Mastering Numerical Methods for A-Level WJEC Mathematics | A-Level WJEC 数学：数值方法考点精讲

📚 Mastering Numerical Methods for A-Level WJEC Mathematics | A-Level WJEC 数学：数值方法考点精讲

Numerical methods form a core component of the A-Level WJEC Mathematics syllabus, equipping students with practical techniques for solving equations and estimating integrals when exact algebraic solutions are unavailable or excessively tedious. These methods bridge theoretical calculus and real‑world computation, demanding not only procedural fluency but also an understanding of error analysis and convergence. This revision guide systematically covers the essential numerical methods you must master, including root‑finding algorithms, fixed‑point iteration, numerical integration, and the interpretation of results in context.

数值方法是 A-Level WJEC 数学大纲的核心模块，它教会学生用实用技巧去求解方程和估算积分——尤其是当精确代数解不存在或过于繁琐时。这些方法将理论微积分与现实计算连接起来，不仅要求操作熟练，还需理解误差分析与收敛性。本复习指南系统梳理了必须掌握的数值方法，涵盖求根算法、不动点迭代、数值积分以及在实际情境中解读结果的能力。

1. Why Numerical Methods Matter | 为什么数值方法如此重要

In pure mathematics, we often seek exact solutions: roots expressed in surds, logarithms, or trigonometric forms. However, many real‑world problems yield equations that cannot be solved symbolically. For instance, eˣ = 3 − x has a solution that cannot be written in closed form. Numerical methods provide approximations to any desired degree of accuracy. WJEC exam questions test both the execution of algorithms and the judgement of when a method is appropriate, how to refine approximations, and how to quantify uncertainty.

在纯数学中，我们常寻求精确解：用根式、对数或三角形式表达的根。然而许多实际问题产生的方程无法用符号求解。例如 eˣ = 3 − x 的解就不能写成封闭形式。数值方法可以给出任意精度的近似值。WJEC 考试既考查算法执行，也考查判断方法适用性、如何改进近似以及如何量化不确定性。

Key benefit: Works when algebra fails.
核心优势： 当代数方法失效时仍可行。
Exam focus: Showing working steps, verifying sign changes, iterating correctly, stating accuracy.
考试重点： 展示运算步骤、验证符号变化、正确迭代、声明精度。

You must internalise that every numerical answer is an estimate; thus the concept of error is inseparable from the method.

必须牢记：每个数值解都是近似值，因此误差概念与方法本身不可分割。

2. Locating Roots: Change of Sign | 确定根的位置：符号变化法

The foundational idea behind all root‑finding is the Intermediate Value Theorem: if a continuous function f(x) changes sign over an interval [a, b], then there exists at least one root α in (a, b) such that f(α) = 0. WJEC expects you to evaluate f(a) and f(b) and conclude that f(a)×f(b) < 0 implies a root lies between a and b.

所有求根方法的基础是介值定理：若连续函数 f(x) 在区间 [a, b] 上符号发生变化，则在 (a, b) 内至少存在一个根 α 使得 f(α) = 0。WJEC 要求你计算 f(a) 和 f(b)，并得出结论：f(a)×f(b) < 0 意味着 a 与 b 之间存在一个根。

Example: Show that the equation x³ − 4x + 1 = 0 has a root between 1 and 2.
f(1) = 1 − 4 + 1 = −2; f(2) = 8 − 8 + 1 = 1. Sign change, so root exists in (1, 2).

示例： 证明方程 x³ − 4x + 1 = 0 在 1 与 2 之间有根。
f(1) = 1 − 4 + 1 = −2；f(2) = 8 − 8 + 1 = 1。符号改变，故 (1, 2) 内存在根。

Always state that the function must be continuous on the interval; WJEC sometimes gives a discontinuous function to test this condition. A sign change alone is not sufficient if there is a discontinuity.

