/ ventory Models, Simulation and Decision Theory © Suppose the probabilities of various sizes of the demana 0.3, 0.5, 0.2. EOL (Small) = (0.3 x 0)+(0.5%2)+(0.2x4)", : is EOL (Medium) = (0.3 x 8)+(0.5 x 0) +(,2x2) = 28 BOL (Large) = (0.3 x 16) + (0.5 x 8)+(0.2x0) ee. 0) Since the small order is connected with the smallest EOL, it is. the. best alternative. Simulation © aur imitation § weahiby . Simulation is a numerical technique, for conducting experiments on a digital computer, which involves certain types of mathematical and logical relationships necessary to describe the.behaviorand:structure, of a complex: real-world system over extended periods of time —T. H Naylor. A simulated model may be defined as one which ails the working: ofa ee scale system of men, “machines, materials and:information operating over a period of time in a simulated. environment of" the. actual: real world: conditions. — Donald G. Malcoln.. * ella Ts veppveten tation a Vealon - ‘toc ure af o msde) or error device which etl 4.9.1, TYPES OF SIMULATI BN is pie Sao rognnan as, noolity. wonder @ Deferiinistie tnodels aa Gi) Slochastic models. iii) Static models: ‘ (i) Deterministic Models, ree ‘Input and output variables. are not permitted to be random variable and models are described by exact funeti ial Fe ic (i) stochastic Models ’ : * Atleast one, of the variables of functional Telationship is given byt the ebay functions.Or ‘Applied Operation, These models do not take variable time into consideration. . (iii) Static Models (iv) Dynamic Models \ These models deal with time varying interaction. \ | “Monte-Carlo simulation technique in whch statistical distribution mations are created by using a series of random numbers.. ee Procedure (i) Clearly define the problem . . (a) Identify the objective of the problem. . (b) Identify the main factors. which have the greatest effect on ‘the objectives of the problem. (i) Construct an approproate model. (a) Specify the variable and parameters. of the models, (b) Formulate the appropriate decision rules, i. e., State the conditions under which the experiment is to the performed. © Identify the type of distribution that will be used. (d) Specify the manner in which time will change. {e) Define the relationship between the variables and Parameters, | (iii) Prepare the model of experimentation . | (a) Define the starting condition for the simulation 3 (4) Specify the number of runs of simulation to be made. (iv) Using (i) to (iii) Experimenting’ with the model : (4) Define a coding system that will correlate the factors defined in step -0(0 | with the random numbers to be generated for the simylation,’* | | (6) Select the random number generator and oreate the random nuanbers to +, be used in the simulation.ventory Models, Simulation and Decision Theory 4.27 te the random numbers with the in (iv) a (v) Summarize and examine the results obtained j in step (). (vi) Evaluate the results of simulation. (c) Associate the generator and creat factories identified in (i) and coded i 4.10.1. SIMULATION’PROCESS () Identify and clearly define the problem, (ii) _ List the statement of, objectiveofthe problem, ¢ (if) Construct an appropriate mathéinatical model of the given Pigblem, (jv) Ensure that the model represents the real situation, (v) Make experiment with the model constructed,’ : (vi) Analyze the simulation activity. foreg ef F (vii) Make-changes ‘in the model: or. the sample and repeat the process until a sufficient number of ‘samples are available, (viii) Tabulate the various values of the. decision criterion sande policy, Ina plete eivacred to iheadedtira 100. cai iets occur due:.to variety: of reasons: The ‘probability associated number of production units Per: Any have. been determin -d using past data. Production day 95 Probability. 0.03%:0.05. 07 0.1045. 0, “102, 103 104-105-196 waidicge OL. 0070.05 00306 2 ees MBA. AL LU, May/June 2011 The ship carrying the cars can decémmodlate up to 101 cars per day. Using the random numbers 97, 2, 80, 66, 96, 55, 50, 29,58, 51, 4, 86, 24, 39, 47. Simulate the production run for a period of 15 days, Also determine the average 'nimber of cars waiting to be sipped and the fe meee number of empty Spaces tn the 'ship..-a a 4.28 — i Applied Operation Sy g Solution: , , . (ss i Assign the random number intervals 00 - 99 manufactures of cars per day, XC | Produetion/Day Probability Phianisee "lide r= K 9% | 003 0.03 00-02 96 0.05 0.18 03-07 | \ 97 0.07 0.15 08-14 98 0.10 0.25 15-24 ‘99 0.15 0.40 25-39 100 0.20 0.60 40-59 101 0.15 0.75 ° 60 - 74 102 0.10 0.85 75 - 84 103 0.07 0.92 85-91 104 0.05: 0.97" 92-96 105 0.03 1.00 97-99 The simulated production of cars for the next 15 days is given in the following table: ©: : . Day Random Production Nawiber of Nember of Empty : Number _|* per Day | Cars Waiting Spaces in the ship . 1 7 105 4 _ 2°] we | 95 : 6 | 3 0 | 102 l : A 66 101 : a ‘3 96 |. 104 magi 6 | 55 190 | mile 100 : 1 7 | 50Models, Simulation and Decision \neory “7 Random | Production| Number of Number of Empty | Day Number per Day Cars Waiting | Spaces in the ship > 8 29 99 : 2 / 9 58 100 : 1 / 10 51 100 : 1 ie 04 96 : 5 12 86 103 2 - B 24 _ 98 : 3 14 39 99 : 2 15, 47 100 : 1 Total 10 23 Average number of cars | _ 10 waiting to be shipped 15 _. = 0.67 per day Average number of } _ 23 empty spaces on the ship 15 = 1.53 per day Example 4.13 | The automobile manufactures around 150 scooters. The daily production varies from 146 to 156 depending upon the availability of raw materials and other working conditions. Production per day | 146 | 147 | 148 | 149 | 150 | 151 | 152 Probability 0.04 | 0.09 | 0.12| 0.14 | 0.11} 0.10 | 0.20 153 | 154 0.12 | 0.18 The finished scooters are transported in a specially arranged lorry [Sreommodting 150 scooters. Using following random numbers.4.30 80, 81, 76, 75, 64, 43, o 18, 26, 10, 12, 65; 68, 69, 61, 57. Simulate the proces, as find out (i) What will be the average number of scooters waiting in the factory? (ii) What will be the average number of empty space on the lorry? Solution: : Assign random number intervals 00 - 99 productions of scooters per day, ive” andom Number Production/Day Probability john R Assigned 146 0.04 0.04 00 - 03 “147 0.09 0.13 04-12 148 0.12 0.25 ° 13-24 149 0.14 0.39 25 - 38 150, 0.11 0.50 39-49 151 0.10 0.60 50-59 152. 0.20 0.80 60-79 153 0.12 0.92 80-91 154 0.08 1,00 92-99 Based on the given numbers, we estate the production Per day as shown in the table below, SIN. Random | Production Number of Number of Empty Number | Per Day | Scooters Waiting Spaces in the Lorry 1 80 153 3 0 {2 81 153 3 L 0 3 76 152 2 0 de 4. 75 152 ° 2 0 Jtes [6 1532 | 2 0“A OL eA EE ye Models, Simutation ana vecisivn sucery 2 A £ e, op e Random | Production Number of Number of Empty 1LNo. Number | PerDay | Scooters Waiting | Spaces in the Lorry 6 43 150 0 0 / 7 18 148 0 2 8 26 149 0 1 9 10 147 0 3 10 } 12 147 0 3 11 65 152 2 0 12 68 152 2: 0 13 69 152 2 0 14 61 152 2 0 15 57 151 1 0 Total 21 9 : Average number of 21 @ Scooter al = 15 = 1.4 per day seg : Average number of 9. (i) empty spaces } is 15 : =-0.6 per day [ Example 4.14 | A tourist car “eperaei finds that iting the past few months, the cars use has varied so~much that the cost maintaining the car varied considerably. During the past 200 weeks the demi, the car tabulated as ‘given below. : . Tripperweek 0° J. 2/3. 4 5 Frequency. 16. 24 30° 60 “40. 30 Using the following random numbers simulate ‘the demand. for a ten week period. 82, 95, 18, 96, 20, 84, 56H, 52; 03,ay ©Solution: Trip per r Cumulative Assigned Week Feauency | Probability Probability | Random Number; 0 16 0.08 0.08 ae 1 24 0.12 0.20 08-19 [2 30 0.15 0.35 20 - 34 3 60 0.30 0.65 35 - 64 4. 40 0.20 0.85 BD) 84 5 30 0.15 1.00: 85 - 90 Week Random Demand ea Number 1” 82 4 2 95 245 2! 3 18 1 _ “4 96 “5 *. 5 20 22 6 “84 4 ’ 4 Te 56 ; > 3 ’ € : = rae ‘i ae 9 52 3 10 03 ov ‘ ee Average Demand ‘= |r oJory Models, Simulation and Decision Theory 433) sperience indication the daily demand as given here. A .U Nov/Dec ~ 2014 - MBA DailyDemand 0 10 20 30 40 50 Probability 0.01 0.20 0.15 050 0.12 0.02 | Considering the following sequence of random number. | 25, 39, 65, 76, 12, 05, 73, 89, 19, 49. Estimate the daily average for the cakes on the basis of simulated data. Solution: . The random number coding for varies demand levels are given in table below Cumulative Random Demand | Probability |». spabitity | Number interval 0 0.01 0.01 00 |_ 10 0.20 1 0.21 i 01-20 20 0,15 0.36 be 21-35 30 0.50 _| 0.86. -}.. 36-85 0 0,12. 0.98 86-95 50 0.02 fo 1.005]; 98-99 Based on the 10 random numbers. We simulate ‘the demand for next 10 days. Days | Random Numbers | Demand 1 ‘ 25 [= 20 2 39 “| 30 3 65 30 4 16 30 5 ee 10 6 05 10Applied Operar. Total demand Expected demand = “number of days 240 _ = 9 = 24 units per day Days Random Numbers Demand | 7 2B 30 | | 8 89 40 _ | 2 19 10 | wv 49 30 | 4.10.2. GENERATION OF RANDOM NUMBER | Random Variable A real valued function defined is the sample space is called Random variable. } } | Random Number | A numerical value assigned to a random variable following uniform probability | density function (Normal, Poisson, exponential, Binomial, geometric uniform etc), Generation of Random Number | Simulation needs the generator of sequence of random number which constitutes _an integral part of the simulation model and also helps in determining random i observations from the probability distribution. i | Mid-Square Method In this method, the given number is squared and the mid figures of the square are selected. It becomes the random number of required digits. ie., (189)? = 35721, 572 is first random number (212)? = 44944, 494 is second random number. 