• 沒有找到結果。

Riemann-Lebesgue ùÜDÿY¹

N/A
N/A
Protected

Academic year: 2022

Share "Riemann-Lebesgue ùÜDÿY¹ "

Copied!
12
0
0

加載中.... (立即查看全文)

全文

(1)

Riemann-Lebesgue ùÜDÿY¹

Šœþ

‡k:

BbÊ¥¹dıø* Riemann - Lebesgue ùÜ}–, ¥Ê¡H}&2rÆ7

”½b5iH 7(}D…óÉí3æ¨

ìó¶(method of stationary phase) DÿY¹Ô u Young’s¿ 5–1

Riemann - Lebesgue ùÜ5–Ä@ç u Fourier b5Y¹4½æD Fourier  }t, ²Æuzÿuƒ Fourier |‚

à}׉b¶jÏfûj˙, |(Ûbû˝

íL½æ (inverse problem), Ĥªcw½ b4 Bbø#_òh<2í„p, 7(

} semi - classical í–1, |(†ùªÿY

¹D Young’s ¿ Ê¡H}&«wuÝ(

4R}j˙2”½bí–1

1. Riemann - Lebesgue ùÜ

} Riemann-Lebesgue ùÜJ.*U

˚u“bç2í˹”—Fourier bOGÿ .øwŸá<2 #ìøƒb f ∈ C[−π, π]

ø5[Ñ Fourier b f (x) a 0

2 +

 n =1

(a n cos

L x+b n sin L x) (1.1)

 

a n = L 1  −L L f (ξ) cos L ξdξ

b n = L 1  −L L f (ξ) sin L ξdξ (1.2)

‚àúiƒbí¸Ï“†,ª[Ñ

f (x) = 1 2L

 L

−L f (ξ)dξ +

 n =1

1 L

 L

−L f (ξ) cos

L (x −ξ)dξ (1.3) BbÛÊ>Eí½æu: ¦ x ì7I L → ∞ † (1.3) }AÑBóá? çÍ Ñ7\„}íæÊ4, Bbílcq f ∈ L 1 ( R ) ¹



−∞ | f(ξ) | dξ < ∞ (1.4) Ĥ (1.3) AÑ (¤v 2L 1  −L L f (ξ)dξ 0)

f (x) = lim

L →∞

 n =1

1 L

·  L

−L f (ξ) cos

L (x − ξ)dξ (1.5) ÛÊ_" Riemann ¸íj¶I

λ n =

L ∆λ = λ n +1 − λ n = π L I(λ, L) =

 L

−L f (ξ) cos λ(x −ξ)dξ (1.6) Ĥ (1.5) ªZŸÑ

f (x) = lim

L →∞

1 π

 n =1

I(λ n , L)∆λ (1.7)

17

(2)

¥£u Riemann ¸5$ I∆λ → 0 (L → ∞) †

f (x) = lim

L →∞

1 π



0 I(λ, L)dλ

= lim

L →∞

1 π



0

 L

−L f (ξ) cos λ(x −ξ)dξ

= 1 π



0



−∞ f (ξ) cos λ(x −ξ)dξ (1.8)

¥_ÿu|±íFourier}t y‚à¸

Ï“ø (1.8) ZŸÑ f (x) = 1

π



0



−∞ f (ξ) cos λ(x −ξ)dξ

= 1 π



0 [a(λ) cos λx+b(λ) sin λx]dλ (1.9) w2

 

a(λ) =  −∞ f (ξ) cos λξdξ

b(λ) =  −∞ f (ξ) sin λξdξ (1.10)

¥£u Fourier ìý (£ý) ‰² wõÿu Fourier b5R2 çÍBbª‚à Euler tøs6¯9Êø–

f (x) = 1 π



0



−∞ f (ξ) cos λ(x −ξ)dξ

= 1



0



−∞ f (ξ)e (x−ξ) + 1



0



−∞ f (ξ)e −iλ(x−ξ)

