Development and Validation of Building Control Algorithm Energy Management

In this paper, a building control algorithm is proposed to reduce the electricity consumption of with variable refrigerant flow (VRF) system. The uses sequence-to-sequence long short-term memory (seq2seq LSTM) set target consumption, and VRF air conditioner system consumption. After setting applied as method updating addition, we propose two methods increase performance seq2seq LSTM model. First, among feature selection methods, random forest used select, numerous features data, only those that are most relevant predicted value. Second, use Bayesian optimization, which selects optimal hyperparameter shows best model performance. order conditioners, priority conditioners designated, prioritization being analytical hierarchy process (AHP). study, comparison without optimization proved achieved good Simulation demonstration experiments using were also conducted, showed decreased in similar manner reduction rate by means algorithm.