A comprehensive analysis method of urban rail transit operation accidents and safety management strategies based on text big data

Yanhui Wang,Kexin Sheng,Penghua Niu,Chenhong Chu,Man Li,Limin Jia
DOI: https://doi.org/10.1016/j.ssci.2023.106400
IF: 6.392
2024-01-03
Safety Science
Abstract:Although the occurrence of urban rail transit operation accidents threatens the safety of people's lives and property, there is little research on how to systematically mine the knowledge behind accident data in China, and there is also lack of methods to support the safety management strategy of urban rail transit operation. To address these shortcomings, this study proposes a comprehensive analysis method that combines qualitative and quantitative analysis of urban rail operation accident factors and safety management strategies based on accident data, which can provide reference for systematic analysis of such accidents in the future and will help urban rail safety researchers and decision makers to take effective safety management measures. The new contributions of the method are as follows: (1) a semantic network framework for accidents and a conceptual model of accident risk control chain are proposed; (2) a method for acquiring and constructing key contents based on semantic analysis is elaborated, and indicators for evaluating the influence degree of key item and the influence possibility of entries are proposed; (3) a management matrix is established to serve safety decision making based on indicators for the degree of attention that key item of accidents should receive. In order to explain the application of the comprehensive analysis method, 910 urban rail transit operation safety accident cases are analyzed in detail, and suggestions for the actual operation management of urban rail transit are proposed in the light of the actual situation of urban rail transit operation in China.
engineering, industrial,operations research & management science
What problem does this paper attempt to address?
The paper focuses on the safety management issues of urban rail transit operation accidents. Most of the current researches mainly focus on the analysis of accident causes and the modeling of accident processes. However, in China, how to systematically excavate the knowledge behind accident data remains to be studied, and there is a lack of methods to support the safety management of urban rail transit operation. In this regard, the paper proposes a comprehensive analysis method that combines qualitative and quantitative analysis to analyze the factors of urban rail transit operation accidents and safety management strategies based on accident data. The aim is to provide reference for the systematic analysis of such accidents in the future and to help researchers and decision-makers in urban rail transit safety to adopt effective safety management measures. The main contributions of the method include: 1) Proposing a semantic network framework for accidents and a concept model for accident risk control chains; 2) Elaborating on the methods of key content acquisition and construction based on semantic analysis, as well as indicators for evaluating the impact degree of key items and the possibility of item impacts; 3) Establishing a management matrix to serve safety decision-making based on the attention indicators of key accident items. To illustrate the application of this method, the paper provides a detailed analysis of 910 urban rail transit operation safety accident cases and proposes practical operation management recommendations based on the actual situation of urban rail transit operation in China. Traditional accident models often consider accidents as a series of events and overlook the interaction of factors at different system levels. This method, through big data analysis, attempts to systematically analyze the knowledge behind a large number of historical accidents from a quantitative perspective and propose corresponding safety management strategies to make up for this shortcoming. The paper verifies the effectiveness of the method by building an accident semantic network framework, a risk control chain model, and conducting material preparation and semantic analysis.