= 1



−∞



−∞ f (ξ)e (x−ξ)

=



−∞

f (λ)e ˆ iλx (1.11) w2

f (λ) = ˆ 1



−∞ f (ξ)e −iλξ (1.12)

ÿuƒb f 5 Fourier ‰² Ĥ Fourier

}t5…”ÿuø/L‰² —Fourier L‰²7¥£ujR}j˙, Ô u(4 4[bR}j˙|½bí5x

,Þ¥<Rû¬˙2_u°íÛïÿ u·|Û6-é5}

 b

a

f (x)

cos nx sin nx

dx,

 b

a

f (x)e iλx dx a.b R

ĤÃ7í„p5‡âln6¥é}í 4”, ¥ÿu Riemenn - Lebesgue ùÜ:

Riemann - Lebesgue ùÜ:

 I:(ä–È)

J f ∈ L 1 ([0, 2π]) †

n →±∞ lim 1



0 f (x) cos nxdx

= lim

n →±∞

1



0 f (x) sin nxdx = 0 C[ѵb5$

n →±∞ lim f (n) ˆ

= lim

n →±∞

1



0 f (x)e −inx dx = 0 (1.13)

 II:(Ìä–È)Jf ∈ L 1 ( R ) †

λ →±∞ lim f (λ) ˆ

= lim

λ →±∞

1



−∞ f (x)e −iλx dx = 0 (1.14) â}5©/4Bbyª!6

f ∈ L 1 ( R ) ⇒ ˆ f ∈ C 0 ( R ) (1.15)

C 0 ( R ): [ýF©/ƒbÅ—Ê̤õÑ 0

5Õ¯

(3)

Ék Riemann - Lebesgue ùÜ|o uâ Riemann k 1876 -„p, 7øOí8

$¹ f ∈ L 1 †uâ Lebesgue k 1903 - F#í Ék¥ìÜí„prÖ/, OBb 1Ì<J¥/jú&¥1(/½b5ìÜ Äуb f 5¸ˇØ2 (Bý©/) FJb }&í.u¥á7u cos nx, ÑBóá? B bõõwÇ$, *Síi Vpëw<2

...

...

...

...

...

...

...

...

...

...

...

...

...

...

...

...

....

...

−1 1

...

cos x (1 _ š) Ç ø

...

...

...

...

...

...

...

...

...

...

...

...

...

...

...

...

...

...

...

−1 1

...

...

...

cos 3x (3 _ š) Ç ù

âÇ$ªø cos nx Ñ n _ìýšÑÊ [0, 2π] ¥_–È, Ĥç n → ∞ v ª;díu cos nx Ê 1 D −1 5 È ( ÄÑ | cos nx| ≤ 1) 0§PÓ

...

...

...

...

...

...

...

...

...

...

...

...

...

...

...

...

...

...

...

...

...

...

...

...

...

...

...

...

...

...

...

...

...

... ...

...

...

...

...

...

...

cos nx(n _ š)

· · ·

7Ékw}  0 cos xdx = 0, †Ä Ñ cos x ÑøU‚ 2π íƒb, w}ø

¶MÊ x W5,j, Çø¶M†Ê x W

5-j ¥s¶Mø£øŠ óJ¾, Ĥ



0 cos xdx = 0 ° Üúƒb cos nxø£

øŠÉuÛÊ)òø<, ˛¤ óJ¾, ]



0 cos nxdx = 0 “0” ¥_Mu<2í,

…ÿu cos x íMÌM 1



0 cos xdx = 0

*Ç$,Võ cos nx ¥<ƒbJ x WÑ2 -7,-PÓ, ]w}¹MÌMÑ 0, °Ü úk}



0 f (x) cos nxdx

5n6D‡Þêró°, ñøÏ íuPÙ Z‰Ñ |f(x)| (Ÿlu1)

...

