°ÇÃ൵½Ã°ø°£¿¬±¸¼Ò

Architecture & Urban Research Institute

pdf¿ø¹®º¸±â ¿¡·¯ ÇØ°á¹æ¹ý ¹Ù·Î°¡±â



¹®ÇåȨ > ¿¬±¸³í¹® > »ó¼¼

[¿ø¹®º¸±â½Ã ¼ÒºñµÇ´Â Æ÷ÀÎÆ® : 100 Æ÷ÀÎÆ®] ¹Ì¸®º¸±â Àοë

´ëÇѼ³ºñ°øÇÐȸ|¼³ºñ°øÇÐ³í¹®Áý 2025³â 6¿ù

³í¹®¸í ¹ÝµµÃ¼ Á¦Á¶¿ë Àü±âŬ¸°·ëÀÇ ½Ç³» ¿Âµµ ¹× ½Àµµ¿¡ ´ëÇÑ Çǵå¹é µ¿½ÃÁ¦¾î ¼öÄ¡½Ã¹Ä·¹ÀÌ¼Ç / Numerical Simulation on Simultaneous Feedback Control of Indoor Air Temperature and Humidity in a Fully Electric Cleanroom for Semiconductor Manufacturing
ÀúÀÚ¸í ¼Û±Ù¼ö(Geun-Soo Song) ; ¾çÁö¼®(Ji-Seok Yang) ; À̰ÇÇü(Kun-Hyung Lee) ; À¯°æÈÆ(Kyung-Hoon Yoo)
¹ßÇà»ç ´ëÇѼ³ºñ°øÇÐȸ
¼ö·Ï»çÇ× ¼³ºñ°øÇÐ³í¹®Áý, Vol.37 No.06 (2025-06)
ÆäÀÌÁö ½ÃÀÛÆäÀÌÁö(288) ÃÑÆäÀÌÁö(12)
ISSN 1229-6422
ÁÖÁ¦ºÐ·ù ȯ°æ¹×¼³ºñ
ÁÖÁ¦¾î ANN; ½Ç³» ¿Âµµ ¹× ½Àµµ; Àΰø½Å°æ¸Á; ¸ðµ¨¿¹Ãø Á¦¾î±â; ´ÙÁßÀÔ·Â ´ÙÁßÃâ·Â ½Ã½ºÅÛ; Çǵå¹é µ¿½ÃÁ¦¾î; °úµµ ÀÀ´ä ; Artificial neural network (ANN); Indoor temperature and humidity; Model predictive control (MPC) controller; Multi-input-multi-output (MIMO) system; Simultaneous feedback control; Transient response
¿ä¾à1 º» ¿¬±¸¿¡¼­´Â ±âÁ¸ÀÇ ¹ÝµµÃ¼ Á¦Á¶¿ë ÇÏÀ̺긮µå Ŭ¸°·ëÀÇ ¿ÏÀüÇÑ Àüµ¿È­¿¡ ÀÇÇÑ FFU ¹æ½ÄÀÇ Àü±âŬ¸°·ë ¸ðµ¨¿¡ ´ëÇØ MATLAB/Simulink Ç÷§ÆûÀ» ÀÌ¿ëÇÑ PID, on/off, ANN, MPC Á¦¾î±âÀÇ Çǵå¹é µ¿½ÃÁ¦¾î ¹æ½Ä¿¡ ´ëÇÑ Å¬¸°·ë ½Ç³»ÀÇ °Ç±¸¿Âµµ ¹× »ó´ë½ÀµµÀÇ ½Ã°£¿¡ µû¸¥ 1½Ã°£ µ¿¾ÈÀÇ °úµµ ÀÀ´ä¿¡ ´ëÇÑ ¼öÄ¡½Ã¹Ä·¹ÀÌ¼Ç °è»ê °á°ú¸¦ ¹ÙÅÁÀ¸·Î ´ÙÀ½°ú °°Àº ³»¿ëÀ» °üÂûÇÒ ¼ö ÀÖ¾ú´Ù.
