📚 5 The Convolution Theorem | 5 卷积定理
The convolution theorem is a cornerstone of mathematical analysis, linking the operation of convolution in the time domain to simple multiplication in the frequency domain. It provides an elegant method for solving differential equations, analysing linear systems, and processing signals. In this article, we explore the theorem in the context of both Laplace and Fourier transforms, walk through its proof, and illustrate its power with detailed worked examples.
卷积定理是数学分析的基石，它将时域中的卷积运算与频域中的简单乘法联系起来。它为求解微分方程、分析线性系统以及处理信号提供了一种优雅的方法。本文将在拉普拉斯变换和傅里叶变换的背景下探讨这一定理，梳理其证明过程，并通过详细的例题展示其威力。
1. What Is Convolution? | 什么是卷积？
At its heart, convolution is a mathematical operation that blends two functions to produce a third. If you think of f(t) as a signal and g(t) as a filter, the convolution (f ∗ g)(t) describes the output when the signal passes through that filter. Intuitively, it measures the area of overlap between f and a reversed and shifted version of g.
从本质上说，卷积是一种将两个函数融合生成第三个函数的数学运算。如果你把 f(t) 想象成一个信号，把 g(t) 想象成一个滤波器，那么卷积 (f ∗ g)(t) 就描述了信号通过该滤波器后的输出。直观地讲，它测量的是 f 与 g 的翻转平移版本之间的重叠面积。
In signal processing, this operation appears whenever a linear time-invariant system is driven by an input: the output is precisely the convolution of the input with the system's impulse response. The same idea carries over to differential equations, where the solution can often be expressed as a convolution of the forcing term with a fundamental solution.
在信号处理中，每当一个线性时不变系统被输入驱动时，这种运算就会出现：输出正是输入与系统冲激响应的卷积。同样的思想也适用于微分方程，其解通常可以表示成强迫项与基本解的卷积。
2. The Convolution Integral Defined | 卷积积分的定义
For functions f(t) and g(t) defined for t ≥ 0, the one-sided convolution is defined by
对于在 t ≥ 0 上定义的函数 f(t) 和 g(t)，单侧卷积定义为
(f ∗ g)(t) = ∫0t f(τ) g(t − τ) dτ
rely mathematical terms, the convolution can also be defined over the whole real line as ∫-∞∞ f(τ) g(t − τ) dτ, but the one-sided version is most natural when using Laplace transforms.
这里 τ 是积分的哑变量，积分限从 0 到 t 反映了因果性：时刻 t 的输出仅依赖于该时刻之前的输入。从纯数学角度看，卷积也可以在整个实数轴上定义为 ∫-∞∞ f(τ) g(t − τ) dτ，但在使用拉普拉斯变换时单侧版本最为自然。
The star notation f ∗ g is standard, and the convolution is commutative, associative and distributive. These properties make it extremely versatile in both theoretical and applied contexts.
星号符号 f ∗ g 是标准的，卷积满足交换律、结合律和分配律。这些性质使其在理论和应用中都极为通用。
3. Geometric Interpretation | 几何解释
To understand what convolution computes, picture the graph of g(τ). Replace τ by t − τ to obtain a reflection about the vertical axis and a shift by t. As t increases, this reflected profile slides across the graph of f(τ). The area under the product curve f(τ) g(t − τ) is exactly the value of the convolution at that t.
要理解卷积计算的内容，可以想象 g(τ) 的图像。将 τ 替换为 t − τ，就得到了关于纵轴的反射并平移了 t。随着 t 增加，这个反射的轮廓在 f(τ) 的图像上滑动。乘积曲线 f(τ) g(t − τ) 下方的面积正是卷积在 t 处的值。
This geometric picture explains why convolution is often described as a "blurring" or "smoothing" operation: it averages f against a moving weight function g. In the context of probability, if f and g are probability densities, their convolution gives the density of the sum of two independent random variables.
