Resumen
The CapTA model of traffic assignment to a large transit network is aimed to deal with traffic phenomena, including a range of capacity constraints, as well as with passenger route choice (Leurent et al, 2014). To this end it has a 3-layer architecture that involves (i) passenger route choice on a network of transit legs and pedestrian links, (ii) line coordination of service missions, vehicle runs with their link running times and station dwelling times, track occupancy and passenger waiting at stations, (iii) vehicle running, including the alighting and boarding of passengers in relation to run dwelling time at the station, in-vehicle passenger load in relation to seat capacity and total vehicle capacity. Yet the line model is macroscopic both for passenger flows and for vehicles serving a given mission, through a frequency variable per station. The paper brings about a novel, hybrid line model that can be used on a standalone basis at the line level or by inclusion in a three-layered model such as CapTA. In the model, a macroscopic representation of passenger flows is hybridized with a microscopic simulation of vehicles. This enables us to simulate headway (ir-)regularity and derive its consequences on vehicle loads, passenger wait times and on-board comfort, track occupancy and vehicle delay for railway modes. The resulting leg generalized times will be used to better simulate optimal travel strategies for path choice in a network in traffic assignment models such as CapTA. After presenting the model and the simulation algorithm, we provide an application to a metro line under two alternative operational policies.