• ๋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 Electrical and Computer Engineering, Pusan National University, Korea.)
  2. (Energy Platform Research Center, Korea Electrotechnology Research Institute, Korea.)



Digital Twin, Machine Learning, Optimization, Power System, Energy Storage System

1. ์„œ ๋ก 

์ตœ๊ทผ ๋””์ง€ํ„ธ๋‰ด๋”œ(Digital new deal), ๊ทธ๋ฆฐ ๋‰ด๋”œ(Green new deal)์ด ๋– ์˜ค๋ฅด๋ฉฐ, ์ •๋ถ€๋Š” ๋ฐ์ดํ„ฐ ๊ฒฝ์ œ ์ด‰์ง„์„ ํ†ตํ•ด ์‹ ์‚ฐ์—… ์œก์„ฑ ๋ฐ ์ฃผ๋ ฅ์‚ฐ์—…์— ๋””์ง€ํ„ธ ์ „ํ™˜์„ ๊ฐ€์†ํ™”ํ•˜๊ณ  ์žˆ๋‹ค. ์นœํ™˜๊ฒฝ, ์ €ํƒ„์†Œ ๋“ฑ ๊ทธ๋ฆฐ ๊ฒฝ์ œ๋กœ์˜ ์ „ํ™˜์„ ์ด‰์ง„ํ•˜์—ฌ ํƒ„์†Œ ์ค‘๋ฆฝ์„ ์ง€ํ–ฅํ•˜๊ณ  ์—๋„ˆ์ง€ ์ ˆ์•ฝ๊ณผ ์‹ ์žฌ์ƒ์—๋„ˆ์ง€ ํ™•์‹  ๋“ฑ์˜ ๊ธฐ๋ฐ˜์ด ๋  ์ˆ˜ ์žˆ๋Š” ์นœํ™˜๊ฒฝ ์—๋„ˆ์ง€ ์ธํ”„๋ผ๋ฅผ ๊ตฌ์ถ•ํ•˜๊ณ  ์žˆ๋‹ค(1). ๋””์ง€ํ„ธ ๋‰ด๋”œ์˜ ํ•ต์‹ฌ ๊ธฐ์ˆ  ์ค‘ ํ•˜๋‚˜์ธ ๋””์ง€ํ„ธ ํŠธ์œˆ์€(Digital Twin) ์‹ค์ œ ๊ณต๊ฐ„๊ณผ ๋™์ผํ•œ ์‹œ์Šคํ…œ์„ ๊ฐ€์ƒ๊ณต๊ฐ„์— ๋ชจ์‚ฌํ•˜์—ฌ ์ด๋ฅผ ๋™๊ธฐํ™”์‹œํ‚ค๋Š” ๊ธฐ์ˆ ๋กœ ๋ฌผ๋ฆฌ์  ๋ฌธ์ œ๋ฅผ ๋ณด๋‹ค ๋นจ๋ฆฌ ๊ฐ์ง€ํ•˜์—ฌ ๋ฌผ๋ฆฌ์  ๋ณ€ํ™”๋ฅผ ๋น ๋ฅด๊ฒŒ ์˜ˆ์ธกํ•  ์ˆ˜ ์žˆ๋‹ค. ๊ตญ๋‚ด์˜ ๊ฒฝ์šฐ์—๋Š” 2016๋…„ ๋ง ์ด์ „๊นŒ์ง€ ๋””์ง€ํ„ธ ํŠธ์œˆ ๊ธฐ์ˆ ์˜ ๊ด€์‹ฌ์ด ๋งค์šฐ ๋‚ฎ์•˜์œผ๋‚˜, 2016๋…„ 10์›” ๊ฐ€ํŠธ๋„ˆ์—์„œ 2017๋…„ 10๋Œ€ ์ „๋žต๊ธฐ์ˆ ๋กœ ๋””์ง€ํ„ธํŠธ์œˆ์„ ๋ฐœํ‘œํ•˜๋ฉด์„œ ๊ธฐ์ˆ ์  ๊ด€์‹ฌ์ด ๊ธ‰์ฆํ•˜์˜€๋‹ค(2). ๋””์ง€ํ„ธ ํŠธ์œˆ์„ ํ†ตํ•ด ๊ฐ€์ƒ๊ณต๊ฐ„์— ์กด์žฌํ•˜๋Š” ํŠธ์œˆ ๋ชจ๋ธ์„ ์‹ค์‹œ๊ฐ„์œผ๋กœ ๋ชจ๋‹ˆํ„ฐ๋งํ•˜์—ฌ ์‹œ์Šคํ…œ์˜ ๊ฑด์ „ํ•œ ์ƒํƒœ๋ฅผ ๊ด€๋ฆฌํ•˜๊ณ  ์ด์ƒ ํ–‰๋™์„ ์‚ฌ์ „์— ๊ฐ์ง€ํ•˜์—ฌ ์‹œ์Šคํ…œ์˜ ์ด์ƒ ์œ ๋ฌด๋ฅผ ์ง„๋‹จํ•˜๋ฉฐ, ์‹œ์Šคํ…œ ๊ตฌ์„ฑ ๋ฐ ์šด์˜ ์•Œ๊ณ ๋ฆฌ์ฆ˜์„ ๋‹ค์–‘ํ•˜๊ฒŒ ๋ณ€ํ™”์‹œ์ผœ ํ…Œ์ŠคํŠธํ•  ์ˆ˜ ์žˆ๋‹ค. ์ด๋Š” ์‹œ์Šคํ…œ์„ ์šด์˜ํ•  ๋•Œ ์„ค๋น„์˜ ๊ตฌ์„ฑ ๋ณ€๊ฒฝ๊ณผ ์šด์˜ ์•Œ๊ณ ๋ฆฌ์ฆ˜ ์„ ํƒ ๋ฐ ์˜์‚ฌ๊ฒฐ์ •์˜ ์–ด๋ ค์›€, ์‹œ์Šคํ…œ ์˜ค์ž‘๋™ ๋“ฑ๊ณผ ๊ฐ™์ด ์•ผ๊ธฐ๋˜๋Š” ๋ฌธ์ œ์ ์„ ๊ทน๋ณตํ•  ์ˆ˜ ์žˆ๋„๋ก ๋„์™€์ค€๋‹ค.

๋””์ง€ํ„ธ ํŠธ์œˆ ๊ธฐ์ˆ ์€ ์ œ์กฐ ๋ฐ ๊ณต์ •์‚ฐ์—…์— ์ฃผ๋กœ ์‚ฌ์šฉ๋˜๊ณ  ์žˆ์ง€๋งŒ, ์ „๋ ฅ์‹œ์žฅ์—์„œ์˜ ๋””์ง€ํ„ธํ™”๋ฅผ ํ†ตํ•ด ์ „๋ ฅ์‚ฐ์—…๋„ ๋ฐœ์ „ํ•จ์— ๋”ฐ๋ผ ์ „๋ ฅ์‹œ์Šคํ…œ์—๋„ ์ ์šฉ์‹œํ‚จ๋‹ค๋ฉด ๋”์šฑ ํšจ์œจ์ ์œผ๋กœ ์ „๋ ฅ์‹œ์Šคํ…œ์„ ์šด์˜ํ•  ์ˆ˜ ์žˆ๋‹ค(3). ์ „๋ ฅ ๋ถ„์•ผ์—์„œ๋Š” ์ธ๊ณต์ง€๋Šฅ(Artificial Intelligence) ๊ธฐ์ˆ ์„ ์ ์šฉํ•œ ์ „๋ ฅ์‹œ์Šคํ…œ์˜ ์šด์˜ ๋ฐ์ดํ„ฐ ์ˆ˜์ง‘ ๋ฐ ๋ถ„์„์— ๋Œ€ํ•œ ์—ฐ๊ตฌ๋Š” ๋ฌด์ˆ˜ํžˆ ๋งŽ์ง€๋งŒ, ์‹ค์ œ ์ „๋ ฅ์‹œ์Šคํ…œ์„ ๊ฐ€์ƒ๊ณต๊ฐ„์— ๋ชจ์‚ฌํ•˜์—ฌ ๊ตฌํ˜„ํ•˜๊ณ  ์‹œ๋ฎฌ๋ ˆ์ด์…˜์„ ์ˆ˜ํ–‰ํ•˜๋Š” ๋””์ง€ํ„ธ ํŠธ์œˆ ์—ฐ๊ตฌ๋Š” ๋ฏธํกํ•˜๋‹ค. ํŠนํžˆ, ์†Œ๊ทœ๋ชจ ์ „๋ ฅ์‹œ์Šคํ…œ์— ๋””์ง€ํ„ธ ํŠธ์œˆ์„ ์ ์šฉํ•˜์—ฌ ์‹ค์ฆํ•œ ์—ฐ๊ตฌ๋Š” ์•„์ง ๋ถ€์กฑํ•œ ์‹ค์ •์ด๋‹ค. ๊ตญ๋‚ด์—์„œ๋Š” ๋Œ€ํ‘œ์ ์œผ๋กœ ์ƒˆ๋งŒ๊ธˆ ์Šค๋งˆํŠธ ์ˆ˜๋ณ€๋„์‹œ์™€ ์„ธ์ข… ์Šค๋งˆํŠธ์‹œํ‹ฐ์— ๋””์ง€ํ„ธ ํŠธ์œˆ ํ”Œ๋žซํผ์„ ๊ตฌ์ถ•ํ•˜์—ฌ ์‹ค์ œ์™€ ์œ ์‚ฌํ•œ ๊ฐ€์ƒ ๋„์‹œ๋ฅผ ๋งŒ๋“ค์–ด ์—ฐ๊ตฌ๋ฅผ ์ง„ํ–‰ํ•˜๊ณ  ์žˆ๋‹ค. ๋””์ง€ํ„ธ ํŠธ์œˆ์„ ํ†ตํ•ด ๊ฐ€์ƒ๊ณต๊ฐ„์—์„œ ๊ธฐ์กด์˜ ์ „๋ ฅ์‹œ์Šคํ…œ ๊ตฌ์„ฑ์„ ๋ณ€๊ฒฝํ•˜๊ฑฐ๋‚˜ ๋‹ค์–‘ํ•œ ์šด์˜ ์•Œ๊ณ ๋ฆฌ์ฆ˜์„ ์„ ์ •ํ•˜๋Š” ๋“ฑ ์‹œ๋ฎฌ๋ ˆ์ด์…˜ํ•จ์œผ๋กœ์จ ์šด์˜ ์‹œ ๋ฐœ์ƒํ•˜๋Š” ์˜์‚ฌ๊ฒฐ์ •์˜ ์–ด๋ ค์›€์„ ํ•ด์†Œํ•˜์—ฌ ์‹œ์Šคํ…œ์˜ ์•ˆ์ •์„ฑ ๋ฐ ์šด์˜์˜ ํšจ์œจ์„ฑ์„ ํ–ฅ์ƒ์‹œํ‚ฌ ์ˆ˜ ์žˆ๋‹ค. ์—๋„ˆ์ง€ ๋ถ„์•ผ์—์„œ์˜ ๋””์ง€ํ„ธ ํŠธ์œˆ ๊ธฐ์ˆ  ์ ์šฉ์€ ์—๋„ˆ์ง€ ๋””์ง€ํ„ธํ™” ๊ธฐ์ˆ ์„ ๋”์šฑ ์ง„ํ™”์‹œ์ผœ ์—๋„ˆ์ง€ ์‚ฐ์—… ์„ฑ์žฅ์— ๊ธฐ์—ฌํ•  ์ˆ˜ ์žˆ์„ ๊ฒƒ์œผ๋กœ ์‚ฌ๋ฃŒ๋œ๋‹ค(3).

