Mastering PDF Joiners: Combining Probability Density Functions | 掌握PDF合并器：组合概率密度函数

📚 Mastering PDF Joiners: Combining Probability Density Functions | 掌握PDF合并器：组合概率密度函数

Have you ever met a revision task labelled ‘pdfjoiner_(4)-196’? It may sound cryptic, but the core idea is simple: it asks you to join two probability density functions (PDFs) together to find the distribution of the sum of two independent random variables. In Edexcel A‑Level Further Statistics, this skill is essential for tackling linear combinations, waiting‑time models, and risk‑management scenarios.

你是否在复习资料中见过一个题为 ‘pdfjoiner_(4)-196’ 的练习？这个名字听起来有些神秘，但其核心思想很简单：它要求你将两个概率密度函数合并，以求出两个独立随机变量之和的分布。在 Edexcel A‑Level 进阶统计中，这项技能对于处理线性组合、等待时间模型以及风险管理场景至关重要。

---

1. What Are PDFs? | 什么是概率密度函数？

For a continuous random variable X, a probability density function f_X(x) describes the relative likelihood near each value x. It must satisfy f_X(x) ≥ 0 for all x, and the total area under the curve is 1: ∫_−∞^∞ f_X(x) dx = 1. Probabilities are found by integration, e.g. P(a < X < b) = ∫_a^b f_X(x) dx.

对于连续型随机变量 X，概率密度函数 f_X(x) 描述了在 x 附近的相对可能性。它必须满足 f_X(x) ≥ 0 对全体 x 成立，并且曲线下的总面积为 1：∫_−∞^∞ f_X(x) dx = 1。概率可通过积分求得，例如 P(a < X < b) = ∫_a^b f_X(x) dx。

In the Edexcel syllabus you will encounter PDFs such as the uniform, exponential, normal, and gamma distributions. Understanding their shapes and parameters is the first step towards joining them.

在 Edexcel 大纲中，你会遇到均匀分布、指数分布、正态分布及伽马分布等概率密度函数。理解它们的形状和参数是合并的第一步。

---

2. Independent Random Variables and Their Sums | 独立随机变量及其和

When X and Y are independent continuous random variables, their sum Z = X + Y is a new random variable whose PDF is not simply the sum of the individual PDFs. Instead, we must combine them using a mathematical tool called convolution — a ‘PDF joiner’.

当 X 和 Y 是独立的连续型随机变量时，它们的和 Z = X + Y 是一个新的随机变量，其 PDF 并非两个密度函数的简单相加。我们必须用名为“卷积”的数学工具将它们合并——这正是一个 PDF 合并器。

Independence means that the joint probability density function factorises: f_X,Y(x,y) = f_X(x) f_Y(y). This property is crucial for deriving the distribution of sums.

独立性意味着联合概率密度函数可以分解：f_X,Y(x,y) = f_X(x) f_Y(y)。这一性质对于推导和的分布至关重要。

---

3. The Convolution Formula: The Core PDF Joiner | 卷积公式：核心 PDF 合并器

For independent X and Y with PDFs f_X and f_Y, the PDF of Z = X + Y is given by the convolution integral:

对于独立的 X 和 Y，其 PDF 为 f_X 和 f_Y，则 Z = X + Y 的 PDF 由下列卷积积分给出：

f_Z(z) = ∫_−∞^∞ f_X(x) f_Y(z − x) dx

f_Z(z) = ∫_−∞^∞ f_X(x) f_Y(z − x) dx

Here, for each possible sum z, we integrate the product of f_X(x) and f_Y(z − x) over all x. In practice, the limits become finite because PDFs are usually zero outside their support.

这里，对于每一个可能的和 z，我们将 f_X(x) 与 f_Y(z − x) 的乘积对所有 x 进行积分。实际计算时，积分限会变为有限区间，因为 PDF 通常在其支撑集之外为零。

The convolution formula is the engine behind the ‘pdfjoiner’ idea: it literally weaves two density functions into one.

卷积公式正是“pdfjoiner”这一概念背后的引擎：它确实将两个密度函数编织成了一个。

---

4. Deriving the Convolution Step by Step | 逐步推导卷积公式

Start from the cumulative distribution function (CDF) of Z: F_Z(z) = P(Z ≤ z) = P(X + Y ≤ z). For continuous independent variables, this becomes a double integral over the region x + y ≤ z.

