📚 Numerical Methods for A-Level AQA Mathematics | A-Level AQA 数学：数值方法 考点精讲
Numerical methods allow mathematicians and scientists to find approximate solutions to problems that cannot be solved analytically. In AQA A-Level Mathematics, this topic covers root finding through sign change and iteration, numerical integration using the trapezium rule, and the derivation and application of the Newton–Raphson method. Understanding error behavior, limitations of iterative schemes, and how to justify a given level of accuracy is essential for exam success. This revision guide unpacks every key concept and technique, providing examples, commentary, and common pitfalls to avoid.
数值方法使数学家和科学家能够为无法解析求解的问题找到近似解。在 AQA A-Level 数学中，该主题涵盖通过符号变化和迭代求根、使用梯形法则进行数值积分，以及牛顿–拉弗森方法的推导和应用。理解误差行为、迭代方案的局限性以及如何证明给定的精度水平对考试成功至关重要。本复习指南详细讲解每一个关键概念和技术，提供示例、点评和常见错误以避免。
1. Locating Roots Using Sign Changes | 通过符号变化定位根
onfirms the interval contains a root when f is continuous. Always state that f(x) is continuous on the interval to justify use of the sign change rule.
对于连续函数 f(x)，如果 f(a) 和 f(b) 具有相反的符号，则区间 [a, b] 内至少存在一个根。这是根定位的基础。应用时计算 f(a) × f(b)；乘积为负确认区间内含根，前提是 f 连续。务必说明 f(x) 在区间上连续，以证明使用符号变化规则的合理性。
Example: show that x³ – 3x + 1 = 0 has a root between 1 and 2. Compute f(1) = –1, f(2) = 3, so sign change; as f is a polynomial (continuous), root exists in (1, 2).
示例：证明方程 x³ – 3x + 1 = 0 在 1 和 2 之间存在一个根。计算 f(1) = –1，f(2) = 3，符号变化；由于 f 是多项式（连续），根存在于 (1, 2)。
2. Iteration and Rearranging Equations | 迭代与方程重组
ue x₀, then apply xₙ₊₁ = g(xₙ). Not all rearrangements converge; a sufficient condition is |g'(x)| < 1 near the root. If iterations diverge, choose a different rearrangement.
迭代生成收敛于根的一系列近似值。首先将 f(x) = 0 重组为 x = g(x) 的形式。选择一个初始值 x₀，然后应用 xₙ₊₁ = g(xₙ)。并非所有重组都收敛；一个充分条件是 |g'(x)| < 1 在根附近。如果迭代发散，应选择不同的重组形式。
A common exam task: show that x = √3x – 1 is a rearrangement of x³ – s clearly. Then use x₀ = 1.5 to compute x₁, x₂, x₃, rounding to a specified decimal place.
一个常见的考试任务：证明 x = √3x – 1 是 x³ – 3x + 1 = 0 的重组。清晰地展示代数步骤。然后使用 x₀ = 1.5 计算 x₁, x₂, x₃，并按指定小数位舍入。
sketching may be required to show the path of iterations on a graph of y = x and y = g(x).
迭代时，保留完整的计算器精度直到最终舍入。可能会要求绘制蛛网图或阶梯图来说明收敛行为。在 AQA 中，可能需要绘制在 y = x 和 y = g(x) 图形上的迭代路径。
3. Convergence Condition and Cobweb Diagrams | 收敛条件与蛛网图
d-point iteration xₙ₊₁ = g(xₙ) converges to a root α if the gradient of g near α satisfies |g'(α)| steeper than 1 in magnitude near the intersection with y = x, the iteration will diverge.
固定点迭代 xₙ₊₁ = g(xₙ) 收敛到根 α，如果 g 在 α 附近的梯度满足 |g'(α)| < 1。实际上，学生可能被要求解释为什么给定的重组有效。如果 y = g(x) 在接近 y = x 交点处的梯度绝对值大于 1，则迭代将发散。
s to the curve y = g(x) and horizontal lines to y = x. Staircase convergence occurs when g'(x) > 0; cobweb convergence when g'(x) < 0. Both indicate |g'(x)| < 1. Understanding these diagrams helps visualise approximation paths.