๋ณธ ๋…ผ๋ฌธ์—์„œ๋Š” ๋””์ง€ํ„ธ ํŠธ์œˆ ๊ธฐ์ˆ ์„ ์‹ค์ฆ ๋งˆ์ดํฌ๋กœ๊ทธ๋ฆฌ๋“œ(Microgrid)์— ์ ์šฉํ•˜์—ฌ ์‹œ์Šคํ…œ์˜ ์ƒํƒœ์™€ ํŠน์„ฑ์„ ๋ฐ˜์˜ํ•œ ๊ฐ€์ƒ ๋ชจ๋ธ์„ ๊ตฌํ˜„ํ•˜์—ฌ ์‹ค์ œ ์‹œ์Šคํ…œ์„ ์œ ์‚ฌํ•˜๊ฒŒ ๋ชจ์‚ฌํ•˜์˜€์œผ๋ฉฐ, ๋ชจ๋“  ์šด์˜ ๋ฐ์ดํ„ฐ์™€ ์ •๋ณด๊ฐ€ ์‹ค์‹œ๊ฐ„์œผ๋กœ ๊ณต์œ  ๊ฐ€๋Šฅํ† ๋ก ํ•˜์˜€๋‹ค. ๋‹ค๋งŒ, ๋งˆ์ดํฌ๋กœ๊ทธ๋ฆฌ๋“œ ๊ตฌ์„ฑ ์ž์› ์ค‘ ์—๋„ˆ์ง€์ €์žฅ์‹œ์Šคํ…œ(Energy Storage System, ESS)์— ๋Œ€ํ•ด์„œ๋งŒ ์ตœ์  ์šด์˜ ๊ณ„ํš ๋ชจ๋ธ์„ ์ˆ˜๋ฆฝํ•˜์—ฌ ์—ฐ๊ตฌ๋ฅผ ์ˆ˜ํ–‰ํ•˜์˜€๋‹ค. ๊ฐ€์ƒ ๊ณต๊ฐ„์— ๊ตฌํ˜„๋œ ESS์˜ ๊ฐ€์ƒ ๋ชจ๋ธ์— ๊ธฐ๊ณ„ํ•™์Šต ๊ธฐ๋ฐ˜ ์ตœ์  ์šด์˜ ๋ชจ๋ธ์„ ์ ์šฉํ•˜์—ฌ ์‹ค์ฆ ESS ์ตœ์  ์šด์˜ ๊ฒฐ๊ณผ์™€์˜ ์œ ์‚ฌ์„ฑ์„ ๋น„๊ตยท๋ถ„์„ํ•˜์˜€๋‹ค. ์‹ค์ œ ์‹œ์Šคํ…œ๊ณผ ๋™์ผํ•œ ํ™˜๊ฒฝ์œผ๋กœ ๊ตฌ์„ฑ๋˜์–ด ์žˆ๋Š” ๊ฐ€์ƒ๊ณต๊ฐ„์—์„œ ์‹œ์Šคํ…œ ์šด์˜์— ํ•„์š”ํ•œ ์˜์‚ฌ๊ฒฐ์ •์„ ์œ„ํ•ด ์‹œ๋ฎฌ๋ ˆ์ด์…˜์„ ์ˆ˜ํ–‰ํ•˜์—ฌ ์‹œ์Šคํ…œ ์šด์˜์˜ ๋ถˆ์•ˆ์ •์„ฑ๊ณผ ๋ถˆํ™•์‹ค์„ฑ์„ ํ•ด์†Œํ•˜๊ณ ์ž ํ•œ๋‹ค.

2. ์‹ค์ฆ ESS์˜ ๋””์ง€ํ„ธ ํŠธ์œˆ ๋ชจ๋ธ

2.1 ๋””์ง€ํ„ธ ํŠธ์œˆ

๋””์ง€ํ„ธ ํŠธ์œˆ์€ ๊ฐ€์ƒ๊ณต๊ฐ„์—์„œ ์‹œ์Šคํ…œ ๋ถ„์„, ์ตœ์ ํ™”, ์‹œ๋ฎฌ๋ ˆ์ด์…˜ ๋“ฑ์„ ์ˆ˜ํ–‰ํ•  ์ˆ˜ ์žˆ์œผ๋ฏ€๋กœ, ์‹ค์ œ ๊ณต๊ฐ„์—์„œ์˜ ๋น„์šฉ๊ณผ ์‹œ๊ฐ„์„ ์ ˆ์•ฝํ•  ์ˆ˜ ์žˆ๊ณ , ์œ„ํ—˜์„ฑ์„ ์ค„์ผ ์ˆ˜ ์žˆ๋‹ค. ESS ์šด์˜์— ํ•„์š”ํ•œ ์ตœ์  ์ถฉ๋ฐฉ์ „ ๊ณ„ํš์„ ์ˆ˜๋ฆฝํ•˜๊ธฐ ์œ„ํ•œ ๋ชจ๋ธ์„ ๋‹ค์–‘ํ•˜๊ฒŒ ์ ์šฉํ•˜์—ฌ ์‹œ๋ฎฌ๋ ˆ์ด์…˜ ํ•  ์ˆ˜ ์žˆ์œผ๋ฉฐ, ์‹คํ—˜ ์กฐ๊ฑด์— ๋”ฐ๋ผ ๊ฒฐ๊ณผ๊ฐ€ ์–ด๋–ป๊ฒŒ ๋ฐ”๋€Œ๋Š”์ง€ ์‹ค์‹œ๊ฐ„์œผ๋กœ ํ™•์ธ์ด ๊ฐ€๋Šฅํ•˜๋‹ค. ๊ฐ€์ƒ๊ณต๊ฐ„์—์„œ์˜ ์‹œ๋ฎฌ๋ ˆ์ด์…˜์œผ๋กœ ํ†ต์ฐฐ๋ ฅ์„ ํ–ฅ์ƒ์‹œ์ผœ ์˜์‚ฌ๊ฒฐ์ •์„ ์ˆ˜๋ฆฝํ•˜๊ณ  ์ด๋ฅผ ํ˜„์‹ค์— ๋ฐ˜์˜ํ•  ๊ฒฝ์šฐ ์‹œ์Šคํ…œ์˜ ๋ถˆ์•ˆ์ •์„ฑ ๋ฐ ๋ถˆํ™•์‹ค์„ฑ์„ ์ค„์ผ ์ˆ˜ ์žˆ์œผ๋ฉฐ ๋”์šฑ ํšจ์œจ์ ์œผ๋กœ ์šด์˜ํ•  ์ˆ˜ ์žˆ๋‹ค. ๊ทธ๋ฆผ 1์€ ๋งˆ์ดํฌ๋กœ๊ทธ๋ฆฌ๋“œ์—์„œ์˜ ๋””์ง€ํ„ธ ํŠธ์œˆ์„ ๋‚˜ํƒ€๋‚ธ๋‹ค. ๊ฐ๊ฐ์˜ ์ž์›๋“ค์˜ ์šด์˜ ๋ฐ์ดํ„ฐ๋ฅผ ๊ฐ€์ƒ ๊ณต๊ฐ„์— ๋™์ผํ•˜๊ฒŒ ๊ตฌํ˜„ํ•˜์—ฌ ์‹ค์‹œ๊ฐ„์œผ๋กœ ๋ฐ์ดํ„ฐ๋ฅผ ๋ฐ›์•„ ์šด์˜ํ•  ์ˆ˜ ์žˆ์œผ๋ฉฐ, ์ด๋ฅผ ํ†ตํ•ด ์‹ค์ œ ์‹œ์Šคํ…œ์— ํ”ผ๋“œ๋ฐฑ์„ ํ•  ์ˆ˜ ์žˆ๋‹ค.

