Eigenvectors | 特征向量

📚 Eigenvectors | 特征向量

In A-Level Further Mathematics, eigenvectors form a beautiful bridge between linear algebra and geometry. Understanding them is essential for mastering matrix transformations, systems of differential equations, and many real-world applications from physics to economics.

在A-Level进阶数学中，特征向量构成了线性代数与几何之间一座美妙的桥梁。理解特征向量对于掌握矩阵变换、微分方程组以及从物理学到经济学的众多现实应用都至关重要。

1. The Intuitive Idea | 直观概念

Imagine applying a matrix transformation to a vector. In general, the vector will change both its direction and its length. However, certain special vectors, called eigenvectors, only have their length changed — their direction remains exactly the same (or exactly opposite).

想象一下对向量施加矩阵变换。一般来说，向量的大小和方向都会改变。然而，某些特殊的向量，称为特征向量，只会改变长度——其方向保持不变（或完全相反）。

The factor by which the length changes is called the eigenvalue (denoted λ). If the eigenvalue is negative, the eigenvector is reversed in direction while still lying on the same straight line.

长度变化的比例因子称为特征值（用 λ 表示）。若特征值为负数，则特征向量的方向被反转，但仍位于同一直线上。

Formal definition: For a square matrix A, a non-zero vector v and a scalar λ satisfying the equation:

正式定义：对于方阵 A，若非零向量 v 和标量 λ 满足方程：

A v = λ v

then v is an eigenvector of A with eigenvalue λ.

则 v 是矩阵 A 的特征向量，λ 是其对应的特征值。

2. The Characteristic Equation | 特征方程

To find eigenvalues, we rearrange the defining equation. Since v ≠ 0, we need the matrix (A − λI) to be singular — that is, its determinant must be zero:

为求特征值，我们改写定义方程。因为 v ≠ 0，矩阵 (A − λI) 必须是奇异矩阵，即其行列式必须为零：

det(A − λI) = 0

This equation is called the characteristic equation, and it always produces a polynomial in λ of degree n for an n × n matrix.

该方程称为特征方程，对 n × n 矩阵而言，它总是产生一个关于 λ 的 n 次多项式。

Example — 2 × 2 matrix: Consider the matrix:

例如 —— 2 × 2 矩阵：考虑矩阵：

A = [3 1; 2 4]

We subtract λI from A:

我们从 A 中减去 λI：

A − λI = [3−λ 1; 2 4−λ]

Now take the determinant:

现在取行列式：

(3−λ)(4−λ) − 2×1 = λ² − 7λ + 10 = 0

Factorising gives (λ − 5)(λ − 2) = 0, so the eigenvalues are λ = 5 and λ = 2.

因式分解得 (λ − 5)(λ − 2) = 0，所以特征值为 λ = 5 和 λ = 2。

3. Finding Eigenvalues: Step-by-Step | 求特征值：逐步解析

For a general 2 × 2 matrix A = [a b; c d], the characteristic equation takes a convenient form:

对于一般的 2 × 2 矩阵 A = [a b; c d]，特征方程具有简便的形式：

λ² − (a + d)λ + (ad − bc) = 0

Notice that (a + d) is the trace of A (sum of diagonal entries), and (ad − bc) is the determinant of A.

注意，(a + d) 是 A 的迹（对角元素之和），而 (ad − bc) 是 A 的行列式。

Key observations:

The sum of the eigenvalues equals the trace: λ₁ + λ₂ = a + d
The product of the eigenvalues equals the determinant: λ₁λ₂ = ad − bc

关键观察：

特征值之和等于迹：λ₁ + λ₂ = a + d
特征值之积等于行列式：λ₁λ₂ = ad − bc

These relationships are excellent for quickly verifying your answers in an exam.

这些关系非常适合在考试中快速验证答案。

4. Finding Eigenvectors: Step-by-Step | 求特征向量：逐步解析

Once each eigenvalue λ is known, we substitute it into the equation (A − λI)v = 0 and solve for v. The solution set is a line of eigenvectors — any non-zero scalar multiple is also an eigenvector.

