Data-Driven Building Energy Consumption Prediction Model Based on VMD-SA-DBN

Prediction of building energy consumption using mathematical modeling is crucial for improving the efficiency utilization, assisting in planning and scheduling, further achieving goal conservation emission reduction. In consideration non-linear non-smooth characteristics time series data, a short-term, hybrid prediction model combining variational mode decomposition (VMD), simulated annealing (SA) algorithm, deep belief network (DBN) proposed this study. VMD-SA-DBN model, VMD algorithm decomposes into different modes to reduce fluctuation data. The SA-DBN built each separately, DBN structure parameters are optimized by SA algorithm. results aggregated reconstructed obtain final output. validity performance evaluated on publicly available dataset, show that new significantly improves accuracy stability compared with several typical machine learning methods. mean absolute percent error (MAPE) 63.7%, 65.5%, 46.83%, 64.82%, 44.1%, 36.3%, 28.3% lower than long short-term memory (LSTM), gated recurrent unit (GRU), VMD-LSTM, VMD-GRU, DBN, SA-DBN, VMD-DBN models, respectively. will help managers formulate more-favorable low-energy reduction plans improve efficiency.