Votre navigateur n'accepte pas le Javascript.La navigation sur ce site risque de ne pas fonctionner correctement.

194. Déterminez la forme de Jordan d’une matrice avec des opérations élémentaires

Soit la matrice suivante :

\begin{align*}
A = \begin{pmatrix}
0 & 1 & 0 & 0 &0 &0 &0 \\
0 &0 &0 &0 &0 &0 &4 \\
0 &0 &0 &1 &0 &0 &0 \\
0 &0 &0 &0 &1 &0 &0 \\
0 &0 &0 &0 &0 &0 &0 \\
0 &0 &0 &0 &0 &0 &1\\
0 &0 &0 &0 &0 &0 &0
\end{pmatrix}.
\end{align*}

Comment obtenir sa forme de Jordan ?

Note. La matrice $A$ est triangulaire supérieure avec uniquement des zéros sur sa diagonale principale. Elle est donc nilpotente : il existe un entier $k$ tel que $A^k=0.$ (Vous pouvez vérifier que $A^7$ est la matrice nulle.)

Première étape : utilisez des dilatations pour changer le $4$ en $1$ dans la dernière colonne

Vous effectuez les deux opérations élémentaires sur la matrice $A$ : $L_2 \leftarrow 1/4 L_2$ et $C_2\leftarrow 4C_2$ ce qui fournit :

\begin{align*}
A_1 = \begin{pmatrix}
0 & 4 & 0 & 0 &0 &0 &0 \\
0 &0 &0 &0 &0 &0 &1 \\
0 &0 &0 &1 &0 &0 &0 \\
0 &0 &0 &0 &1 &0 &0 \\
0 &0 &0 &0 &0 &0 &0 \\
0 &0 &0 &0 &0 &0 &1\\
0 &0 &0 &0 &0 &0 &0
\end{pmatrix}.
\end{align*}

Puis vous effectuez les deux opérations élémentaires sur la matrice $A_1$ : $L_1\leftarrow 1/4 L_1$ et $C_1\leftarrow 4C_1$ ce qui fournit :

\begin{align*}
A_2 = \begin{pmatrix}
0 & 1 & 0 & 0 &0 &0 &0 \\
0 &0 &0 &0 &0 &0 &1 \\
0 &0 &0 &1 &0 &0 &0 \\
0 &0 &0 &0 &1 &0 &0 \\
0 &0 &0 &0 &0 &0 &0 \\
0 &0 &0 &0 &0 &0 &1\\
0 &0 &0 &0 &0 &0 &0
\end{pmatrix}.
\end{align*}

En notant $D_1$ la première matrice de dilatation, vous avez :

\begin{aligned}
D_1 = \begin{pmatrix}
1 & 0 & 0 & 0 &0 &0 &0 \\
0 &4 &0 &0 &0 &0 &0 \\
0 &0 &1 &0 &0 &0 &0 \\
0 &0 &0 &1 &0 &0 &0 \\
0 &0 &0 &0 &1 &0 &0 \\
0 &0 &0 &0 &0 &1 &0\\
0 &0 &0 &0 &0 &0 &1
\end{pmatrix}.
\end{aligned}

Alors $D_1^{-1}AD_1 = A_1.$

En notant $D_2$ la deuxième matrice de dilatation, vous avez :

\begin{align*}
D_2 = \begin{pmatrix}
4 & 0 & 0 & 0 &0 &0 &0 \\
0 &1 &0 &0 &0 &0 &0 \\
0 &0 &1 &0 &0 &0 &0 \\
0 &0 &0 &1 &0 &0 &0 \\
0 &0 &0 &0 &1 &0 &0 \\
0 &0 &0 &0 &0 &1 &0\\
0 &0 &0 &0 &0 &0 &1
\end{pmatrix}.
\end{align*}

Alors $D_2^{-1}A_1D_2 = A_2.$

Soit $D_2^{-1}D_1^{-1}AD_1D_2 = A_2.$

Deuxième étape : utilisez des transvections pour n’avoir qu’un seul $1$ dans la dernière colonne

L’idée consiste à partir du $1$ du dessus, associé à un bloc de Jordan de taille maximale, pour éliminer les autres.

