主要内容

高等代数教学辅导

8.2 标准正交基

建设中!

子节 8.2.1 主要知识点

定义 8.2.1.

设内积空间\(V\) 中一组非零向量\(\alpha_1,\alpha_2,\cdots ,\alpha_m\)满足
\begin{equation*} \left(\alpha_i,\alpha_j\right)=0,\ (i\neq j,\ i,j=1,2,\cdots ,m), \end{equation*}
则称它们为一正交向量组。若还满足
\begin{equation*} \|\alpha_i\|=1,\ (i=1,2,\cdots ,m), \end{equation*}
则称为\(V\)标准正交向量组

备注 8.2.3.

  1. 内积空间中线性无关向量组未必是正交向量组;
  2. \(n\)维内积空间中正交向量组所含向量个数不超过\(n\)
  3. \(n\)维内积空间中任意\(n\)个向量,若构成正交向量组,则它必为\(V\)的一个基。

定义 8.2.4.

内积空间\(V\)中的一个两两正交的向量组构成的基称为正交基。若正交基的每个向量都是单位向量,则称为标准正交基

备注 8.2.5.

  1. 标准正交基\(\xi_1,\xi_2,\cdots ,\xi_n\)下,向量的坐标可用内积表示。
    \(\alpha=a_1\xi_1+a_2\xi_2+\cdots +a_n\xi_n\),则\(a_i=\left(\alpha,\xi_i\right)\),即
    \begin{equation*} \alpha=\left(\alpha,\xi_1\right)\xi_1+\left(\alpha,\xi_2\right)\xi_2+\cdots +\left(\alpha,\xi_n\right)\xi_n. \end{equation*}
  2. 标准正交基下,内积有特别简单的表示式。
    \(\alpha=a_1\xi_1+a_2\xi_2+\cdots +a_n\xi_n, \beta=b_1\xi_1+b_2\xi_2+\cdots +b_n\xi_n\),则
    \begin{equation*} \left(\alpha,\beta\right)=a_1\bar{b}_1+a_2\bar{b}_2+\cdots+a_n\bar{b}_n=\sum\limits_{i=1}^n\left(\alpha,\xi_i\right)\overline{\left(\beta,\xi_i\right)}. \end{equation*}
  3. 标准正交基下向量的长度计算特别简单。
    \begin{equation*} \|\alpha\|=\sqrt{\left|a_1\right|^2+\left|a_2\right|^2+\cdots +\left|a_n\right|^2}. \end{equation*}
Gram-Schmidt正交化过程
  1. 先把线性无关的向量组\(\alpha_1,\alpha_2,\cdots ,\alpha_s\)化成正交向量组\(\beta_1,\beta_2,\cdots ,\beta_s\),令
    \begin{equation*} \begin{array}{ccl} \beta_1&=&\alpha_1,\\ \beta_2&=&\alpha_2-\frac{\left(\alpha_2,\beta_1\right)}{\left(\beta_1,\beta_1\right)}\beta_1,\\ \beta_3&=&\alpha_3-\frac{\left(\alpha_3,\beta_1\right)}{\left(\beta_1,\beta_1\right)}\beta_1-\frac{\left(\alpha_3,\beta_2\right)}{\left(\beta_2,\beta_2\right)}\beta_2,\\ &\vdots&\\ \beta_s&=&\alpha_s-\frac{\left(\alpha_s,\beta_1\right)}{\left(\beta_1,\beta_1\right)}\beta_1-\frac{\left(\alpha_s,\beta_2\right)}{\left(\beta_2,\beta_2\right)}\beta_2-\cdots -\frac{\left(\alpha_s,\beta_{s-1}\right)}{\left(\beta_{s-1},\beta_{s-1}\right)}\beta_{s-1};\\ \end{array} \end{equation*}
  2. 再单位化,令
    \begin{equation*} \gamma_i=\frac{1}{\|\beta_i\|}\beta_i,\ i=1,2,\cdots ,s, \end{equation*}
    则得标准正交向量组\(\gamma_1,\gamma_2,\cdots ,\gamma_s\)满足
    \begin{equation*} \langle\alpha_1,\cdots ,\alpha_r\rangle=\langle\gamma_1,\cdots ,\gamma_r\rangle ,r=1,2,\cdots ,s \mbox{。} \end{equation*}

备注 8.2.8.