从 Z 的累积分布函数出发：F_Z(z) = P(Z ≤ z) = P(X + Y ≤ z)。对于连续独立变量，这转化为在区域 x + y ≤ z 上的二重积分。

After changing the order of integration, we obtain F_Z(z) = ∫_−∞^∞ f_X(x) [∫_−∞^z−x f_Y(y) dy] dx. Differentiating with respect to z (using Leibniz’s rule) yields the convolution formula.

交换积分次序后，可得 F_Z(z) = ∫_−∞^∞ f_X(x) [∫_−∞^z−x f_Y(y) dy] dx。对 z 求导（使用莱布尼茨法则）即得到卷积公式。

This derivation is examinable in Edexcel Further Statistics 1; you should be comfortable illustrating it with a diagram and explaining why the limits change.

这个推导过程在 Edexcel 进阶统计 1 中可能会考查；你应该能熟练地结合图示加以说明，并解释为何积分限会发生变化。

---

5. Example 1: Sum of Two Independent Uniform Distributions | 例1：两个独立均匀分布之和

Let X ~ U(0,1) and Y ~ U(0,1) independently. Their PDFs are f_X(x) = 1 for 0 ≤ x ≤ 1, and f_Y(y) = 1 for 0 ≤ y ≤ 1. We want the PDF of Z = X + Y.

设 X ~ U(0,1) 与 Y ~ U(0,1) 独立。它们的 PDF 为 f_X(x) = 1 (0 ≤ x ≤ 1)，f_Y(y) = 1 (0 ≤ y ≤ 1)。我们要求 Z = X + Y 的 PDF。

Using convolution, f_Z(z) = ∫_{max(0, z−1)}^{min(1, z)} 1·1 dx. The limits are determined by the overlaps of the supports of X and (z−X). After integration we obtain the triangular distribution:

利用卷积，f_Z(z) = ∫_{max(0, z−1)}^{min(1, z)} 1·1 dx。积分限由 X 与 (z−X) 支撑集的重叠决定。积分后得到三角形分布：

f_Z(z) = z for 0 ≤ z ≤ 1; f_Z(z) = 2 − z for 1 ≤ z ≤ 2; and 0 otherwise.

f_Z(z) = z，当 0 ≤ z ≤ 1；f_Z(z) = 2 − z，当 1 ≤ z ≤ 2；其余为零。

This result, often nicknamed ‘the hat function’, is a classic example of a PDF joiner outcome.

这个结果常被戏称为“帽子函数”，是 PDF 合并器产出的经典范例。

---

6. Using Moment Generating Functions (MGF) | 使用矩母函数

When the moment generating function M_X(t) = E[e^tX] exists, it offers a shortcut for joining PDFs. If X and Y are independent, then M_Z(t) = M_X(t) × M_Y(t) for Z = X + Y.

若矩母函数 M_X(t) = E[e^tX] 存在，它可以为合并 PDF 提供捷径。若 X 与 Y 独立，则对于 Z = X + Y，有 M_Z(t) = M_X(t) × M_Y(t)。

For example, the MGF of a normal distribution N(μ, σ²) is exp(μt + ½σ²t²). The product of two such MGFs immediately shows that the sum of two independent normal variables is again normal, with means and variances added.

例如，正态分布 N(μ, σ²) 的 MGF 为 exp(μt + ½σ²t²)。两个这样的 MGF 相乘立即可得：两个独立正态变量之和仍为正态分布，且均值相加、方差相加。

In the Edexcel FM1 exam, you will be expected to quote the MGF of common distributions and use it to identify the distribution of a sum, which is an elegant alternative to convolution.

在 Edexcel FM1 考试中，你需要能够写出常见分布的 MGF，并用之识别和分布，这是卷积的一种简洁替代方案。

---

7. Discrete Sums and Probability Generating Functions (PGF) | 离散和与概率母函数

For non‑negative integer‑valued independent random variables, the probability generating function G_X(s) = E[s^X] plays the role of a PDF joiner. For Z = X + Y, G_Z(s) = G_X(s) G_Y(s).

