Home  Contact Us  Search  Site Map  Downloads  About Unipass

Sample Set Sensitivity in Monte Carlo Simulation

Once the sample set MPP was identified, the associated First-Order Sensitivity Measurements can be derived. It is noted that these sensitivities are only available for each limit state function. They can be viewed in the Sensitivity Plots.

Similarly to First-order Sensitivity for component problems, UNIPASS offers the following five types of sensitivities for Sample Set sensitivity for each limit state function:

1. SENSITIVITY OF LIMIT-STATE FUNCTION WITH RESPECT TO THE RANDOM VARIABLES AT THE MPP IN BOTH THE STANDARD NORMAL SPACE AND THE ORIGINAL SPACE

2. SENSITIVITY OF FAILURE PROBABILITY AND THE RELIABILITY INDEX, WITH RESPECT TO RANDOM VARIABLES

3. SENSITIVITY OF FAILURE PROBABILITY AND THE RELIABILITY INDEX, WITH RESPECT TO MEANS AND STANDARD DEVIATIONS OF RANDOM VARIABLES

4. SENSITIVITY OF FAILURE PROBABILITY AND THE RELIABILITY INDEX, WITH RESPECT TO DISTRIBUTION PARAMETERS, WHICH MAY BE CORRELATION COEFFICIENTS, BOUNDS OR OTHER PARAMETERS USED TO DEFINE THE DISTRIBUTION

5. DIMENSIONLESS (UNIT FREE) SENSITIVITY OF LIMIT STATE FUNCTION, FAILURE PROBABILITY, AND THE RELIABILITY INDEX, WITH RESPECT TO MEAN STANDARD DEVIATIONS OF RANDOM VARIABLES

The accuracy of sample set first-order sensitivity depends on how close of sample set MPP is to the true MPP. If the number of sample points in the Monte Carlo simulation is big enough, the sample set sensitivity should be very accurate.

Last Updated 02/08/10

Send mail to info@predictionprobe.com with questions or comments about this web site.
Copyright © 2001-2004 PredictionProbe, Inc.