...

...

...

...

...

...

...

...

...

...

...

...

...

...

...

...

...

...

...

...

...

...

...

...

...

...

...

...

...

...

...

...

...

...

...

.

...

...

...

...

...

...

...

... ...

...

...

...

...

...

f (x)

−f(x) Ç ú

¤v}  0 f (x) cos nxdx . c)bÑ 0 Oç n → ∞ v£íDŠí¶M˛˛¤J

¾, Ĥ¯Üí“¿u

n lim →∞



0 f (x) cos nxdx = 0 à‹ cos x H²Ñ sin x, 6ó°í!‹

n lim →∞



0 f (x) sin nxdx = 0

¥ÿu Riemann - Lebesgue ùÜ

(4)

ìÜ„p:

M Bbø} J|1zpwÊbç2&, í½b4

(1) Riemann ¸:

ʇÞín62Bb˛%ø−Ê–È [0, 2π], cos nx â n _ìýšF A, Ĥ

Bbø [0, 2π] }A n } (¥_–1ÿu Riemann ¸)

...

...

...

...

...

...

..

0

...

...

...

...

...

..

...

...

...

n

...

...

...

n

...

...

...

2(n−1)π n

Ç û



0  n

k =1



2k n

π

2(k−1)

n

π

(1.16) Ĥ}ª[Ñ



0 f (x) cos nxdx

=

 n k =1



2k n

π

2(k−1)

n

π

f (x) cos nxdx

=

 n k =1

 2kπ

2(k−1)π f ( y

n )(cos y) 1 n dy (x = y

n )

= 1 n

 n k =1



0 f (ξ + 2(k − 1)π

n ) cos ξdξ (y = ξ + 2(k − 1)π)

= 1



0 cos ξ

·(  n

k =1

f ( ξ + 2(k − 1)π

n ).

n )dξ O f Ñ©/ƒb, w Riemann Å—

 n k =1

f ( ξ + 2(k − 1)π

n )

n



0 f (x)dx, n → ∞



0 f (x) cos nxdx

→ ( 1



0 cos ξdξ)(



0 f (x)dx)

= 0 (1.17)

Ê„pí¬˙2ªJüìíu}í”

ÌMkÉ1Ý.Íí (44Bb}¥ó“), 7uÄ Ñ cos ξ 5MÌM (mean) ÑÉíí ] à‹J cos 2 x H  cos x ª)



0 f (x) cos 2 nxdx

→ ( 1



0 cos 2 ξdξ)(



0 f (x)dx)

= 1 2



0 f (x)dx (1.18)

çÍ6ª‚àìÜí!‹V„p



0 f (x) cos 2 nxdx

=



0 f (x) 1 + cos 2nx 2 dx

= 1 2



0 f (x)dx+ 1 2



0 f (x) cos 2nxdx

1 2



0 f (x)dx + 0 = 1 2



0 f (x)dx BbêÛøKí9õ:

( lim

n →∞



0 f (x) cos nxdx) 2

= lim n →∞ 

0 f (x) cos 2 nxdx (1.19)

BbÊ|(øJy‚àÿY¹ (weak con-

vergence) íh1V}¤Ûï

(5)

Ê,Hí„p¬˙2, Bbcàƒ cos x ÑøU‚Ñ 2π íƒb, ĤªòQR2BL

<U‚°Ñ 2π í©/ƒb g(x) :

n lim →∞



0 f (x)g(nx)dx

= 1



0 g(x)dx 

0 f (x)dx (1.20)

¤v cos x â g(x) ¦H; cos nx †â g(nx) ≡ g n (x) ¦H



0 f (x)g n (x)dx

=



0 f (x)g(nx)dx

=

 n k =1



2k

n

π

2(k−1)

n

π

f (x)g(nx)dx

= 1 n

 n k =1



0 f ( ξ +2(k −1)π

n )g(ξ)dξ

= 1



0 g(ξ)(

 n k =1

f ( ξ +2(k −1)π

n )

n )dξ

1



0 g(ξ)dξ · 

0 f (x)dx (2) Weierstrass V¡ìÜ:

