Eclats de vers : Matemat 12 : Analyse complexe - 1

Index des Grimoires

Retour à l’accueil

Table des matières

\( \newcommand{\parentheses}[1]{\left(#1\right)} \newcommand{\crochets}[1]{\left[#1\right]} \newcommand{\accolades}[1]{\left\{#1\right\}} \newcommand{\ensemble}[1]{\left\{#1\right\}} \newcommand{\identite}{\mathrm{Id}} \newcommand{\indicatrice}{\boldsymbol{\delta}} \newcommand{\dirac}{\delta} \newcommand{\moinsun}{{-1}} \newcommand{\inverse}{\ddagger} \newcommand{\pinverse}{\dagger} \newcommand{\topologie}{\mathfrak{T}} \newcommand{\ferme}{\mathfrak{F}} \newcommand{\img}{\mathbf{i}} \newcommand{\binome}[2]{ \left\{ \begin{array}{c} #1 \\ #2 \\ \end{array} \right\} } \newcommand{\canonique}{\mathfrak{c}} \newcommand{\tenseuridentite}{\boldsymbol{\mathcal{I}}} \newcommand{\permutation}{\boldsymbol{\epsilon}} \newcommand{\matriceZero}{\mathfrak{0}} \newcommand{\matriceUn}{\mathfrak{1}} \newcommand{\christoffel}[2]{ \left\{ \begin{array}{c} #1 \\ #2 \\ \end{array} \right\} } \newcommand{\lagrangien}{\mathfrak{L}} \newcommand{\sousens}{\mathfrak{P}} \newcommand{\partition}{\mathrm{Partition}} \newcommand{\tribu}{\mathrm{Tribu}} \newcommand{\topologies}{\mathrm{Topo}} \newcommand{\setB}{\mathbb{B}} \newcommand{\setN}{\mathbb{N}} \newcommand{\setZ}{\mathbb{Z}} \newcommand{\setQ}{\mathbb{Q}} \newcommand{\setR}{\mathbb{R}} \newcommand{\setC}{\mathbb{C}} \newcommand{\corps}{\mathbb{K}} \newcommand{\boule}{\mathfrak{B}} \newcommand{\intervalleouvert}[2]{\relax \ ] #1 , #2 [ \ \relax} \newcommand{\intervallesemiouvertgauche}[2]{\relax \ ] #1 , #2 ]} \newcommand{\intervallesemiouvertdroite}[2]{[ #1 , #2 [ \ \relax} \newcommand{\fonction}{\mathbb{F}} \newcommand{\bijection}{\mathrm{Bij}} \newcommand{\polynome}{\mathrm{Poly}} \newcommand{\lineaire}{\mathrm{Lin}} \newcommand{\continue}{\mathrm{Cont}} \newcommand{\homeomorphisme}{\mathrm{Hom}} \newcommand{\etagee}{\mathrm{Etagee}} \newcommand{\lebesgue}{\mathrm{Leb}} \newcommand{\lipschitz}{\mathrm{Lip}} \newcommand{\suitek}{\mathrm{Suite}} \newcommand{\matrice}{\mathbb{M}} \newcommand{\krylov}{\mathrm{Krylov}} \newcommand{\tenseur}{\mathbb{T}} \newcommand{\essentiel}{\mathfrak{E}} \newcommand{\relation}{\mathrm{Rel}} \newcommand{\strictinferieur}{\ < \ } \newcommand{\strictsuperieur}{\ > \ } \newcommand{\ensinferieur}{\eqslantless} \newcommand{\enssuperieur}{\eqslantgtr} \newcommand{\esssuperieur}{\gtrsim} \newcommand{\essinferieur}{\lesssim} \newcommand{\essegal}{\eqsim} \newcommand{\union}{\ \cup \ } \newcommand{\intersection}{\ \cap \ } \newcommand{\opera}{\divideontimes} \newcommand{\autreaddition}{\boxplus} \newcommand{\autremultiplication}{\circledast} \newcommand{\commutateur}[2]{\left[ #1 , #2 \right]} \newcommand{\convolution}{\circledcirc} \newcommand{\correlation}{\ \natural \ } \newcommand{\diventiere}{\div} \newcommand{\modulo}{\bmod} \newcommand{\pgcd}{pgcd} \newcommand{\ppcm}{ppcm} \newcommand{\produitscalaire}[2]{\left\langle #1 \left|\right\relax #2 \right\rangle} \newcommand{\scalaire}[2]{\left\langle #1 \| #2 \right\rangle} \newcommand{\braket}[3]{\left\langle #1 \right| #2 \left| #3 \right\rangle} \newcommand{\orthogonal}{\bot} \newcommand{\forme}[2]{\left\langle #1 , #2 \right\rangle} \newcommand{\biforme}[3]{\left\langle #1 , #2 , #3 \right\rangle} \newcommand{\contraction}[3]{\left\langle #1 \odot #3 \right\rangle_{#2}} \newcommand{\dblecont}[5]{\left\langle #1 \right| #3 \left| #5 \right\rangle_{#2,#4}} \newcommand{\major}{major} \newcommand{\minor}{minor} \newcommand{\maxim}{maxim} \newcommand{\minim}{minim} \newcommand{\argument}{arg} \newcommand{\argmin}{arg\ min} \newcommand{\argmax}{arg\ max} \newcommand{\supessentiel}{ess\ sup} \newcommand{\infessentiel}{ess\ inf} \newcommand{\dual}{\star} \newcommand{\distance}{\mathfrak{dist}} \newcommand{\norme}[1]{\left\| #1 \right\|} \newcommand{\normetrois}[1]{\left|\left\| #1 \right\|\right|} \newcommand{\adh}{adh} \newcommand{\interieur}{int} \newcommand{\frontiere}{\partial} \newcommand{\image}{im} \newcommand{\domaine}{dom} \newcommand{\noyau}{ker} \newcommand{\support}{supp} \newcommand{\signe}{sign} \newcommand{\abs}[1]{\left| #1 \right|} \newcommand{\unsur}[1]{\frac{1}{#1}} \newcommand{\arrondisup}[1]{\lceil #1 \rceil} \newcommand{\arrondiinf}[1]{\lfloor #1 \rfloor} \newcommand{\conjugue}{conj} \newcommand{\conjaccent}[1]{\overline{#1}} \newcommand{\division}{division} \newcommand{\difference}{\boldsymbol{\Delta}} \newcommand{\differentielle}[2]{\mathfrak{D}^{#1}_{#2}} \newcommand{\OD}[2]{\frac{d #1}{d #2}} \newcommand{\OOD}[2]{\frac{d^2 #1}{d #2^2}} \newcommand{\NOD}[3]{\frac{d^{#3} #1}{d #2^{#3}}} \newcommand{\deriveepartielle}[2]{\frac{\partial #1}{\partial #2}} \newcommand{\dblederiveepartielle}[2]{\frac{\partial^2 #1}{\partial #2 \partial #2}} \newcommand{\dfdxdy}[3]{\frac{\partial^2 #1}{\partial #2 \partial #3}} \newcommand{\dfdxdx}[2]{\frac{\partial^2 #1}{\partial #2^2}} \newcommand{\gradient}{\mathbf{\nabla}} \newcommand{\combilin}[1]{\mathrm{span}\{ #1 \}} \newcommand{\trace}{tr} \newcommand{\proba}{\mathbb{P}} \newcommand{\probaof}[1]{\mathbb{P}\left[#1\right]} \newcommand{\esperof}[1]{\mathbb{E}\left[#1\right]} \newcommand{\cov}[2]{\mathrm{cov} \left( #1 , #2 \right) } \newcommand{\var}[1]{\mathrm{var} \left( #1 \right) } \newcommand{\rand}{\mathrm{rand}} \newcommand{\variation}[1]{\left\langle #1 \right\rangle} \newcommand{\composante}{comp} \newcommand{\bloc}{bloc} \newcommand{\ligne}{ligne} \newcommand{\colonne}{colonne} \newcommand{\diagonale}{diag} \newcommand{\matelementaire}{\mathrm{Elem}} \newcommand{\matpermutation}{permut} \newcommand{\matunitaire}{\mathrm{Unitaire}} \newcommand{\gaussjordan}{\mathrm{GaussJordan}} \newcommand{\householder}{\mathrm{Householder}} \newcommand{\rang}{rang} \newcommand{\schur}{\mathrm{Schur}} \newcommand{\singuliere}{\mathrm{DVS}} \newcommand{\convexe}{\mathrm{Convexe}} \newcommand{\petito}[1]{o\left(#1\right)} \newcommand{\grando}[1]{O\left(#1\right)} \)

