Analysis and optimization of transit network design with integrated routing and scheduling

Abstract

This study presents a methodology for transit network design with integrated routing and scheduling, which minimizes total user travel time, including in-vehicle travel time, waiting time, transfer time, and transfer penalties. Since the network which provides short in-vehicle travel time results in long waiting time, in-vehicle travel time resulting from routing, and waiting time resulting from scheduling are interdependent. Iterative approach was used in this study to generate user total travel time minimization network. The algorithm starts from the minimum in-vehicle travel time network, and then increases in-vehicle travel time while it decreases waiting time. To execute this methodology, TRANED (TRAnsit NEtwork Designer) was programmed. As results of the algorithm, major travel flows have direct connections, while minor travel flows have shorter waiting times with circuitous travel paths. This methodology consists of three major steps: Generation of an initial network, Assignment, and Network improvement. They are followed by a supporting step, Network analysis which calculates the results of the previous three steps. All these steps are iterated until the optimal transit network is generated. A simple example was used to illustrate the procedure of the methodology, and its results were generated and compared with those reached by previous researchers. Input elements for the model are demand volume, transit operating speed, transfer penalties, and vehicle capacity. Ranges of these elements were applied to the TRANED to analyze sensitivity of the generated transit network to each one of them. Networks obtained through these analyses were classified by geometric and operational features into three types: Transfer-oriented, Transfer-avoidance, and Directly-connected transit network. This methodology using an iterative approach offers greater flexibility than most previous studies which used simplified combinatorial approaches. So, the TRANED model can take into account various constraints and conditions, such as variable transit demand, operator's constraints, express service, given schedule information, and integrated service with an existing network. Applications of these factors and the resulting transit networks were also presented.

Cite this work

No public copy is stored for this record, so only the metadata is cited.

Stored only when both papers are in the corpus — 14 citations exist upstream.