对于非负整值独立随机变量，概率母函数 G_X(s) = E[s^X] 扮演着 PDF 合并器的角色。对于 Z = X + Y，有 G_Z(s) = G_X(s) G_Y(s)。

A typical example: if X ~ Po(λ) and Y ~ Po(μ) independently, the resulting PGF is e^λ(s−1) × e^μ(s−1) = e^{(λ+μ)(s−1)}, which shows Z ~ Po(λ+μ). This method is fast and error‑free when the PGF is known.

一个典型示例：若 X ~ Po(λ) 与 Y ~ Po(μ) 独立，乘积 PGF 为 e^λ(s−1) × e^μ(s−1) = e^{(λ+μ)(s−1)}，表明 Z ~ Po(λ+μ)。当已知 PGF 时，这种方法快速且不易出错。

---

8. Common Mistakes When Joining PDFs | 合并 PDF 时的常见错误

One of the biggest pitfalls is forgetting to check independence. The convolution and MGF product rules only hold when the random variables are independent. Always state this assumption clearly.

最大的陷阱之一是忘记检查独立性。卷积与 MGF 乘积法则仅在随机变量独立时成立，务必要明确陈述这一假设。

Another mistake is using incorrect limits for the convolution integral. Students often copy the full (−∞, ∞) range instead of determining the overlap region where both f_X(x) and f_Y(z−x) are non‑zero. Draw a diagram to avoid this.

另一个错误是使用错误的卷积积分上下限。学生常直接照搬 (−∞, ∞) 全区间，而不是确定 f_X(x) 与 f_Y(z−x) 均非零的重叠区域。绘制简图可有效避免此类错误。

Also, remember that the PDF of the sum is not the sum of PDFs; you must use convolution or an equivalent generating function approach.

还要记住，和的 PDF 并非两个密度函数的代数和；必须使用卷积或等价的生成函数方法。

---

9. Worked Practice Question | 典型例题演示

Let X and Y be independent exponential random variables each with rate λ, so f_X(x) = λe^−λx, x ≥ 0. Find the PDF of Z = X + Y.

设 X 与 Y 独立，且均服从参数为λ的指数分布，即 f_X(x) = λe^−λx，x ≥ 0。求 Z = X + Y 的 PDF。

We use convolution: for z ≥ 0, f_Z(z) = ∫₀^z λe^−λx · λe^−λ(z−x) dx = λ²e^−λz ∫₀^z 1 dx = λ² z e^−λz. This is the PDF of a gamma distribution Gamma(2, λ), also called Erlang distribution.

使用卷积：对于 z ≥ 0，f_Z(z) = ∫₀^z λe^−λx · λe^−λ(z−x) dx = λ²e^−λz ∫₀^z 1 dx = λ² z e^−λz。这是形状参数为 2 的伽马分布 Gamma(2, λ)，也称为埃尔朗分布。

Alternatively, using the MGF of exponential (1 − t/λ)⁻¹, you quickly see the product is (1 − t/λ)⁻², confirming the Gamma(2,λ) result. This dual approach reinforces the concept of a PDF joiner.

也可以利用指数分布的 MGF (1 − t/λ)⁻¹，乘积立即得到 (1 − t/λ)⁻²，从而确认 Gamma(2,λ) 的结果。这种双重方法强化了 PDF 合并器的概念。

---

10. Tips for Edexcel A‑Level Further Statistics Exams | Edexcel A‑Level 进阶统计考试技巧

When faced with a ‘pdfjoiner’ style problem, first identify the supports of the two variables. Sketch the region where the joint density is positive. This will guide you in setting correct limits for the convolution.

遇到“pdfjoiner”类型的问题时，首先要确定两个变量的支撑集。画出联合密度为正的区域示意图。这将帮助你设置正确的卷积积分限。

Always state ‘Since X and Y are independent…’ early in your solution. Marks are often allocated for recognising that MGFs or PGFs multiply only under independence.

在解答中尽早写出“由于 X 和 Y 独立…”。识记到 MGF 或 PGF 仅在独立时才能相乘是常见的得分点。

Finally, verify your final PDF — check that it is non‑negative and integrates to 1 over its domain. These simple checks can save you from losing marks.

最后，检查求得的 PDF

Published by TutorHao | A-Level Mathematics Revision Series | aleveler.com

更多咨询请联系16621398022（同微信）