这一节主要是引入了一个新的定义:minimal polynomial。之前看过的教材中对此的定义是degree最低的能让T或者A为0的多项式,其实这个最低degree是有点概念性上的东西,但是这本书由于之前引入了ideal和generator,所以定义起来要严谨得多。比较容易证明的几个结论是: T T T和 [ T ] B [T]_{\mathfrak B} [T]B有相同的minimal polynomial,相似的矩阵有相同的minimal polynomial.
后面的主要内容就是两个定理,Theorem 3说明特征多项式和最小多项式有相同的roots。Theorem 4即有名的Cayley-Hamilton定理, T T T的特征多项式可以annihilate T T T,因此最小多项式整除特征多项式,这一节中对此定理的证明用了行列式的方法。
Exercises
1.Let V V V be a finite-dimensional vector space. What is the minimal polynomial for the identity operator on V V V? What is the minimal polynomial for the zero operator?
Solution: The minimal polynomial for I I I is p = x − 1 p=x-1 p=x−1, since p ( I ) = I − I = 0 p(I)=I-I=0 p(I)=I−I=0. The mininal polynomial for 0 0 0 is p = x p=x p=x.
2.Let a , b , c a,b,c a,b,c be elements of a field F F F, and let A A A be the following 3 × 3 3\times 3 3×3 matrix over F F F:
A = [ 0 0 c 1 0 b 0 1 a ] . A=\begin{bmatrix}0&0&c\\1&0&b\\0&1&a\end{bmatrix}. A=⎣⎡010001cba⎦⎤.
Prove that the characteristic polynomial for A A A is x 3 − a x 2 − b x − c x^3-ax^2-bx-c x3−ax2−bx−c and that this is also the minimal polynomial for A A A.
Solution: We have
det ( x I − A ) = ∣ x 0 − c − 1 x − b 0 − 1 x − a ∣ = x ∣ x − b − 1 x − a ∣ − c ∣ − 1 x 0 − 1 ∣ = x ( x 2 − a x − b ) − c = x 3 − a x 2 − b x − c \begin{aligned}\det (xI-A)&=\begin{vmatrix}x&0&-c\\-1&x&-b\\0&-1&x-a\end{vmatrix}=x\begin{vmatrix}x&-b\\-1&x-a\end{vmatrix}-c\begin{vmatrix}-1&x\\0&-1\end{vmatrix}\\&=x(x^2-ax-b)-c=x^3-ax^2-bx-c\end{aligned} det(xI−A)=∣∣∣∣∣∣x−100x−1−c−bx−a∣∣∣∣∣∣=x∣∣∣∣x−1−bx−a∣∣∣∣−c∣∣∣∣−10x−1∣∣∣∣=x(x2−ax−b)−c=x3−ax2−bx−c
To see this is also the minimal polynomial for A A A, notice that
A 2 = [ 0 0 c 1 0 b 0 1 a ] [ 0 0 c 1 0 b 0 1 a ] = [ 0 c a c 0 b c + a b 1 a b + a 2 ] A^2=\begin{bmatrix}0&0&c\\1&0&b\\0&1&a\end{bmatrix}\begin{bmatrix}0&0&c\\1&0&b\\0&1&a\end{bmatrix}=\begin{bmatrix}0&c&ac\\0&b&c+ab\\1&a&b+a^2\end{bmatrix} A2=⎣⎡010001cba⎦⎤⎣⎡010001cba⎦⎤=⎣⎡001cbaacc+abb+a2⎦⎤
thus for any polynomial f f f of degree 2, we write f = m x 2 + n x + q f=mx^2+nx+q f=mx2+nx+q, then
f ( A ) = m [ 0 c a c 0 b c + a b 1 a b + a 2 ] + n [ 0 0 c 1 0 b 0 1 a ] + q [ 1 0 0 0 1 0 0 0 1 ] f(A)=m\begin{bmatrix}0&c&ac\\0&b&c+ab\\1&a&b+a^2\end{bmatrix}+n\begin{bmatrix}0&0&c\\1&0&b\\0&1&a\end{bmatrix}+q\begin{bmatrix}1&0&0\\0&1&0\\0&0&1\end{bmatrix} f(A)=m⎣⎡001cbaacc+abb+a2⎦⎤+n⎣⎡010001cba⎦⎤+q⎣⎡1