常见的矩阵
设 A A A是 n n n阶矩阵
- 单位阵:主对角元素为 1 1 1,其余元素为 0 0 0的矩阵称为单位阵,记作 E n E_{n} En
- 数量阵:数 k k k与单位阵 E E E的积 k E kE kE称为数量阵
- 对角阵:非对角元素都是 0 0 0的矩阵(即 ∀ i ≠ j 恒有 a i j = 0 \forall i\ne j恒有a_{ij}=0 ∀i=j恒有aij=0)的对角阵,记作 Λ , Λ = d i a g [ a 1 , a 2 , ⋯ , a n ] \Lambda,\Lambda=diag[a_{1},a_{2},\cdots,a_{n}] Λ,Λ=diag[a1,a2,⋯,an]
- 上(下)三角阵:当 i > j ( i < j ) i>j(i<j) i>j(i<j)时,有 a i j = 0 a_{ij}=0 aij=0的矩阵称为上(下)三角阵
- 对称阵:满足 A T = A A^{T}=A AT=A,即 a i j = a j i a_{ij}=a_{ji} aij=aji的矩阵称为对称阵
- 反对称阵:满足 A T = − A A^{T}=-A AT=−A,即 a i j = − a j i , a i i = 0 a_{ij}=-a_{ji},a_{ii}=0 aij=−aji,aii=0的矩阵称为反对称阵
上述补充在## 一、概念、运算 ### 概念
下
运算
例1:设 α = ( 1 2 , 0 , ⋯ , 0 , 1 2 ) T \alpha=(\frac{1}{2},0,\cdots,0, \frac{1}{2})^{T} α=(21,0,⋯,0,21)T是 n n n维列向量,矩阵 A = E − α α T , B = E + a α α T A=E-\alpha \alpha^{T},B=E+a \alpha \alpha^{T} A=E−ααT,B=E+aααT,若 B B B是 A A A的逆矩阵,则 a = ( ) a=() a=()
对于行向量和列向量,只要出现了 a T a a^{T}a aTa或 a a T aa^{T} aaT,可能结果是常数,尤其是 a T a a T a a^{T}aa^{T}a aTaaTa或 a a T a a T aa^{T}aa^{T} aaTaaT,一定有组合为常数
A
B
=
(
E
−
α
α
T
)
(
E
+
a
α
α
T
)
=
E
+
(
a
−
1
)
α
α
T
−
a
α
α
T
α
⏟
常数
α
T
=
E
+
(
a
−
1
−
a
α
T
α
)
α
α
T
=
E
+
(
a
−
1
−
a
(
1
2
0
⋯
0
1
2
)
(
1
2
0
⋮
0
1
2
)
)
α
α
T
=
E
+
(
a
−
1
−
1
2
a
)
α
α
T
\begin{aligned} AB&=(E-\alpha \alpha^{T})(E+a \alpha \alpha^{T})\\ &=E+(a-1)\alpha \alpha^{T}-a \alpha \underbrace{\alpha^{T}\alpha}_{常数} \alpha^{T}\\ &=E+(a-1-a \alpha^{T}\alpha)\alpha \alpha^{T}\\ &=E+(a-1-a \begin{pmatrix} \frac{1}{2} & 0 & \cdots & 0 & \frac{1}{2} \end{pmatrix}\begin{pmatrix} \frac{1}{2} \\ 0 \\ \vdots \\ 0 \\ \frac{1}{2} \end{pmatrix})\alpha \alpha^{T}\\ &=E+(a-1- \frac{1}{2}a)\alpha \alpha^{T} \end{aligned}
AB=(E−ααT)(E+aααT)=E+(a−1)ααT−aα常数
αTααT=E+(a−1−aαTα)ααT=E+(a−1−a(210⋯021)⎝
⎛210⋮021⎠
⎞)ααT=E+(a−1−21a)ααT
又因为
B
B
B是
A
A
A的逆矩阵
E
+
(
a
−
1
−
1
2
a
)
α
α
T
=
E
⇔
(
a
−
1
−
1
2
a
)
α
α
T
=
0
⇔
a
−
1
−
1
2
a
=
0
E+(a-1- \frac{1}{2}a)\alpha \alpha^{T}=E\Leftrightarrow(a-1- \frac{1}{2}a)\alpha \alpha^{T}=0\Leftrightarrow a-1- \frac{1}{2}a=0