๊ทธ๋ฆผ. 1. ๋งˆ์ดํฌ๋กœ๊ทธ๋ฆฌ๋“œ์—์„œ์˜ ๋””์ง€ํ„ธ ํŠธ์œˆ

Fig. 1. Digital Twin in Microgrid

../../Resources/kiee/KIEE.2022.71.6.819/fig1.png

2.2 ๋””์ง€ํ„ธ ํŠธ์œˆ์„ ์œ„ํ•œ ์‹ค์ฆ ๋ฐ์ดํ„ฐ ๋ถ„์„

์‹ค์ œ๋กœ ๊ตฌ์ถ•๋˜์–ด ์žˆ๋Š” ๋งˆ์ดํฌ๋กœ๊ทธ๋ฆฌ๋“œ์˜ ๊ตฌ์„ฑ ์ž์› ์ค‘ ์‹ค์ฆ ESS๋ฅผ ๋Œ€์ƒ์œผ๋กœ ์—ฐ๊ตฌ๋ฅผ ์ˆ˜ํ–‰ํ•˜์˜€์œผ๋ฉฐ, ์‹ค์ฆ ESS๋Š” ๋ฐฐํ„ฐ๋ฆฌ ์šฉ๋Ÿ‰ 500kWh, PCS ์šฉ๋Ÿ‰ 250kW๋กœ ๊ตฌ์„ฑ๋˜์–ด ์žˆ๋‹ค. 2019๋…„ 1์›”๋ถ€ํ„ฐ 2019๋…„ 11์›”๊นŒ์ง€๋Š” ์ •๋ถ€์˜ ์ง€์นจ์— ๋”ฐ๋ผ ESS ์šด์˜์„ ์ค‘๋‹จํ•˜์˜€์œผ๋ฏ€๋กœ, ์žฌ๊ฐ€๋™์ด ๋œ ํ›„ ์šด์˜์˜ ์•ˆ์ •ํ™”๊ฐ€ ๋œ ์‹œ๊ธฐ์ธ 2020๋…„ 1์›”๋ถ€ํ„ฐ 2020๋…„ 12์›”๊นŒ์ง€์˜ ESS ์šด์˜ ๋ฐ์ดํ„ฐ๋ฅผ EMS (Energy Management System)๋กœ๋ถ€ํ„ฐ ์ˆ˜์ง‘ํ•˜์˜€๋‹ค. ๊ทธ๋ฆผ 2๋Š” ์ˆ˜์ง‘๋œ ๋ฐ์ดํ„ฐ ์ค‘ 2020๋…„ ์‹ค์ œ ์šด์˜ ๊ธฐ๊ฐ„ ๋™์•ˆ์˜ ์‹ค์ฆ ESS ์ถฉยท๋ฐฉ์ „๋Ÿ‰์„ ๋‚˜ํƒ€๋‚ด๋ฉฐ, EMS๋กœ๋ถ€ํ„ฐ ์ˆ˜์ง‘๋œ ์‹ค์ฆ ESS์˜ ์šด์˜ ๋ฐ์ดํ„ฐ ๊ฐ„ ์ƒ๊ด€๊ด€๊ณ„๋Š” ๊ทธ๋ฆผ 3๊ณผ ๊ฐ™๋‹ค. ESS์˜ ์ถฉยท๋ฐฉ์ „๋Ÿ‰์€ SOC(State of Charge), SOH(State of Health), ์ „๊ธฐ์š”๊ธˆ์˜ ๊ณ„์‹œ๋ณ„ ์š”๊ธˆ ๋‹จ๊ฐ€์™€ ์‹œ๊ฐ„์— ๋น„๊ต์  ์ƒ๊ด€๊ด€๊ณ„๊ฐ€ ๋†’์€ ๊ฒƒ์„ ํ™•์ธํ•  ์ˆ˜ ์žˆ๋‹ค. ๊ณ„์‹œ๋ณ„ ์š”๊ธˆ์ œ ์ ์šฉ์œผ๋กœ ์ธํ•ด ๊ณ„์ ˆ๋ณ„๋กœ ESS ์ถฉยท๋ฐฉ์ „ ์Šค์ผ€์ค„๊ณผ ์ถฉยท๋ฐฉ์ „๋Ÿ‰์ด ์ƒ์ดํ•˜๋ฏ€๋กœ ๊ทธ๋ฆผ 4์™€ ๊ฐ™์ด ๊ณ„์ ˆ๋ณ„๋กœ ๋ฐ์ดํ„ฐ๋ฅผ ๋ถ„๋ฅ˜ํ•˜์—ฌ ๋ถ„์„ํ•˜์˜€๋‹ค.

๊ทธ๋ฆผ. 2. ์‹ค์ฆ ESS์˜ ์ตœ์  ์ถฉยท๋ฐฉ์ „๋Ÿ‰

Fig. 2. Optimal charging/discharging of ESS

../../Resources/kiee/KIEE.2022.71.6.819/fig2.png

๊ทธ๋ฆผ. 3. ์‹ค์ฆ ESS ์šด์˜ ๋ฐ์ดํ„ฐ์˜ ์ƒ๊ด€๊ด€๊ณ„ ๋ถ„์„

Fig. 3. Correlation coefficient of ESS operation data

../../Resources/kiee/KIEE.2022.71.6.819/fig3.png

๊ทธ๋ฆผ. 4. ๊ณ„์ ˆ๋ณ„ ์‹ค์ฆ ESS์˜ ์ตœ์  ์ถฉ๋ฐฉ์ „๋Ÿ‰

Fig. 4. Optimal charging/discharging of ESS by season

../../Resources/kiee/KIEE.2022.71.6.819/fig4.png

3. ESS ์ตœ์  ์šด์˜ ๊ณ„ํš ๋ชจ๋ธ

ESS๋Š” ์šฉ๋„๊ฐ€ ๋‹ค์–‘ํ•˜๊ธฐ์— ์ตœ์  ์šด์˜ ๊ณ„ํš์„ ์ˆ˜๋ฆฝํ•˜๋ ค๋ฉด ์šฉ๋„๋ณ„๋กœ ๋ชฉ์ ํ•จ์ˆ˜์™€ ๊ทธ์— ๋”ฐ๋ฅธ ์ œ์•ฝ์กฐ๊ฑด๋„ ๋‹ค๋ฅด๊ฒŒ ์„ค์ •ํ•ด์•ผ ํ•œ๋‹ค. ESS์™€ PCS์˜ ํŠน์„ฑ์„ ์ถฉ๋ถ„ํžˆ ํŒŒ์•…ํ•˜์—ฌ ์„ค์ •ํ•œ ๋ชฉ์ ํ•จ์ˆ˜์™€ ์ œ์•ฝ์กฐ๊ฑด์„ ํ†ตํ•ด ์šด์˜ ๊ณ„ํš์„ ๋„์ถœํ•œ๋‹ค. ์‹ค์ฆ ESS๋Š” ์ผ๋ฐ˜์ ์œผ๋กœ ๋งŽ์ด ์‚ฌ์šฉ๋˜๊ณ  ์žˆ๋Š” ์ตœ์ ํ™” ๊ธฐ๋ฐ˜ ์•Œ๊ณ ๋ฆฌ์ฆ˜์„ ์‚ฌ์šฉํ•˜์—ฌ ESS ์ผ๊ฐ„ ์ตœ์  ์ถฉ๋ฐฉ์ „ ๊ณ„ํš์„ ์ˆ˜๋ฆฝํ•˜์˜€์œผ๋ฉฐ, ์ „๊ธฐ์š”๊ธˆ ์ ˆ๊ฐ์•ก์„ ์ตœ๋Œ€๋กœ ํ•˜๋Š” ๋ชฉ์ ํ•จ์ˆ˜๋ฅผ ์ •์˜ํ•˜์˜€๊ณ , ESS์˜ ์ •๊ฒฉ ์ถฉ๋ฐฉ์ „ ์ถœ๋ ฅ, ESS์˜ ์ „์ง€์šฉ๋Ÿ‰ ๋ฐ ์ˆ˜์šฉ๊ฐ€์˜ ์ตœ์†Œยท์ตœ๋Œ€ ์ˆ˜์ „์ „๋ ฅ์„ ์ œ์•ฝ์กฐ๊ฑด์œผ๋กœ ์„ค์ •ํ•˜์˜€๋‹ค(4).

3.1 ์ธ๊ณต์ง€๋Šฅ์„ ์ ์šฉํ•œ ESS ์ตœ์  ์šด์˜ ๊ณ„ํš ๋ชจ๋ธ

์ตœ์ ํ™” ๊ธฐ๋ฐ˜ ์šด์˜ ๊ณ„ํš ๋ชจ๋ธ์€ ์‹œ์Šคํ…œ์— ๋Œ€ํ•œ ์ดํ•ด๊ฐ€ ์ •ํ™•ํ•ด์•ผ ์‹œ์Šคํ…œ์ด ์•ˆ์ •์ ์œผ๋กœ ๋™์ž‘ํ•  ์ˆ˜ ์žˆ๋„๋ก ๊ตฌํ˜„ ๊ฐ€๋Šฅํ•˜์ง€๋งŒ, ๋‹ค์ˆ˜์˜ ์ž์›๊ณผ ๋ณต์žกํ•œ ์‹œ์Šคํ…œ ๊ตฌ์กฐ์—์„œ๋Š” ๋น„์„ ํ˜•์„ฑ๊ณผ ์—ฐ์‚ฐ ๋ถ€๋‹ด, ์ •๋ณด์˜ ๋ถ€์žฌ ๋“ฑ์œผ๋กœ ์ธํ•ด ์ •ํ™•ํ•œ ๋ชจ๋ธ๋ง์ด ์–ด๋ ต๋‹ค. ๋˜ํ•œ ๊ธฐ์กด์˜ ์ถ”๊ฐ€ ์ž์› ๋ฐœ์ƒ ๋˜๋Š” ๊ธฐ์กด ์ž์› ์ œ๊ฑฐ ๋“ฑ๊ณผ ๊ฐ™์ด ์‹œ์Šคํ…œ์˜ ๋ณ€ํ™”๊ฐ€ ์žˆ์„ ๊ฒฝ์šฐ๋Š” ๋ชจ๋ธ์˜ ๊ตฌ์กฐ์ ์ธ ์ˆ˜์ •์ด ๋ถˆ๊ฐ€ํ”ผํ•˜๋‹ค. ์ตœ์ ํ™” ๊ธฐ๋ฐ˜์˜ ์šด์˜์—์„œ๋Š” ๋‹ค์ˆ˜์˜ ์ž์›์„ ๊ณ ๋ คํ•˜๋Š” ๊ณผ์ •์—์„œ ๊ฒฐ์ • ๋ณ€์ˆ˜์˜ ์ฆ๊ฐ€, ์‹œ์Šคํ…œ ๋ณต์žก์„ฑ์˜ ์ฆ๊ฐ€๋กœ ์ธํ•œ ์—ฐ์‚ฐ ์‹œ๊ฐ„ ์ฆ๊ฐ€, ์ˆ˜๋ ด์„ฑ ๊ฐ์†Œ ๋“ฑ์˜ ์—ฐ์‚ฐ ๋ฌธ์ œ๊ฐ€ ๋ฐœ์ƒํ•  ์ˆ˜ ์žˆ๋‹ค(3).

