4 minute read

๋ณต์ˆ˜์ „๊ณตํ•˜๊ณ  ์žˆ๋Š” ์ˆ˜ํ•™๊ณผ์˜ ์กธ์—…์‹œํ—˜์„ ์œ„ํ•ด ํ•™๋ถ€ ์ˆ˜ํ•™ ๊ณผ๋ชฉ๋“ค์„ ๋‹ค์‹œ ๊ณต๋ถ€ํ•˜๊ณ  ์žˆ์Šต๋‹ˆ๋‹ค๋งŒโ€ฆ ๋ฏธ๋ถ„๋ฐฉ์ •์‹์€ ์กธ์—…์‹œํ—˜ ๋Œ€์ƒ ๊ณผ๋ชฉ์ด ์•„๋‹ˆ๋ผ๋Š” ๊ฑธ ๋‚˜์ค‘์— ์•Œ๊ฒŒ ๋˜์—ˆ์Šต๋‹ˆ๋‹คโ€ฆ OTLโ€ฆ ๊ทธ๋ž˜๋„ ์ด์™• ์‹œ์ž‘ํ•œ ๊ฑฐ ๋‹ค์‹œ ๋ณต์Šต ์ข€ ํ•ด๋ด…์‹œ๋‹ค! ๐Ÿƒ ๋ฏธ๋ถ„๋ฐฉ์ •์‹ ํฌ์ŠคํŠธ ์ „์ฒด ๋ณด๊ธฐ

Example 1

Draw the phase portrait of $xโ€™ = A x$ where

\[A = \left( \begin{matrix} 2 & 1 \\ 4 & 2 \end{matrix} \right)\]

.

๋จผ์ €, eigen value๋ฅผ ๊ตฌํ•ด๋ณด์ž.

\[(2 - \lambda)(2-\lambda) - 4 = 0\]

์ด๋ฏ€๋กœ, $\lambda_1 = 4$, $\lambda_2 = 0$์ด ๋œ๋‹ค.

$\lambda_1 = 4$์˜ eigen value๋Š” ํ•˜๋˜๋Œ€๋กœ ๊ตฌํ•˜๋ฉด ๋œ๋‹ค: $v_1 = (1, 2)$.

$\lambda_2 = 0$๋ฅผ ์ฒ˜์Œ ๋ด์„œ ์ข€ ๋‹นํ™ฉํ–ˆ๋Š”๋ฐ, eigen value๊ฐ€ 0์ผ์ง€๋‹ค๋กœ eigen vector๋Š” ๊ตฌํ•  ์ˆ˜ ์žˆ๋‹ค. $A v = 0 \cdot v = 0$์˜ ์‹์„ ํ’€๋ฉด ๋˜๊ธฐ ๋•Œ๋ฌธ์ด๋‹ค. ์ด ๊ฒฝ์šฐ๋Š” $(1, -2)$๋กœ ๋‚˜์˜จ๋‹ค.

๊ทธ๋ฆฌ๊ณ  general solution์„ ๊ตฌํ•˜๋ฉด ์•„๋ž˜์™€ ๊ฐ™๋‹ค.

\[x(t) = c_1 \left( \begin{matrix} 1 \\ 2 \end{matrix} \right) e^{4t} + c_2 \left( \begin{matrix} 1 \\ -2 \end{matrix} \right) e^{0t} = c_1 \left( \begin{matrix} 1 \\ 2 \end{matrix} \right) e^{4t} + c_2 \left( \begin{matrix} 1 \\ -2 \end{matrix} \right)\]

์š” solution์— ๋Œ€ํ•œ Phase Portrait์„ ๊ทธ๋ ค๋ณด๋ฉด ์•„๋ž˜์™€ ๊ฐ™๋‹ค.

$\lambda_1 = 4$์ด ์–‘์ˆ˜์ด๋ฏ€๋กœ ๋ฒกํ„ฐ $v_1$ ์œ„์—์„œ ์›์ ์— ๋Œ€ํ•ด ๋‚˜๊ฐ€๋Š” ๋ฐฉํ–ฅ์œผ๋กœ ๊ถค์ ์ด ์›€์ง์ธ๋‹ค. ๋ฐ˜๋ฉด์— ๋ฒกํ„ฐ $v_2$์—์„œ๋Š” $t$์— ๋Œ€ํ•œ ๋ถ€๋ถ„์ด ์—†๊ณ , ๋‹จ์ˆœํžˆ $v_1$ ๊ถค์ ์„ ํ‰ํ–‰์ด๋™ ์‹œํ‚ค๋Š” ์—ญํ• ์„ ํ•œ๋‹ค.

