Visual walkthrough — Initial value problems (IVP) vs boundary value problems (BVP)

4.7.2 · D2 · Maths › Partial Differential Equations › Initial value problems (IVP) vs boundary value problems (BVP

Hum sab kuch usi ek chhoti si equation, $y^{\prime \prime }+y=0$, ke around banate hain, kyunki yeh sabse chhoti equation hai jo BVP ke teeno outcomes dikhati hai. Aakhir mein tum dekh paoge ki side-conditions ki location hi poori kahani kyun hai.

---

Step 1 — Ek differential equation poore curves ki ek family hoti hai

KYA HAI. IVP vs BVP ki baareekiyon mein jaane se pehle, dekho ki "$y^{\prime \prime }+y=0$" ka matlab kya hai. Ise zor se padho: "$y$ ki second derivative, $y$ ki negative hai." Ek function jiska bend ($y^{\prime \prime }$) hamesha zero ki taraf wapas point karta hai — exactly yahi ek wave karta hai. Solutions ki poori family hai

$y(x)=\underset{\text{knob 1}}{\underbrace{c_1}}\underset{\text{building block }{y}_1}{\underbrace{\cos x}}+\underset{\text{knob 2}}{\underbrace{c_2}}\underset{\text{building block }{y}_2}{\underbrace{\sin x}}.$

• $\cos x$ aur $\sin x$ do independent building-block waves hain — tum ek ko doosre ko scale karke nahi bana sakte.
• $c_1,c_2$ do free knobs hain (numbers jinhe hum choose kar sakte hain). Inhe ghumaane se wave ka shape aur size badalta hai.

DO knobs kyun. Equation second order hai (sabse bada derivative $y^{\prime \prime }$ hai). Har derivative jise "undo" karna hota hai ek arbitrary constant chhod jaata hai, isliye do derivatives → do constants. Dono knobs lock karne ke liye humein bilkul do facts chahiye honge.

PICTURE. $(c_1,c_2)$ ka har choice ek curve hai. Figure mein unka ek bundle dikhaya gaya hai — ek family.

---

Step 2 — IVP: ek point pin karo, uski slope bhi, aur future forced ho jaata hai

KYA HAI. Ek IVP tumhe sab kuch ek location $x_0$ par deta hai. Hamare second-order equation ke liye iska matlab hai usi jagah par ek value aur ek slope:

$\underset{\text{yahan height}}{\underbrace{y(x_0)}}=A,\underset{\text{yahan slope}}{\underbrace{y^{\prime }(x_0)}}=B.$

$x_0=0$, $A=1$, $B=0$ lo: height $1$ par shuru karo, bilkul flat.

• $y(0)=c_1\cos 0+c_2\sin 0=c_1=1$. (Yahan $\cos 0=1$, $\sin 0=0$ selection karte hain.)
• $y^{\prime }(x)=-c_1\sin x+c_2\cos x$, isliye $y^{\prime }(0)=c_2=0$.

Dono knobs fix → ek curve, $y=\cos x$.

YEH HAMESHA KYUN KAAM KARTA HAI. Ek point par height aur slope jaanna ek car ki abhi ki position aur velocity jaanne jaisa hai — physics ke paas agli instant ke baare mein koi choice nahi. Jo tool yeh guarantee karta hai woh Wronskian hai, aur Picard–Lindelöf woh theorem hai jo kehta hai "haan, bilkul ek."

PICTURE. Ek akela dot jiske saath ek chhoti slope-stick lagi hai; family ka sirf ek hi member dono se hokar jaata hai.

---

Step 3 — "Ek point par position + slope" kabhi ambiguous kyun nahi ho sakta

KYA HAI. Do IVP facts ko family mein feed karo aur tumhe knobs mein ek $2\times 2$ system milta hai. Sab kuch decide karne wali matrix building blocks aur unki slopes se, sab kuch ek hi point $x_0$ par bani hoti hai:

$W(x_0)=\underset{\cos x_0}{\underbrace{y_1(x_0)}}\underset{\cos x_0}{\underbrace{y_2^{\prime }(x_0)}}-\underset{\sin x_0}{\underbrace{y_2(x_0)}}\underset{-\sin x_0}{\underbrace{y_1^{\prime }(x_0)}}={\cos }^2{x}_0+{\sin }^2{x}_0=1.$

Yeh number $W$ Wronskian hai. Dhyan do ki yeh $=1$ nikla, kabhi zero nahi, har point par.