1. Exponentielle complexe

1.1. Dépendances

  • Chapitre \ref{chap:complexe} : Les complexes

1.2. Introduction

AFAIRE : ARRANGER LE CHAPITRE

Considérons l'unique solution \(x\) de l'équation différentielle :

\( \OD{x}{t}(t) = z \cdot x(t) \\ x(0) = 1 \)

où \(z \in \setC\). On définit alors l'exponentielle d'un nombre complexe par :

\[\exp( z \cdot t ) = x(t)\]

On a donc simplement :

\[\exp(z) = x(1)\]

1.3. Additivité

Comme les fonctions :

\( s(t) = \exp( (z_1 + z_2) \cdot t ) \\ p(t) = \exp(z_1 t) \cdot \exp(z_2 t) \)

vérifient la même équation différentielle :

\( \OD{s}{t}(t) = (z_1 + z_2) \ s(t) \\ s(0) = 1 \)

\( \OD{p}{t}(t) = z_1 \ p(t) + z_2 \ p(t) = (z_1 + z_2) \ p(t) \\ p(0) = 1 \)

on a \(s(t) = p(t)\) pour tout \(t\). En considérant le cas \(t=1\), on arrive à la propriété d'additivité des exponentielles :

\[\exp(z_1 + z_2) = \exp(z_1) \cdot \exp(z_2)\]

1.4. Lien avec les fonctions trigonométriques

Considérons maintenant le cas particulier où \(u(t) = \exp(\img t)\) :

\( \OD{u}{t}(t) = \img u(t) \\ u(0) = 1 \)

On a alors :

\[\frac{d^2 u}{dt^2}(t) = - u(t) \qquad u(0) = 1 \qquad \OD{u}{t}(0) = \img\]

Soit la fonction \(u : \setR \mapsto \setR\) définie pour tout réel \(t\) par :

\[u(t) = \cos(t) + \img \sin(t)\]

On voit que :

\[\OD{u}{t} = -\sin(t) + \img \cos(t)\]

vérifie la même équation. Par unicité de la solution, on a :

\[\exp(\img t) = \cos(t) + \img \sin(t)\]

On en déduit directement que :

\( \exp(\img \pi/2) = \img \\ \exp(\img \pi) = -1 \\ \exp(3\pi\img/2) = \img \\ \exp(2\pi\img) = 1 \)

ainsi que la périodicité :

\[\exp(t+2\pi) = \exp(t)\]

pour tout \(t\in\setR\).

En utilisant l'additivité, on note que :

\[\exp(z) = \exp(a) \cdot \exp(\img b) = \exp(a) \cdot (\cos(b) + \img \sin(b))\]

De la même façon, si nous prenons \(u(t) = \exp(-i.t)\), nous en déduisons :

\( \frac{d^2 u}{dt^2}(t) = - u(t) \\ u(0) = 1 \qquad \OD{u}{t}(0) = -i \)

Donc :

\[\exp(-\img t) = \cos(t) - \img \sin(t)\]

En additionnant, puis en soustrayant les relations :

\( \exp( \img u ) = \cos(u) + \img \sin(u) \\ \exp( - \img u ) = \cos(u) - \img \sin(u) \)

on obtient :

\( \cos(u) = \frac{ \exp(\img u) + \exp(-\img u) }{2} \\ \sin(u) = \frac{ \exp(\img u) - \exp(-\img u) }{2 \img} \)

Choisissant un angle \(\theta\in\setR\) tel que :

\( \cos(\theta) = \frac{\Re(z)}{\abs{z}} \\ \sin(\theta) = \frac{\Im(z)}{\abs{z}} \)

on peut réexprimer \(z\) comme :

\[z = \abs{z}(\cos(\theta)+\img\sin(\theta)) = \abs{z}\exp(\img\theta)\]

Comme les fonctions \(\cos\) sont \(2\pi\) périodiques, il existe une infinité d'angles \(\theta\) vérifiant cette propriété. On définit l'argument de \(z\) comme l'unique \(\theta\) vérifiant cette propriété et se trouvant dans l'intervalle \([0,2\pi)\).