(1) º» ¿¬±¸ÀÇ ¸ðµç Çǵå¹é Á¦¾î±â´Â Àü±âŬ¸°·ë ½Ç³»ÀÇ »ó´ë½Àµµ ¹× °Ç±¸¿ÂµµÀÇ µ¿½ÃÁ¦¾î¸¦ ÅëÇÑ ´ÙÁßÀÔ·Â ´ÙÁßÃâ·Â(MIMO) ½Ã½ºÅÛ¿¡ ´ëÇØ Á¤»óÀûÀ¸·Î ¸ñÇ¥ ¼³Á¤°ª 45%RH, 23¡É¿¡ ¾ÈÂø½Ãų ¼ö ÀÖ¾ú´Ù.
(2) º» ¿¬±¸ÀÇ Àü±âŬ¸°·ë ¿îÀüÁ¶°Ç¿¡¼­ »ó´ë½Àµµ ¹× °Ç±¸¿ÂµµÀÇ Çǵå¹é Á¦¾î·çÇÁ °¢°¢¿¡ ´ëÇØ °¢ ¼Òºñ·® °á°úµéÀÌ ºñ½ÁÇÏÁö¸¸ ¼öºÐ¹«³ëÁñ °¡½ÀÀåÄ¡ÀÇ ANN Á¦¾î±â, DCCÀÇ PID Á¦¾î±â°¡ ¾ö¹ÐÇÏ°Ô °¡Àå ÀÛÀº °ªÀ» ³ªÅ¸³¿À» °üÂûÇÒ ¼ö ÀÖ¾ú´Ù.
(3) º» ¿¬±¸ÀÇ ÀΰøÁö´ÉÇü AI Á¦¾î±âµéÀÎ ANN ¹× MPC Á¦¾î±â´Â »ó´ë½Àµµ ¹× °Ç±¸¿ÂµµÀÇ ¿À¹ö½´ÆÃ¿¡ ÀÖ¾î PID Á¦¾î±â¿¡ ÀÇÇÑ 3.6%RH, 3.3¡É¿¡ ºñÇØ °ÅÀÇ Á¦·Î ¼öÁØÀ» º¸ÀÌ¸ç ¸ñÇ¥ ¼³Á¤°ª¿¡ ´õ ½Å¼ÓÇÏ°í ºÎµå·´°Ô ¼ÒÇÁÆ®·£µùÇÏ´Â Ãâ·Â ÀÀ´äÀ» Á¦°øÇÑ´Ù´Â °ÍÀ» °üÂûÇÒ ¼ö ÀÖ¾ú´Ù.
(4) º» ¿¬±¸ÀÇ ANN Á¦¾î±â´Â ÀÚ½ÅÀÇ Àΰø½Å°æ¸ÁÀ» ÈÆ·Ã½Ãų º¥Ä¡¸¶Å·¿ë ºòµ¥ÀÌÅ͸¦ PID Á¦¾î±â·ÎºÎÅÍ ¼±Çà È®º¸ÇÔÀ¸·Î½á ¼º°øÀûÀÎ Á¦¾î µ¿ÀÛÀ» ´Þ¼ºÇÒ ¼ö ÀÖ¾ú´Ù. ÇÑÆí, MPC Á¦¾î±â´Â ANN°ú °°Àº PIDÀÇ º¥Ä¡¸¶Å· ºòµ¥ÀÌÅͰ¡ ÇÊ¿ä¾øÀÌ º» ¿¬±¸ÀÇ Àü±âŬ¸°·ëÀÇ ¼öÇÐÀû ¸ðµ¨°ú Á¦¾à Á¶°Ç ¹× ºñ¿ëÇÔ¼ö ÃÖÀûÈ­¸¦ ÅëÇÏ¿© ´Üµ¶À¸·Î ANN¿¡ ¹ö±Ý°¡´Â ¼öÁØÀ¸·Î ÀÛµ¿ÇÔÀ» È®ÀÎÇÏ¿´´Ù.
¿ä¾à2 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¡É by the PID controller. They also provided output responses that soft-landed more quickly and smoothly to target set points, respectively.
¼ÒÀåó ´ëÇѼ³ºñ°øÇÐȸ
¾ð¾î Çѱ¹¾î
DOI https://doi.org/10.6110/KJACR.2025.37.6.288