Fire Risk Assessment of Electric Vehicles in Underground Garages Based on Fuzzy Bayesian Network

In response to the phenomena of fire spread, explosion risk, and toxic gas release from electric vehicles in underground garages, We introduce a method for assessing the fire risk of electric vehicles, utilizing fuzzy approach to bayesian network. In this work, a fire cause-based model is developed based on expert experience and logistic regression analysis of historical samples. Twenty-three high-risk nodes were quantitatively analyzed using Netica to identify the high-risk factors. The experimental results indicate that fire hazard investigation and remediation (with a probability of 81.4%) is a primary factor contributing to the fire. This finding is highly significant for fire prevention, control, and safety management in underground spaces.