\( θ = arg(z) \quad⇔\quad

\begin{cases} z = \abs{z}\exp(\img\theta) \\ \theta\in [0,2\pi) \end{cases}

\)

Inspiré par relation :

\[z = \abs{z}\exp(\img\arg(z))\]

et cherchant à étendre la propriété :

\[\ln(x \cdot y) = \ln(x) + \ln(y)\]

du logarithme sur \(\setR\), on définit le logarithme d'un complexe par :

\( \ln(z) = \ln(\abs{z}) + \ln(\exp(\img\arg(z))) \\ \ln(z) = \ln(\abs{z}) + \img\arg(z) \)

Pour un \(z \in \setC\) donné, l'ensemble des \(y \in \setC\) tels que :

\[\exp(y) = z\]

peut s'écrire :

\[\mathcal{Y} = \{ y_k = \ln(\abs{z}) + \img \arg(z) + 2 \pi \img k : k \in \setZ \}\]

Les complexes permettent de retrouver aisément les propriétés fondamentales des fonctions trigonométriques \(\cos\), \(\sin\). En se rappelant que :

\[\exp(-\img t) \cdot \exp(i t) = \exp(i t - i t) = \exp(0) = 1\]

on retombe sur :

\( ( \cos(t) + \img \sin(t) ) ( \cos(t) - \img \sin(t) ) = 1 \\ \cos(t)^2 + \sin(t)^2 = 1 \)

En dérivant la relation reliant l'exponentielle aux fonctions trigonométriques, on obtient :

\( \OD{}{t}( \cos(t) + \img \sin(t) ) = \OD{}{t}\exp(\img t) \\ \OD{}{t} \cos(t) + \img \OD{}{t}\sin(t) = \img \exp(\img t) \\ \OD{}{t} \cos(t) + \img \OD{}{t}\sin(t) = \img \cos(t) - \sin(t) \)

on a bien :

\( \OD{}{t} \cos(t) = -\sin(t) \\ \OD{}{t}\sin(t) = \cos(t) \)

On a aussi :

\( \cos(u+v) + \img \sin(u+v) = \exp\left[ \img ( u + v ) \right] \\ \cos(u+v) + \img \sin(u+v) = \exp\left[ \img u \right] \cdot \exp\left[ \img v \right] \\ \cos(u+v) + \img \sin(u+v) = (\cos(u) + \img \sin(u))(\cos(v) + \img \sin(v)) \)

ce qui donne, tous calculs faits :

\( \cos(u+v) = \cos(u)\cos(v) - \sin(u)\sin(v) \\ \sin(u+v) = \sin(u)\cos(v) + \sin(v)\cos(u) \)

2. Dérivation et intégration dans le plan complexe

2.1. Dépendances

  • Chapitre \ref{chap:differ} : Les différentielles
  • Chapitre \ref{chap:integ} : Les intégrales

2.2. Dérivées

On définit la dérivée d'une fonction \(f : \setC \mapsto \setC\) par analogie avec la dérivée des fonctions réelles :

\[\OD{f}{z}(z) = \lim_{ \substack{ h \to 0 \\ h \in\setC } } \frac{f(z+h) - f(z)}{h}\]

Si \(f\) est dérivable sur \(A\subseteq\setC\), on dit qu'elle est analytique sur \(A\).

On définit les fonctions \(u,v : \setR^2 \mapsto \setR\) associées à \(f\) par :

\( u(x,y) = \Re\left(f(x + \img y) \right) \\ v(x,y) = \Im\left(f(x + \img y) \right) \)

On a alors \(f(z) = u(x,y) + \img v(x,y)\). Si \(f\) est dérivable, la limite ne peut pas dépendre du chemin suivi pour arriver en \((x,y)\). On peut donc choisir \(h\) de la forme \(\Delta x\) ou \(\img \Delta y\) avec \(\Delta x,\Delta y\in\setR\). Cela donne :

\( \OD{f}{z}(z) = \lim_{\Delta x \to 0} \frac{\PD{u}{x}\Delta x + \img \PD{v}{x}\Delta x}{\Delta x} \\ \OD{f}{z}(z) = \lim_{\Delta y \to 0} \frac{\PD{u}{y}\Delta y + \img \PD{v}{y}\Delta y}{\img \Delta y} \)

et par comparaison :

\( \PD{u}{x} = \PD{v}{y} \\ \PD{u}{y} = -\PD{v}{x} \)

