Unit E6: Maclaurin
Series单元 E6:麦克劳林级数
The capstone of HL calculus. A Maclaurin series is the polynomial that a smooth function would be if its derivatives at zero were Taylor coefficients. The unit is built on five memorised series ($e^x$, $\sin x$, $\cos x$, $\ln(1+x)$, $(1+x)^n$) and three ways to extend them: substitution, term-by-term differentiation and integration, and combination with L'Hopital's rule for $0/0$ limits.本单元是 HL 微积分的收束。麦克劳林级数把光滑函数展成多项式,零点处的各阶导数即为泰勒系数。建立在五个必背级数($e^x$、$\sin x$、$\cos x$、$\ln(1+x)$、$(1+x)^n$)之上,再用三条扩展线索:代换、逐项求导与积分,以及与洛必达法则配合处理 $0/0$ 极限。
How to use this guide本指南使用说明
Every section in Unit E6 carries the HL flag. Maclaurin series is an HL-only super-topic; SL students do not see it. The unit reads in two passes: first memorise the five standard series in E6.1, then learn the three derivation techniques (E6.2, E6.3) and the two application techniques (E6.4, E6.6) that build on them. L'Hopital's rule (E6.5) is the classical alternative to E6.6 and is treated in parallel.E6 全单元都带 HL 标记。麦克劳林级数为 HL 专属,SL 不涉及。本单元按两遍阅读:先把 E6.1 的五个标准级数背熟,再掌握三种推导技巧(E6.2、E6.3)和两种应用技巧(E6.4、E6.6)。洛必达法则(E6.5)作为 E6.6 的经典替代,并列处理。
Memorise the five standard series cold. Practise one substitution problem ($e^{-x^2}$) and one $0/0$ limit problem ($\lim_{x \to 0} (1 - \cos x)/x^2$). Memorise the general Maclaurin formula $f(x) = \sum f^{(n)}(0) x^n / n!$ in case the question asks you to derive a series the standard list does not cover.
把五个标准级数背得滚瓜烂熟。完整做一道代换题($e^{-x^2}$)和一道 $0/0$ 极限题($\lim_{x \to 0} (1 - \cos x)/x^2$)。背熟通项公式 $f(x) = \sum f^{(n)}(0) x^n / n!$,以备题目要你推导标准表外的级数。
For each standard series, know its radius of convergence and the conditions on $x$ for the equality to hold. For L'Hopital, justify the indeterminate form ($0/0$ or $\infty/\infty$) before differentiating top and bottom. For limits via series, keep enough terms so that the leading non-zero behaviour survives after cancellation.
每个标准级数都要记住收敛半径与等式成立的 $x$ 条件。用洛必达前,必须先论证不定型($0/0$ 或 $\infty/\infty$)。用级数求极限时,保留的项数要足够,使主导的非零行为在抵消后仍然存在。
Standard Maclaurin Series标准麦克劳林级数 HL AHL 5.18
The general Maclaurin formula. If $f$ has derivatives of all orders at $0$ and the series converges to $f(x)$ on some interval around $0$, then $$ f(x) \;=\; \sum_{n=0}^{\infty} \frac{f^{(n)}(0)}{n!} x^{n} \;=\; f(0) + f'(0) x + \tfrac{f''(0)}{2!} x^{2} + \tfrac{f'''(0)}{3!} x^{3} + \cdots. $$
通项公式(麦克劳林公式)。若 $f$ 在 $0$ 处各阶可导,且级数在 $0$ 的某邻域内收敛于 $f(x)$,则 $$ f(x) \;=\; \sum_{n=0}^{\infty} \frac{f^{(n)}(0)}{n!} x^{n} \;=\; f(0) + f'(0) x + \tfrac{f''(0)}{2!} x^{2} + \tfrac{f'''(0)}{3!} x^{3} + \cdots。 $$
Derive the first four terms of the Maclaurin series for $f(x) = e^{x}$ directly from the general formula.用通项公式直接推导 $f(x) = e^{x}$ 麦克劳林级数的前四项。
Compute derivatives at $0$. Because $\tfrac{d}{dx} e^{x} = e^{x}$, every derivative of $f$ is $e^{x}$, so $f^{(n)}(0) = e^{0} = 1$ for all $n \ge 0$.
