# SIMULATION IN OPERATION RESEARCH

Presented by:Apsha Gupta Avneet Kaur Sayan Chatterjee Sneha Chhabra Swati Rawat Ujjawal Gupta (67) (70) (96) (99) (103) (104)

SIMULATION DEFINATION
Simulation is a quantitative technique developed for studying alternative courses of actions by building a model of that system and then conducting a series of experiments to predict the behaviour of the system over a period of time. T. H. Taylor defined Simulation as ³A numerical technique for conducting experiments on a digital computer, which involves certain types of mathematical & logical relationships necessary to describe the behaviour and structure of a complex real world system over extended period of time.´ It is possible to simulate the given system for a given set of inputs and study the impact of changes to be incorporated on the behaviour of the system. The simulation approach can be used to study a problem that involves uncertainty.

 Simulation is useful in solving problems where all values of the variables are either not known or partially known.  In situations where it is difficult to predict or identify bottlenecks, Simulation is used to foresee these unknown difficulties.  The simulation approach is useful to study a problem that involves uncertainty.

Types of Simulation
1. System Simulation: This technique is used in situations where business or operating environment is reproduced to study the behaviour of the system under different operating parameters. The impact of alternative management actions on the system can be analysed. This involves Simulation of an Inventory System and Simulation of Queuing System. Monte Carlo Simulation: This technique is based upon probability distribution and the use of random numbers. Also called computer simulation, it can be described as a numerical technique that involves modeling a stochastic system with the objective of predicting the system¶s behaviour. It is more popular in business applications due to its ease of implementation and low costs.

2.

Thus we take random no. depends upon the number of places to which Probability is known. From the random number tables.s 00-99 to represent them.  .If the prob. triple digit etc. This can be done by following any fixed pattern like row wise. Choice of random numbers whether single digit. which add up to 1. have been calculated to two decimal places. diagonal wise. choose a set of required random numbers from any part of the table.Steps Involved in Simulation (Monte Carlo Technique)    Find the cumulative Probability Assign random numbers Interval corresponding to the Probability. Eg. double digit. column wise.00. we need 100 numbers of 2 digit to represent eaach point of probability.

ships etc. costs. military equipments. profits etc. sales. at design stage ± Training of pilots .  Deciding the crew size for breakdowns/queuing systems  Weather forecast  Forecasting demand. contd.Applications of Simulation  Location of emergency vehicles  Inventory policy decisions  Financial planning (portfolio selection. planes. capital budgeting decisions)  Scheduling production processes  Design of sophisticated products such as reactors. cars etc.  Physical simulation of : ± Testing of Brake & Car engine testing ± Aero-dynamic testing of planes.

11 0.02 0.08 11 0.13 0.Example 17.07 12 0.10 Probability distribution of lead time .06 Probability 0. are as follows: Probability distribution of daily demand Units demanded 3 4 5 6 7 8 9 10 0.08 0.19 0.1 (Simulation of Inventory System) Ques) Consider the case of a certain product for which the probability distribution of daily demand and the probability distribution of the lead time. both developed empirically by observations made over a long span of peroid.16 0.

The dealer is anxious to known.20 3 0.15 The ordering cost is known to be Rs 80 per order. what would be the total inventory costs.30 4 0. the holding cost per unit per day is estimated at Rs 2. with a beginning inventory balance 30 units. we shall evaluate a simulation plan which calls for re-order quantity of 40 units and re-order level of 20 units. Soln) For our example. for specific re-order levels and re-order quantities . representing the loss in profits is Rs 20 unit per day. while the unit shortage cost. . ordering and shortage costs). (made up for holding. and therby selecting an appropriate combination of two.Lead time (days) Probability 2 0.35 5 0.

00 00-01 02-09 10-20 21-36 37-55 56-68 69-78 79-86 87-93 94-99 .13 0.19 0.07 0.16 0.10 0.69 0.56 0.87 0. Random Number Coding: Demand Distribution Daily Demand Probability Cumulative Probability Random Number Interval 3 4 5 6 7 8 9 10 11 12 0.08 0.06 0.11 0.94 1.02 0.08 0.21 0.37 0.10 0.79 0.Table 1.02 0.

20 0.85 1. Random Number Coding: Lead Time Distribution Lead Time (days) 2 3 4 5 Probability Cumulative Probability 0.30 0.Table 2.35 0.20 0.50 0.00 Random Number Interval 00-19 20-49 50-84 85-99 0.15 .

Table 3.Time (4) (5) Receipts (6) Balance (7) O (8) H S (9) (10) 0 1 2 3 4 5 6 7 8 9 68 13 09 20 73 07 92 99 93 8 5 4 5 9 4 11 12 11 47 74 - 3 4 - 40 - 30 22 17 13 8 39 35 24 12 1 80 80 - 44 34 26 16 78 70 48 24 2 - . Simulation Worksheet Day (1) R. No (2) Demand (3) R. No L.

