山东科学 ›› 2023, Vol. 36 ›› Issue (5): 67-74.doi: 10.3976/j.issn.1002-4026.2023.05.009

• 交通运输 • 上一篇    下一篇

基于改进的Apriori算法的铁路事故风险源关联分析方法

王宁(), 昌锡铭(), 杨欣, 吴建军   

  1. 北京交通大学 轨道交通控制与安全国家重点实验室,北京 100044
  • 收稿日期:2022-12-11 出版日期:2023-10-20 发布日期:2023-10-12
  • 通信作者: 昌锡铭 E-mail:20120910@bjtu.edu.cn;xmchang@bjtu.edu.cn
  • 作者简介:王宁(1995—),女,研究生,研究方向为交通运输规划与管理。E-mail: 20120910@bjtu.edu.cn
  • 基金资助:
    国家自然科学基金(71942006)

Association analysis method for railway accident hazards based on the improved Apriori algorithm

WANG Ning(), CHANG Ximing(), YANG Xin, WU Jianjun   

  1. State Key Laboratory of Rail Traffic Control and Safety, Beijing Jiaotong University, Beijing 100044
  • Received:2022-12-11 Online:2023-10-20 Published:2023-10-12
  • Contact: CHANG Ximing E-mail:20120910@bjtu.edu.cn;xmchang@bjtu.edu.cn

摘要:

为了预防铁路事故的发生,需要对造成铁路事故的风险源因素进行探索和分析,揭示铁路事故的发生规律。提出利用改进的Apriori算法,对铁路事故与风险源进行数据挖掘分析。考虑铁路事故伤亡的严重程度,提出新的支持度、置信度指标计算方法,对铁路事故因素进行加权量化。同时添加时间约束,探索不同时间的铁路事故风险源关联规则。利用英国铁路事故数据,挖掘铁路事故与风险源之间的关联规则,针对实际案例制定切实有效的预防措施。结果表明,利用改进的Apriori算法能够得到更多的铁路事故与风险源之间的关联规则,对于预防铁路事故的发生具有重要的作用。

关键词: 风险源, 铁路事故, 关联规则, Apriori算法, 数据挖掘分析

Abstract:

The causes of railway accidents are difficult to determine as several hazards can lead to accidents. To prevent the occurrence of railway accidents, the hazards responsible for railway accidents should be analyzed, and the occurrence rules of previous railway accidents should be revealed. In this study,data mining analysis on railway accidents and hazards was conducted using the improved Apriori algorithm.Considering the severity of accident casualties, a new calculation method for support and confidence indicators was proposed to weigh and quantify railway accident factors.Furthermore, time constraints were added to explore association rules of hazards with corresponding railway accidents at different times. Using the actual UK railway accident data, the association rules between railway accidents and hazards were discovered, and effective preventive measures were formulated for actual cases. Results show that the improved Apriori algorithm can explore more association rules between railway accidents and hazards, which can play an important role in preventing railway accidents.

Key words: hazards, railway accidents, association rule, Apriori algorithm, data mining analysis

中图分类号: 

  • U298.5