始终要说明函数在该区间上连续；WJEC 有时会给出有不连续点的函数来考查这一条件。如果存在不连续，仅靠符号变化是不够的。

3. The Bisection Method | 二分法

The bisection method systematically narrows the interval containing the root. Starting with [a, b] where f(a) and f(b) have opposite signs, compute the midpoint m = (a+b)/2. Evaluate f(m). If f(m) = 0, you have found the root exactly. Otherwise, replace a or b with m depending on which subinterval still shows a sign change. Repeat until the interval width is less than the required accuracy.

二分法系统地缩小区间来逼近根。从 f(a) 与 f(b) 异号的区间 [a, b] 开始，计算中点 m = (a+b)/2。计算 f(m)。若 f(m) = 0，则已找到精确根。否则根据哪个子区间仍保持符号变化，将 m 替换为 a 或 b。重复直至区间宽度小于所需精度。

Algorithm summary:
1. Given f, a, b, tolerance.
2. Do while (b − a)/2 > tolerance:
  m = (a+b)/2
  If f(m) = 0, stop.
  Else if f(a)×f(m) < 0 then b = m else a = m.
3. Root ≈ m with error ≤ (b−a)/2.

算法摘要：
1. 给定 f、a、b、容差。
2. 当 (b − a)/2 > 容差时循环：
  m = (a+b)/2
  如果 f(m) = 0，停止。
  否则如果 f(a)×f(m) < 0，则 b = m，否则 a = m。
3. 根 ≈ m，误差 ≤ (b−a)/2。

A typical exam question gives one iteration and asks you to complete the next, recording the midpoint and the sign of f(m). Ensure you state the new interval clearly and the final approximate root to a specified decimal place.

典型考题会给出一次迭代，让你完成下一次，并记录中点及 f(m) 的符号。务必清晰写出新区间，以及按指定小数位数给出最终近似根。

4. Linear Interpolation (False Position) | 线性插值法（试位法）

Linear interpolation improves on bisection by using a straight line between (a, f(a)) and (b, f(b)) to estimate where the root might lie. The x‑intercept of this chord is c = (a f(b) − b f(a)) / (f(b) − f(a)). Then evaluate f(c); depending on the sign, replace a or b with c. This often converges faster than bisection but can behave unpredictably if the function is not nearly linear on the interval.

线性插值法比二分法更进一步，它利用 (a, f(a)) 与 (b, f(b)) 之间的直线来估计根的位置。此弦的 x 截距为 c = (a f(b) − b f(a)) / (f(b) − f(a))。然后计算 f(c)；根据符号将 a 或 b 替换为 c。这通常比二分法收敛更快，但如果函数在区间上不是近似线性的，则可能出现不可预测的行为。

WJEC may ask you to perform one step of linear interpolation and comment on its accuracy compared to bisection. Remember to show the formula, substitution, and the resulting c correctly rounded.

WJEC 可能要求你执行一步线性插值，并与二分法比较精度。记住写出公式、代入过程以及正确舍入后得到的 c。

5. Newton‑Raphson Method | 牛顿-拉夫森法

The Newton‑Raphson method uses the gradient of f(x) to produce an iterative sequence: xₙ₊₁ = xₙ − f(xₙ)/f'(xₙ). Starting from an initial guess x₀, each step moves closer to the root, provided the function is differentiable and the starting value is sufficiently close. This method converges quadratically near a simple root, making it extremely efficient.

牛顿-拉夫森法利用 f(x) 的梯度构造迭代序列：xₙ₊₁ = xₙ − f(xₙ)/f'(xₙ)。从初始猜测值 x₀ 开始，每一步都更靠近根，前提是函数可导且初始值足够接近。在单根附近该方法为二次收敛，效率极高。

Step	Formula / Action	步骤	公式 / 动作
1	Choose x₀ near root	1	选择靠近根的 x₀
2	Compute f(xₙ) and f'(xₙ)	2	计算 f(xₙ) 和 f'(xₙ)
3	xₙ₊₁ = xₙ − f(xₙ)/f'(xₙ)	3	xₙ₊₁ = xₙ − f(xₙ)/f'(xₙ)
4	Repeat until \|xₙ₊₁ − xₙ\| < tolerance	4	重复直到 \|xₙ₊₁ − xₙ\| < 容差

Exam pitfalls: forgetting to differentiate correctly, misapplying the formula, or continuing iteration without recognising divergence. Always check if f'(xₙ) is near zero — the method fails there because the tangent is horizontal.

