Short-term load forecasting based on PCA-ILSTM

Abstract

To obtain accurate load prediction results from massive power data, a short-term load prediction method based on PCA and improved long and short term memory neural network is studied. Firstly, to reduce the degree of data redundancy and eliminate redundant attributes, the principal component analysis method is used to extract the main attributes that affect the power load, and the short-term load prediction model is established by using the obtained main attributes. To improve the prediction accuracy of long and short term memory network, a sparrow search algorithm was proposed to optimize the number of hidden layer neurons, learning rate and iteration times of long and short term memory network. The short-term load prediction model was established by sparrow search optimization of long and short term memory network. The simulation results show that the proposed method has better prediction effect than other algorithms, and the effectiveness of the proposed method is verified.