10 11 12 13 14 15 16 17 18 19 20 21 22 23 18 24 22 07 29 57 33 49 65 92 98 00 57 12 5 6 6 4 6 8 6 7 8 11 12 3 8 5 25 21 - 3 3 - 40 40 0 0 34 30 24 16 10 3 35 24 12 9 1 36 80 80 - 0 0 68 60 48 32 20 6 70 48 24 18 2 72 80 120 - .

24 25 26 27 28 29 30 31 32 33 34 35 36 37 31 96 85 92 91 77 37 34 11 27 10 59 33 87 6 12 10 11 11 9 7 6 5 6 5 8 6 11 47 69 09 - 3 4 2 - 40 40 - 30 18 8 0 29 20 13 7 2 36 31 23 17 6 80 80 80 - 60 36 16 0 58 40 26 14 4 72 62 46 34 12 60 - .

.356.38 39 40 72 73 79 9 9 10 80 4 40 - 37 28 18 80 640 74 56 36 1. the three adding upto Rs 2. the holding cost= Rs 1. we find that the total ordering cost = Rs 640.of cost = Rs 260.456 and the out.456 260 Total O: Ordering Cost H: Holding Cost S: Shortage Cost Completing a 40 day period.

40 0. The company has two different assembly lines to produce its most popular product ³Pressurex´.25 0.05 .The processing time for eaach of the assembly lines is regarded as a random variable and is described by the following distributions.10 0.3 N.15 0.Example 17.20 0.20 0.K Vohra ( Monte Carlo Technique) Ques) The Tit-Fit Scientific Laboratories is engaged in producing different types of high class equipment for use in science laboratories.10 Assembly A2 0.15 0. Process Time (minutes) 10 11 12 13 14 Assembly A1 0.40 0.

read the no. 4134 8343 3602 7505 7428 7476 1183 9445 0089 3424 4943 1915 5415 0880 9309 .Using the random numbers. For the purpose.¶s vertically taking the first two digits for the processing time on assembly A1 and the last two digits for processing time on assembly A2. generate data on the process times for 15 units of the item and compute the expected process time for the product.

we will find the cumulative Prob.10 0.20 0.10 0.20 0.95 1.15 0. Cum.40 0.00 RN Interval 00-09 10-24 25-64 65-89 90-99 Prob. 0.90 1. Assembly A2 Cum.05 .25 0.00 RN Interval 00-19 20-59 60-79 80-94 95-99 10 11 12 13 14 0.60 0. 0. Table 1.65 0. Prob. And random number Intervals to the processing times on each of the assemblies. Time (mts) Assembly A1 Prob.20 0.40 0.80 0.Solution) Firstly.10 0.25 0. Prob.15 0.

Simulation Worksheet Unit 1 2 3 4 5 6 7 8 9 10 11 Assembly A1 R.Number 34 76 43 43 83 15 02 45 15 05 89 Time 11 12 11 11 13 10 10 11 10 10 13 Total Time (mts.Table 2.) 12+11=23 25 23 24 24 21 22 25 22 23 23 .Number 41 74 49 83 11 19 36 94 54 75 00 Time 12 13 12 13 11 11 12 14 12 13 10 Assembly A2 R.

27 The expected completion time for a unit works out to be 23. .Contd. 12 13 14 15 08 74 34 93 10 13 12 14 80 28 24 09 13 11 11 10 23 24 23 24 Expected time= 349/15 = 23.27 minutes.

was sold week 4 5 6 7 8 9 10 Total 6 5 9 12 8 7 3 50 Using random no. .What is the average no.EXAMPLE 17. of heaters demanded per week over the 20 week interval.4 Haggins plumbing and heating maintains a stock of 30 million gallon hot water heaters that it sells to home owners and installs for them. the owner likes the idea of having supply on hand so as to meet all customer demand but he also recognizes that it is expensive to do so . of weeks this no.If Haggins maintains a constant supply of 8 hot water heaters in any given week. given below.how many times will he be out of stock during the 20 week simulation period? 2. simulate demand for 20 weeks and answer the following question 1. He examine hot water heater sales over the past 50 weeks and notes the following Hot water heater sales per No.

74.94 1.40 .95.64 .22 .32.12 .66. 10.06 1.14 .03.00 Random no.24 .96.12 .44 Demand per allocation 4 5 6 7 8 9 10 Total No.03.18 .34. 24.51.46.80 . of weeks 6 5 9 12 8 7 3 50 probability .RANDOM NO.97.59.10 .08.48.00 Cumulative probability .16 . interval 00-11 12-21 22-39 40-63 64-79 80-93 94-99 .34.23.77.

of weeks = 135/20 =6.Day 1 2 3 4 5 6 7 8 9 10 Random no.75 . demand 08 48 66 97 03 96 46 74 77 44 4 7 8 10 4 10 7 8 8 7 No. of weeks e is expecting to be out of stock=3 Average demand for heaters per week= total demand/no. demand 10 24 03 32 23 59 95 34 34 51 4 6 4 6 6 7 10 6 6 7 Day 11 12 13 14 15 6 17 18 19 20 Random no.

