摘要
地铁作为城市公共客运的重要载体,其系统设备在运营过程中难免发生一些故障。因此,应用数据挖掘技术对已有地铁故障数据进行关联规则挖掘,分析其影响,对故障预警与风险危害评估具有重大意义。针对地铁故障数据种类多样、影响程度难以界定等问题,建立考虑故障关联的改进Apriori算法,与经典的FP-Growth算法进行对比,对地铁故障关联规则进行研究,优化该算法的基本思想和流程。选取某地铁2020年设备故障数据为例,对其进行详细地分析,基于Python语言实现建模仿真,输出得到车载ATP故障、信号设备故障等多类故障之间的关联规则结果,为地铁故障影响程度分析、故障诊断、故障预警、风险危害等级划分等提供重要的参考依据。
As an important carrier of urban public transportation,metro system equipment inevitably has some failures during its operation.It is of great significance to apply data mining technology to mine some association rules from existing metro failure data and analyze their influence,so as to evaluate earl yfailure warning and risk hazards.Targeting the variety of metro failure data and the difficulty in defining an impact degree,an improved Apriori algorithm that considers failure association was proposed.Compared with a classic FP-Growth algorithm,metro failure association rules were studied,and subsequently their basic ideas and processes were optimized.Finally,taking metro equipment failure data in 2020 as an example,a model was simulated and established based on the Python.Results showed an association rule among ATP failures,signal equipment failures and other types of failures,which provides an important reference for failure degree analysis,failure diagnosis,failure warning,risk hazard classification,etc.
作者
刘文雅
徐永能
LIU Wenya;XU Yongneng(Nanjing University of Science and Technology, Nanjing 210094, China)
出处
《兵器装备工程学报》
CSCD
北大核心
2021年第12期210-215,共6页
Journal of Ordnance Equipment Engineering
基金
国家自然科学基金项目(52072214)
国家重点研发计划项目(2017YFB1001801)。
