Abstract: In order to facilitate managers to formulate the targeted maintenance strategies for equipment and promote the transformation of rail transit enterprises to refined operations, five characteristics such as type of equipment failure, equipment failure impact, relative maintenance cost, failure reparability and parameter extraction are selected to build a hierarchical analysis model of urban rail equipment characteristics to calculate the equipment characteristic weights, and use hierarchical clustering method to cluster equipment according to equipment characteristics and average distance as the characteristic similarity measurement method, then according to the applicability of each maintenance method and the results of equipment clustering, a logical decision method is used to select the maintenance method for the equipment. Finally, take the Wuxi metro equipment as an example to verify the feasibility of the program.

Key words: urban rail transit equipment, maintenance method, hierarchical clustering, logical decision

摘要： 为便于管理者为设备制定针对性的维修策略，促进轨道交通企业向精细化管理转型，选择设备故障类型、设备故障影响度、相对维修成本、故障可维修度、参数提取度5个特性准则，构建城轨设备特性层次分析模型计算设备特性权重，并采用层次聚类方法根据设备特性，以平均距离为特性相似度度量方法将设备聚类，再根据各维修方式的适用性及设备聚类结果，利用逻辑决断方法为设备选择维修方式。最后，以无锡地铁设备维修为例，验证该方法的可行性。

关键词: 城市轨道交通设备, 维修方式, 层次聚类, 逻辑决断