Optimal control of TABS using internal heat load prediction

Thermally Activated Building System (TABS) has attracted attention, as a means of creating an indoor thermal environment that satisfy both energy-saving and thermal comfort in the office. However, since the thermal response of the ceiling surface temperature is slow due to the ceiling's large thermal mass in TABS, optimal control that considers the load prediction results is desirable. In this paper, we propose a control method that incorporates load prediction by Neural Network with MPC. As a result, in the analysis using the proposed method, the control performance is improved by supplying water from the night before the load is generated, and the integrated water supply flow rate can be reduced.