• ๋Œ€ํ•œ์ „๊ธฐํ•™ํšŒ
Mobile QR Code QR CODE : The Transactions of the Korean Institute of Electrical Engineers
  • COPE
  • kcse
  • ํ•œ๊ตญ๊ณผํ•™๊ธฐ์ˆ ๋‹จ์ฒด์ด์—ฐํ•ฉํšŒ
  • ํ•œ๊ตญํ•™์ˆ ์ง€์ธ์šฉ์ƒ‰์ธ
  • Scopus
  • crossref
  • orcid

  1. (Dept. of Control and Automation Engineering, Korea Maritime and Ocean University, Korea.)



Takagi-Sugeno (T-S) fuzzy model, Sampled-data fuzzy observer, $H_โˆž$, Linear matrix inequality, Immeasurable premise.

1. ์„œ ๋ก 

์ตœ๊ทผ ๋งŽ์€ ์ œ์–ด ์‹œ์Šคํ…œ์ด ์ƒ˜ํ”Œ๋ฐ์ดํ„ฐ ์ œ์–ด ์ด๋ก  (1)์— ๊ทผ๊ฑฐํ•˜์—ฌ ์„ค๊ณ„๋˜๊ณ  ์žˆ๋‹ค. ์ƒ˜ํ”Œ๋ฐ์ดํ„ฐ ์ œ์–ด ์ด๋ก ์€ ์—ฐ์†์‹œ๊ฐ„ ์‹œ์Šคํ…œ์„ ๋””์ง€ํ„ธ ์ปดํ“จํ„ฐ์— ์˜ํ•ด ์ œ์–ดํ•˜๊ธฐ ์œ„ํ•ด ์‚ฌ์šฉ๋˜๋Š” ์ด๋ก ์ด๋‹ค. ๊ธฐ์กด์˜ ์—ฐ์†์‹œ๊ฐ„ ํ˜น์€ ์ด์‚ฐ์‹œ๊ฐ„ ์ œ์–ด ์ด๋ก ๊ณผ๋Š” ๋‹ฌ๋ฆฌ ์ƒ˜ํ”Œ๋ฐ์ดํ„ฐ ์ œ์–ด ์ด๋ก ์€ ์—ฐ์†์‹œ๊ฐ„๊ณผ ์ด์‚ฐ์‹œ๊ฐ„ ์ƒํƒœ ๋ณ€์ˆ˜๊ฐ€ ํ˜ผ์žฌ๋˜์–ด ์žˆ๋Š” ๋™์—ญํ•™ ๋ชจ๋ธ์„ ๋‹ค๋ฃจ๊ธฐ ๋•Œ๋ฌธ์— ๊ธฐ์กด๊ณผ๋Š” ๋‹ค๋ฅธ ์ ‘๊ทผ์„ ์š”๊ตฌํ•œ๋‹ค.

ํ•œํŽธ, ๋น„์„ ํ˜• ๋™์—ญํ•™ ์‹œ์Šคํ…œ์„ ์„ ํ˜• ๋ถ€๋ถ„ ์‹œ์Šคํ…œ์˜ ํผ์ง€ํ•ฉ์œผ๋กœ ํ‘œํ˜„ํ•˜๋Š” ํƒ€์นด๊ธฐ-์ˆ˜๊ฒŒ๋…ธ (Takagi-Sugeno, T-S) ํผ์ง€ ๋ชจ๋ธ (2)์— ๋Œ€ํ•œ ์ƒ˜ํ”Œ๋ฐ์ดํ„ฐ ์ œ์–ด ์ด๋ก  ์—ญ์‹œ ํ™œ๋ฐœํžˆ ์—ฐ๊ตฌ๋˜์–ด ์˜ค๊ณ  ์žˆ๋‹ค (5-14). T-S ํผ์ง€ ์‹œ์Šคํ…œ์˜ ์ƒ˜ํ”Œ๋ฐ์ดํ„ฐ ์ œ์–ด ์ด๋ก ์€ ํฌ๊ฒŒ ์ž…๋ ฅ ์ง€์—ฐ ๋ฐฉ๋ฒ• (3-7)๊ณผ ์™„์ „ ์ด์‚ฐํ™” ๋ฐฉ๋ฒ• (8,9)์œผ๋กœ ๋‚˜๋‰˜์–ด ์—ฐ๊ตฌ๋˜๊ณ  ์žˆ๋‹ค. ์ž…๋ ฅ ์ง€์—ฐ ๋ฐฉ๋ฒ•์€ ์ด์‚ฐ์‹œ๊ฐ„ ์ƒํƒœ ๋ณ€์ˆ˜๋ฅผ ๋“ฑ๊ฐ€์˜ ์‹œ๊ฐ„ ์ง€์—ฐ๋œ ์ƒํƒœ ๋ณ€์ˆ˜๋กœ ๋ณ€ํ™˜ํ•˜๊ณ  Lyapunov-Krasovskii functional (LKF)์„ ์ด์šฉํ•˜์—ฌ ์ด์— ๋Œ€ํ•œ ์•ˆ์ •๋„๋ฅผ ๋ถ„์„ํ•œ๋‹ค. ์ตœ๊ทผ์—๋Š” ์‹œ๊ฐ„ ์ข…์† LKF๋ฅผ ์‚ฌ์šฉํ•˜์—ฌ ์‹œ๋ณ€ ์ƒ˜ํ”Œ๋ง ์ฃผ๊ธฐ์— ๋Œ€์‘ํ•˜๋Š” ์•ˆ์ •๋„ ๋ถ„์„ ์ด๋ก ์ด ์—ฐ๊ตฌ๋˜์—ˆ๋‹ค (4). ๋˜ํ•œ (7)์—์„œ๋Š” ์ œ์–ด ๊ทœ์น™์— ์ง€์ˆ˜ ํ•จ์ˆ˜๋ฅผ ํฌํ•จํ•˜์—ฌ ์ง€์ˆ˜ ์•ˆ์ •๋„์˜ ์ถฉ๋ถ„์กฐ๊ฑด์„ ์ˆ˜์น˜์ ์œผ๋กœ ์™„ํ™”ํ•˜๋Š” ์—ฐ๊ตฌ๋„ ์ˆ˜ํ–‰๋˜์–ด ์˜ค๊ณ  ์žˆ๋‹ค. ๋ฐ˜๋ฉด์—, ์ด์‚ฐํ™” ์ ‘๊ทผ๋ฒ•์€ ์ „์ฒด ์ œ์–ด ์‹œ์Šคํ…œ์„ ๋“ฑ๊ฐ€์˜ ์ด์‚ฐํ™” ๋ชจ๋ธ๋กœ ๋ณ€ํ™˜ํ•˜๊ณ , ์ด์‚ฐ Lyapunov ํ•จ์ˆ˜๋ฅผ ๊ธฐ๋ฐ˜์œผ๋กœ ์ด์— ๋Œ€ํ•œ ์•ˆ์ •๋„ ๋ถ„์„์„ ์ˆ˜ํ–‰ํ•œ๋‹ค. (9)์—์„œ๋Š” ์™„์ „ ์ด์‚ฐํ™” ์ ‘๊ทผ๋ฒ•์„ ๊ธฐ๋ฐ˜์œผ๋กœ ์ƒ˜ํ”Œ๋ฐ์ดํ„ฐ ํผ์ง€ ๊ด€์ธก๊ธฐ๋ฅผ ์„ค๊ณ„ํ•˜๋Š” ์—ฐ๊ตฌ๊ฐ€ ์ˆ˜ํ–‰๋˜์—ˆ๋‹ค.

์ผ๋ฐ˜์ ์œผ๋กœ ์ œ์–ด๊ธฐ๋Š” ์ƒํƒœ ๋ณ€์ˆ˜๋ฅผ ๋˜๋จน์ž„ํ•˜์—ฌ ์‹œ์Šคํ…œ์„ ์•ˆ์ •ํ™”ํ•˜๋ฉฐ, ์ด๋Ÿฌํ•œ ์ œ์–ด ๋ฐฉ๋ฒ•์„ ์ƒํƒœ ํ”ผ๋“œ๋ฐฑ ์ œ์–ด๋ผ๊ณ  ํ•œ๋‹ค. T-S ํผ์ง€ ๋ชจ๋ธ์˜ ์•ˆ์ •ํ™”์—๋„ parallel-distributed- compensation (PDC) ๊ธฐ๋ฒ• (12)์„ ๊ธฐ๋ฐ˜์œผ๋กœ ํ•œ ์ƒํƒœ ํ”ผ๋“œ๋ฐฑ ์ œ์–ด๊ฐ€ ๋„๋ฆฌ ์‚ฌ์šฉ๋˜๊ณ  ์žˆ๋‹ค. PDC ์ œ์–ด ๊ธฐ๋ฒ•์„ ์‚ฌ์šฉํ•˜๊ธฐ ์œ„ํ•ด์„œ๋Š” ์‹œ์Šคํ…œ์˜ ์ƒํƒœ ๋ณ€์ˆ˜ ๋ฐ ์ „์ œ ๋ณ€์ˆ˜๋ฅผ ์ •ํ™•ํ•˜๊ฒŒ ์•Œ๊ณ  ์žˆ์–ด์•ผ ํ•˜์ง€๋งŒ, ์ผ๋ฐ˜์ ์œผ๋กœ ๋น„์šฉ์˜ ์ ˆ๊ฐ์ด๋‚˜ ์ธก์ • ๋ถˆ๊ฐ€๋Šฅ ๋“ฑ์˜ ์ด์œ ๋กœ ์ธํ•ด ๋ชจ๋“  ์ƒํƒœ ๋ณ€์ˆ˜์™€ ์ „์ œ ๋ณ€์ˆ˜๋ฅผ ์ •ํ™•ํ•˜๊ฒŒ ์ธก์ •ํ•˜๋Š” ๊ฒƒ์€ ๋ถˆ๊ฐ€๋Šฅํ•˜๋‹ค. ์ด๋Ÿฌํ•œ ๋ฌธ์ œ๋ฅผ ํ•ด๊ฒฐํ•˜๊ณ ์ž ์ƒํƒœ ์ถ”์ • ์—ฐ๊ตฌ๊ฐ€ ํ™œ๋ฐœํžˆ ์ง„ํ–‰๋˜์–ด ์˜ค๊ณ  ์žˆ๋‹ค. T-S ํผ์ง€ ์‹œ์Šคํ…œ์˜ ์ƒํƒœ ์ถ”์ •์—๋Š” ํผ์ง€ ํ•„ํ„ฐ (10)๋‚˜ ํผ์ง€ ๊ด€์ธก๊ธฐ (11)๊ฐ€ ์ผ๋ฐ˜์ ์œผ๋กœ ์‚ฌ์šฉ๋œ๋‹ค. ํผ์ง€ ํ•„ํ„ฐ๋Š” ์ผ๋ฐ˜์ ์œผ๋กœ ์ ๊ทผ ์•ˆ์ •ํ•œ ์‹œ์Šคํ…œ์˜ ์ƒํƒœ ์ถ”์ •์— ์‚ฌ์šฉ๋˜๋‚˜, ์ตœ๊ทผ์—๋Š” ์ง„๋™ํ•˜๋Š” ์‹œ์Šคํ…œ์— ๊ด€ํ•œ ์ ์šฉ ์—ฐ๊ตฌ๋„ ์ˆ˜ํ–‰๋˜์—ˆ๋‹ค.

