跳转至

第二章 行列式与矩阵的秩

知识点

  • 排列
    • 标准排列(自然序排列)
  • 方阵的行列式
  • 行列式的性质
  • Laplace 定理
  • 矩阵的秩

Laplace 定理

  • 行列式按某行(列)展开

    • 余子式 \(M_{ij}\):去掉第 \(i\) 行第 \(j\) 列后,剩下的 \(n-1\) 阶行列式

    • 代数余子式 \(A_{ij}\xlongequal{\text{def}}(-1)^{i+j}M_{ij}\)

    • \(|A|=\sum_{k=1}^n a_{ik}A_{ik}=\sum_{k=1}^n a_{ki}A_{ki}\)

      证明

      (下证按行展开。按列展开是同理的)

      下证 \(a_{ik}\)\(|A|\) 的贡献是 \(a_{ik}A_{ik}\)

      通过 \(i-1\) 次行交换、\(k-1\) 次列交换,将 \(a_{ik}\) 挪到左上角的位置,且其他元素的相对位置保持不变,得到新矩阵 \(A'\)

      由于 \(a_{ik}\)\(A'_{11}\) 的位置,所以 \(|A'|\) 中所有包含 \(a_{ik}\) 的项,就是 \(a_{ik}\) 乘以“\(A'\) 去掉第一行第一列后的 \(n-1\) 阶行列式”(即 \(M_{ik}\)),所以 \(|A'|\)\(a_{ik}\) 的贡献就是 \(a_{ik}M_{ik}\)

      \(|A|=(-1)^{(i-1)+(k-1)}|A'|=(-1)^{i+k}|A'|\),所以 \(|A|\)\(a_{ik}\) 的贡献是 \(a_{ik}A_{ik}\)

    • \(\sum_{k=1}^na_{ik}A_{jk}=\sum_{k=1}^n a_{ki}A_{kj}=|A|[i=j]\)

      证明

      \(i=j\) 时已证。

      \(i\neq j\) 时:

      \(A\) 的第 \(j\) 行替换为第 \(i\) 行,得到新矩阵 \(A'\)

      • 由于 \(A'\) 的第 \(i\) 行和第 \(j\) 行相同,故 \(|A'|=0\)
      • \(A'\) 按第 \(j\) 行展开,则 \(|A'|=\sum_{k=1}^n a_{ik}A_{jk}\)

      因此 \(\sum_{k=1}^n a_{ik}A_{jk}=0\)

      Tip

      计算 \(k_1A_{i1}+k_2A_{i2}+\cdots+k_nA_{in}\),可以将矩阵的第 \(i\) 行替换为 \(k_1,k_2,\cdots,k_n\),计算新矩阵的行列式。

    • Vandermonde(范德蒙德)行列式

      \(D_n=\begin{vmatrix}1 & 1 & \dots & 1 \\x_1 & x_2 & \dots & x_n \\x_1^2 & x_2^2 & \dots & x_n^2 \\\vdots & \vdots & \ddots & \vdots \\x_1^{n-1} & x_2^{n-1} & \dots & x_n^{n-1}\end{vmatrix}=\prod_{1 \leq i < j \leq n} (x_j - x_i)\)

      证明

      \(n=2\) 时成立。

      假设 \(n-1\) 阶成立,则

      \[ D_n \xlongequal{R_i-x_1R_{i-1}} \begin{vmatrix} 1 & 1 & 1 & \dots & 1 \\ 0 & x_2 - x_1 & x_3 - x_1 & \dots & x_n - x_1 \\ 0 & x_2(x_2 - x_1) & x_3(x_3 - x_1) & \dots & x_n(x_n - x_1) \\ \vdots & \vdots & \vdots & \ddots & \vdots \\ 0 & x_2^{n-2}(x_2 - x_1) & x_3^{n-2}(x_3 - x_1) & \dots & x_n^{n-2}(x_n - x_1) \end{vmatrix} \]

      按第一列展开:

      \[ \begin{aligned} D_n&= \begin{vmatrix} x_2 - x_1 & x_3 - x_1 & \dots & x_n - x_1 \\ x_2(x_2 - x_1) & x_3(x_3 - x_1) & \dots & x_n(x_n - x_1) \\ \vdots & \vdots & \ddots & \vdots \\ x_2^{n-2}(x_2 - x_1) & x_3^{n-2}(x_3 - x_1) & \dots & x_n^{n-2}(x_n - x_1) \end{vmatrix} \\ &= (x_2 - x_1)(x_3 - x_1)\cdots(x_n - x_1) \begin{vmatrix} 1 & 1 & \cdots & 1 \\ x_2 & x_3 & \cdots & x_n \\ \vdots & \vdots & \ddots & \vdots \\ x_2^{n-2} & x_3^{n-2} & \cdots & x_n^{n-2} \end{vmatrix}\\ &=\prod_{1<j\leq n}(x_j-x_1)\times D_{n-1} \end{aligned} \]

      由归纳法得证。

    • Cramer 法则

      对于 \(n\) 元线性方程组 \(\begin{cases} a_{11}x_1 + a_{12}x_2 + \dots + a_{1n}x_n = b_1 \\ a_{21}x_1 + a_{22}x_2 + \dots + a_{2n}x_n = b_2 \\ \vdots \\ a_{n1}x_1 + a_{n2}x_2 + \dots + a_{nn}x_n = b_n \end{cases}\),若其系数行列式 \(D = \begin{vmatrix} a_{11} & a_{12} & \dots & a_{1n} \\ a_{21} & a_{22} & \dots & a_{2n} \\ \vdots & \vdots & \ddots & \vdots \\ a_{n1} & a_{n2} & \dots & a_{nn} \end{vmatrix}\neq 0\),则方组有唯一解 \(x_i=\frac{D_i}D\),其中 \(D_i\) 是将 \(D\) 的第 \(i\) 列元素替换为方程组的常数项 \(b_1,b_2,\dots,b_n\) 后得到的行列式。

      例题

      \(a_i \neq a_j (i \neq j)\),求解线性方程组 \(\begin{cases}x_1 + a_1x_2 + a_1^2x_3 + \cdots + a_1^{n - 1}x_n = 1 \\x_1 + a_2x_2 + a_2^2x_3 + \cdots + a_2^{n - 1}x_n = 1 \\x_1 + a_3x_2 + a_3^2x_3 + \cdots + a_3^{n - 1}x_n = 1 \\\cdots\cdots\cdots\cdots\cdots\cdots\cdots\cdots\cdots\cdots\cdots\cdots \\x_1 + a_nx_2 + a_n^2x_3 + \cdots + a_n^{n - 1}x_n = 1\end{cases}\)

      解答

      \(D=\begin{vmatrix}1 & a_1 & a_1^2 & \cdots & a_1^{n - 1} \\1 & a_2 & a_2^2 & \cdots & a_2^{n - 1} \\1 & a_3 & a_3^2 & \cdots & a_3^{n - 1} \\\cdots & \cdots & \cdots & \cdots & \cdots \\1 & a_n & a_n^2 & \cdots & a_n^{n - 1}\end{vmatrix} = \prod_{1 \leq i < j \leq n} (a_j - a_i) \neq 0\)

      \(D_1 = D\)

      \(D_j = \begin{vmatrix}1 & a_1 & a_1^2 & \cdots & 1 & \cdots & a_1^{n - 1} \\1 & a_2 & a_2^2 & \cdots & 1 & \cdots & a_2^{n - 1} \\1 & a_3 & a_3^2 & \cdots & 1 & \cdots & a_3^{n - 1} \\\cdots & \cdots & \cdots & \cdots & \cdots & \cdots & \cdots \\1 & a_n & a_n^2 & \cdots & 1 & \cdots & a_n^{n - 1}\end{vmatrix} = 0 \ (j = 2,3,\cdots,n)\)

      故方程组唯一解为:\(x_1 = \frac{D_1}{D} = 1\)\(x_2 = \frac{D_2}{D} = 0\)\(\cdots\)\(x_n = \frac{D_n}{D} = 0\)

      Info

      Cramer 法则的使用条件是方程组个数与未知量个数相同。

      若系数行列式 \(D=0\),则线性方程组有无穷多解或无解。

      \(b\)\(A\) 的某列成比例时有奇效。

  • 行列式按多行(列)展开

    • \(k\) 阶子式:\(D\begin{pmatrix} i_1 \!\!&\!\! i_2 \!\!&\!\! \cdots \!\!&\!\! i_k \\ j_1 \!\!&\!\! j_2 \!\!&\!\! \cdots \!\!&\!\! j_k \end{pmatrix}\),是 \(k\) 阶行列式