๋น„์„ ํ˜• ๊ณ ๋ ค, ์—ฐ์‚ฐ ๋ฌธ์ œ, ์‹œ์Šคํ…œ ํ™•์žฅ ๋˜๋Š” ๋ณ€๊ฒฝ ์‹œ ์ˆ˜ํ•™์  ๋ชจ๋ธ๋ง์˜ ์žฌ๊ตฌ์„ฑ, ๋น…๋ฐ์ดํ„ฐ ์ฒ˜๋ฆฌ์˜ ์–ด๋ ค์›€ ๋“ฑ์œผ๋กœ ์ธํ•ด ์ตœ์ ํ™” ๊ธฐ๋ฐ˜์˜ ๋งˆ์ดํฌ๋กœ๊ทธ๋ฆฌ๋“œ ์šด์˜ ๊ธฐ์ˆ ์ด ๊ฐ€์ง€๋Š” ํ•œ๊ณ„๋ฅผ ๊ทน๋ณตํ•  ์ˆ˜ ์žˆ๋Š” ์ƒˆ๋กœ์šด ์šด์˜๊ธฐ๋ฒ•์ด ํ•„์š”ํ•˜๋‹ค. ๋ณธ ์—ฐ๊ตฌ์—์„œ๋Š” ํ•ด๋‹น ๋ฌธ์ œ์ ์„ ๊ทน๋ณตํ•˜๊ธฐ ์œ„ํ•˜์—ฌ ์ธ๊ณต์ง€๋Šฅ ๊ธฐ์ˆ ์„ ๋„์ž…ํ•˜์—ฌ ์šด์˜ ๋ชจ๋ธ์„ ๊ตฌ์ถ•ํ•˜์˜€๋‹ค. ์ œ์•ˆํ•œ ์šด์˜ ๋ชจ๋ธ์€ ๋ฐ์ดํ„ฐ ๊ธฐ๋ฐ˜ ๋ชจ๋ธ๋กœ ์„ค๊ณ„ํ•˜์˜€๊ธฐ์—, ๋ชฉ์ ํ•จ์ˆ˜์™€ ์ œ์•ฝ์กฐ๊ฑด ๋“ฑ์„ ๋”ฐ๋กœ ์„ค์ •ํ•˜์ง€ ์•Š์œผ๋ฏ€๋กœ ๊ธฐ์กด ๊ณ„์‚ฐ๊ณผ์ •์„ ์ถ•์†Œ์‹œํ‚ฌ ์ˆ˜ ์žˆ๋‹ค. ์‹œ์Šคํ…œ์˜ ๋ฌผ๋ฆฌ์  ์ •๋ณด๋ฅผ ์ˆ˜์‹ํ™”ํ•˜์—ฌ ๋ชจ๋ธ์— ๋ฐ˜์˜ํ•˜์ง€ ์•Š๊ณ , ๊ณผ๊ฑฐ์˜ ์šด์˜ ๋ฐ์ดํ„ฐ๋งŒ์œผ๋กœ ๋ชจ๋ธ์„ ํ•™์Šต์‹œํ‚ค๋Š” ๊ธฐ๊ณ„ํ•™์Šต ๋ชจ๋ธ๊ณผ ์„ ํ˜•ํšŒ๊ท€ ๋ชจ๋ธ์„ ์‚ฌ์šฉํ•˜์—ฌ ESS์˜ ์ตœ์  ์šด์˜ ๊ณ„ํš์„ ์ˆ˜๋ฆฝํ•˜์˜€์œผ๋ฉฐ, ๊ทธ ๊ณผ์ •์€ ๊ทธ๋ฆผ 5์™€ ๊ฐ™๋‹ค.

๊ทธ๋ฆผ. 5. ์ธ๊ณต์ง€๋Šฅ ๊ธฐ๋ฐ˜ ESS ์ตœ์  ์šด์˜ ๊ณ„ํš ํ”„๋กœ์„ธ์Šค

Fig. 5. Process of ESS optimal scheduling using AI

../../Resources/kiee/KIEE.2022.71.6.819/fig5.png

์ˆ˜์ง‘๋œ ๋ฐ์ดํ„ฐ๋Š” ์š”์ผ๋ณ„ ํŠน์„ฑ์„ ๊ณ ๋ คํ•˜์—ฌ ๊ทผ๋ฌด์ผ, ํ† ์š”์ผ, ๊ณตํœด์ผ๋กœ ๋ถ„๋ฅ˜ํ•œ ํ›„ ๊ทผ๋ฌด์ผ๋งŒ์„ ์‚ฌ์šฉํ•˜์˜€์œผ๋ฉฐ, 2์›”, 5์›”, 8์›”, 10์›”์„ ๊ฒจ์šธ, ๋ด„, ์—ฌ๋ฆ„, ๊ฐ€์„์„ ๋‚˜ํƒ€๋‚ด๋Š” ํ…Œ์ŠคํŠธ ์›”๋กœ ์„ค์ •ํ•˜์˜€๋‹ค. ๊ณ„์ ˆ๋ณ„๋กœ ๊ฐ๊ฐ 9์ข…์˜ 720๊ฐœ ๋ฐ์ดํ„ฐ๋ฅผ ์‚ฌ์šฉํ•˜์˜€๊ณ , ํŠธ๋ ˆ์ด๋‹ ์…‹๊ณผ ํ…Œ์ŠคํŠธ ์…‹์€ 7:3์œผ๋กœ ๊ตฌ์„ฑํ•˜์˜€์œผ๋ฉฐ, ์ˆ˜์ง‘๋œ ๋ฐ์ดํ„ฐ์˜ ํฌ๊ธฐ๊ฐ€ ๋‹ค๋ฅด๋ฏ€๋กœ (0,1) ์‚ฌ์ด๋กœ ์ •๊ทœํ™”(Normalization)ํ•˜์—ฌ ํ™œ์šฉํ•˜์˜€๋‹ค.

EMS๋กœ๋ถ€ํ„ฐ ์ˆ˜์ง‘๋œ ESS ์šด์˜ ๋ฐ์ดํ„ฐ๋Š” ์ „์ฒ˜๋ฆฌ ๊ณผ์ •์„ ๊ฑฐ์นœ๋‹ค. ๋ฐ์ดํ„ฐ ์ „์ฒ˜๋ฆฌ๋Š” ํŠน์ง•์„ ๋ถ„์„ํ•˜๊ธฐ ์ ํ•ฉํ•˜๊ฒŒ ๋ฐ์ดํ„ฐ๋ฅผ ๊ฐ€๊ณตํ•˜๋Š” ์ž‘์—…์„ ๋งํ•œ๋‹ค. ์ˆ˜์ง‘๋œ ๋ฐ์ดํ„ฐ์—๋Š” ์˜๋ฏธ ์—†๋Š” ๊ฐ’์ด๋‚˜ ๋„(Null) ๊ฐ’์ด ์กด์žฌํ•˜๋Š” ๋“ฑ์œผ๋กœ ์ธํ•ด ์ˆ˜๋งŽ์€ ๋ณ€์ˆ˜๋“ค์ด ๋ฐ์ดํ„ฐ์˜ ํ’ˆ์งˆ์„ ๋–จ์–ด๋œจ๋ฆฌ๊ธฐ ๋•Œ๋ฌธ์— ์ด๋ฅผ ๋ฐฉ์ง€ํ•˜๊ธฐ ์œ„ํ•ด์„œ ํ•ด๋‹น ์ž‘์—…์„ ์ˆ˜ํ–‰ํ•œ๋‹ค. ๋ฐ์ดํ„ฐ ๋ถ„์„์— ๋ถ€์ ํ•ฉํ•œ ๊ตฌ์กฐ, ๋ˆ„๋ฝ๋œ ํ•ญ๋ชฉ, ๊ฒฐ์ธก๊ฐ’ ์กด์žฌ ๋“ฑ์œผ๋กœ ์ธํ•ด ์ „์ฒ˜๋ฆฌ ๊ณผ์ •์„ ํ•„์ˆ˜์ ์ด๋ฉฐ, ๋…ธ์ด์ฆˆ(Noise) ์ œ๊ฑฐ, ์ค‘๋ณต๊ฐ’ ์ œ๊ฑฐ, ๊ฒฐ์ธก๊ฐ’ ๋ณด์ •, ์ด์ƒ์น˜(Outlier) ๊ฒ€์ถœ ๋“ฑ์œผ๋กœ ์›๋ณธ ๋ฐ์ดํ„ฐ์™€ ๋ณด์ •๋œ ๋ฐ์ดํ„ฐ์™€์˜ ๊ด€๊ณ„๋ฅผ ๊ณ ๋ คํ•˜์—ฌ ๋ชฉ์ ์— ์•Œ๋งž์€ ๋ฐ์ดํ„ฐ ์ „์ฒ˜๋ฆฌ ๊ณผ์ •์„ ์ˆ˜ํ–‰ํ•œ๋‹ค. ๊ทธ ์ดํ›„ ์šด์˜ ๋ฐ์ดํ„ฐ๋ฅผ ์šด์˜ ํŠน์„ฑ์— ๋งž๊ฒŒ ๋ถ„๋ฅ˜ํ•˜๊ณ  ๋ถ„์„ํ•œ๋‹ค. ๊ณผ๊ฑฐ ๋ฐ์ดํ„ฐ๋ฅผ ๊ธฐ๋ฐ˜์œผ๋กœ ๋ชจ๋ธ์„ ๊ฐœ๋ฐœํ•˜๋Š” ๋ฐ ์žˆ์–ด ์ด์ƒ์น˜๋‚˜ ๋…ธ์ด์ฆˆ ๋“ฑ์€ ๋ชจ๋ธ์˜ ์„ฑ๋Šฅ์„ ์ €ํ•˜์‹œํ‚ค๊ธฐ ๋•Œ๋ฌธ์— ๋ชจ๋ธ์˜ ์„ฑ๋Šฅ ๊ฐœ์„ ์„ ์œ„ํ•ด์„œ๋Š” ๋ฐ์ดํ„ฐ์˜ ์ „์ฒ˜๋ฆฌ ๊ณผ์ •์ด ํ•„์ˆ˜์ ์ด๋‹ค.

์ „์ฒ˜๋ฆฌ ๊ณผ์ •์„ ๊ฑฐ์นœ ๋ฐ์ดํ„ฐ๋ฅผ ํ•™์Šต, ๊ฒ€์ฆ, ํ…Œ์ŠคํŠธ ๋ฐ์ดํ„ฐ ์…‹(Dataset)์œผ๋กœ ๊ตฌ์„ฑํ•˜๊ณ  ๊ฐœ๋ฐœ๋œ ๋ชจ๋ธ์„ ์„ ํƒํ•˜์—ฌ ์ตœ์ข…์ ์œผ๋กœ ESS ์šด์ „๊ณ„ํš์„ ์ˆ˜๋ฆฝํ•œ๋‹ค. ์ธ๊ณต์ง€๋Šฅ์„ ์ ์šฉํ•œ ESS ์ตœ์  ์šด์˜ ๊ณ„ํš์€ ์ผ๋ฐ˜์ ์ธ ์„ ํ˜•ํšŒ๊ท€ ๊ธฐ๋ฒ•์ธ ์ตœ์†Œ์ž์Šน๋ฒ•(Least Square Estimation, LSE)๊ณผ ๊ธฐ๊ณ„ํ•™์Šต ๊ธฐ๋ฒ•์ธ NARX ๋„คํŠธ์›Œํฌ๋ฅผ ์ ์šฉํ•˜์—ฌ ๋ชจ๋ธ์„ ๊ตฌํ˜„ํ•˜์˜€๋‹ค. ๋ฐ˜๋ณต์ ์ธ ํ•™์Šต์œผ๋กœ ๋ชจ๋ธ์˜ ์„ฑ๋Šฅ์„ ๋†’์ธ ํ• ๋ชจ๋ธ(Fit model)์„ ์ด์šฉํ•˜์—ฌ ์ตœ์ ์˜ ์šด์˜ ๊ณ„ํš ์ˆ˜๋ฆฝ์„ ์œ„ํ•œ ESS ์ถฉ๋ฐฉ์ „๋Ÿ‰์„ ๋„์ถœํ•œ๋‹ค.

