.aa You have either reached a page that is unavailable for viewing or reached your viewing limit for this book.aa You have either reached a page that is unavailable for viewing or reached your viewing limit for this book.aa You have either reached a page that is unavailable for viewing or reached your viewing limit for this book.aa You have either reached a page that is unavailable for viewing or reached your viewing limit for this book.aa You have either reached a page that is unavailable for viewing or reached your viewing limit for this book.aa You have either reached a page that is unavailable for viewing or reached your viewing limit for this book.aa You have either reached a page that is unavailable for viewing or reached your viewing limit for this book.aa You have either reached a page that is unavailable for viewing or reached your viewing limit for this book.aa You have either reached a page that is unavailable for viewing or reached your viewing limit for this book.aa You have either reached a page that is unavailable for viewing or reached your viewing limit for this book.aa You have either reached a page that is unavailable for viewing or reached your viewing limit for this book.aa You have either reached a page that is unavailable for viewing or reached your viewing limit for this book.aa You have either reached a page that is unavailable for viewing or reached your viewing limit for this book.aa You have either reached a page that is unavailable for viewing or reached your viewing limit for this book.aa You have either reached a page that is unavailable for viewing or reached your viewing limit for this book.aa You have either reached a page that is unavailable for viewing or reached your viewing limit for this book.aa You have either reached a page that is unavailable for viewing or reached your viewing limit for this book.aa You have either reached a page that is unavailable for viewing or reached your viewing limit for this book.aa You have either reached a page that is unavailable for viewing or reached your viewing limit for this book.aa You have either reached a page that is unavailable for viewing or reached your viewing limit for this book.aa You have either reached a page that is unavailable for viewing or reached your viewing limit for this book.aa You have either reached a page that is unavailable for viewing or reached your viewing limit for this book.aa You have either reached a page that is unavailable for viewing or reached your viewing limit for this book.aa You have either reached a page that is unavailable for viewing or reached your viewing limit for this book.aa You have either reached a page that is unavailable for viewing or reached your viewing limit for this book.aa You have either reached a page that is unavailable for viewing or reached your viewing limit for this book.aa You have either reached a page that is unavailable for viewing or reached your viewing limit for this book.aa You have either reached a page that is unavailable for viewing or reached your viewing limit for this book.aa You have either reached a page that is unavailable for viewing or reached your viewing limit for this book.aa You have either reached a page that is unavailable for viewing or reached your viewing limit for this book.aa You have either reached a page that is unavailable for viewing or reached your viewing limit for this book.aa You have either reached a page that is unavailable for viewing or reached your viewing limit for this book.aa You have either reached a page that is unavailable for viewing or reached your viewing limit for this book.aa You have either reached a page that is unavailable for viewing or reached your viewing limit for this book.aa You have either reached a page that is unavailable for viewing or reached your viewing limit for this book.aa You have either reached a page that is unavailable for viewing or reached your viewing limit for this book.aa You have either reached a page that is unavailable for viewing or reached your viewing limit for this book.aa You have either reached a page that is unavailable for viewing or reached your viewing limit for this book.aa You have either reached a page that is unavailable for viewing or reached your viewing limit for this book.aa You have either reached a page that is unavailable for viewing or reached your viewing limit for this book.aa You have either reached a page that is unavailable for viewing or reached your viewing limit for this book.aa You have either reached a page that is unavailable for viewing or reached your viewing limit for this book.aa You have either reached a page that is unavailable for viewing or reached your viewing limit for this book.aa You have either reached a page that is unavailable for viewing or reached your viewing limit for this book.aa You have either reached a page that is unavailable for viewing or reached your viewing limit for this book.aa You have either reached a page that is unavailable for viewing or reached your viewing limit for this book.aa You have either reached a page that is unavailable for viewing or reached your viewing limit for this book.aa You have either reached a page that is unavailable for viewing or reached your viewing limit for this book.aa You have either reached a page that is unavailable for viewing or reached your viewing limit for this book.aa You have either reached a page that is unavailable for viewing or reached your viewing limit for this book.aa You have either reached a page that is unavailable for viewing or reached