Problem 1: Service System Capacity

Queue
Customers Server Served
Customers

Problem: Can a server taking an average of x time units meet the demand?
Solution with an Analytical Model:
Must specify the conditions on customer arrival, service completions. Take
measurements or adopt assumptions (clearly specified).
1. Pr{customer arrives in interval (t, t + t]} =  t + o(t) independent of t
2. Pr{service completes in interval (t, t + t]} =  t + o(t) independent of t
3. Customer arrivals and service completions are mutually independent events
4. Population of customers is large and service is in order of arrival
5. The limit of the ratio o(t) / t as t  0 is 0
 Analytical Model of Server Capacity
Let pn(t) = Pr{n customers in the system at time t}
Then
pn(t + t) = pn-1(t) [t + o(t)] + pn(t) [1 - t + o(t)] [1 - t + o(t)]
+ pn+1(t) [t + o(t)] n1
P0(t + t) = p0(t) [1 - t + o(t)] [1 - t + o(t)] + p1(t) [t + o(t)]
Producing the steady-state equations:
pn+1 – ( + ) pn +  pn-1 = 0 (n  1) and p1 -  p0 = 0 (n = 0)
Solving for pn in terms of p0 gives pn = (/)n p0
Using the fact that  pn = 1 and letting  = /, we obtain
pn = n(1 - ) where  < 1
 Problem 1

 Given a high-resolution computer image of a map of an irregularly

shaped lake with several islands, determine the water surface
area. Assume that the x-y coordinates of every point on the map
can be measured at an acceptable level of resolution.

Suggest alternative solution approaches!

 Monte Carlo Simulation
Y

0 X

Step 1:
Enclose the area of interest in the smallest rectangle of
known dimensions X and Y. Set j = 1, S = 0, and choose a
large value for N where:
j = trial number
S = number of trials resulting in a hit on the water surface
area
N = total number of trials
 Monte Carlo Simulation
Y

RN y

0 RN x X

Step 2:
Generate a uniformly distributed random number, RNx
over the length of X.
Step 3:
Generate another uniformly distributed random number, RNy
over the length of Y.
 Monte Carlo Simulation
Y

RN y

0 RN x X

Step 4:
If the point of intersection, (RNx , RNy), falls on the water
surface area, add 1 to S.
Step 5:
Add 1 to j. If j > N, go to Step 6; otherwise, go to Step 2.
 Monte Carlo Simulation

RN y

0 RN x X

Step 6:
A S
The estimate of the water surface area, A, is given by: =
XY N

NOTE: As N  , A  true value of the area

 Problem 2
 Use the fundamental theory and logic of the Monte Carlo
Simulation technique to estimate the area under the Sine curve
over 0 and .

Sin (x)

0 
 Monte Carlo Simulation
y

Sin (x)

0 

Step 1:
Enclose the area of interest in the smallest rectangle of
known dimensions  and 1. Set j = 1, S = 0, and choose a
large value for N where:
j = trial number
S = number of trials with a hit on the area under the Sin (x)
curve
N = total number of trials
 Monte Carlo Simulation

RNy
Sin (x)

0 RNx 

Step 2:
Generate a uniformly distributed random number, RNx
over the length of .
Step 3:
Generate another uniformly distributed random number,
RNy over the length of 1.
 Monte Carlo Simulation

RNy
Sin (x)

0 RNx 

Step 4:
If the point of intersection, (RNx , RNy), falls on or below the
Sin (x) curve (i.e., if RNy ≤ Sin (RNx)), add 1 to S.
Step 5:
Add 1 to j. If j > N, go to Step 6; otherwise, go to Step 2.
 Monte Carlo Simulation

RNy
Sin (x)

0 RNx 

Step 6:
A S
The estimate of the area, A, is given by: =
XY N

NOTE: As N  , A  2 (true value of the area)

 Problem 3
 Use the fundamental theory and logic of the Monte Carlo
Simulation technique to solve the following optimization
problem:

Maximize Z = ( eX1 + X2 ) 2 + 3 ( 1 – X3 ) 2
Subject to:
0 ≤ X1 ≤ 1
0 ≤ X2 ≤ 2
2 ≤ X3 ≤ 3
 Monte Carlo Simulation
 Use the fundamental theory and logic of the Monte Carlo Simulation
technique to solve the following optimization problem:

Maximize Z = ( eX1 + X2 ) 2 + 3 ( 1 – X3 ) 2
Subject to:
0 ≤ X1 ≤ 1
0 ≤ X2 ≤ 2
2 ≤ X3 ≤ 3

 Step 1:
Set j = 1 and choose a large value for N where:
j = trial number
N = total number of trials
 Monte Carlo Simulation
 Use the fundamental theory and logic of the Monte Carlo Simulation
technique to solve the following optimization problem:

Maximize Z = ( eX1 + X2 ) 2 + 3 ( 1 – X3 ) 2
Subject to:
0 ≤ X1 ≤ 1
0 ≤ X2 ≤ 2
2 ≤ X3 ≤ 3

 Step 2:
Generate a proper random number, RN1 , over 0 and 1.
 Step 3:
Generate another proper random number, RN2 , over 0 and 2.
 Step 4:
Generate another proper random number, RN3 , over 2 and 3.
 Monte Carlo Simulation
 Use the fundamental theory and logic of the Monte Carlo Simulation
technique to solve the following optimization problem:

Maximize Z = ( eX1 + X2 ) 2 + 3 ( 1 – X3 ) 2
Subject to:
0 ≤ X1 ≤ 1
0 ≤ X2 ≤ 2
2 ≤ X3 ≤ 3

 Step 5:
Substitute RN1 for X1 , RN2 for X2 , and RN3 for X3 in the
objective function. Store its value in Z(j) and record the
corresponding values for X1 , X2 , and X3.
 Step 6:
Add 1 to j. If j > N, go to Step 7; otherwise, go to Step 2.
 Monte Carlo Simulation
 Use the fundamental theory and logic of the Monte Carlo Simulation
technique to solve the following optimization problem:

Maximize Z = ( eX1 + X2 ) 2 + 3 ( 1 – X3 ) 2
Subject to:
0 ≤ X1 ≤ 1
0 ≤ X2 ≤ 2
2 ≤ X3 ≤ 3

 Step 7:
The approximate solution of the problem is determined by
the values of X1 ( = RN1 ), X2 ( = RN2 ), and X3 ( = RN3 ), which
correspond to the maximum value of { Z(j), j = 1, 2, 3, ..., N }.

 NOTE: As N  , X1  1, X2  2, and X3  3.