计算各阶导数在 $0$ 处的值。由于 $\tfrac{d}{dx} e^{x} = e^{x}$,故 $f$ 的所有阶导数都是 $e^{x}$,从而对所有 $n \ge 0$ 都有 $f^{(n)}(0) = e^{0} = 1$。
Substitute into the Maclaurin formula.
代入麦克劳林公式。
$$ e^{x} \;=\; \sum_{n=0}^{\infty} \frac{f^{(n)}(0)}{n!} x^{n} \;=\; \sum_{n=0}^{\infty} \frac{x^{n}}{n!}. $$First four terms.
前四项。
$$ e^{x} \;=\; 1 + x + \frac{x^{2}}{2!} + \frac{x^{3}}{3!} + \cdots \;=\; 1 + x + \frac{x^{2}}{2} + \frac{x^{3}}{6} + \cdots. $$Remark. The series converges for all real $x$ because the ratio of successive terms $|x|/(n+1) \to 0$, regardless of $x$.
注。该级数对全体实 $x$ 收敛,因为相邻两项之比 $|x|/(n+1) \to 0$(与 $x$ 无关)。
Going deeper: why the formula has this shape深入探讨:为何公式呈此形式
Suppose $f(x) = \sum_{n=0}^{\infty} a_{n} x^{n}$ converges in a neighbourhood of $0$. Differentiate term by term $k$ times:
设 $f(x) = \sum_{n=0}^{\infty} a_{n} x^{n}$ 在 $0$ 的邻域内收敛。逐项求导 $k$ 次:
$$ f^{(k)}(x) \;=\; \sum_{n=k}^{\infty} a_{n} \cdot n (n-1) \cdots (n - k + 1) \cdot x^{n - k}. $$Evaluate at $x = 0$: every term with $n > k$ vanishes, and the term with $n = k$ leaves $a_{k} \cdot k!$. Hence $f^{(k)}(0) = k! \, a_{k}$, which gives $a_{k} = f^{(k)}(0) / k!$. This is the Maclaurin coefficient.
在 $x = 0$ 处求值:$n > k$ 的项全部为零,$n = k$ 的项留下 $a_{k} \cdot k!$。故 $f^{(k)}(0) = k! \, a_{k}$,即 $a_{k} = f^{(k)}(0) / k!$。此即麦克劳林系数。
Caveat. A function can have all derivatives at $0$ yet the resulting series can have radius of convergence zero or can converge to a different function. The standard catalogue of five series is exempt from this pathology: each equals its Maclaurin series on the stated interval.
警告。一个函数虽各阶可导,其麦克劳林级数的收敛半径可能为零,或收敛到别的函数。标准五式不存在此病态:在所述区间上每个函数与其麦克劳林级数恒等。
Series by Substitution into Known Series用代换法导出新级数 HL AHL 5.18
Standard substitution targets.
- $e^{u} \to e^{kx}$, $e^{-x}$, $e^{x^{2}}$, $e^{-x^{2}}$ by $u = kx$, $u = -x$, $u = x^{2}$, $u = -x^{2}$.
- $\sin u \to \sin (kx)$, $\sin (x^{2})$ similarly.
- $\dfrac{1}{1 - u} \to \dfrac{1}{1 + x^{2}}$ by $u = -x^{2}$, valid for $|x| < 1$.
典型代换。
- $e^{u} \to e^{kx}$、$e^{-x}$、$e^{x^{2}}$、$e^{-x^{2}}$:分别取 $u = kx$、$u = -x$、$u = x^{2}$、$u = -x^{2}$。
- $\sin u \to \sin (kx)$、$\sin (x^{2})$ 同法。
- $\dfrac{1}{1 - u} \to \dfrac{1}{1 + x^{2}}$:取 $u = -x^{2}$,在 $|x| < 1$ 上成立。
Find the Maclaurin series for $e^{-x^{2}}$. Write out the first four non-zero terms and the general term.求 $e^{-x^{2}}$ 的麦克劳林级数。写出前四个非零项与通项。
Start from the standard $e^{u}$ series.
