基于CNN-LSTM-AM的电动汽车充电负荷预测

打开文本图片集

中图分类号：U469.72 文献标识码：A 文章编号：1003-8639（2025）10-0062

【Abstract】In view of the challenges brought by large-sale electric vehicle charging loads to the power grid and thelimitationsof traditional prediction methods，this paper proposes ahybrid neural network loadpredictionmodel based on CNN-LSTM-AM，comprehensivelyconsidering theinfluenceofload corelation factors in typical scenariosoncharging load prediction.The method proposed inthis paperisappliedtothepredictionof electric vehiclecharging load in Hainan Province.By comparing it with traditional prediction methods，the efectiveness ofthe proposed method is verified.

【Key Words】electric vehicle;chargingload;CNN-LSTM-AM;forecasting model; time series

0 引言

电动汽车（ElectricVehicle，EV）的快速普及在推动可持续发展的同时，其大规模无序充电行为对电网的稳定性、安全性和经济性构成严峻挑战[1]。（剩余4151字）