Abstract Electric vehicle (EV) loads impact on the electric power utility distribution network (DN) can be best investigated when studying the corresponding EV charging stations (EVCS) in the DN. However, different types of data/metadata need to become available for properly modeling EVCSs for relevant power system studies. One metadata is EV users’ behavior or charging pattern. Generative AI synthetic data can be utilized to replicate EV user behavior as an alternative to metered EV data. In this paper, we first present the EVCS electrical data requirements for cyber-physical modeling and its impact on two main categories of power distribution system studies, i.e., operational planning and electromagnetic transient studies. Distribution planners can benefit by ensuring these data requirements are met prior to delving into relevant studies. We then describe the methods for generating synthetic EV data, along with sample results. Synthetic EV charging data have the potential to be utilized as input EVCS load-shapes to inform distribution grid techno-economic and offline/real-time cyber-physical simulations.
Arzani et al. (Mon,) studied this question.