这种几何图像解释了卷积为什么常被描述为“模糊”或“平滑”操作：它用滑动的权函数 g 对 f 进行加权平均。在概率论的背景下，若 f 和 g 是概率密度，它们的卷积就给出了两个独立随机变量之和的密度。
4. Convolution Properties | 卷积性质
The convolution operation enjoys several algebraic properties that simplify calculations:
卷积运算具有若干代数性质，可简化计算：
Commutativity: f ∗ g = g ∗ f
交换律：f ∗ g = g ∗ f
Associativity: (f ∗ g) ∗ h = f ∗ (g ∗ h)
结合律：(f ∗ g) ∗ h = f ∗ (g ∗ h)
Distributivity: f ∗ (g + h) = f ∗ g + f ∗ h
分配律：f ∗ (g + h) = f ∗ g + f ∗ h
Identity: f ∗ δ = f, where δ is the Dirac delta function.
单位元：f ∗ δ = f，其中 δ 是狄拉克 δ 函数。
These properties mirror those of multiplication, which is no coincidence: the convolution theorem will show that, under a suitable transform, convolution truly becomes multiplication.
这些性质与乘法的性质相似，这并非巧合：卷积定理将表明，在适当的变换下，卷积确实变成了乘法。
5. Laplace Transform Essentials | 拉普拉斯变换要点
The Laplace transform of a function f(t), denoted L{f(t)} or F(s), is defined by
函数 f(t) 的拉普拉斯变换，记作 L{f(t)} 或 F(s)，定义为
F(s) = ∫0∞ f(t) e−st dt
where s is a complex variable. This transform is a powerful tool for solving linear ordinary differential equations because it turns differentiation into multiplication by s. A few basic transform pairs are listed below.
其中 s 是复变量。该变换是求解线性常微分方程的强大工具，因为它将微分运算转化为乘以 s。下面列出几个基本的变换对。
| f(t) | F(s) = L{f(t)} |
|---|---|
| 1 | 1/s |
| tn | n! / sn+1 |
| eat | 1 / (s − a) |
| sin(ωt) | ω / (s2 + ω2) |
| cos(ωt) | s / (s2 + ω2) |
These pairs will be essential when we apply the convolution theorem to invert products of transforms.
当我们将卷积定理应用于变换乘积的逆变换时，这些变换对至关重要。
6. The Convolution Theorem for Laplace Transforms | 拉普拉斯变换的卷积定理
em: if F(s) = L{f(t)} and G(s) = L{g(t)}, then the Laplace transform of the convolution (f ∗ g)(t) is simply the product F(s) G(s). Symbolically,
本文的核心结论是拉普拉斯卷积定理：若 F(s) = L{f(t)} 且 G(s) = L{g(t)}，则卷积 (f ∗ g)(t) 的拉普拉斯变换就是乘积 F(s) G(s)。用符号表示为
L{(f ∗ g)(t)} = F(s) G(s)
The equivalent statement in terms of the inverse Laplace transform is equally useful:
用拉普拉斯逆变换表述的等价命题同样有用：
L−1{F(s) G(s)} = (f ∗ g)(t) = ∫0t f(τ) g(t − τ) dτ
This means that instead of searching for a complicated inverse transform of a product, we can compute a convolution integral of two known functions. Many differential equations yield solutions in precisely this form.
这意味着我们不必去寻找乘积的复杂逆变换，而是可以计算两个已知函数的卷积积分。许多微分方程的解正是以这种形式出现的。
7. Proving the Convolution Theorem | 证明卷积定理
The proof is a beautiful interplay of integration and change of variables. Start with the left-hand side:
该证明是积分与变量替换的巧妙结合。我们从左边开始：
L{(f ∗ g)(t)} = ∫0∞ e−st [∫0t f(τ) g(t − τ) dτ] dt
he order of integration so that τ runs from 0 to ∞ and, for a fixed τ, t runs from τ to ∞:
在 (τ, t) 平面上的积分区域为 0 ≤ τ ≤ t，0 ≤ t < ∞。交换
Published by TutorHao | IB Mathematics Revision Series | aleveler.com
更多咨询请联系16621398022（同微信）
或截图后用「扫一扫 → 相册」选择本图