Fire Risk Analysis in Ship Engine Room Based on Bayesian Networks

A method on fire risk analysis in ship engine room is proposed based-on Bayesian networks and according to the particular properties of the ship engine room fire.Since fires do not occur uniformly at every locations in the engine room space.Four high-risk areas,including main engine,generator,boiler and electrical equipment,are identified.Considering the fire developing process from initial event to ignition,a model of ship engine room fire risk based-on Bayesian networks is built,with which probability inference of the risk factors can be performed.Bayesian inference can help to judge the fire disaster and to give advice on the decision making of fire fighting.A sensitivity analysis of the Bayesian networks model related to evidence is made to sort the fire risk factors.In terms of the sorting sequence,appropriate measure can be taken to prevent fire occurrence.