A Bi-objective Stochastic Hub Location-Routing Model for the Railway Rapid Transit Network Design

This study focuses on the development of a stochastic bi-objective hub location-routing model for a railway rapid transit network design problem. Due to the use of railway rapid transit systems in the hub-level sub-network (i.e., the network among the hub nodes) and the spoke-level sub-network (i.e., the network that connect spoke nodes to each other and to hub nodes), the decisions to make concern the location of hub nodes, spoke nodes, hub edges and spoke edges, and the determination of hub and spoke lines, the percentage of satisfied demands, and the way of routing the demands, simultaneously. Uncertainty is assumed for demands represented by a finite set of scenarios. The problem is formulated through a two-stage stochastic modeling framework. The aim is to maximize the total expected profit and to minimize the total expected service time. The performance of the model is evaluated through computational tests using the well-known AP dataset. The computational results confirm the importance of considering the stochastic model and the conflicting profit and time objectives for the given problem. Some managerial insights are also provided through the analysis of the resulting networks under various parameter settings and the investigation of the effect of these settings on the characteristics of the obtained solutions and the interactions among the different aspects of the studied complex decision problem