Random-Fuzzy Risk Tree Approach for Assessing Fire Hazard in Sub-healthy Electric Vehicles

This paper delves into the application of risk sensitivity analysis using a random-fuzzy risk tree framework for evaluating the fire risk associated with recalled electric vehicles (EVs). The proposed methodology aims to prioritize the significance of fire hazards within the EV system. Specifically, it assesses the extent to which risk probabilities at the basic event level influence the overall system-level risk probability. Recalled EVs often represent sub-healthy vehicles that harbor potential risks, including the threat of fires, which are characterized by latency, randomness, and fuzziness. Furthermore, there exists no clear-cut correlation between fire risk and various influencing factors. To date, there has been a paucity of specialized theoretical research focusing on risk information from the unique perspective of fire hazards in sub-healthy EVs. The present study collects recall bulletins issued in China over a four-year period, identifies fire hazards in diseased, sub-healthy EVs, and constructs a fire risk tree that unravels the fundamental influencing factors and their logical causal relationships. Leveraging random-fuzzy theory, we evaluate the risk sensitivity, accurately determining the influence weights of various risk hazards and prioritizing potential weak links. This combined approach integrates random-fuzzy analysis with sensitivity assessment, providing valuable insights for addressing fire hazards in sub-healthy EVs.