์„ ํ˜•ํšŒ๊ท€๋ถ„์„ ์ค‘ ํ•˜๋‚˜์ธ LSE๋Š” ์„ ํ˜•ํšŒ๊ท€ ๋ฐฉ์ ์‹์˜ ๊ณ„์ˆ˜๋ฅผ ๊ณ„์‚ฐํ•˜๋Š”๋ฐ ์ผ๋ฐ˜์ ์œผ๋กœ ์‚ฌ์šฉ๋œ๋‹ค. LSE๋Š” ํšŒ๊ท€๋ฐฉ์ •์‹์„ ์‚ฌ์šฉํ•˜์—ฌ ์˜ˆ์ธกํ•œ ๊ฐ’๊ณผ ์‹ค์ œ ๊ฐ’์˜ ์˜ค์ฐจ ์ œ๊ณฑ์ด ์ตœ์†Œ๊ฐ€ ๋˜๋„๋ก ์˜ค์ฐจ ๊ณ„์ˆ˜๋ฅผ ๊ฒฐ์ •ํ•˜๋ฉฐ ์•„๋ž˜์˜ ์‹(1)-(3)๊ณผ ๊ฐ™์ด ๊ณ„์‚ฐํ•œ๋‹ค(8,9).

(1)
$A X= B$

(2)
$$ A=\left[\begin{array}{llll} 1 & 2 & \cdots & n \\ 1 & 1 & \cdots & 1 \end{array}\right]^{T}, B=\left[\begin{array}{lllll} x_{1, k} & x_{2, k} & \cdots & x_{n, k} \end{array}\right]^{T} $$

(3)
$X =(A^{T}A)^{-1}A^{T}B$

์ผ๋ฐ˜์ ์ธ ์‹ ๊ฒฝ๋ง ๋„คํŠธ์›Œํฌ์™€๋Š” ๋‹ค๋ฅด๊ฒŒ ์ถœ๋ ฅ ๋‰ด๋Ÿฐ์—์„œ ํ”ผ๋“œ๋ฐฑ ๊ตฌ์กฐ๋ฅผ ํ†ตํ•ด ์™ธ์ƒ์ ์ธ ์ž…๋ ฅ์„ ๊ฐ–๋Š” ๋น„์„ ํ˜• ์ž๊ธฐํšŒ๊ท€(Nonlinear autoregressive with external input, NARX) ๋„คํŠธ์›Œํฌ๋Š” ๊ทธ๋ฆผ 6๊ณผ ๊ฐ™๋‹ค.

๊ทธ๋ฆผ. 6. NARX ๋„คํŠธ์›Œํฌ ๊ตฌ์กฐ๋„

Fig. 6. Architecture of NARX Networks.

../../Resources/kiee/KIEE.2022.71.6.819/fig6.png

NARX ๋„คํŠธ์›Œํฌ๋Š” ์™ธ์ƒ๋ณ€์ˆ˜์ธ u(t)์™€ ๋‚ด์ƒ๋ณ€์ˆ˜์ธ y(t)๊ฐ€ ์กด์žฌํ•˜๋Š” ํ”ผ๋“œ๋ฐฑ ๋”œ๋ ˆ์ด(Feed back delay)๋ฅผ ๊ฐ€์ง„ ์žฌ์ˆœํ™˜ ๋™์  ๋„คํŠธ์›Œํฌ๋กœ, ํ”ผ๋“œ๋ฐฑ ์—ฐ๊ฒฐ์ด ์—ฌ๋Ÿฌ ๊ณ„์ธต์„ ํฌํ•จํ•˜๊ณ  ์žˆ์œผ๋ฉฐ ์ด๋Š” ์‹(4)์™€ ๊ฐ™๋‹ค(5,6). ํ•œ ์‹œ๊ณ„์—ด์˜ ์ด์ „๊ฐ’, ํ”ผ๋“œ๋ฐฑ ์ž…๋ ฅ๊ฐ’, ์™ธ๋ถ€ ์‹œ๊ณ„์—ด์ธ ๋‘ ๋ฒˆ์งธ ์‹œ๊ณ„์—ด์„ ์‚ฌ์šฉํ•˜์—ฌ ์›๋ž˜ ์‹œ๊ณ„์—ด์„ ์˜ˆ์ธกํ•˜๋„๋ก ํ•™์Šตํ•œ๋‹ค.

(4)
\begin{align*} y(t)= f( & y(t-1),\:y(t-2),\:\cdots ,\:y(t-n_{y}),\:\\ & u(t-1),\:u(t-2),\:\cdots ,\:u(t-n_{u})) \end{align*}

3.2 ESS ์ตœ์  ์šด์˜ ๋ชจ๋ธ ์„ฑ๋Šฅํ‰๊ฐ€

์ธ๊ณต์ง€๋Šฅ์„ ์ ์šฉํ•œ ESS ์ตœ์  ์šด์˜ ๊ณ„ํš ๋ชจ๋ธ์˜ ์„ฑ๋Šฅ์€ ํ‰๊ฐ€์ง€ํ‘œ๋ฅผ ํ†ตํ•ด ์ •ํ™•์„ฑ์„ ํ‰๊ฐ€ํ•˜์˜€๋‹ค. ESS ์šด์˜์— ํ•„์š”ํ•œ ์ต์ผ 1์‹œ๊ฐ„ ๋‹จ์œ„ ์ตœ์  ์ถฉ๋ฐฉ์ „ ๊ณ„ํš ๋ชจ๋ธ์˜ ์„ฑ๋Šฅํ‰๊ฐ€์ง€ํ‘œ์ธ ํ‰๊ท ์ œ๊ณฑ๊ทผ์˜ค์ฐจ(Root Mean Square Error, RMSE)์™€ ํ‰๊ท ์ ˆ๋Œ€์˜ค์ฐจ(Mean Absolute Error, MAE)๋ฅผ ์‚ฌ์šฉํ•˜์—ฌ ์ •๋Ÿ‰์ ์œผ๋กœ ์ •ํ™•์„ฑ์„ ํ‰๊ฐ€ํ•˜์˜€๋‹ค. RMSE์™€ MAE๋Š” ๊ฐ’์ด ์ ์„์ˆ˜๋ก ๋ชจ๋ธ์˜ ์ •ํ™•์„ฑ์„ ๋†’์ด ํ‰๊ฐ€ํ•œ๋‹ค.

ํ‘œ 1. ESS ์ตœ์  ์šด์˜ ๋ชจ๋ธ์˜ ์„ฑ๋Šฅํ‰๊ฐ€[RMSE]

Table 1. Performance of ESS optimal scheduling model[RMSE]

Spring

Summer

Fall

Winter

Linear

Regression

1.19

2.02

0.38

0.38

Machine

Learning

0.54

1.19

0.77

0.23

ํ‘œ 2. ESS ์ตœ์  ์šด์˜ ๋ชจ๋ธ์˜ ์„ฑ๋Šฅํ‰๊ฐ€[MAE]

Table 2. Performance of ESS optimal scheduling model[MAE]

Spring

Summer

Fall

Winter

Linear

Regression

0.68

1.81

0.32

0.31

Machine

Learning

0.36

0.87

0.39

0.17

๋””์ง€ํ„ธ ํŠธ์œˆ ๊ธฐ์ˆ ์„ ์ ์šฉํ•˜์—ฌ ๊ฐ€์ƒ๊ณต๊ฐ„์— ์‹ค์ œ ESS๋ฅผ ๋ชจ์‚ฌํ•œ ๊ฐ€์ƒ ESS ๋ชจ๋ธ์˜ ์ตœ์  ์šด์˜๋ชจ๋ธ ์„ฑ๋Šฅ์„ ์‹ค์ œ ์‹œ์Šคํ…œ ์šด์˜ ๊ฒฐ๊ณผ์™€ ๋น„๊ตยท๋ถ„์„ํ•œ ๊ฒฐ๊ณผ๋ฅผ ํ‘œ 1~2๋กœ ํ™•์ธํ•  ์ˆ˜ ์žˆ๋‹ค. RMSE์™€ MAE ์ง€ํ‘œ ๋ชจ๋‘ ๊ธฐ๊ณ„ํ•™์Šต์„ ์ ์šฉํ•œ ๋ชจ๋ธ์ด ๊ฐ€์žฅ ์ ์€ ์ˆ˜์น˜๋กœ ๋ชจ๋ธ์˜ ์„ฑ๋Šฅ์ด ๊ฐ€์žฅ ์šฐ์ˆ˜ํ•˜๊ณ  ์‹ค์ œ ์šด์˜ ๋ชจ๋ธ๊ณผ ๊ฐ€์žฅ ์œ ์‚ฌํ•จ์„ ํ™•์ธํ•  ์ˆ˜ ์žˆ๋‹ค. ๋ชจ๋ธ์˜ ์„ฑ๋Šฅํ‰๊ฐ€์ง€ํ‘œ์ธ RMSE์™€ MAE๋ฅผ ๊ตฌํ•˜๋Š” ๋ฐฉ๋ฒ•์€ ๊ฐ๊ฐ ์‹(5),(6)๊ณผ ๊ฐ™๋‹ค(7).