¥_ìܵsBb“ì2Ê'K–Èí

©/ƒbÏ.ÖÿuÖáƒb”, ĤBb ªJltt

f (x) = x k k N



0 x k cos nxdx

= 1

n (x k sin nx

0 

0 kx k −1 sin nxdx

= k n



0 x k −1 sin nxdx O

| 

0 x k cos nxdx |

2k

n (2π) k −1 → 0 çn → ∞

]ª)



0 x k cos nxdx → 0 çn → ∞ úL<í©/ƒb f(x) , ª¦Öá P k (x) VV¡

max

x ∈[0,2π] |f(x) − P k (x) | → 0 k → ∞ Ĥ



0 f (x) cos nxdx

=



0 (f (x) − P k (x)) cos nxdx +



0 P k (x) cos nxdx

=



0 (f (x) − P k (x)) cos nx + 0

¦"úM

| 

0 f (x) cos nxdx |



0 |f(x) − P k (x) | | cos nx|dx

≤ max

x ∈[0,2π] |f(x) − P k (x) |

· 

0 | cos nx|dx → 0 ç k → ∞

¥_„pj¶.dOø_„pj¶µó

!…6.ñqõ|, J cos x ¦HÑL<U

‚Ñ 2π í©/ g(x) w”ÌMÑS? Oºz

bç2}& (analysis) íxÍ—V¡ (den-

sity) í–1[®ËóçÀU Êõ‰ƒb2

(6)

úí–1:

Ôƒb

(characteristic function)

ÀÓƒb (simple function)

ª¿ƒb

(measurable function)

ªƒb (integrable function) BbÿG#è6 25

(3) Fourier b:

Bbø f WÇÑ Fourier b (ªJl cq f ÑøU‚ƒb7/ f 5 Fourier  bY¹)

f (x) a 0 2 +

 k =1

(a k cos kx + b k sin kx) (1.21) si°v cos nx 7(}1‚àúiƒ bí£>4)

1



0 f (x) cos nxdx = a n , n = 0, ±1, ±2 · · · ObY¹, Ĥ

a n = 1



0 f (x) cos nxdx →0 n → ∞ J cos x ²ÑL<U‚ 2π 5ƒb g(x) ø š‚à Fourier b

g(x) a 0 2 +

 j =1

a j cos jx+

 j =1

b j sin jx

g(nx) a 0 2 +

 j =1

a j cos jnx+

 j =1

b j sin jnx

FJ



0 f (x)g(nx)dx

= a 0 2



0 f (x)dx +

 j =1

(a j c jn + b j d jn ) w2

 

c jn =  0 f (x) cos jnxdx d jn =  0 f (x) sin jnxdx

‚à Cauchy - Schwartz .ø

 j=1

a j c jn ≤ ( 

j=1

a 2 j )

12

(

 j=1

c 2 jn )

12

O

 j =1

c 2 jn 

k =n

c 2 k → 0 ç n → ∞

] 

j =1

a j c jn → 0 ç n → ∞

°Üª)

 j =1

b j d jn → 0 ç n → ∞

FJ

n lim →∞



0 f (x)g n (x)dx

=



0 f (x) a 0 2 dx

= 1



0 g(x)dx 

0 f (x)dx (1.22)

(7)

²Æuz (àÿY¹íxk) g n ( ·) → g(·), g = a 0

2 = 1



0 g(x)dx (4) Ìä–È:

ÄÑ Fourier }Ñø_Y¹/ e −iλx Ñ©/, Ĥª) ˆ f (λ) ú λ 7kuø_©

/, J f ∈ L 2 ( R ) † â Parseval ìÜø



−∞ | ˆ f (λ) | 2 dλ < O ˆ f (λ) Ñø_©/, ]ª!6

f (λ) ˆ → 0 ç λ → ±∞

à‹ f ∈ L 2 ( R ) , BbªJ5? f 5~M ƒb (truncated function)

f n (x) =

 

f (x) |f(x)| ≤ n 0 |f(x)| > n



−∞ |f n (x) | 2 dx ≤ n 

−∞ |f n (x) |dx

≤ n 

−∞ |f(x)|dx < ∞ ] f n ∈ L 2 ( R ), Ä ¤â‡Þ5!‹ø

f ˆ n (λ) → 0 ç λ → ±∞

Oâ~Mƒb5ì2ø



−∞ |f(x) − f n (x) |dx → 0, ç n → ∞ FJ

| ˆ f (λ) − ˆ f n (λ) |

=



−∞ (f (x) − f n (x))e −iλx dx



−∞ |f(x) − f n (x) |dx → 0

ĤJ¦ |λ| óç×U) | ˆ f n (λ) | → 0, †

| ˆ f (λ) | → 0 ç |λ| → ∞

Å: â Fourier bíhõ7k, Rie- mann - Lebesgue ùÜõÒ,ÿuƒb f 5O n á Fourier [b O Fourier b1

³Ììuâ cos nx, sin nx, e inx F A â Sturm-Liouville ½æ7õÉbuø £>ƒb (orthogonal functions) ¹ª

u  + λu = 0 ↔ {e

n

x } [pu  ]  + [λρ(x) − q(x)]u = 0 ↔ {ϕ n (x) } Ĥª“¿íu¤ví Riemann-

Lebesgue ùÜÑ

n lim →∞

< f, ϕ n >

< ϕ n , ϕ n >

= lim

n →∞

 b

a f ϕ n ρdx

 b

a ϕ 2 n ρdx

= 0 (1.23)

2. ìó¶ (method of sta- tionary phase)

Ên6šíóà (dispersion) ½æv, Bb44bû˝6-5}

I(a, b, λ) =

 b

a

f (x)e iλg (x) dx, λ  1 (2.1)

¥}Ê;ç (Acoustics) , Smç

(Geometric Optics) íÏm06”ò

Riemann- Lebesgue ùÜTX7¥}í

ÎhWÑ (macroscopic behavoir) Ob

(8)

nj;ßDmš5f]†âúw‰Y (fluc- tuation) ChWÑ (microscopic behav- ior)

¶ (method of stationary phase)

[ø-«…t (Euler formula) e = cos θ + i sin θ

ĤBbªln6} I(a, b, λ) 5õ¶

ReI(a, b, λ)

=

 b

a f (x) cos(λg(x))dx, λ  1 (2.2)

¥}õÒ,ÿu Riemann - Lebesgue ù Ü5R2

 b

a

f (x) cos λxdx  b

a

f (x) cos λg(x)dx ŸluÀPƒb I(x) = x, Ûʆ²Ñ g(x), O !…, cos λg(x) EÍu cos λx 5$ ĤJƒb f 5Z‰¾±ük cos(λg(x)) íu (àÇFý)

Ç ü

†yàÊ Riemann-Lebesgue ùÜ2F n6, ƒb cos λg(x) £í¶MDŠí¶

M˛¤M© óJ¾, Ĥ¥¶Mú}

ReI(a, b, λ) 7kªz³à, ñøb5?

