唐口煤矿坚硬顶板冲击危险多源信息预测预警研究(2).pdf

分类号分类号 TD324 密密 级级 公公 开开 U D C 单位代码单位代码 10424 学学 位位 论论 文文 唐口煤矿坚硬顶板冲击危险多源唐口煤矿坚硬顶板冲击危险多源 信息预测预警研究信息预测预警研究 严超超严超超 申请学位级别申请学位级别硕士学位硕士学位 专业专业名名称称矿业工程矿业工程 指导教师姓名指导教师姓名顾顾 士士 坦坦 职职 称称副副 教教 授授 副副指导教师姓名指导教师姓名张张 晓晓 职职 称称副研究员副研究员 山山 东东 科科 技技 大大 学学 二二 O 一一八八年年六六月月 万方数据 国家重点研发计划子课题（2016YFC0801403-3） 山东省自然科学基金面上项目（ZR2018MEE009） 国家自然科学基金项目（51374140、51204102） 论文题目论文题目 唐口煤矿坚硬顶板冲击危险多源唐口煤矿坚硬顶板冲击危险多源 信息预测预警研究信息预测预警研究 作者姓名作者姓名 严严 超超 超超 入学时间入学时间 2015 年年 9 月月 专业名称专业名称 矿业工程矿业工程 研究方向研究方向矿山压力与岩层控制矿山压力与岩层控制 指导教师指导教师 顾顾 士士 坦坦 职职 称称 副副 教教 授授 副副指导教师指导教师 张张 晓晓 职职 称称 副副 研研 究究 员员 论文提交日期论文提交日期2018 年年 4 月月 论文答辩日期论文答辩日期2018 年年 5 月月 授予学位日期授予学位日期2018 年年 6 月月 万方数据 STUDY ON PREDICTION AND EARLY-WARNING OF MULTI-SOURCE INATION OF ROCK BURST IN HARD ROOF OF TANG KOU COAL MINE A Dissertation ted in fulfillment of the requirements of the degree of MASTER OF ENGINEERING from Shandong University of Science and Technology by Yan Chaochao Supervisor Associate Professor Gu Shitan Vice Supervisor Associate Professor Zhang Xiao College of mining and safety engineering June 2018 万方数据 万方数据 万方数据 万方数据 山东科技大学硕士学位论文 摘要 I 摘摘 要要 煤炭是我国经济建设的主体能源，随着多年来开采强度和开采规模不断提 升，我国埋藏较浅的煤炭资源逐年减少，深部煤炭资源的开采是未来发展的趋 势，然而，深部矿井开采在灾害防治方面具有严重挑战，矿井冲击地压动力灾 害就是其中之一。 本文以唐口煤矿坚硬顶板工作面诱发冲击及防治为研究背景， 运用理论分析、数值模拟与现场监测数据分析等研究方法，分析了坚硬顶板诱 发冲击危险机理，研究了坚硬顶板弯曲-破断-运动规律，建立了冲击地压监测 预测预警模型，为唐口煤矿坚硬顶板工作面安全回采提供理论依据和指导，主 要研究如下 （1）通过能量理论分析，得到唐口煤矿坚硬顶板工作面发生冲击地压所需 能量临界密度值；分析坚硬顶板冲击失稳力学模型，计算出 5305 坚硬顶板工作 面煤壁前方应力集中影响区及最大弯矩位置。 （2）通过数值模拟建立坚硬顶板工作面仿真模型，研究坚硬顶板在开采过 程中顶板的弯曲-破断-运动规律、工作面煤壁前方超前支承压力分布情况及变 化规律、坚硬顶板岩层中弹性应变能积聚释放演化规律，为工作面安全回采提 供理论指导。 （3）通过对唐口煤矿 5305 工作面冲击地压监测数据的现场收集、数据挖 掘分析，并基于小波神经网络理论，建立了冲击地压小波神经网络预测模型， 基于聚类 K-means 算法原理，对微震能量、钻屑量、应力值进行聚类分类预测 分析，通过建立的冲击地压监测数据预测和数据挖掘分析模型，有效的提高了 矿井冲击地压监测数据合理充分利用。 （4）基于模糊神经网络模型，通过对冲击地压监测数据多源信息特征提取 分析，建立模糊神经网络冲击地压多源信息综合监测预警模型，并将预警模型 在 5305 工作面进行了实践应用。经过现场检验分析，模糊神经网络预警模型预 警效果良好，预警准确率较高。 关键词关键词坚硬顶板；冲击地压；预测模型；聚类算法；监测预警 万方数据 山东科技大学硕士学位论文 摘要 II Abstract Coal is the main energy source for Chinas economic construction. With the continuous increase in mining intensity and scale over the years, Chinas shallow coal resources have been reduced year by year, and the exploitation of deep coal resources is the trend of future development. However, deep mining has serious challenges in disaster prevention and control. One of these is the rock burst geodynamic disaster. This dissertation took the induced impact and prevention of hard roof working face in Tangkou Coal Mine as the research background. The theoretical analysis, numerical simulation and on-site monitoring data analysis s have used to analyze the mechanism of inducing rock burst hazard of hard roof and study the bending, breaking and movement of hard roof. This dissertation established a prediction and early warning model for rockburst monitoring, which provided a theoretical basis and guiding significance for the safe mining of the hard roof working face of mines. The main research is as follows 1 According to energy theory analysis, the critical energy density required for impact rock pressure in the hard roof working face of Tangkou Coal Mine has obtained; The mechanical model of impact instability of hard roof has analyzed; the maximum stress concentration area and bending moment position in front of the coal wall of the 5305 hard roof working face has calculated. 2 This dissertation has establish a simulation model of hard roof working face by numerical simulation, study bending-breaking-movement law of hard roof during mining process and distribution of advance bearing pressure in front of working face coal wall, elastic strain energy ation accumulation and release evolution rules provide guidance for the safe mining of working face. 3 Through on-the-spot collection and analysis of the rock burst monitoring data of 5305 working face in Tangkou Coal Mine, and based on wavelet neural network theory, a prediction model of rock burst has established. based on the principle of clustering algorithm, and the microseism Energy, drill cuttings, and stress values were used for cluster prediction analysis. Through the establishment of impact rock pressure monitoring data prediction and mining analysis models, the mine rock pressure monitoring data was effectively and reasonably used. 4 Based on the theory of fuzzy neural network, the multi-source ination feature extraction and analysis of the rock burst monitoring data has used to establish a fuzzy neural network comprehensive monitoring and early warning model for multi-source ination of rock burst, and the early warning model has applied in 5305 working face. After on-site inspection and analysis, the fuzzy neural network early warning model works well and the accuracy of early warning was high. Key words hard roof; rock burst; prediction model; clustering algorithm; monitoring and warning 万方数据 山东科技大学硕士学位论文 目录 III 目目 录录 摘摘 要要 ......................................................................................................... I 目目 录录 ...................................................................................................... III 1 绪论绪论 ....................................................................................................... 