An Adaptive Sequence-Based Selection Hyper-Heuristic for Application to Electric Bus Scheduling

Abstract

Buses are important for public transportation and beneficial for the environment. However, diesel buses are significant polluters emitting greenhouse gases and particulates. Consequently, with the advent of electric vehicles there has been a drive to transition to electric buses. Key to this transition is to optimise electric bus fleets to reduce distance travelled whilst maintaining service levels. This is complex due to the added constraint of the limited range of electric buses. This paper considers the use of a Sequence-based Selection Hyper Heuristic (SSHH) method to solve this problem. Moreover, an adaptive SSHH (A-SSHH) technique is introduced which significantly improves upon SSHH. Indeed, bus fleet non-service distances and sizes are reduced by as much as 10% using A-SSHH over SSHH. Comparing with an optimised diesel bus fleet electric buses reduce carbon dioxide emissions by over 60% and importantly for fleet operators, energy costs are similarly reduced.