Design and Verification of a Load Forecasting Model for Charging Stations Using Image Classification

This paper proposes a new classification prediction scheme, which applies user profiling technology to load forecasting, uses RF threshold classification method and BIRCH clustering to form a user data feature classifier (RF-BIRCH), and then combines Bi-LSTM and CNN_LSTM to form A hybrid prediction model was developed. The core idea is to perform classification prediction based on user portrait segmentation data, and then compared the performance with some commonly used deep learning prediction models. Finally, it was verified that the prediction scheme proposed in this article has better prediction performance.