求得每个特征值 λ 之后，将其代入方程 (A − λI)v = 0 中解出 v。解集是一条由特征向量构成的直线——任何非零标量倍仍然是特征向量。

For the example above, first take λ = 5:

对于上面的例子，先取 λ = 5：

[3−5 1; 2 4−5] = [−2 1; 2 −1]

Solving (−2)x + y = 0 gives y = 2x. Therefore the eigenvectors have the form:

解方程 (−2)x + y = 0 得 y = 2x。因此特征向量的形式为：

v = [x; 2x] = x[1; 2]

We normally write the simplest non-zero vector: v = [1; 2]. Any multiple of this vector is also an eigenvector.

我们通常写出最简单的非零向量：v = [1; 2]。任何该向量的倍数仍然是特征向量。

Similarly, for λ = 2:

类似地，对于 λ = 2：

[1 1; 2 2] yields x + y = 0, so y = −x

Hence the eigenvector is v = [1; −1] (or any scalar multiple).

因此特征向量为 v = [1; −1]（或其任何标量倍）。

5. Worked Example — Complete Solution | 完整例题解析

Find the eigenvalues and eigenvectors of the matrix:

求以下矩阵的特征值和特征向量：

B = [2 −1; −1 2]

Step 1 — Characteristic equation:

第一步 —— 特征方程：

det(B − λI) = (2−λ)(2−λ) − (−1)(−1) = λ² − 4λ + 3 = 0

Step 2 — Factorise: (λ − 3)(λ − 1) = 0, so λ = 3 or λ = 1.

第二步 —— 因式分解： (λ − 3)(λ − 1) = 0，因此 λ = 3 或 λ = 1。

Step 3 — Eigenvector for λ = 3:

第三步 —— 求 λ = 3 的特征向量：

[−1 −1; −1 −1] v = 0

This gives −x − y = 0, so y = −x. Therefore v = [1; −1].

由此得 −x − y = 0，因此 y = −x。所以 v = [1; −1]。

Step 4 — Eigenvector for λ = 1:

第四步 —— 求 λ = 1 的特征向量：

[1 −1; −1 1] v = 0

This gives x − y = 0, so y = x. Therefore v = [1; 1].

由此得 x − y = 0，因此 y = x。所以 v = [1; 1]。

Summary:

Eigenvalue λ	Eigenvector v
3	[1; −1]
1	[1; 1]

6. Geometric Interpretation | 几何解释

Geometrically, an eigenvector of a 2 × 2 matrix transformation is a direction along which the transformation acts purely as a scaling operation. The eigenvalue tells us the scale factor.

从几何角度看，2 × 2 矩阵变换的特征向量是变换仅作为纯缩放作用的方向。特征值告诉我们缩放因子。

For example, the matrix A = [3 1; 2 4] has eigenvector [1; 2] with eigenvalue 5. This means:

例如，矩阵 A = [3 1; 2 4] 有特征向量 [1; 2] 和对应的特征值 5。这意味着：

A [1; 2] = 5 [1; 2] = [5; 10]

Notice that [5; 10] is exactly in the same direction as [1; 2] — it lies on the same line through the origin, just stretched by a factor of 5.

注意 [5; 10] 与 [1; 2] 方向完全相同——位于经过原点的同一直线上，只是被拉长了 5 倍。

Special cases:

λ = 0: the eigenvector is mapped to the zero vector.
λ = 1: the eigenvector is unchanged (fixed direction and length).
λ < 0: the eigenvector is reflected through the origin.

特殊情况：

λ = 0：特征向量被映射为零向量。
λ = 1：特征向量保持不变（方向和长度均不变）。
λ < 0：特征向量经过原点被反射。

7. Matrices with No Real Eigenvectors | 无实特征向量的矩阵

Some matrices have characteristic equations with no real solutions. Consider the rotation matrix:

某些矩阵的特征方程没有实数解。考虑旋转矩阵：

R = [0 −1; 1 0]

The characteristic equation is λ² + 1 = 0, giving λ = ±i (complex numbers). Such matrices rotate every vector through 90°, so no vector keeps its original direction — hence no real eigenvalues or eigenvectors exist.

