An improved risk assessment method based on a comprehensive weighting algorithm in railway signaling safety analysis. (August 2020)

• Record Type: Journal Article
• Authors: Liu, Chang; Yang, Shiwu; Cui, Yong; Yang, Yixuan
• Is Part Of: Safety science. Volume 128(2020)
• Publication Date: 2020-08
• Subjects: Railway signaling safety -- Risk assessment -- Maximum absolute weighted residual -- Maximum entropy method -- Comprehensive weighting algorithm; Industrial accidents -- Periodicals; Accident Prevention -- Periodicals; Safety -- Periodicals; Travail -- Accidents -- Périodiques; 363.11
• DOI: 10.1016/j.ssci.2020.104768
• Languages: English
• ISSNs: 0925-7535

Abstract

Graphical abstract: Highlights: The core of risk assessment is to obtain the evaluation matrix and weight vector. Most existing weighting methods have obvious subjectivity. Maximum entropy and maximum absolute weighted residual can be used in combination. The weighting result of comprehensive weighting algorithm is more reasonable. Weight vector can be used to calculate the dangerous failure rate and safety level. Abstract: There remains a lack of applied research on risk assessment technology in the safety analysis of the railway signaling system, and the research method in the existing literature mostly is the Analytic Hierarchy Process (AHP). This paper synthetically analyzes and evaluates the merits and demerits of the traditional AHP, the Maximum Absolute Weighted Residual (MAWR), and the Maximum Entropy Method (MEM), finally proposes a comprehensive weighting algorithm organically combining above methods. This algorithm applies to the weighting process under different evaluation matrix conditions. It neutralizes the excessive deviation caused by the human factors in the subjective weighting method and the data factors in the objective weighting method, so it makes the weighting result more reasonable. In addition, based on the comprehensive weighting algorithm, three kinds of typical railway signaling equipment are selected as calculation cases, and the weighting results can be applied in calculating the equipment dangerous failure rate, assessing the equipment safety level, and other risk management processes. This algorithm has practical application value for the risk assessment of the railway signaling system and its apparatus.