Improving Timetable Quality in Scheduled Transit Networks

This work deals with an original problem with regard to the traditionally sequential planning process in public transit networks. This problem aims at modifying the network's timetables without rendering the vehicle and driver schedules obsolete. The objective is to improve the quality of service for passengers through number and quality of transfers. This approach goes in the opposite direction compared to the usual approach which schedules resources once timetables are set. We propose a model and a solution method based on tabu search and a neighborhood specifically developed. Experiments are led on five instances related to a real transit network. Important gains are obtained on the considered case study, allowing for better mobility of users inside the network and on the intermodal level.

• Type: Journal
• Year: 2010
• Where: IEAAIE
• Authors: Valérie Guihaire, Jin-Kao Hao
• Keywords: Artificial Intelligence, IEAAIE 2010, Public Transit Networks, Sequential Planning Process, Transit Network