Ship survivability assessment considering correlation
Zhang Feng-xiang · Ship Science and Technology · 2010
The relationship among ship systems is a difficult problem in ship survivability assessment. In order to make the outcomes of the assessment more effective,an assessment model considering correlation was built by Bayesian networks theory. While building the model,on the one hand,a middle layer composed by the nodes of system capability was built,which is adopted from the analytic hierarchy process;on the other hand,according to the actual situation of the systems,cycle Bayesian networks was adopted,and the cycle was eliminated by computing the limit marginal probabilistic distribution so that cycle Bayesian networks was transformed to classic Bayesian networks. At last,through calculating the model by MSBNX,the correlative influence caused by electric system and damage control system was assessed effectively,meanwhile a conclusion was received that when the systems' survivability was certain the ship survivability could be enhanced by promoting damage control capability.