Application of bayesian network in the investigation of vehicle spontaneous combustion accident

In connection with the characteristics of the vehicle ignition accident, this paper put forward using the bayesian networks to investigate the vehicle ignition accident. According to the characteristics and causes of the accident, these cases can be divided into electric, fuel, exhaust and mechanical wear related four aspects. Accident in electric as an example, the interaction between causal cause accidents and related factors was analyzed, and the bayesian network model is established according to the combustion process. By the Bayesian algorithm, the probability of the event was computed and the probability of fire occurrence and factors was advised.