Robust optimization model for the electric vehicle routing problem under battery energy consumption uncertainty with arc segmentation

Abstract

As the greenhouse gas emission regulations have strengthened, establishing a sustainable transportation system has become more essential. Thus, studies on the transportation system using electric vehicles have received more research attention. However, operation using electric vehicles has obstacles such as technical limitations of vehicle batteries and insufficient number of charging stations, which can be much affected by the traffic flow changes. Therefore, we propose a robust electric vehicle routing model under road traffic flow uncertainty. An arc segmentation method is used to apply the road traffic flow uncertainty more realistically and mitigate the conservativeness of the solutions. We also use the genetic algorithm with three-layer chromosome encoding schemes. Three-layer chromosome encoding schemes guarantee the feasibility of the solutions by checking the product and battery capacity of electric vehicles. The experiments show that the solutions of the robust electric vehicle routing model tend to visit charging stations more often due to the uncertain road traffic flow. However, simulation experiments show that the robust electric vehicle routing model provides better performance.