E+(a−1−21a)ααT=E⇔(a−1−21a)ααT=0⇔a−1−21a=0
所以
a
=
2
a=2
a=2
例2:设 f ( x ) = x 100 + x 99 + ⋯ + x + 1 , A = ( 1 0 0 0 0 1 0 0 0 ) f(x)=x^{100}+x^{99}+\cdots+x+1,A=\begin{pmatrix}1 & 0 & 0 \\ 0 & 0 & 1 \\ 0 & 0 & 0\end{pmatrix} f(x)=x100+x99+⋯+x+1,A=⎝ ⎛100000010⎠ ⎞,求 f ( A ) f(A) f(A)和 f ( A ) − 1 f(A)^{-1} f(A)−1
显然 常数 + 矩阵 常数+矩阵 常数+矩阵的形式不存在,所以,一般的,遇到 常数 + 矩阵 常数+矩阵 常数+矩阵,考虑把常数变为常数倍的单位阵 E E E
A
2
=
(
1
0
0
0
0
1
0
0
0
)
(
1
0
0
0
0
1
0
0
0
)
=
(
1
0
0
0
0
0
0
0
0
)
A^{2}=\begin{pmatrix}1 & 0 & 0 \\ 0 & 0 & 1 \\ 0 & 0 & 0\end{pmatrix}\begin{pmatrix}1 & 0 & 0 \\ 0 & 0 & 1 \\ 0 & 0 & 0\end{pmatrix}=\begin{pmatrix}1 & 0 & 0 \\ 0 & 0 & 0 \\ 0 & 0 & 0\end{pmatrix}
A2=⎝
⎛100000010⎠
⎞⎝
⎛100000010⎠
⎞=⎝
⎛100000000⎠
⎞
又
A
3
=
A
2
A
=
(
1
0
0
0
0
0
0
0
0
)
(
1
0
0
0
0
1
0
0
0
)
=
(
1
0
0
0
0
0
0
0
0
)
=
A
2
A^{3}=A^{2}A=\begin{pmatrix}1 & 0 & 0 \\ 0 & 0 & 0 \\ 0 & 0 & 0\end{pmatrix}\begin{pmatrix}1 & 0 & 0 \\ 0 & 0 & 1 \\ 0 & 0 & 0\end{pmatrix}=\begin{pmatrix}1 & 0 & 0 \\ 0 & 0 & 0 \\ 0 & 0 & 0\end{pmatrix}=A^{2}
A3=A2A=⎝
⎛100000000⎠
⎞⎝
⎛100000010⎠
⎞=⎝
⎛100000000⎠
⎞=A2
可知
A
m
=
A
2
(
m
≥
2
)
A^{m}=A^{2}(m\geq2)
Am=A2(m≥2),故
f
(
A
)
=
A
100
+
⋯
+
A
2
+
A
+
E
=
99
A
2
+
A
+
E
=
(
99
0
0
0
0
0
0
0
0
)
+
(
1
0
0
0
0
1
0
0
0
)
+
(
1
0
0
0
1
0
0
0
1
)
=
(
101
0
0
0
1
1
0
0
1
)
f
(
A
)
−
1
=
(
101
0
0
0
1
1
0
0
1
)
−
1
利用分块矩阵求逆
=
(
1
101
0
0
0
1
−
1
0
0
1
)
\begin{aligned} f(A)&=A^{100}+\cdots+A^{2}+A+E\\ &=99A^{2}+A+E\\ &=\begin{pmatrix}99 & 0 & 0 \\ 0 & 0 & 0 \\ 0 & 0 & 0\end{pmatrix}+\begin{pmatrix}1 & 0 & 0 \\ 0 & 0 & 1 \\ 0 & 0 & 0\end{pmatrix}+\begin{pmatrix}1 & 0 & 0 \\ 0 & 1 & 0 \\ 0 & 0 & 1\end{pmatrix}\\ &=\begin{pmatrix}101 & 0 & 0 \\ 0 & 1 & 1 \\ 0 & 0 & 1\end{pmatrix}\\ f(A)^{-1}&=\begin{pmatrix}101 & 0 & 0 \\ 0 & 1 & 1 \\ 0 & 0 & 1\end{pmatrix}^{-1}利用分块矩阵求逆\\ &=\begin{pmatrix} \frac{1}{101} & 0 & 0 \\ 0 & 1 & -1 \\ 0 & 0 & 1\end{pmatrix} \end{aligned}
f(A)f(A)−1=A100+⋯+A2+A+E=99A2+A+E=⎝