Sales (Unit) 27 28 29 30 31 32 Probability 0. what is the advantage or disadvantage of the company? . 11. 50. Any unsold product is to be disposed off at loss of Rs. 20 If the company decides to produce 29 units per day.  Using the following random numbers.20 0. respectively.05  The production cost and sales price of each unit are Rs. There is a penalty of Rs.15 0. 15. 01. The sale of these items depends upon demand which has the following distribution. 65. 40 and Rs.10 0. 79.35 0. estimate the total profit/loss for the company for the next ten days.15 0. 16. 99. 10.Ques) A company manufactures 30 units/day. 5 per unit if the demand is not met. 99.

Sales (unit) Probability Cumulative probability Random No.20 30 0.05 .15 32 0.10 0.15 29 0.10 28 0.35 31 0. Interval 27 0.

15 0.00 .Sales (unit) Probability Cumulative probability Random No.15 0.25 29 0.10 28 0.95 32 0.35 0.45 30 0.20 0.05 1.80 31 0.10 0. Interval 27 0.

Sales (unit) Probability Cumulative probability Random No.25 0.10 0.As the first step.15 0.05 0.20 0. Interval 27 28 29 30 31 32 0.15 0.35 0.00 00-09 .45 0.95 1.80 0. random numbers 00-99 are allocated to various possible sales values in production to the probabilities associated with them.10 0.

40= Rs. 10 Loss per unit unsold = Rs. From the given following information. first when production is 30 units per day and then when it is 29 units. Also. 5 per unit  Using these inputs. the profit/loss for the 10 days is calculated. the total profit works out to be the same. 2695 when 30 units are produced. we have Profit per unit sold = Rs.  It is evident that the total profit/loss for the 10 days is Rs. 50 ± Rs. Now we simulate the demand for the next 10 days using the given random numbers. 15 Penalty for using demand = Rs. if the company decides to produce 29 units per day. .

Day Random Estimated Numbers Sales (units) 10 99 65 99 95 01 79 1 16 20 28 Profit/Loss per day with production 30 units 29 units 1 2 3 4 5 6 7 8 9 10 .

Day Random Estimated Numbers Sales (units) 10 99 65 99 95 01 79 1 16 20 28 32 30 32 32 27 30 28 28 28 Profit/Loss per day with production 30 units 29 units 1 2 3 4 5 6 7 8 9 10 .

Day Random Estimated Numbers Sales (units) 10 99 65 99 95 01 79 1 16 20 28 32 30 32 32 27 30 28 28 28 Profit/Loss per day with production 30 units 28*10-2*15 = Rs. 250 29 units 1 2 3 4 5 6 7 8 9 10 .

250 Total Profit = Rs. 290 30*10-2*15 = Rs. 250 30*10-2*5 = Rs. 290 30*10 = Rs.Day Random Estimated Numbers Sales (units) 10 99 65 99 95 01 79 1 16 20 28 32 30 32 32 27 30 28 28 28 Profit/Loss per day with production 30 units 28*10-2*15 = Rs. 300 1 2 3 4 5 6 7 8 9 10 30*10-2*5 = Rs. 250 28*10-2*15 = Rs. 300 28*10-2*15 = Rs. 250 28*10-2*15 = Rs. 225 30*10 = Rs. 290 27*10-3*15 = Rs. 2695 .

225 30*10 = Rs. 300 29 units 28*10-1*15 = Rs. 250 30*10-2*5 = Rs. 2695 . 265 1 2 3 4 5 6 7 8 9 10 30*10-2*5 = Rs. 250 Total Profit = Rs. 290 30*10 = Rs. 300 28*10-2*15 = Rs. 290 27*10-3*15 = Rs. 290 30*10-2*15 = Rs. 250 28*10-2*15 = Rs. 250 28*10-2*15 = Rs.Day Random Estimated Numbers Sales (units) 10 99 65 99 95 01 79 1 16 20 28 32 30 32 32 27 30 28 28 28 Profit/Loss per day with production 30 units 28*10-2*15 = Rs.

