For the characteristics of high frequency and strong suddenness of dynamic disasters in deep coal mines, the traditional detection and evaluation techniques applied to shallow coal mine failed to accurately judge the risk degree of disasters. Therefore, it is of great significance to use advanced detection technologies and appropriate evaluation methods to improve the accuracy and efficiency of risk assessment in the process of coal mining. The present paper applies the rapid detection and multi-source dynamic detection technologies used in the field of mining with the purpose of improving the reliability of detection technologies for typical dynamic disaster. In this study, the data fusion technology was used to analyze data obtained from laboratory experiments, engineering survey, detection and historical data, so as to form the final dynamic and static indicators. Then, the new combined evaluation models with time series of coal and gas outburst as well as rock burst were established respectively to carry out the comprehensive risk evaluation using the least-squares method and the time-varying weight method. After the comprehensive analysis on the results of the above two evaluation models, the risk areas of the typical dynamic disasters were judged and classified. Finally, the evaluation models were coded to build an early-warning software platform that could achieve automatic evaluation and the actual 3D visualization of coal mining areas. The early-warning software platform was applied to risk assessment of dynamic disasters in Gengcun Coal Mine in Yima City, Henan Province, China. The results of the 6-month experiment showed that the risk assessment accuracy and reliability of the proposed evaluation models was 100% and 90% respectively, which indicates that the newly developed approach is reliable and can be recommended for applying in more coal mines to improve the process safety risk control.

针对深部煤矿动态灾害频率高、突发性强的特点，应用于浅部煤矿的传统检测评价技术无法准确判断灾害风险程度。因此，采用先进的检测技术和合适的评价方法，对提高煤矿开采过程中风险评估的准确性和效率具有重要意义。本文将快速检测和多源动态检测技术应用于采矿领域，旨在提高典型动态灾害检测技术的可靠性。本研究采用数据融合技术对实验室实验、工程勘察、检测、历史数据等数据进行分析，从而形成最终的动态和静态指标。然后，分别建立新的煤与瓦斯突出时间序列与岩爆组合评价模型，分别采用最小二乘法和时变权重法进行综合风险评价。综合分析以上两种评价模型的结果，对典型动态灾害的风险区域进行判断和分类。最后，对评价模型进行编码，构建了可实现煤矿区自动评价和实际3D可视化的预警软件平台。将该预警软件平台应用于河南省义马市耿村煤矿动态灾害风险评估。