Multiobjective Transit Network Design With Travel Time Reliability for Conventional and Autonomous Electric Vehicles

초록

This study addresses the multiobjective transit network design and frequency setting problem, aiming to improve travel time reliability under urban congestion. Digital tachograph data collected from a large fleet are utilized to assess the variability of roadway travel times across the network. By integrating a multinomial logit mode choice model, this study explores the relationship between automobile demand and public transit use, highlighting the influence of urban congestion on transit performance. In addition, a comparative analysis between the operations of autonomous electric buses and those of conventional vehicles reveal potential improvements in operational benefits as well as enhancements in passenger experience. A sensitivity analysis reveals the network response to changes in operating unit costs and variations in total passenger demand. Experimental results from a large-scale, real-world transit network demonstrate notable improvements in operator's net costs, passenger travel times, and overall system reliability, suggesting promising directions for future transit network planning and sustainable urban mobility.

키워드