ํผ์ง€ ๊ด€์ธก๊ธฐ์˜ ์„ค๊ณ„๋Š” ๊ด€์ธก๊ธฐ ์‹œ์Šคํ…œ๊ณผ ์ถ”์ • ๋Œ€์ƒ ์‹œ์Šคํ…œ์˜ ์ƒํƒœ ๋ฒกํ„ฐ์˜ ์˜ค์ฐจ ๋ฒกํ„ฐ์— ๋Œ€ํ•œ ์˜ค์ฐจ ๋™์—ญํ•™์„ ์ˆ˜๋ฆฝํ•˜๊ณ , ์ด๋ฅผ ์ ๊ทผ ์•ˆ์ •ํ™”ํ•˜๋Š” ๊ด€์ธก๊ธฐ ์ด๋“ ํ–‰๋ ฌ์„ ์ฐพ์Œ์œผ๋กœ์จ ์ด๋ฃจ์–ด์ง„๋‹ค. ์ด๋•Œ ํผ์ง€ ์ œ์–ด๊ธฐ์˜ ์ œ์–ด ์ด๋“๊ณผ ๊ด€์ธก๊ธฐ์˜ ์ด๋“์€ ์„œ๋กœ ๋ถ„๋ฆฌ๋˜์–ด ๋™์‹œ์— ์„ค๊ณ„ํ•  ์ˆ˜ ์žˆ์œผ๋ฏ€๋กœ, ์ด๋ฅผ ๋™์‹œ์— ์„ค๊ณ„ ํ•˜๋Š” ๊ด€์ธก๊ธฐ ๊ธฐ๋ฐ˜ ์ œ์–ด๊ธฐ์— ๋Œ€ํ•œ ์—ฐ๊ตฌ๊ฐ€ ์ˆ˜ํ–‰๋˜์—ˆ๋‹ค (12). ํ•œํŽธ, ์ƒํƒœ ์ถ”์ • ๋งŒ์„ ์œ„ํ•œ ํผ์ง€ ๊ด€์ธก๊ธฐ ์—ฐ๊ตฌ๋„ ํ™œ๋ฐœํžˆ ์ˆ˜ํ–‰๋˜์–ด ์˜ค๊ณ  ์žˆ์œผ๋‚˜, ๊ธฐ์กด ์—ฐ๊ตฌ ๋Œ€๋ถ€๋ถ„์€ ์‹œ์Šคํ…œ์˜ ๋ชจ๋ธ์„ ์ •ํ™•ํžˆ ์•Œ๊ณ  ์žˆ์„ ๋•Œ๋ฅผ ๊ฐ€์ •ํ•˜์—ฌ ์—ฐ๊ตฌ๊ฐ€ ์ง„ํ–‰๋˜์—ˆ๊ณ , (13)์—์„œ๋Š” ํผ์ง€ ๊ด€์ธก๊ธฐ์˜ ํ•˜๋“œ์›จ์–ด ๊ตฌํ˜„ ๊ฐ€๋Šฅ์„ฑ์„ ๊ฒ€์ฆํ•˜๋Š” ์—ฐ๊ตฌ๊ฐ€ ์ˆ˜ํ–‰๋˜์—ˆ๋‹ค. ๊ทธ๋Ÿฌ๋‚˜ ์•ž์„  ์—ฐ๊ตฌ๋Š” ๋Œ€๋ถ€๋ถ„ ํผ์ง€ ๊ทœ์น™์˜ ์ „์ œ ๋ณ€์ˆ˜๊ฐ€ ์ธก์ • ๊ฐ€๋Šฅํ•˜๋‹ค๋Š” ๊ฐ€์ •ํ•˜์— ์—ฐ๊ตฌ๋˜์—ˆ๋‹ค๋Š” ๋‹จ์ ์„ ๊ฐ€์ง„๋‹ค. ์ตœ๊ทผ (14)์—์„œ ์ „์ œ ๋ณ€์ˆ˜๋ฅผ ์ธก์ •ํ•  ์ˆ˜ ์—†๋Š” ๋ชจ๋ธ ๋ถˆํ™•์‹ค์„ฑ์„ ํฌํ•จํ•˜๋Š” ์‹œ์Šคํ…œ์˜ ์ƒํƒœ ์ถ”์ •์„ ์œ„ํ•œ ์ƒ˜ํ”Œ๋ฐ์ดํ„ฐ ๊ด€์ธก๊ธฐ ์„ค๊ณ„ ์—ฐ๊ตฌ๊ฐ€ ์ˆ˜ํ–‰๋˜์—ˆ๋‹ค. ์ด๋ฅผ ์œ„ํ•ด ๊ด€์ธก๊ธฐ๊ฐ€ ์‹œ์Šคํ…œ๊ณผ ๋‹ค๋ฅธ ์ „์ œ๋ถ€๋ฅผ ๊ฐ€์ง€๋ฉฐ, ๊ทธ๋กœ ์ธํ•ด ์‹œ์Šคํ…œ๊ณผ ๊ด€์ธก๊ธฐ์˜ ์†Œ์† ํ•จ์ˆ˜๊ฐ€ ๋ถˆ์ผ์น˜๋˜๋Š” ์ƒํ™ฉ์„ ๊ฐ€์ •ํ•˜์—ฌ ์—ฐ๊ตฌํ•˜์˜€๋‹ค. ๊ทธ๋Ÿฌ๋‚˜ ์ด๋Ÿฐ ๋ถˆ์ผ์น˜ ๋ฌธ์ œ๋ฅผ ํ•ด๊ฒฐํ•˜๊ธฐ ์œ„ํ•ด, (15)์—์„œ ์—ฐ๊ตฌ๋œ ์†Œ์† ํ•จ์ˆ˜์˜ ๊ด€๊ณ„๋ฅผ ์ด์šฉํ•œ slack matrix ์กฐ๊ฑด์„ ์ถ”๊ฐ€ํ•จ์œผ๋กœ์จ ์•ˆ์ •ํ™” ์กฐ๊ฑด์ด ๋งŽ์€ ์ˆ˜์˜ ์„ ํ˜• ํ–‰๋ ฌ ๋ถ€๋“ฑ์‹์„ ํฌํ•จํ•˜๋Š” ๋‹จ์ ์„ ๊ฐ€์ง„๋‹ค. ์ด๊ฒƒ์€ ์—ฌ๋Ÿฌ ๋…ผ๋ฌธ์—์„œ๋„ ๋‚˜ํƒ€๋‚˜ ์žˆ๋‹ค (18-24)

์ด๋Ÿฌํ•œ ๋ถ„์„์— ์ฐฉ์•ˆํ•˜์—ฌ, ๋ณธ ๋…ผ๋ฌธ์—์„œ๋Š” ๋ชจ๋ธ ๋ถˆํ™•์‹ค์„ฑ์„ ๊ฐ–๋Š” ๋น„์„ ํ˜• ์‹œ์Šคํ…œ์˜ ์ƒํƒœ๋ฅผ ์ถ”์ •ํ•˜๊ธฐ ์œ„ํ•œ ์ƒ˜ํ”Œ๋ฐ์ดํ„ฐ ํผ์ง€ ๊ด€์ธก๊ธฐ ์„ค๊ณ„ ๊ธฐ๋ฒ•์„ ์ œ์•ˆํ•œ๋‹ค. ํผ์ง€ ์‹œ์Šคํ…œ์˜ ์ „์ œ๋ถ€์—๋Š” ์ธก์ •ํ•  ์ˆ˜ ์—†๋Š” ์ „์ œ ๋ณ€์ˆ˜๊ฐ€ ์กด์žฌํ•˜๋Š” ์ƒํ™ฉ์„ ๊ฐ€์ •ํ•œ๋‹ค. ์ด๋Ÿฌํ•œ ์ธก์ •ํ•  ์ˆ˜ ์—†๋Š” ์ „์ œ ๋ณ€์ˆ˜๋ฅผ ๋‹ค๋ฃจ๊ธฐ ์œ„ํ•ด ๋ณธ ๋…ผ๋ฌธ์—์„œ ๊ด€์ธก๊ธฐ๋Š” ์‹œ์Šคํ…œ๊ณผ IF-THEN ๊ทœ์น™์˜ ์ „์ œ ๋ถ€๋ถ„์„ ๊ณต์œ ํ•˜์ง€ ์•Š์œผ๋ฉฐ, ๊ด€์ธก๊ธฐ์™€ ์‹œ์Šคํ…œ ๊ฐ„์˜ ์˜ค์ฐจ ๋™์—ญํ•™์„ T-S ํผ์ง€ ๋ชจ๋ธ๋กœ ์œ ๋„ํ•œ๋‹ค. ์ธก์ • ๋ถˆ๊ฐ€๋Šฅํ•œ ์ „์ œ ๋ณ€์ˆ˜, ๋ชจ๋ธ ๋ถˆํ™•์‹ค์„ฑ, ์‹œ์Šคํ…œ ์™ธ๋ž€์ด ์ƒํƒœ ์ถ”์ •์— ๋ฏธ์น˜๋Š” ์˜ํ–ฅ์„ ์ตœ์†Œํ™”ํ•˜๊ธฐ ์œ„ํ•ด $H_{\infty} $ ์„ฑ๋Šฅ ๊ธฐ์ค€์„ ์ด์šฉํ•œ๋‹ค. ๋˜ํ•œ, ์•ˆ์ •๋„ ์กฐ๊ฑด์„ ์ˆ˜์น˜์ ์œผ๋กœ ์™„ํ™”ํ•˜๊ธฐ ์œ„ํ•ด ํผ์ง€ Lyapunov ํ•จ์ˆ˜ (16)๋ฅผ ๊ธฐ๋ฐ˜์œผ๋กœ, ์˜ค์ฐจ ๋™์—ญํ•™์˜ ์ ๊ทผ ์•ˆ์ •๋„์™€ $H_{\infty} $ ์กฐ๊ฑด์„ ๋งŒ์กฑํ•˜๋„๋ก ํ•˜๋Š” ์„ ํ˜• ํ–‰๋ ฌ ๋ถ€๋“ฑ์‹ ํ˜•ํƒœ์˜ ์ถฉ๋ถ„์กฐ๊ฑด์„ ์œ ๋„ํ•œ๋‹ค. ๋งˆ์ง€๋ง‰์œผ๋กœ, ์‹œ๋ฎฌ๋ ˆ์ด์…˜ ์˜ˆ์ œ๋ฅผ ํ†ตํ•ด ์ œ์•ˆ๋œ ๋ฐฉ๋ฒ•์˜ ์šฐ์ˆ˜์„ฑ์„ ๊ฒ€์ฆํ•œ๋‹ค.