    • 余子式:\(M\begin{pmatrix} i_1 \!\!&\!\! i_2 \!\!&\!\! \cdots \!\!&\!\! i_k \\ j_1 \!\!&\!\! j_2 \!\!&\!\! \cdots \!\!&\!\! j_k \end{pmatrix}\),是 \(n-k\) 阶行列式

    • 代数余子式 \(A\begin{pmatrix} i_1 \!\!&\!\! i_2 \!\!&\!\! \cdots \!\!&\!\! i_k \\ j_1 \!\!&\!\! j_2 \!\!&\!\! \cdots \!\!&\!\! j_k \end{pmatrix}\xlongequal{\text{def}}(-1)^{i_1+i_2+\cdots+i_k+j_1+j_2+\cdots+j_k}M\begin{pmatrix} i_1 \!\!&\!\! i_2 \!\!&\!\! \cdots \!\!&\!\! i_k \\ j_1 \!\!&\!\! j_2 \!\!&\!\! \cdots \!\!&\!\! j_k \end{pmatrix}\)

    • Laplace(拉普拉斯)定理

      按第 \(i_1,i_2,\cdots,i_k\,(1\leq i_1<i_2<\cdots<i_k\leq n)\) 展开:

      \(|A|=\sum_{1\leq j_1<j_2<\cdots<j_k\leq n}D \begin{pmatrix} i_1 \!\!&\!\! i_2 \!\!&\!\! \cdots \!\!&\!\! i_k \\ j_1 \!\!&\!\! j_2 \!\!&\!\! \cdots \!\!&\!\! j_k \end{pmatrix} A \begin{pmatrix} i_1 \!\!&\!\! i_2 \!\!&\!\! \cdots \!\!&\!\! i_k \\ j_1 \!\!&\!\! j_2 \!\!&\!\! \cdots \!\!&\!\! j_k \end{pmatrix}\)

    • 推论:

      \(\begin{vmatrix}A_{r\times r}&O_{r\times s}\\C_{s\times r}&B_{s\times s}\end{vmatrix}=\begin{vmatrix}A_{r\times r}&C_{r\times s}\\O_{s\times r}&B_{s\times s}\end{vmatrix}=|A||B|\\\begin{vmatrix}O_{r\times s}&A_{r\times r}\\B_{s\times s}&C_{s\times r}\end{vmatrix}=\begin{vmatrix}C_{r\times s}&A_{r\times r}\\B_{s\times s}&O_{s\times r}\end{vmatrix}=(-1)^{rs}|A||B|\)

      证明

      前一行 \(\text{LHS}=|A|\times (-1)^{(1+2+\cdots r)+(1+2+\cdots r)}|B|=|A||B|\)

      后一行 \(\text{LHS}=|A|\times (-1)^{(1+2+\cdots r)+((s+1)+(s+2)+\cdots+(s+r))}|B|\),而 \((-1)^{(1+2+\cdots r)+((s+1)+(s+2)+\cdots+(s+r))}=(-1)^{2(1+2+\cdots+r)+rs}=(-1)^{rs}\)

矩阵的秩

  • 行阶梯形矩阵(行最简形矩阵),其秩为阶梯头个数。

矩阵的相抵

  • 相抵(等价)

    定义:对于 \(A,B\in\mathbb P^{m\times n}\),若 \(r(A) = r(B)\),则称 \(A\)\(B\) 相抵(或等价),记作 \(A \stackrel{\mathcal{R}}{\sim} B\)

    定理:\(A\stackrel{\mathcal {R}}{\sim}B\) \(\Leftrightarrow\) \(A\) 可经有限次初等变换化为 \(B\)

  • 相抵等价关系

    数学上,称满足自反性、对称性和传递性的关系为一个等价关系。

    对于 \(A,B,C\in \mathbb P^{m\times n}\)

    • 自反性:\(A \stackrel{\mathcal{R}}{\sim} A\)

    • 对称性:若 \(A \stackrel{\mathcal{R}}{\sim} B\),则 \(B \stackrel{\mathcal{R}}{\sim} A\)

    • 传递性:若 \(A \stackrel{\mathcal{R}}{\sim} B\)\(B \stackrel{\mathcal{R}}{\sim} C\),则 \(A \stackrel{\mathcal{R}}{\sim} C\)

    所以矩阵的相抵关系是一种等价关系。

  • 相抵等价类

    若将 \(\mathbb P^{m\times n}\) 中秩相同的矩阵归为一类,可得 \(\min\{m\,n\}+1\) 个相抵等价类,与 \(\begin{pmatrix} E_r & O \\ O & O \end{pmatrix}\)\(A\) 同类的矩阵的秩均为 \(r\,(r\in 0,1,\cdots,\min\{m,n\})\)

  • 相抵(等价)标准形

    \(r(A) = r\) 时,称 \(\begin{pmatrix} E_r & O \\ O & O \end{pmatrix}\)\(A\) 的相抵(等价)标准形。

题目

行列式的计算

  • 化零降阶法

    • 将某一行(列)化成仅剩一个非 \(0\),再行列式展开化成低一阶。

    • 一点两斜形,按”一点“所在的行(列)展开,得到两个三角形

      例题

      计算 \(D=\begin{vmatrix} a_1 & b_1 & & & \\ & a_2 & b_2 & & \\ & & \ddots & \ddots & \\ & & & a_{n - 1} & b_{n - 1} \\ b_n & 0 & \cdots & 0 & a_n \end{vmatrix}\)

      解答
      \[\begin{aligned}D &= a_1\cdot (-1)^{1+1}\begin{vmatrix} a_2 & b_2 & & \\ & \ddots & \ddots & \\ & & a_{n - 1} & b_{n - 1} \\ & & & a_n \end{vmatrix} + b_n\cdot (-1)^{n + 1}\begin{vmatrix} b_1 & & & \\ a_1 & b_2 & & \\ & \ddots & \ddots & \\ & & a_{n - 1} & b_{n - 1} \end{vmatrix}\\ &= a_1a_2\cdots a_n + (-1)^{n + 1}b_1b_2\cdots b_n\end{aligned}\]
  • 化基本形法

    • 行(列)和相等:通过变换让一行(列)元素相同,方便提取公因子,进而造 \(0\)

    • 循环行列式:虽然也是行(列)和相等,但别的做法。

      例题

      计算 \(D=\begin{vmatrix} 1 & 2 & 3 & \cdots & n - 1 & n \\ n & 1 & 2 & \cdots & n - 2 & n - 1 \\ n - 1 & n & 1 & \cdots & n - 3 & n - 2 \\ \vdots & \vdots & \vdots & \ddots & \vdots & \vdots \\ 3 & 4 & 5 & \cdots & 1 & 2 \\ 2 & 3 & 4 & \cdots & n & 1 \end{vmatrix}\)

      解答

      因为同一列是等差的,所以 \(R_i-R_{i+1}\) 可以把很多元素干成一样的。然后稍加变换就能得到爪形行列式。

      \[\begin{aligned}\begin{vmatrix} 1 & 2 & 3 & \cdots & n - 1 & n \\ n & 1 & 2 & \cdots & n - 2 & n - 1 \\ n - 1 & n & 1 & \cdots & n - 3 & n - 2 \\ \vdots & \vdots & \vdots & & \vdots & \vdots \\ 3 & 4 & 5 & \cdots & 1 & 2 \\ 2 & 3 & 4 & \cdots & n & 1 \end{vmatrix} &\xlongequal{R_i-R_{i+1}} \begin{vmatrix} 1 - n & 1 & 1 & \cdots & 1 & 1 \\ 1 & 1 - n & 1 & \cdots & 1 & 1 \\ 1 & 1 & 1 - n & \cdots & 1 & 1 \\ \vdots & \vdots & \vdots & & \vdots & \vdots \\ 1 & 1 & 1 & \cdots & 1 - n & 1 \\ 2 & 3 & 4 & \cdots & n & 1 \end{vmatrix} \\ &\xlongequal{C_i-C_n} \begin{vmatrix} -n & 0 & 0 & \cdots & 0 & 1 \\ 0 & -n & 0 & \cdots & 0 & 1 \\ 0 & 0 & -n & \cdots & 0 & 1 \\ \vdots & \vdots & \vdots & & \vdots & \vdots \\ 0 & 0 & 0 & \cdots & -n & 1 \\ 1 & 2 & 3 & \cdots & n - 1 & 1 \end{vmatrix} \\&= \begin{vmatrix} -n & 0 & 0 & \cdots & 0 & 0 \\ 0 & -n & 0 & \cdots & 0 & 0 \\ 0 & 0 & -n & \cdots & 0 & 0 \\ \vdots & \vdots & \vdots & & \vdots & \vdots \\ 0 & 0 & 0 & \cdots & -n & 0 \\ 1 & 2 & 3 & \cdots & n - 1 & 1 + \frac{1}{n} + \frac{2}{n} + \cdots + \frac{n - 1}{n} \end{vmatrix} \\& = (-n)^{n - 1}[1 + \frac{1}{2}(n - 1)] \\&= \frac{1}{2}(n + 1)(-n)^{n - 1}\end{aligned}\]
    • 爪形(法 1):斜爪消平爪