⎛9900000000⎠
⎞+⎝
⎛100000010⎠
⎞+⎝
⎛100010001⎠
⎞=⎝
⎛10100010011⎠
⎞=⎝
⎛10100010011⎠
⎞−1利用分块矩阵求逆=⎝
⎛1011000100−11⎠
⎞
矩阵的初等行变换
- 倍乘初等矩阵,记作
E
(
i
(
k
)
)
E(i(k))
E(i(k))
例如
E ( 2 ( k ) ) = ( 1 0 0 0 k 0 0 0 1 ) E(2(k))=\begin{pmatrix}1 & 0 & 0 \\ 0 & k & 0 \\ 0 & 0 & 1\end{pmatrix} E(2(k))=⎝ ⎛1000k0001⎠ ⎞
E ( 2 ( k ) ) E(2(k)) E(2(k))表示由单位阵 E E E的第二行(或第二列)乘 k ( k ≠ 0 ) k(k\ne0) k(k=0)倍得到的矩阵 - 互换初等矩阵,记作
E
(
i
,
j
)
E(i,j)
E(i,j)
例如
E ( 1 , 2 ) = ( 0 1 0 1 0 0 0 0 1 ) E(1,2)=\begin{pmatrix}0 & 1 & 0 \\ 1 & 0 & 0 \\ 0 & 0 & 1\end{pmatrix} E(1,2)=⎝ ⎛010100001⎠ ⎞
E ( 1 , 2 ) E(1,2) E(1,2)表示由单位阵 E E E的第一、二行(或一、二列)互换得到的矩阵 - 倍加初等矩阵,记作
E
(
i
j
(
k
)
)
E(ij(k))
E(ij(k))
例如
E ( 13 ( k ) ) = ( 1 0 0 0 1 0 k 0 1 ) E(13(k))=\begin{pmatrix}1 & 0 & 0 \\ 0 & 1 & 0 \\ k & 0 & 1\end{pmatrix} E(13(k))=⎝ ⎛10k010001⎠ ⎞
E ( 13 ( k ) ) E(13(k)) E(13(k))表示由单位阵 E E E的第一行的 k k k倍加到第三行得到的矩阵,当看成列变换时,应是 E E E的第三列的 k k k倍加到第一列得到的矩阵
补充在## 三、初等变换、初等矩阵
下
分块矩阵
例3:已知 B , C B,C B,C分别是 m m m阶与 n n n阶可逆矩阵,证明 ( B O O C ) \begin{pmatrix}B & O \\ O & C\end{pmatrix} (BOOC)可逆,且 ( B O O C ) − 1 = ( B − 1 O O C − 1 ) \begin{pmatrix}B & O \\ O & C\end{pmatrix}^{-1}=\begin{pmatrix}B^{-1} & O \\ O & C^{-1}\end{pmatrix} (BOOC)−1=(B−1OOC−1)
由
B
,
C
B,C
B,C均可逆,有
∣
B
O
O
C
∣
=
∣
B
∣
⋅
∣
C
∣
≠
0
\begin{vmatrix} B&O\\O&C \end{vmatrix}=|B|\cdot|C|\ne0
∣
∣BOOC∣
∣=∣B∣⋅∣C∣=0
所以
(
B
O
O
C
)
\begin{pmatrix}B & O \\ O & C\end{pmatrix}
(BOOC)可逆
设
(
B
O
O
C
)
−
1
=
(
X
Y
Z
W
)
\begin{pmatrix}B & O \\ O & C\end{pmatrix}^{-1}=\begin{pmatrix}X & Y \\ Z & W\end{pmatrix}
(BOOC)−1=(XZYW),则
(
B
O
O
C
)
(
X
Y
Z
W
)
=
(
E
O
O
E
)
\begin{pmatrix}B & O \\ O & C\end{pmatrix}\begin{pmatrix}X & Y \\ Z & W\end{pmatrix}=\begin{pmatrix} E & O \\ O & E\end{pmatrix}
(BOOC)(XZYW)=(EOOE)
即
{
B
X
=
E
(
1
)
B
Y
=
0
(
2
)
C
Z
=
0
(
3
)
C
W
=
E
(
4
)
\left\{ \begin{array} { l l } { B X = E } & { ( 1 ) } \\ { B Y = 0 } & { ( 2 ) } \\ { C Z = 0 } & { ( 3 ) } \\ { C W = E } & { ( 4 ) } \end{array}\right.
