Research on Building Energy Consumption Prediction Method Based on LSTM Network

Building energy consumption forecasting is of great significance to building energy conservation. The accuracy of traditional building energy consumption forecasting methods is inadequate. Based on the extensive application of LSTM neural network in time series forecasting, this paper uses LSTM neural network to establish building energy consumption forecasting model, collects energy consumption training data through building energy consumption monitoring system, and proves LSTM by comparing the forecasting effect of LSTM neural network and BP neural network. LSTM neural network is superior to BP neural network in convergence speed and prediction accuracy.