In-Depth Exploration of the Relationship between Charging Station Market Price and Power Consumption based on Machine Learning Algorithms

How to deeply explore the relationship between the market electricity price and electricity consumption of charging stations has become a problem that must be solved. This paper combines the gated recurrent unit (GRU) neural network with the XGBoost algorithm to construct a combined prediction model. The GRU neural network is used to model the time series data, and then the output of the GRU neural network is used as the input feature of the XGBoost algorithm, and the XGBoost algorithm is used to further explore the complex relationship between the market electricity price and electricity consumption. On multiple data sets, the RMSE value of the GRU-XGBoost model in electricity forecasting is as low as 15kWh, the MAE value of electricity price forecasting is about 2.07 yuan on average, and the determination coefficient remains advantageous. The GRU-XGBoost hybrid model proposed in this paper provides strong support for charging station operation decisions.