Foundations — Chaotic systems — sensitivity to initial conditions, Lyapunov exponents

2.1.25 · D1 · Physics › Analytical Mechanics › Chaotic systems — sensitivity to initial conditions, Lyapuno

Isse pehle ki tum parent topic ka ek bhi formula padh sako, tumhe har symbol apna banana hoga. Yeh page har ek symbol ko bilkul zero se build karta hai — pehle plain words mein, phir ek picture, phir yeh topic uske bina kyun nahi chal sakta. Upar se neeche padho; har block uske upar wale par depend karta hai.

0. "State" aur "Trajectory" — woh stage jahan sab kuch hota hai

Yahan $x$ "ek line par ek distance" nahi hai. Yeh ek point hai jo ek saath kai dimensions mein live kar sakta hai. Agar ball ko position aur velocity dono chahiye, toh $x$ ek pair hai $(\text{position},\text{velocity})$ — ek 2-D room mein ek dot.

Figure s01 — state ek dot hai, uski motion ek curve hai. Neeche ki picture position (horizontal) ko velocity (vertical) ke against plot karti hai; orange dot $t=0$ par state hai, green dot wahi dot baad mein hai, aur blue curve woh trajectory hai jo woh sweep out karta hai. Alt text: ek spiral blue curve jisme ek labelled starting dot aur ek baad ka dot hai.

Topic ko yeh kyun chahiye. Poora subject do trajectories ke baare mein hai jo almost ek doosre ke upar se start karti hain. Agar tum state ko ek moving dot ki tarah ek curve ke saath picture nahi karte, toh "nearby trajectories separate" phrase meaningless hai. States ka yeh room-of-states exactly wohi hai jo Phase space and flows study karta hai.

1. $t$ — time, aur "$t\to \infty$" kya pooch raha hai

Tum ${\lim }_{t\to \infty }$ dekhoge. Plain words mein iska matlab hai: "motion ko hamesha ke liye follow karo aur poochho ki woh number kis par settle hota hai." Yeh koi mystical infinity nahi hai — yeh hai "itna lambe waqt tak ruko ki ek lucky starting spot matter karna band kar de."

Topic ko yeh kyun chahiye. Lyapunov exponent ko system describe karna chahiye, ek specific 5-second clip ko nahi. Saare time par average karna ($t\to \infty$) is accident ko wash out kar deta hai ki tum kahan se start kiye the.

2. $\delta$ (delta) — "ek tiny gap"

Do dots ko ek baal ki duri par release karo. Woh chhota sa arrow draw karo jo ek dot se doosre dot ki taraf point karta hai — uski length $|\delta |$ hai (vertical bars ka matlab hai "ki length", hum inhe aage define karenge). $t=0$ par woh arrow ${\delta }_0$ hai; baad mein woh $\delta (t)$ hai. Kyunki ek state multi-dimensional ho sakti hai, $\delta$ khud ek arrow (ek vector) hai, sirf ek single number nahi — states ke room mein uski ek length aur ek direction dono hain.

Figure s02 — gap arrow time ke saath barhta hai. Do runs (blue aur orange) ek baal ki duri par start karti hain; red double-headed arrows teen moments par gap $\delta$ measure karte hain — left mein tiny ${\delta }_0$, baad mein visibly wider $\delta (t)$. Alt text: do nearly-overlapping curves jo alag ho rahi hain, left se right tak chaudi hoti red gap arrows ke saath.

Topic ko yeh kyun chahiye. Chaos defined hai is ek chhote se arrow ke saath jo hota hai usse. Agar yeh shrink hota hai → shant. Agar yeh barhta hai → chaos. Parent page par har formula $\delta$ ke baare mein ek statement hai.

3. $|\cdot |$ — "ka size", direction ko ignore karte hue

4. $f$ aur $\dot{x}$ — woh rule jo dot ko push karta hai

Figure s03 — rule har point par ek arrow hai. Gray arrows $f(x)$ dikhate hain: har location par, rulebook exactly ek push prescribe karta hai. Red dot ek balance point hai jahan $f=0$ (koi push nahi). Alt text: chhote gray arrows ki ek grid jo ek central red dot ki taraf spiral karti hai. Yeh arrow-field flow hai.

