Fourier Coefficient Computation Techniques | 傅里叶系数的计算技巧

📚 Fourier Coefficient Computation Techniques | 傅里叶系数的计算技巧

Fourier series allow us to represent periodic functions as infinite sums of sines and cosines. The coefficients in these series are determined by integrals over one period, and computing them efficiently requires a combination of symmetry analysis, integration by parts, and clever use of orthogonality. This guide covers the essential techniques for IB Mathematics Higher Level students.

傅里叶级数使我们能够将周期函数表示为正弦和余弦的无穷和。这些级数中的系数通过对一个周期内的积分来确定，而高效计算这些系数需要结合对称性分析、分部积分法以及正交性的巧妙应用。本指南涵盖了IB数学高级水平学生所需的核心技巧。

1. The Fourier Series Formulas | 傅里叶级数公式

For a function f(x) with period 2π, the Fourier series is written as:

f(x) = a₀/2 + Σₙ₌₁^∞ [aₙcos(nx) + bₙsin(nx)]

The Fourier coefficients a₀, aₙ, and bₙ are computed using the following integrals over one full period:

a₀ = (1/π) ∫₋π^π f(x) dx

aₙ = (1/π) ∫₋π^π f(x)cos(nx) dx, n ≥ 1

bₙ = (1/π) ∫₋π^π f(x)sin(nx) dx, n ≥ 1

Note that some textbooks use αₙ or cₙ notation, but the underlying computation is identical. The factor 1/π is crucial — it ensures that the basis functions are correctly normalized.

对于一个周期为2π的函数f(x)，傅里叶级数写作：

f(x) = a₀/2 + Σₙ₌₁^∞ [aₙcos(nx) + bₙsin(nx)]

傅里叶系数a₀、aₙ和bₙ通过以下对完整周期的积分来计算：

a₀ = (1/π) ∫₋π^π f(x) dx

aₙ = (1/π) ∫₋π^π f(x)cos(nx) dx，n ≥ 1

bₙ = (1/π) ∫₋π^π f(x)sin(nx) dx，n ≥ 1

请注意，有些教材使用αₙ或cₙ的记号，但计算本质完全相同。因子1/π至关重要——它确保了基函数的正确归一化。

2. Orthogonality Relations | 正交性关系

The entire theory of Fourier series rests on the orthogonality of trigonometric functions. Over the interval [−π, π], the following relations hold for integers m and n:

∫₋π^π sin(mx)cos(nx) dx = 0 for all m, n
∫₋π^π cos(mx)cos(nx) dx = 0 if m ≠ n, and π if m = n ≠ 0
∫₋π^π sin(mx)sin(nx) dx = 0 if m ≠ n, and π if m = n ≠ 0
∫₋π^π cos(nx) dx = 0 for n ≥ 1, and ∫₋π^π 1 dx = 2π

These relations explain why each coefficient isolates exactly one term in the series. When you multiply f(x) by cos(nx) and integrate, all terms vanish except the one containing aₙ.

傅里叶级数的全部理论建立在三角函数的正交性之上。在区间[−π, π]上，对于整数m和n，以下关系成立：

∫₋π^π sin(mx)cos(nx) dx = 0，对所有m、n成立
∫₋π^π cos(mx)cos(nx) dx = 0（若m ≠ n），等于π（若m = n ≠ 0）
∫₋π^π sin(mx)sin(nx) dx = 0（若m ≠ n），等于π（若m = n ≠ 0）
∫₋π^π cos(nx) dx = 0（n ≥ 1），而∫₋π^π 1 dx = 2π

这些关系解释了为什么每个系数恰好分离级数中的一项。当你将f(x)乘以cos(nx)并积分时，除含有aₙ的项外，所有项都消失为零。

3. Exploiting Symmetry: Even and Odd Functions | 利用对称性：偶函数与奇函数

The most powerful shortcut in Fourier coefficient computation comes from recognizing even and odd symmetry. If f(x) is even (symmetric about the y-axis), then f(x)sin(nx) is odd, so all bₙ = 0. The series contains only cosine terms. If f(x) is odd (antisymmetric about the origin), then f(x)cos(nx) is odd, so all aₙ = 0 (including a₀). The series contains only sine terms.