      \(\begin{vmatrix}a_{11}&a_{12}&\cdots&a_{1n}\\a_{21}&a_{22}& & \\\vdots&&\ddots&\\a_{n1}& & &a_{nn}\end{vmatrix}=(a_{11}-\sum_{i=2}^na_{i1}\frac{a_{1i}}{a_{ii}})\prod_{i=2}^na_{ii}\)

      可以当做一个模板解其他型行列式

      例题

      计算 \(D=\begin{vmatrix} x_1 - a_1 & x_2 & x_3 & \cdots & x_n \\ x_1 & x_2 - a_2 & x_3 & \cdots & x_n \\ x_1 & x_2 & x_3 - a_3 & \cdots & x_n \\ \vdots & \vdots & \vdots & \ddots & \vdots \\ x_1 & x_2 & x_3 & \cdots & x_n - a_n \end{vmatrix}\)

      解答

      \(R_{i+1}-R_1\) 可以得到爪形行列式,\(D=(-1)^n\prod_{i=1}^n a_i(\sum_{i=1}^n\dfrac{x_i}{a_i}-1)\)

      计算 \(D=\begin{vmatrix} a_1b_1 & a_1b_2 & a_1b_3 & \cdots & a_1b_n \\ a_2b_1 & a_2b_2 & a_2b_3 & \cdots & a_2b_n \\ a_3b_1 & a_3b_2 & a_3b_3 & \cdots & a_3b_n \\ \vdots & \vdots & \vdots & \ddots & \vdots \\ a_nb_1 & a_nb_2 & a_nb_3 & \cdots & a_nb_n \end{vmatrix}\)

      解答

      \(R_1\) 提出公因子 \(a_1\),再 \(R_i-a_iR_1\),得到爪形行列式。\(D=a_1b_n\prod_{i=1}^{n-1}(a_{i+1}b_i-a_ib_{i+1})\)

    • 爪形(法 2):按平爪展开,得到 \(n\) 个能直接计算的行列式

    • 异爪形(法 1,特殊可用):通过逐行(列)相消,用主对角线消次对角线

      例题

      计算 \(D_{n+1} = \begin{vmatrix}-a_1 & a_1 & 0 & \cdots & 0 & 0 \\0 & -a_2 & a_2 & \cdots & 0 & 0 \\0 & 0 & -a_3 & \cdots & 0 & 0 \\\vdots & \vdots & \vdots & & \vdots & \vdots \\0 & 0 & 0 & \cdots & -a_n & a_n \\1 & 1 & 1 & \cdots & 1 & 1\end{vmatrix}\)

      解答
      \[\xlongequal{C_{j+1}+C_j} \begin{vmatrix} -a_1 & 0 & 0 & \cdots & 0 & 0 \\ 0 & -a_2 & 0 & \cdots & 0 & 0 \\ 0 & 0 & -a_3 & \cdots & 0 & 0 \\ \vdots & \vdots & \vdots & & \vdots & \vdots \\ 0 & 0 & 0 & \cdots & -a_n & 0 \\ 1 & 2 & 3 & \cdots & n & n+1 \end{vmatrix} = (-1)^n (n+1) a_1 a_2 \cdots a_n\]
    • 异爪形(法 2,通用):把平爪展开。

      \(D_n = \begin{vmatrix}a_n & c_n & 0 & \cdots & 0 & 0 & 0 \\0 & a_{n-1} & c_{n-1} & \cdots & 0 & 0 & 0 \\0 & 0 & a_{n-2} & \cdots & 0 & 0 & 0 \\\vdots & \vdots & \vdots & \ddots & \vdots & \vdots & \vdots \\0 & 0 & 0 & \cdots & a_3 & c_3 & 0 \\0 & 0 & 0 & \cdots & 0 & a_2 & c_2 \\b_n & b_{n-1} & b_{n-2} & \cdots & b_3 & b_2 & a_1\end{vmatrix}\\=\sum_{k=1}^n (-1)^{k-1}b_k\cdot (\prod_{i=k+1}^n a_i)\cdot (\prod_{i=2}^k c_i)\)

      证明

      按最后一行展开,\(D_n = \sum_{k=1}^{n} (-1)^{n+k} b_{n-k+1} M_{nk}\)

      • \(k = 1\) 时,\(M_{n1}=b_n \cdot c_2 c_3 \cdots c_n\)

      • \(k = n\) 时,\(M_{nn}=a_1 \cdot a_2 a_3 \cdots a_n = a_1 a_2 \cdots a_n\)

      • \(1<k<n\) 时,\(M_{nk}\) 可写为分块行列式 \(M_{nk} = \begin{vmatrix}A_{k-1} & 0 \\0 & B_{n-k}\end{vmatrix}\),其中 \(A_{k-1} = \begin{pmatrix}a_n & c_n & 0 & \cdots & 0 \\0 & a_{n-1} & c_{n-1} & \cdots & 0 \\0 & 0 & a_{n-2} & \cdots & 0 \\\vdots & \vdots & \vdots & \ddots & \vdots \\0 & 0 & \cdots & 0 & a_{n-k+2}\end{pmatrix}\)\(B_{n-k} = \begin{pmatrix}c_{n-k+1} & 0 & \cdots & 0 \\a_{n-k} & c_{n-k} & \cdots & 0 \\\vdots & \vdots & \ddots & \vdots \\0 & \cdots & a_2 & c_2\end{pmatrix}\)

        \(M_{nk} =\prod_{i=n-k+2}^{n} a_i\times \prod_{i=2}^{n-k+1} c_i\)

      \(D_n=\sum_{k=1}^n (-1)^{k-1}b_k\cdot (\prod_{i=k+1}^n a_i)\cdot (\prod_{i=2}^k c_i)\)

    • 三对角形:通过逐行(列)相消,把一条斜线消 \(0\)

    • 拉普拉斯展开式

      • X 形:把的 \(2n+1-i\) 行移动到第 \(2i\) 行,第 \(2n+1-i\) 列移动到第 \(2i\) 列(\(i\in[1,n]\)

        \(\begin{aligned}\begin{vmatrix}a_1& & & & &b_1\\&\ddots & & &⋰& \\& &a_n&b_n& & \\& &b_{n+1}&a_{n+1}& & \\& ⋰ & & &\ddots& \\b_{2n}& & & & &a_{2n}\end{vmatrix}_{2n}&=\begin{vmatrix} a_1 & b_1 & & & & \\ b_{2n} & a_{2n} & & & & \\ & & \ddots &\ddots & & \\ & & \ddots& \ddots & & \\ & & & & a_n & b_n \\ & & & & b_{n + 1} & a_{n + 1} \end{vmatrix}_{2n}\\ &=\prod_{i=1}^n\begin{vmatrix}a_i&b_i\\b_{2n-i+1}&a_{2n-i+1}\end{vmatrix}\end{aligned}\)

  • 递推法

    • 爪形(法 3):按只有两个元素的行(列)展开,变成 \(n-1\) 形爪形和另一个行列式,另一个行列式再按只有单个元素的行(列)展开变成三角形

    • 异爪形(法 3):按开口或爪的根部展开,变成 \(n-1\) 阶异爪形和一个 \(n-1\) 阶三角形

      \(D_{n} = \begin{vmatrix} * & * & & & \\ & * & * & & \\ & & \ddots & \ddots & \\ & & & * & * \\ * & * & \cdots & * & * \end{vmatrix}_{n}\) 可以按第 \(1\) 列或第 \(n\) 列展开。

      例题

      计算 \(D_n = \begin{vmatrix}a & -1 \\& a & -1 \\& & a & \cdots \\& & & \cdots & -1 \\a_n & a_{n - 1} & a_{n - 2} & \cdots & a + a_1\end{vmatrix}\)

