You are on page 1of 8

MB0048 SET 1

1. OR has be far in and it i too yo defined authorit Howeve Septemb of the Researc the Ju Society carried followin Operational Research is the application of the methods of science to complex problems arising in the direction and management of large systems of men, machines, materials and money in industry, business, government and defence. The distinctive approach is to develop a scientific model of the system, incorporating measurements of factors such as chance and risk, with which to predict and compare the outcomes of alternative decision strategies and controls. The purpose is to help management, determine its policy and actions scientifically. Operations research (OR) is an analytical method of problem-solving and decision-making that is useful in the management of organizations. In operations research, problems are broken down into basic components and then solved in defined steps by mathematical analysis. Analytical methods used in OR include mathematical logic, simulation, network analysis, queuing theory , and game theory . The process can be broadly broken down into three steps. 1. A set of potential solutions to a problem is developed. (This set may be large.) 2. The alternatives derived in the first step are analyzed and reduced to a small set of solutions most likely to prove workable. 3. The alternatives derived in the second step are subjected to simulated implementation and, if possible, tested out in real-world situations. In this final step, psychology and management science often play important roles.

Characteristics
1. OR approaches problem-solving and decision-making from the total systems perspective. OR does not experiment with the system itself but constructs a model of the system upon which to conduct experiments. 2. OR is not necessarily using interdisciplinary teams, but it is interdisciplinary; it draws on techniques from sciences such as biology, physics, chemistry, mathematics and economics and applies the appropriate techniques from each field to the system being studied.

3. Model building and mathematical manipulation provide the methodology which has been the key contribution of OR. 4. OR is for operations economy. 5. The primary focus is on decision-making and computers are used extensively. b. Operations research is an experimental and applied science devoted to observing, understanding and predicting the behaviour of purposeful man-machine systems; and operations-research workers are actively engaged in applying this knowledge to practical problems in business, government and society. Thus, operations-research workers are engaged in three classical aspects of science: Describing the behaviour of these systems. Analyzing this behaviour by constructing theories (frequently called models) that account for the observed phenomena. Using these theories to predict future behaviour, that is, the effects that will be produced by changes in the systems or in their methods of operation. Since the operating systems studied by operations-research workers arise in a wide variety of practical industrial, military and governmental environments, it follows that the results of their research frequently make important contributions to the solution of problems of choice, policy and planning that arise in these environments; these contributions are characteristically made by presenting the research findings directly to the executives responsible for the operations or systems studied. Limitations: Models are only an attempt in understanding operation and should never be considered as absolute in any sense. Validity of any model with regard to corresponding operation can only be verified by carrying the experiment and relevant data characteristics. Constructions of models require services of subject experts. 2. Linear Programming Problem: Linear programming is a branch of mathematics and statistics that allows researchers to determine solutions to problems of optimization. Linear programming problems are distinctive in that they are clearly defined in terms of an objective function, constraints and linearity. The characteristics of linear programming make it an extremely useful field that has found use in applied fields ranging from logistics to industrial planning. Optimization All linear programming problems are problems of optimization. This means that the true purpose behind solving a linear programming problem is to either maximize or minimize some value. Thus, linear programming problems are often found in economics, business, advertising and many other fields that

value efficiency and resource conservation. Examples of items that can be optimized are profit, resource acquisition, free time and utility. Linearity As the name hints, linear programming problems all have the trait of being linear. However, this trait of linearity can be misleading, as linearity only refers to variables being to the first power (and therefore excluding power functions, square roots and other non-linear functions). Linearity does not, however, mean that the functions of a linear programming problem are only of one variable. In short, linearity in linear programming problems allows the variables to relate to each other as coordinates on a line, excluding other shapes and curves. Objective Function All linear programming problems have a function called the "objective function." The objective function is written in terms of the variables that can be changed at will (e.g., time spent on a job, units produced and so on). The objective function is the one that the solver of a linear programming problem wishes to maximize or minimize. The result of a linear programming problem will be given in terms of the objective function. The objective function is written with the capital letter "Z" in most linear programming problems. Constraints All linear programming problems have constraints on the variables inside the objective function. These constraints take the form of inequalities (e.g., "b < 3" where b may represent the units of books written by an author per month). These inequalities define how the objective function can be maximized or minimized, as together they determine the "domain" in which an organization can make decisions about resources. b. Graphical Method of Solving LPP The graphic solution procedure is one method of solving two variable linear programming problems and involves the following steps: 1. Formulate the problem in terms of a series of mathematical constraints and an objective function. 2. Plot each of the constraints as follows: Each inequality in the constraint equation be written as equality. Give any arbitrary value to one variable and get the value of other variable by solving the equation. Similarly, give another arbitrary value to the variable and find the corresponding value of the other variable. Now plot these two set of values. Connect these points by a straight line. This exercise is to be carried out for each of the constraint equations. Thus, there will be as many straight lines as there are equations; each straight line representing one constraint. 3. Identify the feasible region (or solution space), i.e., the area which satisfies all the constraints simultaneously. For greater than constraints, the feasible region will be the area which lies above the constraint lines. For lies than constraints, this area is generally the region below these lines. For greater than or equal to or less than equal to constraints, the feasible region includes the points on the constraint lines also.