your viewing limit for this book.aa You have either reached a page that is unavailable for viewing or reached your viewing limit for this book.aa You have either reached a page that is unavailable for viewing or reached your viewing limit for this book.aa You have either reached a page that is unavailable for viewing or reached your viewing limit for this book.aa You have either reached a page that is unavailable for viewing or reached your viewing limit for this book.aa You have either reached a page that is unavailable for viewing or reached your viewing limit for this book.aa You have either reached a page that is unavailable for viewing or reached your viewing limit for this book.aa You have either reached a page that is unavailable for viewing or reached your viewing limit for this book.aa You have either reached a page that is unavailable for viewing or reached your viewing limit for this book.aa You have either reached a page that is unavailable for viewing or reached your viewing limit for this book.aa You have either reached a page that is unavailable for viewing or reached your viewing limit for this book.aa You have either reached a page that is unavailable for viewing or reached your viewing limit for this book.aa You have either reached a page that is unavailable for viewing or reached your viewing limit for this book.aa You have either reached a page that is unavailable for viewing or reached your viewing limit for this book.aa You have either reached a page that is unavailable for viewing or reached your viewing limit for this book.aa You have either reached a page that is unavailable for viewing or reached your viewing limit for this book.aa You have either reached a page that is unavailable for viewing or reached your viewing limit for this book.aa You have either reached a page that is unavailable for viewing or reached your viewing limit for this book.aa You have either reached a page that is unavailable for viewing or reached your viewing limit for this book.aa You have either reached a page that is unavailable for viewing or reached your viewing limit for this book.aa You have either reached a page that is unavailable for viewing or reached your viewing limit for this book.aa You have either reached a page that is unavailable for viewing or reached your viewing limit for this book.aa You have either reached a page that is unavailable for viewing or reached your viewing limit for this book.aa You have either reached a page that is unavailable for viewing or reached your viewing limit for this book.aa You have either reached a page that is unavailable for viewing or reached your viewing limit for this book.aa You have either reached a page that is unavailable for viewing or reached your viewing limit for this book.aa You have either reached a page that is unavailable for viewing or reached your viewing limit for this book.aa You have either reached a page that is unavailable for viewing or reached your viewing limit for this book.aa You have either reached a page that is unavailable for viewing or reached your viewing limit for this book.aa You have either reached a page that is unavailable for viewing or reached your viewing limit for this book.aa You have either reached a page that is unavailable for viewing or reached your viewing limit for this book.aa You have either reached a page that is unavailable for viewing or reached your viewing limit for this book.aa You have either reached a page that is unavailable for viewing or reached your viewing limit for this book.aa You have either reached a page that is unavailable for viewing or reached your viewing limit for this book.aa You have either reached a page that is unavailable for viewing or reached your viewing limit for this book.aa You have either reached a page that is unavailable for viewing or reached your viewing limit for this book.aa You have either reached a page that is unavailable for viewing or reached your viewing limit for this book.aa You have either reached a page that is unavailable for viewing or reached your viewing limit for this book.aa You have either reached a page that is unavailable for viewing or reached your viewing limit for this book.aa You have either reached a page that is unavailable for viewing or reached your viewing limit for this book.aa You have either reached a page that is unavailable for viewing or reached your viewing limit for this book.aa You have either reached a page that is unavailable for viewing or reached your viewing limit for this book.aa You have either reached a page that is unavailable for viewing or reached your viewing limit for this book.aa You have either reached a page that is unavailable for viewing or reached your viewing limit for this book.aa You have either reached a page that is unavailable for viewing or reached your viewing limit for this book.aa You have either reached a page that is unavailable for viewing or reached your viewing limit for this book.aa You have either reached a page that is unavailable for viewing or reached your viewing limit for this book.aa You have either reached a page that is unavailable for viewing or reached your viewing limit for this book.aa You have either reached a page that is unavailable for viewing or reached your viewing limit for this book.aa You have either reached a page that is unavailable for viewing or reached your viewing limit for this book.aa You have either reached a page that is unavailable for viewing or reached your viewing limit for this book.aa You have either reached a page that