1 1.1 研究背景及意义 ................................................................................................... 1 1.2 研究矿井概况 ....................................................................................................... 2 1.3 国内外研究现状 ................................................................................................... 4 1.4 主要研究内容及技术路线 ................................................................................... 8 2 坚硬顶板诱发冲击危险理论研究坚硬顶板诱发冲击危险理论研究 .................................................... 10 2.1 引言 ..................................................................................................................... 10 2.2 冲击地压演化过程能量转化规律 ..................................................................... 10 2.3 坚硬顶板冲击失稳力学模型分析 ..................................................................... 17 2.4 本章小结 ............................................................................................................. 24 3 坚硬顶板环境冲击危险演化规律数值模拟分析坚硬顶板环境冲击危险演化规律数值模拟分析 ............................ 26 3.1 引言 ..................................................................................................................... 26 3.2 数值模型的建立 ................................................................................................. 26 3.3 模拟结果分析 ..................................................................................................... 28 3.4 本章总结 ............................................................................................................. 35 4 坚硬顶板冲击地压多源信息预测研究坚硬顶板冲击地压多源信息预测研究 ............................................ 36 4.1 引言 ..................................................................................................................... 36 4.2 冲击地压监测预警方法 ..................................................................................... 36 4.3 小波神经网络预测 ............................................................................................. 39 4.4 聚类算法分类预测 ............................................................................................. 45 4.5 本章小结 ............................................................................................................. 50 5 坚硬顶板冲击地压多源信息综合监测预警及应用坚硬顶板冲击地压多源信息综合监测预警及应用 ........................ 52 万方数据 山东科技大学硕士学位论文 目录 IV 5.1 引言 ..................................................................................................................... 52 5.2 模糊神经网络理论基础 ..................................................................................... 52 5.3 冲击地压模糊神经网络预警模型 ..................................................................... 57 5.4 冲击地压危险多源信息预警模型应用 ............................................................. 62 5.5 本章小结 ............................................................................................................. 66 6 主要结论与展望主要结论与展望 ................................................................................. 67 6.1 主要结论 ............................................................................................................. 67 6.2 创新点 ................................................................................................................. 68 6.3 展望 ..................................................................................................................... 68 参考文献参考文献 ................................................................................................. 69 致致 谢谢 ....................................................................................................... 73 攻读硕士期间攻读硕士期间主要研究成果主要研究成果................................................................. 74 万方数据 山东科技大学硕士学位论文 目录 V Contents Abstract ..................................................................................................................... ⅠⅠ Contents .................................................................................................................... ⅢⅢ 1 Introduction ............................................................................................................. 1 1.1 Research Background and Significance ................................................................................ 1 1.2 Research mine overview ........................................................................................................ 2 1.3 Research status at home and abroad ...................................................................................... 