4.10.3. ADVANTAGES OF SIMULATION @_ Simulation models are comparatively flexible and can be modified to adjus! the variation in the environment of real situation. ‘ oea a ere a i at Seer eee ot <, jory Models, Simulation and Decision Theory 4.35 ¢ \0.3. ADVANTAGES OF SIMULATION /w Simulation models are comparatively flexible and can be modified to adjust the variation in the environment of real situation. (ii) We can investigate the consequences for a system of possible changes in parameters in terms of the model. (iii) Simulation methods are easier to apply than pure analytical. methods. (iv) Simulation is a teaching aid eg., in a business games, case studies, research ete. . (v) Simulation can be used as a pre-service test to try out new policies and decision rules for operating a system before running the risk of experimentation in the real system. 4.10.4, DISADVANTAGES OF SIMULATION (i) Simulation is not precise. (ii) A good simulation model may be too expensive. (iii) Simulation is a time-consuming exercise. (iv) The simulation model needs to be examined and analyzed for decision making. It only creates the alternatives and not the answer or optimal solution by itself.Serre eet ere eee eee ee ere L (y 13.1. INTRODUCTION The technique of simulation has long boen used by the designers and analysts in physical sciences and it promises to become an important tool for tackling the complicated problems of managerial decision-making. Scale models of machines have been used to simulate the plant layouts and models of aircrafts have been tested in wind tunnels to determine their aerodynamic characteristics. Simulation, which can appropriately be called management laboratory, determines the effect of a number of alternate policies without disturbing the real system. It helps in selecting the best policy with the prior assurances that its implementation will be beneficial Probably the first important application of simulation was made by John Von Neumann and Stanislaw Ulam for studying the tedius behaviour of neutrons in a nuclear shielding problem Which was too complex for mathematical analysis. With the remarkable success of the technique ‘on neutron problem, it became popular and found many applications in business and industry Development of digital computers in early 1950s is further responsible for the rapid progress made by the simulation techniques. The range of simulation application varies from simple queuing ‘models to models of large integrated systems of production, 13.2 WHEN TO USE SIMULATION? In the foregoing chapters we have discussed a number of operations research tools and techniques for solving various types of managerial decision-making problems. Techniques like linear programming, dynamic programming, queuing theory, network models, etc, are not sufficient to tackle all the important managerial problems requiring data analysis. Each technique has its ‘own limitations. Linear programming models assume that the data do not alter over the planning horizon. It is one time decision process and assumes average values for the decision variables. If the planning hhorizon is long, say 10 years, the multiperiod linear programming model may deal with the yearly averaged data, but will not take into account the variations over the months and weeks with the result that month to month and week to week variations are left uncovered. Other important limitation of linear programming is that it assumes the data to be known with certainty, In many ‘eal situations the uncertainties about the data are such that they cannot be ignored. In case the Uncertainty relates to only a few variables, the sensitivity analysis can be applied to determine its effect on the decision. But, in situations, where uncertainty pervades the entire model, the sensitivity analysis may become too cumbersome and computationly difficult to determine the ‘impact of uncertainty on the recommended plan Dynamic programming models, however, ean be used to determine optimal strategies. by ‘king into account the uncertainties and can analyse multiperiod planning problems. In other Words, this technique is free from the two main limiations of linear programming. But it has its own shortcomings. Dynamic programming models can be used to tackle very simple situations involving only a few variables, If the number of state variables is a bit larger, the computation task becomes quite complex and involved. Download From - www.cyaspirants.com 1192 fb.com/egaspiranta Download From - www.cgaspirants.com SIMULATION # 1193 ¢ stochastic ions hold good for other rasthematical techniques like dynamic stoc mets ich as mein and sung line sess Gu male ie sc aerte these moves; moreover, by making a amber of assumptions the systems are simplified to scl ‘n extent that in many cases the results obtained are only rough approximations. aay, ove discusion we conc that when the characterises such as uncertainty, develop i fYnamic interaction between the decision and subsequent event, and the need oy inate, tiled procedures and finely divided tne intsrvals combine gets fae ne ion, it becomes 100 complex to be solved by any ofthe techniques of mathematical programming and Pie anaie Models. I must be analysed by some lhe kind of quontcniee technique which may valle agi and reliable results. Many ce techniques are coming Up, but, 50 far, the best availabe is simulation Jn general the simulation technique ‘analysis is either to0 complex o 14 éependable tool in situations where mathematical ” 100 cost 13.3 WHAT IS SIMULATION? Simulation is an imitation of reality, A children cycling park, with various crossings and unas isa simulated model ofthe city trathe space te the laboratories « number of experiments cavinguformed on simulated models to determine tre behaviour of the real system in true We dover A simple illustration i the testing of m sucrat ‘model in @ wind tunnel from which secZemine the performance of the actual ach under real operating conditions. Planetarium Rares fePresent a beautiful simulation of the Vanel system. Environments in a geological gerdon and in a musem of natural history are other In all these examples, it hasbeen tried to imiote the reality to see what might happen under foo berating conditions. This imitation of realty whieh may be in the physical form or in the fonn of mathematical equations may be called sinters defined as one which depicts the working of a large sale syste information operating over a period of time in vinelatch pruitions. According to Shannon simulanon the process of by conducting experiments with this model forthe purpose of understanding the behaviour of the ©peration of the system, : 'm of men, machines, materials and invironment of the actual real world designing a model of the real system 13.4 ADVANTAGES OF THE SIMULATION TECHNIQUE {The simulation technique, when compared with the mathematical Programming and standard Probability analysis, offers a number of advantages over these fechniques; a few important among them can be summarized as follows: 1} Many important managerial decision problems ar 00 intricate to be solved by mathematical aerating and experimentation with the actual system, even if possible. sto ten costly ‘and risky. Simulation offers the solution by allowing experimentation wilt woes of the fb.com/cgaspirant1194 ¢ OPERATIONS RESEARCH system without interfering with the real system, Simulation is, thus, often a bypass for complex mathematical analysis. 2. Through simulation, management can foresee the difficulties and bottlenecks which may come up due {0 the introduction of new machines, equipment or process. It, thus, eliminates the need of costly trial and error methods of trying out the new concept on real methods and equipment 3. Simulation has the advantage of being relatively free from mathematics and, thus, can be easily understood by the operating personnel and non-technical managers. This helps in getting the proposed plans accepted and implemented 4, Simulation models are comparatively flexible and can be modified to accommodate the changing environments of the real situation 5. Computer simulation can compress the performance of a system over several years and involving large calculations into a few minutes of computer running time 6. The simulation technique is easier te use than mathematical models and is considered quite superior to the mathematical analysis 7. Simulation has advantageously been used for training the operating and managerial staiT in the operation of complex plans. It is always advantageous to train people on simulated ‘models before putting into their hands the real system. Simulated exercises have been developed to impart the trainee sufficient exercise and experience, A simulated exercise familiatizes the trainee with the data required and helps in judging what information is really important. Due to his personel involvement into the exercise, the trainee gains sufficient confidence, and moreover becomes familiar with data processing on electronic computer 8. Once a simulation model has been constructed, t may be used time and again to analyse different situations, 9. Itis a valuable and convenient method of breaking down a complicated system into sub- systems and then studying each of these subsystems individually or jointly with others 13.5 LIMITATIONS OF THE SIMULATION TECHNIQUE In spite of all the advantages claimed by the simulation technique, many operations research analysts consider it a method of last resort and use it only when all other techniques fail. If 4 particular type of problem can be well represented by a mathematical model, the analytical approach is considered to be more economical, accurate and reliable. Further, in very large and ‘complex problems simulation may suffer from the same deficiencies as other mathematical models. In brief, the simulation technique suffers from the following limitations: 1. Simulation does not produce optimum results. When the model deals with uncertainties, the results of simulation are only reliable approximations subject to statistical errors. 