Research on Electric Vehicle Charging Load Forecasting

Abstract

The peak to valley ratio of power grid load is increasing day by day, and the peak shaving task and pressure faced by the power grid are becoming increasingly severe. In this case, the charging load prediction is more important. For example, for electric vehicles, with the rapid development of electric vehicles and their supporting technologies, the scale and number of electric vehicles are growing rapidly, and the charging mode is gradually moving towards rapid charging, which has a certain impact on the load characteristics of local areas and even the whole province. In order to accurately grasp the charging characteristics of electric vehicles and prepare for grid planning in advance, the calculation and prediction of charging load has great value for orderly charging of electric vehicles and feedback to the grid as energy storage equipment.A charging load forecasting method is proposed in this paper. It is to predict the charging load in the next 24 hours according to the charging load of the day. The program is developed by python. The core part of the program is completed through the BP neural network. By training the BP neural network and testing the trained model, the accuracy rate is used to test the prediction effect of the model, complete the establishment of the prediction model, and predict the charging load curve in the next 24 hours. The fitting degree of the prediction results is good, which indicates that the prediction results meet the expected requirements.