      解答

      按第一列展开:

      \[\begin{aligned} D_n&= a\cdot (-1)^{1+1}\begin{vmatrix} a & -1 \\ & a & -1 \\ & & a & \cdots \\ & & & \cdots & -1 \\ a_{n - 1} & a_{n - 2} & a_{n - 3} & \cdots & a + a_1 \end{vmatrix}_{n - 1} + a_n\cdot (-1)^{n + 1}\begin{vmatrix} -1 \\ a & -1 \\ & a & -1 \\ & & \cdots & -1 \\ & & & a & -1 \end{vmatrix}_{n - 1} \\&= aD_{n - 1} + (-1)^{n + 1}a_n(-1)^{n - 1}\\&= aD_{n - 1} + a_n \\ D_n &= aD_{n - 1} + a_n \\ &= a(aD_{n - 2} + a_{n - 1}) + a_n = a^2D_{n - 2} + a_n + a \cdot a_{n - 1} \\ &= a^2(aD_{n - 3} + a_{n - 2}) + a_n + a_{n - 1} = a^3D_{n - 3} + a_n + a \cdot a_{n - 1} + a^2 \cdot a_{n - 2} \\ &= \cdots \\ &= a^{n - 1}D_1 + a_n + a \cdot a_{n - 1} + a^2 \cdot a_{n - 2} + \cdots + a^{n - 2}a_2 \\ &= a^n + \sum_{i = 1}^{n} a^{n - i}a_i \end{aligned}\]

      计算 \(D_{n} = \begin{vmatrix} a_{1} & -a_{1} & 0 & \cdots & 0 & 0 \\ 0 & a_{2} & -a_{2} & \cdots & 0 & 0 \\ \vdots & \vdots & \vdots & & \vdots & \vdots \\ 0 & 0 & 0 & \cdots & a_{n} & -a_{n} \\ b & b & b & \cdots & b & b \end{vmatrix}\)

      解答

      按第 \(n\) 列展开:

      \[\begin{aligned}D_{n} &=(-a_n)\cdot (-1)^{(n-1)+n}D_{n-1}+b\cdot (-1)^{n+n}\begin{vmatrix} a_{1} & -a_{1} \\ & a_{2} & -a_{2} \\ & & \ddots & \ddots \\ & & & a_{n - 1} & -a_{n - 1} \\ & & & & a_{n} \end{vmatrix}\\&= a_{n}D_{n - 1}+b\prod_{i = 1}^{n}a_{i}\end{aligned}\]
      • \(\prod_{i = 1}^{n}a_{i} \neq 0\) 时,\(\dfrac{D_{n}}{\prod_{i = 1}^{n}a_{i}} = \dfrac{D_{n - 1}}{\prod_{i = 1}^{n - 1}a_{i}} + b=\cdots=\dfrac{D_1}{a_1}+(n-1)b\),又 \(D_1=2a_1b\),则 \(D_n= (n + 1)b\prod_{i = 1}^{n}a_{i}\)

      • \(\prod_{i = 1}^{n}a_{i} = 0\) 时,则 \(D_{n}\) 中有一行的元素全为 \(0\)\(D_{n} = 0\),也符合上式。

      \(D_{n} = (n + 1)b\prod_{i = 1}^{n}a_{i}\)

    • 三对角形(高阶):按第一行或最后一行展开,得到一个 \(n-1\) 阶三对角形和一个 \(n-2\) 阶三对角形“加边后”的产物。

      例题

      计算 \(D_n = \begin{vmatrix}a & c \\b & a & c \\& b & \ddots & \ddots \\& & \ddots & a & c \\& & & b & a\end{vmatrix}_n\)

      解答
      \[ \begin{aligned} D_n &= a\cdot (-1)^{1+1}\begin{vmatrix} a & c \\ b & a & c \\ & b & \ddots & \ddots \\ & & \ddots & a & c \\ & & & b & a \end{vmatrix}_{n - 1} - c\cdot (-1)^{1+2}\begin{vmatrix} b & c \\ & a & c \\ & b & \ddots & \ddots \\ & & \ddots & a & c \\ & & & b & a \end{vmatrix}_{n - 1} \\ &= a\cdot D_{n - 1} - c\cdot bD_{n - 2}\end{aligned} \]

      可以特征方程法求,也可考虑归纳证明。

      Tip

      递推公式 \(D_n = aD_{n-1} + bD_{n-2}\) 的求解

      特征方程为 \(x^2=ax+b\),解得两个根 \(x_1,x_2\) 称为特征根。若 \(x_1\neq x_2\),则通项为 \(D_n=Ax_1^n+Bx_2^n\);否则,\(D_n=(A\cdot n+B)x_1^n\)。其中 \(A,B\) 可以通过 \(D_1,D_2\) 解出。

      证明
      • \(b = 0\) 时,\(D_n = aD_{n-1} = a^2D_{n-2} = \cdots = a^{n-1}D_1\)

      • \(b \neq 0\) 时,\(D_n - aD_{n-1} - bD_{n-2} = 0\)

        考虑特征方程 \(x^2 - ax - b = 0\)\(\Delta = a^2 + 4b\)

        1. \(\Delta \neq 0\)

          方程有两个不相等的根 \(u_1,u_2\)\(u_1 + u_2 = a\)\(u_1u_2 = -b\),则:

          \(D_n - (u_1 + u_2)D_{n-1} + u_1u_2D_{n-2} = 0\)

          \(D_n - u_1D_{n-1} = u_2(D_{n-1} - u_1D_{n-2})=\cdots=u_2^{n-2}(D_2 - u_1D_1)\)

          同理 \(D_n - u_2D_{n-1} = u_1(D_{n-1} - u_2D_{n-2})=\cdots= u_1^{n-2}(D_2 - u_2D_1)\)

          消去 \(D_{n-1}\),解得:\(D_n = \dfrac{u_1^{n-1}(D_2 - u_2D_1) - u_2^{n-1}(D_2 - u_1D_1)}{u_1 - u_2}\)

        2. \(\Delta = 0\)

          方程有两个相等的根 \(u_1 = u_2 = u\)\(u = \frac{a}{2}\)\(u^2 = -b\),则:

          \(D_n - 2uD_{n-1} + u^2D_{n-2} = 0\)

          \(D_n - uD_{n-1} = u(D_{n-1} - uD_{n-2}) = \cdots=u^{n-2}(D_2 - uD_1)\)

          \(D_n = uD_{n-1} + u^{n-2}(D_2 - uD_1)\\= u(uD_{n-2} + u^{n-3}(D_2 - uD_1)) + u^{n-2}(D_2 - uD_1) = u^2D_{n-2} + 2u^{n-2}(D_2 - uD_1) \\=\cdots=u^{n-2}D_2 + (n - 2)u^{n-2}(D_2 - uD_1)\\=u^{n-2}((n - 1)D_2 - (n - 2)uD_1)\)

    • 大对称形:把其中一个行(列)拆分成一个单元素行(列)和一个方便消其他位置的行(列)。

      例题

      计算 \(D_{n} = \begin{vmatrix}a + x_1 & a & \cdots & a \\a & a + x_2 & \cdots & a \\\vdots & \vdots & & \vdots \\a & a & \cdots & a + x_n\end{vmatrix}\)

      解答
      \[\begin{aligned}D_n &= \begin{vmatrix}a + x_1 & a & \cdots & a & a \\a & a + x_2 & \cdots & a & a \\\vdots & \vdots & & \vdots & \vdots \\a & a & \cdots & a + x_{n-1} & a \\ a & a & \cdots & a & a\end{vmatrix} + \begin{vmatrix}a + x_1 & a & \cdots & a & 0 \\a & a + x_2 & \cdots & a & 0 \\\vdots & \vdots & & \vdots & \vdots \\a & a & \cdots & a + x_{n-1} & 0 \\a & a & \cdots & a & x_n\end{vmatrix}\\&= \begin{vmatrix}x_1 & 0 & \cdots & 0 & a \\0 & x_2 & \cdots & 0 & a \\\vdots & \vdots & & \vdots & \vdots \\0 & 0 & \cdots & x_{n-1} & a \\0 & 0 & \cdots & 0 & a\end{vmatrix} + x_n D_{n-1}\\&= x_1 x_2 \cdots x_{n-1} a + x_n D_{n-1}\end{aligned}\]

      同理 \(D_{n-1} = x_1 x_2 \cdots x_{n-2} a + x_{n-1} D_{n-2}\)\(\cdots\)\(D_2 = x_1 a + x_2 D_1\),故 \(D_n = a \sum_{i=1}^{n} \prod_{j\neq i} x_j + \prod_{j=1}^{n} x_j\)

      计算 \(D_{n} = \begin{vmatrix} a_1 & b & \cdots & b & b \\ b & a_2 & \cdots & b & b \\ \vdots & \vdots & \ddots & \vdots & \vdots \\ b & b & \cdots & a_{n - 1} & b \\ b & b & \cdots & b & a_n \end{vmatrix}_{n}\)

      解答

      \(D_1 = a_1\)\(D_2 = a_1a_2 - b^2\)

