Research On Boarding Choice Behavior Of PDL Passengers Based On SP Data

In recent years, there is a rapid development in railway. The appearance of high-speed trains and passenger dedicated line eases the tension in railway transportation, enhances the level of rail transport service, and promotes personnel and interregional economic and trade exchanges. In this paper, we choose rail passengers as study objects, research on the choice of passenger about transport service products based on Stated Preference data, and build the model of passengers' choice on high-speed trains, as well as transfer and nonstop trains. The results provide a basis for the design of transport service products for passenger dedicated line.Firstly the background of the research is dissertated. The actuality of the related issue is analyzed. The important meaning of the research and the main content and methods in the paper are put forward.Secondly the paper tends to introduce the theory and the implementation of the passengers' travel survey. We focus on the difference between Revealed Preference survey and Stated Preference survey as well as the respective applying conditions. Next the questionnaire design and survey method are expatiated.Thirdly we make a survey on passengers' trip. A detailed analysis is made including the basic information for travelers, the future high-speed train fare, the transportation mode of passengers' options on the long-distance and middle distance travel, the choice about the arrival and departure time of the passengers' trip, the choice about transfer mode and nonstop trains, as well as the choice between high-speed train and the sunset-departure and sunrise-arrival train in existing railways. These data support the built of final model.Last but not least we analyze the important factors that influence the passengers' choice behavior. Disaggregate theories are introduced, containing the utility theory and the familiar models. We make a quantified description on the impact of each factor of passengers' choice behavior. On these bases we establish binary logit model of passengers' choices on high-speed train, as well as transfer mode and nonstop train, meanwhile the models are solved. Finally the experimental results indicate that 87.8% and 82.28% of the predicted data fit the observed data. The results show the higher accuracy of the model.The research in this paper provides the foundation for the design of transport service products in the future passenger dedicated line network. The results provide a certain reference value for railway passenger assignment and the design and optimization of passenger train running-plan.