⎩
⎨
⎧BX=EBY=0CZ=0CW=E(1)(2)(3)(4)
由
(
1
)
(1)
(1)有
X
=
B
−
1
X=B^{-1}
X=B−1
由
(
2
)
(2)
(2),因
B
B
B可逆,有
Y
=
B
−
1
O
=
O
Y=B^{-1}O=O
Y=B−1O=O
类似地
Z
=
O
,
W
=
C
−
1
Z=O,W=C^{-1}
Z=O,W=C−1
所以
(
B
O
O
C
)
−
1
=
(
B
−
1
O
O
C
−
1
)
\begin{pmatrix}B & O \\ O & C\end{pmatrix}^{-1}=\begin{pmatrix}B^{-1} & O \\ O & C^{-1}\end{pmatrix}
(BOOC)−1=(B−1OOC−1)
例4:设 A = ( 0 a 1 0 ⋯ 0 0 0 a 2 ⋯ 0 ⋮ ⋮ ⋮ ⋮ 0 0 0 ⋯ a n − 1 a n 0 0 ⋯ 0 ) A=\begin{pmatrix}0 & a_{1} & 0 & \cdots & 0 \\ 0 & 0 & a_{2} & \cdots & 0 \\ \vdots & \vdots & \vdots & & \vdots \\ 0 & 0 & 0 & \cdots & a_{n-1} \\ a_{n} & 0 & 0 & \cdots & 0\end{pmatrix} A=⎝ ⎛00⋮0ana10⋮000a2⋮00⋯⋯⋯⋯00⋮an−10⎠ ⎞,其中 a i ≠ 0 , i = 1 , 2 , ⋯ , n a_{i}\ne0,i=1,2,\cdots,n ai=0,i=1,2,⋯,n,则 A − 1 = ( ) A^{-1}=() A−1=()
把
A
A
A分块如下
A
=
(
0
a
1
0
⋯
0
0
0
a
2
⋯
0
⋮
⋮
⋮
⋮
0
0
0
⋯
a
n
−
1
a
n
0
0
⋯
0
)
=
(
O
A
1
A
2
O
)
A=\left(\begin{array}{c:cccc}0 & a_{1} & 0 & \cdots & 0 \\ 0 & 0 & a_{2} & \cdots & 0 \\ \vdots & \vdots & \vdots & & \vdots \\ 0 & 0 & 0 & \cdots & a_{n-1} \\\hdashline a_{n} & 0 & 0 & \cdots & 0\end{array}\right)=\begin{pmatrix} O & A_{1} \\ A_{2} & O \end{pmatrix}
A=⎝
⎛00⋮0ana10⋮000a2⋮00⋯⋯⋯⋯00⋮an−10⎠
⎞=(OA2A1O)
由于
(
O
A
B
O
)
−
1
=
(
O
B
−
1
A
−
1
O
)
\begin{pmatrix}O & A \\ B & O\end{pmatrix}^{-1}=\begin{pmatrix}O & B^{-1} \\ A^{-1} & O\end{pmatrix}
(OBAO)−1=(OA−1B−1O),有
A
−
1
=
(
0
0
⋯
0
1
a
n
1
a
1
0
⋯
0
0
0
1
a
2
⋯
0
0
⋮
⋮
⋮
⋮
0
0
⋯
1
a
n
−
1
0
)
A^{-1}=\begin{pmatrix} 0 & 0 & \cdots & 0 & \frac{1}{a_{n}} \\ \frac{1}{a_{1}} & 0 & \cdots & 0 & 0 \\ 0 & \frac{1}{a_{2}} & \cdots & 0 & 0 \\ \vdots & \vdots & & \vdots & \vdots \\ 0 & 0 & \cdots & \frac{1}{a_{n-1}} & 0 \end{pmatrix}
A−1=⎝
⎛0a110⋮000a21⋮0⋯⋯⋯⋯000⋮an−11an100⋮0⎠
⎞