ํ‘œ๊ธฐ๋ฒ•: ์ž„์˜์˜ ํ–‰๋ ฌ $X$์— ๋Œ€ํ•ด ${sym}\{X\}=X+X^{T}$์ด๋ฉฐ, $col\left\{c_{1,\:}c_{2,\:}\cdots ,\: c_{n}\right\}$์€ $c_{1}$์—์„œ $c_{n}$์„ ์›์†Œ๋กœ ๊ฐ–๋Š” ์—ด๋ฒกํ„ฐ, $diag\left\{d_{1,\:}\right.$$\left. d_{2,\:}\cdots ,\: d_{n}\right\}$์€ $d_{1}$์—์„œ $d_{n}$์„ ์›์†Œ๋กœ ๊ฐ–๋Š” ๋Œ€๊ฐ ํ–‰๋ ฌ์„ ์˜๋ฏธํ•œ๋‹ค.

2. ๋ฌธ์ œ ์ œ๊ธฐ

๋ณธ ๋…ผ๋ฌธ์—์„œ๋Š” ๋‹ค์Œ์˜ ์‹์œผ๋กœ ํ‘œํ˜„๋˜๋Š” T-S ํผ์ง€ ์‹œ์Šคํ…œ์˜ ์ƒํƒœ ์ถ”์ •์„ ๋‹ค๋ฃฌ๋‹ค.

(1)
$\dot x(t)=\sum_{i=1}^{r}w_{i}(z(t))\left\{(A_{i}+\Delta A_{i})x(t)+B_{i}\varpi(t)\right\}$, $y(t)=Cx(t)$,

์—ฌ๊ธฐ์„œ $r\in R_{>0}$์€ ํผ์ง€ ๊ทœ์น™์˜ ์ˆ˜, $x(t)\in R^{n_{X}}$์€ ์‹œ์Šคํ…œ์˜ ์ƒํƒœ ๋ณ€์ˆ˜, $\varpi(t)\in R^{n_{W}}$๋Š” ์‹œ์Šคํ…œ ์™ธ๋ž€, $z(t)\in R^{n_{Z}}$๋Š” ์‹œ์Šคํ…œ์˜ ์ „์ œ ๋ณ€์ˆ˜, $y(t)\in R^{n_{Y}}$๋Š” ์‹œ์Šคํ…œ์˜ ์ถœ๋ ฅ์ด๊ณ , $A_{i}\in R^{n_{X}\times n_{X}}$์™€ $B_{i}\in R^{n_{X}\times n_{W}}$, $C\in R^{n_{Y}\times n_{X}}$๋Š” ์•Œ๊ณ  ์žˆ๋Š” ์‹œ์Šคํ…œ ํ–‰๋ ฌ์ด๋ฉฐ, $\Delta A_{i}\in R^{n_{X}\times n_{X}}$๋Š” ์‹œ์Šคํ…œ ๋ชจ๋ธ์˜ ๋ถˆํ™•์‹ค์„ฑ์„ ๋‚˜ํƒ€๋‚ธ๋‹ค. ๋งˆ์ง€๋ง‰์œผ๋กœ, $w_{i}(z(t))\in[0,\: 1]$์€ $i$๋ฒˆ์งธ ๊ทœ์น™์— ๋Œ€ํ•œ ์†Œ์† ํ•จ์ˆ˜๋กœ ๋‹ค์Œ์„ ๋งŒ์กฑํ•œ๋‹ค.

$\sum_{i=1}^{r}w_{i}(z(t))=1$, $\sum_{i=1}^{r}\dot w_{i}(z(t))=0$ ํ•œํŽธ, (1)์˜ ์ƒํƒœ ๋ณ€์ˆ˜๋ฅผ ์ถ”์ •ํ•˜๊ธฐ ์œ„ํ•œ ์ƒ˜ํ”Œ๋ฐ์ดํ„ฐ ํผ์ง€ ๊ด€์ธก๊ธฐ์˜ ์‹œ์Šคํ…œ ๋ชจ๋ธ์€ ๋‹ค์Œ๊ณผ ๊ฐ™๋‹ค.

(2)
$\dot{\hat{x}}(t)=\sum_{i=1}^{r} m_{i}(\hat{z}(t))\left\{A_{i} \hat{x}(t)+L_{i}\left(y\left(t_{k}\right)-\hat{y}\left(t_{k}\right)\right)\right\},$ $\hat y(t_{k})=C\hat x(t_{k})$,

์—ฌ๊ธฐ์„œ $\hat x(t)\in R^{n_{X}}$๋Š” ๊ด€์ธก๊ธฐ์˜ ์ƒํƒœ ๋ฒกํ„ฐ, $\hat y(t_{k})\in R^{n_{Y}}$๋Š” ๊ด€์ธก๊ธฐ์˜ ์ถœ๋ ฅ, $\hat z(t)\in R^{n_{\hat Z}}$๋Š” ๊ด€์ธก๊ธฐ์˜ ์ „์ œ ๋ณ€์ˆ˜์ด๊ณ , $t_{k}$๋Š” $k$๋ฒˆ์งธ ์ƒ˜ํ”Œ๋ง ์‹œ๊ฐ„์œผ๋กœ ์ƒ˜ํ”Œ๋ง ์ฃผ๊ธฐ $h$์™€ $t_{k}= kh$์˜ ๊ด€๊ณ„๋ฅผ ๊ฐ–๋Š”๋‹ค. $L_{i}\in R^{n_{X}\times n_{Y}}$๋Š” ๊ด€์ธก๊ธฐ์˜ ์ด๋“ ํ–‰๋ ฌ๋กœ ๊ฒฐ์ •๋  ๊ฐ’์ด๋‹ค. ๊ทธ๋ฆฌ๊ณ , $m_{i}(\hat z(t))\in[0,\:1]$์€ ๊ด€์ธก๊ธฐ์˜ ์†Œ์† ํ•จ์ˆ˜๋กœ ์‹œ์Šคํ…œ์˜ ์†Œ์† ํ•จ์ˆ˜์™€ ๋งˆ์ฐฌ๊ฐ€์ง€๋กœ ๋‹ค์Œ์˜ ์กฐ๊ฑด์„ ๋งŒ์กฑํ•œ๋‹ค.

$\sum_{i=1}^{r}m_{i}(\hat z(t))=1$, $\sum_{i=1}^{r}\dot m_{i}(\hat z(t))=0$

ํ‘œํ˜„์„ ๊ฐ„๋‹จํžˆ ํ•˜๊ธฐ ์œ„ํ•ด, ์ž„์˜์˜ ์Šค์นผ๋ผ ํ•จ์ˆ˜ $s_{i}(t)$์™€ ํ–‰๋ ฌ $X_{i}$์— ๋Œ€ํ•ด์„œ ๋‹ค์Œ์˜ ํ‘œ๊ธฐ๋ฒ•์„ ์‚ฌ์šฉํ•œ๋‹ค.

(3)
$\sum_{i=1}^{r}s_{i}(t)X_{i}= X_{s}(t)$.

์ด์ œ ์‹œ์Šคํ…œ๊ณผ ๊ด€์ธก๊ธฐ์˜ ์ถ”์ • ์˜ค์ฐจ์— ๋Œ€ํ•œ ๋™์—ญํ•™์„ ์œ ๋„ํ•˜๊ธฐ ์œ„ํ•ด $e(t):=x(t)-\hat x(t)$๋กœ ์ •์˜ํ•˜๋ฉด, (1)๊ณผ (2)๋กœ๋ถ€ํ„ฐ (3)์˜ ํ‘œ๊ธฐ๋ฒ•์„ ์‚ฌ์šฉํ•˜์—ฌ ๋‹ค์Œ๊ณผ ๊ฐ™์ด ์˜ค์ฐจ ๋™์—ญํ•™์„ ์œ ๋„ํ•  ์ˆ˜ ์žˆ๋‹ค.