is unavailable for viewing or reached your viewing limit for this book.aa You have either reached a page that is unavailable for viewing or reached your viewing limit for this book.aa You have either reached a page that is unavailable for viewing or reached your viewing limit for this book.aa You have either reached a page that is unavailable for viewing or reached your viewing limit for this book.aa You have either reached a page that is unavailable for viewing or reached your viewing limit for this book.aa You have either reached a page that is unavailable for viewing or reached your viewing limit for this book.aa You have either reached a page that is unavailable for viewing or reached your viewing limit for this book.aa You have either reached a page that is unavailable for viewing or reached your viewing limit for this book.aa You have either reached a page that is unavailable for viewing or reached your viewing limit for this book.aa You have either reached a page that is unavailable for viewing or reached your viewing limit for this book.aa You have either reached a page that is unavailable for viewing or reached your viewing limit for this book.aa You have either reached a page that is unavailable for viewing or reached your viewing limit for this book.aa You have either reached a page that is unavailable for viewing or reached your viewing limit for this book.aa You have either reached a page that is unavailable for viewing or reached your viewing limit for this book.aa You have either reached a page that is unavailable for viewing or reached your viewing limit for this book.aa You have either reached a page that is unavailable for viewing or reached your viewing limit for this book.aa You have either reached a page that is unavailable for viewing or reached your viewing limit for this book.aa You have either reached a page that is unavailable for viewing or reached your viewing limit for this book.aa You have either reached a page that is unavailable for viewing or reached your viewing limit for this book.aa You have either reached a page that is unavailable for viewing or reached your viewing limit for this book.aa You have either reached a page that is unavailable for viewing or reached your viewing limit for this book.aa You have either reached a page that is unavailable for viewing or reached your viewing limit for this book.aa You have either reached a page that is unavailable for viewing or reached your viewing limit for this book.aa You have either reached a page that is unavailable for viewing or reached your viewing limit for this book.aa You have either reached a page that is unavailable for viewing or reached your viewing limit for this book.aa You have either reached a page that is unavailable for viewing or reached your viewing limit for this book.aa You have either reached a page that is unavailable for viewing or reached your viewing limit for this book.aa You have either reached a page that is unavailable for viewing or reached your viewing limit for this book.aa You have either reached a page that is unavailable for viewing or reached your viewing limit for this book.aa You have either reached a page that is unavailable for viewing or reached your viewing limit for this book.aa You have either reached a page that is unavailable for viewing or reached your viewing limit for this book.aa You have either reached a page that is unavailable for viewing or reached your viewing limit for this book.aa You have either reached a page that is unavailable for viewing or reached your viewing limit for this book.aa You have either reached a page that is unavailable for viewing or reached your viewing limit for this book.aa You have either reached a page that is unavailable for viewing or reached your viewing limit for this book.aa You have either reached a page that is unavailable for viewing or reached your viewing limit for this book.aa You have either reached a page that is unavailable for viewing or reached your viewing limit for this book.aa You have either reached a page that is unavailable for viewing or reached your viewing limit for this book.aa You have either reached a page that is unavailable for viewing or reached your viewing limit for this book.aa You have either reached a page that is unavailable for viewing or reached your viewing limit for this book.aa You have either reached a page that is unavailable for viewing or reached your viewing limit for this book.aa You have either reached a page that is unavailable for viewing or reached your viewing limit for this book.aa You have either reached a page that is unavailable for viewing or reached your viewing limit for this book.aa You have either reached a page that is unavailable for viewing or reached your viewing limit for this book.aa You have either reached a page that is unavailable for viewing or reached your viewing limit for this book.aa You have either reached a page that is unavailable for viewing or reached your viewing limit for this book.aa You have either reached a page that is unavailable for viewing or reached your viewing limit for this book.aa You have either reached a page that is unavailable for viewing or reached your viewing limit for this book.aa You have either reached a page that is unavailable for viewing or reached your viewing limit for this book.aa You have either reached a page that is unavailable for viewing or reached your viewing limit for this book.aa You have either reached a page that is unavailable for viewing or reached your viewing limit for this book.aa