由标准 $e^{u}$ 级数出发。
$$ e^{u} \;=\; \sum_{n=0}^{\infty} \frac{u^{n}}{n!}, \qquad \text{all real } u. $$Substitute $u = -x^{2}$. This is valid for all real $x$ because $-x^{2}$ ranges over all real numbers as $x$ does.
代入 $u = -x^{2}$。当 $x$ 取遍实数时 $-x^{2}$ 亦遍历实数,故对全体实 $x$ 成立。
$$ e^{-x^{2}} \;=\; \sum_{n=0}^{\infty} \frac{(-x^{2})^{n}}{n!} \;=\; \sum_{n=0}^{\infty} \frac{(-1)^{n} x^{2n}}{n!}. $$First four non-zero terms.
前四个非零项。
$$ e^{-x^{2}} \;=\; 1 - x^{2} + \frac{x^{4}}{2!} - \frac{x^{6}}{3!} + \cdots \;=\; 1 - x^{2} + \frac{x^{4}}{2} - \frac{x^{6}}{6} + \cdots. $$Remark. The function $e^{-x^{2}}$ has no elementary antiderivative; the series form is precisely how the error function $\operatorname{erf}(x) = \tfrac{2}{\sqrt{\pi}} \int_{0}^{x} e^{-t^{2}} \, dt$ is computed.
注。$e^{-x^{2}}$ 没有初等原函数;误差函数 $\operatorname{erf}(x) = \tfrac{2}{\sqrt{\pi}} \int_{0}^{x} e^{-t^{2}} \, dt$ 正是通过此级数计算的。
Series by Differentiation and Integration用求导与积分导出级数 HL AHL 5.18
The geometric series anchor. $$ \frac{1}{1 - x} \;=\; \sum_{n=0}^{\infty} x^{n} \;=\; 1 + x + x^{2} + x^{3} + \cdots, \qquad |x| < 1. $$ Differentiating or integrating this anchor (and its substitution variants) produces a whole family of series: $\dfrac{1}{(1-x)^{2}}$, $-\ln(1-x)$, $\arctan x$, $\arcsin x$.
几何级数为根基。 $$ \frac{1}{1 - x} \;=\; \sum_{n=0}^{\infty} x^{n} \;=\; 1 + x + x^{2} + x^{3} + \cdots, \qquad |x| < 1。 $$ 对此根基(及其代换变体)求导或积分可产生一整族级数:$\dfrac{1}{(1-x)^{2}}$、$-\ln(1-x)$、$\arctan x$、$\arcsin x$。
Use the geometric series for $1/(1 - x)$ to derive the Maclaurin series for $-\ln(1 - x)$, and state the interval of validity.用 $1/(1 - x)$ 的几何级数推导 $-\ln(1 - x)$ 的麦克劳林级数,并指出有效区间。
Start with the geometric series.
从几何级数出发。
$$ \frac{1}{1 - x} \;=\; \sum_{n=0}^{\infty} x^{n}, \qquad |x| < 1. $$Integrate both sides from $0$ to $x$. The antiderivative of $1/(1 - x)$ with respect to $x$ is $-\ln|1 - x|$, and $-\ln|1 - 0| = 0$, so
两边在 $0$ 到 $x$ 上积分。$1/(1 - x)$ 关于 $x$ 的原函数为 $-\ln|1 - x|$,且 $-\ln|1 - 0| = 0$,故
$$ -\ln(1 - x) \;=\; \int_{0}^{x} \frac{dt}{1 - t} \;=\; \int_{0}^{x} \sum_{n=0}^{\infty} t^{n} \, dt \;=\; \sum_{n=0}^{\infty} \frac{x^{n+1}}{n + 1}. $$Rewrite by shifting the index. Let $m = n + 1$:
换指标改写。令 $m = n + 1$:
$$ -\ln(1 - x) \;=\; \sum_{m=1}^{\infty} \frac{x^{m}}{m} \;=\; x + \frac{x^{2}}{2} + \frac{x^{3}}{3} + \frac{x^{4}}{4} + \cdots. $$Interval. Term-by-term integration preserves the radius of convergence ($|x| < 1$). At the endpoints $x = -1$ and $x = 1$: at $x = -1$, the series becomes $\sum (-1)^{m}/m$, which converges (alternating); at $x = 1$, the series becomes $\sum 1/m$, the harmonic series, which diverges. So the equality holds for $-1 \le x < 1$.