\(\langle\alpha_1,\alpha_2,\cdots,\alpha_i\rangle =\langle\gamma_1,\gamma_2,\cdots,\gamma_i\rangle ,(i=1,2,\cdots ,n)\)可知,若
\begin{equation*} (\gamma_1,\gamma_2,\cdots,\gamma_n)=(\alpha_1,\alpha_2,\cdots,\alpha_n)T, \end{equation*}
则过渡矩阵\(T=\left(t_{ij}\right)\)是上三角矩阵(即\(t_{ij}=0,\ i>j\))。
  • \(\xi_1,\xi_2,\cdots ,\xi_n\)\(\eta_1,\eta_2,\cdots,\eta_n\)\(n\)维内积空间\(V\)的两个标准正交基,\(Q\)是从标准正交基\(\xi_1,\xi_2,\cdots ,\xi_n\)到标准正交基\(\eta_1,\eta_2,\cdots,\eta_n\)的过渡矩阵, 即
    \begin{equation*} (\eta_1,\eta_2,\cdots,\eta_n)=(\xi_1,\xi_2,\cdots ,\xi_n)Q, \end{equation*}
  • \(Q=\left(q_{ij}\right)\),则\(Q\)\(n\)阶复方阵且
    \begin{equation*} \overline{Q}^TQ=E\mbox{。} \end{equation*}

定义 8.2.10.

\(Q\)是实\(n\)阶方阵,若\(Q^TQ=E\),则称\(Q\)正交矩阵
\(U\)是复\(n\)阶方阵,若\(\overline{U}^TU=E\),则称\(U\)酉矩阵

定义 8.2.18.

\(U,W\)是内积空间\(V\)的子空间,\(\beta\in V\)
  • 若对任意\(\alpha\in U\),有\((\alpha,\beta)=0\),则称\(\beta\)与子空间\(U\)正交,记为\(\beta\bot U\)
  • 若对任意\(\alpha\in U,\gamma\in W\),有
    \begin{equation*} (\alpha,\gamma)=0, \end{equation*}
    则称\(U\)\(W\)正交,记为\(U\bot W\)

定义 8.2.21.

如果内积空间\(V\)的子空间\(U,W\)满足
  1. \(U\bot W\)
  2. \(U+W=V\)
则称\(W\)\(U\)正交补空间

备注 8.2.23.

子空间\(U\)的正交补记为\(U^{\bot}\),即
\begin{equation*} U^{\bot}=\{\alpha\in V|\ \alpha\bot U\}\mbox{。} \end{equation*}

定义 8.2.24.

\(V, W\)是两个内积空间,\(\varphi:V\rightarrow W\)是线性映射。如果\(\varphi\) 是线性空间同构且保持内积, 即对任意\(\alpha ,\beta\in V\),总成立
\begin{equation*} (\varphi(\alpha),\varphi (\beta))=(\alpha,\beta), \end{equation*}
则称\(\varphi\)内积空间的同构映射,也称\(V, W\)同构的内积空间, 并记为\(V\cong W\)

备注 8.2.25.

内积空间的同构关系满足:
  • 反身性;
  • 对称性;
  • 传递性。

练习 8.2.2 练习

1.

\(\mathbb{C}^{n\times n}\)上,定义内积为\(\left(A,B\right)=\mbox{tr}(A\overline{B}^T)\),试证:\(E_{ij}(i,j=1,2,\cdots ,n)\)是关于此内积的一个标准正交基。
解答.
因为
\begin{equation*} \begin{array}{ccl}\left(E_{ij},E_{kl}\right)&=&\mbox{tr}(E_{ij}\overline{E_{kl}}^T)\\&=&\mbox{tr}(E_{ij}E_{lk})\\&=&\left\{\begin{array}{ll} \mbox{tr}(E_{ik})&\mbox{当}j=l,\\ \mbox{tr}(0)&\mbox{当}j\neq l, \end{array}\right.\\&=&\left\{\begin{array}{ll} 1&\mbox{当}i=k\mbox{且}j=l,\\ 0&\mbox{当}i\neq k\mbox{或}j\neq l, \end{array}\right.\end{array} \end{equation*}
所以\(E_{ij}(i,j=1,2,\cdots ,n)\)是关于此内积的一个标准正交基。

2.