Partez de la matrice $A_2$ ci-dessous :

\begin{align*}
A_2 = \begin{pmatrix}
0 & 1 & 0 & 0 &0 &0 &0 \\
0 &0 &0 &0 &0 &0 &1 \\
0 &0 &0 &1 &0 &0 &0 \\
0 &0 &0 &0 &1 &0 &0 \\
0 &0 &0 &0 &0 &0 &0 \\
0 &0 &0 &0 &0 &0 &1\\
0 &0 &0 &0 &0 &0 &0
\end{pmatrix}.
\end{align*}

Effectuez alors les opérations élémentaires $L_6\leftarrow L_6-L_2$ et $C_2\leftarrow C_2+C_6$ ce qui fournit :

\begin{align*}
A_3 = \begin{pmatrix}
0 & 1 & 0 & 0 &0 &0 &0 \\
0 &0 &0 &0 &0 &0 &1 \\
0 &0 &0 &1 &0 &0 &0 \\
0 &0 &0 &0 &1 &0 &0 \\
0 &0 &0 &0 &0 &0 &0 \\
0 &0 &0 &0 &0 &0 &0\\
0 &0 &0 &0 &0 &0 &0
\end{pmatrix}.
\end{align*}

Selon le positionnement des $1$ à éliminer, plusieurs transvections peuvent être utilisées. Dans le cas présent, une seule suffit.

La matrice de transvection associée est :

\begin{align*}
T = \begin{pmatrix}
1 & 0 & 0 & 0 &0 &0 &0 \\
0 &1 &0 &0 &0 &0 &0 \\
0 &0 &1 &0 &0 &0 &0 \\
0 &0 &0 &1 &0 &0 &0 \\
0 &0 &0 &0 &1 &0 &0 \\
0 &1 &0 &0 &0 &1 &0\\
0 &0 &0 &0 &0 &0 &1
\end{pmatrix}.
\end{align*}

Vous avez $T^{-1}A_2T = A_3$ et $T^{-1}D_2^{-1}D_1^{-1}AD_1D_2T = A_3.$

Troisième et dernière étape : utilisez une matrice de permutation pour obtenir la matrice de Jordan

Repartez de la matrice suivante :

\begin{align*}
A_3 = \begin{pmatrix}
0 & 1 & 0 & 0 &0 &0 &0 \\
0 &0 &0 &0 &0 &0 &1 \\
0 &0 &0 &1 &0 &0 &0 \\
0 &0 &0 &0 &1 &0 &0 \\
0 &0 &0 &0 &0 &0 &0 \\
0 &0 &0 &0 &0 &0 &0\\
0 &0 &0 &0 &0 &0 &0
\end{pmatrix}.
\end{align*}

Appliquez la matrice de permutation suivante :

\begin{align*}
P = \begin{pmatrix}
1 & 0 & 0 & 0 &0 &0 &0 \\
0 & 1 & 0 & 0 &0 &0 &0 \\
0 & 0 & 0 & 1 &0 &0 &0 \\
0 & 0 & 0 & 0 &1 &0 &0 \\
0 & 0 & 0 & 0 &0 &1 &0 \\
0 & 0 & 0 & 0 &0 &0 &1 \\
0 & 0 & 1 & 0 &0 &0 &0
\end{pmatrix}.
\end{align*}

Vous obtenez alors :

P^{-1}A_3P = J =  \begin{pmatrix}
0 & 1 & 0 & 0 &0 &0 &0 \\
0 &0 &1 &0 &0 &0 &0 \\
0 &0 &0 &0 &0 &0 &0 \\
0 &0 &0 &0 &1 &0 &0 \\
0 &0 &0 &0 &0 &1 &0 \\
0 &0 &0 &0 &0 &0 &0\\
0 &0 &0 &0 &0 &0 &0
\end{pmatrix}.

Calculez la matrice de passage associée au processus et son inverse

Vous cherchez une matrice inversible $Q$ telle que $Q^{-1}AQ = J.$

D’après ce qui précède, vous pouvez choisir $Q = D_1D_2TP$, ce qui est égal à $ID_1D_2TP$ où $I$ est la matrice identité.

