基于矿山被动CT监测的矿震冲击危险预警基础研究

As rockburst frequently happen in China, the prediction should be firstly made to guide the prevention. A kind of high-resolution and wide-range monitoring and prediction method is the most important basis to effectively guide rockburst prevention. In view of the problem and shortage of present strong mine tremor and rockburst danger prediction method, this research is concentrated on the improvement of the accuracy, efficiency and resolution of passive velocity tomography method using mine tremor in coal mine. The relationship between stress and longitudinal wave velocity is also be improved in this research as a basis to create the grading warning system of rockburst danger level by passive velocity tomography method using microseismic monitoring system. Based on the evolutionary algorithm, mechanism of rockburst, inversion theory, microseismology and statistics, large sample statistical test is adopted to build the quantitative assessment model of passive velocity tomography and then the genetic algorithm is applied to study the influence of network layout on the calculation accuracy; the principal component analysis method is used to study the relationship among adaptive unequal spacing gridding method, computational efficiency, resolution and adjustment mechanism; the ultrasonic testing experiment is taken to study the influence of changing angles between longitudinal wave and stress, and the regression analysis technology is used to build the new model between stress and velocity, which is applied to construct the rockburst prediction index and indentify the relationship between prediction index and the danger level. This reasearch will further improve the prediction efficiency of seismological observation system and provide the technical support for coal mine safety production.

我国冲击地压灾害频发表明应以预警指导防治为主。一种高分辨率和大范围监测预警方法是有效指导防治工作的重要基础。针对目前矿震冲击危险预警技术存在的问题与不足，课题以提高矿山被动CT计算精度、效率和分辨率,完善应力与纵波波速关系模型，建立基于微震监测的矿山被动CT危险性分级预警体系为研究主线，基于微震学、冲击地压机理、统计学、进化算法和反演理论，采用大样本统计试验建立矿山被动CT计算精度的定量评价模型，进而利用遗传算法研究台网布设对提高计算精度的影响及调整机制；采用主成分分析法研究自适应不等间距网格划分与计算效率、分辨率的关系及划分机制；通过超声波测试实验手段研究纵波与轴向应力载荷夹角变化对波速的影响，并利用回归分析技术建立改进模型，从而用于构建冲击危险预警指标和划分指标与冲击危险程度对应关系。研究成果将进一步提升微震监测系统预警效能，为煤矿安全开采提供技术保障。

冲击地压是煤矿井下采掘空间周围煤岩体急剧破坏，并突然释放巨大能量的一种强烈动力现象。冲击地压灾害影响因素多，实现有效预警难度大，严重制约了煤矿安全高效生产。因此开发有效的探测预警技术，是煤矿安全生产的重大需求。以往采用钻屑法等冲击危险探测技术以“点探测”为主，存在范围小、效率低、劳动强度大等不足，研发大范围、高效、区域应力场探测方法是迫切需要解决的工程科学问题。实验研究了煤岩三轴加卸载全过程震动波波速的演化规律，发现震动波波速与应力大小具有显著的幂函数关系，依此建立了震动波波速与煤岩应力的耦合模型，准确描述了波速与应力的关系，建立了波速异常指标及评价准则，用于判别煤岩应力集中程度，为现场冲击危险的分区分级奠定了理论基础。创建了矿山被动CT监测精度评价模型，提出了基于主成份分析法的自适应不等间距网格法和主动CT实验中的雷管延时消除方法，形成了一套显著提高矿山CT反演监测精度和效率的技术方法体系。研究揭示了震动波动载扰动的时空分布规律，构建了评价区域动载扰动强度的指标D；基于载荷与波速耦合模型及动静载诱冲判别准则，建立了冲击危险动静载叠加分区分级的评价模型，确定了动静载叠加强度指标I和判别准则，I值越大，冲击危险性越高。解决了冲击危险大范围、高效、区域探测技术难题。该内容属矿业安全学科，通过本项目研究共发表标注基金号论文10篇，其中SCI收录论文4篇，EI论文1篇，核心期刊论文5篇。已获授权发明专利2项，申请专利2项。获软件著作权2项。培养硕士研究生2名。通过本项目研究，获批企业合作项目1项，依托研究成果获山东省煤炭工业科学技术厅颁发的科技进步二等奖。