线性代数 03.04 初等矩阵

$\color{blue}{\S 第三章 第四节 初等矩阵}$

$\color{blue}{一、初等矩阵的概念}$

$定义4.由单位矩阵E经过一次初等变换得到的矩阵称为初等矩阵.$

$1.对调两行(列)$
$E[i, j] = \begin{pmatrix} 1 & & & & \\ & \ddots & & \\ & & 1 & & \\ & & & 0 & \cdots & & & 1 & \\ & & & & 1 & & & \\ & & & \vdots & & \ddots & & \vdots & \\ & & & 1 & & \cdots & & 0 & & \\ & & & & & & & & & 1 & & \\ & & & & & & & & & & \ddots & \\ & & & & & & & & & & & 1 \\ \end{pmatrix} \begin{matrix} & \\ & \\ \leftarrow 第i行 & \\ & \\ & \\ \leftarrow 第j行 & \\ & \\ & \\ \end{matrix} \\$

$2.以数k \neq 0乘以某行(列)$
$E[i(k)] = \begin{pmatrix} 1 & & & & & & \\ & \ddots & & & & & \\ & & 1 & & & & & \\ & & & k & & & \\ & & & & 1 & & \\ & & & & & \ddots & \\ & & & & & & 1 \\ \end{pmatrix} \leftarrow 第i行$

$3.以数k乘以某行(列)加到另一行(列)上去.$
$E[i, j(k)] = \begin{pmatrix} 1 & & & & & & \\ & \ddots & & & & & \\ & & 1 & \cdots & k & & \\ & & & \ddots & \vdots & & \\ & & & & 1 & & \\ & & & & & \ddots & \\ & & & & & & 1 \\ \end{pmatrix}\begin{matrix} & \\ & \\ \leftarrow 第i行 & \\ & \\ \leftarrow 第j行 & \\ & \\ & \\ \end{matrix}$

$注意:初等方阵是可逆矩阵,且其逆矩阵仍然是初等方阵.\\ E[i,j]^{-1} = E[i, j] \\ E[i(k)]^{-1} = E[i(\dfrac{1}{k})] \quad k \neq 0 \\ E[i, j(k)]^{-1} = E[i, j(-k)]$

$例1.设A = \begin{pmatrix} a_{11} & a_{12} & a_{13} & a_{14} \\ a_{21} & a_{22} & a_{23} & a_{24} \\ a_{31} & a_{32} & a_{33} & a_{34} \\ \end{pmatrix} \\ 计算E[1, 3]A; AE[1, 3]; E[2(k)]A; AE[2(k)]; E[1, 3(k)]A; AE[1,3(k)].$
$解: E[1, 3]A = \begin{pmatrix} 0 & 0 & 1 \\ 0 & 1 & 0 \\ 1 & 0 & 0 \\ \end{pmatrix}\begin{pmatrix} a_{11} & a_{12} & a_{13} & a_{14} \\ a_{21} & a_{22} & a_{23} & a_{24} \\ a_{31} & a_{32} & a_{33} & a_{34} \\ \end{pmatrix} = \begin{pmatrix} a_{31} & a_{32} & a_{33} & a_{34} \\ a_{21} & a_{22} & a_{23} & a_{24} \\ a_{11} & a_{12} & a_{13} & a_{14} \\ \end{pmatrix} \\ AE[1, 3] = \begin{pmatrix} a_{11} & a_{12} & a_{13} & a_{14} \\ a_{21} & a_{22} & a_{23} & a_{24} \\ a_{31} & a_{32} & a_{33} & a_{34} \\ \end{pmatrix} \begin{pmatrix} 0 & 0 & 1 & 0 \\ 0 & 1 & 0 & 0 \\ 1 & 0 & 0 & 0 \\ 0 & 0 & 0 & 1 \\ \end{pmatrix} = \begin{pmatrix} a_{13} & a_{12} & a_{11} & a_{14} \\ a_{23} & a_{22} & a_{21} & a_{24} \\ a_{33} & a_{32} & a_{31} & a_{34} \\ \end{pmatrix} \\ E[2(k)]A = \begin{pmatrix} a_{11} & a_{12} & a_{13} & a_{14} \\ ka_{21} & ka_{22} & ka_{23} & ka_{24} \\ a_{31} & a_{32} & a_{33} & a_{34} \\ \end{pmatrix} \\ AE[2(k)] = \begin{pmatrix} a_{11} & ka_{12} & a_{13} & a_{14} \\ a_{21} & ka_{22} & a_{23} & a_{24} \\ a_{31} & ka_{32} & a_{33} & a_{34} \\ \end{pmatrix} \\ E[1, 3(k)]A = \begin{pmatrix} a_{11}+ka_{31} & a_{12}+ka_{32} & a_{13}+ka_{33} & a_{14}+ka_{34} \\ a_{21} & a_{22} & a_{23} & a_{24} \\ a_{31} & a_{32} & a_{33} & a_{34} \\ \end{pmatrix} \\ AE[1,3(k)] = \begin{pmatrix} a_{11} & a_{12} & ka_{11} + a_{13} & a_{14} \\ a_{21} & a_{22} & ka_{21} + a_{23} & a_{24} \\ a_{31} & a_{32} & ka_{31} + a_{33} & a_{34} \\ \end{pmatrix}$