Moreover, for even functions, the integral from −π to π equals twice the integral from 0 to π, which often simplifies the computation significantly:

aₙ = (2/π) ∫₀^π f(x)cos(nx) dx (for even f)

bₙ = (2/π) ∫₀^π f(x)sin(nx) dx (for odd f)

傅里叶系数计算中最强大的捷径来自对偶对称和奇对称的识别。如果f(x)是偶函数（关于y轴对称），则f(x)sin(nx)是奇函数，因此所有bₙ = 0，级数只包含余弦项。如果f(x)是奇函数（关于原点反对称），则f(x)cos(nx)是奇函数，因此所有aₙ = 0（包括a₀），级数只包含正弦项。

此外，对于偶函数，从−π到π的积分等于从0到π积分的两倍，这通常能显著简化计算：

aₙ = (2/π) ∫₀^π f(x)cos(nx) dx（f为偶函数时）

bₙ = (2/π) ∫₀^π f(x)sin(nx) dx（f为奇函数时）

4. Half-Range Expansions | 半区间展开

When a function is defined only on (0, L), we can construct either a cosine series (even extension) or a sine series (odd extension). For a function defined on (0, π), the half-range cosine series has coefficients:

a₀ = (2/π) ∫₀^π f(x) dx, aₙ = (2/π) ∫₀^π f(x)cos(nx) dx

The half-range sine series has coefficients:

bₙ = (2/π) ∫₀^π f(x)sin(nx) dx

This technique is valuable because it lets us represent non-periodic functions on a finite interval with a smooth, periodic extension. Note the factor 2/π here, not 1/π — this compensates for integrating over only half the period.

当函数仅在(0, L)上定义时，我们可以构造余弦级数（偶延拓）或正弦级数（奇延拓）。对于定义在(0, π)上的函数，半区间余弦级数的系数为：

a₀ = (2/π) ∫₀^π f(x) dx, aₙ = (2/π) ∫₀^π f(x)cos(nx) dx

半区间正弦级数的系数为：

bₙ = (2/π) ∫₀^π f(x)sin(nx) dx

这种技术的价值在于，它允许我们在有限区间上用平滑的周期延拓来表示非周期函数。请注意这里是因子2/π而非1/π——这是为了补偿只对半个周期进行积分。

5. Integration by Parts for Piecewise Polynomials | 分段多项式的分部积分

For functions defined piecewise, such as f(x) = x on (0, π) and f(x) = 0 on (−π, 0), the coefficient integrals must be split into subintervals. The key formula for integration by parts is:

∫ u dv = uv − ∫ v du

When computing ∫ x cos(nx) dx, choose u = x and dv = cos(nx) dx, giving du = dx and v = sin(nx)/n:

∫₀^π x cos(nx) dx = [x sin(nx)/n]₀^π − (1/n)∫₀^π sin(nx) dx = 0 + (cos(nπ) − 1)/n² = ((−1)ⁿ − 1)/n²

Similarly, ∫ x sin(nx) dx = [−x cos(nx)/n]₀^π + (1/n)∫₀^π cos(nx) dx = −π cos(nπ)/n + 0 = (−1)ⁿ⁺¹ π/n.

