Deck 14: Simulation Cdrom Modules

Question 1

Using simulation one should be able to find the optimal solution in a reasonable amount of computation time.

Answer 1

False

Question 2

For simple waiting line systems, simulation makes it easier to find the operating characteristics than waiting line theory.

Answer 2

False

Question 3

An advantage of simulation is that decision makers can see the effects of a policy over several years before making a decision.

Answer 3

False

Question 4

In simulation, several alternatives are evaluated and one chooses the best among the alternatives evaluated.

Question 5

Most practical applications of simulation can be solved using a paper and pencil approach.

Question 6

Probabilistic simulation is akin to sampling, each run representing one sample.

Question 7

In general, the higher the variation in the output values of parameters of interest in repeated runs, the smaller the number of runs needed to estimate the expected values of these parameters with a given confidence level.

Question 8

Most practical applications of simulation can be solved using a paper and pencil approach.

Question 9

The purpose of validation is to make sure that the computer program does what was intended to be done.

Question 10

The purpose of validation is to check whether model results adequately represent real system performance.

Question 11

If the variability of the parameter of interest as per simulation is higher, greater is the likelihood that the real system differs from the simulated system.

Question 12

In theory, the numbers on a random number table follow these properties: all numbers are equally likely, and if you take the numbers in a row only (left to right or right to left, but not top to bottom or across), no pattern appears in the sequence of numbers.

Question 13

An important benefit of simulation is the ability to conduct the experiment under controlled conditions.

Question 14

In a fertilizer plant, there is an intermediate storage tank for slurry. A new intern developed a simulation model to study the relationship between volume of slurry in the tank and time of day. This is an example of discrete simulation.

Question 15

Fixed interval simulation is appropriate when the user is interested in how many events occurred, rather than exactly at what time or at which location the event occurred.

Question 16

In simulation several alternatives are evaluated, and one chooses the best among the alternatives evaluated.

Question 17

Probabilistic simulation is suitable for problems that do not have an inherent fixed answer.

Question 18

In the 7-step process for simulation, 'Run Simulations' comes before 'Validating the Model'.

Question 19

Comparing the results of a simulation model under construction with the known system performance under identical circumstances will help you to validate the system.

Question 20

All deterministic simulations incorporate mechanisms for mimicking a random behavior of one or more variables.

Question 21

Using simulation, one can exhaust all possible options for a problem and thus find the optimal solution, though it may take a little more computation time.
The only method to validate a simulation model of a non-existing system is to use the test of reasonableness.

Question 22

Using the equation $T=-\frac{1}{\lambda} \ln (\cdot R N)$ large random numbers produce small time values.

Question 23

In a normally distributed random number table, about half the numbers will be negative.

Question 24

In a normally distributed random number table, most numbers will be between -3 and +3 .

Question 25

In a normally distributed random number table, about $33 \%$ of the numbers will be between -1 and +1 .

Question 26

Using excel, once we create a set of random numbers with the formula Rand(), then the numbers will remain the same even if we refresh the sheet.

Question 27

In excel, the 'Paste Special' function (value only) helps us to preserve the set of random numbers so that calculations based on random numbers can be verified.

Question 28

Simulation is a particularly well-suited technique for multiple server Poisson arrival, exponential service waiting line problems.

Question 29

Simulation is a particularly well-suited technique for multiple server General arrival and general service waiting line problems.

Question 30

Two properties of random number tables are

A) uniformity and randomness
B) normality and order
C) independence and normality
D) independence and uniformity

Question 31

Important benefits of simulation include

A) handling of problems that cannot be solved analytically
B) always finding the optimal solution
C) relatively fast and cheap calculations
D) evaluating and finding an exact solution for the problem at hand

Question 32

If you are interested in predicting the time of day a road block may occur, you would use

A) fixed interval simulation
B) next event simulation
C) deterministic simulation
D) any one of $\mathrm{A}, \mathrm{B}$ , or $\mathrm{C}$ can be used

Question 33

Aspects of validation include

A) test of reasonableness
B) generating results similar to known systems
C) checking assumptions and values of parameters used in the model
D) all of the above

Question 34

There is a backyard lake in your house with a very odd-shaped boundary. You drew a computer map of it, enclosed it in a rectangle, and used simulation to estimate its surface area. You must have used

A) fixed interval simulation
B) next event simulation
C) deterministic simulation
D) probabilistic simulation

