Optimization of demand-responsive transit systems using zonal strategy

Abstract

In this thesis, two mathematical models were developed and applied to optimize demandresponsive transit systems using zonal strategies. Demand-responsive transit systems provide services between a residential area and a terminal. The first model considers a scenario where the terminal is located outside the service area, which optimizes the area of zones and bus capacity by minimizing the costs occurring within the entire service area; while the second model designs the transit system with its terminal sitting within the service area by minimizing the costs generated in service zones separately. Case studies for the application of both models were conducted for a presumptive demand-responsive transit system operating in northwest Calgary,showing that the models can successfully assist in designing demand-responsive transit systems.Furthermore, sensitivity analysis was conducted to identify the sensitivity of the model output to each model parameter.iii Acknowledgements

A survey on demand-responsive public bus systems

“…The objective is to minimize the total cost, including operator and user costs. In a variation of a demand-responsive system with the terminal inside the serving area, Wang et al (2018) optimize how a large area should be divided into zones for each cycle of service. Each zone is served by a feeder demandresponsive system line assigned to a selected terminal within the area.…”

Joint Optimization of Zone Area and Headway for Demand Responsive Transit Service under Heterogeneous Environment

“…Most studies did not consider the heterogeneity of the service region and oversimplified environment, which may lead to biased estimation of system performance. Few studies considered heterogeneous environment of the service region while optimizing transit systems (Chien and Schonfeld, 1997;Chien and Yang, 2000;Chien and Qin, 2004;Wang, 2017;Kim and Roche, 2021).…”

“…Step 1: Select a community located adjacent to the region boundary farthest away from the terminal (Wang, 2017), and assign the community as a part of zone i (e.g. i = 1).…”

Designing Zonal-Based Flexible Bus Services Under Stochastic Demand

“…Karabuk (2009) introduced a cluster-first route-second zonal heuristic that clusters customers to receive service together and arranges vehicle routes based on the zonal clustering. Recently, there were some studies of DRT that divided the service zone before planning, such as Lu et al (2017) and Wang et al (2018). In the former paper, each zone is served by one operator, and vehicles cannot traverse boundaries unless they pick up or drop off customers.…”