2. Quantification of the variables is enother difficulty. In a number of situations it is not possible to quantify all the variables that affect the behaviour of the system 3. In very lerge and complex problems, the large number of variables and the inter- relationships between them make the problem very unwieldy and hard to program. The number of variables may be too large and may exceed the capacity of the available ‘computer. 4. Simulation is, by no means, a cheap method of analysis. In a number of situations, such 48s corporate planning, simulation is comparatively costlier and time consuming ‘5. Other important limitations stem from too much tendency to rely on the simulation models, ‘This results in application of the tecliique to some simple problems which can more appropriately be handled by other techniques of mathematical programming, fb.com/egaspirantDownload From - sww.cgaspiramts-com SIMULATION & 1195 13.6 APPLICATIONS OF SIMULATION _ Simulations quite versatile and commonly applied technique for solving decision PODIETS as been applied successfully to a wide range of problems of science and technology as given below: 1. In the field of basic sciences, it has been used to evaluate the area under @ curve, estimate the value of x, in matrix inversion and study of particle diffusion 2. In industrial problems ineluding the shop floor management, design of computer seme, design of queuing systems, inventory control, communication networks, chemical processes, nuclear reactors and scheduling of production processes. 4m business and economic problems, mcluding customer behaviour, price determination, economic forecasting, portfolio selection and capital budgeting. In social problems, including population growth, effect of environment on health and group behaviour 5. In biomedical systems, and brain activities. In the design of weapon systems, war strategies and tactics In the study of projects involving risky investments. 13.7 MONTE CARLO SIMULATION including fluid balance, distribution of electrolyte in human body 6 7 The Monte Carlo method of simulation was developed by the two mathematicians John ‘Von Neumann and Stainslaw Ulam, during World War II to study how far neutrons would travel through different materials. The technique provided an approximate but quite workable solution to the problem With the remarkable success of this technique on neutron problem, it soon became Popular and found many applications in business and industry and at present forms avery important {00 of operation researcher's too] kit, “Monte Carlo 1s the code number given to the technique by the above two mathematicians ‘The technique employ’ random numbers and is and wherein physical experimentation is im, impossible. 1113 a method of simulation ‘out Monte Carlo simulation are: 1. Select the measure of effectiveness (objective function) of the problem. Iti either to be maximized or minimized, For example, it may be ide time of service facility in a queuing Problem or numberof shortages or the total inventory cost in an inventory contra problem 2. Identify the variables tht affect the measure of effectiveness significantly. For example, number of service facilites in a queving problem or demand, lead time and safety stock in an inventory control problem, Determine the cumulative probability distribution of each variable selected in step 2. Plot these distributions with values of the variables along x-axis and cumulative probability values along the y-axis, Get a set of random numbers from a random number generator. Consider each random number as a decimal value ofthe cumulative probability distribution Enter the cumulative distribution plot along the y-axis. Project this point horizontally tll it meets the distribution curve. Then project the point of intersection down on the x-axis, Record the value (or values if several variables are being simulated) generated in step 5 Substitute in the formule chosen for measure of effectiveness and find its simulated value. Repeat steps 5 and 6 until sample is large enough to the satisfaction of the decision make. ‘Though in any real problem to be solved by simulation, the variables in the problem are probabilistic in nature, we first consider below a problem where they are fixed and constant and see how the simulation technique which involves repetitive experimentation can be used to analyse such a problem. The remaining problems involve variables that are probabilistic im nature. used to solve problems that involve probability practicable and formulation of mathematical model is by sampling technique, The steps involved in carrying 6 1, fo.com/cgaspirantMRR ert eee eer eae tty From - www.cgaspirants.com 1196 @ OPERATIONS RESEARCH EXAMPLE 13.7-1 (Event-Type Simulation) Customers arrive at a service facility o get the required service. The interarrival and service times are constant and are 1.8 minutes and 4 minutes respectively. Simulate the system for 14 ‘minutes. Determine the average waiting time of a customer and idle time of the service facility. [Andhra Prod. Council. 1978} Solution ‘The arrival times of customers within 14 minutes period will be Customer 12 3 4 S$ 6 7 8 Arrival time (mimes); 0 18 36 S4 72 90 108 126 ‘The time at which the service begins and ends within time period of 14 minutes is shown below. Waiting time of customers and idle time of service facility are also calculated, Customer Service Waiting time Idle time of begins ends of customer service facility 1 0 4 ° 0 2 4 8 4-18-22 0 3 8 2 8-36-44 0 4 if 12-5.4=66 0 The waiting time of the first four customers is calculated above, For the remaining, it is caleulated below Customer 5 oy Waiting time (minuies) : 14-72-68 50 32 14 Average waiting time of a customer 0422444 +66+6845+32414 8 29.6 2 =37 minutes . mim Idle time of facility = nil EXAMPLE 13.7-2 Find the value of x experimentally by simulation. Solution Draw the coordinate axes OX and OY. With centre O draw an are PR of unit radius as shown in Fig. 13.1 and complete the square OPQR. Equation of the circle is x? + y? =I. From random number table C-1 (at the end of the book) select any two random numbers, say 0.2068 and 0.7298 (first two four-digit numbers with decimal from the second column) and let x= 0.2068 and y= 0.7295, Plot the point P, (0.2068, 0.7295). Obviously, if 3+ y? s 1, P, will lie inside or on ‘arc of the circle but if x?+y#>1, the point P, will lie outside the arc but within the square. Figs fb.com/egaspirants.com Download From - www.cpaspirunts.co SIMULATION & 1197 In ths manner hundreds or thousands of pais of random numbers are selected and it's ascertained whether the points representmg them he ann the are oF Beyond the are but Ince AE squire Suppose N ts the teal umber of pots considered out of which fe infon the are. Then in area enclosed by the arc Wi” ~ area of the square The above equation gives the experimental value of x. Obviously, the larger the sample size N. closer will be the obtained value to the true value of . EXAMPLE 13,7-3 Three points are chosen at random on the circumference of a circle. Find by Monte Carlo ‘methods the probability that they lie on the same semi-circle. Solution Draw a circle of circumference unity ie. of radius z as shown in Fig 13.2. Consider a ‘nplet of three random numbers (frst three two-digit numbers with decimal from the last column) (48.051 and 0.06 from the random number table C.1 These are poted as points A, B and C in igure, the distance of a point from O along the circumference anticlockwise being the value of a random number. Since for this triplet, the difference between the maxima (0.51) and minima (0.06) is tess than 0.50, the triplet lies obviously on a semicircle. 026 Fig. 19.2 ‘The general rule for ascertaining whether a triplet of random numbers lies on a semicircle fr not is as follows (®) Find the difference between the maxima and minima. If this difference is $ 0:50, the Inplet lies on a semicircle (i) If this difference 1s > 0.50, add unity to those random number(s) an the triplet which is (are) less than 0 50. Now find the difference between the new maxima and the minima, 1 this new difference is < 0.50, the triplet lies on # semicircle, otherwise not fo.com/egaspirantSe aa eee eee eee eee eee Download From - www.cgaspirants.com 1198 ¢ OPERATIONS RESEARCH Following the above rule it is ascertained whether each of the following 15 triplets lies (7) cof does not lie (*) on a semicircle. So Trplet Difference Difference between between macima new maxima and minima and minima 1 0790.56 006 079-006 = 073 106 - 0.56 = 0.50 ¥ 2 091-051 O13 091-013 = 078 113-051 =062 3065059051 065-051 = 0147 4 050013 094 094 -0.13 = 081 113-050 = 063 * 5 1087) 026 O78 078-026 = 0.52 126-057 = 0.69 * 6 033-060 031 060-031 = 0294 7 015 064 089 089-015 = 074 1.15 -064= 051 * 8 074099 063 0.99 0.63 = 0.36% 9 058 083. 