240 29*10-1*5 = Rs. 265 29*10-3*5 = Rs. 275 29*10-3*5 = Rs. 250 . 265 27*10-2*15 = Rs. 290 27*10-3*15 = Rs. 300 28*10-2*15 = Rs. 225 30*10 = Rs. 275 29*10-1*5 = Rs.Day Random Estimated Numbers Sales (units) 1 2 3 4 5 6 7 8 10 99 65 99 95 01 79 1 28 32 30 32 32 27 30 28 Profit/Loss per day with production 30 units 28*10-2*15 = Rs. 285 28*10-1*15 = Rs. 285 29*10-3*5 = Rs. 290 30*10 = Rs. 265 30*10-2*5 = Rs. 300 29 units 28*10-1*15 = Rs. 250 30*10-2*5 = Rs. 290 30*10-2*15 = Rs.

2695 28*10-1*15 = Rs. 250 Total Profit = Rs. 2695 . 265 = Rs. 265 10 20 28 28*10-2*15 = Rs. 250 28*10-1*15 = Rs.9 16 28 28*10-2*15 = Rs.

we proceed as follows. perturbed about the waiting time of mechanics . The manager. he wishes to ascertain how many attendants may be employed to minimise the total cost involved. The foll.2 N.K Vohra (simulation of queuing system) Ques) In a large workshop undertaking servicing jobs. is in the process of determining whether more attendants be hired for the store for raising the level of service. Foe helping the manager solve his problem using simulation. . Data on the times between successive arrivals and the service times for the mechanics have been obtained from the past 200 observations made on the system presently in operation. the mechanics obtain their tools and requirements from a central store. The idle time cost for the mechanics and the wages required to be paid to the attendants being known.Example 17.

16 0.07 .25 0.06 0.37 0.Distribution of inter-arrival time: Time (Minutes) 0 3 6 9 12 15 Total Frequency 12 18 50 74 32 14 200 Probability 0.09 0.

10 0.24 .Distribution of service time: Time (Minutes) 4 6 8 10 12 Total Frequency 6 20 36 88 48 200 Probability 0.44 0.04 0.18 0.

40 0.09 0.00 Random Number Interval 00-05 06-14 15-39 40-76 77-92 93-99 0.16 0.77 0.) Step 1: Assign random numbers to each observed arrival time interval based on the likelihood of the occurrence of each time interval Time (Minutes) 0 3 6 9 12 15 Probability Cumulative Probability 0.37 0.06 0.15 0.Sol.25 0.07 .93 1.06 0.

04 0.24 0.00 00-03 04-13 14-31 32-75 76-99 4 6 8 10 12 .10 0.04 0.76 1.Step 2: Follow the same procedure for the service time distribution Time (Minutes) Probability Cumulative Random Probability Number Interval 0.14 0.18 0.44 0.32 0.

01 9.33 9.21 9.41 9.21 8.23 9.21 am 8.31 8.51 9.53 .33 8. (5) 87 39 28 97 69 87 52 52 15 85 Time (6) 12 10 8 12 10 12 10 10 8 12 Waiting Time (min) (7) Service Queue Length (10) 1 1 2 2 2 2 2 2 2 2 Begins (8) 00 03 07 06 12 11 16 14 12 11 8.48 8.23 9.13 9. No.09 9.Step 3: Now Simulate the operation S.57 9.13 9.39 8.39 8.01 9. Arrivals No.30 Service R.24 8.09 am 8.31 8.33 9.41 Ends (9) 8.18 8.09 am 8. R. Time O¶ (1) (2) (3) clock (4) 1 2 3 4 5 6 7 8 9 10 58 47 23 69 35 55 69 90 86 74 9 9 6 9 6 9 9 12 12 9 8.51 9.39 8. No.

09 11.48 Service R.19 11.03 10.36 10. No.27 10.27 10.47 10. Time (5) (6) 41 82 98 99 23 77 42 60 22 91 10 12 12 12 8 12 10 10 8 12 Waiting Time (min) (7) Service Queue Length (10) 2 3 4 4 4 5 6 6 6 5 Begins (8) 17 21 21 18 24 29 32 33 37 39 9.54 10.27 10.53 10.42 9.42 10.39 10. Time O¶ (1) (2) (3) clock (4) 11 12 13 14 15 16 17 18 19 20 39 15 90 98 39 16 52 56 21 23 6 6 12 15 6 3 9 9 6 6 9.27 11.15 10.S. Arrivals No.09 10.39 .09 11.47 10.39 10.36 9.15 10.03 10.59 11.18 10.19 11.27 Ends (9) 10.15 10. No.59 11. R.

24 34 10 10 8 12 10 51 49 53 52 55 6.29 4 3 2 1 0 73 . Time (5) (6) Waiting Time (min) (7) Service Queue Length (10) Begins (8) Ends (9) 21 22 23 24 25 Total 00 87 20 40 73 0 12 6 9 9 10. Arrivals No. No.39 11.07 pm 12.06 22 11.49 11.59 12.15 92 11.49 11.07 pm 12.19 12.19 11.00 36 11.23 mins 11. R.S. Time O¶ (1) (2) (3) clock (4) Service R.48 68 11. No.59 12.