基于优化LSTM网络的巷道工作面矿压预测  

Prediction of Rock Pressure in Roadway Working Face Based on Optimized LSTM Network

作　　者：赵保兵[1] ZHAO Baobing(Shanxi Polytechnic College,Taiyuan 030006,China)

机构地区：[1]山西职业技术学院,太原030006

出　　处：《煤炭技术》2024年第2期54-56,共3页Coal Technology

摘　　要：冲击地压事故作为一种动力效应明显、破坏力强的矿井灾害,其预防方法一直是煤矿安全生产的难题。通过将LSTM神经网络技术与矿压规律进行结合,研究一种基于优化LSTM网络的巷道工作面矿压预测模型,通过对矿压历史数据进行标准化处理后进行LSTM网络训练的方法,实现工作面矿压的预测。现场试验结果显示:该模型对于定点矿压的预测准确率能够达到85%以上,并且能够对不同时段、不同测点的矿压数据进行预测,以表征巷道工作面未来某一时段内的矿压变化趋势,实现预防矿压事故的目的。As a mine disaster with obvious dynamic effects and strong destructive power,the prevention method of rockburst accidents has always been a difficult problem for coal mine safety production.Combines LSTM neural network technology with mine pressure law to study a mine pressure prediction model for roadway working faces based on optimized LSTM network.By standardizing the historical data of mine pressure and training the LSTM network,the prediction of mine pressure in the working face is achieved.The on-site test results show that the accuracy of the model in predicting fixed point rock pressure can reach over 85%,and it can predict rock pressure data at different time periods and measurement points to characterize the trend of rock pressure changes in the future at a certain time period of the tunnel working face,and achieve the goal of preventing rock pressure accidents.

关 键 词：LSTM网络 冲击地压 神经网络 煤矿安全 

分 类 号：TD323[矿业工程—矿井建设]