Original scientific paper
https://doi.org/10.7307/ptt.v37i1.659
Importance Analysis of Causative Nodes for Accident Chains of Railway Locomotive Operation Based on STPA-PageRank Method
Ping WAN
; East China Jiaotong University, School of Transportation Engineering
*
Wei-Lun YANG
; East China Jiaotong University, School of Transportation Engineering
Jie-Wen LUO
; East China Jiaotong University, School of Transportation Engineering
Xiao-Feng MA
; Wuhan University of Technology, Intelligent Transport Systems Research Centre
* Corresponding author.
Abstract
Nowadays, in terms of complex and random incidents for locomotive operation, the prevention and control for every tiny and possible influencing factor is not only costly, but also brings great psychological burden to locomotive drivers. Firstly, 68 sets of data of railway locomotive operation accidents happened in recent two years were collected and compiled. Secondly, the system theory process analysis (STPA) method was adopted to extract 68 accident chains based on those data. Then, the complex network theory and PageRank algorithm were utilised to calculate the importance of every node in directed-weighted network formed by those accident chains. The results showed that the importance of human factors is significantly higher than other layers including environment, facility and management. Especially, no effective control behaviour (H7) and false control behaviour (H10) are the top two important causative nodes among all human factors. Besides, being forced to stop (D39) and overrunning of signal (D42) are the top two important causative nodes among unsafe events. For those nodes with high value of PageRank, some targeted security measures should be adopted, so as to save risk management investment and improve the overall safety level of the locomotive operation system.
Keywords
railway locomotive operation; risk management; complex network; STPA; PageRank
Hrčak ID:
327637
URI
Publication date:
6.2.2025.
Visits: 362 *