【作者】 王向前；

【导师】 孟祥瑞；

【作者基本信息】 安徽理工大学 ， 采矿工程， 2014， 博士

【摘要】 煤矿回采巷道支护参数的合理选择,决定着煤矿回采工作面的能否正常运转,也是保障煤矿安全生产的重要因素之一。由于煤矿回采巷道支护参数影响因素具有主观和客观上的不确定性和复杂性,有些参数值的获取也较为困难,采用理论计算、数值模拟、相似材料模拟等方法进行支护参数设计,结果并不尽理想。近年来随着计算机技术的快速发展,很多高校和科研院所把专家系统、决策支持系统引入到巷道支护设计应用中,取得了一定的成果,一定程度上提高了设计效率和效果,但由于传统的规则推理机制存在知识获取的“瓶颈”、规则库维护和更新困难、灵活性和适应性差等缺陷,制约着这些系统一直没能很好的推广应用。为此,本论文研究基于案例推理的煤矿回采巷道支护决策系统,克服传统的围岩分类和规则获取等困难,为煤矿回采巷道支护参数选择的智能决策探索新的方法,同时也拓展了案例推理技术理论在复杂案例推理中的应用。论文从探讨案例推理的基本原理入手,论述了案例的表示方法、案例检索策略、案例相似度的计算方法、案例属性权重的确定方法、案例的修正技术以及案例的学习和维护,并在此基础上结合煤矿回采巷道支护参数选择的特点,构建了基于案例推理的煤矿回采巷道支护决策系统模型。在收集大量回采巷道支护案例及其围岩参数的基础上,采用统计学分析方法,分析了煤矿回采巷道围岩的物理力学参数分布规律及各参数之间的相关关系。结果显示回采巷道围岩参数总体上表现出如下规律：(1)围岩的泊松比、密度和内摩擦角呈现较好的标准正态分布形态；(2)围岩的抗压强度与抗拉强度、内聚力和密度具有明显的正相关性,而与弹性模量,泊松比,内摩擦角的相关性较弱；(3)泥岩和粉砂岩的弹性模量随着埋藏深度的逐渐增加呈现出明显的增大趋势,而泊松比则呈现减小的趋势。并在此基础上针对回采巷道围岩参数缺失的具体情况,提出了基于岩石描述属性文本相似度和围岩参数相关性为基础的围岩参数取值算法,构建了煤矿回采巷道围岩参数数据库系统。最后采用面向对象编程语言和SQL Server2000数据库,设计开发了基于案例推理的煤矿回采巷道支护决策系统,实现了回采巷道支护参数智能选择、矿压监测、分析与预测及围岩参数取值等功能,并通过系统应用,在案例库不充分完备的情况下,取得了较好的效果。图表参更多还原

【Abstract】 Rational selection of supporting parameters of coal mining roadway is one of important factors influencing the normal performing of coal mining workface, also is one of the important safeguards of safety in coal mine production. As influencing factors of choosing coal mining roadway supporting parameters are uncertain and complex, and some parameter values are more difficult to obtain, it is not satisfactory and effective to choose the support parameters by common methods such as theoretical calculation, numerical simulation, and similar material simulation. In recent years, with the rapid development of computer technology, many researchers of universities and institutes continuously try to apply expert system and decision support system to roadway support design and have acquired many achievements, which improve efficiency and effect. But with some shortcomings in traditional rule reasoning mechanism, such as bottleneck of knowledge acquisition, difficult in updating and maintenance of rule base, poor flexibility and adaptability and so on, these achievements have not been put into wide application.Therefore, this paper studies the coal mine roadway supporting decision-making system based on case-based reasoning, to overcome such difficulties as in traditional surrounding rock classification and rule acquisition, and to offer a new way to establish coal mining roadway support intelligent decision system. The study also expands the application fields of case-based reasoning technology.From the basic principle of case-based reasoning, this paper explores case representation, case retrieval, calculation method of case similarity, how to determinate the attribute weights of case, case modification technology, case learning and maintenance, based upon those explorations, combining the characteristics of coal mining roadway support parameter selection, the coal roadway supporting decision-making system based on case-based reasoning model has been established. On the basis of collecting numerous mining roadway support cases with parameters of their surrounding rock, employing the statistical analysis method, this paper analyzes the physical and distributional law of mechanical parameters of coal roadway surrounding rock and the relationship between the parameters. Results show general laws of the mining roadway surrounding rock parameters are as follows:(1) the poisson’s ratio, density of the surrounding rock and the internal friction angle show a better standard normal distribution;(2) the compressive strength of surrounding rock has obvious positive correlations to tensile strength cohesion and density, but has weaker correlations to elastic modulus, poisson’s ratio, internal friction angle;(3) elastic modulus of mudstone and siltstone increase with the gradual increase in buried depth, while the poisson’s ratio shows the tendency of decreasing with the increase in buried depth. And according to the specific condition of lack of roadway surrounding rock parameters, the algorithm of parameter selection of surrounding rock is proposed based on similarity of rock feature description text and correlation of rock parameters, moreover, surrounding rock parameters database system for coal mining roadway is constructed.Finally, using object-oriented programming language and SQL Server2000database, coal mining roadway supporting decision system based on CBR has been developed, which can function properly in mining roadway support parameters selection, data analysis of mine ground pressure monitoring and so on; Further, despite the inadequate cases, the mining roadway support parameters selection engineering tested in the decision system still get the satisfactory results.更多还原