围绕拟解决的重大科学问题“煤矿典型动力灾害风险判识及监控预警技术研究”，以煤与瓦斯突出、冲击地压等典型煤矿动力灾害为研究对象，鉴于煤矿典型动力灾害诱发机理不清、风险判识不明、监控预警技术不足等现状。

通过研究：① 冲击地压孕育机理与风险判识及监测预警;② 煤与瓦斯突出灾变机理及监测预警;③ 煤矿典型动力灾害信号采集传输和智能化分析;④ 煤矿典型动力灾害监控预警系统平台等4项内容，开发了大尺度、真三维、全封闭自动开挖煤与瓦斯突出物理模拟实验装置，研制了包括光纤光栅微震传感、三轴应力传感、分布式多点激光甲烷检测等在内的动力灾害前兆信息新型感知与多网融合传输传感装置，建立了井下传感器数据的多元海量动态信息的聚合理论与方法，构建了基于漂移特征的潜在煤矿典型动力灾害预测方法与多粒度知识挖掘方法，形成了基于大数据分析和数据挖掘的煤矿动力灾害风险判识和监控预警模型。

通过现场实践，表明能实现全面采集人机环参数，且采集传感器具有故障自诊断、响应时间短、标校周期长等优点。监控预警系统稳定运行无故障率达到了99%，抗干扰等级不低于3级，系统监控预警准确率大于90%，实现了煤矿典型动力灾害隐患在线监测、智能判识、实时预警。

Focusing on a major scientific problem to be solved,i. e. ,“research on the risk identification and monitoring and early warning technology of typical coal mine dynamic disasters”,this paper investigates the typical coal mine dynamic disasters such as coal and gas outburst and rock burst. In view of the current situation of unclear mechanism of typical coal mine dynamic disasters,unclear risk identification and warning technology of monitoring and early warning,etc. The research covers ① the development mechanism of rock burst and risk identification and monitoring and early warning,② the coal and gas outburst disaster mechanism and monitoring and early warning,③ the coal mine typical dynamic disaster signal acquisition and transmission and intelligent analysis,and ④ the coal mine typical dy- namic disaster monitoring and early warning system platform. A large-scale,true three-dimensional,fully closed and automatic experimental device for the physical simulation of coal and gas outburst is developed. A new type of sensing and fusion transmission sensor device for dynamic disaster precursor information,including fiber Bragg grating micro- seismic sensor,tri-axial stress sensor,and distributed multi-point laser methane detection,is developed. The aggregation theory and method of multi-dimensional and massive dynamic information of underground sensor data are established. The prediction method of typical dynamic disasters and the multi granularity knowledge mining method based on drift characteristics are constructed. A model of judgment,recognition and warning of major coal mine disasters based on big data analysis and data mining is established. Through the field applications,it is shown that the acquistion sensor can realize the comprehensive acquisition of man-machine ring parameters,and has the advantages of self- diagnosis on fault,short response time and long calibration cycle. The non-fault rate of the monitoring and warning sys- tem has reached 99% in a stable operation,the anti-interference level is no less than level 3,and the system's monitoring and warning accuracy is more than 90% . The system has realized the online monitoring,intelligent judgment and real-time warning of typical power hazards in coal mines.