モデル予測制御を用いた躯体蓄熱型放射空調システムの最適制御に関する研究

抄録

 Thermo-Active Building System (TABS) is air conditioning system which utilizes the building frame as thermal storage part and radiation surface, and has been introduced in Japan in recent years. TABS offers higher energy efficiency, a more comfortable environment for workers and has advantages such as peak shift, reduction of heat source capacity and cost reduction by using large thermal mass. Also, control performance is important to create a comfortable thermal environment because the thermal response of the ceiling surface temperature is slow due to the large thermal mass.

 Therefore, research and development of new control methods is progressing. In particular, Model Predictive Control (MPC), that determines the current value of the optimal manipulated variable pattern while predicting the behavior of a future controlled variable, is attracting attention.

 In this study, we proposed MPC as optimal control method for an air conditioning system with large thermal mass. Moreover, we showed the effectiveness of this method by comparing conventional control method using coupled analysis of CFD analysis integrating general-purpose control simulator.

 The following results were obtained:

 1. MPC had optimized the manipulated variable (water flow rate) to satisfy the constraint conditions while taking into account the future controlled variable, internal load and set point. Therefore, the control error of ceiling temperature from set point in the case of MPC was smaller than ON/OFF control.

 2. MPC has the effect of reducing about 12% of the energy consumption required for cooling and transport power of piping water.

 3. It was confirmed that not only the sensor location but also the working space satisfied the comfortable range at 1.1m above the floor.