      \(n \geq 3\) 时,

      \[\begin{aligned}D_{n} &=\begin{vmatrix} a_1 & b & \cdots & b & b \\ b & a_2 & \cdots & b & b \\ \vdots & \vdots & \ddots & \vdots & \vdots \\ b & b & \cdots & a_{n - 1} & b \\ b & b & \cdots & b & b \end{vmatrix}_{n}+\begin{vmatrix} a_1 & b & \cdots & b \\ b & a_2 & \cdots & b \\ \vdots & \vdots & \ddots & \vdots \\ b & b & \cdots & a_{n - 1} \\ b & b & \cdots &b& a_n - b \end{vmatrix}_{n}\\&=\begin{vmatrix} a_1 - b & & & &b \\ & a_2 - b & & &b \\ & & \ddots & &\vdots \\ & & & a_{n - 1} - b & b \\ & & & & b \end{vmatrix}_{n}+(a_n - b)D_{n - 1} \\&=b\prod_{i = 1}^{n - 1}(a_i - b)+ (a_n - b)D_{n - 1}\end{aligned}\]

      \(n=2\) 时也符合上式。

      • \(\prod_{i = 1}^{n}(a_i - b) \neq 0\) 时:

        \[\begin{aligned}\dfrac{D_{n}}{\prod_{i = 1}^{n}(a_i - b)}& = \dfrac{D_{n - 1}}{\prod_{i = 1}^{n - 1}(a_i - b)} + \dfrac{b}{a_n - b}\\&=\cdots\\&=\dfrac{a_1}{a_1 - b} + \sum_{i = 2}^{n}\dfrac{b}{a_i - b}\\ &=1+\sum_{i=1}^n\dfrac{b}{a_i-b}\\\Rightarrow D_n&= (1 + \sum_{i = 1}^{n}\frac{b}{a_i - b} )\prod_{i = 1}^{n}(a_i - b)\\& = \prod_{i = 1}^{n}(a_i - b) + b\sum_{i = 1}^{n}\prod_{j\neq i}(a_j - b)\end{aligned}\]
      • \(\prod_{i = 1}^{n}(a_i - b) = 0\) 时,不妨设 \(a_k = b\),则

        \[\begin{aligned}D_{n} &= \begin{vmatrix} a_1 & b & \cdots & b & \cdots & b \\ b & a_2 & \cdots & b & \cdots & b \\ \vdots & \vdots & \ddots & \vdots & & \vdots \\ b & b & \cdots & a_k & \cdots & b \\ \vdots & \vdots & & \vdots & \ddots & \vdots \\ b & b & \cdots & b & \cdots & a_n \end{vmatrix}_{n} \\&= \begin{vmatrix} a_1 - b & & & b \\ & a_2 - b & & b \\ & & \ddots & \vdots \\ & & & b \\ & & & \vdots & \ddots \\ & & & b & a_n - b \end{vmatrix}_{n} \\&\xlongequal{按R_k展开} b\cdot (-1)^{k+k}\prod_{i\neq k}(a_i - b)\end{aligned}\]

        也符合上式。

      \(D_{n} = \prod_{i = 1}^{n}(a_i - b) + b\sum_{i = 1}^{n}\prod_{j\neq i}(a_j - b)\)

      计算 \(D_{n} = \begin{vmatrix}x & b & b & \cdots & b \\a & x & b & \cdots & b \\a & a & x & \cdots & b \\\vdots & \vdots & \vdots & & \vdots \\a & a & a & \cdots & x\end{vmatrix}\),其中 \(a\neq b\)\(a=b\) 是行/列和相等型)

      解答
      \[\begin{aligned} D_{n}&=\begin{vmatrix} x & b & b & \cdots & b \\ a & x & b & \cdots & b \\ a & a & x & \cdots & b \\ \vdots & \vdots & \vdots & & \vdots \\ a & a & a & \cdots & b \end{vmatrix}_{n}+\begin{vmatrix} x & b & b & \cdots & 0 \\ a & x & b & \cdots & 0 \\ a & a & x & \cdots & 0 \\ \vdots & \vdots & \vdots & & \vdots \\ a & a & a & \cdots & x - b \end{vmatrix}_{n}\\&=b\begin{vmatrix} x & b & b & \cdots & 1 \\ a & x & b & \cdots & 1 \\ a & a & x & \cdots & 1 \\ \vdots & \vdots & \vdots & & \vdots \\ a & a & a & \cdots & 1 \end{vmatrix}_{n}+(x - b)D_{n - 1}\\&=b\begin{vmatrix} x - a & b - a & b - a & \cdots & 1 \\ 0 & x - a & b - a & \cdots & 1 \\ 0 & 0 & x - a & \cdots & 1 \\ \vdots & \vdots & \vdots & & \vdots \\ 0 & 0 & 0 & \cdots & 1 \end{vmatrix}_{n}+(x - b)D_{n - 1}\\&=b(x - a)^{n - 1}+(x - b)D_{n - 1}\end{aligned}\]
      • 法 1:

        \(D_{n} - \dfrac{b}{b - a}(x - a)^{n} = (x - b)[D_{n - 1} - \dfrac{b}{b - a}(x - a)^{n - 1}]\)\(= \cdots=(x - b)^{n-1}[x-\dfrac{b}{b-a}(x-a)]=-\dfrac{a}{b-a}(x-b)^n\),即 \(D_{n} = \dfrac{b(x - a)^{n} - a(x - b)^{n}}{b - a}\)

        Tip

        待定系数法

        \(a_n=pa_{n-1}+f(n)\)

        \(f(n)=C\):设 \(a_n+A=p(a_{n-1}+A)\)

        \(f(n)=qr^{n-1}\):设 \(a_n+Ar^n=p(a_{n-1}+Ar^{n-1})\)\(a_n=pa_{n-1}+pAr^{n-1}-Ar^n\),则 \(A=\dfrac{qr^{n-1}}{pr^{n-1}-r^n}=\dfrac{q}{p-r}\)

        \(f(n)=xn+y\):设 \(a_n+An+B=p(a_{n-1}+A(n-1)+B)\)

        \(f(n)=xn^2+yn+z\):设 \(a_n+An^2+Bn+C=p(a_{n-1}+A(n-1)^2+B(n-1)+C)\)

        \(f(n)\) 是多项式:设成同次数多项式

      • 法 2:

        同理可得 \(D_{n}^{\text{T}}=a(x - b)^{n - 1}+(x - a)D_{n - 1}^{\text{T}}\)

        \(D_{n}=D_{n}^{\text{T}},\ D_{n - 1}=D_{n - 1}^{\text{T}}\),得到 \(\dfrac{D_{n}-b(x - a)^{n - 1}}{x - b}=\dfrac{D_{n}-a(x - b)^{n - 1}}{x - a}\),解得 \(D_{n}=\dfrac{b(x - a)^{n}-a(x - b)^{n}}{b - a}\)

      计算 \(D_{n} = \begin{vmatrix}x_1 & b & b & \cdots & b \\a & x_2 & b & \cdots & b \\a & a & x_3 & \cdots & b \\\vdots & \vdots & \vdots & & \vdots \\a & a & a & \cdots & x_n\end{vmatrix}\)

      解答

      \(D_n=b\prod_{i=1}^{n-1}(x_i - a)+(x_n - b)D_{n - 1}\)

      同理(转置)有 \(D_n=a\prod_{i=1}^{n-1}(x_i-b)+(x_n-a)D_{n-1}\)

      \(a\neq b\),解得 \(D_n=\dfrac 1 {a-b}[a\prod_{i=1}^n (x_i-b)-b\prod_{i=1}^n (x_i-a)]\)

      \(a=b\),由递推可得 \(D_n=\prod_{i=1}^n (x_i-b)+b\sum_{i=1}^n \prod_{j\neq i}(x_j-b)\)

      计算 \(D_n = \begin{vmatrix} a_1 + x_1^2 & x_2x_1 & \cdots & x_{n-1}x_1 & x_nx_1 \\ x_1x_2 & a_2 + x_2^2 & \cdots & x_{n-1}x_2 & x_nx_2 \\ \vdots & \vdots & \ddots & \vdots & \vdots \\ x_1x_{n-1} & x_2x_{n-1} & \cdots & a_{n-1} + x_{n-1}^2 & x_nx_{n-1} \\ x_1x_n & x_2x_n & \cdots & x_{n-1}x_n & a_n + x_n^2 \end{vmatrix}_n\)