Pour obtenir $Q$ vous partez donc de la matrice identité et effectuez les opérations élémentaires suivantes dans cet ordre : $C_2\leftarrow 4C_2$, $C_1\leftarrow 4C_1$, $C_2\leftarrow C_2+C_6$ et la dernière qui place la colonne 7 à la place de la troisième et décale toutes les autres sur la droite, à savoir $(C_1,C_2,C_3,C_4,C_5,C_6,C_7)\rightarrow (C_1,C_2,C_7, C_3,C_4,C_5,C_6).$

\begin{align*}
 I = \begin{pmatrix}
1 & 0 & 0 & 0 &0 &0 &0 \\
0 &1 &0 &0 &0 &0 &0 \\
0 &0 &1 &0 &0 &0 &0 \\
0 &0 &0 &1 &0 &0 &0 \\
0 &0 &0 &0 &1 &0 &0 \\
0 &0 &0 &0 &0 &1 &0\\
0 &0 &0 &0 &0 &0 &1
\end{pmatrix} &\xrightarrow[C_2\leftarrow 4C_2]{}  \begin{pmatrix}
1 & 0 & 0 & 0 &0 &0 &0 \\
0 &4 &0 &0 &0 &0 &0 \\
0 &0 &1 &0 &0 &0 &0 \\
0 &0 &0 &1 &0 &0 &0 \\
0 &0 &0 &0 &1 &0 &0 \\
0 &0 &0 &0 &0 &1 &0\\
0 &0 &0 &0 &0 &0 &1
\end{pmatrix} \xrightarrow[C_1\leftarrow 4C_1]{}  \begin{pmatrix}
4 & 0 & 0 & 0 &0 &0 &0 \\
0 &4 &0 &0 &0 &0 &0 \\
0 &0 &1 &0 &0 &0 &0 \\
0 &0 &0 &1 &0 &0 &0 \\
0 &0 &0 &0 &1 &0 &0 \\
0 &0 &0 &0 &0 &1 &0\\
0 &0 &0 &0 &0 &0 &1
\end{pmatrix}\\
&\xrightarrow[C_2\leftarrow C_2+C_6]{}  \begin{pmatrix}
4 & 0 & 0 & 0 &0 &0 &0 \\
0 &4 &0 &0 &0 &0 &0 \\
0 &0 &1 &0 &0 &0 &0 \\
0 &0 &0 &1 &0 &0 &0 \\
0 &0 &0 &0 &1 &0 &0 \\
0 &1 &0 &0 &0 &1 &0\\
0 &0 &0 &0 &0 &0 &1
\end{pmatrix} \\
&\xrightarrow[(C_1,C_2,C_3,C_4,C_5,C_6,C_7)\rightarrow (C_1,C_2,C_7, C_3,C_4,C_5,C_6)]{}  Q=\begin{pmatrix}
4 & 0&0 & 0 & 0 &0 &0 \\
0 &4&0 &0 &0 &0 &0  \\
0 &0&0 &1 &0 &0 &0  \\
0 &0&0 &0 &1 &0 &0  \\
0 &0&0 &0 &0 &1 &0  \\
0 &1&0 &0 &0 &0 &1 \\
0 &0&1 &0 &0 &0 &0 
\end{pmatrix}.
\end{align*}

D’après ce qui précède, vous avez $Q^{-1} = (D_1D_2TP)^{-1} = P^{-1}T^{-1}D_2^{-1}D_1^{-1}$, ce qui est égal à $P^{-1}T^{-1}D_2^{-1}D_1^{-1}I.$

Pour obtenir $Q^{-1}$ vous partez donc de la matrice identité et effectuez les opérations élémentaires suivantes dans cet ordre : $L_2\leftarrow 1/4 L_2$, $L_1\leftarrow 1/4 L_1$, $L_6\leftarrow L_6-L_2$ et la dernière qui place la ligne 7 à la place de la troisième et décale toutes les autres sur le fond, à savoir $(L_1,L_2,L_3,L_4,L_5,L_6,L_7)\rightarrow (L_1,L_2,L_7, L_3,L_4,L_5,L_6).$

