Coupled design of urban and interurban transit networks with a park-and-ride system: Analytical formulation and case study in Barcelona

Abstract

This research presents an analytical model for the coupled design of an urban transit network and an interurban one supported by park-and-ride facilities. The former network facilitates the mobility in the centre of the metropolitan area while the interurban services satisfy the centripetal connections from the metropolitan outskirts where park-and-rides are scattered. These three components of the system are interconnected. Both networks share common transfer stops in the main corridors of the central city while park-and-rides are the access points to the interurban network in the periphery. Consequently, the model introduces the concept of coupled networks to define the interrelations between two transit services and the park-and-ride facilities. The variables of density of park-and-ride stations, spacings among corridors and operating headways, control the design of the entire system. Moreover, the model is adaptive to diverse cities with different geographical constraints (e.g. sea or mountain boundaries).

The upper-level of a bi-level formulation minimizes user and agency costs of the system searching for the optimal joint configurations of urban and interurban networks and density of park-and-rides. This optimal design depends on a lower-level that computes the optimal location of the last park-and-ride ring and the resulting modal split derived from the performance of the transit system. This problem has been solved using a grid search method.

Lastly, the model is applied with real data from the metropolis of Barcelona (Spain), revelling that the optimal location for the park-and-ride rings is further from the central city when demand at its entrances is higher. In addition, the model prioritizes accessibility for the urban network and maintains low waiting times for the interurban services, establishing a clear hierarchy between them. The coupled optimization of both networks reduces general costs by 4.6%, emphasizing the significance of considering the entire transit system.