      解答
      \[\begin{aligned}D_n &= \begin{vmatrix} a_1 + x_1^2 & x_2x_1 & \cdots & x_{n-1}x_1 \\ x_1x_2 & a_2 + x_2^2 & \cdots & x_{n-1}x_2 \\ \vdots & \vdots & \ddots & \vdots \\ x_1x_{n-1} & x_2x_{n-1} & \cdots & a_{n-1} + x_{n-1}^2 \\ x_1x_n & x_2x_n & \cdots & x_{n-1}x_n & a_n \end{vmatrix} + x_n \begin{vmatrix} a_1 + x_1^2 & x_2x_1 & \cdots & x_{n-1}x_1 & x_1 \\ x_1x_2 & 1 + x_2^2 & \cdots & x_{n-1}x_2 & x_2 \\ \vdots & \vdots & \ddots & \vdots & \vdots \\ x_1x_{n-1} & x_2x_{n-1} & \cdots & a_{n-1} + x_{n-1}^2 & x_{n-1} \\ x_1x_n & x_2x_n & \cdots & x_{n-1}x_n & x_n \end{vmatrix}\\&=a_nD_{n-1}+x_n \begin{vmatrix} a_1 & &\cdots & &x_1 \\ & a_2 & \cdots & &x_2 \\ & & \ddots & &\vdots \\ & & \cdots & a_{n-1} & x_{n-1} \\ & & \cdots & & x_n \end{vmatrix}_n\\&=a_nD_{n-1}+x_n^2\prod_{i=1}^{n-1} a_{i}\end{aligned}\]

      \(\dfrac{D_n}{\prod_{i=1}^n a_i}=\dfrac{D_{n-1}}{\prod_{i=1}^{n-1}a_i}+\dfrac{x_n^2}{a_n}\)

      \(D_n=\prod_{i=1}^n a_i(1+\sum_{i=1}^n \dfrac{x_i^2}{a_i})\)

    • Cauchy 行列式

      例题

      计算:\(D_n= \begin{vmatrix}(a_1 + b_1)^{-1} & (a_1 + b_2)^{-1} & \cdots & (a_1 + b_n)^{-1} \\(a_2 + b_1)^{-1} & (a_2 + b_2)^{-1} & \cdots & (a_2 + b_n)^{-1} \\\vdots & \vdots & & \vdots \\(a_n + b_1)^{-1} & (a_n + b_2)^{-1} & \cdots & (a_n + b_n)^{-1}\end{vmatrix}\)

      解答
      \[ \begin{aligned} D_n &\xlongequal{C_i-C_n} \begin{vmatrix} \frac{b_n - b_1}{(a_1 + b_1)(a_1 + b_n)} & \cdots & \frac{b_n - b_{n-1}}{(a_1 + b_{n-1})(a_1 + b_n)} & \frac{1}{a_1 + b_n} \\ \vdots & & \vdots & \vdots \\ \frac{b_n - b_1}{(a_{n-1} + b_1)(a_{n-1} + b_n)} & \cdots & \frac{b_n - b_{n-1}}{(a_{n-1} + b_{n-1})(a_{n-1} + b_n)} & \frac{1}{a_{n-1} + b_n} \\ \frac{b_n - b_1}{(a_n + b_1)(a_n + b_n)} & \cdots & \frac{b_n - b_{n-1}}{(a_n + b_{n-1})(a_n + b_n)} & \frac{1}{a_n + b_n} \end{vmatrix} \\ &= \frac{\prod_{k=1}^{n-1} (b_n - b_k)}{\prod_{j=1}^{n} (a_j + b_n)} \begin{vmatrix} \frac{1}{a_1 + b_1} & \cdots & \frac{1}{a_1 + b_{n-1}} & 1 \\ \vdots & & \vdots & \vdots \\ \frac{1}{a_{n-1} + b_1} & \cdots & \frac{1}{a_{n-1} + b_{n-1}} & 1 \\ \frac{1}{a_n + b_1} & \cdots & \frac{1}{a_n + b_{n-1}} & 1 \end{vmatrix} \\ &\xlongequal{R_i-R_n} \frac{\prod_{k=1}^{n-1} (b_n - b_k)}{\prod_{j=1}^{n} (a_j + b_n)} \begin{vmatrix} \frac{a_n - a_1}{(a_1 + b_1)(a_n + b_1)} & \cdots & \frac{a_n - a_1}{(a_1 + b_{n-1})(a_n + b_{n-1})} & 0 \\ \vdots & & \vdots & \vdots \\ \frac{a_n - a_{n-1}}{(a_{n-1} + b_1)(a_n + b_1)} & \cdots & \frac{a_n - a_{n-1}}{(a_{n-1} + b_{n-1})(a_n + b_{n-1})} & 0 \\ \frac{1}{a_n + b_1} & \cdots & \frac{1}{a_n + b_{n-1}} & 1 \end{vmatrix} \\ &= \frac{\prod_{i=1}^{n-1} (a_n - a_i)(b_n - b_i)}{\prod_{j=1}^{n} (a_j + b_n) \prod_{k=1}^{n-1} (a_n + b_k)} \begin{vmatrix} \frac{1}{a_1 + b_1} & \cdots & \frac{1}{a_1 + b_{n-1}} \\ \vdots & & \vdots \\ \frac{1}{a_{n-1} + b_1} & \cdots & \frac{1}{a_{n-1} + b_{n-1}} \end{vmatrix} \\ &= \frac{\prod_{i=1}^{n-1} (a_n - a_i)(b_n - b_i)}{\prod_{j=1}^{n} (a_j + b_n) \prod_{k=1}^{n-1} (a_n + b_k)} \cdot D_{n-1}. \end{aligned} \]

      不断递推下去即得 \(D_n = \dfrac{\prod_{1 \leq i < j \leq n} (a_j - a_i)(b_j - b_i)}{\prod_{i,j=1}^{n} (a_i + b_j)}\)

  • 拆分法:很多相同元素时用,拆分出来消。大对称形就是一个例子。

    例题

    \(D_n = \begin{vmatrix}a_{11} & a_{12} & \cdots & a_{1n} \\a_{21} & a_{22} & \cdots & a_{2n} \\\cdots & \cdots & \cdots & \cdots \\a_{n1} & a_{n2} & \cdots & a_{nn}\end{vmatrix}\),证明:\(D_n'=\begin{vmatrix}a_{11} + x_1 & a_{12} + x_2 & \cdots & a_{1n} + x_n \\a_{21} + x_1 & a_{22} + x_2 & \cdots & a_{2n} + x_n \\\cdots & \cdots & \cdots & \cdots \\a_{n1} + x_1 & a_{n2} + x_2 & \cdots & a_{nn} + x_n\end{vmatrix} = D_n + \sum_{j=1}^{n} x_j \sum_{i=1}^{n} A_{ij}\),其中 \(A_{ij}\)\(D_n\) 中元素 \(a_{ij}\)\(D_n\) 中的代数余子式

    解答
    \[ \begin{aligned} D_n' &=\begin{vmatrix} a_{11} & a_{12} + x_2 & \cdots & a_{1n} + x_n \\ a_{21} & a_{22} + x_2 & \cdots & a_{2n} + x_n \\ \cdots & \cdots & \cdots & \cdots \\ a_{n1} & a_{n2} + x_2 & \cdots & a_{nn} + x_n \end{vmatrix} + x_1 \begin{vmatrix} 1 & a_{12} + x_2 & \cdots & a_{1n} + x_n \\ 1 & a_{22} + x_2 & \cdots & a_{2n} + x_n \\ \cdots & \cdots & \cdots & \cdots \\ 1 & a_{n2} + x_2 & \cdots & a_{nn} + x_n \end{vmatrix} \\ &= \begin{vmatrix} a_{11} & a_{12} + x_2 & \cdots & a_{1n} + x_n \\ a_{21} & a_{22} + x_2 & \cdots & a_{2n} + x_n \\ \cdots & \cdots & \cdots & \cdots \\ a_{n1} & a_{n2} + x_2 & \cdots & a_{nn} + x_n \end{vmatrix} + x_1 \begin{vmatrix} 1 & a_{12} & \cdots & a_{1n} \\ 1 & a_{22} & \cdots & a_{2n} \\ \cdots & \cdots & \cdots & \cdots \\ 1 & a_{n2} & \cdots & a_{nn} \end{vmatrix} \\ &= \begin{vmatrix} a_{11} & a_{12} + x_2 & \cdots & a_{1n} + x_n \\ a_{21} & a_{22} + x_2 & \cdots & a_{2n} + x_n \\ \cdots & \cdots & \cdots & \cdots \\ a_{n1} & a_{n2} + x_2 & \cdots & a_{nn} + x_n \end{vmatrix} + x_1 \sum_{i=1}^{n} A_{i1} \\ &=D_n+\sum_{j=1}^n x_j\sum_{i=1}^n A_{ij} \end{aligned} \]
  • 归纳法

    例题

    证明:\(D_n=\begin{vmatrix} 2\cos \alpha & 1 & 0 & \cdots & 0 & 0 \\ 1 & 2\cos \alpha & 1 & \cdots & 0 & 0 \\ 0 & 1 & 2\cos \alpha & \cdots & 0 & 0 \\ \vdots & \vdots & \vdots & & \vdots & \vdots \\ 0 & 0 & 0 & \cdots & 2\cos \alpha & 1 \\ 0 & 0 & 0 & \cdots & 1 & 2\cos \alpha \end{vmatrix}_n = \dfrac{\sin(n + 1)\alpha}{\sin \alpha} \quad (\sin \alpha \neq 0)\)

