Electricity Load Prediction Based on WOA-BiLSTM-Attention

Abstract

Accurate power load forecasting helps to rationally improve the efficiency of the power system. In this paper, the WOA-BiLSTM-Attention power load forecasting model is used to combine the bidirectional long and short-term memory neural network, the attention mechanism and the whale optimisation algorithm model, to take the strengths and complement the weaknesses to form a new power load forecasting model. As the factors affecting the power load are far more than one, intricate and complex, a single aspect of the prediction model is obviously unable to accurately and comprehensively meet the existing prediction requirements, the common form of prediction model combination is to fully combine the advantages of the major algorithmic models, cleverly combined to effectively improve the prediction accuracy of the prediction model combination. The values of MAPE, R ² of the WOA-BiLSTM-Attention model used in this paper are 1.6712, 0.8147 respectively. Compared with the traditional model, the simulation effect of the WOA-BiLSTM-Attention power load forecasting model is significantly improved.