A Positive Model of Departure Time and Peak Spreading Dynamics

This paper develops a positive (in contrast to normative) approach for modeling departure time dynamics at the individual levels, and analyzes the consequent system-level peak spreading effects. The positive modeling approach avoids assumptions of substantial rationality, and focuses on how individuals actually make departure time choices. The proposed analytical framework theorizes how heterogeneous users learn time-dependent travel conditions, accumulate relevant spatial-temporal knowledge, form subjective beliefs,search for alternative departure times under sufficiently large stimuli, and adjust departure times based on subjective beliefs and decision rules. Following the theoretical framework, the authors specify learning with Bayesian methods, empirically estimates search start and stopping conditions that vary among users, and empirically derive search and decision rules from a joint reveal/stated-preference survey dataset. The resulting quantitative model is demonstrated with a numerical example. To enable the application of the proposed positive modeling approach, a low-cost and practical memory-recall survey method has also been developed to provide necessary behavioral process data for model estimation and validation. In addition, the individual-level departure time choice model is ready for real-world applications, and can be integrated with microscopic traffic simulation, simulation-based dynamic traffic assignment, and/or activity/agent-based demand models.