íuÅ— dx d cos λg(x) = 0 íµ<õ x, Ä Ñ¤v1.ßÞPÓ (oscillation) 6ÿ³ J¾ (cancellation) Tà (ÄÑ cos λg(x)

¤vÑ4b)

Ç ý

âk¤ŸÄBb˚Å— g  (x i ) = 0 íõ x i

Ñìõ (point of stationary phase), ÿ uç λ óç× vú}õ.íõ

Ç þ (1) g  (x) = 0, α < x < β

I(α, β, λ) =

 β

α

f (x) exp[iλg(x)]dx

=

 β

α

f (x)

iλg  (x) d[e iλg (x) ]

(9)

â}¶}) iλI(α, β, λ) = f

g  exp(iλg) β

α

 β

α

f  g  −fg 

(g  ) 2 exp(iλg)dx

‚àúi.

|x + y| ≤ |x| + |y|,

|  β

α

F (x)dx | ≤  β

α |F (x)|dx ø

λI(α, β, λ)

f (β) g  (β)

+ f (α) g  (α)

+

 β

α

f  g  −fg 

(g  ) 2

dx Ĥª!6

 β

α

f (x) exp[iλg(x)]dx = O( 1

λ ), λ  1 (2.3)

¥u³ìóõ (point of stationary phase) v5!‹, wŸBbn6ìóõ 58$, çÍ´u*ø_õÇá:

(2) g  (x i ) = 0, g  (x i ) = 0

â©/4Bbcâ5?–È[x i −δ, x i + δ]

I i =

 x

i

x

i

−δ f (x) exp[iλg(x)]dx

â Taylor WÇø I i  x

i

x

i

−δ f (x)

· exp{iλ[g(x i )+ 1

2 g  (x i )(x −x i ) 2 }dx

= 2f (x i ) exp[iλg(x i )]

·  x

i

x

i

exp[iλA(x − x i ) 2 ]dx

w2 A = 1 2 g  (x i ) > 0 ÛÊ5?‰b‰² λA(x − x i ) 2 = ξ 2

†,5}Ñ

 x

i

x

i

exp[iλA(x − x i ) 2 ]dx

= 1

λA

 λAδ

0 exp iξ 2

¥}ÿu|±í Fresnel }



0 exp(iξ 2 )dξ

=



0 cos ξ 2 dξ + i



0 sin ξ 2

= 1 2

πe i

π4

=

 π

8 (1 + i) (2.4) Ĥª!6

I i ≈ 2f(x i ) exp[iλg(x i )] 1

λA 1 2

πe i

π4

= (

λg  (x i ) )

12

e iλg (x

i

) e i

π4

f (x i ) à‹ g  (x i ) < 0 , ?ö, cÏ_¯U

I i ≈ (

λ |g  (x i ) | )

12

e iλg (x

i

) e −i

π4

f (x i )

¥uø_õí8$; øO†Ñ I(a, b, λ)

=

 b

a

f (x)e iλg (x) dx

= 

j :g



(x

j

)>0

[

λg  (x j ) ]

12

e iλg (x

j

) e i

π4

f (x j )

+ 

j :g



(x

j

)<0

[

λ |g  (x j ) | ]

12

e iλg (x

j

)

·e −i

π4

f (x j ) + O( 1

λ ) (2.5)

¥ÿu (method of stationary phase), …

µsBb} I(a, b, λ) ú λ ¦Ú¡WÇ

(10)

, wOøá3bu§ìóõ (stationary point) Fà

úk n ½}Bb6ó°í!‹:

g  (x 0 )( ùŸ})



2

g

∂x

2

(x 0 ) (Hessian ä³)

I(λ)  · · ·  f (x)e iλg (x) dx 1 · · · dx n

∼ (

λ )

n2

| det 2 g(x 0 )

∂x 2

12

f (x 0 )

· exp[iλg(x 0 )+

4 sgn 2 g(x 0 )

∂x 2 ] (2.6) w2

sgn x x

|x|

/

∂g(x 0 )

∂x = 0, det( 2 g(x 0 )

2 x 2 ) = 0

¥_!‹ª@àƒFeynman ˜}

(Feynman path integral), Ä ¤BbªJz Riemann - Lebesgue ùÜ6TX7 Feyn- man ˜}5|Ÿá$, ¤v¡b λ ² Ñ s4b (Plank constant)  û˝ç

