Risk Assessment of Subway Fire Based on Genetic Neural Network

Abstract

In this paper, the training of BP neural network and optimization of genetic algorithm are based on MATLAB. The metro fire risk assessment model was established by MATLAB toolbox, training results of BP neural networks and optimized BP neural networks were obtained. By comparing the results of BP neural networks and genetic algorithms optimized neural network, it summed up genetic algorithms optimized neural network can get the training target more quickly. It is 48.98% faster than the BP neural network. It shows the genetic algorithm can improve the training speed effective.