A study of the predictive control of the Ondol System in apartments

The objective of this study is to improve the control performance of Ondol heating system in apartment buildings. For this purpose, prevalent control systems and relevant researches are evaluated. The control system of Ondol should easily be adapted to thermal mass characteristics and building loads variation without complication in real application. In this study, predictive control using the ANN model is proposed and the possibility of this predictive control is investigated through experimental research. The results show that the performance of this predictive control is better than that of the current 2- position on/off control. And through the dynamic analysis of thermal mass model using a computer, the adaptability of the predictive control to several load variations and thermal characteristics are evaluated.