    解答
    • \(n = 1\) 时:\(D_1 = 2\cos \alpha=\dfrac{2\cos\alpha}{\sin\alpha}=\dfrac{\sin(1+1)\alpha}{\sin\alpha}\)

    • \(n = 2\) 时:\(D_2 = \begin{vmatrix} 2\cos \alpha & 1 \\ 1 & 2\cos \alpha \end{vmatrix}= 4\cos^2 \alpha - 1=2\cos^2\alpha+\cos 2\alpha\)\(=\dfrac{2\cos^2 \alpha \sin \alpha + \cos 2\alpha \sin \alpha}{\sin \alpha} = \dfrac{\sin 2\alpha \cos \alpha + \cos 2\alpha \sin \alpha}{\sin \alpha}=\dfrac{\sin(2 + 1)\alpha}{\sin \alpha}\)

    • 假设当 \(n \leq k\) 时结论成立,\(D_{k-1} = \dfrac{\sin((k - 1) + 1)\alpha}{\sin \alpha} = \dfrac{\sin k \alpha}{\sin \alpha}\)\(D_k = \dfrac{\sin(k + 1)\alpha}{\sin \alpha}\)

      \[ \begin{aligned} D_{k+1} &= \begin{vmatrix} 2\cos \alpha & 1 & 0 & \cdots & 0 & 0 \\ 1 & 2\cos \alpha & 1 & \cdots & 0 & 0 \\ 0 & 1 & 2\cos \alpha & \cdots & 0 & 0 \\ \vdots & \vdots & \vdots & & \vdots & \vdots \\ 0 & 0 & 0 & \cdots & 2\cos \alpha & 1 \\ 0 & 0 & 0 & \cdots & 1 & 2\cos \alpha \end{vmatrix}_{k+1} \\ & = 2\cos \alpha \cdot D_k - \begin{vmatrix} 1 & 1 & 0 & \cdots & 0 & 0 \\ 0 & 2\cos \alpha & 1 & \cdots & 0 & 0 \\ 0 & 1 & 2\cos \alpha & \cdots & 0 & 0 \\ \vdots & \vdots & \vdots & & \vdots & \vdots \\ 0 & 0 & 0 & \cdots & 2\cos \alpha & 1 \\ 0 & 0 & 0 & \cdots & 1 & 2\cos\alpha \end{vmatrix}_k\\ &=2\cos\alpha \cdot \dfrac{\sin (k+1)\alpha}{\sin \alpha}-\dfrac{\sin k\alpha}{\sin \alpha}\\ &= \frac{2\cos \alpha \sin(k + 1)\alpha - [ \sin(k + 1)\alpha \cos \alpha - \cos(k + 1)\alpha \sin \alpha]}{\sin \alpha} \\ &= \frac{\cos \alpha \sin(k + 1)\alpha + \cos(k + 1)\alpha \sin \alpha}{\sin \alpha}\\ &= \frac{\sin(k + 2)\alpha}{\sin \alpha} \\ &= \frac{\sin((k + 1) + 1)\alpha}{\sin \alpha} \end{aligned} \]

      \(n=k+1\) 时也成立。

    根据归纳法知,对任意 \(n\) 都成立。

  • 范德蒙德行列式:注意正负交替可能是负数的幂次。

    例题

    计算 \(D=\begin{vmatrix}a&b&c\\a^2&b^2&c^2\\b+c&c+a&a+b\end{vmatrix}\)

    解答
    \[\begin{aligned}D&\xlongequal{R_3+R_1}(a+b+c)\begin{vmatrix}a&b&c\\a^2&b^2&c^2\\1&1&1\end{vmatrix}\\&\xlongequal{R_{32},R_{21}}(a+b+c)\cdot (-1)^2\begin{vmatrix}1&1&1\\a&b&c\\a^2&b^2&c^2\end{vmatrix}\\&=(a+b+c)(b-a)(c-a)(c-b)\end{aligned}\]

    计算 \(D=\begin{vmatrix}1&1&2&1\\1&2&2&4\\1&4&10&16\\1&8&26&64\end{vmatrix}\)

    解答
    \[\begin{aligned}D&=\begin{vmatrix}1&1&1&1\\1&2&3&4\\1&4&9&16\\1&8&27&64\end{vmatrix}+\begin{vmatrix}1&1&1&1\\1&2&-1&4\\1&4&1&16\\1&8&-1&64\end{vmatrix}\\&=\begin{vmatrix}1&1&1&1\\1&2&3&4\\1&4&9&16\\1&8&27&64\end{vmatrix}+\begin{vmatrix}1&1&1&1\\1&2&-1&4\\1&4&(-1)^2&16\\1&8&(-1)^3&64\end{vmatrix}\\&=3!2!1!+1\cdot (-2)\cdot 3\cdot (-3)\cdot 2\cdot 5\\&=192\end{aligned}\]

    计算 \(D=\begin{vmatrix}a & a^2 & bc \\b & b^2 & ac \\c & c^2 & ab\end{vmatrix}\)

    解答
    \[ \begin{aligned} D&\xlongequal{C_3+(a+b+c)C_1} \begin{vmatrix} a & a^2 & a^2 + ab + bc + ac \\ b & b^2 & b^2 + ab + bc + ac \\ c & c^2 & c^2 + ab + bc + ac \end{vmatrix} \\ &=\xlongequal{C_3-C_2} \begin{vmatrix} a & a^2 & ab + bc + ac \\ b & b^2 & ab + bc + ac \\ c & c^2 & ab + bc + ac \end{vmatrix} \\ &=(ab + bc + ac) \begin{vmatrix} a & a^2 & 1 \\ b & b^2 & 1 \\ c & c^2 & 1 \end{vmatrix}\\ &\xlongequal{C_{23},C_{12}} (-1)^2 (ab + bc + ac) \begin{vmatrix} 1 & a & a^2 \\ 1 & b & b^2 \\ 1 & c & c^2 \end{vmatrix}\\ &=(ab + bc + ac)(b - a)(c - a)(c - b) \end{aligned} \]

    计算 \(D_n = \begin{vmatrix} 1 & 1 & \cdots & 1 \\ x_1 & x_2 & \cdots & x_n \\ \vdots & \vdots & & \vdots \\ x_1^{n-2} & x_2^{n-2} & \cdots & x_n^{n-2} \\ x_1^n & x_2^n & \cdots & x_n^n \end{vmatrix}\,(n\geq 2)\)

    解答

    构造 \(n+1\) 阶范德蒙德行列式 \(D'_n = \begin{vmatrix} 1 & 1 & \cdots & 1 & 1 & 1 \\ x_1 & x_2 & \cdots & x_{n-1} & x_n & y \\ \vdots & \vdots & & \vdots & \vdots & \vdots \\ x_1^{n-2} & x_2^{n-2} & \cdots & x_{n-1}^{n-2} & x_n^{n-2} & y^{n-2} \\ x_1^{n-1} & x_2^{n-1} & \cdots & x_{n-1}^{n-1} & x_n^{n-1} & y^{n-1} \\ x_1^n & x_2^n & \cdots & x_{n-1}^n & x_n^n & y^n \end{vmatrix}\)

    则原行列式 \(D_n\) 就是 \(D'_n\) 中元素 \(y^{n-1}\) 的余子式 \(M_{n-1,n}\)

    \(D_n'\) 按第 \(n\) 列展开,\(D_n'=\sum_{i=1}^n y^i(-1)^{i+n}M_{i,n}\),把 \(D_n'\) 看成关于 \(y\) 的多项式 \(D_n'(y)\),那么 \(M_{n-1,n}=\dfrac{[y^{n-1}]D_n'(y)}{(-1)^{(n-1)+n}}\)

    \(D_n'(y) = \prod_{1 \leq i < j \leq n} (x_j - x_i) \cdot \prod_{i=1}^n (y - x_i)\),可得 \([y^{n-1}]D_n'(y)=\prod_{1 \leq i < j \leq n} (x_j - x_i) \cdot ( -\sum_{i=1}^n x_i)\)

    因此 \(D_n=\prod_{1 \leq i < j \leq n} (x_j - x_i) \cdot ( \sum_{i=1}^n x_i)\)

  • 加边法:用新加的行(列)同时消其他的行(列)。注意可以多次加边

    例题

    计算 \(D=\begin{vmatrix} b_1 + a_1 & a_1 & \cdots & a_1 \\ a_2 & b_2 + a_2 & \cdots & a_2 \\ \vdots & \vdots & & \vdots \\ a_n & a_n & \cdots & b_n + a_n \end{vmatrix}\)

