A Regional Electricity Load Growth Prediction Method Based on LSTM+Attention Model
In refrigeration and air conditioning systems,electricity consumption is influenced by various factors such as external temperature,building insulation,and indoor personnel activities,forming a complex electricity dependency network. If we only focus on the growth value of electricity load and ignore these dependency relationships,it will significantly increase the loss of predicted load. Therefore,a regional electricity load growth prediction method based on LSTM+Attention model is proposed. Fit the historical electricity load data of the analysis area,combined with the calculation of electricity dependency residual value,analyze the periodic characteristics of electricity load growth,introduce LSTM+Attention model to identify the influencing factor characteristics of electricity load,and obtain the predicted regional electricity load growth value by scaling the linear regression equation. The experimental results show that the prediction results obtained after the application of the proposed method exhibit small load loss and high prediction accuracy,meeting the power dispatch decision-making needs of regional power supply.
万方数据
维普