对于分段定义的函数，例如在(0, π)上f(x) = x而在(−π, 0)上f(x) = 0，系数的积分必须分割为子区间进行计算。分部积分的关键公式为：

∫ u dv = uv − ∫ v du

当计算∫ x cos(nx) dx时，选取u = x和dv = cos(nx) dx，得到du = dx和v = sin(nx)/n：

∫₀^π x cos(nx) dx = [x sin(nx)/n]₀^π − (1/n)∫₀^π sin(nx) dx = 0 + (cos(nπ) − 1)/n² = ((−1)ⁿ − 1)/n²

类似地，∫ x sin(nx) dx = [−x cos(nx)/n]₀^π + (1/n)∫₀^π cos(nx) dx = −π cos(nπ)/n + 0 = (−1)ⁿ⁺¹ π/n。

6. Handling Piecewise Functions | 分段函数的处理

For a piecewise function, the coefficient integrals are summed over each piece. Consider the square wave defined by f(x) = 1 for 0 < x < π and f(x) = −1 for −π < x < 0. This is an odd function, so aₙ = 0 for all n. The sine coefficients are:

bₙ = (1/π)[∫₋π^0 (−1)sin(nx) dx + ∫₀^π (1)sin(nx) dx]

Evaluating each integral separately:

bₙ = (1/π)[(cos(nx)/n)₋π^0 − (cos(nx)/n)₀^π] = (1/π)[(1 − cos(−nπ))/n − (cos(nπ) − 1)/n] = (2/nπ)(1 − (−1)ⁿ)

Thus bₙ = 4/(nπ) when n is odd, and bₙ = 0 when n is even. This yields the classic square wave series: (4/π)(sin x + sin 3x/3 + sin 5x/5 + …).

对于分段函数，系数积分需要对每一段分别求和。考虑方波：当0 < x < π时f(x) = 1，当−π < x < 0时f(x) = −1。这是一个奇函数，因此所有aₙ = 0。正弦系数为：

bₙ = (1/π)[∫₋π^0 (−1)sin(nx) dx + ∫₀^π (1)sin(nx) dx]

分别计算每个积分：

bₙ = (1/π)[(cos(nx)/n)₋π^0 − (cos(nx)/n)₀^π] = (1/π)[(1 − cos(−nπ))/n − (cos(nπ) − 1)/n] = (2/nπ)(1 − (−1)ⁿ)

因此当n为奇数时bₙ = 4/(nπ)，当n为偶数时bₙ = 0。这得到了经典方波级数：(4/π)(sin x + sin 3x/3 + sin 5x/5 + …)。

7. The Use of Trig Identities | 三角函数恒等式的应用

When f(x) itself contains trigonometric expressions, identities such as 2cosAcosB = cos(A+B) + cos(A−B) simplify the integrands dramatically. For example, if f(x) = cos²x, we can rewrite:

cos²x = (1 + cos 2x)/2

This expression already IS the Fourier series of cos²x with a₀/2 = 1/2 and a₂ = 1/2, all other coefficients zero. No integration is needed at all!

Similarly, f(x) = sin³x = (3sin x − sin 3x)/4 yields immediate coefficients. Recognizing these identities can save substantial time in exams.

当f(x)本身含有三角函数表达式时，诸如2cosAcosB = cos(A+B) + cos(A−B)这样的恒等式能极大地简化被积函数。例如，若f(x) = cos²x，我们可以改写为：

cos²x = (1 + cos 2x)/2

这个表达式本身就是cos²x的傅里叶级数，其中a₀/2 = 1/2，a₂ = 1/2，所有其他系数为零。根本不需要进行任何积分！

类似地，f(x) = sin³x = (3sin x − sin 3x)/4可以直接得到系数。在考试中识别这些恒等式可以节省大量时间。

8. Tabular Integration by Parts | 表格法分部积分

For products of polynomials with trig functions, the tabular method (also called the DI method) is extremely efficient. To compute ∫ x² cos(nx) dx, create a table with columns for derivative and integral:

Derivative column	Integral column	Sign
x²	cos(nx)	+
2x	sin(nx)/n	−
2	−cos(nx)/n²	+
0	−sin(nx)/n³	−

The integral equals x² sin(nx)/n + 2x cos(nx)/n² − 2 sin(nx)/n³. Evaluating at endpoints (with boundary terms) gives the final coefficient. This systematic approach eliminates sign errors that commonly occur in repeated integration by parts.