En dérivant de nouveau ces équations, on obtient :

\( \frac{\partial^2 u}{\partial x^2} + \frac{\partial^2 u}{\partial y^2} = 0 \\ \frac{\partial^2 v}{\partial x^2} + \frac{\partial^2 v}{\partial y^2} = 0 \)

Les puissances de \(z\) sont dérivables. On vérifie que :

\[\OD{z^n}{z}=n z^{n-1}\]

2.3. Extension de l'intégrale au plan complexe

Soit une fonction \(f : A \mapsto \setC\). Comme \(u = \Re(f)\) et \(v = \Im(f)\) sont des fonctions à valeurs réelles :

\[u,v : A \mapsto \setR\]

leur intégrale est bien définie. On définit alors l'intégrale de \(f\) par :

\[\int_A f d\mu = \int_A u d\mu + \img \int_A v d\mu\]

2.4. Intégrale de ligne

Soit une fonction \(\gamma : [a,b] \mapsto \setC\) et la courbe :

\[l = \{ \gamma(t) : t \in [a,b] \} \subset \setC\]

On définit l'intégrale de ligne de \(f\) sur \(l\) par :

\[\int_l f(z) dz = \int_a^b (f \circ \gamma)(t) \OD{\gamma}{t}(t) dt\]

On peut également écrire :

\( \int_l f(z) dz = \int_l (u + \img v)(dx + \img dy) \\ \int_l f(z) dz = \int_l (u dx - v dy) + \img \int_l (u dy + v dx) \)

où les intégrales des membres de droites sont des intégrales de ligne classiques de fonctions \(\setR^2 \mapsto \setR\).

2.4.1. Théorème fondamental

Soit une fonction analytique \(F : \setC \mapsto \setC\). La dérivée d'une composée de fonction nous donne :

\[\OD{}{t}(F \circ \gamma)(t) = (\OD{F}{z} \circ \gamma)(t) \OD{\gamma}{t}(t)\]

On en déduit, en appliquant le théorème fondamental que :

\[\int_l \OD{F}{z}(z) dz = F(\gamma(b))-F(\gamma(a))\]

2.4.2. Contour fermé

Dans la suite, lorsque \(l\) est une courbe fermée et est donc la frontière d'un certain ensemble \(l = \partial S\) inclus dans \(\setC\), nous notons :

\[\oint_l f(z) dz = \int_{\partial S} f(z) dz\]

Soit \(f\) une fonction analytique et \(l = \partial S\) une courbe fermée. On a :

\[\oint_l f(z) dz = \oint_l (u dx - v dy) + \img \oint_l (u dy + v dx)\]

En appliquant le théorème de Stokes, il vient :

\( \oint_l f(z) dz = \int_S (-\PD{v}{x}-\PD{u}{y}) dS + \img \int_S (\PD{u}{x}-\PD{v}{y}) dS \\ \)

Mais comme \(f\) est analytique :

\( \PD{u}{x} = \PD{v}{y} \\ \PD{u}{y} = -\PD{v}{x} \)

Les termes du membre de droite s'annulent donc et :

\[\oint_l f(z) dz = 0\]

2.4.3. Cercle

Soit \(C\) le cercle de centre \(a\) et de rayon \(R\) :

\[C = \{ \gamma(\theta) = a + R \exp(\img\theta) : \theta \in [ 0 , 2 \pi ) \}\]

On a alors :

\[\OD{\gamma}{\theta}(\theta) = R \img \exp(\img\theta)\]

et :

\[(z - a)^k = \left( R \exp(\img\theta) \right)^k\]

pour tout \(z\in C\) et \(k \in \setZ\). Donc :

\( \oint_C \frac{dz}{(z-a)^k} = \int_0^{2\pi} \frac{R \img \exp(\img\theta)}{R^k \exp(\img k\theta)} \\ \oint_C \frac{dz}{(z-a)^k} = \img R^{1-k} \int_0^{2\pi} \exp[\img(1-k)\theta] d\theta \)

Mais les propriétés de périodicité des fonctions trigonométriques et de l'exponentielle associée nous permettent de vérifier que :

\[\int_0^{2\pi} \exp[\img l \theta] d\theta = 2 \pi \delta_{l,0}\]

pour tout \(l\in\setZ\). On a donc :

\[\oint_C \frac{dz}{(z-a)^k} = 2 \pi \img \delta_{k,1}\]

pour tout \(k \in \setZ\). L'intégrale s'annule pour tout les \(k \ne 1\).