์ด๋•Œ, $v_2$๊ฐ€ ํ‰ํ–‰์ด๋™ํ•  ๋•Œ, ์›์  $O$๋Š” $v_2$ ์ง์„  ์œ„๋ฅผ ์›€์ง์ด๋‹ค. ๋”ฐ๋ผ์„œ, Phase Portrait์ด ์œ„์˜ ๊ทธ๋ฆผ๊ณผ ๊ฐ™์ด ์Šฌ๋ผ์ด์ด๋”ฉ? ํ•˜๋Š” ๊ฒƒ์ฒ˜๋Ÿผ ๋ณด์ธ๋‹ค. ์ด๋•Œ, $v_2$๋ฅผ ํ‰ํ˜•(equilibrium)์ด๋ผ๊ณ  ํ•˜๊ณ , ์ด๋ ‡๊ฒŒ ํ‰ํ˜•์—์„œ ๋ฉ€์–ด์ง€๋Š” ๋ฐฉํ–ฅ์œผ๋กœ ์›€์ง์ด๋Š” ๊ฒƒ์„ unstable์ด๋ผ๊ณ  ํ•œ๋‹ค.

Example 2 - Consider bifurcation

Draw the phase portrait of $xโ€™ = A x$ where

\[A = \left( \begin{matrix} a & 1 \\ 2a & 2 \end{matrix} \right)\]

.

And find a bifurcation about $a$.

์ด๋ฒˆ์—๋„ eigen value๋ฅผ ๊ตฌํ•ด๋ณด๋ฉด, ํ•˜๋‚˜๊ฐ€ 0์œผ๋กœ ๋‚˜์˜จ๋‹ค.

\[(a - \lambda)(2 - \lambda) - 2a = 0\]
  • $\lambda_1 = a + 2$
  • $\lambda_2 = 0$

๊ทธ๋ฆฌ๊ณ  ๋Œ€์‘ํ•˜๋Š” eigen vector๋ฅผ ๊ตฌํ•ด๋ณด๋ฉด,

  • $\lambda_1 = a + 2$
    • $v_1 = (1, 2)$
  • $\lambda_2 = 0$
    • $v_2 = (1, -2)$

eigen vector๋Š” ์œ„์˜ ์˜ˆ์ œ๋ฅผ ํ’€์—ˆ์„ ๋•Œ์™€ ๋™์ผํ•˜๊ฒŒ ๋‚˜์˜จ๋‹ค!!

๋งค๊ฐœ๋ณ€์ˆ˜ $a$์˜ ๊ฐ’์— ๋”ฐ๋ผ ์—ฌ๋Ÿฌ Phase Portrait๊ฐ€ ๊ทธ๋ ค์ง„๋‹ค: bifurcation์ธ ์…ˆ!


[case 1: $a = -2$]

$\lambda_1$๋„ $0$์ด ๋˜์–ด ๋ฒ„๋ฆฐ๋‹ค!!! ๊ทธ๋Ÿฐ๋ฐ, ๋‹คํ–‰์ธ ์ ์€ $v_1 \ne v_2$์ด๋‹ค. ๋งŒ์•ฝ $v_1 = v_2$ ์˜€๋‹ค๋ฉด, generalized eigen vector๋ฅผ ๊ตฌํ•ด์ค˜์•ผ ํ•œ๋‹ค. (generalized eigen vector๋ฅผ ๊ตฌํ•ด์•ผ ํ•˜๋Š” ๋ฌธ์ œ๋Š” ๋‹ค์Œ ํฌ์ŠคํŠธ์—์„œโ€ฆ)

general solution์€ ์•„๋ž˜์™€ ๊ฐ™๋‹ค.

\[x(t) = c_1 \left( \begin{matrix} 1 \\ 2 \end{matrix} \right) + c_2 \left( \begin{matrix} 1 \\ -2 \end{matrix} \right)\]

์ด๋ฒˆ์—๋Š” ๋ฒกํ„ฐ $v_1$, $v_2$์— ๋‘˜๋‹ค $t$์— ๋Œ€ํ•œ ๋ถ€๋ถ„์ด ์—†๋‹ค. ๋”ฐ๋ผ์„œ, ์ ์ด ๊ทธ๋ƒฅ $x_1 - x_2$ plane ์œ„์—์„œ ๊ณ ์ •์ ์œผ๋กœ ์กด์žฌํ•œ๋‹ค.


[case 2: $a > - 2$]

$\lambda_1 > 0$์ด ๋˜๋ฏ€๋กœ, ์œ„์—์„œ ๋ณธ ์˜ˆ์ œ์™€ ๋™์ผํ•œ ๊ฒฝ์šฐ๋‹ค. ์Šคํ‚ต!


[case 3: $a < -2$]

$\lambda_2 < 0$์ด ๋˜๋ฏ€๋กœ, ์œ„์—์„œ ๋ณธ ์˜ˆ์ œ์™€ Phase Portrait์ด ๋™์ผํ•˜์ง€๋งŒ, ํ‰ํ˜•์ธ ์ง์„  $v_2$ ์ชฝ์œผ๋กœ ์ˆ˜๋ ดํ•˜๋ฏ€๋กœ stable ์ผ€์ด์Šค๋‹ค.

Reference