4. Identify each of the corner (or extreme points) of the feasible region either by visual inspection or the method of simultaneous equations. 5. Compute the profit/cost at each corner point by substituting the co-ordinates of that point into the objective function. 6. Identify the optimal solution at that corner point which shows highest profit (in a maximization problem) or lowest cost (in a minimization problem)

3. Computational procedure of simplex method Let the linear programme be Max z = c1x1 +c2x2 +..+cnxn Subject to the constraints: a11x1+a12x2+.+ a1nxn = b1 a21x1+a22x2 ++ a2nxn = b2 .. .. .. . .. ..

am1x1+am2x2+.+ amnxn = bm
x1,x2, xn

Step 1. First, observe whether all the right side constants of the constraints are non-negative. If not, it can be changed into positive calue on multiplying both sides of the constraints by 1 . Step 2 Next convert the inequality constraints to equations by introducing the non-negative slack or surplus variables. The coefficients of slack or surplus variables are always taken zero in the objective function. Step 3 Now present the constraint equations in matrix form Step 4

Construct the starting simplex table using the notations. It should be remembered that the values of non-basic variables are always zero at each iteration. Column XB gives the values of basic variables as indicated in the first column. Step 5 Now, proceed to test the basic feasible solution for optimality by the rules given below. This is done by computing the net evaluation for each variable xj by the formula Zj cj = CBXj - cj Optimality test: i. ii. iii. Step 6 In order to improve this basic feasible solution, the vector entering the basis matrix and the vector to be removed from the basis matrix are determined by the incoming vectors and outgoing vectors respectively. Step 7 The above steps should be repeated until an optimum solution is obtained. 3.b Use of artificial variables LPP, in which constraints may also have and = signs after ensuring that all bi are o, are considered in this section. In such problems, basis matrix is not obtained as an identity matrix in the starting simplex table, therefore we introduce a new type of variable, called artificial variable. These variables are fictitious and cannot have any physical meaning. The artificial variable technique is merely a device to get the starting basic feasible solution, so that simplex procedure may be adopted as usual until the optimal solution is obtained. Artificial variables can be eliminated from the simplex table as and when they become zero (non-basic). The process of eliminating artificial variables is performed in Phase I of the solution, and Phase II is used to get an optimal solution. Since the solution of the LPP is completed in two phases it is called Two phase Method. 4.a Economic interpretation of dual Consider the following primal problem: Maximize c1x1 + : : : + cnxn subject to all xi 0 if all zj cj 0, the solution under test will be optimal, if at least on zj cj is negative, the solution under test is not optimal, then proceed to improve the in the next step. If corresponding to any negative zj cj, all elements of the column Xj are negative or zero, then the solution under test will be unbounded.

a11x1 + : : : a1nxn b1 ::: am1x1 + : : : amnxn bm: Economic interpretation: n economic activities, m resources cj is revenue per unit of activity j bi is maximum availability of resource i aij is consumption of resource i per unit of activity j The dual Minimize b1y1 + : : : + bmym subject to all yi 0 a11y1 + : : : a1mym c1 ::: an1x1 + : : : anmym cn:

Interpreting the dual variables


If (x1; : : : ; xn) is optimal for the primal, and (y1; : : : ; ym) is optimal for the dual, then we know: c1x1 + : : : + cnxn = b1y1 + : : : + bmym Left-hand side: Maximal revenue Right-hand side: resources i (availibility of resource i) X (revenue per unit of resource i) In other words: Value of yi at optimal is dual price of resource i Away from optimality, we have c1x1 + : : : + cnxn < b1y1 + : : : + bmym Left-hand side: current (suboptimal) revenue Right-hand side: resources i (worth of resource i) Solution is not optimal because resources are not being fully utilized

Interpreting the dual constraints


If (x1; : : : ; xn) is feasible (not necessarily optimal) for the primal, and (y1; : : : ; ym) is the corresponding collection of dual values, then we know: Current objective coefficient of xj = (Left-hand side of dual constraint j) (Right-hand side) = (a1jy1 + : : : + amjyj) cj cj is a measure of revenue per unit (of activity j) So a1jy1 + : : : + amjyj is an imputed (implicit) cost per unit (of act. j) Also, yi is imputed cost per unit of resource i in a unit of activity j If objective coefficient of xj (= cost revenue, = reduced cost) is strictly negative, then revenue > cost, so it makes since to increase activity j this is the pivoting process of the simplex method.

4.b

Primal problem It is the original linear programming problem. Consider the following primal problem: Maximize c1x1 + : : : + cnxn subject to all xi 0 a11x1 + : : : a1nxn b1 ::: am1x1 + : : : amnxn bm:
DEFINITION OF THE DUAL PROBLEM

The duality principles can be stated formally in general terms. Let the primal problem be: Primal Maximize z =c j x j , subject to: ai j x j bi (i = 1, 2, . . . ,m), (3) x j 0 ( j = 1, 2, . . . , n). Associated with this primal problem there is a corresponding dual problem given by: Dual Minimize v =bi yi , subject to: ai j yi c j ( j = 1, 2, . . . , n), (4) yi 0 (i = 1, 2, . . . ,m). i.e. Dual problem corresponding to a LPP is another LPP formulated from the parameters of the original problem. - A new LPP, derived from the primal according to a set of transformation rules. i. transposing the coefficients matrix. ii. interchanging the role of constant terms and the coefficients of the objective function iii. reverting the inequalities iv. minimizing the objective function instead of maximizing it. 5. Initial Basic Feasible Solution by North West Corner Rule The simplest of the procedures used to generate an initial feasible solution is NWC rule. It is so called because we begin with the north west or upper left corner cell of our transportation table. Various steps of this method can be summarized as under: Steps 1. select the north west (upper left-hand) cornercell of the transportation table and allocate as many as possible equal to the minimum between available supply and demand requirement, ie min (s1,d1). 2. adjust the supply and demand numbers in the respective rows and columns allocation. 3. a. if the supply for the first row is exhausted, then move down to the first celll in the second row and first column and go to step 2 b. if the demand for the first column is satisfied, then move horizontally to the next cell in the second column and first row and go to step 2. 4. if for any cell, supply equals demand, then the next allocation can be made in cell either in the next row or column.

5. continue the procedure until the total available quantity is fully allocated to the cells as required.