Abstract
Cable overload is one of the most critical contributors to early cable fires. This study proposes a hybrid Bayesian network (BN)-based fire risk analysis model, to investigate the evolution of overload-induced early cable fire risks. In particular, the fire risk transmission paths caused by cable overload are reported, considering the critical factors that likely lead to fires. A BN with a specific structure was considered using the fire risk transmission paths. Later, given the risk index system, a hybrid fire risk assessment model caused by cable overload was developed based on the entropy weight method. Subsequently, the corresponding risk levels were evaluated based on the evolution of the fire risk, using numerical simulations. Finally, a case study was conducted to validate the proposed methods, and the results indicated that the proposed methods can effectively evaluate the state of the cable and explain the causes of fire risk, which can be used for early fire warnings.