考试易错点：求导错误、公式套错，或未识别发散仍继续迭代。始终检查 f'(xₙ) 是否接近零——若切线水平，方法失效。

Example: Find √2 by solving x² − 2 = 0 using x₀ = 1.5.
f(x) = x² − 2, f'(x) = 2x.
x₁ = 1.5 − (2.25 − 2)/(3) = 1.5 − 0.08333… = 1.416667.
x₂ = 1.416667 − (2.006944 − 2)/(2.833334) ≈ 1.414216. Rapid convergence.

示例： 通过解 x² − 2 = 0 求 √2，x₀ = 1.5。
f(x) = x² − 2，f'(x) = 2x。
x₁ = 1.5 − (2.25 − 2)/(3) = 1.5 − 0.08333… = 1.416667。
x₂ = 1.416667 − (2.006944 − 2)/(2.833334) ≈ 1.414216。收敛迅速。

6. Fixed‑Point Iteration | 不动点迭代

To solve f(x) = 0, one can rearrange the equation into the form x = g(x). Starting from x₀, generate a sequence via xₙ₊₁ = g(xₙ). If the sequence converges, its limit is a fixed point of g, i.e., a root of the original equation. Convergence requires |g'(x)| < 1 near the root (the gradient of the iteration function must be less than 1 in magnitude).

为求解 f(x) = 0，可将方程改写为 x = g(x) 的形式。从 x₀ 出发，通过 xₙ₊₁ = g(xₙ) 生成序列。若序列收敛，其极限为 g 的不动点，也就是原方程的根。收敛需要在根附近满足 |g'(x)| < 1（迭代函数的梯度绝对值小于 1）。

WJEC frequently asks candidates to identify a suitable rearrangement, produce a cobweb or staircase diagram (if graphical), and determine whether the iteration will converge based on the gradient of g at the approximate root. You should be able to test |g'(α)| < 1 after estimating α.

WJEC 常要求考生找出合适的重组形式，绘制蛛网图或阶梯图（如涉及图形），并根据 g 在近似根处的梯度判断迭代是否收敛。你需要能在估计 α 后检验 |g'(α)| < 1。

Convergence condition: |g'(x)| < 1 in the neighbourhood of the root.

收敛条件：在根附近 |g'(x)| < 1。

For example, solving x² + x − 3 = 0. Possible rearrangements: x = √(3 − x) or x = 3/x − 1. Check which one converges for a given starting value.

例如解 x² + x − 3 = 0。可能的重组：x = √(3 − x) 或 x = 3/x − 1。检验哪个在给定初值下收敛。

7. Numerical Integration: Trapezium Rule | 数值积分：梯形法则

When definite integrals cannot be evaluated analytically, the trapezium rule estimates the area under a curve by dividing the interval [a, b] into n strips of equal width h = (b−a)/n, then approximating each strip as a trapezium. The composite formula is:

∫ₐᵇ f(x) dx ≈ h/2 [y₀ + yₙ + 2(y₁ + y₂ + … + yₙ₋₁)]

当定积分无法解析求解时，梯形法则通过将区间 [a, b] 等分为 n 个宽度 h = (b−a)/n 的小条，并将每个小条近似为梯形来估计曲线下方面积。复合公式为：

∫ₐᵇ f(x) dx ≈ h/2 [y₀ + yₙ + 2(y₁ + y₂ + … + yₙ₋₁)]

WJEC expects you to produce a table of x and y = f(x) values, apply the trapezium rule correctly, and interpret the result. The error in the trapezium rule is bounded by (b−a)³/(12n²)×max|f”(x)|, but you are more likely to be asked about whether the estimate is an over‑ or underestimate based on the concavity of the graph.

