基于区间直觉犹豫模糊集的高速列车系统关键部件辨识

Identification of critical components of high-speed train system based on interval-value intuitionistic hesitant fuzzy set

中文关键词  高速列车系统,网络模型,区间直觉犹豫模糊集,关键部件

英文关键词  High-speed  train system, Network  model, Interval-value  intuitionistic hesitant  fuzzy set, Critical  components

中文摘要

      关键部件作为系统可靠性分析的重要组成以及系统维修策略制定的重要基础, 其辨识方法与识别结果的准确性一直是研究热点与难点. 为有效识别关键部件保证高速列车系统安全可靠性运行, 本文提出一种新的基于区间直觉犹豫模糊集的关键部件辨识方法. 首先通过回顾现有复杂系统网络建模方法及其优缺点, 归纳总结部件的提取规则与部件间作用关系的定义和类型, 提出由局部到整体的高速列车系统全局拓扑网络模型构建方法. 其次鉴于现有针对同一高速列车同一节车辆同一部件的故障维修记录数据量少、随机性大等特点, 提出将节点属性值区间化, 构建节点属性的区间直觉犹豫模糊集, 其中节点属性包括由故障数据分析获得的节点可靠性属性值以及由所建全局拓扑网络得到的拓扑属性. 最后利用区间直觉犹豫模糊积分算子综合节点的多种属性辨识系统中的关键部件. 以转向架系统的拓扑结构以及历史故障数据为基础的实例分析表明, 所提出的辨识方法能够准确的识别出系统中的关键部件, 且与专家经验的分析结果基本一致.

英文摘要

      Critical component is an integral part to system reliability analysis and a mainstay in establishing system maintenance strategy. The identification methods of critical component and their accuracy of the results are always one of the most difficult and hottest problems. In order to identify critical components synthetically and effectively and guarantee safety of high-speed train system, a novel identification method of critical components based on interval-value intuitionistic hesitant fuzzy set is presented in this paper. First, the existing system network modeled methods and their advantages and disadvantages are summarized. Based on above research, the holistic topological network model is proposed from part to whole, according to physical structure and reliability properties, and then the extraction rules of nodes and the types and definition of edges are supposed. The interval of components'' attributes is considered and interval-value intuitionistic hesitant fuzzy set of nodes'' properties are constructed, in view of the characteristics of high speed train system, such as the small amount and randomness of failure data. To utilize interval-value intuitionistic hesitant fuzzy Choquet operator, multi-properties of nodes are aggregated and finally critical components can be identified. The results show that the holistic topological network model can be described the structure of the whole system, and the proposed identification method can give quantitatively the ranking of importance components for bogie system.