Study on sensitivity of electric bus systems under simultaneous optimization of charging infrastructure and vehicle schedules

Abstract

The transition from traditional fuel-based bus transportation towards electric bus systems is regarded as a beacon of hope for emission-free public transport. In this study, we focus on battery electric bus systems, in which charging is possible at a variety of locations distributed at terminal stations over the entire bus network. In such systems, two intertwined planning problems to be considered are charging location planning and electric vehicle scheduling. We account for the interdependent nature of both planning problems by adopting a simultaneous optimization perspective. Acknowledging the existence of parameter uncertainty in such complex planning situations, which is rooted in potential changes of values of several environmental factors, we analyze the solution sensitivity to several of these factors in order to derive methodological guidance for decision makers in public transportation organizations. Based on the formulation of a new mathematical model and the application of a variable neighborhood search metaheuristic, we conduct sensitivity analysis by means of numerical experiments drawing on real-world data. The experiments reveal that it is not possible to identify persistent structures for charging locations by an a priori analysis of the problem instances, so that rather a simultaneous optimization is necessary. Furthermore, the experiments show that the configuration of electric bus systems reacts sensitively to parameter changes.