在4维酉空间\(\mathbb{C}^4\)中,求与向量组
\begin{equation*} \alpha_1=(1,-1,i,1)^T,\alpha_2=(2,i,-1+i,1)^T \end{equation*}
等价的一个标准正交向量组。
解答.
先正交化,令
\begin{equation*} \begin{array}{l} \beta_1=\alpha_1=(1,-1,i,1)^T,\\ \begin{array}{ccl}\beta_2&=&\alpha_2-\frac{(\alpha_2,\beta_1)}{(\beta_1,\beta_1)}\beta_1\\&=&(2,i,-1+i,1)^T-\frac{2\times 1+ i\times (-1)+ (-1+i)\times (-i)+ 1\times 1}{1\times 1+(-1)\times (-1)+i\times (-i)+1\times 1}(1,-1,i,1)^T\\ &=&(1,1+i,-1,0)^T, \end{array} \end{array} \end{equation*}
再单位化,令
\begin{equation*} \begin{array}{l} \gamma_1=\frac{1}{\|\beta_1\|}\beta_1=(\frac{1}{2},-\frac{1}{2},\frac{i}{2},\frac{1}{2})^T,\\ \gamma_2=\frac{1}{\|\beta_2\|}\beta_2=(\frac{1}{2},\frac{1+i}{2},-\frac{1}{2},0)^T, \end{array} \end{equation*}
得与向量组\(\alpha_1,\alpha_2\)等价的标准正交向量组
\begin{equation*} \gamma_1=(\frac{1}{2},-\frac{1}{2},\frac{i}{2},\frac{1}{2})^T,\gamma_2=(\frac{1}{2},\frac{1+i}{2},-\frac{1}{2},0)^T. \end{equation*}

3.

在4维欧氏空间\(\mathbb{R}^4\)中,求与向量组
\begin{equation*} \alpha_1=(1,2,2,-1)^T,\alpha_2=(1,1,-5,3)^T,\alpha_3=(3,2,8,-7)^T \end{equation*}
等价的一个标准正交向量组。
解答.
先正交化,令
\begin{equation*} \begin{array}{ccl} \beta_1&=&\alpha_1=(1,2,2,-1)^T,\\ \beta_2&=&\alpha_2-\frac{(\alpha_2,\beta_1)}{(\beta_1,\beta_1)}\beta_1\\ &=&(1,1,-5,3)^T-\frac{1+2-10-3}{1+4+4+1}(1,2,2,-1)^T\\ &=&(2,3,-3,2)^T,\\ \beta_3&=&\alpha_3-\frac{(\alpha_3,\beta_1)}{(\beta_1,\beta_1)}\beta_1-\frac{(\alpha_3,\beta_2)}{(\beta_2,\beta_2)}\beta_2\\ &=&(3,2,8,-7)^T-\frac{3+4+16+7}{1+4+4+1}(1,2,2,-1)^T-\frac{6+6-24-14}{4+9+9+4}(2,3,-3,2)^T\\ &=&(2,-1,-1,-2)^T, \end{array} \end{equation*}
再单位化,令\(\gamma_1=\frac{1}{\|\beta_1\|}\beta_1=\frac{1}{\sqrt{10}}(1,2,2,-1)^T,\) \(\gamma_2=\frac{1}{\|\beta_2\|}\beta_2=\frac{1}{\sqrt{26}}(2,3,-3,2)^T,\)
\begin{equation*} \gamma_3=\frac{1}{\|\beta_3\|}\beta_3=\frac{1}{\sqrt{10}}(2,-1,-1,-2)^T, \end{equation*}
得与向量组\(\alpha_1,\alpha_2,\alpha_3\)等价的标准正交向量组\(\gamma_1,\gamma_2,\gamma_3\)

4.