其特征方程为 λ² + 1 = 0，即 λ = ±i（复数）。这类矩阵将每个向量旋转 90°，因此没有任何向量保持原本方向——所以不存在实数特征值和特征向量。

This highlights a crucial fact: a matrix may have no real eigenvectors at all. In AQA A-Level questions, you will usually be given matrices known to have real eigenvalues.

这揭示了一个关键事实：矩阵可能根本没有实特征向量。在 AQA A-Level 考试题中，通常会给出已知具有实数特征值的矩阵。

8. Diagonalisation and Matrix Powers | 对角化与矩阵幂

If a 2 × 2 matrix A has two linearly independent eigenvectors, we can write:

如果 2 × 2 矩阵 A 有两个线性无关的特征向量，则我们可以写成：

A = P D P⁻¹

where P is the matrix whose columns are the eigenvectors, and D is the diagonal matrix containing the corresponding eigenvalues:

其中 P 是以特征向量为列构成的矩阵，D 是由相应特征值构成的对角矩阵：

D = [λ₁ 0; 0 λ₂]

Why is this useful? Because powers of A become easy to compute:

这有什么用？因为 A 的幂变得易于计算：

Aⁿ = P Dⁿ P⁻¹ = P [λ₁ⁿ 0; 0 λ₂ⁿ] P⁻¹

Instead of multiplying A by itself many times, we simply raise the eigenvalues to the power n — a huge saving in computation.

无需将 A 反复相乘多次，只需将特征值取 n 次幂——计算量大大节省。

Example: For the matrix A = [3 1; 2 4], we found λ₁ = 5 with v₁ = [1; 2], and λ₂ = 2 with v₂ = [1; −1]. Thus:

例如：对于矩阵 A = [3 1; 2 4]，我们求得 λ₁ = 5 对应 v₁ = [1; 2]，λ₂ = 2 对应 v₂ = [1; −1]。因此：

P = [1 1; 2 −1], D = [5 0; 0 2]

If we need A¹⁰⁰, we compute P D¹⁰⁰ P⁻¹ rather than performing 99 matrix multiplications.

若我们需要 A¹⁰⁰，只需计算 P D¹⁰⁰ P⁻¹，而无需进行 99 次矩阵乘法。

9. Real-World Applications | 实际应用

Eigenvectors are not merely abstract mathematical objects — they appear throughout science and engineering.

特征向量不仅仅是抽象的数学对象——它们广泛应用于科学和工程领域。

Population dynamics: In the Leslie matrix model, the dominant eigenvalue determines the long-term growth rate of an animal population, and the corresponding eigenvector describes the stable age distribution.

种群动态：在 Leslie 矩阵模型中，主特征值决定了动物种群的长期增长率，而相应的特征向量描述了稳定的年龄分布。

Vibration analysis: The natural frequencies and mode shapes of a mechanical system are found via eigenvectors. Bridges and buildings are designed so that no resonant frequencies match expected vibrational inputs.

振动分析：机械系统的固有频率和振型通过特征向量求得。桥梁和建筑物的设计要确保没有任何共振频率与预期的振动输入相匹配。

PageRank algorithm: Google’s original PageRank algorithm used eigenvector computation to rank web pages — the principal eigenvector of a huge matrix determines page importance.

PageRank 算法：Google 最初的 PageRank 算法利用特征向量计算来为网页排序——一个巨型矩阵的主特征向量决定了页面的重要性。

Quantum mechanics: In quantum physics, every measurable quantity corresponds to an operator (a matrix), and the possible measurement outcomes are precisely its eigenvalues.

