Travel demand model is an integration of mathematical models to mimic travelers' choice behavior and various data stored in computer (e.g. transportation networks, spatial distribution of population, employment and land use), in order to simulate travel behaviors of the entire population and predict both vehicle and passenger numbers on transportation networks. Since the model mimics travelers' behavior, it can simulate travelers' behavioral changes in response to Travel Demand Management (TDM) policies or new transportation facilities. Thus, travel demand model can be applied to scientifically evaluate the impacts from transportation policies or investments on the entire transportation network.
Traditional travel demand model is developed based on trip frequencies at a zonal level. Individual traveler' trips are initially aggregated into zones, called Traffic Analysis Zones (TAZ), and then form OD (Origin-Destination) trip matrices to represent the total number of trips between two TAZs. Then, the total number of trips will be converted to vehicle and passenger numbers and finally assigned onto transportation networks.
Consequently, traditional model will be facing many challenges when applied to evaluate TDM policies. For example, it cannot evaluate the impact of flexible work schedule policies since the model just considers travelers' trips but ignores the impacts from their activity schedules; it does not well accommodate population heterogeneities; the trips made by the same traveler will be incoherent in time, space and mode.
To overcome the shortcomings of traditional models, the United States started the theoretical research on activity-based travel demand models as early as the 70s of the last century. Currently, the activity-based model is being moved from theory to practice. Many US Metropolitan Planning Organizations (MPO) and State Departments of Transportation are investing on the development and application of activity-based models. Activity-based models break through the limit of TAZ and treat each individual traveler as analytical unit, simulate daily activity schedules of travelers living in different environments, then derive their travel behaviors and estimate the demand on the entire transportation network. Since breaking through the limit of TAZ, the activity-based model can overcome shortcomings of traditional models.
In recent years, the rise of big data technology brings new opportunities for the development of travel demand models. Big data from different sources, such as floating vehicles, transit smart cards, license plate recognition and mobile phone signaling, have given the researchers a new global view of urban travel demand, making it possible to significantly improve model accuracy and precision. The research team's goal is to explore the travel demand model theory for the current situation of our country with the help of mobile Internet and big data environment, develop a new generation of travel demand model and apply for transportation planning and management. Our team focuses on interdisciplinary collaboration , lays emphasis on international and domestic academic exchanges, provides a good learning platform for researchers and students , and trains high-end talents in travel demand modeling theory and practice .