(5)
$RMSE =\sqrt{\sum_{ {i}=1}^{ {n}}(\hat {y}_{ {i}}- {y}_{ {i}})^{2}/ {n}}$

(6)
$MAE =\left |\sum_{i=1}^{n}\left |\hat y_{i}-y_{i}\right | /n\right |$

๊ทธ๋ฆผ 7์€ ์‹ค์ œ ESS ์ตœ์  ์šด์˜ ๊ฒฐ๊ณผ์™€ ๊ฐ€์ƒ ESS ๋ชจ๋ธ์„ ํ†ตํ•ด ๋„์ถœํ•œ ์ตœ์  ์šด์˜ ๊ฒฐ๊ณผ์˜ ์˜ค์ฐจ๋ฅผ ๋ณด์—ฌ์ฃผ๋ฉฐ ๊ฐ€๋กœ์ถ•์€ ์‹ค์ œ ESS ์šด์˜ ๊ฒฐ๊ณผ๋ฅผ ๋‚˜ํƒ€๋‚ด๊ณ , ์„ธ๋กœ์ถ•์€ ์‹œ๋ฎฌ๋ ˆ์ด์…˜์„ ํ†ตํ•ด ๋„์ถœํ•œ ESS ์ตœ์  ์šด์˜ ๊ฒฐ๊ณผ๋ฅผ ๋‚˜ํƒ€๋‚ธ๋‹ค.

๊ทธ๋ฆผ. 7. ESS ์ตœ์  ์šด์˜ ๋ชจ๋ธ ์„ฑ๋Šฅ ๊ฒฐ๊ณผ

Fig. 7. ESS optimal scheduling model performance

../../Resources/kiee/KIEE.2022.71.6.819/fig7.png

๊ทธ๋ฆผ. 8. ESS ์ตœ์  ์šด์˜ ๊ณ„ํš ๊ฒฐ๊ณผ

Fig. 8. ESS optimal scheduling results

../../Resources/kiee/KIEE.2022.71.6.819/fig8.png

๊ทธ๋ฆผ 8์€ ๊ณ„์ ˆ๋ณ„ ์‹ค์ œ ESS ์šด์˜ ๊ฒฐ๊ณผ์™€ ์‹œ๋ฎฌ๋ ˆ์ด์…˜์„ ํ†ตํ•ด ๋„์ถœํ•œ ESS ์šด์˜ ๊ฒฐ๊ณผ์˜ ํŒจํ„ด์ด ์œ ์‚ฌํ•จ์„ ํ™•์ธํ•  ์ˆ˜ ์žˆ๋‹ค.

๊ทธ๋ฆผ. 9. ์ผ๊ฐ„ ESS ์ตœ์  ์šด์˜ ๊ณ„ํš ๊ฒฐ๊ณผ(๋ด„)

Fig. 9. Daily ESS optimal scheduling (Spring)

../../Resources/kiee/KIEE.2022.71.6.819/fig9.png

๊ทธ๋ฆผ. 10. ์ผ๊ฐ„ ESS ์ตœ์  ์šด์˜ ๊ณ„ํš ๊ฒฐ๊ณผ(์—ฌ๋ฆ„)

Fig. 10. Daily ESS optimal scheduling (Summer)

../../Resources/kiee/KIEE.2022.71.6.819/fig10.png

๊ทธ๋ฆผ. 11. ์ผ๊ฐ„ ESS ์ตœ์  ์šด์˜ ๊ณ„ํš ๊ฒฐ๊ณผ(๊ฐ€์„)

Fig. 11. Daily ESS optimal scheduling (Fall)

../../Resources/kiee/KIEE.2022.71.6.819/fig11.png

๊ทธ๋ฆผ. 12. ์ผ๊ฐ„ ESS ์ตœ์  ์šด์˜ ๊ณ„ํš ๊ฒฐ๊ณผ(๊ฒจ์šธ)

Fig. 12. Daily ESS optimal scheduling (Winter)

../../Resources/kiee/KIEE.2022.71.6.819/fig12.png

๊ทธ๋ฆผ 9~12๋Š” ๋ด„, ์—ฌ๋ฆ„, ๊ฐ€์„, ๊ฒจ์šธ์— ๋Œ€ํ•œ ์ผ๊ฐ„ ESS ์ตœ์  ์šด์˜ ์Šค์ผ€์ค„๋ง ๊ฒฐ๊ณผ๋ฅผ ๋ณด์—ฌ์ค€๋‹ค. ๊ณ„์ ˆ๋ณ„๋กœ ์š”๊ธˆ์ œ๊ฐ€ ๋‹ค๋ฅด๊ณ  ๊ฒฝ๋ถ€ํ•˜, ์ค‘๊ฐ„๋ถ€ํ•˜, ์ตœ๋Œ€๋ถ€ํ•˜์˜ ์‹œ๊ฐ„๋Œ€๊ฐ€ ๋‹ค๋ฅด๊ธฐ ๋•Œ๋ฌธ์— ๋ฐ์ดํ„ฐ๋ฅผ ๊ณ„์ ˆ๋ณ„๋กœ ๋ถ„๋ฅ˜ํ•˜์—ฌ ํ•™์Šตํ•˜์˜€๊ณ , ์ด์— ๋”ฐ๋ผ ๊ณ„์ ˆ๋ณ„ ํŠน์„ฑ์„ ๋ฐ˜์˜ํ•œ ๊ฒฐ๊ณผ๋ฅผ ๋„์ถœํ•˜์˜€๋‹ค.

3.3 ESS ์ตœ์  ์šด์˜ ๋ชจ๋ธ ๊ฒฝ์ œ์„ฑ ๋ถ„์„

์‹ค์ฆ ESS๊ฐ€ ์„ค์น˜๋œ ์‚ฐ์—…์šฉ ์ˆ˜์šฉ๊ฐ€๋Š” ๊ฒฝ๋ถ€ํ•˜, ์ค‘๊ฐ„๋ถ€ํ•˜, ์ตœ๋Œ€๋ถ€ํ•˜์˜ ์‹œ๊ฐ„๋Œ€๋ฅผ ๊ฐ€์ง€๋Š” ๊ณ„์‹œ๋ณ„ ์š”๊ธˆ์ œ๋ฅผ ์‚ฌ์šฉํ•œ๋‹ค. ๊ณ„์‹œ๋ณ„ ์š”๊ธˆ์ œ๋ฅผ ๋ฐ˜์˜ํ•˜์—ฌ ๊ฒฝ๋ถ€ํ•˜, ์ค‘๊ฐ„๋ถ€ํ•˜, ์ตœ๋Œ€๋ถ€ํ•˜ ๊ตฌ๊ฐ„ ์•ˆ์—์„œ ์ถฉ์ „ ๋˜๋Š” ๋ฐฉ์ „ํ•˜๋Š” ๊ทœ์น™ ๊ธฐ๋ฐ˜์˜ ํŠน์ง•์ด ์žˆ๋‹ค. ์ด๋ฅผ ์‚ฌ๋ก€ ์—ฐ๊ตฌ ๊ฒฐ๊ณผ์™€ ์‹ค์ฆ ESS๋ฅผ ๋น„๊ตํ•˜์—ฌ ์œ ์‚ฌํ•จ์„ ํ™•์ธํ•˜์˜€๊ณ  ์‚ฐ์—…์šฉ ์ˆ˜์šฉ๊ฐ€์˜ ๊ณ„์‹œ๋ณ„ ์š”๊ธˆ์ œ ์ •๋ณด๋ฅผ ํ™œ์šฉํ•˜์—ฌ ์š”๊ธˆ ์ ˆ๊ฐ์•ก์„ ์‚ฐ์ถœํ•˜์˜€๋‹ค.

ํ‘œ 3. ESS ์ตœ์  ์šด์˜๋ชจ๋ธ๋ณ„ ์ „๋ ฅ๋Ÿ‰ ์š”๊ธˆ ์ ˆ๊ฐ์•ก[์›]

Table 3. Economic analysis by optimal operating ESS[KRW]

Optimal

Linear

Regression

Machine

Learning

Spring

180,557

170,125

175,826

Summer

498,914

536,068

491,748

Fall

215,713

204,404

211,417

Winter

539,950

546,208

541,396

Total

1,435,134

1,456,805

1,420,387

ํ‘œ 3์€ ESS๋ฅผ ์šด์˜ํ•˜์˜€์„ ๊ฒฝ์šฐ ์ ˆ๊ฐ๋˜๋Š” ์ „๋ ฅ๋Ÿ‰ ์š”๊ธˆ์„ ๋‚˜ํƒ€๋‚ด๋ฉฐ, ์ „๊ธฐ์š”๊ธˆ ์ค‘ ๊ธฐ๋ณธ์š”๊ธˆ์€ ๊ณ ๋ คํ•˜์ง€ ์•Š์€ ๊ธˆ์•ก์ด๋‹ค. ๊ณ„์ ˆ๋ณ„๋กœ 9์ผ์”ฉ ์ด 36์ผ์— ๋Œ€ํ•œ ์ „๋ ฅ๋Ÿ‰ ์š”๊ธˆ ์ ˆ๊ฐ์•ก์ด๋ฉฐ Optimal์€ ์‹ค์ฆ ESS๋ฅผ ์šด์˜ํ•˜์—ฌ ์‹ค์ œ๋กœ ์ ˆ๊ฐ๋œ ์ „๋ ฅ๋Ÿ‰ ์š”๊ธˆ์ด๊ณ , Linear Regression๊ณผ Machine Learning์€ ๊ฐ๊ฐ ์„ ํ˜•ํšŒ๊ท€ ๋ชจ๋ธ๊ณผ ๊ธฐ๊ณ„ํ•™์Šต ๋ชจ๋ธ์„ ์‹ค์ฆ ESS์— ์ ์šฉํ•˜์—ฌ ์ตœ์  ์šด์˜ ๊ณ„ํš์„ ์ˆ˜๋ฆฝํ•˜์—ฌ ์‚ฐ์ถœํ•œ ์ „๋ ฅ๋Ÿ‰ ์š”๊ธˆ ์ ˆ๊ฐ์•ก์„ ๋‚˜ํƒ€๋‚ธ๋‹ค.