区间。逐项积分保持收敛半径($|x| < 1$)。端点:$x = -1$ 时级数为 $\sum (-1)^{m}/m$,交错收敛;$x = 1$ 时级数为 $\sum 1/m$(调和级数),发散。故等式在 $-1 \le x < 1$ 上成立。
Consistency check. Replacing $x$ with $-x$ recovers the standard series $\ln(1 + x) = \sum (-1)^{n+1} x^{n} / n$ on $-1 < x \le 1$, exactly as listed in E6.1.
一致性检验。把 $x$ 换为 $-x$ 即可还原 E6.1 列出的标准级数 $\ln(1 + x) = \sum (-1)^{n+1} x^{n} / n$($-1 < x \le 1$)。
Going deeper: $\arctan x$ from $1/(1 + x^{2})$深入探讨:由 $1/(1 + x^{2})$ 推 $\arctan x$
Substitute $u = -x^{2}$ into the geometric series:
把 $u = -x^{2}$ 代入几何级数:
$$ \frac{1}{1 + x^{2}} \;=\; \frac{1}{1 - (-x^{2})} \;=\; \sum_{n=0}^{\infty} (-x^{2})^{n} \;=\; \sum_{n=0}^{\infty} (-1)^{n} x^{2n}, \qquad |x| < 1. $$Integrate from $0$ to $x$. Since $\arctan(0) = 0$ and $\tfrac{d}{dx} \arctan x = 1/(1 + x^{2})$:
在 $0$ 到 $x$ 上积分。由 $\arctan(0) = 0$、$\tfrac{d}{dx} \arctan x = 1/(1 + x^{2})$:
$$ \arctan x \;=\; \sum_{n=0}^{\infty} \frac{(-1)^{n} x^{2n+1}}{2n + 1} \;=\; x - \frac{x^{3}}{3} + \frac{x^{5}}{5} - \frac{x^{7}}{7} + \cdots, \qquad |x| \le 1. $$At $x = 1$, this is the famous Gregory-Leibniz formula $\pi/4 = 1 - 1/3 + 1/5 - \cdots$. The endpoint $x = 1$ is included because the alternating series converges there (Abel's theorem).
$x = 1$ 时即著名的格雷戈里-莱布尼茨公式 $\pi/4 = 1 - 1/3 + 1/5 - \cdots$。端点 $x = 1$ 可纳入,因交错级数在该处收敛(阿贝尔定理)。
Approximation and Truncation Error逼近与截断误差 HL AHL 5.18
The dominant-term rule. For $|x|$ small, the error is dominated by the first omitted term, $\tfrac{f^{(N+1)}(0)}{(N+1)!} x^{N+1}$. As $|x|$ grows you need more terms to keep the error small. As $N$ grows (at fixed $x$ inside the radius of convergence) the error tends to zero.
Alternating series bound. If the series is alternating with decreasing absolute terms, the truncation error is at most the absolute value of the first omitted term. This applies cleanly to $\sin x$, $\cos x$, $\ln(1 + x)$ on their convergence intervals.
