基于事故数据的轨道交通运行安全风险辨识方法

Safety Risk Identification of Rail Transit Signaling System Based on Accident Data

摘要:

针对轨道交通运行信号系统安全风险定量化辨识问题，首先将原始的信号事故按照海因里希法则分为不同类型事故数据，再考虑事故多因素影响，在一般统计法、统计识别法基础上，提出Management?Machine?Man?Media?Mission factor，即管理?设备?人员?环境?功能 5M要素模型的因子分析安全风险辨识方法，分别辨识254件较大伤亡事故和220件较小伤亡事故的主因素。实验结果表明，自然灾害、材质不良、施工管理不善、防护不力造成电务人员伤亡是导致轨道交通信号系统事故的最主要因素。基于5M的因子分析法可在多尺度综合计量下计算得到事故因素重要性排序，与一般统计法和统计识别法相比，基于5M的因子分析法辨识效果最优，覆盖率适中，影响率提高了106%。

Abstract:

To quantitatively identify the safety risk factors in rail transit signaling system accidents， an accident data mining method is presented. First， the original accidents were divided into different types of accident data in accordance with the law of Heinrich. Then， a 5M（management-machine-man-media-mission factor） based factor analysis identification method was adopted to identify respectively the 254 larger-casualty accidents and the 220 minor-casualty accidents considering the accident causation multi-factors after the general-statistic method and the statistic identification method analysis. The results indicate that the natural disaster， the signal equipment material fault， the poor construction management， and the inadequate staff protection are the most important major factors leading to the signaling system accident. The 5M based factor analysis method can be used to sort quantitatively the importance ranking of accident causation factors under multi-scale comprehensive measurement. A comparison of the general-statistic method and the statistic identification method suggests that the 5M based factor analysis method has the best identification effect， with a relatively higher influencing rate of 106% increase and moderate coverage rate.