You have either reached a page that is unavailable for viewing or reached your viewing limit for this book.aa You have either reached a page that is unavailable for viewing or reached your viewing limit for this book.aa You have either reached a page that is unavailable for viewing or reached your viewing limit for this book.aa You have either reached a page that is unavailable for viewing or reached your viewing limit for this book.aa You have either reached a page that is unavailable for viewing or reached your viewing limit for this book.aa You have either reached a page that is unavailable for viewing or reached your viewing limit for this book.aa You have either reached a page that is unavailable for viewing or reached your viewing limit for this book.aa You have either reached a page that is unavailable for viewing or reached your viewing limit for this book.aa You have either reached a page that is unavailable for viewing or reached your viewing limit for this book.aa You have either reached a page that is unavailable for viewing or reached your viewing limit for this book.aa You have either reached a page that is unavailable for viewing or reached your viewing limit for this book.aa You have either reached a page that is unavailable for viewing or reached your viewing limit for this book.aa You have either reached a page that is unavailable for viewing or reached your viewing limit for this book.aa You have either reached a page that is unavailable for viewing or reached your viewing limit for this book.aa You have either reached a page that is unavailable for viewing or reached your viewing limit for this book.aa You have either reached a page that is unavailable for viewing or reached your viewing limit for this book.aa You have either reached a page that is unavailable for viewing or reached your viewing limit for this book.aa You have either reached a page that is unavailable for viewing or reached your viewing limit for this book.aa You have either reached a page that is unavailable for viewing or reached your viewing limit for this book.aa You have either reached a page that is unavailable for viewing or reached your viewing limit for this book.aa You have either reached a page that is unavailable for viewing or reached your viewing limit for this book.aa You have either reached a page that is unavailable for viewing or reached your viewing limit for this book.aa You have either reached a page that is unavailable for viewing or reached your viewing limit for this book.aa You have either reached a page that is unavailable for viewing or reached your viewing limit for this book.aa You have either reached a page that is unavailable for viewing or reached your viewing limit for this book.aa You have either reached a page that is unavailable for viewing or reached your viewing limit for this book.aa You have either reached a page that is unavailable for viewing or reached your viewing limit for this book.aa You have either reached a page that is unavailable for viewing or reached your viewing limit for this book.aa You have either reached a page that is unavailable for viewing or reached your viewing limit for this book.aa You have either reached a page that is unavailable for viewing or reached your viewing limit for this book.aa You have either reached a page that is unavailable for viewing or reached your viewing limit for this book.aa You have either reached a page that is unavailable for viewing or reached your viewing limit for this book.aa You have either reached a page that is unavailable for viewing or reached your viewing limit for this book.aa You have either reached a page that is unavailable for viewing or reached your viewing limit for this book.aa You have either reached a page that is unavailable for viewing or reached your viewing limit for this book.aa You have either reached a page that is unavailable for viewing or reached your viewing limit for this book.aa You have either reached a page that is unavailable for viewing or reached your viewing limit for this book.aa You have either reached a page that is unavailable for viewing or reached your viewing limit for this book.aa You have either reached a page that is unavailable for viewing or reached your viewing limit for this book.aa You have either reached a page that is unavailable for viewing or reached your viewing limit for this book.aa You have either reached a page that is unavailable for viewing or reached your viewing limit for this book.aa You have either reached a page that is unavailable for viewing or reached your viewing limit for this book.aa You have either reached a page that is unavailable for viewing or reached your viewing limit for this book.aa You have either reached a page that is unavailable for viewing or reached your viewing limit for this book.aa You have either reached a page that is unavailable for viewing or reached your viewing limit for this book.aa You have either reached a page that is unavailable for viewing or reached your viewing limit for this book.aa You have either reached a page that is unavailable for viewing or reached your viewing limit for this book.aa You have either reached a page that is unavailable for viewing or reached your viewing limit for this book.aa You have either reached a page that is unavailable for viewing or reached your viewing limit for this book.aa You have either reached a page that is unavailable for viewing or reached your viewing limit for this book.aa You have either reached a page that is unavailable for viewing or reached