(4)
$\dot e(t)=\left\{A_{w}(t)-A_{m}(t)+A_{m}(t)\right\}x(t)+\Delta A_{w}(t)x(t)$ $+B_{w}(t)\varpi(t)-A_{m}(t)\hat x(t)-L_{m}(t)Ce(t_{k})$. $=A_{m}(t)e(t)-L_{m}(t)Ce(t_{k})+ B_{w}(t)\varpi(t)+\Delta(t)x(t)$ $=\left\{A_{m}(t)-L_{m}(t)C\right\}e(t_{k})+A_{m}(t)\overline{e}(t)+B_{w}(t)\varpi(t)$ $+\Delta(t)x(t)$,

์—ฌ๊ธฐ์„œ $\overline{e}(t)= e(t)-e(t_{k})$์ด๊ณ , $\Delta(t)=A_{w}(t)-A_{m}(t)$$+\Delta A_{w}(t)$์ด๋‹ค.

์ฐธ๊ณ  1: ๋ณธ ๋…ผ๋ฌธ์—์„œ๋Š” ์‹œ์Šคํ…œ (1)๊ณผ ๊ด€์ธก๊ธฐ (2)์˜ ์†Œ์† ํ•จ์ˆ˜ ๋ถˆ์ผ์น˜๋ฅผ ํ•ด๊ฒฐํ•˜๊ธฐ ์œ„ํ•ด (4)์—์„œ์™€ ๊ฐ™์ด ๋‘ ์†Œ์† ํ•จ์ˆ˜์˜ ์ฐจ์ด๋ฅผ ์‹œ๋ณ€์˜ ์™ธ๋ž€์œผ๋กœ ์ •์˜ํ–ˆ๋‹ค. ๋”ฐ๋ผ์„œ ๊ธฐ์กด์˜ ๋ฐฉ๋ฒ• (15) ๋ณด๋‹ค ์ ์€ ์ˆ˜์˜ ์„ ํ˜• ํ–‰๋ ฌ ๋ถ€๋“ฑ์‹์œผ๋กœ ์•ˆ์ •ํ™” ์กฐ๊ฑด์„ ์œ ๋„ํ•  ์ˆ˜ ์žˆ๋‹ค.

๋ณธ ๋…ผ๋ฌธ์—์„œ๋Š” ๋‹ค์Œ์˜ ๊ฐ€์ •์„ ๊ธฐ๋ฐ˜์œผ๋กœ ๊ด€์ธก๊ธฐ ์„ค๊ณ„ ์กฐ๊ฑด์„ ์ˆ˜๋ฆฝํ•œ๋‹ค.

๊ฐ€์ • 1: ํผ์ง€ ๊ด€์ธก๊ธฐ์˜ ์†Œ์† ํ•จ์ˆ˜์˜ ๋ฏธ๋ถ„์€ ๋ชจ๋“  ์‹œ๊ฐ„ $t\ge 0$์—์„œ ๋‹ค์Œ์„ ๋งŒ์กฑํ•œ๋‹ค.

(5)
$\left |\dot m_{i}(\hat z(t))\right |\le\phi_{i}$,

์—ฌ๊ธฐ์„œ $\phi_{i}$๋Š” ์–‘์˜ ์Šค์นผ๋ผ์ด๋‹ค.

๊ฐ€์ • 2: ๋ชจ๋“  $i\in\{1,\:2,\:\cdots ,\: r\}$์— ๋Œ€ํ•ด $(A_{i,\:}C)$์€ ๊ฐ€๊ด€์ธก์ด๋‹ค.

๊ฐ€์ • 3: ์‹œ์Šคํ…œ์˜ ์ƒํƒœ ๋ฒกํ„ฐ $x(t)$์™€ ์ „์ œ ๋ณ€์ˆ˜ $z(t)$๋Š” ์ธก์ • ๋ถˆ๊ฐ€๋Šฅํ•˜๋ฉฐ, ์˜ค์ง $t= t_{k}$์—์„œ์˜ ์‹œ์Šคํ…œ ์ถœ๋ ฅ $y(t_{k})$๋งŒ์ด ์ธก์ • ๊ฐ€๋Šฅํ•˜๋‹ค.

๋ณธ ๋…ผ๋ฌธ์˜ ํผ์ง€ ๊ด€์ธก๊ธฐ ์„ค๊ณ„ ๋ฌธ์ œ๋Š” ๋‹ค์Œ๊ณผ ๊ฐ™์ด ์ •๋ฆฌํ•  ์ˆ˜ ์žˆ๋‹ค.

๋ฌธ์ œ 1: ์ฃผ์–ด์ง„ ํผ์ง€ ์‹œ์Šคํ…œ (1)์˜ ์ƒํƒœ ๋ณ€์ˆ˜๋ฅผ ์ถ”์ •ํ•˜๊ธฐ ์œ„ํ•œ ์ƒ˜ํ”Œ๋ฐ์ดํ„ฐ ํผ์ง€ ๊ด€์ธก๊ธฐ (2)์„ ์„ค๊ณ„ํ•˜๊ธฐ ์œ„ํ•ด ๋ฏธ๋ฆฌ ์ •์˜๋œ ์ƒ˜ํ”Œ๋ง ์ฃผ๊ธฐ $h$์— ๋Œ€ํ•ด์„œ ๋‹ค์Œ ๋‘ ์กฐ๊ฑด์„ ๋งŒ์กฑํ•˜๋Š” ๊ด€์ธก๊ธฐ ์ด๋“ ํ–‰๋ ฌ $L_{i}$๋ฅผ ๊ฒฐ์ •ํ•˜๋ผ.

์กฐ๊ฑด 1) $\Delta(t)= 0$, $\varpi(t)=0$์ผ ๋•Œ (4)์˜ ํ‰ํ˜•์ ์ด ์ ๊ทผ ์•ˆ์ •ํ•˜๋‹ค.

์กฐ๊ฑด 2) $e(0)=0$์ผ ๋•Œ ๋‹ค์Œ์˜ $H_{\infty}$ ์„ฑ๋Šฅ์„ ๋งŒ์กฑํ•œ๋‹ค.

(6)
$\int_{0}^{t_{f}}e^{T}(t)e(t)dt\le\gamma^{2}\int_{0}^{t_{f}}W^{T}(t)W(t)dt$,

์—ฌ๊ธฐ์„œ $\gamma >0$์€ ๊ตฌํ•ด์ง€๋Š” ๊ฐ์‡  ๊ณ„์ˆ˜์ด๊ณ , $t_{f}>0$์€ ์ข…๋ฃŒ ์‹œ๊ฐ„์ด๋ฉฐ, $W(t)={col}\{x(t),\:\varpi(t)\}$์ด๋‹ค.

๋งˆ์ง€๋ง‰์œผ๋กœ ์ œ์•ˆํ•˜๋Š” ๋ฐฉ๋ฒ•์˜ ์ฆ๋ช…์„ ์œ„ํ•ด ์‚ฌ์šฉ๋˜๋Š” ๋‹ค์Œ์˜ ๋ณด์กฐ ์ •๋ฆฌ๋ฅผ ์†Œ๊ฐœํ•˜๋ฉฐ ๋ณธ ์žฅ์„ ๋งˆ๋ฌด๋ฆฌํ•œ๋‹ค.

๋ณด์กฐ ์ •๋ฆฌ 1 (17): ์ฃผ์–ด์ง„ ์ ์ ˆํ•œ ์ฐจ์›์˜ ๋ฒกํ„ฐ $\eta(t)$, $\dot e(t)$, ํ–‰๋ ฌ $N_{m}(t)$, $Q$์— ๋Œ€ํ•ด์„œ ๋‹ค์Œ์˜ ๋ถ€๋“ฑ์‹์ด ์„ฑ๋ฆฝํ•œ๋‹ค.

(7)
$-2\eta^{T}(t)N_{m}(t)\int_{t_{k}}^{t}\dot e(\tau)d\tau$$\le h\eta^{T}(t)N_{m}(t)Q^{-1}N_{m}^{T}(t)\eta(t)$ $+\int_{t_{k}}^{t}\dot e^{T}(\tau)Q\dot e(\tau)d\tau$.

3. ์ฃผ์š” ๊ฒฐ๊ณผ

๋ณธ ์žฅ์—์„œ๋Š” ๋ฌธ์ œ 1์— ๋Œ€ํ•œ ์„ค๊ณ„ ์กฐ๊ฑด์„ ์œ ๋„ํ•œ๋‹ค. ์ด๋ฅผ ์œ„ํ•ด ๋‹ค์Œ์˜ ํผ์ง€ LKF๋ฅผ ์ •์˜ํ•œ๋‹ค.

(8)
$V(t)=e^{T}(t)P_{m}(t)e(t)$$+\int_{t_{k}}^{t}(h-t+\tau)\dot e^{T}(\tau)Q\dot e(\tau)d\tau$,

์—ฌ๊ธฐ์„œ $P_{m}(t)=\sum_{i=1}^{r}m_{i}(\hat z(t))P_{i}$, $P_{i}\in R^{n_{X}\times n_{X}}$์™€ $Q\in R^{n_{X}\times n_{X}}$๋Š” ๊ฒฐ์ •๋˜์–ด์งˆ ์–‘๋ถ€ํ˜ธ ๋Œ€์นญ ํ–‰๋ ฌ์ด๋‹ค.

์•ž์„  ๋‚ด์šฉ์„ ์ •๋ฆฌํ•˜์—ฌ ๋ฌธ์ œ 1์˜ ์กฐ๊ฑด์„ ๋งŒ์กฑํ•˜๋Š” ์ถฉ๋ถ„์กฐ๊ฑด์„ ๋‹ค์Œ์˜ ์ •๋ฆฌ์™€ ๊ฐ™์ด ์–ป์„ ์ˆ˜ ์žˆ๋‹ค.

