Research on Green Building Energy Consumption Prediction Model Based on LSTM Neural Networks

Aiming at the problem that it is difficult for the current green building energy consumption prediction methods to take into account the temporal and nonlinear nature of the energy consumption data at the same time, a prediction method based on the long and short-term memory network is proposed. Firstly, the null values and outliers in the historical data are processed by means of mean padding, and the normalization operation is carried out to complete the preprocessing step of the data; then, the processed data are transformed to convert the time series problem into a supervised learning problem, so as to obtain the sample data for the training and verification of the model; finally, based on the long- and short-term memory network algorithm, the energy consumption prediction model is constructed. The experimental results show that the method can effectively perform energy consumption prediction, in addition, compared with the BP neural network algorithm, the method has a higher prediction accuracy.