0.44 083-044 = 0394 10 064 059 0.03 0.64 - 0.03 1.03 -0.59= 044 7 ul 059 0.30 0.16 059 -0.16 = 0.43 ¥ 2 057 «087 021 087-021 121-057 = 064 * 3 036 060 082 0,82 - 0.36 4 037° 072 033 15 090 076 0.29 129 - 0.76 = 0.53 * ‘Thus out of 15 triplets 9 lie on a semicircle, yielding the required probability = 7; =0 is However, if « large number of triplets are considered, the probability obtained will be very close to the theoretical value of 0.75. EXAMPLE 13.7-4 ‘A town has six wards and they contain 170, 510, 640, 75, 250 and 960 houses respectively. Make a random selection of 8 houses using the table of random numbers. Explain the procedure ‘adopted by you. [CA May, 1989) Solution Since the total number of houses is 2,605, first of all random numbers 0-2,604 are allocated in proportion to the number of houses in each of the six wards es shown in the table below: TABLE 13.1 Tillocation of random minbers to the houses w 2) a) oO oy Ward | No. of houses | Cumulative no. Range Random nos, fitted a. of houses T 170 170 1-169 2 510 680 170.679 (0590 (7), 0354 (8) 3 640, 1.320 680-1319 | 1128 (2), 0764 (4), 1292 (5) 4 5 1.395 1,320-1,394 1340 (6) 5 250 Loa 1395-1644 6 960) 505 L645-2.6004 2181 (). 1749 @) “The first random number picked up trom the randoin number table 1s 2181 (fist 4-diguts of the first random number in table C.1). Since it lies within the interval 1,645-2,604, i is fitted ‘against ward number 6 in column 5. ‘The next random number is 1128, which lies in the interval 680-1,319 and is, therefore, fitted against ward number 3 in column 5. The next random number 7112 is > 2,604 in column 4 and is, therefore, dropped from consideration. {n this manner the following random numbers are ether fitted in column 5 oF dropped; 6557 (D or dropped), 4199 (D), 3545 (D), 1749 (F3), 9103 (D), 0764 (Fa), 3493 (D), 1292 (FS), 4397 (D), 3807 (D), 4984 fh.com/cgaspiranta ve Download From - www.cgaspirants.com SIMULATION # 1199 (D> 1340 (F6), 0590 7), 9566 (), 7615 (Dp, 8808 (D), 6970 (D). 5799 (D), 6343 (D), 4165 (D), 0354 (F5) We stop because 8 houses have been selected. The eight houses selected belong {0 ward number 6, 3, 6, 3, 3,4, 2 and 2 respectively, EXAMPLE 13,7-5 & bakery keeps stock ofa popular brand of cake. Daily demand based on past experience ts given below: Daily demand 0 Is 25 35 45 50 srobabiliy 00h GIS 30. «30.2 a0? Consider the following sequence of random numbers 48, 78, 09, $1, $6, 77, 15, 14, 68 and 09 (i, jing the sequence, simulate the demand for the next 10 days. (i) Find the stock sitation i the owner ofthe bakery decides to make 35 cakes every day. ‘Also estimate the daily average demand forthe cakes on ihe back of the simulated data (PLU. B.Tech. (Me ch.) Dec., 2011; B.E., 2001; Nellore MBA, 2001; P.U.BBA, 2001 Solution below,” TM Simulsted demand forthe cakes forthe next 10 days can be obtained from the table low. TABLE 13,2 Allocation of random numbers to demand of cakes Pemand | Probability | Camuotnve | Random miner | Ransom inunibers fited probability | imierval 7 oor oor 00 5 o1s O16 ols 98). 15(7). 1408. 69C10) 25 036 16-35 35 ORG 36-85 48(1). 7802). $144), 5645), THO). 689) 4s 012 0198 6-97 50 002 Lo 98-99 In order to simulate the demand, the 10 days. Serial no. of these random numbers are shown in the parentheses, Number of cakes demanded in the next 10 days are 35, 35, 15, 35, 35, 35, 15, 15, 35 and 15 respectively, to ae nack situation for various days ifthe decision is made to make 35 cakes everyday is given in the table below: TABLE 13.3 Duy. Demand No. of cakes made Stock T 5 33 = 2 35 35 3 13 35 0 4 35 35 ey 5 35 35 20 o 35 35 20 ; 7 15 35 40 8 15 6 60 9 35 35 wo i 10 13 35 x0 fb.com/cgaspirantMeera eee eee eee eee eee eee eee eee Download From - www.cgaspirants.com 4200 & OPERATIONS RESEARCH ‘Average daily demand = + [354354 15 435.435 4.354 15 +15 +354 15) 270 wy = BT eakes. EXAMPLE 13.7-6 ‘A company manufactures around 200 mopeds. Depending upon the availability of ra materials ard other conditions, the daily production has been varying from 196 mopeds to 204 mopeds, whose probability distribution is as given below: Groduetionday 196 «(197 198 199 200 201 202 203, 204 Probability 005 009 0.12 O14 0.20 0.15 O11 0.08 0.06 The finshed mopeds are transported ina specially designed three-storeyed lorry thax ca acconmadate only 200 mopeds. Using the following 15 random numbers 82, 89, 78, 24, $3. 61. $8, 45, 04, 23, 50, 77, 27, 54 and 10, simulate the process to find out i) what will be the average number of mopeds waiting in the factory? (ii) what will be the number of empty spaces in the lorry? [CA. Dec. 1990: P.T-U. B.Tech, 2001) Solution ‘The random numbers are established as in the table below. TABLE 13.4 Fraduetion day [Probably | Cumulative probability | Random number interval 196 0.05 005 00 - of 197 009 oa 03-13 198 on 026 14-25 199 ow ao 26-9 200 020 060 40-59 201 os 075, 0-74 202 oul 086, 5 85 203 008 oss 86 -- 93 204 0.06 100, 4 = 99 Based on the 15 random numbers given, we simulate the production per day in the table below. TABLE 13.5 Darna | Random mmber | Production per day | No. of mopeds | Empry spaces in waning the lorry T 2 202 2 = 2 89 203 3 3 B 202 7 - 4 4 198 5 : 3 33 200 3 6 ol 20 6 7 Ix 198 4 z 8 45 20 4 e 9 4 196 - 7 lo 2 19K 2 u 50 2K | = R 1 202 2 B 2 19 t 14 54 200 ' 15 Wo 17 - 2 fh.com/cgaspirantSIMULATION % 1201 Average number of mopeds w ing inthe factory ab aeseres4s46ra tsar el + i=28 3 4 ‘Average numberof empty spaces inthe Tony = = 027 EXAMPLE 13.7-7 (Queuing Problem) Two persons X and ¥ work om a two-station assembly line. The distributions of activity times {at their stations are TABLE 13.6 Time 1 seconds | Time jrequency for X | Time frequency for Y 0 7 2 20 7 3 30 10 6 40 Is 8 30 35 2 60. Is 9 70 8 7 so 3 3 (@) Simulate operation of the line for eight items. (0) Assuming ¥ must wait unit X compl letes the frst item before starting work, will he have 0 wait to process any of the other seven ite -ms ? What is the average waiting time of tems for Y. Use the following random mumbers: [ ne For X 83, 70, 02, 12, 59, 46, $4 and 03, For¥ 51, 99. 84, 81, 15, 36, 12 and 54. (6) Determine the inventory of tems between the two stations (d) What is the average production rate? (RTM, Nagpur U, B.Tech. June, 2006; Kuru. U. BE. (Mech) June, 2012; 1993} Solution (@) Table below shows the cumulative frequency distribution for X. Eight random numbers given for person X are also fitted. The serial numbers of random numbers are shown in the parentheses, TABLE 13.7 Time in | Time frequency | Cumulative Range Random numbers fited seconds forX frequency 10 4 a 00-03 | 02). ONE) » 7 nl 04-10 w 10 a n-20 | aay 4 Is 26 21-38 0 35 1" Yo | 702), 5945), 460. 5407) ou Ik x9) Tews | xM) 0 8 ” BY. 96 80 3 hoo 97-99 5 wely Thus the shi umes for X are GU, 50, 10, 30, 50, 5U, 50 and 10 seconds respecti\ Likewise, te eight times fr Y are derived from bi cumulative datrbution below. fb.com/cgaspiranta et eet Download From - www.cgaspirants.com 1202 4 OPERATIONS RESEARCH TABLE 13.8 Time m | Time frequency | Cumulative frequency | Range | Random numbers seconds jor wwe en fied To 2 2 4 Oa 20 3 3 10 04.09 30 6 u 2 1021 156), 120), 40 8 9 38 7 3616) 30 R 3L a 38.01 511), S408) 60. 9 40 80 62-79 Ww 7 47 4 80.93 840). 814) 80 3 50. 100 94.99) 992) “Thus the eight times for Y are 50, 80, 70, 70, 30, 40, 30 and 50 seconds respectively, Note that the cumulative frequency has been multiplied by 2 to make it 100. (b) The above times for persons X and Y are used to calculate the waiting time, if any. TABLE 13.9 Trem 10. Person Person D Training time on | Waiting time on Time in Time out | Time in Tame out | _ the part of ¥_| the part of item 7 0 oo @ 110 oo = 2 60 10 no 190 - - 3 ho 120 190 260 - 70 4 120 150 200 330 - 10 5 150 20 3H 300 - 130 6 20 25 364 400 = 10 7 250 300 400 430 - 100 8 300. 310. 430 xo aH 120 ‘Thus person Y will not have to wait for the remaining seven items. 0+0+70+110 +130 +110 +100+120 _ 640 8 ay (©) In all there are 6 items waiting between the two stations. (@ Total time taken to process 8 items = 480 secs = 8 minutes. 8 Average production rate = 5 Average waiting time of items = item/minute, EXAMPLE 13.7-8 (Queuing Problem) A dentist schedules all her patients for 30 minutes appointments. Some of the patients take ‘more or less than 30 minules depending on the type of dental work 10 be done. The following summary shows the various categories of work, their probabilities and the time needed 10 complete the work. TABLE 13.10 Category Time required (minutes) _[ Probability of eategory Filling 6 a0 Crown oo as Cleaning 1S als Extraction 45 alo Checkup is 20 fb.com/cgaspirant24 Download From - www.cgaspirants.com SIMULATION # 1203 Simulate the dentists clinic for four hours and determine the average waiting time for the crac rhea the idleness ofthe doctor Assume tha ll he petona cron up at the clinic at hana titi scheduled arrival times, starting at 8AM. Une me Lollowing random numbers for handing the above problem: 40,82. 