量子力学：在量子物理学中，每个可观测物理量对应一个算子（矩阵），而可能的测量结果正是其特征值。

10. Common Mistakes and Exam Tips | 常见错误与考试技巧

Students often make the following errors when working with eigenvectors:

学生在处理特征向量时经常会犯以下错误：

Forgetting v ≠ 0: The zero vector satisfies Av = λv but is never an eigenvector. Always state a non-zero eigenvector.
Losing rows when solving: When solving (A − λI)v = 0, both rows are identical in the 2 × 2 case — use only one row to find the relationship between x and y.
Incorrect signs: Check the characteristic equation by substituting back the eigenvalues and confirming det(A − λI) = 0.
Forgetting the eigenvalue ordering: When writing P and D together, ensure the i-th column of P corresponds to the i-th diagonal entry of D.

忘记 v ≠ 0：零向量满足 Av = λv 但永远不是特征向量。务必写出非零特征向量。
求解时丢失行：在求解 (A − λI)v = 0 时，2 × 2 情况下两行相同——只用一行来求 x 与 y 的关系即可。
符号错误：将特征值代回验证 det(A − λI) = 0，以检查特征方程是否正确。
特征值顺序不匹配：当同时写出 P 和 D 时，确保 P 的第 i 列对应 D 的第 i 个对角元素。

Exam tip: Questions often ask you to use eigenvectors to compute Aⁿ for large n. Always write down P, D and P⁻¹ explicitly before showing the final result — this earns method marks even if the arithmetic goes wrong.

考试技巧：考题常要求利用特征向量计算 A 的大次数幂 Aⁿ。务必先明确写出 P、D 和 P⁻¹ 再展示最终结果——即使计算有误，也能获得方法分。

11. Perfect Square Matrices and Repeated Eigenvalues | 完全平方矩阵与重特征值

Sometimes a 2 × 2 matrix has a repeated eigenvalue. For example:

有时 2 × 2 矩阵具有重特征值。例如：

C = [2 0; 0 2]

Here the characteristic equation is (λ − 2)² = 0, so λ = 2 (a repeated eigenvalue). Every non-zero vector is an eigenvector because C = 2I, meaning Cv = 2v for all v.

这里特征方程为 (λ − 2)² = 0，因此 λ = 2（重特征值）。因为 C = 2I，即对所有 v 都有 Cv = 2v，所以每个非零向量都是特征向量。

Distinguish between:

Algebraic multiplicity: the multiplicity of λ as a root of the characteristic polynomial.
Geometric multiplicity: the dimension of the eigenspace (number of linearly independent eigenvectors for that λ).

区分两种重数：

代数重数：λ 作为特征多项式根的重数。
几何重数：特征空间的维数（对应于该 λ 的线性无关特征向量的个数）。

If the geometric multiplicity equals the algebraic multiplicity for every eigenvalue, the matrix is diagonalisable.

若每个特征值的几何重数都等于其代数重数，则该矩阵可对角化。

12. Summary | 总结

Mastering eigenvectors requires a systematic approach:

掌握特征向量需要系统化的方法：

Form the characteristic equation det(A − λI) = 0.
Solve to find the eigenvalues λ.
Substitute each λ into (A − λI)v = 0.
Solve the resulting system to find at least one non-zero eigenvector.
Verify by checking Av = λv for each pair.

建立特征方程 det(A − λI) = 0。
求解得到特征值 λ。
代入每个 λ 到 (A − λI)v = 0。
求解方程组，找到至少一个非零特征向量。
验证每对是否满足 Av = λv。

Remember: the eigenvector equation Av = λv is the heart of the concept. Everything else — the characteristic polynomial, diagonalisation, matrix powers — follows from this single elegant relationship.

记住：特征向量方程 Av = λv 是整个概念的核心。其他一切——特征多项式、对角化、矩阵幂——都源自于这一个简洁而优美的关系。

Practice with a variety of 2 × 2 matrices: symmetric, triangular, and matrices with zero or negative eigenvalues. The more comfortable you are with the procedure, the faster and more accurate you will become under exam conditions.

多练习不同类型的 2 × 2 矩阵：对称矩阵、三角矩阵，以及含零或负特征值的矩阵。你对这一流程越熟悉，在考试条件下就会越快越准确。

Published by TutorHao | Mathematics Revision Series | aleveler.com

更多咨询请联系16621398022（同微信）

Eigenvectors | 特征向量