YEH KYUN IMPORTANT HAI. Ek $2\times 2$ system ka exactly ek answer tab hota hai jab uska determinant nonzero ho. Independent building blocks ke liye Wronskian kabhi zero nahi hota — isliye ek IVP hamesha uniquely solvable hota hai. Yeh baat yaad rakho: BVP ek alag matrix use karega jo zero hit kar sakta hai.

PICTURE. Do arrows ("value info" aur "slope info") jo genuinely alag-alag directions mein point karte hain — yeh poore plane ko span karte hain, isliye koi bhi target reachable hai.

---

Step 4 — BVP: do facts ko do alag points par split karo

KYA HAI. Ek BVP tumhe information do alag jagahon $a\ne b$ par deta hai:

$y(a)=\alpha ,y(b)=\beta .$

Koi slope nahi diya jaata — balki hum maang karte hain ki curve do targets hit kare. Family mein plug karne par knobs mein ek aisa system milta hai jiska matrix ab same building blocks se bana hota hai lekin do alag points par evaluate kiya hua:

$$\underset{\text{ise }M\text{ kaho}}{\underbrace{\left(\begin{array}{cc}{y}_1(a)& y_2(a)\\ y_1(b)& y_2(b)\end{array}\right)}}\left(\begin{array}{c}{c}_1\\ c_2\end{array}\right)=\left(\begin{array}{c}\alpha \\ \beta \end{array}\right).$$

• Row 1 kehti hai "$x=a$ par mixed wave $\alpha$ ke barabar honi chahiye."
• Row 2 kehti hai "$x=b$ par yeh $\beta$ ke barabar honi chahiye."

YEH FAIL KYUN HO SAKTA HAI. Wronskian ke unlike, yeh matrix $M$ do locations ko mix karta hai, aur is baat ki koi guarantee nahi ki iska determinant nonzero rahe. Magic number hai

$\det M=y_1(a)y_2(b)-y_2(a)y_1(b).$

PICTURE. Do pins $x=a$ aur $x=b$ par khade hain; humein ek family curve ko dono se guzaarna hai.

---

Step 5 — Case A: $\det M\ne 0$ → bilkul ek curve fit hoti hai

KYA HAI. $a=0$, $b=\frac{\pi }{2}$ (quarter-wave apart), $\alpha =\beta =0$ lo. Calculate karo

$\det M=\cos 0\cdot \sin \frac{\pi }{2}-\sin 0\cdot \cos \frac{\pi }{2}=1\cdot 1-0\cdot 0=1\ne 0.$

Nonzero determinant → system ka ek aur sirf ek answer hai: $c_1=0$, $c_2=0$, yaani flat line $y=0$. Boring, lekin unique — yahi point hai.

KYUN. Jab do pins wave ke upar jaate hisse par lagte hain, to family unke beech itna bend karti hai ki sirf ek hi member dono ko chhoo sakta hai. Wahi reliable behaviour jo IVP hamesha enjoy karta hai.

PICTURE. Do pins quarter-wave apart; ek akeli curve se hokar jaati hai, baaki sab miss kar jaati hain.

---

Step 6 — Case B: $\det M=0$ matching target ke saath → infinitely many

KYA HAI. Ab pins ko half-wave apart rakho: $a=0$, $b=\pi$, dono targets $0$.

$\det M=\cos 0\cdot \sin \pi -\sin 0\cdot \cos \pi =1\cdot 0-0\cdot (-1)=0.$

Kaam karke dekho: $y(0)=c_1=0$, aur $y(\pi )=-c_1=0$ — jo automatically true hai aur $c_2$ ke baare mein kuch nahi kehta. Isliye $c_2$ ek free knob hai:

$y=c_2\sin x\text{kisi bhi }{c}_2\text{ ke liye.}$

KYUN. $\sin x$ $0$ se shuru hota hai aur exactly half wave ke baad $0$ par wapas aata hai. Uski har scaled copy dono boundary demands maanti hai. Yeh bilkul wahi eigenvalue situation hai Sturm-Liouville Theory ki: $\det M=0$ woh equation hai jiske special lengths eigenvalues hain, aur bacha hua $\sin x$ eigenfunction hai.