4 1.4 Main Research Contents and Technical Routes ..................................................................... 8 2 Theoretical Study on Induced Impact Danger of Hard Roof ........................... 10 2.1 Introduction ......................................................................................................................... 10 2.2 The law of energy transation in the process of impact rock pressure evolution .......... 10 2.3 Analysis of mechanical model of impact instability of hard roof ........................................ 17 2.4 Brif Summary ...................................................................................................................... 24 3 Numerical simulation of the evolution law of rockburst pressure in hard roof ............................................................................................................................ 26 3.1 Introduction ......................................................................................................................... 26 3.2 The establishment of numerical models .............................................................................. 26 3.3 Analysis of simulation results .............................................................................................. 28 3.4 Brif Summary ...................................................................................................................... 35 4 Study on Prediction of Multi-source Ination for Hard Roof Impact Rock Pressure .............................................................................................. 36 4.1 Introduction ......................................................................................................................... 36 4.2 Impact pressure monitoring and early warning ...................................................... 36 4.3 Wavelet neural network prediction ...................................................................................... 39 4.4 Clustering algorithm prediction ........................................................................................... 45 4.5 Brif Summary ...................................................................................................................... 50 万方数据 山东科技大学硕士学位论文 目录 VI 5 Comprehensive Monitoring and Early Warning and Application of Multi-source Ination for Hard Roof Impacting ............................................ 52 5.1 Introduction ......................................................................................................................... 52 5.2 Fuzzy neural network theory basis ...................................................................................... 52 5.3 Fuzzy neural network early warning model for rockburst ................................................... 57 5.4 Application of multi-source ination early warning model for rockburst ...................... 62 5.5 Brif Summary ...................................................................................................................... 66 6 Main Conclusions and Prospects ......................................................................... 67 6.1 Main Conclusions ................................................................................................................ 67 6.2 Innovation Points ................................................................................................................. 68 6.3 Prospects .............................................................................................................................. 68 References ................................................................................................................. 69 Acknowledgements .................................................................................................. 73 Main Achievements During the Master ................................................................. 74 万方数据 山东科技大学硕士学位论文 1 绪论 1 1 绪论绪论 1.1 研究背景研究背景及及意义意义 煤炭是我国经济建设的主体能源，随着多年来开采强度和开采规模不断提 升，我国埋藏较浅的煤炭资源逐年减少，深部煤炭资源是未来开采发展的趋势， 然而，深部矿井的开采在灾害防治方面中具有严重挑战，矿井冲击地压动力灾 害就是其中之一，在我国东部矿区尤其明显。冲击地压动力灾害是我国深部矿 井资源开采面临的重大安全技术难题之一，严重影响矿井安全高效开采。目前， 人们对冲击地压发生机理和防治技术研究已取得一定进展，但由于其发生的前 兆信息、危险源辨识及控制因素极其复杂性，在诱冲机理、监测预警和防治措 施等方面仍需要进行更加深入的研究和探索实践才能取得突破性进展。 坚硬顶板冲击地压作为矿井典型冲击地压类型之一，其危险源及主控因素 主要是开采工作面上方赋存着一层或数层厚度较大的坚硬岩层，坚硬岩层一旦 大范围剪切破断滑移失稳，极易诱发冲击危险。在我国超过 50的矿区存在坚 硬顶板工作面开采问题[1]，并且各个矿井开采地质构造条件不同，开采技术水 平不同，导致冲击危险诱冲机理复杂，对冲击危险防治具有极大挑战。 冲击地压发生的过程具有瞬时性、危害程度极大，一旦诱发冲击，极易 造成工作面设备严重损坏和人员伤亡的重大安全生产事故。坚硬顶板的存在 易诱发矿井动力灾害的发生，对现场工作人员人身安全造成极大的威胁。因 此，分析坚硬顶板工作面上覆岩层运动规律，研究坚硬顶板诱冲机理具有重 要意义。 随着人们对冲击地压灾害重视程度的提高，矿井将冲击地压监测技术应用 到防冲日常工作中，使冲击地压预测预警水平得到进一步提升，但由于冲击地 压监测预警技术单一，监测数据得不到充分挖掘，监测数据指标特征提取不够， 预警智能化程度低的特点日益突出。在大数据、智能化时代背景下，依据认知 神经网络、数据挖掘技术研究冲击地压综合预警智能判别技术，建立冲击地压 综合预警云平台和优化改良模型算法成为研究冲击地压动力灾害防治的热点， 对提高煤矿智能化安全生产及灾害防治具有重要的意义。