Research on Random Modelling of Computer Multi-sample Redundancy in Nonlinear Mathematics Stochastic Finite Element

Abstract

The stochastic problems of engineering structures are generally simulated by stochastic finite element method. Based on the non-linear mathematical random method, the paper proposes the concept of multi-sample redundancy based on Bayesian network and the concept of redundancy compression algorithm. A Monte Carlo stochastic finite element method based on Bayesian network redundancy compression algorithm is established. The modelling process of Bayesian network redundancy compression algorithm is explained in detail through two examples. The Bayesian network redundancy compression algorithm model provides a new strategy for multi-sample system reliability analysis, which can conveniently implement system reliability calculation, fault diagnosis, importance analysis and other applications. If the constructed Bayesian network redundancy compression algorithm satisfies that all non-root nodes have two parent nodes, the reliability solving process of the multi-sample system only needs O(Nm3) computational complexity.