煤层突出危险性综合预测及验证  

Comprehensive prediction and verification of coal seam outburst risk

作　　者：张尔辉 朱权洁[1] 刘衍 高林生[1] 张震 谷雷 ZHANG Erhui;ZHU Quanjie;LIU Yan;GAO Linsheng;ZHANG Zhen;GU Lei(Safety Engineering College,North China Institute of Science and Technology,Yanjiao,065201,China;Shiyan Company of Hubei Tobacco Company,Shiyan,442000,China)

机构地区：[1]华北科技学院安全工程学院,北京东燕郊065201 [2]湖北省烟草公司十堰市公司,湖北十堰442000

出　　处：《华北科技学院学报》2020年第4期19-26,42,共9页Journal of North China Institute of Science and Technology

基　　金：国家自然科学基金资助项目(51174016);贵州省科技计划资助项目(黔科合重大专项字[2018]3003-1,[2018]3003-2);煤与瓦斯突出动态监测与智能预警技术研究(黔科合平台人才[2019]5675)。

摘　　要：为了克服煤与瓦斯突出危险性评价指标多、危险程度划分标准不同等复杂因素对评价结果的影响,提高突出危险性预测的准确性,本文提出基于BP神经网络和灰色关联分析法的煤与瓦斯突出危险性区域预测和区域验证。以河北某矿2#煤层采掘工作面为例,基于BP神经网络对该煤层采掘工作面煤与瓦斯突出危险性进行区域预测;采掘推进过程中,利用灰色关联分析法对该煤层突出危险性进行区域验证。研究表明:BP神经网络预测了2#煤层采掘工作面5个实测区域突出危险程度为一般或较弱;灰色关联分析法确定了钻屑解吸指标Δh 2作为2#煤层采掘工作面突出预测敏感指标的临界值,且均小于突出临界值;综合揭煤情况确定2#煤层采掘工作面为无突出危险工作面。应用结果证明了BP神经网络和灰色关联分析法在区域预测和区域验证的合理性和可行性。In order to overcome the influence of complicated factors,such as too many evaluation indexes and different classification standards for coal and gas outburst risk,and improve the accuracy of outburst risk prediction,regional prediction and verification of coal and gas outburst risk based on BP neural network and gray correlation analysis are proposed.Taking the mining face of No.2 coal seam in Hebei province as an example,based on BP neural network,the coal and gas outburst risk of this coal seam mining face is predicted in the region.In the process of mining advancement,the coal seam outburst risk is verified regionally by grey correlation analysis.The results show that the BP neural network predicts that the outburst risk degree of the 5 measured areas of No.2 coal seam mining face is moderate or weak.The gray correlation analysis method determined the drilling cuttings desorption indexΔh 2 as the critical value of the outburst prediction sensitivity indexes of No.2 coal seam mining face and they are all less than the outburst critical value.The mining face of No.2 coal seam is determined to be the face without outburst danger.The application results prove the rationality and feasibility of BP neural network and gray correlation analysis in regional prediction and regional verification.

关 键 词：BP神经网络 灰色关联分析 煤与瓦斯突出 区域预测 区域验证 

分 类 号：TD713[矿业工程—矿井通风与安全]