| Title |
Numerical Simulation on Simultaneous Feedback Control of Indoor Air Temperature and Humidity in a Fully Electric Cleanroom for Semiconductor Manufacturing |
| Authors |
Geun-Soo Song ; Ji-Seok Yang ; Kun-Hyung Lee ; Kyung-Hoon Yoo |
| DOI |
https://doi.org/10.6110/KJACR.2025.37.6.288 |
| Keywords |
ANN; 실내 온도 및 습도; 인공신경망; 모델예측 제어기; 다중입력 다중출력 시스템; 피드백 동시제어; 과도 응답 Artificial neural network (ANN); Indoor temperature and humidity; Model predictive control (MPC) controller; Multi-input-multi-output (MIMO) system; Simultaneous feedback control; Transient response |
| Abstract |
This paper presents transient responses in one hour of indoor temperature and relative humidity according to various simultaneous feedback control schemes for a semiconductor manufacturing electric cleanroom. The MIMO (multi-input-multi-output) feedback was used to control schemes include on/off switching, proportional-integral-derivative (PID), artificial neural network (ANN), and model predictive control (MPC) controllers. Simultaneous feedback control block diagrams for the electric cleanroom were created using the MATLAB/Simulink platform. Simulation results showed that all of the present MIMO feedback control schemes were able to control the indoor temperature and relative humidity to stably settle to set points. In the present electric cleanroom, the ANN controller for the water spraying nozzle humidifier and the PID controller for the DCC were strictly the most water- and energy-saving, respectively. The present ANN and MPC controllers are artificial intelligence-type controllers. They showed almost zero levels of overshooting of indoor relative humidity and temperature compared to 3.6%RH and 3.3℃ by the PID controller. They also provided output responses that soft-landed more quickly and smoothly to target set points, respectively. |