The Research of back propagation neural network based on genetic algorithm in the gas concentration prediction

The prediction of gas concentration is an extremely complicated nonlinear dynamic system which cannot fully use precise mathematical language to describe. Only using BP neural network algorithm is easy to converge to the local minimal point in the gas concentration prediction, so this paper presents the idea of GA-BP of neural network, which set up the GA - BP neural network model by optimizing the weights and threshold of BP neural network and its application in coal gas concentration prediction. This method combining genetic algorithm and neural network, and use genetic algorithm to optimize neural network to the initial value, so that the BP network can convergence to the optimal solution fast and can achieve higher precision in a shorter period, and the rate of convergence, accuracy and stability is superior to the BP network model, Thus verified the rationality and effectiveness of the GA - BP neural network.