主导项原则。$|x|$ 较小时,误差由首个被舍去的项 $\tfrac{f^{(N+1)}(0)}{(N+1)!} x^{N+1}$ 主导。$|x|$ 越大,需保留项数越多。固定 $x$(在收敛半径内)时,$N$ 越大误差越小。
交错级数界。若级数为绝对值递减的交错级数,截断误差不超过首个被舍弃项的绝对值。$\sin x$、$\cos x$、$\ln(1 + x)$ 在其收敛区间上正满足该条件。
Approximate $e^{0.1}$ using the first four terms of the Maclaurin series for $e^{x}$. Compare with the true value.用 $e^{x}$ 麦克劳林级数的前四项近似 $e^{0.1}$,并与真值比较。
Write the first four terms.
写出前四项。
$$ e^{x} \;\approx\; 1 + x + \frac{x^{2}}{2!} + \frac{x^{3}}{3!} \;=\; P_{3}(x). $$Substitute $x = 0.1$.
代入 $x = 0.1$。
$$ P_{3}(0.1) \;=\; 1 + 0.1 + \frac{0.01}{2} + \frac{0.001}{6} \;=\; 1 + 0.1 + 0.005 + 0.000\,166\,6\overline{6}\ldots \;\approx\; 1.105\,167. $$True value (calculator). $e^{0.1} = 1.105\,170\,918\ldots$. The approximation matches to five decimal places.
真值(计算器)。$e^{0.1} = 1.105\,170\,918\ldots$。近似值与真值前五位小数完全吻合。
Error estimate via the next term. The first omitted term is $\tfrac{x^{4}}{4!} = \tfrac{(0.1)^{4}}{24} = \tfrac{10^{-4}}{24} \approx 4.17 \times 10^{-6}$. This matches the observed error $1.105\,170\,918 - 1.105\,166\,667 \approx 4.25 \times 10^{-6}$ to a fraction of a percent. The dominant-term rule works exactly as advertised.
用下一项估计误差。首个被舍弃项为 $\tfrac{x^{4}}{4!} = \tfrac{(0.1)^{4}}{24} = \tfrac{10^{-4}}{24} \approx 4.17 \times 10^{-6}$。与实际误差 $1.105\,170\,918 - 1.105\,166\,667 \approx 4.25 \times 10^{-6}$ 差不到一个百分点。主导项规则名副其实。
L'Hopital's Rule洛必达法则 HL AHL 5.18
Use the rule when:
- Direct substitution gives $0/0$ or $\infty / \infty$.
- The new ratio $f'/g'$ is easier than the original.
Other indeterminate forms. $0 \cdot \infty$ reshape to $0/0$ or $\infty / \infty$ by rewriting; $\infty - \infty$ combines to a single fraction; $1^{\infty}$, $0^{0}$, $\infty^{0}$ take logarithm first.
何时使用:
- 直接代入得 $0/0$ 或 $\infty / \infty$。
- 新的比 $f'/g'$ 比原比简单。
其他不定型。$0 \cdot \infty$ 改写为 $0/0$ 或 $\infty / \infty$;$\infty - \infty$ 合并为单一分式;$1^{\infty}$、$0^{0}$、$\infty^{0}$ 先取对数。
Evaluate $\lim_{x \to 0} \dfrac{\sin x}{x}$ using L'Hopital's rule.用洛必达法则求 $\lim_{x \to 0} \dfrac{\sin x}{x}$。
Check the indeterminate form. As $x \to 0$, the numerator $\sin x \to 0$ and the denominator $x \to 0$, so the limit has form $0/0$.
检查不定型。$x \to 0$ 时,分子 $\sin x \to 0$、分母 $x \to 0$,故极限为 $0/0$。
Differentiate numerator and denominator separately.
分子分母分别求导。
$$ \lim_{x \to 0} \frac{\sin x}{x} \;=\; \lim_{x \to 0} \frac{(\sin x)'}{(x)'} \;=\; \lim_{x \to 0} \frac{\cos x}{1} \;=\; \cos 0 \;=\; 1. $$Result. $\lim_{x \to 0} \sin x / x = 1$.
