ABSTRACT

In the context of the global energy crisis and economic crisis, energy conservation and emission reduction has become a key social goal. Campus buildings are a building group with full energy-saving potential, and reasonable campus energy management policies can fully tap the energy-saving potential of campus buildings. Based on the collected daily power consumption of four campus buildings, this study adopted the time-series prediction model, ARMA to forecast the daily power consumption in 30 days. To evaluate the accuracy of the predicted model, MAE, RMSE, and MAPE are adopted. Through verification, it is found that the prediction performance of the model is good, and the accuracy can meet the needs of campus building energy management.