| Title |
Development of a Thermal and Electrical Performance Prediction Model for a Photovoltaic-thermal System Using an Artificial Neural Network |
| Authors |
Jinhwan Oh ; Sangmu Bae ; Yujin Nam |
| DOI |
https://doi.org/10.6110/KJACR.2022.34.4.155 |
| Keywords |
실증실험; 인공신경망; 태양광열 시스템; 열 효율; 전력 생산량 Real-scale experiment; Artificial neural network; Photovoltaic-thermal system; Thermal efficiency; Electrical production |
| Abstract |
Recently, Zero Energy Building (ZEB) that reduces the energy consumption and energy demand is becoming major issue. To realize ZEB, high energy efficiency renewable energy systems are actively applied in the building. Among the renewable energy systems, the photovoltaic-thermal (PVT) systems can respond heat and electricity demands of the buildings with one facility. Although, it is necessary to accurately predict the performance of the PVT system, most studies are conducted at the mock-up test level, through energy simulation and laboratory-scale experiment. In this study, the real-scale experiment plant that comprises the PVT module, heat storage tank, and actual building was constructed to collect the database of the PVT system under actual conditions. Moreover, based on the real-scale experimental data, the performance prediction model was constructed using artificial neural network (ANN). To verify the accuracy of the performance prediction model based on ANN, the coefficient of variation root mean square error (Cv(RMSE)) proposed by ASHRAE Guideline 14 was used. The Cv(RMSE) for the predicted result by ANN was calculated as 10% of the thermal efficiency, and 16% of the electrical production. Through the Cv(RMSE) result, it was confirmed that the performance prediction model based on ANN is reliable. |