134,95, 66, Ty a IPTU. B.Tech. April, 2012, REM. Negpur UBE. (Mech) 2011: £7, 2009: ne, 2008; 4 (Final) Now, 1990) Solution is calor time taken by the dentist o teat the ight patients arriving im four hours at the clinic 's calculated in the table below TABLE 13.11 Category Time Probabiliy” T” Cuma | Random mo] Randoarvo counites) probabitiy | imerval Sited Filling 6 oa a0 0039 TT(3), 3404p, . 25(5), 1700) Crown oo os oss 4001) Cleaning 15 O15 070 6616) Extraction 45 10 080 70-79 | 748) Checkup 15 0.20 00 80-99 | 82) g hs the times taken by the dentin to woat he HEN patients are 60, 15, 45, 45, 45, 15,45 and 45 minutes respectively Let us simulate the dentist's clinic (for ight patients staring at 8 A.M. TABLE 13.12 Puiem T” Arvival | Denist® traiment Nain time on the) —Tlle me Jor Tie , me} Stars Ends _| pat of he ponent denist 7 8001800 900 5 = 2 830 | 900 ons 30 3 a0 | 915 ow 15 z 4 930) 1000 104s 30 . 5 Hou | 104s 30 43 7 6 030 7 130 nas oo s i Noo} tas 1230 45 - 8 1130 | 1230 134s “o = foc wate tie rte pts = 00 +154 59545 se) ooey - * = 35.625 minutes Average idleness of the dentist =Nil EXAMPLE 13.7-9 4 company manufactures 30 items per day. The sale of these tems depends upon demand which has the following distribution: Sales (units) Probability 7 0.10 2B Os 29 0.20 30 0.35 3 Os 32 0.05 fh.com/cgaspirant 2Reet ee reer eee ree ee eee ee eee eee eee Download-From—www.cgaspirants.com 1204 & OPERATIONS RESEARCH The production cost and sale price of each unit are & 40 and & 50 respectively. Any unsold product 18 to be disposed of ata loss of @ 15 per unit. There is a penalty of 85 per. unit if the Gemand is not met. Using the following random numbers estimate total profit/loss for the company {or the next 10 days : 10, 99, 65, 99, 95, 01, 79. 11. 16, 20. if the company decides to produce 29 items per day, what isthe advantage or disadvantage 10 the company ? [PT.U. B.Tech. April, 2012; M.D.U. Rohtak B.E. (Mech) Dec.. 2006) Solution First of all, random numbers 00-99 are allocated in proportion to the probabilities associated with the sale of the items as given below: TABLE 13.13 Tues mand | Probability [Cumulative probabiliy | Random mumber 1merval 27 010 010) oy » ons 033 to. » 020 045 25d 30 035 0.80 45.79 31 ous 095 80-94 2 0.05 100, 95.99 } Let us now simulate the demand for next 10 days using the given numbers in order to estate the total profi/loss for the company Since the production cost of each item is € 40 and sale price is & 50, therefore the profit per unit of the sold item will be & 10. There is a loss of 15 per unit associated with each unsold umit and a penalty of & S per unit if the demand is not inet. Accordingly, the profivloss for next ten days are calculated in column (iv) ofthe table below if the company manufactures 30 items per day. “TABLE 13.14 wo] a cn Ww) 0 Day | Random | Esumated | Profirloss per dav when Profit toss por day when ‘uber | sate | Production ~ 30 items per day | Production _ 29 tems per day T[ 28 2B E102 « F1S = 250 2+ WAL RAS = 263 2) » 32 W > 810-2 «85 = 290 29. 3 TSIM a] 6s 30 30» & 10 = 300 29 « FW-d « VS= INS 4] 9 32 30s 810-2 B5= 290 29» FW0-3 BS= 275 3 | 9s 22 30 « 810-2 x BS = 290 29. TW BS= IS «| oo 7 276 T1915 27 B12 + VAS =o a) 30 wee 29 TW» RS = INS a|ou »® Ws TIS DW TIO» FAS = 265 9 | 6 28 DW TMD VIS = 250] Wx VIL x VIS = 265 w | w 28 DTD TIF = 250] 2B + TOL «TAS = 265, Total profit, 22.095, 22.695 “The total proiit for next 10 days will be & 2,695 if the company manufactures 30 stems per day. In case, the company decides to produce 29 items per day, then the profit of the company for next 10 days is calculated in column (v) of the above table It is evident from this table that there is no additional profit or loss ifthe production 1s reduced to 29 items per day since the total profit remains unchanged Le, € 2.695 eaters EEEDownload From - www.cgaspirants.com 1214 & OPERATIONS RESEARCH EXAMPLE 13.7-17 (Queuing Problem) "A-a.amall store of ready made garmems, there is one clerk at the counter who ts to check the bills, receive payments and place the packed garments nto fancy bags. ete: The customers vmival at the check counter is a random phenomenon and the time between the arrivals varies From one minute to five mines, the frequency dstnbution for which ts given in table 13.37, The sence time time taken by the counter clerk) varies from one mime to three minues. The manager Of the store feels that the counter clerk 18 not suficiently loaded with work and wants fo 00g ci eee adltional work. But before taking the decision he likes to know precisely by what percentage of time the counter clerk is idle. TABLE 13.37 Frequency distribution of imierarrival tines Time benveen Frequency % Cnmutanve | Random no range arrevals (minutes) freyuency (ad T 5 35 = 2 23 oo 35-39 5 » 80 60-79 4 2 Dy 80-91 3 8 100 92-99) TABLE 13.38 Frequency disinibuiion of service times Service time Frequency % Cumulative | Random no. range amnutes) Srequency (26) 10 20 D 00-19 15 35 20-34 20 3 si 35-79 25 15 95 0-94 30) 5 too 95-99 Solution ‘As the random numbers are not provided in the problem, we select them from the random number table C-1 as indicated in table 13.39. The process of arrivals and service is simulated for 20 customers. TABLE 13.39 Grvals Random mimber Tnterarrival time Random no.(Last_ Service time 11) (Furst two digus of (mirmtes! (3) twa digits of last (minutes) (3) ast col, in table col, in table C-1) Ca a. T ae 2 2 TS 2 31 2 2 2 3 06 1 25 13 4 2 1 rT 13 5 19 3 2B 15 6 56 2 07 It 7 06 1 n 23 8 Py 4 44 15, 9 31 2 2 1 10 n 1 26 13 n 65 3 93 25 fb.com/cgaspirantDownload From - www.cgaspirants.com SIMULATION @ 1215 ee eM ATION 11215 2 59 2 13 31 2 M4 50 2 1s 13 i 16 94 s 7 37 2 18 26 1 19 78 3 20 33 1 seis io wait for 05 minute before the service can begin, Ths wey oe computations are the custo, a FO" 29 casomer aval thecounier clk side for 10 Seer he She customers have to wat for 7.5 minutes. The tal time taken by 30 orig minutes (Cumming up the inerarrival times), Therefore, the counter leks idle For about ssxt09 =25. Of the time. Based on this specific information, the mana some additional work to the counter clerk TABLE 13.40 Ber can lake the decision of assigning Interarrivl” Aewul Service Sere Service Coumer Crstomer’ Taree ra ened ee ind cura sags od wage win) nin Lime (min) time (min) z z TS 2 3s z 2 4 2 4 6 os 1 5 15 6 738, : 1 \ 6 13 15 9 : 1s 3 ° 13 9 Ws: 7 ‘ 2 u 1 " 2 os - 1 12 25 2 145 : - 4 16 13 16, 175 15 : 2 Is 1 8 v 05 : 1 19 is 9 203 : : 3 2 25 2 43 1s = 2 24 1 45 255 : os 2 %6 1 % ” os : 2 28 15 R 293 1 - 1 2» 2 293 313 : 03 5 uM 3 u ” 25 - 2 36 2 7 3» : ! 1 v7 1s » 40s 2 3 4 1 405 as : 0s 1 41 2 415, 435 - 05 105 75 fb.com/cgaspirantPRE eee eee eee eee eee ee eee eee eee eee eee Download From - www.cgaspirants.com 1216 @ OPERATIONS RESEARCH EXAMPLE 13.7-18 (Queuing Problem) A small scale factory has a machinist o process the jobs. The interarrival time for the job 4s not fixed. The processing time on the jobs also varies. These times are described by frequency distributions given below: TABLE 13.41 Tnterarrival time distribution Processing tne distribution Tnterarrival time (min) | __ Frequency Processing time (mn) | __ Frequency 4 7 3 3 5 20 4 30 6 40 5 30. 7 20 6 30 s 10 7 5 Too To0_ Determine the idle ume of the machinist and the waiting time of jobs in a period of 2 hours. If the machinist’ wages are % 20 per hour and the job’ waiting time costs & 30 per hour, would it be economical to engage a second attendant ? (PLU. B.Tech. April, 2012] Solution Table below shows the cumulative frequency distribution for interarrival times. Twenty five random nos, picked from random no. table are also plotted. The serial numbers of the random nos. are shown in the parentheses. TABLE 13.42 Tnierarrivel Frequency Cumulative Range Random nos. ted lime (min) frequency 4 10 10 00-09 ONS). 09(17). 02(23) 3 20 30 10-29 221), 1902). 16 >. 2518), 15(9), 18(15), 128) 6 40 70 30469 S810), STII). 4812). 6113), 36(14). 38(20), 5321), 40(22) 1 20 90 70-89 78(4), 78(7), 88(16), 85(19), 85(25) 8 10 100 90-99 9346), 95(24) ‘Thus the interarrival times of the jobs are 5, 5, 5,7, 4, 8, 7, 5, 5.6, 6, 6, 6, 6, 5.7. 4.5.7. 6, 6,6,4, 8, and 7 minutes respectively. Likewise, the twenty five jobs will have the processing times derived from the following table: TABLE 13.43 Processing Frequency Cumulative Range Random nos, fited time (min) Srequeney 7 3 3 0-04 4 30 35 (S34 1342), 099), 2014), 180, 24(11. 22112). O7(19), 29 (14), 3316), 1223), 3104, 3 30 65 35e64 ——57(15), A917), O4(I8). 57(22) 30 95 63-94 OBL), THS). B16), 9217). 9309), 9219) 5 5 100 95.9) __991R)_ 98120), 92H). 96025) fb.com/cgaspirant‘Download From - www.cgaspirants.com SIMULATION @ 1217 ‘Thus the processing times forthe jobs ae 6,4 4, 4, 6, 6, 6,7, 6,4, 4, 4s 4s 4 5. 4.5.5, 6 7,7, 8.4.4 and 7 minutes respectively “Table 13.37 can now be prepared by linking the interarival times and processing times of the jobs: TABLE 13.44 Tarerarrival Actual Processing Processing Processing Machinist Jobs fume min.) arrival tne time (min) — starts ends (man) ale time waiting nnd min) min.) time (nin) 3 5 © 5 T 3 - s w 4 " 1s : 1 5 1s 15 9 : : 7 2 4 2 6 3 : 4 %6 6 26 2 : : 8 34 6 aM 40 2 : 7 4 6 41 a 1 E 3 6 7 a7 34 - 1 5 sl 6 34 o : 3 6 37 4 60 of : 3 6 a 4 ot 68 : 1 6 ° 4 “ R 1 a 6 8 4 5 ” 2 = 6 8! 4 xt 85 2 : 5 86 5 % 1 1 : a 9% 4 23 7 2 z 4 7 5 7 102 : . 