Evaluation of chloride diffusion in concrete using PSO-BP and BP neural network

Abstract

Chloride diffusion is the major causes of deterioration of concrete structures in engineering. Because chloride diffusion experiments are time consuming, it is desired to develop a model to predict the chloride diffusion in concrete. In this paper, the optimizing of particle swarm algorithm (PSO) on BP neural network is adopted to predict the chloride penetration in concrete. For purpose of building these models, training and testing pattern is gathered from the technical literature. PSO-BP neural network can improve BP disadvantage. PSO-BP neural network is better precision than BP neural network through the results of PSO-BP, BP and experiments. The research results demonstrate that PSO-BP neural network is an effective tool in the prediction of chloride diffusion.