Incorporating Activity-based Special Generator Data into a Conventional Planning Model

Special generators need special attention in developing travel demand models since the standard trip generation and distribution model in the conventional four-step approach do not provide reliable estimates of their travel patterns. New modeling approaches such as activity-based and tour-based models, considering travel behavior of individual household or person, seem to be more appropriate for those special generators. However, only a few practical applications have been made since these approaches usually require a lot of data resources and computing time to solve their complicated model structure. The primary objectives of this research are to improve the trip generation and trip distribution of special generators (e.g., university) by applying an activity-based approach, and to provide a transitional methodology for practically incorporating the activity-based data into a conventional planning model. The research developed a spatial and temporal activity-based model dealing with special generator data of North Carolina State University (NCSU). Also, the research tested the transferability of university student travel data by using statistical approach and indicated that the university students' travel data can be transferred for the two cases considered. The NCSU activity-based model provided the estimates of trip generation at the disaggregated level of individual buildings by hours of the day - a disaggregation was not obtainable from a conventional planning model. The model estimates, student building presence and trip generation, compared well to field data from student registration records and student trips observed at sample buildings. The results revealed that the activity-based model well replicated both building presence and trip generation. In addition, the research compared the estimated trip generation of the activity-based model to that of a traditional planning model and discussed findings in terms of model accuracy, structure, data requirements, and capability of model application. The insights gained from this study will serve as the basis of activity-based Triangle Regional model in North Carolina.