3 102 3 102 107 a e 7 109 6 109 us 2 2 6 U5 a us 122 : 7 From table 13.44, total idle time for the machinist in 2-hour period = 21 minutes, {otal waiting time for the jobs in 2-hour period = 9 minutes 2 +. Average ile time forthe machinist /hour = ==10.5 minutes and idle ime cost of the machinist / hour 95 ,20=83.50 9 Average waiting time forthe jobs / hour = £=4.5 minutes, and waiting ime cost ofthe 45 hour = 4x 30=% 2.25, jobs hour = ‘Since the idle time cost of the machinis/hour is more than the waiting time cost of the jobs! hour, it is not economical to engage the second attendant. EXAMPLES 13.7-19 (Queuing Problem) Ina workshop, fitters report at the tool crib after 8 A.M. to get the tools where the tool room attendant is present to issue them. The arrival and service time distributions are shown below: fb.com/egaspirant cmDownload From - www.cgaspirants.com 1218 ¢ OPERATIONS RESEARCH TABLE 13.45 Arrival disibnition ‘Service distribution Trrerartival ume tin) | _ Frequency | Service ume timmy | Frequency 35-45 008 1545 010 45-55 020 4 020 55-63 035 35465 040 65-75 025 65.75 020 75-85 o10 75-85 010 5-95 0.05 queue, What is the average time he ipeviod of 2 hours. Also find average busy in this ‘Determine the average waiting time of a fitter in the spends in the system? Simulate the queuing phenomenon for 0, ‘raitng time in non-empty queue and per cent time the tool room attendant remains period. Solution First, mean values of interarival times and servic frequency distributions for interarrival times are caleulated, Random numbers ‘their serial nos. are shown in parentheses, .e times are computed. Then cumulative are also plotted; TABLE 13.46 Teun inierarrival Frequency Cumulative Range — Random numbers fited me Canin) (a) frequency 4 3 3 o-04 OT (15) 3 20 28 05-24 15(1), 09). 20 (12) 6 35 ou 25.59 41.3). 55.7). 35 (9). 41 LO), 45 (13). 38 (14), 39 (18), 55.19) 7 25 NS Ged 74 (a), 72 (5). 67 (6). TH (8). 67 (16), 63 (17) 8 10 95 85-94 9 3 100 95.99 96 (11) 6,7.7,7,6, 7, 6,6,% 56,6, 4 7.7.6 Thus the interarrival times of the fitters are 5, and 6 minutes respectively. Likewise, the service times for the 19 fitters can be obtained from the following table: TABLE 13.47 Team serviee Frequency Cumalannre Range Random numbers fie Se oe ee 4 0 10 90211), 07 (12) 5 rm w 1029 20 (1), 22).23 15) 6 40 0 W469 HG), 54 (4), 408). 48 (8), 64 (13). 48 (17), 35 (18) 20 on 8972.2), 74 (9). 7610) im to on.) 95.14), 91 (16), 9119) “Thus the service times For the fiers are 5,7, 6, 6,6, 6, 3.6.7, 7,4, 4, 6, 8, 5, 8,6, 6 and 8 minutes respectively. Table 13.48 can now be prepared by linking the interarrival and service times of the fitters .com/cgaspirant Se ea aban) | 1 ean ma ee enn er naTeDownload From - wow.cgaspirants.com SIMULATION # 1219 TABLE 13.48 ia. _Tnterarrival Ach Service Service Service Fitters Antendam's vee Tact aon ae Set oe “tne time (min) (min (mm) (mind _teme (mun) (min) 7 3 B05 3 805 ‘8:10 - 3 a 5 R10 a io 817 : : 3 6 816 6 ai as 1 : 4 7 823 6 823829 : : 3 7 830 6 830836 : 1 6 7 47 6 ean has : 1 a 6 was 5 84 Rat : : 8 7 850 6 850856 : 2 9 6 856 a 856903 - : 0 6 902 7 90 910 1 : N 9 ou 4 Si ss - 1 2 s 916 4 916 920 : 1 B 6 922 6 922 928 : 2 Ma 9.28 8 92 936 : - 15 4 922 5 936 oat 4 : 16 939 8 om 949 2 : " 7 946 6 949935 3 : 18 6 932 6 935 wor 3 . y 6 938 Tos 14 1 From table 13.38, total waiting time on the part of all the 18 fitters = 14 minutes + Average waiting time of a fitter in the queue x 5 minute, 08 29 Average time a fitter spends in the system = Average waiting time of a fitter in non-empty queue 4 6 total at 3 minutes, Percentage time the too! room attendant remains busy ( 121-13 121 EXAMPLE 13.7-20 (Maintenance Problem) ) 10 108 9 ting time no, of eases when fitter has to p+ 62 mim tpg 78> minutes. «100 = 89.26%. 121 Determine the optimum number of mechanics for 100 semi-automatic machine tools. The ‘operation of the machine tools is automatic and warrants attention of the mechanics only when ‘here is break down. The breakdowns have been seen to occur at the following times: Break down Time (howr) 1 0 2 12 3 21 4 24 5 26 6 38 7 43 8 9 S160 fb.com/cgaspirantee aca Download From - www.cgaspirants.com Solution SIMULATION & 1225 ‘As observed from Fig. 13.3, there are two paths, namely 1-2-4-5 and 1-3-4-5 connecting the first to the last event, In tial 1, random numbers 8, 1. 8, 3 and 6 are selected for activities 1-2, 1-3, 2-4, 3-4 and 4-5 for which durations of 8, 2, 6, 6 and 4 days respectively are obtained ‘These durations are entered in Fig. 13.3. The critical path is 1-2-4-5 with duration o = 18 days. These data form row | of table 13.59. The remaining rows represent the data for tr 2,3, 10 respectively. TABLE 13.59 £84644 Frnal| aemwiy 2 | Aeuwigs 13 | Aetweny 24 | deuty 3d [ Aetna 4-5 | Proect | Creal no [RX [Durorion [RN [Puration | RN [Duration [RN [Duration RAN [Durarion| duration path soe sores ems eee ie [ras el 4 w fies s{s] 4 foe] a [7] 6 fat 6 4 frees Laas af7{ os fs] a frfo2 fof sx 3s 16 frsas sto} 1 jo s}oo fut a fs ne fr2as' Laas «| @ | 8 a fof2 fate fads B ql eaetesaoe |g | ae |r| ee [eee ae |e u @ sla: |e | |e eee] 6) ” of2] a fs] 4 fof 6 fof x fal 3 BR Tg ee (ee ee Ee pe ee 12 From table 13.59 itis observed that there is 70 per cent chance of path 1-2-4-5 to be critical, while there is 60 per cent chance that path 1-3-4-5 is critical ‘Average duration of the project = UGH e so H+ eT 417413 +1] 139 10 13.9 days. 13.8 GENERATION OF RANDOM NUMBERS In the illustrative examples 13.7.2 through 13.7.23 random numbers were either given in the problem or taken from random number table for carrying out hand computations. However, in ‘computer simulation, the random numbers can be obtained by the following methods: 1. Random numbers may be drawn from random number table stored in the memory of the computer. This process, however, is neither practicable nor economical. It is a very slow process and moreover, the random numbers occupy a large portion of the computer memory. Even on methodological grounds itis objectionable to use the same set of random numbers again and again, Therefore, several methods of generating the random numbers internally by the computer have been evolved. The necessary formulae occupy relatively litle space. 2. Anelectronic device may be constructed as part of the digital computer to generate true random numbers. The method, however, is considered to be very expensive. 3. Mid-square method of generating pseudo-random numbers {).com/cgaspirantDownload From - www.cgaspirants.com 1226 ¢ OPERATIONS RESEARCH A four digit number is taken. By squaring it a high digit figure is obtained, from which the ‘middle four digits are picked up. This yields the second random number and the process is repeated and a sequence of pseudo-random numbers is obtained. For instence, it x)= 2421, then x7 = 05861241 x)= 8612 x? = 74166544 = 1665, x = 02772225, = 4" 722 and 50 on. However, one may come across the following situations {} Tee series may vanish because random number obtained is 0000 () A random number reprodi 7600, xy = 7600 = 2100, 1, = 4100, x,,= 8100, x,,= 6100 and the ‘ype as in expression (13.2) "141 =(, + ) (modulo m) (132) IF, the congruence method is called the multiplicative ‘ype, as in equation (13.3), Fy ar, (modulo m), (133) The multiplicative methods are considered better t nixed methods, The generation of random numbers by in example 13.8.1 ‘selection of the values forthe constants a,b and mm is very important, because on them cpends the length ofthe sequence of random numbers afer which the sequence repeats. It is ‘ot possible to generate @ non- are ease rea rea rate aera reat ee eae ease eet eet eae ere eee eee eee eee eee eee Download From - www.cgaspirants.com SIMULATION 1227 Most of the computer systems have a subroutine available for generating random numbers. In DEC-20 computer system, a random number (soy Y) is generated by using the following statement. Y=RAN @). here x may be any number of alphabet EXAMPLE 13.81 The procedure of random number generation can be illustrated by taking some values for the constants a, b and m in equations (13.1) 0 (133). Let a= 16, b = 18 and m = 23. (@) Mixed Congruence Method ar, + b) (modulo m) Taking “1, x14 18 _ 3 Woxts18 M+ semsinder 1, a 8 lox 11418 _ 194 yo WH +18 194 4 semainder 10, yo ENB TE 8 ems 16x10+18_178 16x10 18 178 > semsinder 17, 0218 UE 7+ reminder | p= OATS 290 19 + remainder 14, see lox 14418 _ 242 +182 49 2, HE18 «22 = 10 + remainder 12, 6x12418_ 210 16x 12418. 