Simulating one-way electric carsharing systems with a multi-agent model

Abstract

Abstract With improved technology and supported public policy, electric vehicles (EVs) are coming back in one-way carsharing from the 2010s. Although the addition of EVs offers a way to achieve carbon neutrality, the shorter vehicle range and longer charging time of EVs pose a greater challenge to the operation of one-way carsharing than fuel vehicles. Common methods (e.g., mathematical model) are difficult to deal with such a complex system composed of "user-station-parking space-vehicle-energy", where supply and demand are restricted by mutual feedback. Therefore, this paper builds a one-way electric carsharing model and integrates it into an open-source multi-agent transport simulation platform (MATSim) to study its supply and demand relationship. A Shanghai baseline scenario was built to validate the model and test the impacts of vehicle range, charging rate, and power supply mode on carsharing demand. The results show that: (1) Vehicle range expansion and charging rate improvement have less impact on carsharing demand. The current vehicle range and charging rate can meet the daily use in Shanghai. (2) When the power supply mode changes from charging piles to battery swapping, the carsharing usage decreases slightly (-3%), while the carsharing trip characteristics remain almost the same. This means that operators could use battery swapping for power supply. This study provides suggestions for electric carsharing operators in Shanghai, as well as a simulation tool for more operators to test the supply and demand relationship of electric carsharing.