证明:\(\begin{pmatrix} \cos\theta&\sin\theta\\ -\sin\theta&\cos\theta \end{pmatrix},\begin{pmatrix} \cos\theta&\sin\theta\\ \sin\theta&-\cos\theta \end{pmatrix}\)是正交矩阵且二阶正交矩阵只能是如上两种形式。
解答.
因为
\begin{equation*} \begin{pmatrix} \cos\theta&\sin\theta\\ -\sin\theta&\cos\theta \end{pmatrix}^T\begin{pmatrix} \cos\theta&\sin\theta\\ -\sin\theta&\cos\theta \end{pmatrix}=E_2, \begin{pmatrix} \cos\theta&\sin\theta\\ \sin\theta&-\cos\theta \end{pmatrix}^T\begin{pmatrix} \cos\theta&\sin\theta\\ \sin\theta&-\cos\theta \end{pmatrix}=E_2, \end{equation*}
所以\(\begin{pmatrix} \cos\theta&\sin\theta\\ -\sin\theta&\cos\theta \end{pmatrix},\begin{pmatrix} \cos\theta&\sin\theta\\ \sin\theta&-\cos\theta \end{pmatrix}\)是正交矩阵。 设\(Q=\begin{pmatrix} q_{11}&q_{12}\\q_{21}&q_{22} \end{pmatrix}\)是正交矩阵,由\(Q^TQ=E\)可知
\begin{equation*} q_{11}^2+q_{21}^2=1,q_{12}^2+q_{22}^2=1,q_{11}q_{12}+q_{21}q_{22}=0, \end{equation*}
不妨设\(q_{11}=\cos\theta ,q_{21}=\sin\theta\),则
\begin{equation*} q_{12}=\sin\theta,q_{22}=-\cos\theta\mbox{或}q_{12}=-\sin\theta,q_{22}=\cos\theta. \end{equation*}

5.

\(\alpha_1,\alpha_2,\cdots ,\alpha_n\)是内积空间\(V\)\(n\)个线性无关向量,\(\beta_1,\beta_2,\cdots ,\beta_n\)是这组向量经过正交化得到的向量组,证明:
\begin{equation*} \begin{vmatrix} \left(\alpha_1,\alpha_1\right)&\left(\alpha_1,\alpha_2\right)&\cdots&\left(\alpha_1,\alpha_n\right)\\ \left(\alpha_2,\alpha_1\right)&\left(\alpha_2,\alpha_2\right)&\cdots&\left(\alpha_2,\alpha_n\right)\\ \cdots&\cdots&\cdots&\cdots\\ \left(\alpha_n,\alpha_1\right)&\left(\alpha_n,\alpha_2\right)&\cdots&\left(\alpha_n,\alpha_n\right) \end{vmatrix}=\prod\limits_{i=1}^n\left(\beta_i,\beta_i\right)\mbox{。} \end{equation*}
解答.
由正交化过程知,\((\alpha_1,\alpha_2,\cdots ,\alpha_n)=(\beta_1,\beta_2,\cdots ,\beta_n)A\),其中
\begin{equation*} A=(a_{ij})_{n\times n}=\begin{pmatrix} 1&a_{12}&\cdots&a_{1,n-1}&a_{1n}\\ 0&1&\cdots&a_{2,n-1}&a_{2n}\\ \cdots&\cdots&\cdots&\cdots&\cdots\\ 0&0&\dots&1&a_{n-1,n}\\ 0&0&\cdots&0&1 \end{pmatrix}, \end{equation*}