WJEC 要求你列出 x 与 y = f(x) 的表格，正确应用梯形法则并解释结果。梯形法则的误差界为 (b−a)³/(12n²)×max|f”(x)|，但更常问的是根据图形凹凸性判断估值为高估还是低估。

Concavity rule: If f”(x) > 0 on [a, b] (convex), the trapezium rule overestimates. If f”(x) < 0 (concave), it underestimates.

凹凸法则： 若在 [a, b] 上 f”(x) > 0（下凸），梯形法则高估；若 f”(x) < 0（上凸），则低估。

8. Simpson’s Rule | 辛普森法则

Simpson’s rule provides a more accurate estimate by approximating the curve with quadratic segments over pairs of strips. It requires an even number of strips (n must be even). The formula is:

∫ₐᵇ f(x) dx ≈ h/3 [y₀ + yₙ + 4(y₁ + y₃ + …) + 2(y₂ + y₄ + …)]

where the odd‑indexed ordinates are multiplied by 4 and the even‑indexed (except first and last) by 2.

辛普森法则通过用二次线段在小条对上近似曲线，提供更精确的估计。它要求条数为偶数（n 必须是偶数）。公式为：

∫ₐᵇ f(x) dx ≈ h/3 [y₀ + yₙ + 4(y₁ + y₃ + …) + 2(y₂ + y₄ + …)]

其中奇数下标的纵坐标乘以 4，偶数下标（除首尾）乘以 2。

WJEC may ask you to apply Simpson’s rule with a small number of strips (n = 2 or 4) and compare the results with the trapezium rule or the exact value. Always check that n is even, list the multipliers clearly, and maintain accuracy to the required decimal places.

WJEC 可能让你用少量分条（n = 2 或 4）应用辛普森法则，并与梯形法则或精确值比较。务必确认 n 是偶数，清晰列出乘数，并按所需小数位保持精度。

9. Error Analysis and Accuracy | 误差分析与精度

Every numerical method produces an approximate answer; WJEC questions often conclude by asking “Give your answer to 2 decimal places” or “State the accuracy of your approximation.” Understanding errors is vital:

Absolute error = |approximation − true value|.
Relative error = absolute error / |true value|.
In iterative methods, the difference between successive iterates can indicate convergence: stop when |xₙ₊₁ − xₙ| < ½×10⁻ᵏ for k decimal place accuracy, usually 5×10⁻⁽ᵏ⁺¹⁾ in WJEC mark schemes.

每种数值方法都会给出近似解；WJEC 题目常以“给出答案精确至 2 位小数”或“说明近似的精度”收尾。理解误差至关重要：

绝对误差 = |近似值 − 真值|。
相对误差 = 绝对误差 / |真值|。
在迭代方法中，连续迭代值之差可指示收敛：通常当 |xₙ₊₁ − xₙ| < ½×10⁻ᵏ 时停止，可获得 k 位小数精度，WJEC 评分方案常用 5×10⁻⁽ᵏ⁺¹⁾。

You should also relate error to the method’s order of convergence. Newton‑Raphson is second order (quadratic), meaning the number of correct digits roughly doubles each step near a simple root; fixed‑point iteration is usually first order (linear).

你还需要将误差与方法的收敛阶联系起来。牛顿-拉夫森法为二阶（二次收敛），意味着在单根附近每步正确位数大约翻倍；不动点迭代通常为一阶（线性收敛）。

10. Graphical Interpretation and Cobweb Diagrams | 图形解释与蛛网图

For fixed‑point iteration, WJEC expects you to sketch a cobweb or staircase diagram to illustrate convergence or divergence. Starting from x₀ on the x‑axis, you move vertically to y = g(x), horizontally to y = x, and repeat. If the cobweb spirals into the intersection, the iteration converges; if it moves outward, it diverges.

