Electric Vehicle Load Prediction Based on Fuzzy Decision Making of Charging Willingness and Spatio-Temporal Markov Chain

Electric vehicle (EV) charging load is not only affected by travelling spatio-temporal behaviour, but also by the owner's subjective willingness, this paper proposes a fuzzy decision-making on the owner's charging willingness and a spatio-temporal Markov chain EV load prediction model, which takes into account the travelling spatio-temporal characteristics and the cost of charging. Firstly, the multi-segment transfer characteristics of EVs are analysed, and the Markov chain spatio-temporal transfer model is established using the transfer probability matrix and travel characteristic quantities of each region at each time of the day; secondly, after each segment transfer of EVs, a fuzzy method is used to make decisions on the owner's charging willingness by comprehensively considering the factors such as the battery charge level (SOC), the charging cost, and the mileage of the next segment, and the charging load prediction model is constructed; and lastly. Finally, through simulation verification and comparison with other methods, the proposed model is better than traditional methods in simulating owners' willingness and predicting charging load, and the prediction results are more effective.