\(\det A=1\)\((\alpha_i,\alpha_j)=(\sum\limits_{k=1}^n a_{ki}\beta_k,\sum\limits_{l=1}^n a_{lj}\beta_l)=\sum\limits_{k,l=1}^n a_{ki}\overline{a_{lj}}(\beta_k, \beta_l)\)
因此,
\begin{equation*} \begin{array}{cl} &\begin{vmatrix} \left(\alpha_1,\alpha_1\right)&\left(\alpha_1,\alpha_2\right)&\cdots&\left(\alpha_1,\alpha_n\right)\\ \left(\alpha_2,\alpha_1\right)&\left(\alpha_2,\alpha_2\right)&\cdots&\left(\alpha_2,\alpha_n\right)\\ \cdots&\cdots&\cdots&\cdots\\ \left(\alpha_n,\alpha_1\right)&\left(\alpha_n,\alpha_2\right)&\cdots&\left(\alpha_n,\alpha_n\right) \end{vmatrix}\\=&\begin{vmatrix} \sum\limits_{k,l=1}^n a_{k1}\overline{a_{l1}}(\beta_k, \beta_l)&\sum\limits_{k,l=1}^n a_{k1}\overline{a_{l2}}(\beta_k, \beta_l)&\cdots&\sum\limits_{k,l=1}^n a_{k1}\overline{a_{ln}}(\beta_k, \beta_l)\\ \sum\limits_{k,l=1}^n a_{k2}\overline{a_{l1}}(\beta_k, \beta_l)&\sum\limits_{k,l=1}^n a_{k2}\overline{a_{l2}}(\beta_k, \beta_l)&\cdots&\sum\limits_{k,l=1}^n a_{k2}\overline{a_{ln}}(\beta_k, \beta_l)\\ \cdots&\cdots&\cdots&\cdots\\ \sum\limits_{k,l=1}^n a_{kn}\overline{a_{l1}}(\beta_k, \beta_l)&\sum\limits_{k,l=1}^n a_{kn}\overline{a_{l2}}(\beta_k, \beta_l)&\cdots&\sum\limits_{k,l=1}^n a_{kn}\overline{a_{ln}}(\beta_k, \beta_l) \end{vmatrix}\\ =&\det\left(A^T\begin{pmatrix} \left(\beta_1,\beta_1\right)&\left(\beta_1,\beta_2\right)&\cdots&\left(\beta_1,\beta_n\right)\\ \left(\beta_2,\beta_1\right)&\left(\beta_2,\beta_2\right)&\cdots&\left(\beta_2,\beta_n\right)\\ \cdots&\cdots&\cdots&\cdots\\ \left(\beta_n,\beta_1\right)&\left(\beta_n,\beta_2\right)&\cdots&\left(\beta_n,\beta_n\right) \end{pmatrix}\overline{A}\right)\\ =&\det A^T \cdot\det\left({\rm diag}(\left(\beta_1,\beta_1\right),\left(\beta_2,\beta_2\right),\cdots ,\left(\beta_n,\beta_n\right))\right)\cdot\det\overline{A}=\prod\limits_{i=1}^n\left(\beta_i,\beta_i\right).\end{array} \end{equation*}