210 3 E818 200 9 remainder 3, 16x3 +18 _ 6 rym ASRS IS SS oo sea . y= SHIR LS 2+ remainder 20 1620418 _ 338 y= ARRAS BS 14 + remainder 16, 16x16+18 _ 274 Sento 18 _ FA 11 + remainder 21, 16x21418 _ 354 rg= RRA 354 15 + remainder 9, 169418 _ 162 _ yo SEIS I 7 + remainder 1 The string of random numbers obtained is 1, 11,10, 17, 14, 12,3, 20, 16,21, 9, after which the sequence starts repeating, (8) Muluiplicative Congruential Method 71 = 16r, (modulo 23). Taking nel 16x1 r= S21 = 0 + remainder 16, 16x16 ne = 11 + remainder 3, 16x3 = 2+ remainder 2, and so on 2 fb.com/egaspirantDownload From - www.cgaspirants.com 1228 ¢ OPERATIONS RESEARCH This way the following sequence of random numbers can be generated 1, 16, 3,2,9,6, 4, 18, 12,8 13, 1 (©) Additive Congruential Method F012 (F,* 18) (modulo 23) Taking +1 Leg ‘ = 0 + remainder 19, By 70 + remainder | = 1 + remainder 14, = 1 + remainder 9, a= Sy = 1+ remainder 4, and so on, TPismettod results in the following string of 23 random numbers 119, 14,9, 4,22, 17, 12.7, 2, 20015, 10,5,0.18, 8 3 21, 16, 11, 6,1 13.9 SIMULATION LANGUAGES min specialised simulation languages, which are designed 9 ‘meet the following objectives: — t conveniently describe the elements whi generation of random variates for most flexibility of changing the design confi ich commonly appear in simulation, such as the of the statistical distributions iguration of the system so as to consider altemate ‘configurations, — ittemal timing and contro! mechanism for book keeping ofthe vital infonnation during the simulation run, {o obiain conveniently the data and statistics about the behaviour ofthe system [G.Provide simple operational procedures, such as altering the inital state ofthe system, and kind of output data to be generated, etc GASP and SIMSCRIPT are two widely used general simulation languages, which can easily do the job of FORTRAN or PL I. These are FORTRAN based languages and henoe the knowledge of FORTRAN is a pre-requisite for leaming GASP and SIMSCRIPT ‘The mest commonly used simulation language is GPSS (General purpose simulation sytem), Which was developed by IBM. It 1s easy to lear and incorporates all the features which are {nique to simulation. GPSS is problem oriented language, but has a wide range of applications It employs the next event incrementing time flow mechanism and uses integral time units, The calculations in integer arithmetic help to keep the round off errors to minimum. In GPSS, the System to be simulated is low charted in the form of block diagrams, and the blocks are then written in GPSS statements, fb.com/egaspirantCHAPTER — 13 Simulation 13.1, INTRODUCTION Simulation is the most important technique used in analyzing a number of complex systems where the methods discussed in previous chapters are not adequate, There are many real world problems which cannot be represented by a mathematical model due to stochastic nature ofthe problem, the complexity in problem formulation and many values ofthe variables are not known in advance and there is n0 easy way to find these values. ‘Simulation has become an important tool for tackling the complicated problem of managerial ‘decision-making, Simulation determines the effect of a number of alternate policies without disturbing the real system. Recent advances in simulation methodologies, technical development and software availability have made simulation as one of the most widely and popularly accepted tool in Operation Research. Simulation is a quantitative technique that utilizes a computerized mathematical model in ‘order to represent actual decision-making under conditions of uncertainty for evaluating alternative courses of action based upon facts and assumptions. John Von Newmann and Stainslaw Ulam made first important application of simulation for determining the complicated behaviour of neutrons in a nuclear shielding problem, being too complicated for mathematical analysis. After the remarkable success of this technique on neutron problem, it has become more popular and has many applications in business and industry. The development of digital computers has further increased the rapid progress in the simulation technique. Designers and analysts have long used the techniques of simulation by physical sciences. Simulation is the representative model of real situation. Fore example, ina city, a children’s park with Various signals and crossing is a simulated model of city traffic. A planetarium isa simulated model of the solar system. In laboratories we perform a number of experiments on simulated model to predict the behaviours of the real system under true environment, For training a pilot, Might simulators are used. The simulator under the control of computers gives the same readings as the pilot gets in real flight. The trainee can intervene when there is signal, like engine failure etc. Simulation is the process of generating values using random number without really conducting experiment, Whenever the experiments ae cosly or infeasible or ine-consuing simulation is used to generate the required 13.2. DEFINITION 1. Simulation is a representation of reality through the use of model or other device, which will react in the same manner as reality under @ given set of vonditionsOperations Ressearch 618 2. Simulation is the use of system model that has the designed characteristic of reality in order to produce the essence of actual operation, 3. According to Donald G. Malcolm, simulation model may be defined as one which depicts the working of a large scale system of men, machines, materials and informat jon operating over a period of time in a simulated environment of the actual real world conditions 4+ According to Naylor, eta. simulation isa numerical technique for conducting experiments one digital computer, which involves certain types of mathematical and logical relationships catemded oo, describe the behaviour and structure of a complex real world system over extended period of time, ‘There are two types of simulation, they are: {- Analog Simulation: Simulating the reality in physical form (e.g.: Children’s park, planetarium, tc.) is known as analog simulation, 2. Computer Simulation: For problems of complex ‘managerial decision: Computers then solves the problem. Such type ofs or system simulation, 13.3. CLASSIFICATION OF SIMULATION MODELS Simulation models are classified as: (a) Simulation of Deterministic models: In the case ofthese models, the input and output variables are not Permitted to be random variables and models are described by exact functional relationship. (6) Simutation of Probabilistic models: In such cases method of random sampling is used, The techniques used for solving these ‘models are termed as Monte-Carlo technique. (c) Simulation of Static Models: These models do not take variable time into consideration, (A Simulation of Dynamic Models: ‘These models deal with time varying interaction, 13.4. ADVANTAGES OF SIMULATION Simulation isa widely accepted technique of operations research due to the following reasons: * Itis straightforward and flexible, * Itcan be used to analyze large and complex real world situations that cannot be solved by conventional quantitative analysis models, * Itis the only method sometimes available. *Itstudies the interactive effect of individual components or variables in order to determine which ones are important. * Simulation model, once constructed, may be used over and again to analyze all kinds of different situations,Simulation. on + It is the valuable and convenient method of breaking down a complicated system into subsystems and their study, Each of these subsystems works individually or jointly with others. 43.5. LIMITATIONS OF SIMULATION TECHNIQUE Jn model mostly deals with uncertainties, ns involving statistical errors, opti the results of simulation are only * Since simul jum results eannot be produced by reliable approximat simulation, + In many situations, it is not possible t0 i of the system, + In-very large and complex problems. it is very difficult to make the computer program in view ofthe large number of variables and the involved inter-relationship among them. + For problems requiring the use of computer, simulation may be comparatively costlier and time consuming in many cases. Each solution model is unique and its solutions and inferences are not usually other problems, which can be solved by other techniques. identify all the variables, which affect the behaviour iransferable 10 43.6. MONTE-CARLO SIMULATION “The Monte-Carlo method isa simulation technique in which statistical distribution fun using a series of random numbers. Working on the digital computer fora few minutes we can create data for months or years. The method is generally used to solve problems which cannot be adequately represented by mathematical models or where solution of the mode! is not possible by analytical method. Monte-Carlo simulation yields solution, which should be very close to the optimal, but not necessarily the exact solution, But this technique yields a solution, which converges to the optimal solution as the ‘numberof simulated trials tends to infinity. The Monte-Carlo simulation procedure can be summarized inthe following steps: Step 1: Clearly define the problem: (a). Identify the objectives of the problem. (b) Identify the main factors, which have the greatest effect on the objective of the problem. Step 2: Construct an approximate modi (a) Specify the variables and parameters of the mode. (0). Formulate the appropriate decision rules, ie. state the conditions under which the experiment is to be performed. (©). Identify the type of distribution that will be used. Models use either theoretical distributions ‘or empirical distributions to state the patterns ofthe occurrence associated withthe variables. (4). Specify the manner in which time will change. mas are created Problem 13.1. With the help of a single server queuing model having inter-arival and service times constantly 1.4 minutes and 3 minutes respectively, explain discrete simulation technique taking 10 minutes as the simulation period. Find from this average waiting time and percentage of idle time of the facility of @ ‘customer. Assume that initially the system is empty and the first customer arrives at time t = 0. { |620 Operations Research Solution Data: System i initially empty. Service starts as soon as frst customer arrives. First customer arrives at t= 0 ‘The departure time of first customer = the table. The second customer arrives at | first customer leaves the system, ‘waiting time for second custom the time of arrival 0 + 3 .