your viewing limit for this book.aa You have either reached a page that is unavailable for viewing or reached your viewing limit for this book.aa You have either reached a page that is unavailable for viewing or reached your viewing limit for this book.aa You have either reached a page that is unavailable for viewing or reached your viewing limit for this book.aa You have either reached a page that is unavailable for viewing or reached your viewing limit for this book.aa You have either reached a page that is unavailable for viewing or reached your viewing limit for this book.aa You have either reached a page that is unavailable for viewing or reached your viewing limit for this book.aa You have either reached a page that is unavailable for viewing or reached your viewing limit for this book.aa You have either reached a page that is unavailable for viewing or reached your viewing limit for this book.aa You have either reached a page that is unavailable for viewing or reached your viewing limit for this book.aa You have either reached a page that is unavailable for viewing or reached your viewing limit for this book.aa You have either reached a page that is unavailable for viewing or reached your viewing limit for this book.aa You have either reached a page that is unavailable for viewing or reached your viewing limit for this book.aa You have either reached a page that is unavailable for viewing or reached your viewing limit for this book.aa You have either reached a page that is unavailable for viewing or reached your viewing limit for this book.aa You have either reached a page that is unavailable for viewing or reached your viewing limit for this book.aa You have either reached a page that is unavailable for viewing or reached your viewing limit for this book.aa You have either reached a page that is unavailable for viewing or reached your viewing limit for this book.aa You have either reached a page that is unavailable for viewing or reached your viewing limit for this book.aa You have either reached a page that is unavailable for viewing or reached your viewing limit for this book.aa You have either reached a page that is unavailable for viewing or reached your viewing limit for this book.aa You have either reached a page that is unavailable for viewing or reached your viewing limit for this book.aa You have either reached a page that is unavailable for viewing or reached your viewing limit for this book.aa You have either reached a page that is unavailable for viewing or reached your viewing limit for this book.aa You have either reached a page that is unavailable for viewing or reached your viewing limit for this book.aa You have either reached a page that is unavailable for viewing or reached your viewing limit for this book.aa You have either reached a page that is unavailable for viewing or reached your viewing limit for this book.aa You have either reached a page that is unavailable for viewing or reached your viewing limit for this book.aa You have either reached a page that is unavailable for viewing or reached your viewing limit for this book.Table S Present Values Year 1% 2% «3% «48M OM«7%H——BM——«MH—*10% 1990 9807962 952943933 92H 91T_ 909 2-980 951943925907 890873857842 8G 3 Ol 92 SIS RRO 86H 840 BIG 7947725 4 96124888 RSS 823-92 T6338 T0808. 5 951 906863822784 747713 68165062 6 942888 BT 790746705666 63059654 7 933 BT 81360, TH 665 62358347513 8 938538 TBL 677627582 $4002 467 S914 83766 1034S 5925445006048 10 865820744676 14558 SOR A642 386, 1 896804722650, 585. 527 47S. 429388380 12 (887.789 OLS, 557 497 444397356319. 12 8797 BLL 530 HIS 368 2G 200. 14 81078577505 ARR 3d 200 263 16 8613.5 BL AIT 362 31S S28. 16 833 B53 SB 39433922522 84 74 S513, 3637137272318 1% 860887494. 416350288 Da 19 8857S 306A 2 «8063554456377 312 2582S ATR 2 Bil 66053839359 td 29S 2 «BT RGIS BS OT 4053262? KL 27RD 4923903102? SSG OD 2 O10 TR 37S 2952811092 3000 42524123082 0997S 0ST 35.706 5003552583 181008068036 40 672453307, 208142 T6705 0382022. 4 9 M0264 073 083g $0.86 372k] 087 S434 oO .009OPERATIONS RESEARCH Pf ea eS LTO Sy Third Edition This revised edition elucidates the key concepts and methods of operations Re eee an ne eee cole upgrade student’s knowledge and skills in the subject. enti * Updatedand suffused with numerous and varied solved examples to provide integrated view of theory and applications of operations research. Oe are eet es use et Sufficient hints and answers to unsolved problems to enable students learn at Sey ce Complete conformity to the latest trend of questions. appearing in universities and professional examinations. This book meets the requirement of students pursuing MBA/PGDM, BCom (Hons), MCom, CA, ICWA, BBA, BISIBIT, BCA, AMIE courses. Ithelps them understand the basic concepts of OR and the application of their results in real-life business problems. Professor | K Sharma formerly, Professor, Faculty of Management Studies, University of Dathi, has more than 30 years of teaching experience in Operations ee eee ea uc ee aa authored 17 books, which have been widely appreciateo by students of undergraduate and postgraduate classes of all the Indian University/Management Institutes, and has Ree ite ere CS actively involved in guiding doctoral students, and is also involved in conducting earned eae ee Re eens His research interests in the application of operations research inthe functional areas Sir eee en nr eee nen tee tees ESSEC (A Gradvate School of Management) in France during 1992-93. a ean ‘Macmillan Publishers India Ltd www.macmillanindia.com