6.

\(A\)\(n\)阶实或复可逆矩阵,证明:\(A\)的QR-分解是唯一的。
解答.
\(A=Q_1R_1=Q_2R_2\),其中\(Q_1,Q_2\)是酉矩阵,\(R_1,R_2\)是对角元均为正实数的上三角矩阵,则
\begin{equation*} Q_1^{-1}Q_2=R_1R_2^{-1} . \end{equation*}
一方面,酉矩阵的逆仍是酉矩阵,两个酉矩阵的积也是酉矩阵,所以\(Q_1^{-1}Q_2\)是酉矩阵。另一方面,对角元均为正实数的上三角矩阵的逆矩阵仍是对角元均为正实数的上三角阵,两个对角元均为正实数的上三角矩阵的乘积仍是对角元均为正实数的上三角矩阵,所以\(R_1R_2^{-1}\)是对角元均为正实数的上三角矩阵。注意到对角元均为正实数的上三角酉矩阵只能是单位矩阵,所以
\begin{equation*} Q_1^{-1}Q_2=R_1R_2^{-1}=E_m, \end{equation*}
\(Q_1=Q_2,\ R_1=R_2\)

7.

\(A=\begin{pmatrix} 1&1&0\\1&0&1\\-1&0&0 \end{pmatrix}\),求正交矩阵\(Q\)和上三角矩阵\(R\)(对角元均大于0),使得\(A=QR\)
解答.
\(A=(\alpha_1,\alpha_2,\alpha_3)\),其中\(\alpha_1=(1,1,-1)^T,\alpha_2=(1,0,0)^T,\alpha_3=(0,1,0)^T\),令
\begin{equation*} \begin{array}{l} \beta_1=\alpha_1=(1,1,-1)^T,\\ \beta_2=\alpha_2-\frac{(\alpha_2,\beta_1)}{(\beta_1,\beta_1)}\beta_1=\alpha_2-\frac{1}{3}\beta_1=(\frac{2}{3},-\frac{1}{3},\frac{1}{3})^T,\\ \beta_3=\alpha_3-\frac{(\alpha_3,\beta_1)}{(\beta_1,\beta_1)}\beta_1-\frac{(\alpha_3,\beta_2)}{(\beta_2,\beta_2)}\beta_2=\alpha_3-\frac{1}{3}\beta_1+\frac{1}{2}\beta_2=(0,\frac{1}{2},\frac{1}{2})^T, \end{array} \end{equation*}
再单位化,得
\begin{equation*} \begin{array}{c}\gamma_1=\frac{1}{\|\beta_1\|}\beta_1=\frac{\sqrt{3}}{3}\beta_1=(\frac{\sqrt{3}}{3},\frac{\sqrt{3}}{3},-\frac{\sqrt{3}}{3})^T,\\ \gamma_2=\frac{1}{\|\beta_2\|}\beta_2=\frac{\sqrt{6}}{2}\beta_2=(\frac{\sqrt{6}}{3},-\frac{\sqrt{6}}{6},\frac{\sqrt{6}}{6})^T,\\ \gamma_3=\frac{1}{\|\beta_3\|}\beta_3=\sqrt{2}\beta_3=(0,\frac{\sqrt{2}}{2},\frac{\sqrt{2}}{2})^T, \end{array} \end{equation*}
\(\gamma_1,\gamma_2,\gamma_3\)\(\mathbb{R}^3\)的一个标准正交基。
\(Q=(\gamma_1,\gamma_2,\gamma_3)=\begin{pmatrix} \frac{\sqrt{3}}{3}&\frac{\sqrt{6}}{3}&0\\ \frac{\sqrt{3}}{3}&-\frac{\sqrt{6}}{6}&\frac{\sqrt{2}}{2}\\ -\frac{\sqrt{3}}{3}&\frac{\sqrt{6}}{6}&\frac{\sqrt{2}}{2} \end{pmatrix}\),则\(Q\)是正交矩阵。注意到
\begin{equation*} \begin{array}{l} \alpha_1=\beta_1=\sqrt{3}\gamma_1,\\ \alpha_2=\frac{1}{3}\beta_1+\beta_2=\frac{\sqrt{3}}{3}\gamma_1+\frac{\sqrt{6}}{3}\gamma_2,\\ \alpha_3=\frac{1}{3}\beta_1-\frac{1}{2}\beta_2+\beta_3=\frac{\sqrt{3}}{3}\gamma_1-\frac{\sqrt{6}}{6}\gamma_2+\frac{\sqrt{2}}{2}\gamma_3, \end{array} \end{equation*}
所以\(A=QR\),其中
\begin{equation*} Q=(\gamma_1,\gamma_2,\gamma_3)=\begin{pmatrix} \frac{\sqrt{3}}{3}&\frac{\sqrt{6}}{3}&0\\ \frac{\sqrt{3}}{3}&-\frac{\sqrt{6}}{6}&\frac{\sqrt{2}}{2}\\ -\frac{\sqrt{3}}{3}&\frac{\sqrt{6}}{6}&\frac{\sqrt{2}}{2} \end{pmatrix},R=\begin{pmatrix} \sqrt{3}&\frac{\sqrt{3}}{3}&\frac{\sqrt{3}}{3}\\ 0&\frac{\sqrt{6}}{3}&-\frac{\sqrt{6}}{6}\\ 0&0&\frac{\sqrt{2}}{2} \end{pmatrix}. \end{equation*}

8.

