Travel demand models are undergoing a transition to more advanced paradigms for both travel demand and network simulation. The new demand models, referred to as Activity-Based Models (ABMs), are based on the principle of microsimulation of individual activity schedules. In practice, however, these demand microsimulation models are still only loosely coupled with aggregate static network procedures. Recent advances and successful applications of simulation-based Dynamic Traffic Assignment (DTA) models opened a way for an integrated ABM-DTA approach.
This presentation reports the first attempt to create a complete microsimulation model for the Chicago region whereby an ABM is integrated with a multimodal DTA. This integration required many innovations to both the ABM and DTA models. The Chicago ABM required an enhanced level of temporal resolution and more elaborate methods for conversion of the person trips into vehicle trips. The Chicago DTA required modifications of the route choice and path finding procedures to address individual value-of-time, value-of-reliability, occupancy, and other user characteristics. Another major effort included development of a multimodal dynamic transit assignment tool, which includes walk and bicycle access modes and is integrated with the highway-based DTA. The integration also included innovative theoretical and software solutions for handling individual trip trajectories instead of aggregate Level-of-Service (LOS) skims. The integrated model tracks each individual in time and space with full consistency between the activity and trip schedules. We believe that the integrated ABM-DTA model represents a prototype of a travel demand model that will be widely applied in practice in the future.