Download MTBF

Survey
yes no Was this document useful for you?
   Thank you for your participation!

* Your assessment is very important for improving the workof artificial intelligence, which forms the content of this project

Document related concepts
no text concepts found
Transcript
MAT 4830
Mathematical Modeling
05
Mean Time Between
Failures
http://myhome.spu.edu/lauw
Reading for Next Wednesday


MCM Contest Rules, Registration and
Instructions
http://www.comap.com/undergraduate/c
ontests/mcm/instructions.php
Preview

Use Monte Carlo Method to estimate the
life of products
Emachine…
Mean Time Between Failures
Suppose that we make a product composed of
n components where the ith component is
guaranteed by its manufacturer to possess a
time before failure that is normally distributed
about the mean i with standard deviation i.
ith component
Ti ~ N ( i ,  i )
Mean Time Between Failures
Assume that our product will fail when any one
of its components fail.
Questions: What guarantee can we offer our
customers for our product? What is our
product's mean time between failures (MTBF)?
Analytic Solution
Ti ~ N ( i ,  i )
T  min(Ti )
1i  n
MTBF  

 tf (t )dt

Analytic Solution
Fi (t )  Cumulative Distribution Function of Ti


n
 tf (t )dt  



 t j 1 1  Fj (t )  Fi (t )dt
i 1 
n
j i
Monte Carlo Solution

3 components
T1 ~ N (11, 1)
T2 ~ N (12, 2)
T3 ~ N (13, 3)
Monte Carlo Solution

3 components
T1 ~ N (11, 1)
T2 ~ N (12, 2)
T3 ~ N (13, 3)

Let us watch this product fail over and
over by simulation
Idea
i T1
T2
T3 T  min(T1 , T2 , T3 )
1 11.5 11.4 13.0
t1  11.4
2 10.6 12.0 10.2
t2  10.2
n 11.0 10.9 15.5
tn  10.9
Idea
i T1
T2
T3 T  min(T1 , T2 , T3 )
1 11.5 11.4 13.0
t1  11.4
2 10.6 12.0 10.2
t2  10.2
n 11.0 10.9 15.5
tn  10.9
Idea
i T1
T2
T3 T  min(T1 , T2 , T3 )
1 11.5 11.4 13.0
t1  11.4
2 10.6 12.0 10.2
t2  10.2
n 11.0 10.9 15.5
tn  10.9
Idea
i T1
T2
T3 T  min(T1 , T2 , T3 )
1 11.5 11.4 13.0
t1  11.4
2 10.6 12.0 10.2
t2  10.2
n 11.0 10.9 15.5
tn  10.9
1 n
1 n 2
   ti , m   ti ,  2  m   2
n i 1
n i 1
Programming Outline
Within a loop:
1. Generate random numbers T1 , T2 , T3
(with the corresponding distribution)
2. Compute ti  min(T1 , T2 , T3 )
n
n
i 1
i 1
3. Keep track of  ti and  ti2
1 n
1 n 2
2
2
Then compute:    ti , m   ti ,   m  
n i 1
n i 1
Sample Output
> MTBF(1000);
The mean is approximately 10.138361
The standard deviation is approximately 1.355540
Classwork

Individual (Each of you need to think
through the process)
Problem 1
The example above
Problem 1
Maple Commands:
HW Problem 1
A manufactured product consists of three
components. Suppose the time before
failure for the parts are given by
Ti ~ N (10, 1), i  1, 2,3
HW Problem 1
The first component is a backup for the
second; i.e., either both must fail or the
third component must fail before the
construct fails.
HW Problem 1


Write a program to estimate the mean
and standard deviation of the life of the
product.
Briefly explain the key underlying
methodology of your program.
Related documents