Intelligent classification strategy for flammable-gases base on BP neural network

In the chemical and industrial field, the technology of detecting flammable or poisonous gases had always been important for product-safety control. Since the most of the work environment is complex and the gas always is complicated mixture, the precision of detection is rather low. IF the advanced pattern recognition and classification technology can be used in this field, we can effectively improve the detection accuracy. This paper firstly introduce the basics knowledge of BP neural network, then use MATLAB neural network toolbox to build network models and used experimental data to train the net. The simulation results show the network´s performance is good.