์ •๋ฆฌ 1: ์ฃผ์–ด์ง„ ์ƒ˜ํ”Œ๋ง ์ฃผ๊ธฐ $h$์™€ ์–‘์˜ ์Šค์นผ๋ผ $\alpha$, $\beta$, $\lambda_{B}$, $\lambda_{\Delta}$, $\phi_{i}$์— ๋Œ€ํ•ด ๋‹ค์Œ์˜ ์„ ํ˜• ํ–‰๋ ฌ ๋ถ€๋“ฑ์‹ ์กฐ๊ฑด์„ ๋งŒ์กฑํ•˜๋Š” ๋Œ€์นญ ํ–‰๋ ฌ $0prec P_{i}\in R^{n_{X}\times n_{X}}$, $0prec Q\in R^{n_{X}\times n_{X}}$, $X\in R^{n_{X}\times n_{X}}$, ์ž„์˜์˜ ํ–‰๋ ฌ $N_{1i}\in R^{n_{X}\times n_{X}}$, $N_{2i}\in R^{n_{X}\times n_{X}}$, $N_{3i}\in R^{n_{X}\times n_{X}}$, $M\in R^{n_{X}\times n_{X}}$, $\overline{L}_{i}\in R^{n_{X}\times n_{Y}}$์ด ์กด์žฌํ•˜๋ฉด ์‹(4)์˜ ์˜ค์ฐจ ๋™์—ญํ•™์€ ๋ฌธ์ œ 1์˜ ์กฐ๊ฑด์„ ๋งŒ์กฑํ•œ๋‹ค.

$\quad$$\quad$minimize $\gamma^{2}$

$\quad$$\quad$subject to

(9)
$P_{i}+X SUCC 0$,

(10)
$\begin{bmatrix}\chi_{i}& *& *\\N_{i}^{T}& -\dfrac{1}{h}Q& *\\\widetilde M & 0& -\Gamma\end{bmatrix}PREC 0$,

์—ฌ๊ธฐ์„œ $i\in\{1,\:2,\:\cdots ,\: r\}$,

$\chi_{i}=\chi_{i}^{T}=\left[\chi_{i}^{ab}\right]$, $(a,\:b)\in\{1,\:2,\:3\}\times\{1,\:2,\:3\}$,

$\chi_{i}^{11}=P_{\phi}+{sym}\left\{M^{T}A_{i}-\overline{L}_{i}C\right\}+I$,

$\chi_{i}^{21}=P_{\phi}+\alpha\left\{M^{T}A_{i}-\overline{L}_{i}C\right\}+A_{i}^{T}M +N_{1i}^{T}+I$,

$\chi_{i}^{22}=P_{\phi}+{sym}\left\{\alpha M^{T}A_{i}+N_{2i}\right\}+I$,

$\chi_{i}^{31}=P_{i}+\beta\left\{M^{T}A_{i}-\overline{L}_{i}C\right\}- M$,

$\chi_{i}^{32}=P_{i}+ N_{3i}+\beta M^{T}A_{i}-\alpha M$,

$\chi_{i}^{33}= h Q -\beta\left(M+M^{T}\right)$,

$\widetilde M =\begin{bmatrix}M&\alpha M&\beta M\\M&\alpha M&\beta M\end{bmatrix}$, $\Gamma =\begin{bmatrix}-\dfrac{\gamma^{2}}{\lambda_{B}}I& 0\\0& -\dfrac{\gamma^{2}}{\lambda_{\Delta}}\end{bmatrix}$,

$N_{i}=\begin{bmatrix}N_{1i}& N_{2i}& N_{3i}\end{bmatrix}$์ด๋‹ค. ๊ทธ๋ฆฌ๊ณ  ๊ด€์ธก๊ธฐ ์ด๋“ ํ–‰๋ ฌ์€ $L_{i}=M^{-T}\overline{L}_{i}$๋กœ๋ถ€ํ„ฐ ์–ป๊ฒŒ ๋œ๋‹ค.

์ฆ๋ช…: ์‹(8)์˜ ํผ์ง€ LKF์˜ ์‹œ๊ฐ„ ๋ฏธ๋ถ„์€ ๋‹ค์Œ๊ณผ ๊ฐ™๋‹ค.

(11)
$\dot V(t)=2\dot e^{T}(t)P_{m}(t)e(t)+\sum_{i=1}^{r}\dot m_{i}(\hat z(t))e^{T}(t)P_{i}e(t)$ $+h\dot e^{T}(t)Q\dot e(t)-\int_{t_{k}}^{t}\dot e^{T}(\tau)Q\dot e(\tau)d\tau$.

ํ•œํŽธ, ์†Œ์† ํ•จ์ˆ˜์˜ ์กฐ๊ฑด์œผ๋กœ๋ถ€ํ„ฐ ์ž„์˜์˜ ๋Œ€์นญ ํ–‰๋ ฌ $X$์— ๋Œ€ํ•ด ๋‹ค์Œ์ด ์„ฑ๋ฆฝํ•จ์„ ์•Œ ์ˆ˜ ์žˆ๋‹ค.

(12)
$\sum_{i=1}^{r}\dot m_{i}(\hat z(t))X = 0$.

์‹(11)์— (12)์„ ๋”ํ•˜๊ณ , (5)์˜ ์กฐ๊ฑด์„ ์ ์šฉํ•˜๋ฉด, ๋‹ค์Œ์„ ์–ป๋Š”๋‹ค.

(13)
$\dot V(t)\le 2\dot e^{T}(t)P_{m}(t)e(t)+e^{T}(t)P_{\phi}e(t)+h\dot e^{T}(t)Q\dot e(t)$ $-\int_{t_{k}}^{t}\dot e^{T}(\tau)Q\dot e(\tau)d\tau$,

์—ฌ๊ธฐ์„œ $P_{\phi}=\sum_{i=1}^{r}\phi_{i}(P_{i}+ X)$์ด๋‹ค.

ํ•œํŽธ, ์ž„์˜์˜ ํ–‰๋ ฌ $N_{m}(t)=\begin{bmatrix}N_{1m}(t)& N_{2m}(t)& N_{3m}(t)\end{bmatrix}$์™€ $M$, ์–‘์˜ ์Šค์นผ๋ผ $\alpha$, $\beta$์— ๋Œ€ํ•ด ๋‹ค์Œ์ด ์„ฑ๋ฆฝํ•จ์€ ์ž๋ช…ํ•˜๋‹ค.

(14)
$2\eta^{T}(t)N_{m}(t)\left\{\overline{e}(t)-\int_{t_{k}}^{t}\dot e(\tau)d\tau\right\}=0$,

(15)
$2\left\{M e(t_{k})+\alpha M\overline{e}(t)+\beta M\dot e(t)\right\}^{T}$ $\times\left\{-\dot e(t)+\overline{A}_{m}(t)e(t_{k})+A_{m}(t)\overline{e}(t)\right.$ $\left. +B_{w}(t)\varpi(t)+\Delta(t)x(t)\right\}=0$,

์—ฌ๊ธฐ์„œ $\overline{A}_{m}(t)= A_{m}(t)+ L_{m}(t)C$, $\eta(t)= col\left\{e(t_{k}),\:\overline{e}(t)\right .$$,\:\dot e(t)\}$์ด๋‹ค.

์‹(13)์— (14)์™€ (15)๋ฅผ ๋”ํ•˜๊ณ , (7)์˜ ๋ถ€๋“ฑ์‹์„ ์ ์šฉํ•˜๋ฉด, $\dot V(t)$์— ๋Œ€ํ•ด ๋‹ค์Œ์ด ์„ฑ๋ฆฝํ•œ๋‹ค.

(16)
$\dot V(t)\le 2\dot e^{T}(t)P_{m}(t)\left\{e(t_{k})+\overline{e}(t)\right\}$ $+\left\{e(t_{k})+\overline{e}(t)\right\}^{T}P_{\phi}\left\{e(t_{k})+\overline{e}(t)\right\}$$+h\dot e^{T}(t)Q\dot e(t)$ $+2\eta^{T}(t)N_{m}(t)\overline{e}(t)+h\eta^{T}(t)N_{m}(t)Q^{-1}N_{m}^{T}(t)\eta(t)$ $+2\left\{Me(t_{k})+\alpha M\overline{e}(t)+\beta M\dot e(t)\right\}^{T}$ $\times\left\{-\dot e(t)+\overline{A}_{m}(t)e(t_{k})+A_{m}\overline{e}(t)\right\}$ $+2\left\{Me(t_{k})+\alpha M\overline{e}(t)+\beta M\dot e(t)\right\}^{T}$ $\times\left\{B_{w}(t)\varpi(t)+\Delta(t)x(t)\right\}$,

์—ฌ๊ธฐ์„œ $e(t)= e(t_{k})+\overline{e}(t)$๋ฅผ ์‚ฌ์šฉํ–ˆ๋‹ค.

์ด์ œ ์‹(6)์˜ $H_{\infty}$ ์กฐ๊ฑด์„ ๋ณด์žฅํ•˜๊ธฐ ์œ„ํ•ด, ๋‹ค์Œ ์‹์„ ๊ณ ๋ คํ•˜์ž.

(17)
$\dot e^{T}(t)\dot e(t)-\gamma^{2}W^{T}(t)W(t)\le 0$.

์‹(16)๊ณผ (17)๋กœ๋ถ€ํ„ฐ ๋‹ค์Œ์„ ์–ป์„ ์ˆ˜ ์žˆ๋‹ค.

(18)
$\dot\upsilon(t):=\dot V(t)+\dot e^{T}(t)\dot e(t)-\gamma^{2}W^{T}(t)W(t)$ $\le\eta^{T}(t)\chi_{m}(t)\eta(t)+h\eta^{T}(t)N_{m}(t)Q^{-1}N_{m}^{T}(t)\eta(t)$ $+2\left\{Me(t_{k})+\alpha M\overline{e}(t)+\beta M\dot e(t)\right\}^{T}$ $\times\left\{B_{w}(t)\varpi(t)+\Delta(t)x(t)\right\}$ $-\gamma^{2}\varpi^{T}(t)\varpi(t)-\gamma^{2}x^{T}(t)x(t)$,

์—ฌ๊ธฐ์„œ $\chi_{m}(t)=\chi_{m}^{T}(t)=\left[\chi_{m}^{ab}(t)\right]$, $(a,\:b)\in\{1,\:2,\:3\}\times\{1,\:2,\:3\}$,

$\chi_{m}^{11}(t)=P_{\phi}+I+{sym}\left\{M^{T}\overline{A}_{m}(t)\right\}$,

$\chi_{m}^{21}(t)=P_{\phi}+\alpha M^{T}\overline{A}_{m}(t)+A_{m}^{T}(t)M+N_{1m}^{T}(t)+I$,

$\chi_{m}^{22}(t)=P_{\phi}+{sym}\left\{\alpha M^{T}A_{m}(t)+ N_{2m}(t)\right\}$,

$\chi_{m}^{31}(t)=P_{m}(t)+\beta M^{T}\overline{A}_{m}(t)-M$,

$\chi_{m}^{32}(t)= P_{m}(t)+N_{3m}(t)+\beta M^{T}A_{m}(t)-\alpha M$,

$\chi_{m}^{33}(t)=h Q-\beta(M+M^{T})$์ด๋‹ค.

