Scribd
Upload
35 views9 pages

Simulation Run Statistics Explained

Simulation run statistics involve collecting data to measure performance and behavior in simulations, particularly to analyze outcomes and make decisions. It addresses issues such as autocorrelation and initial bias that can affect the accuracy of results. As sample sizes increase, biases diminish, but achieving steady state values may still take significant sample lengths.

Uploaded by

playbooksforme
Copyright
© All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PPTX, PDF, TXT or read online on Scribd
35 views9 pages

Simulation Run Statistics Explained

Simulation run statistics involve collecting data to measure performance and behavior in simulations, particularly to analyze outcomes and make decisions. It addresses issues such as autocorrelation and initial bias that can affect the accuracy of results. As sample sizes increase, biases diminish, but achieving steady state values may still take significant sample lengths.

Uploaded by

playbooksforme
Copyright
© All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PPTX, PDF, TXT or read online on Scribd

SIMULATION RUN

STATISTICS
Presented By:

Aayush Koirala Aayush Khadka


(021-304) (021-303)
01
Introduction
What is simulation run statistics ?
1. Data collected to measure performance and
behavior in a simulation
2. Method to handle problems that arise in measuring
statistics from simulation runs
3. Its purpose is to analyze outcomes and make
decisions.
4. In most of the simulation study, the assumptions of
stationary and mutually independent observations
do not apply so correlation is necessary to analyze
such scenario
Example
Consider a system with Kendall’s notation M/M/1/FIFO and
the objective is to measure the mean waiting time.
In simulation run approach, the mean waiting time is
estimated by

(n): Sample Mean (average waiting time for n samples).


​: Individual waiting times.
Continued..
● Such series of data in which one value affect other
values is said to be autocorrelated.
● Waiting time is dependent
● Positive term is added for this system but in other
system it can be negative
Problems
• The distribution may not be stationary.

• A simulation run is started with the system in some initial


idle state.

• In this case, the early arrivals will obtain service quickly


deviating from normal distribution.

• Hence, the sample means of the early arrivals is known as


initial bias.

• As the sample size increases and the length of run is long,


the effect of bias dies and the normal distribution is again
established.
Description
• Sample mean depends upon the sample length for
M/M/I system

• Server utilization=0.9

• Steady state mean=10

• Mean value biased below steady state mean

• As sample size increase bias diminishes but even


sample = 2000 mean only reached 95% of steady
state value.
Thank You !
Any further
questions ?

You might also like