Optimal์˜ ์ „๋ ฅ๋Ÿ‰ ์š”๊ธˆ ์ ˆ๊ฐ์•ก์„ ๋ถ„์„ํ•˜์˜€์„ ๋•Œ ์ „๋ ฅ ์‚ฌ์šฉ๋Ÿ‰์ด ๋งŽ๊ณ  ์š”๊ธˆ ๋‹จ๊ฐ€๊ฐ€ ๋น„๊ต์  ๋น„์‹ผ ์—ฌ๋ฆ„๊ณผ ๊ฒจ์šธ์— ์ „๋ ฅ๋Ÿ‰ ์š”๊ธˆ์ด ๋งŽ์ด ์ ˆ๊ฐ๋˜์—ˆ์œผ๋ฉฐ, ์‹ค์ฆ ESS๋ฅผ ์‹ค์ œ๋กœ ์šด์˜ํ•˜์—ฌ 1,435,134์›์˜ ์ „๋ ฅ๋Ÿ‰ ์š”๊ธˆ์ด ์ ˆ๊ฐ๋˜์—ˆ๋‹ค. ์„ ํ˜•ํšŒ๊ท€ ๋ชจ๋ธ์„ ํ™œ์šฉํ•˜์—ฌ ESS ์ตœ์  ์šด์˜ ๊ณ„ํš์„ ์ˆ˜๋ฆฝํ•˜์˜€์„ ๋•Œ ๋ชจ๋ธ์˜ ์„ฑ๋Šฅ์€ ๊ธฐ๊ณ„ํ•™์Šต ๋ชจ๋ธ๋ณด๋‹ค ๋–จ์–ด์กŒ์ง€๋งŒ, ์‚ฐ์ถœ๋œ ์ „๋ ฅ๋Ÿ‰ ์š”๊ธˆ ์ ˆ๊ฐ์•ก์€ ์ด 1,456,805์›์œผ๋กœ ๋‹ค๋ฅธ ๋ชจ๋ธ๋ณด๋‹ค ๊ฐ€์žฅ ๋งŽ์ด ์ ˆ๊ฐ๋˜์—ˆ๋‹ค. ์‹ค์ฆ ESS์— ์ ์šฉ๋œ ์ตœ์ ํ™” ๊ธฐ๋ฐ˜์˜ ๋ชจ๋ธ๊ณผ ๊ฐ€์žฅ ์œ ์‚ฌํ•œ ๋ชจ๋ธ์€ ๊ธฐ๊ณ„ํ•™์Šต ๋ชจ๋ธ์ด์ง€๋งŒ, ๊ฒฝ์ œ์„ฑ ๋ถ„์„ ๊ฒฐ๊ณผ ์‚ฐ์ถœ๋œ ์š”๊ธˆ ์ ˆ๊ฐ์•ก์€ 1,420,387์›์œผ๋กœ ๊ฐ€์žฅ ์ ๊ฒŒ ์ ˆ๊ฐ๋˜์—ˆ๋‹ค. ๋‹ค๋งŒ, ์‹ค์ œ๋กœ ESS๋ฅผ ์šด์˜ํ•˜์—ฌ ์–ป์€ ์ „๋ ฅ๋Ÿ‰ ์š”๊ธˆ ์ ˆ๊ฐ์•ก๊ณผ ๊ฐ ๋ชจ๋ธ์˜ ํ™œ์šฉํ•˜์—ฌ ์‚ฐ์ถœํ•œ ์ „๋ ฅ๋Ÿ‰ ์š”๊ธˆ ์ ˆ๊ฐ์•ก์„ ๋น„๊ตํ•˜์˜€์„ ๋•Œ ์„ ํ˜•ํšŒ๊ท€ ๋ชจ๋ธ์€ 21,671์›์ด๊ณ , ๊ธฐ๊ณ„ํ•™์Šต์„ ์ ์šฉํ–ˆ์„ ๊ฒฝ์šฐ์—๋Š” 14,747์›์œผ๋กœ ์„ ํ˜•ํšŒ๊ท€ ๋ชจ๋ธ๋ณด๋‹ค ์ ˆ๊ฐ์•ก์˜ ์ฐจ์ด๊ฐ€ ์ ์—ˆ๋‹ค.

๊ฐ ๋ชจ๋ธ๋ณ„๋กœ ์‚ฐ์ถœํ•œ ์ „๋ ฅ๋Ÿ‰ ์š”๊ธˆ ์ ˆ๊ฐ์•ก๊ณผ ์‹ค์ œ ์ „๋ ฅ๋Ÿ‰ ์š”๊ธˆ ์ ˆ๊ฐ์•ก์˜ ์ฐจ์ด๋Š” ์šด์˜ ๊ณ„ํš ๋ชจ๋ธ์˜ ์œ ์‚ฌํ•จ์„ ๋‚˜ํƒ€๋‚ด๋ฏ€๋กœ, ์ „๋ ฅ๋Ÿ‰ ์š”๊ธˆ์ด ๊ฐ€์žฅ ๋งŽ์ด ์ ˆ๊ฐ๋˜์—ˆ๋”๋ผ๋„ ์‹ค์ œ ์ „๋ ฅ๋Ÿ‰ ์š”๊ธˆ ์ ˆ๊ฐ์•ก๊ณผ ์ฐจ์ด๊ฐ€ ๋งŽ์ด ๋‚œ๋‹ค๋Š” ๊ฒƒ์€ ์‹ค์ฆ ESS ์šด์˜ ๊ณ„ํš ๋ชจ๋ธ๊ณผ ์ƒ์ดํ•œ ์šด์˜ ๊ณ„ํš ๋ชจ๋ธ์ž„์„ ํ™•์ธํ•  ์ˆ˜ ์žˆ๋‹ค. ์ด๋ฅผ ๋‹ค๋ฅด๊ฒŒ ํ•ด์„ํ•˜๋ฉด ์‹ค์ฆ ESS ์šด์˜ ์‹œ ์ ์šฉํ•˜๊ณ  ์žˆ๋Š” ์ตœ์ ํ™”๊ธฐ๋ฒ• ๊ธฐ๋ฐ˜ ์šด์˜ ๊ณ„ํš ๋ชจ๋ธ๊ณผ ๊ฐ€์žฅ ์œ ์‚ฌํ•œ ๋ชจ๋ธ์ด ๊ธฐ๊ณ„ํ•™์Šต ๊ธฐ๋ฒ•์„ ์ ์šฉํ•œ ๋ชจ๋ธ์ž„์„ ๋œปํ•œ๋‹ค.

4. ๊ฒฐ ๋ก 

๋งˆ์ดํฌ๋กœ๊ทธ๋ฆฌ๋“œ ์šด์˜ ์‹œ ๊ธฐ์ˆ ์ ์œผ๋กœ ๊ฐ€์žฅ ํฐ ๋ฌธ์ œ์ ์ธ ์‹œ์Šคํ…œ์˜ ๋น„์„ ํ˜•์„ฑ, ๋ณต์žก๋„ ์ฆ๊ฐ€๋กœ ์ธํ•ด ๋ฐœ์ƒํ•  ์ˆ˜ ์žˆ๋Š” ์—ฐ์‚ฐ ์‹œ๊ฐ„ ์ฆ๊ฐ€, ์ˆ˜๋ ด์„ฑ ์ €ํ•˜, ๋น…๋ฐ์ดํ„ฐ ์ฒ˜๋ฆฌ์˜ ์–ด๋ ค์›€๊ณผ ๊ฐ™์ด ๊ธฐ์กด์˜ ์ตœ์ ํ™” ๊ธฐ๋ฐ˜ ๋งˆ์ดํฌ๋กœ๊ทธ๋ฆฌ๋“œ ์šด์˜ ์‹œ ๋ฐœ์ƒํ•˜๋Š” ๋ฌธ์ œ๋ฅผ ์ธ๊ณต์ง€๋Šฅ ๊ธฐ์ˆ ์„ ๋„์ž…ํ•˜์—ฌ ํ•ด๊ฒฐํ•˜๊ณ ์ž ํ•˜์˜€๋‹ค.

๋ณธ ๋…ผ๋ฌธ์—์„œ๋Š” ๋””์ง€ํ„ธ ํŠธ์œˆ ๊ธฐ์ˆ ์„ ์ ์šฉํ•˜์—ฌ ๊ฐ€์ƒ๊ณต๊ฐ„์— ์‹ค์ œ ๋งˆ์ดํฌ๋กœ๊ทธ๋ฆฌ๋“œ์— ๊ตฌ์ถ•๋˜์–ด ์žˆ๋Š” ์‹ค์ฆ ESS์˜ ํ™˜๊ฒฝ๊ณผ ํŠน์„ฑ์„ ์ •ํ™•ํ•˜๊ฒŒ ๋ชจ์‚ฌํ•œ ๊ฐ€์ƒ ๋ชจ๋ธ์„ ๊ตฌํ˜„ํ•˜์˜€๋‹ค. ์„ ํ˜•ํšŒ๊ท€ ๋ชจ๋ธ๊ณผ ๊ธฐ๊ณ„ํ•™์Šต ๋ชจ๋ธ์„ ์ ์šฉํ•˜์—ฌ ESS ์šด์˜์— ํ•„์š”ํ•œ ์ต์ผ 1์‹œ๊ฐ„ ๋‹จ์œ„ ESS ์ตœ์  ์ถฉ๋ฐฉ์ „๋Ÿ‰ ๊ณ„ํš์„ ์ˆ˜๋ฆฝํ•˜์˜€๊ณ , ์ด๋ฅผ ์‹ค์ œ ESS์— ์ ์šฉํ•˜์—ฌ ๋ชจ๋ธ์˜ ์„ฑ๋Šฅ์„ ํ‰๊ฐ€ํ•˜์˜€๋‹ค. RMSE์™€ MAE๋กœ ๊ฐ ๋ชจ๋ธ์˜ ์„ฑ๋Šฅ์„ ํ‰๊ฐ€ํ•˜์˜€์œผ๋ฉฐ, ๊ฐ€์ƒ ๊ณต๊ฐ„์— ์ ์šฉ์‹œํ‚จ ESS ์ตœ์  ์šด์˜ ๋ชจ๋ธ์˜ ๊ฒฐ๊ณผ์™€ ์ตœ์ ํ™” ๊ธฐ๋ฐ˜์˜ ์‹ค์ฆ ESS ์ตœ์  ์šด์˜ ๊ฒฐ๊ณผ๋ฅผ ๋น„๊ตํ•˜์—ฌ ์ œ์•ˆํ•œ ๋ชจ๋ธ์˜ ์„ฑ๋Šฅ์„ ๊ฒ€์ฆํ•˜์˜€๋‹ค. ๊ธฐ๊ณ„ํ•™์Šต ๊ธฐ๋ฐ˜์˜ ์ตœ์  ์šด์˜ ๊ณ„ํš ๋ชจ๋ธ์ด ์ตœ์ ํ™” ๊ธฐ๋ฐ˜์˜ ์šด์˜ ๊ณ„ํš ๋ชจ๋ธ๊ณผ ์œ ์‚ฌํ•จ์„ ํ™•์ธํ•˜์˜€๊ณ , ๋˜ํ•œ ์ „๋ ฅ๋Ÿ‰์š”๊ธˆ์€ ์„ ํ˜•ํšŒ๊ท€ ๋ชจ๋ธ์ด ๋” ๋งŽ์ด ์ ˆ๊ฐ๋˜์—ˆ์Œ์„ ํ™•์ธํ•  ์ˆ˜ ์žˆ์—ˆ๋‹ค. ์‚ฌ์šฉ์ž์˜ ํŽธ์˜์— ๋”ฐ๋ผ ๊ธฐ๋ฒ•์„ ์ ์ ˆํ•˜๊ฒŒ ์„ ์ •ํ•˜๊ณ  ๋ฐ์ดํ„ฐ๋ฅผ ์ถฉ๋ถ„ํžˆ ํ•™์Šต์‹œ์ผœ ์‚ฌ์šฉํ•œ๋‹ค๋ฉด, ์ตœ์ ํ™” ๊ธฐ๋ฐ˜์˜ ๋ชฉ์ ํ•จ์ˆ˜ ๋ฐ ์ œ์•ฝ์กฐ๊ฑด ์„ค์ • ์—†์ด ๋ฐ์ดํ„ฐ๋งŒ์œผ๋กœ๋„ ์‹ค์ฆ ESS๋ฅผ ์šด์˜ํ•œ ๊ฒฐ๊ณผ์™€ ์œ ์‚ฌํ•˜๊ฒŒ ESS๋ฅผ ์šด์˜ํ•  ์ˆ˜ ์žˆ์„ ๊ฒƒ์œผ๋กœ ์‚ฌ๋ฃŒ๋œ๋‹ค.

