Structural Reliability Analysis Based on Subset Simulation Method and Universal Generating Function
The traditional structural reliability analysis method is difficult to obtain satisfactory accuracy at low cost when dealing with the problems of multidimensional non-normal distribution,small failure probability and high nonlinear function.In order to over-come the shortcomings of existing methods,the UGF-SS method for structural reliability analysis was proposed by combining the subset simulation method(SS)with the universal generating function(UGF).The intermediate failure probabilities were obtained by using the universal generating function according to the idea of subset simulation,thus the structural failure probability was obtained.The random variables were discretized in low density to obtain the structural universal generating function,the intermediate events were divided and the failure probabilities were calculated.In subsequent simulations,a non-uniform adaptive subdivision was performed for the intermedi-ate events to obtain the intermediate failure probabilities for each layer.In the last layer of simulation,the intermediate event samples were expanded using Latin hypercube sampling to calculate the failure probabilities.After the simulation,each intermediate failure prob-ability was multiplied to obtain the structural failure probability.The reasonableness and feasibility of the proposed method are verified through the analysis of examples,and the proposed method retains the high efficiency of the subset simulation method while exerting the universality of the universal generating function to arbitrary random variables and functional functions,the accuracy of the results is en-sured.
NETL
NSTL
万方数据
