【作者】 丁宝成；

【导师】 沈玉志；

【作者基本信息】 辽宁工程技术大学 ， 管理科学与工程， 2010， 博士

【摘要】 煤炭工业做为我国的基础产业,在现在及未来的国民经济发展中都起着举足轻重的作用。从目前我国能源现状、经济发展水平和世界能源格局来看,相当长的时期内以煤为主的能源消费结构难以改变。煤炭的重要地位给煤炭工业的发展提出了更高的要求。然而近年来煤矿事故频发,说明我国煤矿的安全形势并未从根本上有所好转,安全生产状况是我国煤矿企业生产过程中的头等大事,已成为全社会观注的热点。因此,加强煤矿安全预警研究,对于加强煤矿安全管理的科学性和可靠性,预防及减少煤矿事故的发生,减少事故造成的损失,对于贯彻“安全第一,预防为主”的生产方针,促进煤炭工业健康有序、安全高效、可持续发展,适应煤炭工业发展对煤矿安全提出的更高要求,具有重要的现实意义。本文以井工开采煤矿企业生产系统为研究对象,综合运用安全科学、预警管理理论、人工智能理论、计算机技术和网络技术等相关学科的理论和方法,融合现代先进管理理念,对煤矿安全预警问题进行研究。在分析国内外煤矿安全预警研究现状及存在问题的基础上,针对我国煤矿企业的安全特点,从影响煤矿安全的特征分析入手,构建完善的煤矿安全预警管理体系及预警模型,旨在对煤矿企业生产中存在的事故及隐患进行及时有效的预警,达到预防和减少煤矿事故发生的目的。根据安全预警管理理论,构建了煤矿企业安全预警管理的完整框架体系,分别从组织保障、预警核心内容、预警职能、运行模式等方面进行了探讨,通过对煤矿企业安全预警理论的研究,探索了企业在安全状态中的识错防错机理和非安全状态下的纠错治错机理,为煤矿企业的安全管理工作提供了理论基础。对我国煤矿安全生产现状进行了统计分析,探索了事故发生的一般规律性。根据事故致因理论,建立了以“事故类型–部门结构–事故成因”为三维的煤矿事故分析立体模型,并从“人-机-环-管”方面对煤矿事故发生的本质原因进行了探讨,为事故原因的定性、定量分析,从理论上提供了新的思路。探讨了煤矿安全预警系统的技术方法。通过对影响煤矿企业系统安全的各种因素及因素间的相互关系的分析,将影响煤矿安全的因素划分为自然地质因素、人员因素、设备因素、环境因素、管理因素五大类,建立了科学合理的预警指标体系。以模糊数学、神经网络理论为基础,构建了基于模糊AHP的煤矿安全预警模型及基于补偿模糊神经网络的煤矿安全预警模型,对预警指标进行综合评价,根据预警指标值及预警界限,进行单指标预警及系统综合预警。每种评价方法都有各自的适用范围、优点和缺点。单一的评价方法在适用范围和评价精度、评价的客观性等方面都有各自的局限性。根据两模型的特点,构建了基于模糊AHP及补偿模糊神经网络的煤矿安全预警组合模型,在应用过程中通过对模糊AHP模型和网络模型两种单一预警模型进行合理的有机组合,综合利用两模型方法包含的有用信息,改变了因选择单一模型不合理而带来的预警误差,有效地提高了预警精度,使得预警工作更具科学性,进一步提高了预警结果的可靠性。结合实际生产矿井,将以上的模型、方法等在矿井生产过程中进行了实际应用,并达到了预期的效果。更多还原

【Abstract】 The coal industry plays a important role in the present and future national economic development as a basic industry.According to the current status of national energy structure and economic development level and pattern of world energy pattern,The status of coal-dominated energy consumption is difficult to change for a long term.It is put forward higher requirements for the coal industry’s development as its important position. However,accidents had been happend frequently in the coal mine in recent years, it shows that the security situation doesn’t get better,safety in production is a major event and hot topics of whole society. It has an important practical significance to strengthen the study of coal mine safety early warning for strengthen reliability and scientific of the management of coal mine, To prevent and reduce coal mine accidents, reduce the losses caused by the accident,Implement the "safety first, prevention crucial " policy of safety production, to promote healthy, orderly, safe and effective, and sustainable development, adapted to the coal industry’s development。The production system of coal mine as the research object, it is researched on the safty early warning system of coal mine baseed on Safety science、security warning management theory、Artificial intelligence theory、Computer and network technology and the theory of related,Discussed the causes of the coal mine accidents from the perspective of safety management, established a scientific and reasonable indicators and model of early warning, established a Decision Support System to monitoring、diagnosising、controlling and correction, in order to pre-warning effectively and in time during the production of coal mining enterprises, to prevent and reduce coal mine accidents.It build a complete framework of early warning system depended on the security warning theory,Discussed the organizational security、the core content、functions、operating modes and other aspects in detail, to provide a theoretical basis for the management of the coal enterprises.It Explore the general rule of the accident through statistical analysis of the production situation in coal mine. According to the Accident-causing Theory, build the“Accident Type - Sector structure - cause of the incident”as the three-dimensional analysis model of coal mine accidents,research on the cause of the accident from the aspects of man、machine、Environment and management factor, to offer a new way which analysis the cause of the accident qualitatively and quantitatively in theory.It studied on the technical methods of early warning systems of coal mine safety. According to exploring the various factors and the interrelationship of them that will affect the safety, These factors are divided into geological、human、equipment、environmental、and management factors,established a scientific and reasonable indicators of early warning system,according to the fuzzy mathematical theory and Artificial intelligence theory, The model of fuzzy comprehensive evaluation based on AHP and the model of Compensatory Fuzzy Neural Network which to make a comprehensive evaluation of early warning indicators and to safety prewarning are established. According to the values and limits of early warning indicators to make single and comprehensive early warning.Each evaluation method have their advantages and disadvantages in application,and single evaluation method have its limitations when it is used in aspect of applicability、Accuracy、Objectivity. According to the characteristics of two models, Combination Model of mine safty early-warning based on fuzzy and AHP and Compensatory Fuzzy Neural Network is established, the above two model is combined reasonably and organically, useing useful information included in two model, early-warning errors that caused by single model selected are improvement,the scientificity and reliability of the early warning is further improve.The above models and methods are used in the mines and the desired results are achieved.更多还原