Vous êtes ici : Version française > Parcours du doctorant > Formation
-
Partager cette page
Colloque / Séminaire
Deux Séminaires LICIT - 2/ Activity, Travel and Flow Dynamics: Equilibration on Congestible Networks
Le 31 octobre 2019
Deux séminaires : 14h30-16h30
Amphi du bas - ENTPE – Rue Maurice Audin, 69120 Vaulx-en-Velin
Langue / language: the presentation will be in English
Langue / language: the presentation will be in English
Présenté par : Dr. Hani S. Mahmassani, Northwestern University
Dr. Hani S. Mahmassani, Northwestern University
Résumé au format pdf + biographie de l'orateur
The travel modeling field has evolved from an exclusive focus on flows, to recognizing that these flows result from travel decisions made by people, and ultimately that travel choices are generally made to satisfy people’s engagement in a set of economic and social activities. Modeling activity choices in a largescale network context remains one of the challenges for transportation scientists and modelers. This talk presents a recent example of computational structures for integrating activity-based models of agent choices
within multimodal network simulation-assignment framework.
A fixed-point formulation and solution procedure are described for the joint activity schedule adjustment and dynamic traffic assignment problem. Activity patterns are defined in terms of activity start time, end time, sequence, and number of intermediate and final destinations. In the iterative solution scheme, planned schedules are loaded to the network to obtain the dynamic traffic conditions. Individuals who experience schedules inconsistent with the planned ones (due to congestion) may adjust their activity durations or trip
departure times to accommodate the fluctuations. In some cases, however, the travelers’ schedule inconsistencies may not be resolved through departure time or duration adjustments; therefore, the traveler may cancel an activity/trip to accommodate the fluctuations. We discuss strategies for selecting activities to be cancelled, and present the influence of incorporating activity cancellation on numerical convergence. Recent developments in integrated DTA modeling will also be discussed, including incorporating the effect of privately owned/controlled Autonomous Vehicles in the household activity travel process.
Résumé au format pdf + biographie de l'orateur
The travel modeling field has evolved from an exclusive focus on flows, to recognizing that these flows result from travel decisions made by people, and ultimately that travel choices are generally made to satisfy people’s engagement in a set of economic and social activities. Modeling activity choices in a largescale network context remains one of the challenges for transportation scientists and modelers. This talk presents a recent example of computational structures for integrating activity-based models of agent choices
within multimodal network simulation-assignment framework.
A fixed-point formulation and solution procedure are described for the joint activity schedule adjustment and dynamic traffic assignment problem. Activity patterns are defined in terms of activity start time, end time, sequence, and number of intermediate and final destinations. In the iterative solution scheme, planned schedules are loaded to the network to obtain the dynamic traffic conditions. Individuals who experience schedules inconsistent with the planned ones (due to congestion) may adjust their activity durations or trip
departure times to accommodate the fluctuations. In some cases, however, the travelers’ schedule inconsistencies may not be resolved through departure time or duration adjustments; therefore, the traveler may cancel an activity/trip to accommodate the fluctuations. We discuss strategies for selecting activities to be cancelled, and present the influence of incorporating activity cancellation on numerical convergence. Recent developments in integrated DTA modeling will also be discussed, including incorporating the effect of privately owned/controlled Autonomous Vehicles in the household activity travel process.