📚 Eigenvectors and Eigenvalues | 特征向量与特征值
Eigenvectors and eigenvalues are among the most beautiful and powerful concepts in linear algebra, and they form a key part of the AQA A-Level Further Mathematics specification. Whether you are analysing matrix transformations, solving systems of differential equations, or preparing for a demanding exam question, mastering eigenvalues and eigenvectors is essential. In this article, we will explore their definitions, the characteristic equation, worked examples, diagonalisation, powers of matrices, and the geometric intuition behind them — with everything you need to succeed in your AQA exam.
特征向量与特征值是线性代数中最优美、最强大的概念之一，也是 AQA A-Level 进阶数学大纲中的关键内容。无论你是在分析矩阵变换、求解微分方程组，还是在备战高难度的考试题目，掌握特征值与特征向量都至关重要。在本文中，我们将探讨它们的定义、特征方程、实例演算、对角化、矩阵的幂以及背后的几何直觉——涵盖你在 AQA 考试中取得成功的全部要点。
1. What Are Eigenvectors and Eigenvalues? | 什么是特征向量与特征值
Let A be a square matrix. A non-zero vector v is called an eigenvector of A if multiplying A by v produces a vector that is simply a scalar multiple of v. The scalar is called the corresponding eigenvalue. In symbols:
Av = λv, where v ≠ 0
设 A 是一个方阵。若非零向量 v 满足：A 乘以 v 后得到的向量恰好是 v 的某个标量倍，则 v 称为 A 的特征向量，该标量称为对应的特征值。用符号表示：
Av = λv，其中 v ≠ 0
The word ‘eigen’ comes from German, meaning ‘own’ or ‘characteristic’. These vectors are called characteristic vectors because, under the transformation defined by A, they retain their own direction — only their length (and possibly their orientation) changes. For example, if λ = 3, the vector is stretched to three times its length; if λ = -1, it is reflected; if λ = 0, it is mapped to the zero vector.
“eigen” 一词源自德语，意为”本征的”或”特征的”。这些向量之所以被称为特征向量，是因为在 A 所定义的变换下，它们保持自身方向不变——只有长度（以及可能的方向）发生改变。例如，若 λ = 3，向量被拉伸为原来的三倍；若 λ = -1，向量被反射；若 λ = 0，向量被映射为零向量。
2. The Characteristic Equation | 特征方程
To find eigenvalues, we manipulate the defining equation Av = λv. First, bring every term to one side:
Av – λv = 0, so (A – λI)v = 0
为了求特征值，我们变形基本方程 Av = λv。首先将所有项移到一侧：
Av – λv = 0，即 (A – λI)v = 0
Here I is the identity matrix of the same size as A. We require v to be non-zero. For the equation (A – λI)v = 0 to have a non-trivial solution, the matrix (A – λI) must be singular — that is, its determinant must equal zero:
det(A – λI) = 0
这里的 I 是与 A 同阶的单位矩阵。我们要求 v 非零。要使方程 (A – λI)v = 0 存在非零解，矩阵 (A – λI) 必须是奇异的——即它的行列式必须为零：
det(A – λI) = 0
This equation is called the characteristic equation of A. Expanding det(A – λI) produces a polynomial in λ, called the characteristic polynomial. For an n×n matrix, the characteristic polynomial has degree n, so there are at most n eigenvalues (counting multiplicities).
