Study On Travelers' Route Choice Behavior During Traffic Incident

Traffic incident is the non-normal events what happens on the road and results in reduction of the capacity and the service level of the road. In case of incident, highway capacity would lead to the sharp reduction, which not only causes traffic delay and increases travel cost, but induces secondary incident and environmental pollution. The applied study of travel behavior during traffic incident has not been reported in china, so the analysis for travel behavior during traffic incident is not only for excellent understanding of traffic characteristic, but also for evaluating the feasibility of emergency management and the influence of traffic system, environment and society. Research on travel behavior during traffic incident could raise emergency incident management level and upgrade the service level of road network by dispatching resources appropriately to the fullest extent possible.In this paper, based on the analysis of the travel behavior under the influence of traffic incident, a route choice model based on mixed logit model is presented and calibrated in this paper. Firstly, traffic incident has been classified by different standards, then the travelers' dynamic cognitive and decision process is given out in the thesis, which is a continuing learning and readjusting process. Behaviour trait of traveler is analyzed in terms of psychological characteristics, incident type, trip state, incident position. Secondly, according to the SP survey, the descriptive statistical analysis and relation characteristics analysis between variables are studied. Finally, on this basis, the mixed logit model and the multinomial logit model are builded, which can describe travelers'route choice during traffic incident. By sample analysis, the route choice model based on mixed logit is much closer to the actual trip.The results show that the main factors which affect the route choice behavior are incident position and traffic information. The mixed logit is applied to route choice research during traffic incident, where the coefficient of sex, location and status are uniformly distributed, the coefficient of the factor of route choice and traffic information are normally distributed and the rest are fixed. The results for devising emergency incident management tactics and information release strategies have certain practical significance.