Power Load Prediction Model Based on Long Short Term Memory and Sparrow Search Algorithm

Abstract

Power load forecasting is an important reference for power demand evaluation. In order to promote the accuracy and stability of load short-term prediction, this paper brings up an innovative power load prediction method based on improved Long Short Term Memory(LSTM) recurrent neural network. Sparrow Search Algorithm(SSA) is employed to optimize the initial weights and bias of LSTM layers, which can boost the accuracy and convergence rate of the model. Finally, the processed dataset is used to train and test the SSA-LSTM model and the simulation demonstrates superiority of the proposed method.