Deploying battery swap stations for electric freight vehicles based on trajectory data analysis

Abstract

This article proposed a biobjective model to deploy battery swap stations for electric freight vehicles (EFVs) based on big data analysis. We particularly extracted trip, parking, and charging information of EFVs in Beijing from a one-week dataset containing trajectories of 17 716 EFVs (with a sample rate of 99.8%) in 2019 to define rules in the model and parameterize the model, so as to improve the model realism and accuracy. The biobjective model aimed to minimize the total cost of building battery swap stations and maximize operational efficiency of EFVs. The model was solved by a genetic algorithm. Parameter sensitivity analysis was also conducted. The test case of Beijing suggested that the biobjective model, together with genetic algorithm, could help freight companies find a set of Pareto optimal solutions to the deployment of battery swap stations. Among the solutions, the one with the highest investment in battery swap stations could reduce the average charging time of EFVs by 96.56%. In addition, the sensitivity analysis results suggested that the parameters related to battery, infrastructure, and number of EFVs were influential to both the total costs and operational efficiency of EFVs and should be considered carefully in the deployment of battery swap stations.