์ตœ์ ํ™” ๋ชจ๋ธ์„ ํ™œ์šฉํ•˜์—ฌ ESS๋ฅผ ์šด์˜ํ•  ๊ฒฝ์šฐ ์˜ฌ๋ฐ”๋ฅธ ๋ชฉ์ ํ•จ์ˆ˜ ์ •์˜ ๋ฐ ์ œ์•ฝ์กฐ๊ฑด์„ ์„ค์ •ํ•ด์•ผ ๋ฌธ์ œ ํ•ด๊ฒฐ์ด ๊ฐ€๋Šฅํ•˜์ง€๋งŒ ๋ฐ์ดํ„ฐ ๊ธฐ๋ฐ˜์˜ ๊ธฐ๊ณ„ํ•™์Šต ๋ชจ๋ธ์„ ํ™œ์šฉํ•  ๊ฒฝ์šฐ์—๋Š” ๋ฐ˜๋ณต์„ ๊ฑฐ๋“ญํ•˜์—ฌ ํ•™์Šตํ•˜๋ฏ€๋กœ ๋ชฉ์ ํ•จ์ˆ˜ ๋ฐ ์ œ์•ฝ์กฐ๊ฑด ์„ค์ •๊ณผ ๊ฐ™์€ ์–ด๋ ค์›€์„ ํ•ด๊ฒฐํ•  ์ˆ˜ ์žˆ๋‹ค. ๋˜ํ•œ ๋””์ง€ํ„ธ ํŠธ์œˆ ๊ธฐ์ˆ ์„ ํ†ตํ•ด ๊ฐ€์ƒ๊ณต๊ฐ„์—์„œ ์‹œ๋ฎฌ๋ ˆ์ด์…˜์„ ํ†ตํ•ด ์–ป์€ ๊ฒฐ๊ณผ๋ฅผ ์‹ค์ œ ์‹œ์Šคํ…œ์— ์ ์šฉํ•˜๊ธฐ ์ „ ๊ฒ€์ฆ์ด ๊ฐ€๋Šฅํ•˜๋‹ค. ์ด๋ฅผ ํ†ตํ•ด ์‹œ์Šคํ…œ ์šด์˜์— ์žˆ์–ด์„œ ์˜์‚ฌ๊ฒฐ์ • ์ˆ˜ํ–‰ ์‹œ ์‹œ์Šคํ…œ์˜ ๋ถˆ์•ˆ์ •์„ฑ ๋ฐ ๋ถˆํ™•์‹ค์„ฑ์„ ํ•ด์†Œํ•  ์ˆ˜ ์žˆ๋‹ค.

ํ–ฅํ›„ ์ง€์†์ ์ธ ๋ฐ์ดํ„ฐ ์ถ•์ ์„ ํ†ตํ•ด ๋ชจ๋ธ์˜ ์˜ˆ์ธก ๋ณ€์ˆ˜ ๋˜๋Š” ํŒŒ๋ผ๋ฏธํ„ฐ ๋“ฑ์„ ๋ณ€ํ™”์‹œ์ผœ ๋ชจ๋ธ์˜ ์„ฑ๋Šฅ์„ ๋”์šฑ ๊ณ ๋„ํ™”ํ•  ์˜ˆ์ •์ด๋‹ค. ๋” ๋‚˜์•„๊ฐ€ ๋””์ง€ํ„ธ ํŠธ์œˆ ๋Œ€์ƒ์„ ESS ๋ฟ ์•„๋‹ˆ๋ผ ๋‹ค์–‘ํ•œ ๋ถ„์‚ฐ์ž์›์„ ํฌํ•จํ•œ ๋งˆ์ดํฌ๋กœ๊ทธ๋ฆฌ๋“œ๋กœ ํ™•๋Œ€ํ•˜์—ฌ ์šด์˜ ๊ธฐ์ˆ ์„ ์—ฐ๊ตฌํ•  ๊ณ„ํš์ด๋‹ค. ์‹ค์ฆ ๋งˆ์ดํฌ๋กœ๊ทธ๋ฆฌ๋“œ์˜ ์ƒํƒœ ๋ฐ ์šด์˜ ํŠน์„ฑ์„ ์ •ํ™•ํ•˜๊ฒŒ ํŒŒ์•…ํ•˜๊ณ  ์‹ค์‹œ๊ฐ„์œผ๋กœ ์ •๋ณด๋ฅผ ์ˆ˜์ง‘ํ•˜์—ฌ ๊ฐ€์ƒ๊ณต๊ฐ„์— ๋งˆ์ดํฌ๋กœ๊ทธ๋ฆฌ๋“œ๋ฅผ ๋””์ง€ํ„ธ ํŠธ์œˆ์„ ๊ตฌ์ถ•ํ•˜์—ฌ ๋งˆ์ดํฌ๋กœ๊ทธ๋ฆฌ๋“œ ์šด์˜์˜ ํšจ์œจ์„ฑ์„ ํ–ฅ์ƒ์‹œํ‚ฌ ๊ฒƒ์œผ๋กœ ๊ธฐ๋Œ€ํ•œ๋‹ค.

Acknowledgements

This research was supported by Korea Institute of Energy Technology Evaluation and Planning(KETEP) grant funded by the Korea government(MOTIE)(No.20212020900530, Development and Demonstration of Cloud Energy Management System for Distributed Buildings).

References

1 
Korean relevant Ministryannounced, Jul 14, 2020, The Korean New Deal, related website https://www.knewdeal.go.kr/Google Search
2 
The technical definition and detailed evolution of the digital twin 5-stage model, Institute of Information & Communications Technology Planning & Evaluation website noticed Feb 10, 2021, https://m.iitp.kr/kr/1/knowledge/ periodicalViewA.it?masterCode=publication&searClassCode=B_ITA_01&identifier=1166Google Search
3 
H.A. Park, G.S. Byeon, W.B. Son, J.Y. Kim, 2021, Digital Twin-based ESS optimal operating technology using Machine Learning, The Korean Institute of Electrical Engineers Conf, pp. 766-797Google Search
4 
S.K. Kim, K.H. Cho, J.Y. Kim, G.S. Byeon, 2019, Field study on operational performance and economics of lithium-polymer and lead-acid battery systems for consumer load management, Renewable and Sustainable Energy Reviews, Vol. 113, No. 109234DOI
5 
H.A. Park, G.S. Byeon, W. Son, H.C. Jo, J.Y. Kim, 2020, Digital Twin for Operation of Microgrid: Optimal Scheduling in Virtual Space of Digital Twin, Energies, Vol. 13, No. 20, pp. 5504DOI
6 
H.A. Park, Y.R. Lee, W. Son, J.Y. Kim, 2020, Data-driven Load Forecasting Using NARX Networks, The Korean Institute of Electrical Engineers Conf, pp. 903-904Google Search
7 
C.J. Willmott, K. Matsuura , 2005, Advantages of the mean absolute error (MAE) over the root mean square error (RMSE) in assessing average model performance, Climate research, Vol. 30, No. 1, pp. 79-82DOI
8 
B. Jaeger, 2006, The method of least squares, Hand book of research on informatics in healthcare and biomedicine, Vol. igi global, pp. 181-185Google Search
9 
H.A. Park, W. Son, J.Y. Kim, K.H. Cho, E.S. Kim, J.H. Jeon, 2019, Improving the performance of industrial load prediction models using optimized weighting factor of WMA, The Korean Institute of Electrical Engineers Conf, pp. 85-86Google Search

์ €์ž์†Œ๊ฐœ

๋ฐ•ํ–ฅ์•„(Hyang-A Park)
../../Resources/kiee/KIEE.2022.71.6.819/au1.png

She received her M.S. degrees in department of electrical engineering from Pusan National University, South Korea in 2015.

Currently, she is pursing the Ph.D. degree in department of electrical engineering from Pusan National University, South Korea.

She is also working at Energy Platform Research Center, Korea Electrotechnology Research Institute.

๋ณ€๊ธธ์„ฑ(Gil-Sung Byeon)
../../Resources/kiee/KIEE.2022.71.6.819/au2.png

He received the M.S and Ph.D. degree in department of electrical engineering form the Korea National University, South Korea in 2006, 2013.

He is currently a Principal Researcher at Energy Platform Research Center, Korea Electrotechnology Research Institute.

๊น€์ข…์œจ(Jong-Yul Kim)
../../Resources/kiee/KIEE.2022.71.6.819/au3.png

He received the M.S and Ph.D. degree in department of electrical engineering form the Pusan National University, South Korea in 1999, 2011.

He is currently a Director and Principal Researcher at Energy Platform Research Center, Korea Electrotechnology Research Institute.

๊น€์„ฑ์‹ (Sung-Shin Kim)
../../Resources/kiee/KIEE.2022.71.6.819/au4.png

He received the M.S. degree in department of electrical engineering from Yonsei University, South Korea in 1986.

He received the Ph.D. degree in department of electrical engineering from Georgia Institute of Technology, U.S.A. in 1996.

He is currently a Professor at the School of Electrical Engineering, Pusan National University, South Korea.

His research interests include Intelligent System, Intelligent Robot, Fault Diagnosis and Prediction.