ํ•œํŽธ, ๋‹ค์Œ์ด ์„ฑ๋ฆฝํ•จ์€ ์ž๋ช…ํ•˜๋‹ค.

(19)
$X^{T}Y +Y^{T}X\le\sigma X^{T}X +\sigma^{-1}Y^{T}Y$,

์—ฌ๊ธฐ์„œ $X$์™€ $Y$๋Š” ์ž„์˜์˜ ํ–‰๋ ฌ์ด๊ณ , $\sigma$๋Š” ์–‘์˜ ์Šค์นผ๋ผ๋‹ค.

์‹(18)์˜ ๋งˆ์ง€๋ง‰ ์‹์— (19)์˜ ๋ถ€๋“ฑ์‹์„ ์ ์šฉํ•˜๋ฉด, ๋‹ค์Œ์„ ์–ป์„ ์ˆ˜ ์žˆ๋‹ค.

(20)

$2\left\{Me(t_{k})+\alpha M\overline{e}(t)+\beta M\dot e(t)\right\}^{T}$ $\times\left\{B_{w}(t)\varpi(t)+\Delta(t)x(t)\right\}$ $-\gamma^{2}\varpi^{T}(t)\varpi(t)-\gamma^{2}x^{T}(t)x(t)$

$\le$$\left\{Me(t_{k})+\alpha M\overline{e}(t)+\beta M\dot e(t)\right\}^{T}$$\left(\sigma_{1}I+\sigma_{2}I\right)$ $\times\left\{Me(t_{k})+\alpha M\overline{e}(t)+\beta M\dot e(t)\right\}$ $+\varpi^{T}(t)\left\{\sigma_{1}^{-1}B_{w}^{T}(t)B_{w}(t)-\gamma^{2}I\right\}\varpi(t)$ $+x^{T}(t)\left\{\sigma_{2}^{-1}\Delta^{T}(t)\Delta(t)-\gamma^{2}I\right\}x(t)$

$=\eta^{T}(t)\widetilde M^{T}\Sigma\widetilde M\eta(t)$$+\varpi^{T}(t)\left\{\sigma_{1}^{-1}B_{w}^{T}(t)B_{w}(t)-\gamma^{2}I\right\}\varpi(t)$ $+x^{T}(t)\left\{\sigma_{2}^{-1}\Delta^{T}(t)\Delta(t)-\gamma^{2}I\right\}x(t)$,

์—ฌ๊ธฐ์„œ $\Sigma ={diag}\left\{\sigma_{1}I,\:\sigma_{2}I\right\}$์ด๋‹ค.

์ด์ œ (20)์˜ ๊ฒฐ๊ณผ๋ฅผ (18)์— ์ ์šฉํ•˜๋ฉด,

(21)
$\dot\upsilon(t)\le\eta^{T}(t)\left\{\chi_{m}(t)+h N_{m}(t)Q^{-1}N_{m}^{T}(t)+\widetilde M^{T}\Sigma\widetilde M\right\}\eta(t)$ $+\varpi^{T}(t)\left\{\sigma_{1}^{-1}B_{w}^{T}(t)B_{w}(t)-\gamma^{2}I\right\}\varpi(t)$ $+x^{T}(t)\left\{\sigma_{2}^{-1}\Delta^{T}(t)\Delta(t)-\gamma^{2}I\right\}x(t)$ .

๋”ฐ๋ผ์„œ $\dot\upsilon(t)\le 0$๋ฅผ ๋ณด์žฅํ•˜๊ธฐ ์œ„ํ•ด์„œ๋Š” ๋‹ค์Œ์˜ ์กฐ๊ฑด์ด ๋งŒ์กฑํ•˜์—ฌ์•ผ ํ•จ์„ ์•Œ ์ˆ˜ ์žˆ๋‹ค.

(22)
$\chi_{m}(t)+h N_{m}(t)Q^{-1}N_{m}^{T}(t)+\widetilde M^{T}\Sigma\widetilde M PREC 0$,

(23)
$\sigma_{1}^{-1}B_{w}^{T}(t)B_{w}(t)-\gamma^{2}I = 0$,

(24)
$\sigma_{2}^{-1}\Delta^{T}(t)\Delta(t)-\gamma^{2}I = 0$.

๋ชจ๋“  ์‹œ๊ฐ„ $t\ge 0$์—์„œ $B_{w}^{T}(t)B_{w}(t)$์™€ $\Delta^{T}(t)\Delta(t)$์˜ ์ตœ๋Œ€ ๊ณ ์œณ๊ฐ’์„ ๊ฐ๊ฐ $\lambda_{B}$์™€ $\lambda_{\Delta}$๋ผ๊ณ  ํ•˜๋ฉด, (23)๊ณผ (24)๋กœ๋ถ€ํ„ฐ,

(25)
$\sigma_{1}^{-1}=\dfrac{\gamma^{2}}{\lambda_{B}}$, $\sigma_{2}^{-1}=\dfrac{\gamma^{2}}{\lambda_{\Delta}}$

๋ฅผ ์–ป์„ ์ˆ˜ ์žˆ๋‹ค.

์ด์ œ (22)์— Schur complement๋ฅผ ์ ์šฉํ•˜๋ฉด,

(26)
$\begin{bmatrix}\chi_{m}(t)& *& *\\N_{m}^{T}(t)& -\dfrac{1}{h}Q& *\\\widetilde M & 0& -\Sigma^{-1}\end{bmatrix}PREC 0$

์„ ์–ป์„ ์ˆ˜ ์žˆ์œผ๋ฉฐ, ๋งˆ์ง€๋ง‰์œผ๋กœ $\Sigma^{-1}$์— (25)๋ฅผ ๋Œ€์ž…ํ•˜๋ฉด, $\Sigma^{-1}=\Gamma$์ž„์„ ์•Œ ์ˆ˜ ์žˆ๋‹ค. ๋”ฐ๋ผ์„œ (26)์œผ๋กœ๋ถ€ํ„ฐ (10)์˜ ์„ ํ˜• ํ–‰๋ ฌ ๋ถ€๋“ฑ์‹์„ ์–ป์„ ์ˆ˜ ์žˆ๋‹ค.

์ด๋กœ์จ ์ •๋ฆฌ 1์˜ ์„ ํ˜• ํ–‰๋ ฌ ๋ถ€๋“ฑ์‹์„ ๋งŒ์กฑํ•˜๋Š” ํ•ด๋ฅผ ์ฐพ์„ ์ˆ˜ ์žˆ๋‹ค๋ฉด, $\dot\upsilon(t)\le 0$์ด ๋ณด์žฅ๋˜๋ฏ€๋กœ ๋ฌธ์ œ 1์˜ ์กฐ๊ฑด์„ ๋งŒ์กฑํ•˜๋Š” ์ƒ˜ํ”Œ๋ฐ์ดํ„ฐ ํผ์ง€ ๊ด€์ธก๊ธฐ๋ฅผ ์„ค๊ณ„ํ•  ์ˆ˜ ์žˆ์Œ์„ ์•Œ ์ˆ˜ ์žˆ๋‹ค. ์ด๊ฒƒ์œผ๋กœ ์ฆ๋ช…์„ ๋งˆ๋ฌด๋ฆฌํ•œ๋‹ค. โ– 

์ฐธ๊ณ  2: ์ •๋ฆฌ 1์€ ์ธก์ • ๋ถˆ๊ฐ€๋Šฅํ•œ ์ „์ œ ๋ณ€์ˆ˜๋ฅผ ๊ฐ–๋Š” ๋ถˆํ™•์‹คํ•œ ์‹œ์Šคํ…œ์˜ ์ƒํƒœ ์ถ”์ •์„ ์œ„ํ•œ ์„ ํ˜• ํ–‰๋ ฌ ๋ถ€๋“ฑ์‹ ๊ธฐ๋ฐ˜์˜ ์ƒ˜ํ”Œ๋ฐ์ดํ„ฐ ํผ์ง€ ๊ด€์ธก๊ธฐ ์„ค๊ณ„ ์กฐ๊ฑด์„ ์ œ๊ณตํ•œ๋‹ค. ์ •๋ฆฌ 1์—์„œ๋Š” ์†Œ์† ํ•จ์ˆ˜์˜ ๋ถˆ์ผ์น˜ ๋ฌธ์ œ๋ฅผ $H_{\infty}$ ์กฐ๊ฑด์„ ์‚ฌ์šฉํ•˜์—ฌ ํ•ด๊ฒฐํ•˜์—ฌ ๊ธฐ์กด์˜ ๋ฐฉ๋ฒ•์— ๋น„ํ•ด ์„ ํ˜• ํ–‰๋ ฌ ๋ถ€๋“ฑ์‹์˜ ์ˆ˜๋ฅผ ์ค„์ผ ์ˆ˜ ์žˆ๋‹ค. ๋˜ํ•œ, ํผ์ง€ LKF๋ฅผ ์‚ฌ์šฉํ•˜์—ฌ ๊ธฐ์กด ์—ฐ๊ตฌ๋ณด๋‹ค ์ˆ˜์น˜์ ์œผ๋กœ ์™„ํ™”๋œ ์กฐ๊ฑด์„ ์ œ๊ณตํ•˜๋ฉฐ, ์ด๋ฅผ ํ†ตํ•ด ๋”์šฑ ๊ฐœ์„ ๋œ $H_{\infty}$ ์„ฑ๋Šฅ์„ ์–ป์„ ์ˆ˜ ์žˆ๋‹ค.