$\color{blue}{二、初等矩阵的有关定理}$

$定理4.设A是一个m \times n矩阵,\\ 对A实施一次初等行变换,\\ 相当于在A的左边乘以相应的m阶初等矩阵;\\ 对A实施一次初等列变换,\\ 相当于在A的右边乘以相应的n阶初等矩阵.$

$定理5.设A为可逆矩阵,则存在有限个初等矩阵P_1,P_2,\cdots,P_l,使\\ A = P_1P_2 \cdots P_l$
$证:因为A \sim E，故E经过有限次初等变换可变成A,\\ 也就是说,存在有限个初等矩阵P_1, P_2, \cdots, P_l使\\ P_1P_2 \cdots P_r EP_{r + 1} \cdots P_l = A \\ 即\\ A = P_1 P_2 \cdots P_l$

$推论：m \times n矩阵A \sim B的充分必要条件是:存在m阶可逆矩阵P及n阶可逆矩阵Q,使PAQ = B.$

$\color{blue}{三、用初等矩阵求逆矩阵}$

1.利用初等变换求逆
$当|A| \neq 0,由A = P_1P_2 \cdots P_l \\ A^{-1} = (P_1P_2 \cdots P_l)^{-1} = P_l ^{ -1} \cdots P_ 2^{-1} P_1^{-1} \\ 故\\ A^{-1}(A|E) = (A^{-1}A | A^{-1}E) = (E | A^{-1}) \\ 即 P_l^{-1} \cdots P_2^{-1}P_1^{-1}(A | E) = (E | A^{-1})$

$例2.设A = \begin{pmatrix} 1 & 2 & 3 \\ 2 & 2 & 1 \\ 3 & 4 & 3 \\ \end{pmatrix},求A^{-1}$
$解: \\ (A | E) = \begin{pmatrix} 1 & 2 & 3 & 1 & 0 & 0 \\ 2 & 2 & 1 & 0 & 1 & 0 \\ 3 & 4 & 3 & 0 & 0 & 1 \\ \end{pmatrix} \sim \begin{pmatrix} 1 & 2 & 3 & 1 & 0 & 0 \\ 0 & -2 & -5 & -2 & 1 & 0 \\ 0 & -2 & -6 & -3 & 0 & 1 \\ \end{pmatrix} \\ \sim \begin{pmatrix} 1 & 0 & -2 & -1 & 1 & 0 \\ 0 & -2 & -5 & -2 & 1 & 0 \\ 0 & 0 & -1 & -1 & -1 & 1 \\ \end{pmatrix} \sim \begin{pmatrix} 1 & 0 & 0 & 1 & 3 & -2 \\ 0 & -2 & 0 & 3 & 6 & -5 \\ 0 & 0 & 1 & 1 & 1 & -1 \\ \end{pmatrix} \\ \sim \begin{pmatrix} 1 & 0 & 0 & 1 & 3 & -2 \\ 0 & 1 & 0 & -\dfrac{3}{2} & -3 & \dfrac{5}{2} \\ 0 & 0 & 1 & 1 & 1 & -1 \\ \end{pmatrix} \\ 所以A^{-1} = \begin{pmatrix} 1 & 3 & -2 \\ -\dfrac{3}{2} & -3 & \dfrac{5}{2} \\ 1 & 1 & -1 \\ \end{pmatrix}$

