Electric Vehicles Charging Optimization Method Considering Spatial and Temporal Distribution Charging Demands Prediction

A spatial and temporal distribution prediction model is built considering electric vehicles (EVs) trip characteristics, charging area difference and users' charging habits in this paper. Charging demands forecasting for EVs based on Monte Carlo simulation is proposed by simulating real-time charging behavior under multiple charging scenarios. Based on the preliminary prediction results, optimal charging strategy for EVs based on evolutionary algorithm is established with the night charging scenario as an example. Furthermore, the influence of different charging state (SOC) thresholds on the optimal charging is conducted. The results show that the proposed prediction model provides preferable charging load distribution and the optimal charging strategy achieves the real-time valley filling. This paper has significant guidance for load optimization of electric vehicles.