Moment and sensitivity analysis for dynamic characteristic of stochastic structure

The uncertainty of structural parameters will lead to the uncertainty of structural dynamic characteristics (such as the natural frequency, mode shape). If using the determinacy analysis for the designing may make the product in the dangerous state, then the dynamic characteristic analysis of the stochastic structure is very necessary. The statistic moments of dynamic characteristics of stochastic structure include the statistical properties of the dynamic characteristics. Through these statistical properties, the statistical information of reliability analysis and the distribution regularity of dynamic characteristic can be further obtained. And the sensitivity analysis of statistic moment can distinguish the degree of importance of random parameters with respect to the stochastic structure inherent properties, and then provide an effective way for improving the structural design model. In this work, MCMC based stratified sampling method is presented to estimate the statistics moment. And by collecting the information of random samples, the sensitivity of statistics moment is analyzed without increasing computing calculation. And the results of examples illustrate the advantages of the presented method.