Bayesian Network modelling for safety management of electric vehicles transported in RoPax ships. (May 2021)

• Record Type: Journal Article
• Authors: Wu, Bing; Tang, Yuheng; Yan, Xinping; Guedes Soares, Carlos
• Is Part Of: Reliability engineering & system safety. Volume 209(2021)
• Publication Date: 2021-05
• Subjects: Transporting electric vehicles -- Data-driven Bayesian network -- Safety management -- Fire accidents; Reliability (Engineering) -- Periodicals; System safety -- Periodicals; Industrial safety -- Periodicals; Fiabilité -- Périodiques; Sécurité des systèmes -- Périodiques; Sécurité du travail -- Périodiques; 620.00452
• DOI: 10.1016/j.ress.2021.107466
• Languages: English
• ISSNs: 0951-8320

Abstract

Highlights: A data-driven Bayesian Network to analyse the effects of the influencing factors on the consequences. The proposed model is applied to safety management of RoPax ships transporting electric vehicles. Results demonstrate that electric cars should not be allowed to charge during being transported by RoPax ships. Adjustment on the external temperature is an effective measure for consequence reduction. Abstract: The safety management of electric vehicles when being transported by RoPax ships is addressed, as this form of maritime transportation is gaining relevance owing to the increased production of this type of vehicles, which develop a complex chemical reaction mechanism and hazard characteristics (e.g. initial exothermic temperature, self-heating rate, pressure rise rate). This paper develops a data-driven Bayesian Network to analyse the effects of the influencing factors on the consequences, and to propose appropriate countermeasures. The kernel of this model is built from the analysis of a sample of 132 accidents of fire accidents in electric vehicles, to derive the quantitative and qualitative relationships amongst influencing factors by using mutual information and Expectation-maximization algorithm, respectively, and to further analysis the marginal probability to discover the effects of influencing factors on the consequences. Afterwards, the findings from sensitivity analysis are used to discover the key failure patterns, and the results demonstrate that it would better not to allow the electric cars to charge when being transported by RoPax ships because the occurrence probability of explosion will increase. Moreover, adjustment on the external temperature is also an effective measure for consequence reduction.