Chapter 13

Management Science:
Operations Research(OR)
Chapter 13
Simulation
24-01-13 Lecture 1
 The main purpose of simulation is the
prediction of what is going to happen as a
result of our policies (that have already be
decided upon).
 Simulation does not optimize anything.

Deterministic vs. stochastic
simulation.
 Depends on the data fed into the system
that may be deterministic or stochastic.
Here: stochastic simulation with random
elements (Monte Carlo simulation).
Continuous vs. Discrete event
simulation.
 Flight simulators: continuous
(movements & reactions)
 Discrete events: Cases of measles

reported to & confirmed by a given
hospital, number of truckloads of a
product delivered to a retailer
Main steps of a discrete-event
simulation
 Building of the model,
 Assigning numbers to uncertain events
according to their likelihoods,
 Generation of uncertain events,
 Application of the predetermined policies,
 Evaluation of the results including
verification of the model.
Assign numbers to random
events.
 General rule: Assign numbers so as to preserve
probabilities.
Example: Assignment of random numbers to

demands
Demand 20 35 50 60
Probability .1 .3 .5 .1
Numbers assigned to event 01-10 11-40 41-90 91-00
Random numbers:
 Different distributions It is possible to
start with uniformly distributed random
numbers (all digits are equally likely) &
then use a formula to convert them to
random numbers that follow other
distributions.
Midsquare method (John von
Neumann)
 Idea: Take the middle part of a number,
square it, & repeat.
 Example: Take the seed x0 = (3456). The

square of this number is (3456)2 =
11943936, whose middle part is the next
random number, i.e., x1 = 9439.
 (9439)2 = 89094721 → x2 = 0947,
 (0947)2 = 896809 → x3 = 9680,
 (9680)2 = 93702400 → x4 = 7024,
 Problem with some sequences:
x0 = 3,792. Now (3,792)2 = 14,379,264, so that
x1 = 3,792 & the sequence has converged.

Linear congruence methods
xi = (a + bxi1) mod c
 Example: Let a = 7, b = 9, & c = 93.
x0 = 19, we obtain
x1 = [7 + 9(19)] mod 93 = 178 mod 93 = 85,
x2 = [7 + 9(85)] mod 93 = 772 mod 93 = 28,
x3 = [7 + 9(28)] mod 93 = 259 mod 93 = 73,
x4 = [7 + 9(73)] mod 93 = 664 mod 93 = 13,
13.3.1 Simulation of a Waiting
Line System
 Purpose: Evaluate the system. Potential tools:
 (1) probability that a customer will have to wait,
 (2) the average waiting time, &
 (3) the average number of customers in the system
 (really a proxy for the congestion of the
system).
 (4) the average idle time during a workday, &
 (5) the cost of the system.
 One channel, interarrival times are
uniformly distributed on the integers in [4,
9], service times are uniformly distributed
on the integers in [5, 7].
Random digits:
2049 9135 6601 5112 5266
6728 2188 3846 3734 4017
7087 2825 8667 8831 1617
7239 9622 1622 0409 5822
6187 0189 5748 0380 8820
3606 7316 4297 2160 8973

(1) (2) (3) (4) (5) (6) (7) (8) (9) (10)
1 4 9:04 9:11 0 4 9:04 7 0 4
2 9 9:13 9:20 0 2 9:13 7 0 6
3 9 9:22 9:27 0 2 9:22 5 0 8
4 5 9:27 9:33 0 0 9:27 6 0 8
5 6 9:33 9:39 0 0 9:33 6 0 8
6 6 9:39 9:46 0 0 9:39 7 0 8
7 5 9:44 9:52 2 0 9:46 6 2 8
8 5 9:49 9:59 3 0 9:52 7 5 8
9 6 9:55 10:06 4 0 9:59 7 9 8
10 6 10:01 10:12 5 0 10:06 6 14 8
11 6 10:07 10:18 5 0 10:12 6 19 8
12 7 10:14 10:23 4 0 10:18 5 23 8
13 8 10:22 10:29 1 0 10:23 6 24 8
14 8 10:30 10:37 0 1 10:30 7 24 9
15 8 10:38 10:43 0 1 10:38 5 24 10
24-01-13 Lecture 15
1) Customer number
2) Interarrival time (a random number from
the list)
3) The arrival time of the customer
4) The time at which service performed for the
customer is finished and the customer
leaves
5) The time the customer had to wait (not
including service time)
6) The time that the service station is idle
before the customer arrives
7) Starting time of service for the customer
8) Service time for the customer (a random
number from the list)
9) The cumulative waiting time,
10) The cumulative idle time.
 Evaluation: Note some idle time in the
beginning, later the system is very busy.
Not surprising: the average interarrival
time is 6.5 minutes, the average service
time is 6 minutes → average traffic
intensity  = .9231. this is a very busy
system.
24-01-13 Lecture 18
 Total waiting time for all 15 customers
(column 9) is 24 minutes. Hence, the
average waiting time is 24/15 = 1.6
minutes per customer. (Getting longer as
time progresses).
 Idle time of the system: The system has
been operating from 9:00 until the last
customer left at 10:43, i.e., for a total of
103 minutes.
 The total idle time (row 15 of column 10) is

ten minutes → the total idle time is 10/103
 10% of the time. (Idle time mostly during
the early stages).
24-01-13 Lecture 20
2-21
Processes & Operations
Internal and
external customers
Inputs
• Workers Processes and
• Managers operations
• Equipment Outputs
• Facilities 1 3 • Services
• Materials 5 • Goods
• Services
• Land 2 4
• Energy
Information on
Figure 1.1 performance
1-22
24-01-13 Lecture 23
MRP II Figure 13.11
Customer orders Forecasts
Master production schedule
Bills of materials
MRP Inventory records
Routings
Inventory transactions
Time standards explosion
Material requirements plan
Cost and
Manufacturing resource plan
financial data
Financial/ Sales and Human

Purchasing Manufacturing
accounting marketing resource
reports reports
reports reports reports
13-24
24-01-13 Lecture 25
Lean Systems
 Operations systems that maximize the
value added by each of a company’s
activities by paring unnecessary resources
and delays from them
 Just-in-time (JIT) philosophy
 The belief that waste can be eliminated by
cutting unnecessary capacity or inventory and
removing non-value-added activities in
operations
 JIT system
 A system that organizes the resources,
information flows, and decision rules that
enable a firm to realize the benefits of JIT
principles
10-26

Chapter 13

Uploaded by

Copyright:

Available Formats

Chapter 13

Uploaded by

Document Information

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Chapter 13

Uploaded by

Copyright:

Available Formats

Management Science:

 Simulation does not optimize anything.

 Discrete events: Cases of measles

Example: Assignment of random numbers to

 Example: Take the seed x0 = (3456). The

x1 = 3,792 & the sequence has converged.

 Example: Let a = 7, b = 9, & c = 93.

 The total idle time (row 15 of column 10) is

Customer orders Forecasts

Master production schedule

Material requirements plan

Financial/ Sales and Human

You might also like