Vu que la fonction \(f(z) = (z-a)^{-k}\) est analytique partout sauf en \(a\), on obtient le même résultat pour toute courbe fermée entourant \(a\) :

\[\oint_{\partial S} \frac{dz}{(z-a)^k} = 2 \pi \img \delta_{k,1} \delta_S(a)\]

2.4.4. Théorème de Cauchy

Si \(f\) est analytique, la fonction

\( g(z) =

\begin{cases} \frac{f(z)-f(a)}{z-a} & \mbox{si } z \ne a \\ \OD{f}{z}(a) & \mbox{si } z = a \end{cases}

\)

est analytique aussi. On a donc :

\[\oint_l g(z) dz = 0\]

Soit \(a\in\setC\) et une courbe fermée entourant \(a\) telle que \(a\notin l\). On a alors :

\[\oint_l \frac{f(z)-f(a)}{z-a} dz = \oint_l g(z) dz = 0\]

On en déduit que :

\[f(a) \oint_{\partial S} \frac{1}{z-a} dz = \oint_{\partial S} \frac{f(z)}{z-a} dz\]

ce qui nous donne la valeur de \(f(a)\) en fonction d'une intégrale :

\[f(a) = \frac{1}{2 \pi \img} \oint_{\partial S} \frac{f(z)}{z-a} dz\]

En dérivant \(k\) fois cette dernière relation par rapport à \(a\), on obtient :

\[\OD{f}{z}(a) = \frac{k !}{2 \pi \img} \oint_{\partial S} \frac{f(z)}{(z-a)^{k+1}} dz\]

2.4.5. Séries de Laurent

Supposons que \(f\) puisse s'écrire comme une combinaison linéaire de puissances entière de \(z-a\). On a :

\[f(z) = \sum_{n=-\infty}^{+\infty} \alpha_n (z-a)^n\]

Multipliant le tout par \((z-a)^{-k-1}\) et intégrant sur un contour entourant \(a\), on obtient :

\[\oint_l f(z)(z-a)^{-k-1} dz = \sum_{n=-\infty}^{+\infty} \alpha_n \oint_l (z-a)^{n-k-1} dz\]

Mais nous avons vu que seule l'intégrale de \((z-a)^{-1}\) ne s'annule pas. On en déduit que :

\[\alpha_k = \frac{1}{2 \pi \img} \oint_l \frac{f(z)}{(z-a)^{k+1}} dz\]

2.4.6. Théorème des résidus

Considérons le cas où :

\[f(z) = \sum_{n=-p}^{+\infty} \alpha_n (z-a)^n\]

pour un certain \(p \in \setZ\). On dit alors que \(f\) à un pôle d'ordre \(p\) en \(a\). En multpliant le tout par \((z-a)^p\) et en dérivant \(p-1\) fois, on obtient :

\[\frac{d^{p-1}}{dz^{p-1}}[(z-a)^p f(z)] = (p-1) ! \left[a_{-1} + a_0 (z-a) + a_1 (z-a)^2 + ...\right]\]

On a donc :

\[\lim_{z \to a} \frac{d^{p-1}}{dz^{p-1}}[(z-a)^p f(z)] = (p-1) ! a_{-1}\]

Mais comme \(a_{-1}\) n'est rien d'autre, à un facteur \(2\pi\img\) près, que l'intégrale de \(f\) sur un contour entourant \(a\), on a : \[\] \ointl f(z) dz = 2 π \img limz → a \frac{1}{(p-1)!} \frac{dp-1}{dzp-1}[(z-a)p f(z)]

3. Polynômes et exponentielles

\label{chap:polyexpo}

3.1. Chebyshev

Les polynômes de Chebyshev \(T_n\) sont défini par :

\( T_n : x\mapsto\cos(n \arccos(x) ) \\ T_n(\cos(\theta)) = \cos(n \theta) \)

Il est clair que :

\( T_0(x) = 1 \\ T_1(x) = x \)

Les propriétés des fonctions trigonométriques :

