Load Prediction Algorithm Applied with Indoor Environment Sensing in University Buildings

Recently, building automation system (BAS) and energy management (BEMS) technologies have been applied to efficiently reduce the consumption of buildings. In addition, studies on utilizing large quantities data actively conducted using artificial intelligence machine learning. However, high cost installation difficulties limit use measuring devices sense indoor environment all Therefore, this study developed a comprehensive sensor module with relatively inexpensive sensors measure university building. an algorithm for predicting load in real time through learning based measurement is proposed. When reliability number occupants according CO2 concentration was quantitatively assessed, mean squared error (MSE), root square deviation (RMSD), absolute (MAE) were calculated be 23.1, 4.8, 2.5, respectively, indicating accuracy algorithm. Since used economical can easily existing buildings, it expected favorable dissemination prediction technology.