《粗糙集在矿井局部通风设备故障诊断中的应用_》由会员分享，可在线阅读，更多相关《粗糙集在矿井局部通风设备故障诊断中的应用_（2页珍藏版）》请在金锄头文库上搜索。
1、 论文题目：粗糙集在矿井局部通风设备故障诊断中的应用 专 业：计算机应用技术 硕 士 生：杨 阳 （签名） 指导教师：付 燕 （签名） 摘 要 近年来，由于煤矿的大量开采，瓦斯爆炸事故频繁发生，给人们的生命财产安全造成了极大危害。矿井局部通风设备故障是导致掘进工作面较易发生瓦斯爆炸事故的主要原因之一。矿井局部通风设备系统包含多个组成部分，产生故障的原因和种类较多。工作人员无法及时对所发生的故障进行准确地分析判断，延长了停风时间，为瓦斯爆炸事故的发生埋下了隐患。 粗糙集理论作为人工智能领域中一个新的学术热点，它无需任何先验信息，能有效地分析和处理不精确、不一致、不完备性数据，通过发现数据间的关系
2、，揭示潜在的规律，从而提取有用信息，简化信息的处理。因此，本文应用粗糙集理论提取出矿井局部通风设备的故障诊断规则，从而快速准确地判断故障发生原因，减少停风时间，降低瓦斯爆炸事故发生的概率。 本文首先介绍了粗糙集理论的基础知识，分析了几种基于粗糙集理论的约简算法。其次，针对基于区分矩阵的属性频度算法不一定能找到正确属性约简的问题，结合决策表是否相容的因素，对基于区分矩阵的属性频度算法提出了改进，避免计算的盲目性，使约简结果更加准确。通过实例分析，对改进算法的有效性进行了验证。最后，针对收集的矿井局部通风设备故障信息，建立故障决策信息系统，并采用本文的改进算法对其进行属性约简和值约简，确定了关键故
3、障征兆，提取出故障诊断规则。实验证明，使用粗糙集理论获取的故障诊断规则应用在矿井局部通风设备故障诊断中，可以提高故障诊断的效率。 关 键 词： 粗糙集理论；区分矩阵；属性约简；矿井局部通风设备；故障诊断 研究类型： 应用研究 Subject : Research about Diagnosis for Mine Local Ventilation Facilities Based on Rough Sets Specialty : Computer Application Technology Name : Yang Yang (Signature) Instructor : Fu Yan (
4、Signature) ABSTRACT Due to heavy coal exploitation in recent years, many serious accidents caused by gas explosion in coal mines frequently happened, which jeopardized to peoples lives and wealth. Mine local ventilation facilities failure is one of the main reason which leads to the accident of gas
5、explosions in heading face. The mine local ventilation facilities incorporates a number of components, involving many types and reasons of the failure. The results that mine staff cannot timely carry out the problem, extended the wind stopped time and increased the probability of the gas explosion.
6、Rough set theory, as a new academic hot spot in artifitial intelligence, without any prior information, can efficiently analyze and dispose of imprecise, inconsistent, incomplete data. It reveals the hidden laws by the discovery of nexus between the data, in order to pick up the useful information a
7、nd simplify the information disposal. Based on the rough set theory, this paper carries out research on the rules of the failure diagnosis in partial mine ventilation facilities. It instructs how to estimate the cause of the failure quickly and exactly, shorten the time of stopping the wind, and red
8、uce the probability of the gas explosion. Firstly, in this paper, rough sets are elaborated. Attribute reduction algorithms based on rough set theory are analyzed. Secondly, aim at the problem that the frequency property of the algorithm based on discernibility matrix may not be able to find the rig
9、ht attribute reduction, and the factor that the decision table whether or not compatible, the frequency property of the algorithm based on Discernibility matrix proposed are be improved, to avoid the blindness of the calculation, so that more accurately reducted results are obtained. The effectiveness of the algorithms in this paper is clearly demonstrated by the experiment results. Finally, according to the fault of the whole facilities, the fault decision-making information system for mine local ventilation facilities is established based on rough sets. Adopt the improved algorithm in