结果。$\lim_{x \to 0} \sin x / x = 1$。
Caveat. Strictly, L'Hopital's rule presupposes the derivative of $\sin x$ is $\cos x$, whose standard proof uses precisely $\lim \sin x / x = 1$. So this is a circular argument as a foundational proof; treat it instead as a quick recheck once the derivative is known. The series approach in E6.6 avoids the circularity.
警告。严格来说,洛必达需要 $\sin x$ 的导数 $\cos x$,而该导数的标准证明恰用 $\lim \sin x / x = 1$。故此论证作为基础证明是循环的;只能视为已知导数后的快速复核。E6.6 的级数法可避免循环。
Going deeper: a repeated-application example深入探讨:反复使用的例子
Evaluate $\lim_{x \to 0} \dfrac{1 - \cos x}{x^{2}}$. As $x \to 0$, both numerator and denominator approach $0$. Apply L'Hopital once:
求 $\lim_{x \to 0} \dfrac{1 - \cos x}{x^{2}}$。$x \to 0$ 时分子分母同趋于 $0$。第一次用洛必达:
$$ \lim_{x \to 0} \frac{1 - \cos x}{x^{2}} \;=\; \lim_{x \to 0} \frac{\sin x}{2 x}. $$This is still $0/0$. Apply L'Hopital again:
仍为 $0/0$。再用一次:
$$ \lim_{x \to 0} \frac{\sin x}{2 x} \;=\; \lim_{x \to 0} \frac{\cos x}{2} \;=\; \frac{1}{2}. $$Compare with the series approach in E6.6, which arrives at the same answer in one step.
将此与 E6.6 的级数法比较:后者一步即得同一结果。
Limits via Maclaurin Series用麦克劳林级数求极限 HL AHL 5.18
- Substitute the Maclaurin series for the numerator and denominator (or for whichever factors are causing the indeterminacy).
- Identify the leading non-zero powers of $x$ in numerator and denominator.
- Divide numerator and denominator by the lowest common power of $x$ (so the leading term in each becomes a constant).
- Take $x \to 0$. Only the constant terms survive.
When the series method beats L'Hopital. Series shine when L'Hopital requires multiple applications (e.g. $\lim (1 - \cos x)/x^{2}$ needs two L'Hopital passes; one series substitution gives the answer immediately).
- 用麦克劳林级数代换分子分母(或制造不定的部分)。
- 找出分子分母中首个非零的 $x$ 次方。
- 分子分母同除以最低公共 $x$ 次方,使各自首项化为常数。
- 令 $x \to 0$,只剩常数项。
级数法何时胜过洛必达。洛必达需多次重复时级数法显优。例如 $\lim (1 - \cos x)/x^{2}$ 需洛必达两次,而代入级数一次即得。
Evaluate $\lim_{x \to 0} \dfrac{1 - \cos x}{x^{2}}$ by substituting the Maclaurin series for $\cos x$.用 $\cos x$ 的麦克劳林级数代入求 $\lim_{x \to 0} \dfrac{1 - \cos x}{x^{2}}$。
Substitute the standard series for $\cos x$.
代入 $\cos x$ 的标准级数。
$$ \cos x \;=\; 1 - \frac{x^{2}}{2!} + \frac{x^{4}}{4!} - \frac{x^{6}}{6!} + \cdots. $$Form the numerator.
构造分子。
$$ 1 - \cos x \;=\; 1 - \Bigl[ 1 - \frac{x^{2}}{2} + \frac{x^{4}}{24} - \cdots \Bigr] \;=\; \frac{x^{2}}{2} - \frac{x^{4}}{24} + \frac{x^{6}}{720} - \cdots. $$Divide by $x^{2}$.
