Methods Based On The Bayesian Networks For Traffic Accidental Event Situation Assessment

The urban transportation relates the city social and economic activity, is the life vein of the urban development. The urban transportation system is not only the source of traffic accidental event but also is the lifeline of emergency guarantee. The traffic accidental event's handling delay will bring the heavy loss to personnel's life and the social economy. It is an important aspect of Intelligent Traffic System to quickly response and transaction.In order to quickly process the traffic accidental event and reduce delay, this paper presents application of situation assessment which is widely applied in military domain to deal with traffic accidental event. The main jobs can be summarized as follows:This paper has researched situation assessment and the principium of Bayesian networks, and analyzed the feasibility of situation assessment which was used to deal with traffic accident event, and presented the overall workflow of situation assessment. This paper has researched the theory how to ensure node and structure of Bayesian networks, and combined actual situation of traffic accidental event to ensure them, and researched EM algorithm. This paper has researched information transform algorithm, and presented the Bayesian network structure of accident reason inference according to reversion inference. On the base of shortest path algorithm in time-dependent network, this paper has researched resource scheduling algorithm based on time window theory, and presented specific description. Based on above theory research, this paper has compacted the event which the Dalian traffic police crew processed the traffic accidental event to develop a prototype system" the traffic accidental event situation assessment and scheduling scheme generation system".This paper's research indicates methods which based on the Bayesian networks for traffic accidental event situation assessment can provide decision supports to quickly response and process of event, and shorten the processing time in which decision department made emergency process decision, and win time to rescue the personnel life in event and reduce society's economic loss.