Research on Ming Pressure Prediction of Working Face Based on Informer Neural Network
In order to effectively improve the problems of low accuracy and insufficient generalization ability of mine pressure prediction in the working face,a time-series prediction model of the mine pressure was established based on the Informer neural network,and the historical mine pressure data collected by hydraulic supports was taken as input to realize the prediction of the mine pressure for a period of time in the future.The established model was based on the mine pressure input sequence information extracted by the ProbSpare self-attention mechanism,which can capture the long-term dependence of the input sequence,and model the complex nonlinear relationships among the influencing factors,thereby improving Informer model prediction accuracy.The mine pressure data of the XV1307 working face of Chengzhuang Mine was used for model training and testing,and the obtained prediction results were compared with those of particle swarm optimization BP neural network(PSO-BP)and Long Short Term Memory network(LSTM).The results show that,for the prediction of mine pressure in the next 1-4 days,the root mean square error,mean absolute error and determination coefficient of the Informer neural network are all the optimal,and a good prediction effect has been achieved.
万方数据
维普
