4.7.3 · HinglishPartial Differential Equations

Fourier series — motivation from periodic functions

1,752 words8 min readRead in English

4.7.3 · Maths › Partial Differential Equations


HUM YEH CHAHTE HI KYUN HAIN?

Periodic function KIYA HOTA HAI?

Sabse simple periodic functions aur hain. Ye sab period share karte hain (ye ek period ke andar pure cycles fit karte hain). Isliye ye period ki kisi bhi cheez ke liye perfect Lego bricks hain.


SERIES KAISE BANATE HAIN? (Derivation from scratch)

Hum claim karte hain ki period ki ek periodic function (to ) ko likha ja sakta hai:

Poora problem yeh hai: numbers kya hain? Hum inhe derive karte hain, yaad nahi karte.

Step 1 — Orthogonality trick (WHY it works)

Jinke integrals chahiye (sab se tak):

Ye sach kyun hain? Product-to-sum use karo, jaise . Har piece hota hai integer ke saath; par ek poori number of cycles integrate karne se milta hai, siwaaye jab (ek constant) ho, jo survive karta hai.

Step 2 — nikaalte hain (HOW)

Series ko se multiply karo aur se tak integrate karo:

Yeh step kyun? Orthogonality har term ko khatam kar deti hai siwaaye us ek cosine term ke, jo bachata hai. Isliye:

ki choice se = average se double, isliye = average value. Neat.

Step 3 — nikaalte hain

Wahi trick ke saath:

Figure — Fourier series — motivation from periodic functions

Symmetry shortcut (80/20 ka fayda)


Worked Examples


Forecast-then-Verify

Recall Aage padhne se pehle forecast karo

Q: Ek even triangle wave ke liye, compute karne se pehle — kaun se coefficients zero hain, aur kyun?

A: Sab , kyunki function even hai aur sines odd hain; sirf cosines (jo even symmetry match karte hain) aate hain. Symmetry rule se verify hota hai.


Common Mistakes (Steel-manned)


Flashcards

ka period ke saath periodic hone ka kya matlab hai?
sab ke liye; aisa sabse chhota fundamental period hai.
Woh kaunsi ek property hai sines/cosines ki jo Fourier coefficients extract karne deti hai?
Orthogonality: ek period par alag-alag sines/cosines ke products ke integrals zero hote hain.
ka formula (period )?
.
ka formula (period )?
.
Constant term kyun likha jaata hai?
Taaki baaki jaala same formula use kare; tab ka matlab ka average hota hai.
Agar even hai, to kaun se coefficients vanish hote hain?
Sab (sirf cosines bachte hain).
Agar odd hai, to kaun se coefficients vanish hote hain?
Sab (sirf sines bachte hain).
Jump discontinuity par Fourier series kahan converge karti hai?
Midpoint par.
Gibbs phenomenon kya hai?
Jump ke paas partial sums ka persistent ~9% overshoot jo terms badhne par bhi nahi jaata.
Square wave (±1, period ) ki Fourier series?
— sirf odd harmonics.
PDE solvers ko Fourier series ki zaroorat kyun hai?
Separation of variables sine/cosine modes deta hai; arbitrary initial condition ko match karne ke liye use un modes mein expand karna padta hai.

Recall Feynman: ek 12-saal ke bacche ko samjhao

Socho tumhare paas ek weird wiggly line hai jo forever repeat hoti hai, jaise heartbeat monitor. Tumhare paas ek "pure musical notes" ka box bhi hai, har ek ek smooth wave. Fourier ki amazing discovery: us weird wiggle ko un pure notes ka sahi mix bajaakar perfectly recreate kar sakte ho — kuch loud, kuch soft, kuch high-pitched, kuch low. Mix ki "recipe card" ek clever sliding-and-adding trick (integration) se milti hai jo poochti hai "is note ka kitna hissa meri wiggle mein chhupa hai?" Kyunki notes kabhi ek doosre ke saath confuse nahi hote (orthogonality), har sawaal ka ek clean jawab milta hai.


Connections

  • Separation of Variables — woh sine/cosine modes produce karta hai jo Fourier series assemble karti hai.
  • Heat Equation — initial temperature profile ko Fourier (sine) series ke roop mein expand kiya jaata hai.
  • Wave Equation — plucked-string ki shape harmonics mein decompose hoti hai.
  • Orthogonality of Functions — inner-product structure .
  • Even and Odd Functions — coefficients ke liye symmetry shortcuts.
  • Fourier Transform — Fourier series ki limit jab period ho.
  • Gibbs Phenomenon — jumps par convergence behaviour.

Concept Map

separation of variables

must match

needs

expresses

defined by

stacks

share

coefficients a_n b_n

found via

integrate product over period

proven by

multiply and integrate

Solve PDEs on interval

Solutions are sines and cosines

Arbitrary initial condition f x

Fourier series

Periodic function period 2L

f x+T = f x

Pure waves cos and sin

Unknown numbers

Orthogonality trick

Different waves give 0

Product-to-sum identities