Multi-Agent Trips Simulation-Based Charging Demand Forecasting

Abstract

As the market penetration of electrical vehicle (EV) in modern transportation sector, the availability of public charging stations has become a major concern for EV owners. The lack of public charging infrastructure is hindering the further adoption of EVs in Japan. Therefore, a practical forecast is needed to analyze the spatiotemporal distribution of charging demand and aid in related fields such as infrastructure allocation and electricity grid management. This study aims to address this problem by generating high-resolution daily trips of each agent from real-life personal trips survey and deploy on multi-agent simulation software, MATsim. Subsequently, according to the proposed energy consumption model and charging rules for different types of EVs, the spatiotemporal distribution of charging demand is aggregated in the studied regions. We then compare the existing public charging stations with the predicted charging demand to identify overloads in current charging stations. The analysis revealed significant variability in charging demand across individual stations. These findings underscore the uneven utilization of charging infrastructure and emphasize the need for targeted interventions in both high-demand urban areas and underutilized suburban stations. Our framework effectively models agents' trip trajectories and generates the spatiotemporal distribution of charging demands in a bottom-up fashion, even with limited information. © 2025 The Author(s). IEEJ Transactions on Electrical and Electronic Engineering published by Institute of Electrical Engineers of Japan and Wiley Periodicals LLC.