Application and performance evaluation of long short-term memory network based on particle swarm optimization in building energy consumption prediction
MA Chao
PAN Song
CUI Ying
LIU Yiqiao
CUI Tong
WANG Haoyu
WAN Taocheng
Abstract：The traditional long short-term memory network(LSTM)is prone to fall into local optimum during parameter optimization.Introducing the particle swarm optimization(PSO)algorithm can enhance the global search ability of the model.A long short-term memory network model based on particle swarm optimization(PSO-LSTM)is proposed to conduct experiments on the energy consumption data of a certain building in northern China.Compare PSO-LSTM with six mainstream machine learning algorithms,namely convolutional neural network(CNN),LSTM,bidirectional long short-term memory network(BiLSTM),backpropagation(BP),extreme learning machine(ELM),and radial basis function(RBF).The results show that the coefficient of determination(R2)of the PSO-LSTM model reaches 0.95,which is 6.74%to 31.94%higher than that of the control model.It also performs well in indicators such as mean absolute error(MAE),mean deviation error(MBE),mean absolute percentage error(MAPE),and root mean square error(RMSE).It has good predictive ability and stability,providing an effective solution for the prediction of building energy consumption.
Keywords：building energy consumption predictionmachine learning algorithmPSO-LSTMperformance evaluationpredictive model
Publication Date：2025-07-25
Online Publishing Date：2025-09-15(First online date of this platform, not the publication date of the document)
Pages：5( 1-5 )