Short-Term Bus Load Forecasting Based on Combined Feature Selection and GRU-Attention Model

Short-term bus load forecasting is of great importance for power system dispatch and operation. In order to improve the accuracy of short-term bus load forecasting, a bus load forecasting method based on two feature selection algorithms and a gated cyclic unit with attention mechanism is proposed. This method firstly uses the pearson correlation coefficient method and the distance correlation coefficient method to obtain the correlation coefficient between weather feature quantities and bus load, thereby establishing a comprehensive correlation coefficient. Then, strongly correlated weather feature quantities are obtained based on the size of the comprehensive correlation coefficient, and input them into the gated cycle unit with attention mechanism model along with the historical bus load, and output the final prediction result. Through the verification of the actual bus load data validation, the method proposed in this paper achieves higher accuracy than the conventional forecasting methods.