\( \cos[(n+m)\theta]= \cos[n\theta]\cos[m\theta]+\sin[n\theta]\sin[m\theta] \\ \cos[(n-m)\theta]= \cos[n\theta]\cos[m\theta]-\sin[n\theta]\sin[m\theta] \)

nous montrent que :

\[T_{n+m}(x) + T_{n-m}(x) = 2 T_n(x) T_m(x)\]

On en déduit entre autre que :

\[T_{n+1}(x) + T_{n-1}(x) = 2 x T_n(x)\]

Les fonctions \(T_n\) sont donc des polynômes. Comme :

\[T_n \left[ \cos\left(\frac{2k+1}{2n}\right) \right] = 0\]

les racines sont données par :

\[x_k = \cos\left(\frac{2k+1}{2n}\right)\]

pour \(k=0,1,...,n-1\). On déduit aussi la propriété suivante de la définition :

\[(T_n \circ T_m)(x) = T_{m \cdot n}(x) = (T_m \circ T_n)(x)\]

Nous allons voir que les \(T_n\) sont orthogonaux moyennant un certain poids. Si \(m = n = 0\), il est clair que :

\[\int_0^\pi T_0(\cos(\theta)) T_0(\cos(\theta)) d\theta = \int_0^\pi d\theta = \pi\]

Si \(m, n \ne 0\), on a :

\( \int_0^\pi T_m(\cos(\theta)) T_n(\cos(\theta)) d\theta = \int_0^\pi \cos(m\theta) \cos(n\theta) d\theta \\ \int_0^\pi T_m(\cos(\theta)) T_n(\cos(\theta)) d\theta = \unsur{2} \int_0^\pi \cos[(m+n)\theta] d\theta + \unsur{2} \int_0^\pi \cos[(m-n)\theta] d\theta \)

Si \(m = n\), on a alors :

\[\int_0^\pi T_m(\cos(\theta)) T_n(\cos(\theta)) d\theta = \unsur{2n} [\sin(2 n \pi) - \sin(0)] + \frac{\pi}{2}\]

La première intégrale s'annule donc par périodicité de la fonction \(\sin\). A présent, si \(m \ne n\), on a :

\( \int_0^\pi T_m(\cos(\theta)) T_n(\cos(\theta)) d\theta = \\ \unsur{2(m+n)} [\sin[(m+n)\pi] - \sin(0)] + \\ \unsur{2(m-n)} [\sin[(m-n)\pi] - \sin(0)] = 0 \)

Rassemblant ces résultats, on obtient :

\( ∫0^π Tm(cosθ) Tn(cosθ) dθ =

\begin{cases} \pi & \mbox{si } m = 0 \\ \pi/2 & \mbox{si } m = n \ne 0 \\ 0 & \mbox{si } m \ne n \end{cases}

\)

Le changement de variable \(x = \cos(\theta)\),

\( dx = -\sin(\theta) d\theta \\ d\theta = \frac{dx}{\sqrt{1-x^2}} \)

nous donne :

\[\int_0^\pi T_m(\cos\theta) T_n(\cos\theta) d\theta = \int_{-1}^1 \frac{T_m(x) T_n(x)}{\sqrt{1-x^2}}dx\]

On en déduit que :

\[\int_{-1}^1 \frac{T_m(x) T_n(x)}{\sqrt{1-x^2}} dx = \delta_{mn} \frac{\pi}{2} (1+\delta_{m,0})\]

pour tout \(m,n\in\setN\).

3.2. Hermite

Les polynômes de Hermite sont orthogonaux pour le produit scalaire :

\[\int_{-\infty}^{+\infty} H_n(x) H_m(x) \exp(-x^2) dx = 2^n n ! \sqrt{\pi} \delta_{mn}\]

Ils obéissent à la récurrence :

\( H_0(x) = 1 \\ H_1(x) = 2 x \\ H_{n+1}(x) = 2 x H_n(x) - 2 n H_{n-1}(x) \\ \)

3.3. Laguerre

Les polynômes de Laguerre sont orthogonaux pour le produit scalaire :

\[\int_{0}^{+\infty} L_n(x) L_m(x) \exp(-x) dx = \delta_{mn}\]

Ils obéissent à la récurrence :

\( L_0(x) = 1 \\ L_1(x) = 1-x \\ (n+1) L_{n+1}(x) = (2 n + 1 - x) L_n(x) - n L_{n-1}(x) \)

Auteur: chimay

Created: 2023-05-10 mer 16:45

Validate