Forecast of urban EV charging load and smart control concerning uncertainties

Abstract

Popularizing electric vehicles (EV) can be one of the most effective ways to deal with the ever severe air pollution. However, the accumulated charging power from the increasing integration of EVs could add large pressure to the peak of power grid. In this paper, a novel EV load forecasting model is formulated based on Markov chain, allowing for the stochasticity of user behavior, traffic and weather. The impact of EV integration is assessed in a typical medium voltage (MV) system while the resulting peak load and power loss are mitigated with charging control schemes. The proposed model could help to forecast the future charging demands with probabilistic uncertainties. The smart control schemes shall instruct the aggregator to make optimal charging plan concerning security and efficiency issues.