Method for Predicting Charging Load Range of Electric Vehicle Based on CTA-GPR
Aiming at the high randomness of large capacity batteries and individual charging behavior,which made it difficult to obtain accurate and reliable prediction accuracy for electric vehicle(EV)charging loads,at the same time there was a lack of uncertainty in quantifying charging load prediction,a new short-term single step electric vehicle charging load range prediction method was proposed.Firstly,a hybrid model combining spatiotemporal feature fusion and improved attention mechanism was constructed;then,Gaussian process regression was introduced for optimization,and a hybrid improvement method was put forward for range prediction analysis of EV charging load;finally,a verification was made based on the 35 d electric vehicle charging load data of a certain city.The experimental results show that the electric vehicle charging load range prediction method based on CTA-GPR achieves higher point prediction accuracy,appropriate prediction confidence interval,and reliable probability prediction results,indicating the effectiveness of the proposed method.
deep learningattention mechanismelectric vehicle(EV)charging loadrange prediction
万方数据
维普