¢. arrival time + service time = 3 minutes (Dep) in 4 minutes and third arrives at 2.8 minutes (Arr). Until the: We can calculate first customer and. - 1.6 minutes. The procedure is shown in the table Time | Event Arr = arrival | Customer Waiting time. Dep = departure | Number 0.0 Ar, 1 ~ 14 Ar, 2 : 28 Ar 3 S 3.0 Dep 1 | 3.00~ 1.40116 min, for customer 2. 42 Arr, 4 36 An. 5 6.00) Dep 2 __ | 600-280 3.2 min. for customer 3, 7.00) An. 6 84 Ant, 7 9.00 Dep. 3__| 9.00-4.20=4.8 min. for customer 9.80 Arr. 8 40.00 | End of given time 10.00 ~ 5.60 = 4.4 min, for customer 40.00 - 7.00 = 3.0 min, for customer 6 10.00 8.4= 1.6 min, for customer? 10.00 -9.80 = 0.2 min, or customer 8. Average waitin ig time per customer for those who must wait = / number of waiting times taken=(I.4+28+42 1564708, Percentage of idle time of server = ‘Sum of waiting time ofall customers 4+9.8)/7= 18.8 /7= 2.7 minutes. Sum of idle time of server / total time = 0%,‘Simulation 21 43.7, RANDOM NUMBERS Random number is a number in a sequence of numbers whose probability of occurrence is the same as that of any other number in that sequence. 13.7.1, Pseudo-Random Numbers ical Random numbers which are generated by some deterministic process but which satisfy statisti test for randomness are called Pseudo-random numbers. 43.7.2. Generation of Random Numbers ion one can generate Pseudo-random numbers. These methods ‘where starting with an intial number called seed is generates the imber of recursive procedure are available, {due method. This method is described Using some arithmetic operati ‘most commonly specify a procedure, second number and from that a third number and so on. A nu the most common being the congruence method or the resi by the expression: = (ar; +) (modulo m), ‘a ae the ith and (j + 1th random numbers. “The expression implies multiplication of « by r, and addition of b and then division by m. Then The canainder or residue. To begin the process of random number generation, in addition to a,b is also required. It may be any random number and is called seed. he Where a, b and m are constants, r; and r; Problem 13.2 ‘With the help of an example explain the additive multiplicative and mixed types ofthe congruence random number generators, im, the value of ro Solution “The ongruence random number generator is described by the recursive expression revi (ar; +4) (modulo m), ‘Wherea,b and m are constants. The selection of these constant is very important as it determines the starting of random number, which can be obtained by this method. The above expression is for a tnixed type congruential method as it comprises both multiplication of a and r, and addition of ar, and 6, If expression, ‘When 6 = 0, the expression obtained is r,, 4 = (ar 1) (modulo m), this is known as multiplicative method. ‘To illustrate the different types of the congruence methods, let us take and let the starting random number or seed be ry (a) Mixed Congruential method: 1, = (ar; + 6) (modulo am), therefore, the expression reduces to 7). 1 = (Fi+4) (modulo m). This is known as additive type 16, b= 18 and m = 23622 | ries (ar, +b) (modulo m), | = Residue n (16 «1+ 189/23, 34/23 | 1 residue IT n (16% 11 +18) 23 194/23 | 8 + residue 10 " (16 * 10+ 18)/23 178/23 | 7+ residue 17 t% (16 * 17+ 18)/23 290/23 | 12+ residue 14 *s (16 14 + 18) 123 242/23 | 10+ residue 12 re (16 12 + 18)/23 210/23 | 9+ residue 3 6 (16 * 3+ 18)/23 66/23 | 2+ residue 20 ts (16 * 20+ 18)/23 338/23 | 14+ residue 16 % (06 « 16+ 18)/23 274123 | 11 + residue 21 ro (16 21 + 18)/23 354/23 | 15+ residue 9) nm (16 «9 + 18/23 162/23 | 7+ residue 1 ‘The random numbers generated by this method ar (0) Multiplicative Congruential Method: The string of random numbers obtained b) 4, 18, 12, 8, and 13, tie1 "ar, (modulo m) £5 | rerear, (modulo m) | Random Number fe (161/23 (0+ Residue 16 rs | 06%169/23 [+ Residues 6 (U6 * 39/23 2+ Residue 2 % (U6 * 29/23 1+ Residue 9 6 (U6 * 9723 6+ Residue 6 ts (16 « 6)/23 44 Residue 4 a (16 * 4)/2: 2+ Residue 18 re | 619/23 | 12+ Residue 12 ry | (ex 129/23 84 Residue 8 ra | (16 *8)/23 5+ Residue 13 ru 6% 139/23 9+ residue | Operations Ressearch re: 1, 11, 10, 17, 14, 12, 3, 20, 16, 21, and 9, 'y multiplicative congruential method is 1, 16, 3, 2,9, 6,Simulation (c)_ Additive Congruential Metho 4 = (ys 6) (modulo m). 623 7 | fier = (61+) (modulo m) | Random Number a (14 18)/23 OF residue 19 A (9+ 18/23 1 residue 14 1 (14+ 189/23 1 esidue 9 1 (+ 189/23 residue 4 6 (4+ 18)/23 (0+ residue 22 ts (2+ 18/23 1+ residue 17 » (17+ 18)/23 1+ residue 12 ts (12+ 18)/23 1 residue 7 % (7+ 18/725 1 residue 2 to (2+ 18)/23, (0+ residue 20 mi (20+ 18/23 1 residue 15 ‘The random numbers generated are: 1, 19, 14, 19, 4, 22, 17, 12, 7, 2, 20, and 15. Problem 13.3. “The distribution of inter-arrival time ina single server model is 2 2 % | % z 2 ‘And the distribution of service time is % the two digit random numbers as 12, 40, 48, 93, 61, 17, 55, ‘Complete the following table using t 21, 85, 68 to generate arrivals and $4, 90, 18, 38, 16, 87, 91, 41, 54,11 to generate the corresponding service times: ‘Arrival | Random | Arsivat | THE | Random | Ti | waiting time Number | Number | Time} SEV | umber | Sele) in Queue Begins ends Solution ‘The distribution of inter-arrival times and the two-digit random numbers assigned to different values of T is as below:624 Operations Ressearch tT | mo ry) ‘Random numbers | 025 025 001024 2 | 050 075 251074 3 | oas 1.00 751099 Inter-arrival times are Given 1, 2, 2, 3, 2, corresponding to random numbers 12, 40, 48, 93, 61, 17, 5 two-digit random numb 5, 21, 85 and 68 » 1. 2. 1, 3, 2 respectively. Similarly, the distribution of service times and ers assigned to different values of $ are as follows: S| fe Le Random number 1 | 050 050 001049 2 | oas 075 251074 3 | 02s 1.00 751099 ‘The simulation is done as follows: Tine Tine Arrival | Random | Arival | Service | Random | Serice | Waiting smumber \ umber | tine | Begins in | ‘number | Endean | Tite in Mins. Mins, | Quewe 1 2 1 U 4 3 7 2 «0 3 3 90 6 : 3 8 5 6 1s 7 1 4 3 & 8 38 8 : 5 6 ‘0 10 16 i : 6 7 ur u 87 14 : 7 35 B 4 3 " 1 3 2 4 " a 18 3 9 85 " 18 34 20 1 10 @ 9 20 uw 2 1 ‘The working of the above table is as below: The simulation of the single-server system starts at zero time. First customer arrives at | time unit after that and the service immediately begins. Since the service time for the first customer is 2 time units, service ends at 3 time units. The second Customer arrives after an inter-arrival time of 2 time units and goes to service immediately at 3 time tunits. The third customer who arrives at 5 time units has to wait tll the service of 2nd customer ends at 6 units of time. The other entries are also filled on the same logic.Simulation = Problem 13.4. A coffee house in a busy market operates counter servi hhas approached you with the problem of determining the num ‘counter. He wants that the average waiting time ofthe customer sl recording the data for a number of days, the following frequency distrib tomers and the service time atthe counter are established. Simulate the system for 1e suitable answer to the problem. ice, The proprietor of the coffee house ber of bearers he should employ at the hhould not exceed 2 minutes. After ution of inter-arrival time of 0 arrivals of | various alternative number of bearers and determine the Tmterarrival mein mins. | Frequency (29) | Service me in mins. | Frequency 0) 0 5 10 5 0s 35 20 25 10 25 3.0 35 1s 15 40 20 20 10 5.0 15 25 7 3.0 3 Solution Itis queuing sit number of bearers, the waiting time of th System and waiting customer can enter any of the service cham By taking tvoedigit random number interarrival and interservice times areas follows: Random number for arrivals: fon where customers arrive at counter for taking coffee. Depending upon the 'e customers will vary. It is like a single queue multi-channel inel as and when one becomes available. “earrivat time in minutes | Frequency | Cumulative frequency | Random numbers o 5 5 (0010 04 os 35 0 051039 10 25 6 4010 64 15 15 80 651079 20 10 %0 801089 25 7 7 90 10 96 3.0" 3 100 97 10 100 Random number for Service: ‘Service time in minutes | Frequency | Cumulative frequency | Random number 1a 5 5 ‘01004 a 25 30 05 1029 aa 35 65 30 10 64 ae 20 85 65 10 84 ea 15 100 85 10 99Operations Ressearch tor | Nor | 47! | Ker | Tome’ | “Tine sence | San | seme | Seer | in 2 « [3 [10 1.00 3 sv Lo 24 2 20 30 50 19 7 6 1.0) 32 3 60 80 11.00 20 [ : ‘The customer waiting time with two servers is sometimes greater than 2 minutes. Hence let us tty with one more bearers. The table below shows the waiting time of customers with three bearers With two bearers, total waiting tim 1.8 minutes. is 18 minutes. Hence average wi Server | Sener | Sever | Sonar | Sener [Somme | o Arrival | One One Two Two | Three | Three pant amber | Serie | Sere | Service| Semice| Some | Sernce | ating Begins | Ends | Begins | Ends | Begins | Ends | Time 1 | 00 [30 o 2 to | 40 o 3 i 20 | ao |e a Ee 05 5 «9 | 70 10 6 5.0 9.0 0 7 | «0 | 90 Q s 70 | 90 Q 9 9.0 12.0 o 10 [10 [20 0 With three bearers, the total waiting time is 1.5 minutes. Average waiting time is 0.15 minutes. Similarly, we can also calculate the average waiting time of the bearers