4. ์‹œ๋ฎฌ๋ ˆ์ด์…˜ ์˜ˆ์ œ

๋ณธ ์žฅ์—์„œ๋Š” ์ œ์•ˆํ•˜๋Š” ๋ฐฉ๋ฒ•์˜ ํƒ€๋‹น์„ฑ์„ ๊ฒ€์ฆํ•˜๊ธฐ ์œ„ํ•ด ๋‹ค์Œ์˜ ์Šคํ”„๋ง ์‹œ์Šคํ…œ์— ๋Œ€ํ•œ ์ƒ˜ํ”Œ๋ฐ์ดํ„ฐ ํผ์ง€ ๊ด€์ธก๊ธฐ๋ฅผ ์„ค๊ณ„ํ•œ๋‹ค (12).

(27)
$\ddot x(t)= 0.01x(t)+ 0.67x^{3}(t)$, $y(t_{k})= x(t_{k})$.

์œ„์˜ ๋น„์„ ํ˜• ๋™์—ญํ•™ ๋ชจ๋ธ์„ ์‹(1)์˜ T-S ํผ์ง€ ๋ชจ๋ธ๋กœ ํ‘œํ˜„ํ•˜๋ฉด, ๋‹ค์Œ๊ณผ ๊ฐ™๋‹ค.

$\dot x(t)=\sum_{i=1}^{r}w_{i}(z(t))\left\{(A_{i}+\Delta A_{i})x(t)+B_{i}\varpi(t)\right\}$,

์—ฌ๊ธฐ์„œ $r=2 $, $z(t)= x_{1}(t)$, $w_{1}(z(t))= 1-x_{1}^{2}(t)$, $w_{2}(z(t))$$= x_{1}^{2}(t)$, $A_{1}=\begin{bmatrix}0& 1\\-0.01& 0\end{bmatrix}$, $A_{2}=\begin{bmatrix}0& 1\\-0.68& 0\end{bmatrix}$, $B_{1}=\begin{bmatrix}0.1\\0\end{bmatrix}$, $B_{2}=\begin{bmatrix}0.1\\0\end{bmatrix}$, $C=\begin{bmatrix}1& 0\end{bmatrix}$,์ด๋‹ค.

$\Delta A$๋Š” ์‹œ์Šคํ…œ ๋ชจ๋ธ์˜ ๋ถˆํ™•์‹ค์„ฑ์„ ๋‚˜ํƒ€๋‚ด๋Š” ํ–‰๋ ฌ๋กœ ์ •ํ™•ํ•œ ๊ฐ’์„ ์•Œ์ง€ ๋ชปํ•˜๋Š” ํ–‰๋ ฌ์ด์ง€๋งŒ, ์‹œ๋ฎฌ๋ ˆ์ด์…˜์„ ์œ„ํ•ด ๋ณธ ๋…ผ๋ฌธ์—์„œ๋Š” $\Delta A=\begin{bmatrix}0& 0\\0.05& 0\end{bmatrix}$์œผ๋กœ ์„ค์ •ํ•˜์˜€๋‹ค.

ํ•œํŽธ, ์‹(2)์˜ ํผ์ง€ ๊ด€์ธก๊ธฐ์˜ ์ „์ œ ๋ณ€์ˆ˜์™€ ์†Œ์† ํ•จ์ˆ˜๋Š” ๋‹ค์Œ๊ณผ ๊ฐ™์ด ์„ค์ •ํ•˜์˜€๋‹ค.

$\hat z(t)=\hat x_{1}(t)$, $m_{1}(\hat z(t))=0.6-0.5\hat x_{1}^{2}(t)$,

$m_{2}(\hat z(t))=1$$-m_{1}(\hat z(t))$.

๋˜ํ•œ, $h=0.05$, $\varpi(t)=0.1 e^{-0.5 t}$, $x(0)= col\{0.5,\: 0.3\}$, $\hat x(0)={col}\{0,\: 0\}$, $\alpha =1$, $\beta =1$, $\lambda_{B}= 0.1$, $\lambda_{\Delta}=10$์œผ๋กœ ์„ค์ •ํ•˜๊ณ  ์ •๋ฆฌ 1์˜ ์„ ํ˜• ํ–‰๋ ฌ ๋ถ€๋“ฑ์‹์„ ํ†ตํ•ด ๋‹ค์Œ์„ ์–ป์„ ์ˆ˜ ์žˆ๋‹ค.

$L_{1}=\begin{bmatrix}11.1673\\21.9618\end{bmatrix}$, $L_{2}=\begin{bmatrix}11.2156\\21.3962\end{bmatrix}$.

๋™์ผํ•œ ์กฐ๊ฑด์—์„œ (14)์˜ ์ •๋ฆฌ 2๋ฅผ ํ†ตํ•ด ์„ค๊ณ„ํ•œ ๊ด€์ธก๊ธฐ ์ด๋“ ํ–‰๋ ฌ์€ ๋‹ค์Œ๊ณผ ๊ฐ™๋‹ค.

$L_{1}^{[14]}=\begin{bmatrix}9.8832\\13.1487\end{bmatrix}$, $L_{2}^{[14]}=\begin{bmatrix}9.8864\\12.5037\end{bmatrix}$.

๊ทธ๋ฆผ 1๊ณผ 2์— ์‹œ์Šคํ…œ์˜ ์ƒํƒœ ๋ณ€์ˆ˜์™€ ์ œ์•ˆ๋œ ๋ฐฉ๋ฒ•๊ณผ (14)์˜ ๋ฐฉ๋ฒ•์„ ํ†ตํ•ด ์ถ”์ •๋œ ์ƒํƒœ ๋ณ€์ˆ˜์˜ ์‹œ๊ฐ„ ์‘๋‹ต์„ ๋‚˜ํƒ€๋‚ด์—ˆ๋‹ค.

๊ทธ๋ฆผ 1 $x_{1}(t)$์™€ $\hat x_{1}(t)$์˜ ์‹œ๊ฐ„ ์‘๋‹ต ๊ทธ๋ž˜ํ”„.

Fig. 1 The time responses of $x_{1}(t)$ and $\hat x_{1}(t)$.

../../Resources/kiee/KIEE.2019.68.11.1403/fig1.png

๊ทธ๋ฆผ 2 $x_{2}(t)$์™€ $\hat x_{2}(t)$์˜ ์‹œ๊ฐ„ ์‘๋‹ต ๊ทธ๋ž˜ํ”„.

Fig. 2 The time responses of $x_{2}(t)$ and $\hat x_{2}(t)$.

../../Resources/kiee/KIEE.2019.68.11.1403/fig2.png

๊ทธ๋ฆผ์„ ํ†ตํ•ด ์ œ์•ˆํ•˜๋Š” ๋ฐฉ๋ฒ•์ด ์‹œ์Šคํ…œ์˜ ๋ถˆํ™•์‹ค์„ฑ๊ณผ ์†Œ์† ํ•จ์ˆ˜์˜ ๋ถˆ์ผ์น˜๊ฐ€ ์žˆ๋Š” ํ™˜๊ฒฝ์—์„œ ๊ธฐ์กด์˜ ๋ฐฉ๋ฒ•๋ณด๋‹ค ๋‚˜์€ ์ƒํƒœ ์ถ”์ • ์„ฑ๋Šฅ์„ ์ œ๊ณตํ•˜๋Š” ๊ฒƒ์„ ์•Œ ์ˆ˜ ์žˆ์œผ๋ฉฐ, ์ด๋Š” ํผ์ง€ LKF์˜ ์‚ฌ์šฉ๊ณผ $H_{\infty}$ ์„ฑ๋Šฅ ์ง€ํ‘œ๋ฅผ ์ด์šฉํ•ด ๊ธฐ์กด์˜ ๋ฐฉ๋ฒ•์„ ์ˆ˜์น˜์ ์œผ๋กœ ์™„ํ™”ํ–ˆ๊ธฐ ๋•Œ๋ฌธ์œผ๋กœ ์ƒ๊ฐํ•  ์ˆ˜ ์žˆ๋‹ค.

5. ๊ฒฐ ๋ก 

๋ณธ ๋…ผ๋ฌธ์—์„œ๋Š” ๋ชจ๋ธ ๋ถˆํ™•์‹ค์„ฑ์„ ๊ฐ–๋Š” ๋น„์„ ํ˜• ์‹œ์Šคํ…œ์˜ ์ƒํƒœ๋ฅผ ์ถ”์ •ํ•˜๊ธฐ ์œ„ํ•œ ์ƒ˜ํ”Œ ๋ฐ์ดํ„ฐ ํผ์ง€ ๊ด€์ธก๊ธฐ ์„ค๊ณ„ ๊ธฐ๋ฒ•์„ ์ œ์•ˆํ–ˆ๋‹ค. ํผ์ง€ ์‹œ์Šคํ…œ์˜ IF-THEN ๊ทœ์น™์—๋Š” ์ธก์ •ํ•  ์ˆ˜ ์—†๋Š” ์ „์ œ ๋ณ€์ˆ˜๊ฐ€ ํฌํ•จ๋˜์–ด ์žˆ๋‹ค๊ณ  ๊ฐ€์ •๋˜์—ˆ๋‹ค. ๋ณธ ๋…ผ๋ฌธ์—์„œ ๊ด€์ธก๊ธฐ๋Š” ์ธก์ •ํ•  ์ˆ˜ ์—†๋Š” ์ „์ œ ๋ณ€์ˆ˜๋ฅผ ๋‹ค๋ฃจ๊ธฐ ์œ„ํ•ด ์‹œ์Šคํ…œ๊ณผ IF-THEN ๊ทœ์น™์˜ ์ „์ œ ๋ถ€๋ถ„์„ ๊ณต์œ ํ•˜์ง€ ์•Š๋Š” ๊ฒƒ์œผ๋กœ ๊ฐ€์ •ํ•œ๋‹ค. ๊ทธ ํ›„, ๊ด€์ธก๊ธฐ์™€ ๋ชจ๋ธ ๋ถˆํ™•์‹ค์„ฑ์„ ํฌํ•จํ•œ ์‹œ์Šคํ…œ ๊ฐ„์˜ ์˜ค์ฐจ ๋™์—ญํ•™์„ T-S ํผ์ง€ ๋ชจ๋ธ๋กœ ํ‘œํ˜„ํ–ˆ๋‹ค. ์ผ์น˜ํ•˜์ง€ ์•Š๋Š” ์ „์ œ ๋ถ€๋ถ„์˜ ์˜ํ–ฅ, ๋ชจ๋ธ ๋ถˆํ™•์‹ค์„ฑ ๋ฐ ์™ธ๋ž€์ด ์ƒํƒœ ์ถ”์ •์— ๋ฏธ์น˜๋Š” ์˜ํ–ฅ์„ ์ตœ์†Œํ™”ํ•˜๊ธฐ ์œ„ํ•ด $H_{\infty} $ ์„ฑ๋Šฅ ๊ธฐ์ค€์ด ์ •์˜๋˜์—ˆ๋‹ค. ํผ์ง€ ๋ฆฌ์•„ํ‘ธ๋…ธํ”„ ํ•จ์ˆ˜๋ฅผ ๊ธฐ๋ฐ˜์œผ๋กœ, ์ƒํƒœ ์ถ”์ • ์˜ค์ฐจ ๋™์—ญํ•™์ด ์ ๊ทผ ์•ˆ์ •๋˜๊ณ  $H_{\infty} $ ์กฐ๊ฑด์„ ๋งŒ์กฑํ•˜๋Š” ๊ฒƒ์„ ๋ณด์žฅํ•˜๋Š” ์„ ํ˜• ํ–‰๋ ฌ ๋ถ€๋“ฑ์‹ ํ˜•ํƒœ์˜ ์ถฉ๋ถ„์กฐ๊ฑด์„ ์œ ๋„ํ–ˆ๋‹ค. ๋งˆ์ง€๋ง‰์œผ๋กœ, ์‹œ๋ฎฌ๋ ˆ์ด์…˜ ์˜ˆ์ œ๋ฅผ ํ†ตํ•ด ์ œ์•ˆ๋œ ๋ฐฉ๋ฒ•์˜ ์šฐ์ˆ˜์„ฑ์„ ๊ฒ€์ฆํ–ˆ๋‹ค.