对于不动点迭代，WJEC 要求你绘制蛛网图或阶梯图来说明收敛或发散。从 x 轴上的 x₀ 开始，竖直移动到 y = g(x)，再水平移动到 y = x，如此重复。若蛛网旋向交点，则迭代收敛；若向外移动，则发散。

You must be able to identify the gradient behaviour of g near the fixed point: if −1 < g'(α) < 0, you get a cobweb (spiral); if 0 < g'(α) < 1, a staircase. Both converge. If |g'(α)| > 1, divergence occurs.

你必须能辨别不动点附近 g 的梯度行为：若 −1 < g'(α) < 0，得到蛛网（螺旋形）；若 0 < g'(α) < 1，得阶梯形。两者都收敛。若 |g'(α)| > 1，则发散。

11. Common Mistakes and Examiner Tips | 常见错误与考官提示

Top errors in WJEC numerical methods exams include:

Rounding too early: Always keep full accuracy in your calculator and round only final answers.
Misapplying formulas: Forgetting to multiply by 2 or 4 in Simpson’s rule, or using the wrong n.
Sign errors: Missing a negative sign when evaluating f(a).
Assuming convergence: Not checking the gradient condition for fixed‑point iteration.
Misreading accuracy instructions: Stopping iteration one step too soon or giving answer to one decimal place when two were required.

WJEC 数值方法考试中最常见的错误包括：

过早舍入： 始终在计算器中保留完整精度，仅对最终答案舍入。
公式套用错误： 辛普森法则忘记乘以 4 或 2，或使用错误的 n。
符号错误： 计算 f(a) 时遗漏负号。
假设收敛： 未检验不动点迭代的梯度条件。
精度要求读错： 过早停止迭代，或要求两位小数却只给出一位。

Examiners recommend: always show a clear table of values for integration, label rows and columns, and write the formula before substituting numbers. For iterative methods, state x₁ = …, x₂ = … with at least 5 significant figures for intermediate steps.

考官建议：积分别忘了画清晰的数值表，标注行和列，先写公式再代入数字。迭代方法中，中间步骤至少保留 5 位有效数字，写明 x₁ = …、x₂ = …。

12. Summary and Exam Strategy | 总结与应试策略

Numerical methods in WJEC A-Level Mathematics are not merely a set of routines; they embody the transition from exact algebra to approximate computation. Mastery requires you to:

Know the algorithm for each method and when to apply it.
Perform calculations accurately without formula sheets in the non‑calculator paper.
Understand error, convergence, and graphical interpretations.
Present your solution with clarity: formula, substitution, result, check.

WJEC A-Level 数学中的数值方法不仅是一套流程；它们体现了从精确代数到近似计算的转变。要掌握这部分，你需要：

熟悉每种方法的算法及适用时机。
在非计算器卷中不依赖公式表进行精确计算。
理解误差、收敛和图形解释。
清晰地呈现解答：公式、代入、结果、检验。

Practice past papers focusing on the structured steps WJEC demands. For the bisection method, tables with columns for a, b, m, f(a), f(b), f(m) and sign change are expected. For Newton‑Raphson, show the derivative and then the iterative formula with values substituted. By internalising these patterns, you will turn numerical methods from a potential pitfall into a reliable mark‑winner.

通过练习真题，专注于 WJEC 要求的步骤化表述来熟练。二分法需列出包含 a、b、m、f(a)、f(b)、f(m) 和符号变化的表格。牛顿-拉夫森法则要展示导数及代入值后的迭代公式。将这些模式内化，就能让数值方法从潜在陷阱变为可靠的得分点。

Published by TutorHao | Mathematics Revision Series | aleveler.com

更多咨询请联系16621398022（同微信）

Mastering Numerical Methods for A-Level WJEC Mathematics | A-Level WJEC 数学：数值方法 考点精讲