 → 0 v¾ä‰çD©t‰ç5É[ ¥ÿ

uF‚“semi-classical limit”

3. ÿY¹ — Y¹h1íTô

ÊOøJÉk Riemann - Lebesgue ùÜ2, øé̓bbJ {cos nx} C {sin nx}(g n (x) = g(nx)) 5 ”Ì1.æ Ê, wŸÄÿu¡H}&2Fû˝í3æ5

ø — PÓ (oscillation) 7 Riemann - Lebesgue ù܆TX7û˝¤ø3æ íOø_5x, °v6µsBb¥<ƒb (cos nx) 5 Y¹b<2†.âŒk},

¥£uÿY¹ (weak convergence) íŸá

<2 7w”̆˚Ñÿ”Ì (weak limit) Wà:

cos nx & 0 w sin nx & 0 w

g n (x) = g(nx) (3.1)

&q = w 1



0 g(x)dx

≡ w − lim g n ( ·) øO7k

g n ( ·) & g w g n 2 ( ·) w & (g) 2

¥µsBbÿY¹íÌ„D˚Ø

w − lim n →∞ (q n ( ·)) 2 = (w−lim q n ( ·)) 2 (3.2)

«wÊÝ(45½æ ªJ“¿íƒu (w −lim g n ( ·)) 2 ≤ w−lim(g n )(.)) 2 (3.3) Bbø½æ[Ñy22:

g n ( ·) & g w ⇒ F (g n ) &F (g) w (3.4) F ÑL<ƒb Ék¥½æªœêcí bçÜ6ÿT6Føÿu—compensated compactness j¶ ¥u‚à Young’s ¿ V·HPÓ (oscillation) 7w–Äÿu Riemann-Lebesgue ùÜ:

cos nx & w 1



0 cos xdx

(11)

BbTø-‰b‰² (λ = cos x) ª) cos nx & w

 1

−1 λ 1 π

1 − λ 2

≡ < λ, ν(x) > (3.5) 7 ν(x) ÿu Young’s ¿ â\}ƒb ÑøJƒbĤéÍ

cos nx & w

 1

−1 λ 1 π

1 − λ 2 dλ = 0 (3.6) O Young’s ¿ µsBbíyÖk¤

cos 2 nx & w

 1

−1 λ 2 1 π

1 − λ 2

= < λ 2 , ν(x) > (3.7) Y¤éR

cos k nx & w

 1

−1 λ k 1 π

1 − λ 2

= < λ k , ν(λ) > (3.8) mÍÖá·Ah7, µóª‚&íuúk L<í©/ƒb f(x) @

f (cos nx) & < f (λ), ν(λ) > w (3.9)

°ÜúL<ÿY¹åJ {g n } 6ó°5!

f (g n ) & < f (λ), ν(λ) > w (3.10) çͤv5 Young’s ¿ ν(λ) uÓ {g n } 7ìí ¥ÿu Young’s ¿ í!…ìÜ (fundamental theorem of Young’s mea- sure)

ìÜ: I K ⊂ R m , Ω R n Ñä5 ÇÕ¯, 7/

u  : Ω R m

Ѫ¿ƒbí/Å— u  (y) ∈ K a.e. † æÊøíœ0¿ (probability measure) ν y ∈ Prob( R m ), y ∈ Ω U)

suppν y ⊂ K, y ∈ Ω

/úL<©/ƒb f : R m R , á jæÊ äbJ u  Å—

w − lim f(u  ) = < f (λ), ν y (λ) >

 f (λ)dν y (λ) (3.11) Ék¤ìÜ5„p~¡5 [1] [3] [4], BbøÊÇÕOd«n Young’s ¿ D

^k'K¶ (compensated compactness method) £w @à vìÜõÒ,ÿuø/

[ÛìÜ (representation theorem) µs Bb‚à Young’s ¿ Dƒb f [ý|V 7/6sg7Ý(4áíÿY¹½æ, 6Ä

¤ÑS¥ìÜÊÝ(4}&2rÆO½bi H ßZM)øTíu Young’s ¿ yuû

˝ Homogenization(ÌG“) í½b5x N¬¥<½æíû˝, éBbyÀUÊ©t }& (classical analystic) 2FIíJæ

¶M(singular part) õÒ,u½bí, 7.

