基于案例推理的煤矿瓦斯应急决策研究与应用

发布时间：2018-06-21 02:42

本文选题：煤矿瓦斯 + 案例推理　；参考：《中国矿业大学》2014年硕士论文

【摘要】：煤矿瓦斯事故给我国煤炭行业造成了严重的损失，制约了煤矿企业的可持续发展。对导致煤矿瓦斯危险的因素进行辨识和分析以及构建有效的煤矿瓦斯应急决策案例推理应急决策系统是当前煤矿安全研究的重要课题之一。基于案例推理通过检索案例库中过去同类的相似问题从而获得当前问题的解决方案，对于系统的构建具有指导意义。案例库的建立和案例检索是案例推理中最关键的两个问题，它们直接影响到案例推理的精度与速度。综上所述，建立一个好的案例库以及实现对具有相关属性的案例进行推理具有重要意义。针对上述问题，本文的具体研究如下：首先，对煤矿瓦斯事故危险源和事故发生时的应急处理流程进行了分析；着重研究了案例推理中案例库的索引和组织方法，提出了案例库的多级索引结构，以突发公共事件中的事故灾难为案例库的一级属性，依照案例库的多级索引结构，分级索引到煤矿瓦斯事故案例的案例库；提出了案例库的两级组织结构，，利用聚类分析思想，运用K-means聚类算法对案例库进行聚类，找到典型案例，将案例库划分为多个子案例库，针对聚类后子案例库间有交叉案例，使得检索结果不精确的问题，提出了一种改进K-means案例聚类算法，构造了代表案例库，把案例库划分为多个无交叉案例的子案例库。然后，讨论了煤矿瓦斯事故应急处理案例的内容，用框架的方法对煤矿瓦斯事故应急进行了案例表示，运用三角模糊数对案例的模糊数属性进行数值化；详细研究了案例检索的过程，提出一种模糊层次熵引擎的方法求案例特征属性权重，运用K近邻算法进行案例之间相似度的计算，检索出与目标案例最相似的k个案例。最后，给出了煤矿瓦斯爆炸事故实例，运用本文给出的权重计算方法和案例检索算法对该实例进行了详细的计算和分析，可以看出该算法是有效的，该算法更能找到相似的案例，远离那些不相似的案例，使检索出的案例更符合要求；构建了煤矿瓦斯爆炸事故应急处理案例推理系统，详细介绍了系统的工作流程、模块及功能，给出了相应的设计界面。
[Abstract]:The coal mine gas accident has caused serious loss to the coal industry of our country and restricted the sustainable development of the coal mine enterprise. The identification and analysis of the factors leading to coal mine gas hazard and the construction of an effective coal mine gas emergency decision case reasoning emergency decision system is one of the most important topics in the research of coal mine safety. Case-based reasoning (CBR) provides a solution to the current problem by retrieving the similar problems in the case base, which is of guiding significance to the system construction. The establishment of case base and case retrieval are the two most important problems in CBR, which directly affect the accuracy and speed of CBR. To sum up, it is important to establish a good case base and to realize case reasoning with relevant attributes. In view of the above problems, the specific research of this paper is as follows: firstly, the analysis of the coal mine gas accident hazard source and the emergency treatment flow when the accident occurs, especially the case base index and organization method in case-based reasoning, The multilevel index structure of the case base is put forward. The first class attribute of the case base is the accident disaster in the sudden public event. According to the multilevel index structure of the case base, the index is graded to the case base of the coal mine gas accident case. The two-level organization structure of case base is put forward. By using the idea of clustering analysis, K-means clustering algorithm is used to cluster the case base, find out the typical cases, divide the case base into several sub-case databases, and aim at the cross-cases among the sub-case bases after clustering. In this paper, an improved K-means case clustering algorithm is proposed to construct the representative case base and divide the case base into several sub-case databases without cross cases. Then, the content of coal mine gas accident emergency cases is discussed, the case representation of coal mine gas accident emergency is carried out with the method of frame, and the fuzzy number attribute of the case is numerically analyzed by using triangular fuzzy number. The process of case retrieval is studied in detail, and a fuzzy hierarchical entropy engine is proposed to calculate the weight of feature attributes of cases. The K-nearest neighbor algorithm is used to calculate the similarity between cases, and the most similar case to the target case is retrieved. Finally, an example of coal mine gas explosion accident is given, which is calculated and analyzed in detail by using the weight calculation method and case retrieval algorithm given in this paper, and it can be seen that the algorithm is effective. The algorithm can find similar cases, far away from those dissimilar cases, so that the retrieval of the case more in line with the requirements; build a coal mine gas explosion emergency handling case reasoning system, detailed introduction of the system work flow, Module and function, the corresponding design interface is given.
【学位授予单位】：中国矿业大学
【学位级别】：硕士
【学位授予年份】：2014
【分类号】：TD712

【参考文献】