该方程称为 A 的特征方程。展开 det(A – λI) 可得到关于 λ 的多项式，称为特征多项式。对于 n×n 矩阵，特征多项式的次数为 n，因此最多有 n 个特征值（按重数计）。
3. Finding Eigenvalues: A Worked Example | 求特征值：实例演示
Let us compute the eigenvalues of the matrix:
A = [ [2, 1], [1, 2] ]
让我们计算矩阵
A = [ [2, 1], [1, 2] ]
的特征值。首先构造 A – λI：
First construct A – λI:
A – λI = [ [2 – λ, 1], [1, 2 – λ] ]
Now take its determinant and set it to zero:
det(A – λI) = (2 – λ)(2 – λ) – (1)(1) = (2 – λ)² – 1 = 0
现在求其行列式并令其为零：
det(A – λI) = (2 – λ)(2 – λ) – (1)(1) = (2 – λ)² – 1 = 0
Expanding the square gives λ² – 4λ + 4 – 1 = 0, which simplifies to:
λ² – 4λ + 3 = 0, so (λ – 1)(λ – 3) = 0
展开完全平方得 λ² – 4λ + 4 – 1 = 0，化简为：
λ² – 4λ + 3 = 0，即 (λ – 1)(λ – 3) = 0
Therefore the eigenvalues are λ = 1 and λ = 3. Notice that the sum of the eigenvalues is 1 + 3 = 4, which equals the trace of A (2 + 2), and the product is 1 × 3 = 3, which equals det(A) = 4 – 1. These checks are quick and valuable in an exam.
因此特征值为 λ = 1 和 λ = 3。注意到特征值之和为 1 + 3 = 4，恰好等于 A 的迹（2 + 2）；特征值之积为 1 × 3 = 3，恰好等于 det(A) = 4 – 1。这些检验在考试中既快捷又实用。
4. Finding Eigenvectors | 求特征向量
Once the eigenvalues are known, we substitute each one back into (A – λI)v = 0 to find the corresponding eigenvectors. Let us continue with the same matrix A.
求出特征值后，我们将每个特征值代回 (A – λI)v = 0，即可找到对应的特征向量。我们继续使用同一个矩阵 A。
For λ = 3, we have A – 3I = [ [-1, 1], [1, -1] ]. We solve:
[ [-1, 1], [1, -1] ] [x, y]ᵀ = [0, 0]ᵀ
对于 λ = 3，有 A – 3I = [ [-1, 1], [1, -1] ]。我们求解：
[ [-1, 1], [1, -1] ] [x, y]ᵀ = [0, 0]ᵀ
The first row gives -x + y = 0, so y = x. A second equation is redundant, so the solution is one-dimensional:
v = k [1, 1]ᵀ, for any non-zero constant k
第一行给出 -x + y = 0，即 y = x。第二个方程是冗余的，因此解是一维的：
v = k [1, 1]ᵀ，其中 k 为任意非零常数
For λ = 1, we have A – I = [ [1, 1], [1, 1] ]. The equation x + y = 0 gives y = -x, so:
v = k [1, -1]ᵀ, for any non-zero constant k
对于 λ = 1，有 A – I = [ [1, 1], [1, 1] ]。方程 x + y = 0 给出 y = -x，因此：
v = k [1, -1]ᵀ，其中 k 为任意非零常数
Any non-zero scalar multiple of an eigenvector is still an eigenvector corresponding to the same eigenvalue. In exam answers, you may write the simplest representative, such as [1, 1]ᵀ or [1, -1]ᵀ, and state that the general form includes an arbitrary constant. When a matrix has two distinct eigenvalues for a 2×2 matrix, the two eigenvectors are automatically linearly independent.