Acknowledgements

This work was partially supported by the Basic Science Research Program through the National Research Foundation of Korea (NRF) funded by the Ministry of Education (NRF-2016R1A6A1A03013567, NRF-2018R1A2A2A14023632, NRF-2019R1G1A1099286).

References

1 
T. Chen, B. A. Francis, 2012, Optimial sampled-data control systems, Springer Science & Business MediaGoogle Search
2 
T. Takagi, M. Sugeno, 1985, Fuzzy identification of systems and its applications to modeling and control, IEEE Transactions on System, Man, and Cybernetics, Vol. SMC-15, No. 1, pp. 116-132DOI
3 
L. Shanmugam, Y. H. Joo, 2019, Design of Interval Type-2 Fuzzy-based Sampled-Data Controller for Nonlinear Systems Using Novel Fuzzy Lyapunov Functional and Its Application to PMSM, IEEE Transactions on Systems, Man, and Cybernetics: Systems. Online Publication.DOI
4 
E. Fridman, 2010, A refiend input delay approach to sampleddata control,, Automatic, Vol. 46, pp. 421-427DOI
5 
H. S. Kim, J. B. Park, Y. H. Joo, 2018, A fuzzy LyapunoKrasovskii functional approach to sampled-data outputfeedback stabilization of polynomial fuzzy systems,, IEEE Transactions on Fuzzy Systems, Vol. 26, No. 1, pp. 366-373DOI
6 
H. S. Kim, J. B. Park, Y. H. Joo, 2019, Fuzzy-model-based sampled-data chaotic synchronisation under the input constraints consideration, IET Control Theory and Applications, Vol. 13, No. 2, pp. 288-296DOI
7 
Y. Wang, H. R. Karimi, H. K. Lam, H. Shen, 2018, An improved result on exponential stabilization of sampleddata fuzzy systems, IEEE Transactions on Fuzzy Systems, Vol. 26, No. 6, pp. 3875-3883DOI
8 
D. W. Lee, H. J. Lee, M. Tomizuka, 2010, Fuzzy stabilization of nonlinear systems under sampled-data feedback: An exact discrete-time model approach, IEEE Transactions on Fuzzy Systems, Vol. 18, No. 2, pp. 251-260DOI
9 
H. J. Lee, 2012, Sampled-data observer-based output-feedback fuzzy stabilization of nonlinear systems: Exact discretetime design approach, Fuzzy Sets and Systems, Vol. 201, No. , pp. 20-39DOI
10 
H. J. Kim, J. B. Park, Y. H. Joo, 2018, Decentralized $H_โˆž$ fuzzy filter for nonlinear large-scale sampled-data systems with uncertain interconnections, Fuzzy Sets and Systems, Vol. 344, No. , pp. 145-162DOI
11 
G. B. Koo, J. B. Park, Y. H. Joo, 2016, Decentralized sampled-data fuzzy observer design for nonlinear interconnected systems, IEEE Transactions on Fuzzy Systems, Vol. 24, No. 3, pp. 661-674DOI
12 
K. Tanaka, T. Ikeda, H. O. Wang, 1998, Fuzzy regulators and fuzzy observers: Relaxed stability conditions and LMI-based designs, IEEE Transactions on Fuzzy Systems, Vol. 6, No. 2, pp. 250-265DOI
13 
H. S. Kim, J. B. Park, Y. H. Joo, 2017, Sampled-data fuzzy observer design for an attitude and heading reference system and its experimental validation, Journal of Electrical Engineering and Technology, Vol. 12, No. 6, pp. 2399-2410DOI
14 
H. J. Kim, J. B. Park, Y. H. Joo, 2018, Sampled-data $H_โˆž$ fuzzy observer for uncertain oscillating systems with immeasurable premise variables, IEEE Access, Vol. 7, pp. 58075-58085DOI
15 
S. H. Hwang, J. B. Park, Y. H. Joo, August 2019, Disturbance observer-based integral fuzzy liding-mode control and its application to wind turbine system, IET Control Theory and Applications, Vol. 13, No. 12, pp. 1891-1900DOI
16 
L. A. Mozelli, R. M. Palhares, G. S. C. Avellar, 2009, A systematic approach to improve multiple Lyapunov function stability and stabilization conditions for fuzzy systems, Information Sciences, Vol. 179, pp. 1149-1162DOI
17 
X. L. Zhu, Y. Wang, 2011, Stabilization for sampled-data neural-network-based control systems, IEEE Transactions on System, Man, and Cybernetics B, Vol. 41, No. 1, pp. 210-221DOI
18 
D. W. Kim, J. B. Park, Y. H. Joo, 2008, Theoretical justifi- cation of approximate norm minimization method for intel- ligent digital redesign, Automatica, Vol. 44, No. 3, pp. 851-856DOI
19 
D. H. Lee, J. B. Park, Y. H. Joo, 2011, Further improvement of periodic control approach for relaxed stabilization condition of discrete-time Takagi-Sugeno fuzzy systems, Fuzzy Sets and Systems, Vol. 174, No. 1, pp. 50-65DOI
20 
H. J. Kim, G. B. Koo, J. B. Park, Y. H. Joo, 2015, Decentralized sampled-data H fuzzy filter for nonlinear largescale systems, Fuzzy Sets and Systems, Vol. 273, pp. 68-86DOI
21 
M. K. Song, J. B. Park, Y. H. Joo, 2015, Robust stabilization for uncertain Markovian jump fuzzy systems based on free weighting matrix method, Fuzzy Sets and Systems, Vol. 277, pp. 81-96DOI
22 
N. Gnaneswaran, Y. H. Joo, 2019, Event-triggered stabilization for T-S fuzzy systems with asynchronous premise constraints and its application to wind turbine system, IET Control Theory and Applications, Vol. 13, No. 10, pp. 1532-1542DOI
23 
N. Gunasekaran, Y. H. Joo, 2019, Stochastic sampled-data controller for T S fuzzy chaotic systems and its applications, IET Control Theory and Applications, Vol. 13, No. 12, pp. 1834-1843DOI
24 
P. Mani, J. H. Lee, K. W. Kang, Y. H. Joo, 2019, Digital controller design via LMIs for direct-driven surface mounted PMSG-based wind energy conversion system, IEEE Transactions Cybernetics Online Publication.DOI

์ €์ž์†Œ๊ฐœ

๊น€ํ•œ์†” (Han Sol Kim)
../../Resources/kiee/KIEE.2019.68.11.1403/au1.png

Han Sol Kim received B.S. degree in Electronic and Computer Engineering from Hanyang University, Korea, in 2011 and M.S. and Ph.D degree in Electrical and Electronic Engineering, Yonsei University, Korea, in 2012 and 2018, respectively.

He joined Samsung Electronics Co. from 2018.

His current research interests include sampled-data control of fuzzy systems, fuzzy-model-based control, and interconnected fuzzy systems.

์ฃผ์˜ํ›ˆ (Young Hoon Joo)
../../Resources/kiee/KIEE.2019.68.11.1403/au2.png

Young Hoon Joo received the B.S., M.S., and Ph.D. degrees in electrical engineering from Yonsei University, Seoul, South Korea, in 1982, 1984, and 1995, respectively.

He was a Project Manager with Samsung Electronics Company, Seoul, from 1986 to 1995.

He was a Visiting Professor with the Department of Electrical and Computer Engineering, University of Houston, Houston, TX, USA, from 1998 to 1999.

He is currently a Professor with the School of IT Information and Control Engineering, Kunsan National University, Gunsan, South Korea.

His current research interests include intelligent robot, intelligent control, wind energy systems, and computer vision.

Dr. Joo served as the President for the Korea Institute of Intelligent Systems in 2009, the Editor-in-Chief for the Intelligent Journal of Control, Automation, and Systems from 2014 to 2017, and the Vice-President for Institute of Control, Robot and Control from 2016 to 2017.

He is serving as the President for the Korean Institute of Electrical Engineers in 2019 and the Director for Research Center of Wind Energy Systems funded by the Korean Government, Kunsan University.