PICTURE. Alag-alag heights ki $\sin$-waves ka ek fan, sab dono pins se hokar jaate hain.

---

Step 7 — Case C: $\det M=0$ clashing target ke saath → koi solution nahi

KYA HAI. Pins ko half-wave apart rakho lekin doosra target badal do: $a=0$, $b=\pi$, $\alpha =0$, $\beta =1$.

• $y(0)=c_1=0$.
• $y(\pi )=-c_1=0$, lekin humne maanga tha ki yeh $1$ ke barabar ho. Isliye $0=1$. Impossible.

Same singular $\det M=0$, phir bhi family target tak pahunch hi nahi sakti — koi curve exist nahi karti.

KYUN. Har family member jo $x=0$ par $0$ hai woh $x=\pi$ par $0$ par wapas jaane ke liye majboor hai. Usse $1$ par pahunchne ko kehna ek wapas aati wave ko uska khud ka rhythm todne ko kehna hai. Woh nahi kar sakti.

PICTURE. Sabhi candidate curves $x=\pi$ par zero cross karti hain; amber target dot unke upar akela float karta rehta hai.

---

Step 8 — Woh rule jo teeno cases summarize karta hai

KYA HAI. Upar sab kuch ek single decision mein collapse hota hai:

KYUN. Side-conditions ki location hi woh cheez hai jo deciding matrix ko hamesha-safe Wronskian se kabhi-kabhi-singular $M$ mein badal deti hai. Woh single change IVP-vs-BVP ki poori kahani hai. Yahi machinery exactly wahi hai jo Separation of Variables (PDE) Heat Equation aur Wave Equation solve karte waqt feed karti hai, aur isliye Green's Functions ko exist karne ke liye $\det M\ne 0$ chahiye hota hai.

PICTURE. Verdict ka ek flowchart, blueprint style mein.

---

Ek-picture summary

Ek frame, teen pin-configurations, ek equation $y^{\prime \prime }+y=0$: quarter-wave gap (unique), matching ends ke saath half-wave gap (infinite fan), clashing end ke saath half-wave gap (empty target).

Recall Feynman retelling — poora walkthrough seedhe shabdon mein

Humne ek hilti hui curve family se shuru kiya: "bend hamesha ghar ki taraf point karta hai" sines aur cosines deta hai, ghoomane ke liye do dials ke saath. IVP: ek jagah khado, batao kitna uuncha aur kitna steep — curve ke paas koi freedom nahi bachi, isliye exactly ek wave fit hoti hai, hamesha. Iska reason Wronskian naam ka ek number hai jo us ek jagah par bana hota hai aur zyiddi se kabhi zero nahi hota. BVP: steepness ki jagah, tum do flags do alag-alag jagahon par gaado aur maango ki wave dono ko chhue. Ab deciding number ek alag wala hai, $\det M$, same wave-blocks se bana lekin do jagahon par padha gaya. Agar flags quarter-wave apart hain, to wave unke beech itna bend karti hai ki sirf ek fit hoti hai — saaf. Lekin unhe half-wave apart gaado aur sine apne aap zero par wapas aa jaata hai: agar dono flags zero par hain, har height ka sine kaam karta hai (infinitely many); agar ek flag zero se upar uthaa hua hai, koi bhi sine wahan pahunch nahi sakta (none). Same equation, facts ki same count — sirf placement badla, aur yahi IVP aur BVP ke beech ka poora fark hai.

Connections

• Wronskian and Linear Independence — IVP uniqueness ke peeche never-zero number (Steps 3, 8).
• Sturm-Liouville Theory — jahan $\det M=0$ eigenvalues ban jaata hai (Step 6).
• Picard-Lindelöf Existence and Uniqueness Theorem — IVP ki guarantee (Step 2).
• Separation of Variables (PDE), Heat Equation, Wave Equation, Green's Functions — jahan yeh BVP verdicts use hote hain.