基于Bayesian网络的坚硬顶板条件下煤与瓦斯突出预警控制机理研究

Over 80 percent of gas disasters were concerned with coal and gas outburst. The proportion of coal and gas outburst occurred in the hard roof coalmines was higher than others. It was very theoretically and practically significant to research on control mechanism of coal and gas outburst early warning system on hard roof condition. Many theories such as early warning theory, extension theory, et al. were used in the study. Theoretical analysis, laboratory experiment, computer simulation and field experiment were adopted. The relationship between gas emission or outburst and influencing factors, the quantitative relationship between outburst omen and outburst, operational mechanism and monitoring and control mechanism would be deeply researched. Comprehensive relationship between gas emission or outburst and primary influencing factors on hard roof condition will be ascertained. Discrimination method of outburst omen based on rough set-extensible pattern recognition will be put forward. Quantitative relationship between outburst omen and outburst will be built up. Dynamic evaluation model used to outburst risk analysis based on Bayesian network-extension theory will be established. Operational mode, operational mechanism and monitoring and control mechanism will be put forward. Countermeasures to prevent outburst will be improved further. Organizational model of outburst early warning system will be constructed. All research findings will be theoretical basis to execute coal and gas outburst early system in coalmines.

我国超过80%的瓦斯灾害和煤与瓦斯突出有关，坚硬顶板条件矿井发生的煤与瓦斯突出占有较高的比重，因此对坚硬顶板条件下煤与瓦斯突出预警系统控制机理进行研究具有重要的理论意义和现实意义。本项目运用预警理论和可拓理论等多学科交叉理论，采用理论分析、实验室试验、计算机仿真和现场试验，对坚硬顶板条件下瓦斯涌出（突出）与主要影响因素之间的关系、煤与瓦斯突出预兆信息和突出之间的量化关系以及预警系统的运行机制和监控机制进行研究。以期确定出坚硬顶板条件下瓦斯涌出（突出）与主要影响因素的综合关系；提出基于粗糙集-可拓模式识别的煤与瓦斯突出预兆信息甄别方法；建立突出预兆信息与突出之间的量化关系；构建基于Bayesian网络-可拓理论的突出风险评价动态模型；提出突出预警系统运转模式、运行机制和监控机制；进一步完善突出预控对策库；建立突出预警系统的组织模型。从而为煤炭企业成功实施煤与瓦斯突出预警系统提供理论支持。

随着煤矿开采深度的增加，瓦斯已经成为制约煤矿安全高效生产的首要因素。坚硬顶板条件下煤与瓦斯突出事故占有较高的比重，而煤与瓦斯突出灾害预警可以提前发出警报。因此针对当前煤与瓦斯突出预警研究存在的不足，对坚硬顶板条件下煤与瓦斯突出预警控制机理进行研究具有重要的理论意义和重大的现实意义。本项目主要对坚硬顶板条件下瓦斯涌出（突出）与影响因素之间的关系、突出预兆信息甄别新方法、预警系统的运行机制与监控机制等内容进行了研究。测定了实验区平禹四矿二1煤层的瓦斯基础参数。理论分析推导出了煤层中径向不稳定流和球向不稳定流的瓦斯压力场方程，推导出了瓦斯流量方程和煤层透气性系数λ的方程，它们都是关于时间t和距钻孔（或巷道）中心的距离r的函数。利用理论推导与现场试验结合的方法分析了煤层瓦斯含量与埋藏深度之间的定量关系，煤层埋藏深度在一定范围内，煤层瓦斯含量与煤层埋藏深度呈正比关系，但随着埋藏深度的增加，正比例系数逐渐减少，瓦斯含量趋于常数。研究得出瓦斯涌出初速度、瓦斯涌出量等参数具有相似的或相反的变化规律，它们是关于距离x或者时间t的S型曲线和倒S型曲线的组合。利用FLAC3D数值模拟等方法进行了采煤工作面瓦斯涌出（突出）影响因素的综合分析，结果发现渗透性是影响工作面煤壁瓦斯涌出的关键因素；瓦斯压力和煤层围岩透气性之间呈现一定的负相关性；周期来压与工作面瓦斯涌出量具有一定的相关性；当覆岩关键层周期性破断失稳后，瓦斯将快速大量涌向采场；采场的采动应力与煤层瓦斯压力之间存在一定的关联，并且它们都对煤体的孔裂隙结构有影响。构建了基于粗糙集-可拓模式识别的煤与瓦斯突出识别模型。构建了煤矿瓦斯灾害预警管理系统的组织模型，对瓦斯灾害预警管理系统的功能体系进行了重构；构建了基于Bayesian网络-可拓理论的风险分析模型；提出了瓦斯灾害预警管理系统的运转模式、运行机制、监控机制与管理制度体系。通过对煤与瓦斯突出预警控制机理的研究，可以为煤矿安全生产提供有力的理论支撑。