\(V_1,V_2\)\(n\)维内积空间\(V\)的子空间,证明:
  1. \(\left(V_1^\bot\right)^\bot=V_1\)
  2. \(V_1\subseteq V_2\),则\(V_2^\bot\subseteq V_1^\bot\)
  3. \(\left(V_1+V_2\right)^\bot=V_1^\bot\bigcap V_2^\bot\)
  4. \(\left(V_1\bigcap V_2\right)^\bot=V_1^\bot +V_2^\bot\)
解答.
  1. 一方面,因为\(V=V_1\oplus V_1^\bot, V=V_1^\bot\oplus \left(V_1^\bot\right)^\bot\),所以
    \begin{equation*} \dim\left(V_1^\bot\right)^\bot=n-\dim(V_1^\bot)=\dim V_1. \end{equation*}
    另一方面,对任意\(\alpha\in V_1, \beta\in V_1^\bot\),有\((\alpha,\beta)=0\),所以\(\alpha\in \left(V_1^\bot\right)^\bot\),即\(V_1\subseteq \left(V_1^\bot\right)^\bot\)。从而\(\left(V_1^\bot\right)^\bot=V_1\)
  2. 对任意\(\alpha\in V_2^\bot,\beta\in V_1\),由\(V_1\subseteq V_2\)\(\beta\in V_2\),则\((\alpha,\beta)=0\),故\(\alpha\in V_1^\bot\)。由\(\alpha\)的任意性得\(V_2^\bot\subseteq V_1^\bot\)
  3. 对任意\(\alpha\in\left(V_1+V_2\right)^\bot,\beta\in V_1\),由于\(\beta=\beta+0\in V_1+V_2\),所以\((\alpha,\beta)=0\),故\(\alpha\in V_1^\bot\)。同理,\(\alpha\in V_2^\bot\)。因此\(\left(V_1+V_2\right)^\bot\subseteq V_1^\bot\bigcap V_2^\bot\)
    另一方面,对任意\(\alpha\in V_1^\bot\bigcap V_2^\bot,\beta+ \gamma\in V_1+V_2\),有\((\alpha,\beta)=0=(\alpha,\gamma)\)。从而\((\alpha,\beta+\gamma)=0\)。于是\(\alpha\in \left(V_1+V_2\right)^\bot\),即\(V_1^\bot\bigcap V_2^\bot\subseteq\left(V_1+V_2\right)^\bot\)
    综上,\(\left(V_1+V_2\right)^\bot=V_1^\bot\bigcap V_2^\bot\)
  4. \((1)\)\((3)\)知,\(\left(V_1^\bot+V_2^\bot\right)^\bot=\left(V_1^\bot\right)^\bot\bigcap \left(V_2^\bot\right)^\bot=V_1\bigcap V_2\),从而有\(\left(V_1\bigcap V_2\right)^\bot=V_1^\bot +V_2^\bot\)

9.

\(U\)是下列齐次线性方程组的解空间:
\begin{equation*} \left\{\begin{array}{l} x_1-x_3+x_4=0,\\ x_2+x_3=0, \end{array}\right. \end{equation*}
试求:
  1. \(U^\bot\)
  2. \(U^\bot\)适合的线性方程组。
解答.
  1. \(\displaystyle U^\bot=\langle(1,0,-1,1)^T,(0,1,1,0)^T\rangle.\)
  2. 因为该齐次线性方程组的基础解系为
    \begin{equation*} \alpha_1=(1,-1,1,0)^T,\alpha_2=(-1,0,0,1)^T, \end{equation*}
    所以\(U^\bot\)适合的线性方程组为
    \begin{equation*} \left\{\begin{array}{l} x_1-x_2+x_3=0,\\ -x_1+x_4=0, \end{array}\right. \end{equation*}

10.

\(A\in\mathbb{R}^{m\times n},\beta\in\mathbb{R}^m\),证明:线性方程组\(AX=\beta\)有解的充分必要条件是\(\beta\)\(A^TX=0\)的解空间正交。
解答.
\(V\)\(A^TX=0\)的解空间,\(U\)\(A^T\)的行向量的转置所生成的线性空间,即\(A\)的列向量所生成的线性空间,则\(V^\bot=U\),故
\begin{equation*} \begin{array}{ccl} \beta\mbox{与}A^TX=0\mbox{的解空间正交}&\Leftrightarrow&\beta\in V^\bot\\ &\Leftrightarrow &\beta\in U\\ &\Leftrightarrow&\beta\mbox{可由的列向量组线性表出}\\ &\Leftrightarrow&AX=\beta\mbox{有解} \end{array} \end{equation*}

11.

写出PPT例2.3中的欧氏空间\(\mathbb{R}[x]_2\)\(\mathbb{R}^3\)之间的一个同构映射。
解答.
\(\varepsilon_1,\varepsilon_2,\varepsilon_3\)\(\mathbb{R}^3\)的一个标准正交基,\(\frac{\sqrt{2}}{2},\frac{\sqrt{6}}{2}x,\frac{3\sqrt{10}}{4}x^2-\frac{\sqrt{10}}{4}\)\(\mathbb{R}[x]_2\)的一个标准正交基,所以
\begin{equation*} \varphi:\mathbb{R}^3\rightarrow \mathbb{R}[x]_2,\ (a_1,a_2,a_3)^T\mapsto \frac{3\sqrt{10}}{4}a_3x^2+\frac{\sqrt{6}}{2}a_2x+\frac{\sqrt{2}}{2}a_1-\frac{\sqrt{10}}{4}a_3 \end{equation*}
\(\mathbb{R}^3\)\(\mathbb{R}[x]_2\)的同构映射。