特征向量的任意非零标量倍仍然是对应同一特征值的特征向量。在考试作答中，你可以写出最简单的代表向量，如 [1, 1]ᵀ 或 [1, -1]ᵀ，并说明一般形式中包含一个任意常数。当 2×2 矩阵具有两个不同的特征值时，对应的两个特征向量自动线性无关。
5. The Trace and Determinant Connection | 迹与行列式的关系
For any 2×2 matrix A = [ [a, b], [c, d] ], the characteristic equation takes a special form:
λ² – (a + d)λ + (ad – bc) = 0
对于任意 2×2 矩阵 A = [ [a, b], [c, d] ]，特征方程有如下特殊形式：
λ² – (a + d)λ + (ad – bc) = 0
Here a + d is the trace of A (denoted tr(A)), and ad – bc is the determinant (det(A)). Comparing this with a quadratic whose roots are λ₁ and λ₂, we obtain two important identities:
- λ₁ + λ₂ = tr(A) — the sum of eigenvalues equals the trace
- λ₁λ₂ = det(A) — the product of eigenvalues equals the determinant
这里 a + d 是 A 的迹（记作 tr(A)），ad – bc 是行列式（det(A)）。将其与以 λ₁ 和 λ₂ 为根的二次方程比较，得到两个重要恒等式：
- λ₁ + λ₂ = tr(A)——特征值之和等于迹
- λ₁λ₂ = det(A)——特征值之积等于行列式
These relationships hold for matrices of any size, provided all eigenvalues are counted with multiplicity. They offer a fast way to check your work: if your eigenvalues do not sum to the trace, you have made an error somewhere. For example, in Section 3 we found eigenvalues 1 and 3 for A = [ [2, 1], [1, 2] ]; indeed tr(A) = 4 = 1 + 3 and det(A) = 3 = 1 × 3.
这些关系对任意阶数的矩阵都成立，前提是特征值按重数计算。它们提供了一种快速检验的方法：如果你的特征值之和不等于迹，那么某处一定出错了。例如，在第 3 节中我们求得 A = [ [2, 1], [1, 2] ] 的特征值为 1 和 3；确实 tr(A) = 4 = 1 + 3，det(A) = 3 = 1 × 3。
6. Repeated Eigenvalues | 重特征值
Sometimes the characteristic polynomial has a repeated root. For example, consider the matrix:
B = [ [1, 1], [0, 1] ]
有时特征多项式具有重根。例如，考虑矩阵：
B = [ [1, 1], [0, 1] ]
Its characteristic equation is (1 – λ)² = 0, so λ = 1 is a repeated eigenvalue of algebraic multiplicity 2. Solving (B – I)v = 0 gives:
[ [0, 1], [0, 0] ] [x, y]ᵀ = [0, 0]ᵀ, hence y = 0
其特征方程为 (1 – λ)² = 0，故 λ = 1 是代数重数为 2 的重特征值。求解 (B – I)v = 0：
[ [0, 1], [0, 0] ] [x, y]ᵀ = [0, 0]ᵀ，故 y = 0
Thus every eigenvector has the form v = k[1, 0]ᵀ. There is only one linearly independent eigenvector, even though the eigenvalue has multiplicity two. Such a matrix is called defective, and it cannot be diagonalised. In your AQA examination, you should be able to recognise this situation: if an eigenvalue of multiplicity m yields fewer than m linearly independent eigenvectors, diagonalisation is impossible.
因此每个特征向量都具有 v = k[1, 0]ᵀ 的形式。尽管特征值的重数为 2，却只有一个线性无关的特征向量。这样的矩阵称为亏损矩阵，它不能被对角化。在 AQA 考试中，你应当能够识别这种情况：如果重数为 m 的特征值只能产生少于 m 个线性无关的特征向量，则该矩阵不可对角化。
7. Properties of Eigenvalues | 特征值的性质
The following properties of eigenvalues are frequently tested or useful in simplifying long calculations. We assume A is an n×n matrix with eigenvalue λ and corresponding eigenvector v.
以下特征值的性质经常被考查，或有助于简化复杂的计算。我们假设 A 是 n×n 矩阵，具有特征值 λ 和对应的特征向量 v。
| Matrix | Eigenvalue | Remark |
|---|---|---|
| A + kI | λ + k | Shift by k |
| kA | kλ | Scale by k |
| A² | λ² | Same eigenvector v |
更多咨询请联系16621398022（同微信）
CommentsMore posts |
屏轩国际教育cambridge primary/secondary checkpoint, cat4, ukiset,ukcat,igcse,alevel,PAT,STEP,MAT, ibdp,ap,ssat,sat,sat2课程辅导，国外大学本科硕士研究生博士课程论文辅导Cancel reply