蛛网图通过画垂线连接到曲线 y = g(x) 和水平线连接到 y = x 来构造。当 g'(x) > 0 时出现阶梯收敛；当 g'(x) < 0 时为蛛网收敛。两者都表明 |g'(x)| < 1。理解这些图示有助于可视化近似路径。
4. The Newton–Raphson Method | 牛顿–拉弗森方法
The Newton–Raphson method uses tangents to approximate roots. The iteration formula is:
牛顿–拉弗森方法使用切线来近似根。迭代公式为：
xₙ₊₁ = xₙ – f(xₙ) / f'(xₙ)
It follows from the tangent line approximation: f'(xₙ) ≈ (f(xₙ) – 0) / (xₙ – le roots, meaning it is very fast once we are close. However, it requires the derivative f'(x) and may fail if f'(xₙ) is zero or very small.
它源自切线近似：f'(xₙ) ≈ (f(xₙ) – 0) / (xₙ – xₙ₊₁)。整理后得到公式。该方法在简单根附近以二次速度收敛，即一旦接近根就非常快。然而，它需要导数 f'(x)，如果 f'(xₙ) 为零或非常小，可能会失败。
Example: find an improved estimate for a root of x² – 5 = 0 starting with x₀ = 2. f(x) = x² – 5, f'(x) = 2x. Then x₁ = 2 – (4 – 5)/(4) = 2 – (–1/4) = 2.25. The sequence converges rapidly to √5 ≈ 2.23607.
示例：以 x₀ = 2 开始，求方程 x² – 5 = 0 根的改进估计值。f(x) = x² – 5，f'(x) = 2x。则 x₁ = 2 – (4 – 5)/(4) = 2 – (–1/4) = 2.25。序列迅速收敛至 √5 ≈ 2.23607。
5. Derivation of Newton–Raphson from First Principles | 从基本原理推导牛顿–拉弗森
ive the formula by considering the tangent at (xₙ, f(xₙ)). The tangent line has equation y – f(xₙ) = f'(xₙ)(x – xₙ). To find where it crosses the x-axis, set y = 0. Then –f(xₙ) = f'(xₙ)(x – xₙ), so x – xₙ = –f(xₙ)/f'(xₙ), giving x = xₙ – f(xₙ)/f'(xₙ). This x becomes the next approximation xₙ₊₁.
AQA 可能要求学生通过考虑在 (xₙ, f(xₙ)) 处的切线来推导公式。切线方程为 y – f(xₙ) = f'(xₙ)(x – xₙ)。要找到它与 x 轴的交点，设 y = 0。则 –f(xₙ) = f'(xₙ)(x – xₙ)，所以 x – xₙ = –f(xₙ)/f'(xₙ)，得到 x = xₙ – f(xₙ)/f'(xₙ)。该 x 成为下一个近似值 xₙ₊₁。
Understanding the geometry reinforces why the method works and why it fails at stationary points. You must be able to derive this quickly and apply it to a variety of functions, including exponentials, logarithms, and trigonometric functions.
理解几何意义可以强化该方法为何有效以及为何在驻点处失败。你必须能够快速推导并将其应用于各种函数，包括指数函数、对数函数和三角函数。
6. Failure Cases of Newton–Raphson | 牛顿–拉弗森方法的失败情形
Newton–Raphson can fail in several ways: if f' d never meets the x-axis, leading to division by zero. If the initial guess is near a turning point, the next estimate can jump far away, potentially diverging. Sometimes the sequence can oscillate between two values without settling on a root. Also, if the root is not simple (multiplicity > 1), convergence may be very slow.
牛顿–拉弗森方法可能在以下情况失败：如果 f'(xₙ) = 0，切线水平且永远不会与 x 轴相交，导致除以零。如果初始猜测靠近一个转折点，下一个估计值可能跳跃到远处，可能发散。有时序列可能在两个值之间振荡而不稳定于根。此外，如果根不是单根（重数大于 1），收敛可能非常缓慢。
lways reference the derivative and the geometry. For instance, starting at a point where f'(x) ≈ 0 leads to a very large change in x, moving the estimate far away.
