Fuzzy reasoning approach and fuzzy analytical hierarchy process for expert judgment capture and process in risk analysis

An, M ORCID: https://orcid.org/0000-0002-1069-7492 and Chen, Y 2018, 'Fuzzy reasoning approach and fuzzy analytical hierarchy process for expert judgment capture and process in risk analysis' , in: Handbook of RAMS in Railway Systems: Theory and Practice , CRC PRess (Taylor & Francis), Boca Raton, USA, pp. 441-474.

[img] PDF (Chapter proofs) - Published Version
Restricted to Repository staff only

Download (4MB) | Request a copy


Many risk assessment techniques currently used in the railway industry are comparatively mature tools, which have been developed on the basis of probabilistic risk analysis (PRA), for example, fault tree analysis, event tree analysis, Monte Carlo simulation, consequence analysis, and equivalent fatality analysis. The results of using these tools heavily rely on the availability and accuracy of the risk data. However, in many circumstances, these methods often do not cope well with uncertainty of information. Furthermore, statistical data do not exist, and these must be estimated on the basis of expert knowledge and experience or engineering judgment. Therefore, railway risk analysts often face circumstances where the risk data are incomplete or there is a high level of uncertainty involved in the risk data. Additionally, railways are a traditional industry, whose history extends for at least two centuries. The existing databases contain a lot of data and information; however, the information may be both an excess of other information that cannot be used in risk analysis and a shortage of key information of major failure events. In many circumstances, it may be extremely difficult to conduct PRA to assess the failure frequency of hazards, the probability, and the magnitude of their possible consequences, because of the uncertainty in the risk data. Although some work has been conducted in this field, no formal risk analysis tools have been developed and applied to a stable environment in the railway industry. Therefore, it is essential to develop new risk analysis methods to identify major hazards and assess the associated risks in an acceptable way in various environments where those mature tools cannot be effectively or efficiently applied. The railway safety problem is appropriate for examination by fuzzy reasoning approach (FRA) and fuzzy analytical hierarchy process (fuzzy-AHP). The FRA method provides a useful tool for modeling risks and other risk parameters for risk analysis involving risks with incomplete or redundant safety information. Because the contribution of each hazardous event to the safety of a railway system is different, the weight of the contribution of each hazardous event should be taken into consideration in order to represent its relative contribution to the risk level (RL) of the railway system. Therefore, the weight factor (WF) is introduced, which indicates the magnitude of the relevant importance of a hazardous event or hazard group to its belongings in a risk tree. Modified fuzzy-AHP has been developed and then employed to calculate the WFs. This has been proven to facilitate the use of fuzzy-AHP and provide relevant reliable results. This chapter presents a development of a railway risk assessment system using FRA and modified fuzzy-AHP. The outcomes of risk assessment are represented as risk degrees, defined risk categories of RLs with a belief of percentage, and risk contributions.They provide safety analysts, managers, engineers, and decision makers with useful information to improve safety management and set safety standards.

Item Type: Book Section
Editors: Mahboob, Q and Zio, E
Schools: Schools > School of the Built Environment > Centre for Urban Processes, Resilient Infrastructures & Sustainable Environments
Publisher: CRC PRess (Taylor & Francis)
ISBN: 9781351978798
Related URLs:
Depositing User: Professor Min An
Date Deposited: 01 Apr 2019 11:43
Last Modified: 16 Feb 2022 01:33
URI: http://usir.salford.ac.uk/id/eprint/50816

Actions (login required)

Edit record (repository staff only) Edit record (repository staff only)


Downloads per month over past year