Research On Individual Travel Behavior And Travel Demand Forecasting Model

Urban traffic congestion has been a worldwide hot spot and difficult issue, the fundamental reason of which is the unbalance between the rigidity of traffic supply and the flexibility of traffic demand, that is the coupling conflicts of traffic supply and demand in mode, time and space. Faced with such contradictions, utilizing reasonable travel management measures to guide or restrict the dynamic generation of traffic demand while appropriately increasing transportation facilities, which is an effective approach to solve urban traffic congestion. As a powerful tool to analyze transportation management policies, current trip based travel demand forecasting and behavior analysis method refers to single trip as a basic analysis unit, doesn't consider the relationship between travels, particularly lacks the characteristics of travel decision-making, and can't explain several phenomenon in traffic behavior aspect, therefore it has been provided with significant deficiencies.According to theories of human activities and behavior, travel essentially derives from the demand of activity participation, and activity and travel decisions of individuals are influenced by families, time, space and life cycle. From the perspective of activity and travel decisions, this thesis investigates activity-based travel demand analysis and forecasting theories, models and case applications. In summary, this study mainly includes six aspects as follows.First of all, in terms of the theories of human activity and behavior, the representation modes of individual travel behavior were summarized. Activity-based travel demand analysis theory foundation was then reviewed, the framework of activity and travel decision was discussed, and the influential factors and choice process of activity decision-making were analyzed systematically. Based on that, the requirements of activity-based travel demand forecasting models was proposed, and the modeling methods and model structure of four representative activity-based travel models were also introduced.Second, due to its potential advantage in the field of activity-based travel behavior analysis, Time Geography method was introduced into the research of human activity and travel behavior in 3D space-time environment. A space-time path oriented 3D temporal GIS framework were then designed and developed, so that activity-travel patterns of individuals and the communication modes of people were visualized and analyzed in the integrated environment of space-time constraints.Third, based on the thought of disaggregated microsimulation, the architecture of activity-based travel demand forecasting model system was illustrated, including activity-based residents travel choice model and traffic equilibrium assignment model.Four, with the choice process of human activity and travel decision, as well as the random utility theory of econometrics, the module frame of activity-based residents travel choice model was established, which consists of a set of different levels and decision scale of Logit models. In addition, according to the time characteristics of residents travel choice, activity-travel choice model system was broke down into population simulator, family decision based residents long-term choice model, and day activity-travel schedule model, which were respectively implemented with C language in Delphi 2005 compiler. The estimation and verification of each sub-models were completed under the application framework of Cube Voyager transportation planning software, and the model fitting results are full day of activity-travel plans for all travelers.Five, traffic equilibrium assignment model were established based on Traffic Analysis and Simulation System (TRANSIMS). Besides, the simulation results of this model system were compared with those of traditional 4-step model, and demonstrated that the model developed by author is practical and advanced.Finally, the calibrated and validated activity-based travel demand forecasting model system were applied in the Ambassador Bridge Gateway Reconstruction project in metro Detroit, U.S., combined with 4 specific scenarios, which involve the roadway closure events and implemented operational mitigation measures. Simulation results were analyzed with emphasis on work zone mobility impacts and regional travel patterns changes, as well as departure time shifts by travelers.