首页 | 本学科首页   官方微博 | 高级检索  
     

基于贝叶斯神经网络的冲击地压预测与影响因素权重分析
引用本文:乔建永,王志强,罗健侨,武超,李敬凯,林陆.基于贝叶斯神经网络的冲击地压预测与影响因素权重分析[J].中国矿业,2022,31(5).
作者姓名:乔建永  王志强  罗健侨  武超  李敬凯  林陆
作者单位:中国矿业大学北京,中国矿业大学北京,中国矿业大学北京,中国矿业大学北京,中国矿业大学北京,中国矿业大学北京
基金项目:国家科技部973项目“深部开采中的动力灾害机理与防治基础研究”(2010CB226800);国家自然科学基金面上项目(51774289,52074291);
摘    要:针对华丰煤矿4#煤层冲击事故频发的问题,采用理论分析及现场实验相结合的方法,对于复杂开采技术类因素和多种典型性地质条件耦合作用下的冲击地压启动机理、能量来源及防治措施展开研究。首先分析4#煤层1411工作面冲击启动能量来源,通过建立工作面前方煤体及围岩结构承受动、静载荷的力学模型,得到冲击启动区煤体内部任意一点的应力解析解,其次基于弹塑性理论,通过主应力平面偏转变换得出的最大、中间和最小主应力求得冲击启动区任意处单元煤体储存的弹性应变能的解析解,并根据现场实际工程参数,分析各种因素对煤体内部应力和能量分布规律的影响。结合力学模型解析解、数值模拟实验和现场记录参数的结果对工作面前方冲击地压的启动、传递和显现进行预测和分析。通过对冲击地压影响因素权重的分析发现了通过水力压裂降低顶板强度或水射流切割顶板降低老顶悬臂梁长度等对工作面防冲治理的优先级应高于保护层开采卸压等手段。

关 键 词:冲击地压  神经网络  冲击启动与传递  弹性应变能  冲击防治
收稿时间:2022/3/5 0:00:00
修稿时间:2022/5/10 0:00:00

Forecast of Rock Burst and Weight Analysis of Influencing Factors Based on Bayesian Neural Network
QIAO Jianyong,WANG Zhiqiang,LUO Jianqiao,WU Chao,LI Jingkai,LIN Lu.Forecast of Rock Burst and Weight Analysis of Influencing Factors Based on Bayesian Neural Network[J].China Mining Magazine,2022,31(5).
Authors:QIAO Jianyong  WANG Zhiqiang  LUO Jianqiao  WU Chao  LI Jingkai  LIN Lu
Affiliation:China University of Mining and Technology Beijing,Beijing,China University of Mining and Technology Beijing,Beijing,China University of Mining and Technology Beijing,Beijing,China University of Mining and Technology Beijing,Beijing,China University of Mining and Technology Beijing,Beijing,China University of Mining and Technology Beijing,Beijing
Abstract:In view of the frequent occurrence of rockburst accidents in Huafeng coal mine 4# coal seam, the starting mechanism, energy source and prevention measures of rockburst under the coupling of complex mining technical factors and various typical geological conditions are studied by using the method of theoretical analysis and field experiment. Firstly, the energy source of impact start-up in 1411 working face of 4# coal seam is analyzed. By establishing the mechanical model of dynamic and static load borne by the coal body and surrounding rock structure in front of the working face, the analytical solution of stress at any point in the coal body in the impact start-up area is obtained. Secondly, based on the elastic-plastic theory, the maximum The intermediate and minimum principal stress shall strive to obtain the analytical solution of the elastic strain energy stored in the unit coal at any place in the impact start-up area, and analyze the influence of various factors on the internal stress and energy distribution law of the coal according to the actual engineering parameters on site. Combined with the analytical solution of mechanical model, numerical simulation experiment and the results of field recorded parameters, the start, transmission and appearance of rockburst in front of the working face are predicted and analyzed. Through the analysis of the weight of influencing factors of rock burst, it is found that the priority of anti scour treatment of working face should be higher than that of protective layer mining and pressure relief by reducing the strength of roof by hydraulic fracturing or cutting the roof by water jet and reducing the length of cantilever beam of main roof.
Keywords:rockburst  neural Networks  rockburst occrurence and transfer  elastic strain energy  rock burst prevention
点击此处可从《中国矿业》浏览原始摘要信息
点击此处可从《中国矿业》下载全文
设为首页 | 免责声明 | 关于勤云 | 加入收藏

Copyright©北京勤云科技发展有限公司  京ICP备09084417号