Neural-network-based reliability analysis: a comparative study

A study on the applicability of different kinds of neural networks for the probabilistic analysis of structures, when the sources of randomness can be modeled as random variables, is summarized. The networks are employed as n umerical devices for substituting the finite element code needed by Monte C arlo simulation. The comparison comprehends two network types (multi-layer perceptrons and radial basis functions classifiers), cost functions (sum of square errors and cross-entropy), optimization algorithms (back-propagatio n, Gauss-Newton, Newton-Raphson), sampling methods for generating the train ing population (using uniform and actual distributions of the variables) an d purposes of neural network use (as functional approximators and data clas sifiers). The comparative study is performed over four examples, correspond ing to different types of the limit state function and structural behaviors . The analysis indicates some recommended ways of employing neural networks in this field. (C) 2001 Elsevier Science B.V. All rights reserved.