Robust design for underground metro systems with modular vehicles

Abstract

The asymmetric demands of metro lines in megacities can cause high passenger wait times and substantial underutilization of vehicle capacity. The problem is difficult to address because of passenger flow uncertainties and random delays. We propose a modular transit system (MTS) that allows a metro fleet to be dynamically dissembled and assembled in identical modules (or carriages) on metro terminals. A formal formulation of this issue is provided with a nonlinear programming (NLP) model that considers train power, greenhouse gas emissions, wind resistance, and operational economics. Then, a linearization of the NLP further facilitates its fast solution. By utilizing numerical experiments based on Shenzhen Metro data, we illustrate the mathematical model’s viability and confirm the model's usefulness in terms of the economic, low-carbon, and ecological consequences. Then, the robustness of the proposed model and the sensitivity analysis with various parameter values are reported.