    解答
    \[ \begin{aligned} D&=\begin{vmatrix} 1 & 0 & 0 & \cdots & 0 \\ a_1 & b_1 + a_1 & a_1 & \cdots & a_1 \\ a_2 & a_2 & b_2 + a_2 & \cdots & a_2 \\ \vdots & \vdots & \vdots & & \vdots \\ a_n & a_n & a_n & \cdots & b_n + a_n \end{vmatrix} \\ &=\begin{vmatrix} 1 & -1 & -1 & \cdots & -1 \\ a_1 & b_1 & & & \\ a_2 & & b_2 & & \\ \vdots & & & \ddots & \\ a_n & & & & b_n \end{vmatrix} \\ &=(1 + \sum_{i=1}^n\frac{a_i}{b_i} )b_1b_2\cdots b_n \end{aligned}\]

    计算 \(D=\begin{vmatrix}a_1+x_1^2&x_1x_2&\cdots&x_1x_n\\x_2x_1&a_2+x_2^2&\cdots&x_2x_n\\\vdots&\vdots&\ddots&\vdots\\x_nx_1&x_nx_2&\cdots&a_n+x_n^2\end{vmatrix}_n\)

    解答
    \[\begin{aligned}D&=\begin{vmatrix}1&x_1&x_2&\cdots&x_n\\0&a_1+x_1^2&x_1x_2&\cdots&x_1x_n\\0&x_2x_1&a_2+x_2^2&\cdots&x_2x_n\\0&\vdots&\vdots&\ddots&\vdots\\0&x_nx_1&x_nx_2&\cdots&a_n+x_n^2\end{vmatrix}_{n+1}\\ &\xlongequal{R_{i+1}-x_iR_1} \begin{vmatrix}1&x_1&x_2&\cdots&x_n\\-x_1&a_1\\-x_2&&a_2\\\vdots&&&\ddots\\-x_n&&&&a_n\end{vmatrix}_{n+1}\\ &=(1+\sum_{i=1}^n \dfrac{x_i^2}{a_i})\prod_{i=1}^n a_i \end{aligned} \]

    计算 \(D=\begin{vmatrix}1 + x_1 & 1 + x_1^2 & \cdots & 1 + x_1^n \\1 + x_2 & 1 + x_2^2 & \cdots & 1 + x_2^n \\\vdots & \vdots & \ddots &\vdots \\1 + x_n & 1 + x_n^2 & \cdots & 1 + x_n^n\end{vmatrix}_n\)

    解答
    \[ \begin{aligned} D&= \begin{vmatrix} 1 & 0 & 0 & \cdots & 0 \\ 1 & 1 + x_1 & 1 + x_1^2 & \cdots & 1 + x_1^n \\ 1 & 1 + x_2 & 1 + x_2^2 & \cdots & 1 + x_2^n \\ \vdots & \vdots & \vdots & & \vdots \\ 1 & 1 + x_n & 1 + x_n^2 & \cdots & 1 + x_n^n \end{vmatrix}_{n+1}\\ &= \begin{vmatrix} 1 & -1 & -1 & \cdots & -1 \\ 1 & x_1 & x_1^2 & \cdots & x_1^n \\ 1 & x_2 & x_2^2 & \cdots & x_2^n \\ \vdots & \vdots & \vdots & & \vdots \\ 1 & x_n & x_n^2 & \cdots & x_n^n \end{vmatrix}_{n+1}\\ &= \begin{vmatrix} 2 & 0 & 0 & \cdots & 0 \\ 1 & x_1 & x_1^2 & \cdots & x_1^n \\ 1 & x_2 & x_2^2 & \cdots & x_2^n \\ \vdots & \vdots & \vdots & & \vdots \\ 1 & x_n & x_n^2 & \cdots & x_n^n \end{vmatrix} + \begin{vmatrix} -1 & -1 & -1 & \cdots & -1 \\ 1 & x_1 & x_1^2 & \cdots & x_1^n \\ 1 & x_2 & x_2^2 & \cdots & x_2^n \\ \vdots & \vdots & \vdots & & \vdots \\ 1 & x_n & x_n^2 & \cdots & x_n^n \end{vmatrix}\\ &= 2\cdot (-1)^{1+1}\cdot x_1x_2\cdots x_n \begin{vmatrix} 1 & x_1 & \cdots & x_1^{n-1} \\ 2 & x_2 & \cdots & x_2^{n-1} \\ \vdots & \vdots & & \vdots \\ n & x_n & \cdots & x_n^{n-1} \end{vmatrix} + (-1)\begin{vmatrix} 1 & 1 & 1 & \cdots & 1 \\ 1 & x_1 & x_1^2 & \cdots & x_1^n \\ 1 & x_2 & x_2^2 & \cdots & x_2^n \\ \vdots & \vdots & \vdots & & \vdots \\ 1 & x_n & x_n^2 & \cdots & x_n^n \end{vmatrix}\\ &=[2\prod_{i=1}^n x_i-\prod_{i=1}^n (x_i-1)]\prod_{1\leq i<j\leq n}(x_j-x_i) \end{aligned} \]

    计算 \(D_{n}=\begin{vmatrix}0&a_{1}+a_{2}&\cdots&a_{1}+a_{n}\\a_{2}+a_{1}&0&\cdots&a_{2}+a_{n}\\\vdots&\vdots&\ddots&\vdots\\a_{n}+a_{1}&a_{n}+a_{2}&\cdots&0\end{vmatrix}\)

    解答
    \[ \begin{aligned}D_{n}&=\begin{vmatrix} 1&a_{1}&a_{2}&\cdots&a_{n}\\ 0&0&a_{1}+a_{2}&\cdots&a_{1}+a_{n}\\ 0&a_{2}+a_{1}&0&\cdots&a_{2}+a_{n}\\ \vdots&\vdots&\vdots&\ddots&\vdots\\ 0&a_{n}+a_{1}&a_{n}+a_{2}&\cdots&0 \end{vmatrix}_{n + 1}\\&\xlongequal{R_i-R_1}\begin{vmatrix} 1&a_{1}&a_{2}&\cdots&a_{n}\\ -1&-a_{1}&a_{1}&\cdots&a_{1}\\ -1&a_{2}&-a_{2}&\cdots&a_{2}\\ \vdots&\vdots&\vdots&\ddots&\vdots\\ -1&a_{n}&a_{n}&\cdots&-a_{n} \end{vmatrix}_{n + 1}\\&=\begin{vmatrix} 1&0&0&0&\cdots&0\\ 0&1&a_{1}&a_{2}&\cdots&a_{n}\\ a_{1}&-1&-a_{1}&a_{1}&\cdots&a_{1}\\ a_{2}&-1&a_{2}&-a_{2}&\cdots&a_{2}\\ \vdots&\vdots&\vdots&\vdots&\ddots&\vdots\\ a_{n}&-1&a_{n}&a_{n}&\cdots&-a_{n} \end{vmatrix}_{n + 2}\\&\xlongequal{C_{3\sim n + 2}-C_{1}}\begin{vmatrix} 1&0&-1&-1&\cdots&-1\\ 0&1&a_{1}&a_{2}&\cdots&a_{n}\\ a_{1}&-1&-2a_{1}&0&\cdots&0\\ a_{2}&-1&0&-2a_{2}&\cdots&0\\ \vdots&\vdots&\vdots&\vdots&\ddots&\vdots\\ a_{n}&-1&0&0&\cdots&-2a_{n} \end{vmatrix}_{n + 2} \\&\xlongequal[C_{2}-\sum\limits_{i = 3}^{n + 2}\frac{1}{2a_{i - 2}}C_{i}]{C_{1}+\sum\limits_{i = 3}^{n + 2}\frac{1}{2}C_{i}}\begin{vmatrix} 1 - \frac{n}{2}&\frac{1}{2}\sum\limits_{i = 1}^{n}\frac{1}{a_{i}}&-1&-1&\cdots&-1\\ \frac{1}{2}\sum\limits_{i = 1}^{n}a_{i}&1 - \frac{n}{2}&a_{1}&a_{2}&\cdots&a_{n}\\ 0&0&-2a_{1}&0&\cdots&0\\ 0&0&0&-2a_{2}&\cdots&0\\ \vdots&\vdots&\vdots&\vdots&\ddots&\vdots\\ 0&0&0&0&\cdots&-2a_{n} \end{vmatrix}_{n + 2}\\&= (-2)^{n}a_{1}a_{2}\cdots a_{n}[(1 - \frac{n}{2})^{2} - \frac{1}{4}(\sum_{i = 1}^{n}a_{i})(\sum\limits_{i = 1}^{n}\frac{1}{a_{i}})]\end{aligned} \]
  • 函数法