AQA 考题经常展示一个图表，并要求解释该方法为何对给定的起始值会失败。始终提及导数和几何意义。例如，在 f'(x) ≈ 0 的点开始会导致 x 发生非常大的变化，将估计值移得很远。
7. Accuracy Verification Using Bounds and Decimal Places | 使用界限和小数位验证精度
To justify that an approximation is correct to a given number of decimal places, we must show that the root lies between two values that round to that approximation. For example, to prove 1.532 correct to 3 decimal places (3 d.p.), we need to show the root is in the interval [1.5315, 1.5325). This is done by evaluating f(x) at the lower and upper bounds and confirming a sign change.
要证明一个近似值在给定的小数位数下是正确的，我们必须显示根位于四舍五入到该近似值的两个值之间。例如，要证明 1.532 精确到 3 位小数 (3 d.p.)，我们需要显示根在区间 [1.5315, 1.5325) 内。这通过在下界和上界处评估 f(x) 并确认符号变化来完成。
If f(1.5315) and f(1.5325) have opposite signs and the function is continuous on that interval, then the root is indeed between them, guaranteeing the approximation. This technique is heavily examined in AQA papers.
如果 f(1.5315) 和 f(1.5325) 具有相反符号且函数在该区间上连续，则根确实在它们之间，保证了近似值。该技术在 AQA 试卷中被重度考查。
8. Numerical Integration: The Trapezium Rule | 数值积分：梯形法则
e area under a curve y = f(x) between x = a and x = b using n strips of equal width h = (b – a)/n. The formula is:
梯形法则使用 n 个等宽条带 h = (b – a)/n 近似计算曲线 y = f(x) 下 x = a 到 x = b 之间的面积。公式为：
∫ₐᵇ f(x) dx ≈ ½ h [y₀ + yₙ + 2(y₁ + y₂ + … + yₙ₋₁)]
Here y₀ = f(a), yₙ = f(b), and y₁, y₂, … areas of trapeziums formed by connecting adjacent points on the curve with straight lines.
这里 y₀ = f(a)，yₙ = f(b)，y₁, y₂, … 是中间纵坐标。该法则将曲线上相邻点用直线连接起来所形成的梯形面积求和。
For increasing functions, the trapezium rule overestimates the area; for decreasing functions, it underestimates. This is because the curve lies above or below the straight line segments. Understanding this bias helps in interpreting the approximation error.
对于递增函数，梯形法则高估面积；对于递减函数，它低估面积。这是因为曲线位于直线段的上方或下方。理解这种偏差有助于解释近似误差。
9. Applying the Trapezium Rule to Exam-Style Questions | 将梯形法则应用于考试题型
A common problem provides a table of values for f(x) at equally spaced x. Compute h = (b – . Often a subsequent part asks how the estimate could be improved: by increasing n (using more strips) the approximation becomes more accurate as the linear segments better match the curve.
一个常见问题提供 f(x) 在等距 x 处的值表。计算 h = (b – a) / n，其中 n 是条带数。小心应用公式，将所有内部纵坐标值加倍。将最终答案舍入到所需精度。通常后续部分会问如何改进估计：通过增加 n（使用更多条带），由于线性段更好地匹配曲线，近似值变得更精确。
Simpson' n and asked to calculate the percentage error of the trapezium rule estimate. Use: error = |approximate – exact| / exact × 100%.
辛普森法则不在 AQA 大纲中，因此仅考查梯形法则进行数值积分。但是，可能会给你通过积分获得的更精确值，并要求计算梯形法则估计值的百分比误差。使用：误差 = |近似值 – 精确值| / 精确值 × 100%.
10. Comparing Iterative Methods and Error Propagation | 比较迭代方法与误差传播
Different iterative methods have different speeds and reliability. Fixed-point iteration (rearrangement) converges linearly, while Newton–Raphson converges quadratically, meaning the number of correct digits roughly doubles with each step near a simple root. Understanding this helps when analyzing multiple methods: Newton–Raphson reaches high accuracy with fewer steps, provided it doesn't fail.