$注意：亦可利用矩阵的初等列变换求逆矩阵.\\ 事实上:因为\\ A^{-1} = (P_1P_2 \cdots P_l)^{-1} \\ = P_l^{-1} \cdots P_2^{-1}P_1^{-1} \\ \begin{pmatrix} A \\ E \\ \end{pmatrix}A^{-1} = \begin{pmatrix} AA^{-1} \\ EA^{-1} \\ \end{pmatrix} =\begin{pmatrix} E \\ A^{-1} \\ \end{pmatrix} \\ 所以：\begin{pmatrix} A \\ E \\ \end{pmatrix} \sim \begin{pmatrix} E \\ A^{-1} \\ \end{pmatrix}$

2.利用矩阵的初等行变换求解矩阵方程
$事实上,对于AX = B,若A可逆,则有 \\ X = A^{-1}AX = A^{-1}B \\ 对应于: A^{-1}(A | B) = (E | A^{-1}B) \\ 即 \\ (A | B) \sim (E | A^{-1} B)$

$例3.设AX = B,求X.其中\\ A = \begin{pmatrix} 1 & 2 & 3 \\ 2 & 2 & 1 \\ 3 & 4 & 3 \\ \end{pmatrix}, B = \begin{pmatrix} 2 & 5 \\ 3 & 1 \\ 4 & 3 \\ \end{pmatrix}$
$解: 若A可逆,则X = A^{-1}B \\ (A | B) = \begin{pmatrix} 1 & 2 & 3 &2 & 5 \\ 2 & 2 & 1 & 3 & 1 \\ 3 & 4 & 3 & 4 & 3 \\ \end{pmatrix} \sim \begin{pmatrix} 1 & 2 & 3 & 2 & 5 \\ 0 & -2 & -5 & -1 & -9 \\ 0 & -2 & -6 & -2 & -12 \\ \end{pmatrix} \\ \sim \begin{pmatrix} 1 & 0 & -2 & 1 & -4 \\ 0 & -2 & -5 & -1 & -9 \\ 0 & 0 & -1 & -1 & -3 \\ \end{pmatrix} \sim \begin{pmatrix} 1 & 0 & 0 & 3 & 2 \\ 0 & -2 & 0 & 4 & 6 \\ 0 & 0 & 1 & 1 & 3 \\ \end{pmatrix} \\ \sim \begin{pmatrix} 1 & 0 & 0 & 3 & 2 \\ 0 & 1 & 0 & -2 & -3 \\ 0 & 0 & 1 & 1 & 3 \\ \end{pmatrix} \\ X = A^{-1}B = \begin{pmatrix} 3 & 2 \\ -2 & -3 \\ 1 & 3 \\ \end{pmatrix}$

$同理亦可求解矩阵方程 \\ YA = C \\ 若A可逆,则有 \\ Y = CA^{-1} \\ 即 \\ \begin{pmatrix} A \\ C \\ \end{pmatrix} \sim \begin{pmatrix} E \\ CA^{-1} \\ \end{pmatrix}$

$例4.设A的伴随矩阵\\ A^{*} = \begin{pmatrix} 1 & 0 & 0 & 0 \\ 0 & 1 & 0 & 0 \\ 1 & 0 & 1 & 0 \\ 0 & 0 & 0 & 8 \\ \end{pmatrix},\\ 且有ABA^{-1} = BA^{-1} + 3E,求B.$
$解:在等式两边左乘A^{*},右乘A,得 \\ A^{* }(ABA^{-1})A = A^{*}(BA^{-1} + 3E)A \\ 即 A^{* }ABA^{-1}A = A^{* }BA^{-1}A + 3A^{*}A \quad ① \\ 因为A^{* }A = |A|E,而|A|^3 = |A^{ * }| = 8, \\ 从而有 |A| = 2 \\ 故①式可改写为 2B = A^{*}B + 6E \\ 即(2E - A^{*})B = 6E \\ 所以\\ B = 6(2E - A^{*})^{-1} = 6\begin{pmatrix} 1 & 0 & 0 & 0 \\ 0 & 1 & 0 & 0 \\ -1 & 0 & 1 & 0 \\ 0 & 0 & 0 & -6 \\ \end{pmatrix} ^ {-1} \\ = \begin{pmatrix} 6 & 0 & 0 & 0 \\ 0 & 6 & 0 & 0 \\ 6 & 0 & 6 & 0 \\ 0 & 0 & 0 & -1 \\ \end{pmatrix}$