对于多项式与三角函数的乘积，表格法（也称为DI法）极为高效。要计算∫ x² cos(nx) dx，创建一个包含导数列和积分列的表格：

导数列	积分列	符号
x²	cos(nx)	+
2x	sin(nx)/n	−
2	−cos(nx)/n²	+
0	−sin(nx)/n³	−

积分等于x² sin(nx)/n + 2x cos(nx)/n² − 2 sin(nx)/n³。在端点处求值（包含边界项）即得最终系数。这种系统化的方法消除了在重复分部积分中常见的符号错误。

9. Evaluating at Boundaries | 边界点的求值

After computing the integrals, you must evaluate the antiderivative at the limits. Key trigonometric values to memorize:

sin(nπ) = 0, sin(nπ/2) = 0 for even n, ±1 for odd n
cos(nπ) = (−1)ⁿ, cos(2nπ) = 1
cos(nπ/2) = 0 for odd n, ±1 for even n

Be especially careful: sin(−nπ) = 0, and cos(−nπ) = cos(nπ) = (−1)ⁿ. Also note that (1/π)((−1)ⁿ − 1) equals 0 for even n and −2/π for odd n. These evaluations are where most arithmetic slips occur — slow down and check each substitution.

在计算积分之后，需要在积分限处求反导数的值。需要记忆的关键三角函数值：

sin(nπ) = 0，sin(nπ/2)在n为偶数时为0，n为奇数时为±1
cos(nπ) = (−1)ⁿ，cos(2nπ) = 1
cos(nπ/2)在n为奇数时为0，n为偶数时为±1

要格外小心：sin(−nπ) = 0，cos(−nπ) = cos(nπ) = (−1)ⁿ。同时注意(1/π)((−1)ⁿ − 1)在n为偶数时等于0，在n为奇数时等于−2/π。这些代入求值是最容易出错的地方——放慢速度，逐一检查每个代入。

10. Numerical Coefficient Cases | 特殊系数情况

The constant term a₀ requires special attention. Because the constant function 1 has integral 2π over [−π, π], the formula for a₀ is:

a₀ = (1/π) ∫₋π^π f(x) dx

This is NOT the same as the aₙ formula with n = 0 substituted naively, since (1/π)∫₋π^π f(x)cos(0x) dx incorrectly gives (1/π)∫₋π^π f(x) dx — actually this IS the same! However, the series writes the constant term as a₀/2, not a₀. This is a convention quirk: for n = 0, the general formula with cos(0x) = 1 yields the correct value of a₀, but the series representation divides it by 2.

A common exam mistake is computing a₀ correctly but then forgetting the factor 1/2 in the series. Remember: the constant term in the series is a₀/2, not a₀.

常数项a₀需要特别关注。因为常数函数1在[−π, π]上的积分为2π，a₀的公式为：

a₀ = (1/π) ∫₋π^π f(x) dx

这并不等同于天真地将n = 0代入aₙ的公式，因为(1/π)∫₋π^π f(x)cos(0x) dx错误地给出(1/π)∫₋π^π f(x) dx——实际上这确实相同！然而，级数将常数项写为a₀/2而非a₀。这是一个约定上的怪癖：对于n = 0，cos(0x) = 1时的通式确实得到正确的a₀值，但级数表示将其除以2。

考试中一个常见错误是正确计算了a₀，却忘记级数中的因子1/2。记住：级数中的常数项是a₀/2，而非a₀。

11. Worked Example: Full Computation | 完整算例

Let us compute the Fourier coefficients for f(x) = x² on [−π, π]. Since f is even, all bₙ = 0. The a₀ coefficient is:

a₀ = (1/π) ∫₋π^π x² dx = (1/π) × [x³/3]₋π^π = (1/π)(π³/3 − (−π)³/3) = 2π²/3

For aₙ with n ≥ 1, using even symmetry:

aₙ = (2/π) ∫₀^π x²cos(nx) dx

Using the tabular method with u = x² and dv = cos(nx)dx:

∫₀^π x²cos(nx) dx = [x²sin(nx)/n]₀^π + [2x cos(nx)/n²]₀^π − [2sin(nx)/n³]₀^π

= 0 + 2π cos(nπ)/n² − 0 = 2π(−1)ⁿ/n²

Therefore aₙ = (2/π) × 2π(−1)ⁿ/n² = 4(−1)ⁿ/n². The full Fourier series is:

x² = π²/3 + 4Σₙ₌₁^∞ [(−1)ⁿ cos(nx)/n²]

Setting x = π gives π² = π²/3 + 4Σₙ₌₁^∞ (1/n²), from which we derive the famous result Σₙ₌₁^∞ 1/n² = π²/6 — the Basel problem. This shows the power of Fourier series in evaluating infinite sums.

让我们计算[−π, π]上f(x) = x²的傅里叶系数。由于f是偶函数，所有bₙ = 0。a₀系数为：

a₀ = (1/π) ∫₋π^π x² dx = (1/π) × [x³/3]₋π^π = (1/π)(π³/3 − (−π)³/3) = 2π²/3

对于n ≥ 1的aₙ，利用偶对称性：

aₙ = (2/π) ∫₀^π x²cos(nx) dx

使用表格法，取u = x²和dv = cos(nx)dx：

∫₀^π x²cos(nx) dx = [x²sin(nx)/n]₀^π + [2x cos(nx)/n²]₀^π − [2sin(nx)/n³]₀^π

= 0 + 2π cos(nπ)/n² − 0 = 2π(−1)ⁿ/n²

因此aₙ = (2/π) × 2π(−1)ⁿ/n² = 4(−1)ⁿ/n²。完整的傅里叶级数为：

x² = π²/3 + 4Σₙ₌₁^∞ [(−1)ⁿ cos(nx)/n²]

令x = π，得到π² = π²/3 + 4Σₙ₌₁^∞ (1/n²)，由此推导出著名结果Σₙ₌₁^∞ 1/n² = π²/6——巴塞尔问题。这展示了傅里叶级数在求解无穷级数中的威力。

12. Common Pitfalls and Exam Tips | 常见陷阱与考试建议

Mastering these techniques requires attention to detail. Here are the most frequent errors and how to avoid them:

Forgetting the period: All formulas above assume period 2π. For period 2L, replace x by πx/L and multiply integrals by 1/L. Check the period first!
Missing the 1/π factor: Every coefficient has either 1/π or 2/π. Losing this factor changes all answers.
Confusing a₀ with the constant term: The series writes a₀/2 as the constant, not a₀.
Sign errors in integration by parts: Always double-check the alternating signs in the tabular method.
Not using symmetry: Before integrating, always ask: is f even, odd, or neither? This can cut work in half.
Endpoint evaluation errors: Carefully substitute limits; cos(nπ) = (−1)ⁿ is your friend.

When practicing, time yourself on problems like f(x) = x, f(x) = |x|, and f(x) = x³. These exercises build fluency with all the techniques above and cover most exam variations.

掌握这些技巧需要注重细节。以下是最常见的错误及如何避免它们：

忘记周期：上述所有公式都假设周期为2π。对于周期2L，将x替换为πx/L并将积分乘以1/L。首先检查周期！
丢失1/π因子：每个系数都包含1/π或2/π。丢失这个因子会改变所有答案。
混淆a₀与常数项：级数将a₀/2作为常数项，而非a₀。
分部积分中的符号错误：始终仔细检查表格法中的交替符号。
未利用对称性：在积分之前，总是先问：f是偶函数、奇函数还是都不是？这可以将工作量减半。
端点求值错误：小心地代入极限；cos(nπ) = (−1)ⁿ是你的得力助手。

在练习时，给自己计时完成f(x) = x、f(x) = |x|和f(x) = x³这类问题。这些练习能帮助你熟练掌握上述所有技巧，并覆盖大多数考试变体。

Published by TutorHao | Mathematics Revision Series | aleveler.com

更多咨询请联系16621398022（同微信）

Fourier Coefficient Computation Techniques | 傅里叶系数的计算技巧