Topic ko yeh kyun chahiye. Yeh equation wohi hai jo system ko deterministic banati hai: same position $x$ ⇒ same push $f(x)$ ⇒ same future. Determinism "deterministic yet unpredictable" ka surprising half hai. Chhote arrows ka field Phase space and flows mein flow hai, aur uske balance points Fixed points and linear stability analysis mein study kiye jaate hain.

5. $f^{\prime }(x)$ — derivative, yaani "local stretch factor"

Ab sabse important newcomer. Jab hum poochhte hain ki ek gap kaise barhta hai toh ek derivative kyun aata hai?

Number padhna:

• $|f^{\prime }|>1$ → nearby points alag push ho jaate hain (stretch).
• $|f^{\prime }|<1$ → nearby points aapas mein khinch jaate hain (squeeze).
• $|f^{\prime }|=1$ → gap unchanged.

Figure s04 — derivative ek multiplier ki tarah. Ek chhota green "input nudge" ek red "output change" produce karta hai jo $f^{\prime }(x)$ times nudge ke barabar hai; orange line blue rule ki local slope hai. Alt text: ek curve apni tangent line ke saath, ek horizontal green input arrow aur ek vertical red output arrow.

Topic ko yeh kyun chahiye. Parent ki derivation ka Step 2, $\dot{\delta }\approx f^{\prime }(x)\delta$ (ya higher dimensions mein $Df(x)\delta$), sirf "gap ki growth rate = local stretch factor × gap" hai. $f^{\prime }$ ko ek stretch multiplier ki tarah jaane bina, woh line hieroglyphics hai.

6. $e^{\lambda t}$ — exponential, aur sirf yeh kyun fit karta hai

• Agar $\lambda >0$: $e^{\lambda t}$ upar explode karta hai → gap chaura → chaos.
• Agar $\lambda =0$: $e^0=1$, gap wahi rehta hai → periodic/marginal.
• Agar $\lambda <0$: $e^{\lambda t}$ zero ki taraf decay karta hai → gap khatam → calm attractor.

7. $\lambda$ — Lyapunov exponent, aakhirkar package kiya gaya

Hum ise do moves se extract karte hain jo tum ab samajhte ho: $\lambda ={\lim }_{t\to \infty }{\lim }_{{\delta }_0\to 0}\frac{1}{t}\ln \frac{|\delta (t)|}{|{\delta }_0|}.$

Do limits, words mein:

• Pehle ${\delta }_0\to 0$: starting gap ko infinitely tiny rakho taaki "only-the-slope-matters" approximation ($f^{\prime }$) honest rahe.
• Phir $t\to \infty$: poore journey par average karo taaki $\lambda$ system ko describe kare, kisi ek lucky spot ko nahi.

8. $\Delta$ — tolerance, aur predictability time

Ek baar jab tum $\Delta$ choose kar lete ho, predictability time sirf "kitni der lagti hai barhte gap ko tumhara threshold reach karne mein" hai: $\Delta ={\delta }_0{e}^{\lambda t_{\text{pred}}}\Rightarrow t_{\text{pred}}\approx \frac{1}{\lambda }\ln \frac{\Delta }{{\delta }_0}.$ Yeh exponential ko invert karta hai ($\ln$ use karke) woh time $t_{\text{pred}}$ solve karne ke liye jis par $\delta$ pehli baar $\Delta$ hit karta hai. Kyunki $\ln$ itna slowly barhta hai, tumhara starting error ${\delta }_0$ ek hazaar guna shrink karna $t_{\text{pred}}$ ko barely nudge karta hai — "weather ~2 weeks ke baad hopeless hai" ke peeche woh brutal logarithm.

9. Sab kuch saath laana — prerequisite map

Aage sab kuch — SDIC, predictability time $t_{\text{pred}}=\frac{1}{\lambda }\ln (\Delta /{\delta }_0)$, Lyapunov spectrum, strange attractors — in boxes se bana hai. Stretch-aur-shrink balance directly Liouville's theorem (phase-space volume) aur Strange attractors and fractal dimension mein feed karta hai; discrete cousin $|f^{\prime }(x_n)|$ The logistic map and period-doubling route to chaos ko power deta hai; aur bounded-yet-chaotic picture Driven damped pendulum aur Hamiltonian chaos and the KAM theorem mein appear karta hai.

Equipment checklist

Khud test karo — reveal karne se pehle jawab zor se bolo.