A comprehensive review on stochastic modeling of electric vehicle charging load demand regarding various uncertainties
Payam Farhadi, Seyed Masoud Moghaddas‐Tafreshi, Amir Shahirinia
Abstract
The transportation sector is undergoing an ever-increasing fast development, shifting from traditional fossil fuel-based transport to ultra-low emission electrified transport. To accelerate this transformation, appropriate planning and operation of charging stations (CSs) especially fast and ultra-fast CSs is of utmost importance. The initial and most important step toward optimum planning of CSs is to develop models for the prediction of electric vehicle (EV) load demand in order to estimate future charging profiles. Thus, a stochastic EV load model should be employed to get an accurate estimation of the total EV charging load regrading numerous interdependent uncertainties. The paper specially provides a critical review regarding different uncertainties related to EV load demand and various stochastic modeling approaches. For this purpose, related research works reported between 2005 and 2023 were gathered, screened, and summarized. Then, selected research papers are evaluated in terms of stochastic modeling of EV load demand. The stochastic approaches were categorized in two main groups of conventional and fast charging modes with regard to probability density functions, Monte Carlo Simulation, Fuzzy method, Markov chain, artificial neural networks, Copulas, and hybrid modes. Next, details of each methodology by highlighting related cons and pros were provided. It was obtained that most research works took into account three to five random variables (RVs) in-average for stochastic studies. In addition, various test and real-world networks throughout the world were employed to validate the obtained results. Finally, some potential future research areas in the field of stochastic CS planning and operation are presented.