EVCCS: Realistic Simulation Framework for Electric Vehicle Commute and Charge

With the growing number of electric vehicles (EVs) on the roads, it is necessary to understand EV charging patterns within a network of public charging stations. This is needed for optimal allocation of chargers and coordination of charging requests. However, existing datasets do not reflect the charging patterns of EVs in a network of charging stations. Hence, this paper develops a realistic framework for EV commute and charge simulation (EVCCS). For a given U.S. city, the EVCCS uses publicly available data and statistics to create a realistic database of the city that reflects: (a) the population of the city and its distribution among residential neighborhoods, (b) the population of employees within each field, and their distribution among the workplaces within the city, (c) the locations of shopping and entertainment regions and public EV charging stations, (d) the average commute time within the city, and (e) the human daily travel motifs and real-time traffic conditions for weekdays and weekends. Using the established database, the EVCCS mimics EV movement within the city's roads and assigns them to public charging stations when needed. As a case study, the EVCCS is used to simulate the commute and charging of 54 EVs in the city of Cookeville, TN, USA, for one year. The generated dataset from the simulation is used to establish a probabilistic model that reflects the charging patterns among the charging stations. Some insights are highlighted based on the results.