Question 35

The first step in simulation modeling is to

A) gather data
B) develop the model
C) define the problem and set objectives
D) validate the model to make sure that it can be used for the problem

Question 36

The second step in simulation modeling is to

A) gather data
B) develop the model
C) define the problem and set objectives
D) validate the model to make sure that it can be used for the problem

Question 37

The number of runs needed to estimate the expected value of a parameter of interest

A) increases with the variance of the observed variables of interest
B) increases with the confidence level demanded for the expected value
C) Decreases with the variance of the observed variables of interest
D) A and B only

Question 38

An analyst is simulating demand, which is hypothesized to follow a uniform distribution in the range of $[20,39]$ . Allowing only integer values, picking 2-digit random numbers and associating 00-04 with 20, 05-09 with 21, etc., what will be the simulated demand corresponding to a random number choice of 43 ?

A) 28
B) 27
C) 26
D) 29

Question 39

All of the following are advantages of simulation except that

A) it can experiment with system behavior in a controlled setting
B) it can solve even problems which have analytical solution techniques
C) it is well-suited for even problems intractable using other methods
D) allows time compression to study long term effects

Question 40

All of the following are reasons to be wary of using simulation except

A) probabilistic simulation results are approximate rather than optimal
B) can be costly and time consuming
C) requires expertise and training
D) can solve even simple problems, such as single server Poisson arrival and exponential service E) waiting line problems

Question 41

All of the following approaches can be used to simulate an exponentially distributed random variable (.RN is a random number between 0 and 1.0) except

A) $\left(T=-\frac{1}{\lambda} \ln (\cdot R N)\right)$
B) $\left(T=-\frac{1}{\lambda} \ln (1-. R N)\right)$
C) $\left(T=\frac{1}{\lambda} \ln (\cdot R N)\right)$
D) $(T=-\mu \ln (\cdot R N))$ , where $\mathrm{T}$ is average time

Question 42

Using a standard approach to simulating random variables which follow Uniform Distribution UN [20,30] , what would be the simulated value if 30 is the two-digit random number picked from the random number table?

A) 23
B) 22
C) 24
D) 6

Question 43

Time between failures of a critical machine follow normal distribution, with a mean of 30 hours and standard deviation of 5 hours. If the random normal deviate chosen from a table of normally distributed random numbers is 2.5 , the number of simulated hours before the next failure will be:

A) 17.5
B) 32.5
C) 27.5
D) 42.5

Question 44

Using standard approach to simulating random variables that follow negative exponential distribution with an expected time of 3 minutes, what would be the simulated time (in minutes) if 30 is the two-digit random number picked from the random number table? $\left(T=-\frac{1}{\lambda} \ln (\cdot R N)\right)$

A) 0.060199
B) 0.4013
C) 3.6119
D) 24.079

Question 45

Simulating a discrete distribution with two events- $\mathrm{H}$ with probability 0.35 and $\mathrm{T}$ with probability 0.65 - using two-digit random numbers is usually done by associating a random number from 00 to 34 with $\mathrm{H}$ and 35 to 99 with $\mathrm{T}$ . A new analyst wants to associate 00 to 09 and 75 to 99 with $\mathrm{H}$ and the remaining random numbers with $\mathrm{T}$ . Which of the following is true?

A) The new method will not work properly since it will create a bias for tails
B) The new method will work most of the times, but not in relatively smaller simulations
C) The new method is as good as the old method
D) The new method should only be used if we have an enormous number of runs to offset bias

Question 46

An analyst wants to simulate weekly demand during a work week for a product whose daily demand follows normal distribution with mean 300 and standard deviation 100. The following 5 random normal deviates were chosen to help simulate the demand for five working days: 1.2, -.7, -2.2, 1.4, and 1.8. The weekly demand will be

A) 420
B) 1650
C) 1500
D) 1710

Question 47

Daily demand for a product is described in the table below. Profit per unit sold is $\$ 10.00$ . Unmet demand costs are $\$ 5.00$ in goodwill. The demand is not carried forward. Excess stock is liquidated on a daily basis at a cost of $\$ 3.00$ per unit. That is, excess stock is not carried forward in stock. Orders have to be placed using a fixed policy of ordering $X$ number of units each day. Lead-time is 0 , but only one order can be placed each day. What should be the value of $\mathrm{X}$ in order to maximize profit? Try two "good" ordering policies (i.e. two order quantities), simulate the profit corresponding to the two policies for 10 days, and choose the best quantity.