\begin{align*}
 I = \begin{pmatrix}
1 & 0 & 0 & 0 &0 &0 &0 \\
0 &1 &0 &0 &0 &0 &0 \\
0 &0 &1 &0 &0 &0 &0 \\
0 &0 &0 &1 &0 &0 &0 \\
0 &0 &0 &0 &1 &0 &0 \\
0 &0 &0 &0 &0 &1 &0\\
0 &0 &0 &0 &0 &0 &1
\end{pmatrix} &\xrightarrow[L_2\leftarrow 1/4 L_2]{}  \begin{pmatrix}
1 & 0 & 0 & 0 &0 &0 &0 \\
0 &1/4 &0 &0 &0 &0 &0 \\
0 &0 &1 &0 &0 &0 &0 \\
0 &0 &0 &1 &0 &0 &0 \\
0 &0 &0 &0 &1 &0 &0 \\
0 &0 &0 &0 &0 &1 &0\\
0 &0 &0 &0 &0 &0 &1
\end{pmatrix} \xrightarrow[L_1\leftarrow 1/4 L_1]{}  \begin{pmatrix}
1/4 & 0 & 0 & 0 &0 &0 &0 \\
0 &1/4 &0 &0 &0 &0 &0 \\
0 &0 &1 &0 &0 &0 &0 \\
0 &0 &0 &1 &0 &0 &0 \\
0 &0 &0 &0 &1 &0 &0 \\
0 &0 &0 &0 &0 &1 &0\\
0 &0 &0 &0 &0 &0 &1
\end{pmatrix}\\
&\xrightarrow[L_6\leftarrow L_6-L_2]{}  \begin{pmatrix}
1/4 & 0 & 0 & 0 &0 &0 &0 \\
0 &1/4 &0 &0 &0 &0 &0 \\
0 &0 &1 &0 &0 &0 &0 \\
0 &0 &0 &1 &0 &0 &0 \\
0 &0 &0 &0 &1 &0 &0 \\
0 &-1/4 &0 &0 &0 &1 &0\\
0 &0 &0 &0 &0 &0 &1
\end{pmatrix} \\
&\xrightarrow[(L_1,L_2,L_3,L_4,L_5,L_6,L_7)\rightarrow (L_1,L_2,L_7, L_3,L_4,L_5,L_6)]{}  Q^{-1}=\begin{pmatrix}
1/4 & 0 & 0 & 0 &0 &0 &0 \\
0 &1/4 &0 &0 &0 &0 &0 \\
0 &0 &0 &0 &0 &0 &1\\
0 &0 &1 &0 &0 &0 &0 \\
0 &0 &0 &1 &0 &0 &0 \\
0 &0 &0 &0 &1 &0 &0 \\
0 &-1/4 &0 &0 &0 &1 &0
\end{pmatrix}.
\end{align*}

Concluez

Vous avez obtenu $Q^{-1}AQ = J$ avec :

\begin{align*}
A = \begin{pmatrix}
0 & 1 & 0 & 0 &0 &0 &0 \\
0 &0 &0 &0 &0 &0 &4 \\
0 &0 &0 &1 &0 &0 &0 \\
0 &0 &0 &0 &1 &0 &0 \\
0 &0 &0 &0 &0 &0 &0 \\
0 &0 &0 &0 &0 &0 &1\\
0 &0 &0 &0 &0 &0 &0
\end{pmatrix} \qquad J = \begin{pmatrix}
0 & 1 & 0 & 0 &0 &0 &0 \\
0 &0 &1 &0 &0 &0 &0 \\
0 &0 &0 &0 &0 &0 &0 \\
0 &0 &0 &0 &1 &0 &0 \\
0 &0 &0 &0 &0 &1 &0 \\
0 &0 &0 &0 &0 &0 &0\\
0 &0 &0 &0 &0 &0 &0
\end{pmatrix}
\end{align*}
\begin{align*}
Q = \begin{pmatrix}
4 & 0&0 & 0 & 0 &0 &0 \\
0 &4&0 &0 &0 &0 &0  \\
0 &0&0 &1 &0 &0 &0  \\
0 &0&0 &0 &1 &0 &0  \\
0 &0&0 &0 &0 &1 &0  \\
0 &1&0 &0 &0 &0 &1 \\
0 &0&1 &0 &0 &0 &0 
\end{pmatrix} \qquad Q^{-1} = \begin{pmatrix}
1/4 & 0 & 0 & 0 &0 &0 &0 \\
0 &1/4 &0 &0 &0 &0 &0 \\
0 &0 &0 &0 &0 &0 &1\\
0 &0 &1 &0 &0 &0 &0 \\
0 &0 &0 &1 &0 &0 &0 \\
0 &0 &0 &0 &1 &0 &0 \\
0 &-1/4 &0 &0 &0 &1 &0
\end{pmatrix}.
\end{align*}

Partagez !

Diffusez cet article auprès de vos connaissances susceptibles d'être concernées en utilisant les boutons de partage ci-dessous.

Aidez-moi sur Facebook !

Vous appréciez cet article et souhaitez témoigner du temps que j'y ai passé pour le mettre en œuvre. C'est rapide à faire pour vous et c'est important pour moi, déposez un j'aime sur ma page Facebook. Je vous en remercie par avance.

Lisez d'autres articles !

Parcourez tous les articles qui ont été rédigés. Vous en trouverez sûrement un qui vous plaira !