除以 $x^{2}$。
$$ \frac{1 - \cos x}{x^{2}} \;=\; \frac{1}{2} - \frac{x^{2}}{24} + \frac{x^{4}}{720} - \cdots. $$Take the limit. As $x \to 0$, every term with a positive power of $x$ vanishes:
取极限。$x \to 0$ 时,含正幂的项全部消失:
$$ \lim_{x \to 0} \frac{1 - \cos x}{x^{2}} \;=\; \frac{1}{2}. $$Comparison. The same answer comes from L'Hopital after two applications (see the "Going deeper" block in E6.5). The series method gets there in one substitution, which is why it dominates in HL exam problems with iterated indeterminacy.
对照。洛必达两次也得同一答案(见 E6.5 的"深入探讨")。级数法一次代入即可,因此在多次不定型反复出现的 HL 试题中更占优。
Going deeper: how many terms is enough?深入探讨:保留多少项才够?
Suppose you keep only $1 - x^{2}/2$ for $\cos x$. Then $1 - \cos x \approx x^{2}/2$ exactly, and $(1 - \cos x)/x^{2} \to 1/2$. The single dropped term $x^{4}/24$ would, after dividing by $x^{2}$, become $x^{2}/24$, which vanishes as $x \to 0$. So two terms of $\cos x$ are already enough for this problem.
若 $\cos x$ 只保留 $1 - x^{2}/2$,则 $1 - \cos x \approx x^{2}/2$,从而 $(1 - \cos x)/x^{2} \to 1/2$。被舍弃的 $x^{4}/24$ 经除以 $x^{2}$ 后变为 $x^{2}/24$,$x \to 0$ 时为零。故 $\cos x$ 保留两项已够此题。
General principle. Keep enough terms so the leading non-zero behaviour in both numerator and denominator is captured. If a question asks for $\lim (\sin x - x)/x^{3}$, the $\sin x = x - x^{3}/6 + \cdots$ truncated at three terms gives $\sin x - x \approx -x^{3}/6$, hence the limit is $-1/6$. Two terms of $\sin x$ would have given $\sin x - x \approx 0$, which is too coarse.
一般原则。保留到分子分母首个非零行为都能体现为止。若问 $\lim (\sin x - x)/x^{3}$:$\sin x = x - x^{3}/6 + \cdots$ 保留三项给 $\sin x - x \approx -x^{3}/6$,极限为 $-1/6$。只保留两项则 $\sin x - x \approx 0$,太粗。
Exam Strategy and Common Pitfalls考试策略与常见陷阱
- Write the series, then the validity. Examiners reward both. "$\sin x = x - x^{3}/3! + \cdots$ for all real $x$" earns the full A1, whereas only the formula loses the validity mark.
- 写级数,也写收敛区间。两者都给分。"$\sin x = x - x^{3}/3! + \cdots$(全体实 $x$)"得满 A1;只写公式漏掉区间要扣分。
- The signs in $\sin x$ and $\cos x$ alternate; check parity. $\sin$ is odd (only odd powers) and $\cos$ is even (only even powers). A series with both odd and even powers cannot be $\sin x$ or $\cos x$.
- $\sin x$ 与 $\cos x$ 符号交替;检查奇偶性。$\sin$ 为奇函数(只含奇次幂)、$\cos$ 为偶函数(只含偶次幂)。同时含奇偶次幂的级数不可能是 $\sin x$ 或 $\cos x$。
- Translate the validity interval through the substitution. $\ln(1 + u)$ valid on $-1 < u \le 1$ becomes $\ln(1 + 2x)$ valid on $-1/2 < x \le 1/2$, not $-1 < x \le 1$.
- 把收敛区间同步翻译。$\ln(1 + u)$ 在 $-1 < u \le 1$ 有效,故 $\ln(1 + 2x)$ 在 $-1/2 < x \le 1/2$ 有效,不是 $-1 < x \le 1$。
- Integration introduces a constant of integration. Fix it by evaluating both sides at $x = 0$. For example, integrating $1/(1 - x) = \sum x^{n}$ gives $-\ln(1 - x) = \sum x^{n+1}/(n+1) + C$, and $x = 0$ forces $C = 0$.