不同的迭代方法具有不同的速度和可靠性。固定点迭代（重组）以线性速度收敛，而牛顿–拉弗森以二次速度收敛，意味着在简单根附近每次正确数字的数量大约翻倍。理解这一点有助于分析多种方法：牛顿–拉弗森以较少的步骤达到高精度，前提是它不会失败。
Error propagation is also a key idea: if approximations are truncated prematurely, final results may be inaccurate. Always work with full precision until the final rounding step. In numerical integration, the error of the trapezium rule is roughly proportional to 1/n², so doubling the strips reduces the error by about a factor of four.
误差传播也是一个关键概念：如果过早截断近似值，最终结果可能不准确。始终保留完整精度直到最终舍入步骤。在数值积分中，梯形法则的误差大约与 1/n² 成正比，因此增加一倍的条带数大约将误差减少到原来的四分之一。
11. Sketching Iterations and Interpreting Charts | 绘制迭代图与解读图表
ubsequent x₁, x₂, x₃ along the x-axis. If the lines spiral inward, the iteration converges; if they move outward, it diverges. The gradient of g(x) at the intersection determines the pattern.
考题可能展示 y = g(x) 和 y = x 的图形，并标出起始值 x₀。你需要通过交替画垂线连接 y = g(x) 和水平线连接 y = x 来产生蛛网或阶梯图案。在 x 轴上标记每个后续的 x₁, x₂, x₃。如果线条向内螺旋，迭代收敛；如果向外移动，则发散。g(x) 在交点处的梯度决定了模式。
If g'(α) is positive and less than 1, a staircase forms; if negative and between –1 and 0, a cobweb appears. Both indicate convergence. If |g'(α)| > 1, the cobweb/staircase moves away from the root. Understanding these diagrams is directly tested in AQA, often worth several marks.
如果 g'(α) 为正且小于 1，形成阶梯；如果为负且在 –1 和 0 之间，出现蛛网。两者都表明收敛。如果 |g'(α)| > 1，蛛网/阶梯从根远离。理解这些图表在 AQA 中被直接测试，通常值几分。
12. Common Pitfalls and Examiner Tips | 常见错误与考官提示
te continuity when using the sign change rule; misapplying the trapezium rule by missing the factor ½ or doubling the end terms; rounding intermediate values instead of keeping full accuracy; incorrectly rearranging equations for iteration; failing to verify convergence condition |g'(x)| < 1; and using an initial guess for Newton–Raphson where f' read the question to see if it specifies a method or asks for a particular rearrangement.
常见错误包括：在使用符号变化规则时忘记说明连续性；在应用梯形法则时遗漏因子 ½ 或误将端点项加倍；舍入中间值而不是保留完整精度；为迭代错误地重组方程；未能验证收敛条件 |g'(x)| < 1；以及在 f'(x) ≈ 0 处使用牛顿–拉弗森的初始猜测。务必阅读题目是否指定了方法或要求特定的重组。
ated correct to 3 d.p., show the two bounds that would round to that number. If a question says “use x₀ = 1” it may mean show the next three iterations; always present them clearly, often in a table.
当证明根在给定小数位数下正确时，明确显示 f(bound1) × f(bound2) < 0。使用正确记号：如果说明根精确到 3 位小数，显示会舍入到该数的两个界限。如果题目说“使用 x₀ = 1”，这可能意味着显示后续三次迭代；始终清楚呈现，通常在表格中。
Finally, remember that numerical methods problems often link to calculus concepts. Newton–Raphson requires differentiation, and trapezium rule interpretation links to curve shape. Revising integration and differentiation thoroughly will support your numerical methods success.
最后，请记住数值方法题目常常与微积分概念相关联。牛顿–拉弗森需要求导，梯形法则的解释与曲线形状相关。彻底复习积分和微分将支持你在数值方法上的成功。
Published by TutorHao | AQA 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