Risk analysis of human evacuation aboard passenger ships based on fuzzy DEMATEL-ISM-BN

Accidents involving passenger ships present unique and considerable challenges for evacuation and rescue, posing significant risks to life and property. However, existing studies rarely analyse the risk influential factors (RIFs) affecting the evacuation process of passenger ships and the interconnections among these RIFs. To address these gaps, this study proposes a risk analysis framework for the emergency evacuation of passenger ships, aiming to dissect and quantify the interdependencies of RIFs within the context of Human Evacuation from Passenger Ships (HEPS). Firstly, this study conducts a comprehensive review of existing literatures and extensive investigations into passenger ship accidents in order to identify the RIFs of HEPS. Secondly, combining fuzzy language rating data on the relationships between RIFs, the Fuzzy Decision-making trial and evaluation laboratory (DEMATEL) and Interpretative Structural Modeling (ISM) methods are employed to build a hierarchical network model. Thirdly, the hierarchical network model is mapped to a Bayesian Network (BN), facilitating a detailed investigation into the RIFs influencing HEPS. Finally, forward causal reasoning is employed to determine the probabilities of various RIFs occurring and reverse diagnostic reasoning is used to deduce the probability changes of RIFs under different accident severities. This study is of significant importance for enhancing the emergency evacuation capabilities of passenger ships and ensuring public safety.