- 积分要引入常数。用 $x = 0$ 两边对照定常数。如积分 $1/(1 - x) = \sum x^{n}$ 得 $-\ln(1 - x) = \sum x^{n+1}/(n+1) + C$,$x = 0$ 给 $C = 0$。
- Verify the indeterminate form first. L'Hopital applies only to $0/0$ or $\infty/\infty$. State the form before differentiating; the M1 is awarded for the verification step.
- 必须先验不定型。洛必达只用于 $0/0$ 或 $\infty/\infty$。求导前先写出形式;M1 即给在该验证步骤。
- Pick the right tool. One-step indeterminacies favour L'Hopital; multi-step indeterminacies favour Maclaurin series. For exam answers worth both methods, write the series approach when iterating L'Hopital looks tedious.
- 选对工具。单次不定型用洛必达;多次反复用级数。当洛必达需多次而显繁琐时,改用级数法作答。
- Keep enough terms in the series. Underkeeping gives $0/0$ again; overkeeping wastes algebra. The rule is "the lowest power that survives cancellation".
- 级数保留足够项。少了得 $0/0$,多了浪费代数。规则:保留至抵消后仍存在的最低幂为止。
Flashcards闪卡
Unit E6 Practice Quiz单元 E6 练习测验
Readiness Checklist备考清单
Tick each item when you can do it cold, without notes, on your first attempt.
每一条都要"裸做"做对(不看笔记、一次过)才打勾。
- HL Write the Maclaurin series for $e^{x}$, $\sin x$, $\cos x$ with the correct general term and validity (all real $x$)写出 $e^{x}$、$\sin x$、$\cos x$ 的麦克劳林级数,含正确通项与收敛域(全体实 $x$)
- HL Write the Maclaurin series for $\ln(1 + x)$ with validity $-1 < x \le 1$写出 $\ln(1 + x)$ 的麦克劳林级数,含 $-1 < x \le 1$ 收敛域
- HL Write the binomial series for $(1 + x)^{n}$ with validity $|x| < 1$ when $n$ is non-integer写出 $(1 + x)^{n}$ 的二项级数,$n$ 非整数时 $|x| < 1$
- HL Apply the general Maclaurin formula $f(x) = \sum f^{(n)}(0) x^{n}/n!$ to derive a series from scratch用通项公式 $f(x) = \sum f^{(n)}(0) x^{n}/n!$ 从零推导级数
- HL Derive a series by substitution (e.g. $e^{-x^{2}}$, $\sin (2x)$), with validity translated through the substitution用代换法导出级数(如 $e^{-x^{2}}$、$\sin (2x)$),收敛域同步翻译
- HL Differentiate a power series term by term inside its radius of convergence在收敛半径内逐项求导幂级数
- HL Integrate a power series term by term, fixing the constant of integration by evaluating at $x = 0$逐项积分幂级数,用 $x = 0$ 定积分常数
- HL Truncate a Maclaurin series to compute a numerical approximation and estimate the error from the first omitted term截断麦克劳林级数算数值近似,并用首个被舍弃项估计误差
- HL State and apply L'Hopital's rule, verifying the indeterminate form first陈述并应用洛必达法则,事先验证不定型
- HL Apply L'Hopital's rule repeatedly when the new ratio is still indeterminate当新比仍为不定型时反复使用洛必达
- HL Evaluate a $0/0$ limit by substituting Maclaurin series for numerator and denominator通过代入分子分母的麦克劳林级数算 $0/0$ 极限
- HL Decide when the series method beats L'Hopital (multi-step indeterminacies)判断何时级数法胜过洛必达(多次不定型)
- HL Keep enough series terms so the leading non-zero behaviour survives cancellation in the limit保留足够级数项使主导非零行为在极限中存留
IB Paper-Style PracticeIB 试卷风格练习
E6 Practice and Solutions are on the roadmap. They will ship under Practice Questions/Unit_E6_*.html with the bilingual built-in pattern.
E6 配套的 Practice 与 Solutions 已在排期,上线后位于 Practice Questions/Unit_E6_*.html,采用双语内嵌格式。