시뮬레이션을 통한 융복합 히트펌프 시스템의 적응형 예측제어 알고리즘 성능평가

• Published in: KIEAE Journal Vol. 21; no. 6; pp. 55 - 62
• Main Authors: 조혜운(Hye Un Cho), 최영재(Young Jae Choi), 최은지(Eun Ji Choi), 김태원(Tae Won Kim), 문진우(Jin Woo Moon)
• Format: Journal Article
• Language: Korean
• Published: 한국생태환경건축학회 01.12.2021
• Subjects: 건축공학; 인공신경망; Artificial Neural Network; 최적 제어; Hybrid Heat Pump System; 실내온도 예측; 융복합 히트펌프 시스템; Indoor Temperature Prediction; Optimal Control
• ISSN: 2288-968X; 2288-9698
• DOI: 10.12813/kieae.2021.21.6.055

Purpose: An artificial neural network (ANN) based adaptive & predictive control algorithm was developed and performance was tested. The adaptive & predictive control algorithm increases the control stability and energy efficiency of the hybrid heat pump system by deriving optimal flow rate of fan coil unit circulating water. Method: A target building was modeled using TRNSYS 18 computer software to generate dataset for training the predictive models and performance evaluation of developed algorithm. In order to confirm the superiority of the adaptive & predictive control algorithm, On/Off controller and non-adaptive & predictive control algorithm were employed for a comparative simulation. Result: Both the prediction accuracy and control stability of the adaptive & predictive control algorithm for the heating and cooling period were higher than the comparison group. In addition, as a result of energy consumption analysis, it was confirmed that energy savings are possible when optimal control is implemented for high partial loads. Therefore, if the adaptive & predictive control algorithm is employed to the various environment, rapid and stable adaptation is expected to be possible without additional optimization process. KCI Citation Count: 0