Reliability Analysis of Structures with Interval Parameters Based on a Radial Basis Function (RBF) Neural Network

In this study, we introduce a radial basis function (RBF) neural network for the reliability analysis of structures with interval parameters. The RBF neural network is employed to approximate the limit state function of the structure. The training samples are generated using an optimal Latin hypercube design. The RBF neural network is trained with the samples, and afterwards linked with a Monte Carlo simulation to evaluate the reliability of a structure with interval parameters. The purpose of applying the RBF neural network to interval reliability analysis is to reduce the number of runs using finite element analysis, especially when the structure is complicated. A numerical example is used to demonstrate the accuracy and efficiency of the proposed method. The proposed method can be applied to engineering applications.