Study On Safety Assessment Of Fire And Explosion Of Industrial Park Based On Genetic Neural Networks

The processes of production and business of chemical enterprises are important to nation and social for normal functioning. Their safety status not only related to people’s life and property, but also on the harmony and stability of society. The safety issues are more important to chemical industrial park, which as a gathering place for the chemical enterprises. Since the chemical enterprises are concentrated in the chemical industrial park, once an accident of fire and explosion occurred, it would have caused heavy casualties and property losses. For this reason, the safety assessment of chemical industrial park with fire and explosion risk has aroused widespread attention all over the world.This study mainly introduced the several factors of a fire and explosion accident in chemical industry park. Set up the assessment indicators system of chemical industrial park with fire and explosion risk. Then introduced and analyzed both advantages and disadvantages of genetic algorithm and BP neural network. Introduced gray cluster analysis and used this analysis to 12 samples.Firstly combined the advantages of genetic algorithm and BP neural network, and found a way to avoid the inadequate of these two algorithms. Used MATLAB for the programming. Then used gray cluster analysis and the assessment indicators system to cluster the 12 samples. Obtained the genetic neural network training samples and ran MATLAB to train a network. Finally assessed a real chemical industrial park, and got an accurate evaluation. Because the program is based on GA, BPNN and MATLAB, this method has some universalness, it offers other evaluations a new opinion and method.