摘要
随着矿井开采深度不断增加,矿井突水灾害威胁日益增大,如何精准预测矿井突水风险是值得深入研究的课题。针对传统方法未对监测数据中隐藏的有效信息进行深度挖掘的问题,以潘二煤矿11023工作面为工程背景,提出一种基于LSTM的矿井涌水量预测模型和矿井水位预测模型,采用微震监测数据与水文观测数据对矿井涌水量和水位进行预测。结果表明,预测结果与真实测量值偏差较小,精度较高,具有较好的应用价值,可为矿井防治水工作提供技术指导。
With the increasing mining depth,the threat of mine water inrush disaster is increasing day by day.How to accurately predict the risk of mine water inrush is a topic worthy of in-depth research.Aiming at the problem that the traditional methods do not dig deeply the effective information hidden in the monitoring data,taking the 11023 working face of Pan'er Coal Mine as the engineering background,a mine water inrush prediction model and mine water level prediction model based on LSTM are proposed,which use microseismic monitoring data and hydrological observation data to predict the mine water inrush and water level.The results show that the prediction results have small deviation from the real measured value,and have high accuracy,which has good application value and can provide technical guidance for the mine water prevention and control work.
作者
徐一帆
韩云春
黄刚
童政
高翔
Xu Yifan;Han Yunchun;Huang Gang;Tong Zheng;Gao Xiang(State Key Laboratory for Safe Mining of Deep Coal and Environment Protection,Huainan Mining(Group)Co.,Ltd.,Huainan,China;Ping'an Coal Mining Engineering Technology Research Institute Co.,Ltd.,Huainan,China)
出处
《科学技术创新》
2024年第10期183-186,共4页
Scientific and Technological Innovation
关键词
煤矿开采
矿井突水风险预测
长短时记忆网络
coal mining
prediction of mine water inrush risk
long short-term memory network