?µóÖiËz“˛TTÑÉ” BbøÊ

„Vídı2n6 Young ¿ , H-¿ ´ FbÊÝ(4j˙D Homogenization , 5@à

¡5’e

1. G.-Q. Chen, The compensated compactness

method and the system of isentropic gas

dynamics, PreprintMSRI-00527-91 Math-

ematical Science Research Institute, Berke-

ley (1990).

(12)

2. Bernard Dacorogna, Weak continuity and Weak Lower Semicontinuity of Non-Linear Functions. LNM, 922. Springer-Verlag, 1982.

3. L. C. Evans, Weak Convergence methods for Nonlinear Partial Differential Equa- tions, CBMS Regional Conf. Ser. in Math., 74, Amer. Math. Soc. Providence, RI, (1990).

4. P. Gerard, Microlocal Defect measures, Commun. in Partial Differential Equa- tions, 16, 1991(1761-1794)

5. M. A. Pinsky, Partial Differential Equa- tions and Boundary Value Problems with Application, 2nd Ed., McGraw-Hill 1991.

6. M. Struwe, Variational Methods: Ap- plications to Nonlinear Partial Differen-

tial Equations and Hamiltonian System, Springer-Verlag(1990).

7. L. Tartar, Compensated compactness and applications to partial differential equa- tions, Heriot-Watt Symposium, IV, (ed. R.

J. KNOPS), Pitman, New York, (136-211) 1979.

8. L. Tartar, H-measures, a new approach for studying homogenization, oscillations and concentration effects in partial differential equations, Proceedings of the Royal Society of Edinburgh, 115A, 1990 (193-230)

— …dT6L`kÅ AŠ×çbçÍ

ÿü2Ûû˝Í

¶ %ÝçÀ)ݱÀ

˚3Y («×bûF) †p[ (2Û×ç)

") (m¸×ç) ’ó% (2£×ç)

†ŒŒ (2˙@bF) ’Àó (2˙@bF) ÏQö (\µ×ç) £SG (\×@bF) Sï (ÀM×ç) Š Ÿ (òm×) Ït˚ (A×@bF) Ù? (µ±×ç)

ËÅ: 2Ûû˝Í¶ %ÝçÀ C~’£Ÿ¶~¡c……¥³Ü 2Ûû

˝Í¶ %ÝçÀı˙

參考文獻

相關文件

We work over the complex number field C.. Let X be a projective minimal Gorenstein 3-fold of general type.. The above sum runs over all those exceptional divisors of p that lie over

Two-Dimensional Riemann Problems for Ideal Gas Dynamics. Configuration 16 Configuration 17 Configuration 18

This work is a contribution to the comprehension on the impact of biodiesel blends of different origin and properties on the unregulated emissions of carbonyl compounds, PAH,

There have been important progresses during the 1960-1990's on the well-posedness theory for Navier-Stokes equations in the gas dynamics, done mostly by the Japanese School of

43) Calculate the energy, in kilojoules, of one mole of photons of yellow laser light of wavelength 570 nm.. 49) Which of the following bonds is more ionic?. A) C-F B) K-Cl C) Mg-N

 Promote project learning, mathematical modeling, and problem-based learning to strengthen the ability to integrate and apply knowledge and skills, and make. calculated

refined generic skills, values education, information literacy, Language across the Curriculum (

In Phase 2, a Research Team from Hong Kong Institute of Educational Research and the Faculty of Education of the Chinese University of Hong Kong undertook to identify