Positive Feedback: Exploring Current Approaches in Iterative Travel Demand Model Implementation (0-6632-1)

Currently, the models that TxDOT's Transportation Planning and Programming Division (TPP) developed are traditional three-step models (i.e., trip generation, trip distribution, and traffic assignment) that are sequentially applied. A limitation of this sequential approach is an inconsistency between the travel time data used in the different stages of the process that may result in: (1) TDMs, which do not accurately reflect system-wide or corridor-level travel patterns, (2) travel times in alternative analyses that may not reflect accurate results, and (3) inaccurate results being used for the air quality determination process. To resolve these differences in the model sequence, researchers proposed an iterative feedback mechanism as an approach. This project researched current trends, practices, and tools in implementing a feedback approach for potential implementation in the TxDOT TDMs. In general, past research and practice underscores the importance of incorporating feedback loops in the travel demand process, especially in regions with moderate to high levels of traffic congestion during certain times of the day. Since the current TxDOT TDM model does not include any impedance or accessibility measures in the trip generation step, the best current approach is to feed back the output from the traffic assignment step to the trip distribution step. Overall, critical consideration should be given to the relative impact that implementing feedback might have on existing TxDOT TPP practices. A feedback mechanism necessarily includes additional training and increases the calibration time associated with delivering base year travel models; and possibly could include additional data processing (and collection) related to time-of-day implementation.