Prediction of Gas Emission based on Principal Factor Analysis and Improved BP Neural Network

In order to improve the prediction efficiency and accuracy of the gas emission in stope working face,based on reverse BP neural network,principal factor analysis was used to reduce the dimension of variable;combined genetic algorithm (GA) and additional momentum method,the initial weights and thresholds of BP neural network were optimized by genetic algorithm,the GA-BP neural network prediction model based on the principal factor analysis was established,and momentum term was introduced in the process of weight reverse update.The monitoring data of gas emission in Qianjiaying Coal Mine of Kailuan Mining Group was selected as tag data and input data,and the simulation and analysis were carried out for different network models.The results show that the improved GA-BP neural network model converged in 603 time steps,the average relative error is about 0.58%,the prediction accuracy and efficiency are better than other neural network models,and the accurate prediction of gas emission can be realized more effectively.