极限可以说是高等数学的基础,其无限逼近但是不可达的样式奠定了微积分。因此学好极限乃重中之重。
定义
$\underset{n\rightarrow \textit{∞}}{\lim}x_n=A$定义为$\forall \varepsilon >0,\ \exists N>0$,当$n>N$时,恒有$|x_n-A|<\varepsilon$
证明:$\underset{n\rightarrow \textit{∞ }}{\lim}f\left( x \right) =M$
对$\forall \varepsilon >0$,$|f(x)-M|< \varepsilon $,等价于$x > \psi (\varepsilon )$
取$N=[\psi (\varepsilon )]+1$,则对$\forall \varepsilon >0$,当$x>N$时,有$|f(x)-M|<\varepsilon $
由极限定义可得$\underset{n\rightarrow \textit{∞ }}{\lim}f\left( x \right) =M$
极限性质
一般性质
唯一性:极限必定唯一
保号性: 设$\underset{n\rightarrow \textit{∞ }}{\lim}f\left( x \right) =A\left\{ \begin{array}{l}>0\\ <0\\ \end{array} \right. $,则$\exists \delta>0$,当$0<|x-a|<\delta$时,$$f\left( x \right) \left\{ \begin{array}{l} >0\\ <0\\ \end{array} \right. $$
推论:若${a_n}$从某项起$a_n≥0$且$\underset{n\rightarrow∞}{\lim}a_n=a$,则$a≥0$
第一保号性:极限正,去心领域正;极限负,去心领域负
第二保号性:函数不负,极限不负;函数不正,极限不正——$f(x)\ge 0 \Rightarrow \lim f(x)\ge0$
$f(x)>0$,且$\lim f(x)>0$不一定(可$0$)第三保号性:函数大小次序与极限大小次序一致——$g(x)\ge g(x)\Rightarrow \lim f(x)\ge \lim g(x)$
有界性:
- $\underset{n\rightarrow \textit{∞ }}{\lim}a_n =A \Rightarrow \exists M>0, s.t. |a_n|<M$
反之不成立 - 若$f’(x)$在有限区间有界,则$f(x)$在该区间有界
- 若$\underset{x \rightarrow a^+}{\lim}f(x)$存在,$\underset{x\rightarrow b^-}{\lim}f(x)$存在,$f(x)$在$(a,b)$连续,则$f(x)$在$(a,b)$有界
部分性: $\underset{n \rightarrow ∞}{\lim }x_n=a \Leftrightarrow \underset{n \rightarrow ∞}{\lim }x_{2k-1}=\underset{n \rightarrow ∞}{\lim }x_{2k}=a$
海涅定理(归结原则):$\underset{x\rightarrow a}\lim f(x)=b\Leftrightarrow$取$f(x)$定义域内的任意数列${a_n}$,$\underset{n\rightarrow\infty}\lim a_n=a$,且$a_n\ne a$,有$\underset{n\rightarrow \infty }\lim f(a_n)=b$
极限存在性质
夹逼定理
$\left\{ \begin{array}{l} a_n\le b_n\le c_n\\ \underset{n\rightarrow \infty \,\,}{\lim}a_n=\underset{n\rightarrow \infty \,\,}{\lim}c_n=A\\ \end{array} \right. \Rightarrow \underset{n\rightarrow \infty \,\,}{\lim}b_n=A$夹逼取函数:$\sum \min f(x)\le \sum f(x)\le \sum\max f(x)$
基本型(凑成$\dfrac{i}{n}$)
- $an+bi=n(a+b\dfrac{i}{n})$
- $n^2+i^2=n^2[1+(\dfrac{i}{n})^2]$
- $n^2+ni=n^2(1+\dfrac{i}{n})$
- $\dfrac{i}{n}$
放缩型(凑不成$\dfrac{i}{n}$)
夹逼准则($n^2+i$)
放缩后再凑 $ \dfrac{i}{n} $
$(\dfrac{i}{n})^2\le \dfrac{i^2+1}{n^2} \le (\dfrac{i+1}{n})^2$
变量型
$$\underset{n \rightarrow ∞}{\lim}\sum_{i=1}^nf(a+\dfrac{b-a}{n}i)\dfrac{b-a}{n}=\int _a^bf(x)\text{d}x$$
放缩步骤:
- 写成求和形式
- 从通式中找出影响最小的因素($i$、常数)
- 放缩最小因素(一般化为0,最小值,最大值)
单调有界数列必有极限
单调
- $a_{n+1}-a_n\left\{ \begin{array}{l} \ge 0\text{单调递增}\\ \le 0\text{单调递减}\\ \end{array} \right. $
- $f\left( n \right) =a_n$ $f'\left( x \right) \left\{ \begin{array}{l}>0\text{单调递增}\\ <0\text{单调递减}\\ \end{array} \right. $
$a_{n+1}=f\left( a_n \right) $
$f'\left( x \right) \left\{ \begin{array}{l}>0\Rightarrow \left\{ \begin{array}{l}a_2>a_1\text{单调递增}\\a_2 < a_1\text{单调递减}\\\end{array} \right.\\<0\Rightarrow \text{不单调}\\\end{array} \right. $
有界
归纳法
- 当$n=1$时,证明$a_1$有界
- 当$n=k$时,假设$a_k$有界
- 当$n=k+1$时,证明$a_{k+1}$有界
无穷性质
$\text o(x^m)+\text o(x^n)=\text o(x^{min{m,n}})$
$\text o(x^m) \cdot \text o(x^n)=\text o(x^{m+n})$
无穷大
- $\left. \begin{array}{r} A\rightarrow \infty\\ B\rightarrow \infty\\ \end{array} \right\} \Rightarrow AB\rightarrow \infty $
- $\left. \begin{array}{r} A\text{无界}\\ B\rightarrow \infty\\ \end{array} \right\} \nRightarrow AB\rightarrow \infty $ 反例:$a_n=n^2\ \ b_n=\left\{ \begin{array}{l} n\,\,\,\,n=1,3,5,\cdots\\ 0\,\,\,\,n=2,4,6,\cdots\\ \end{array} \right. $
无穷小的比阶
- $f(x)\sim ax^k$
- $\underset{x\rightarrow 0}{\lim}\dfrac{f(x)}{x^k}=c$(常用洛必达)
- 若$f(x)=a_0+a_1x+…+a_kx^k+a_{k+1}x^{k+1}$(低次方说了算)
- $x\rightarrow 0,\left. \begin{array}{r} f\left( x \right) \sim x^m\\ g\left( x \right) \sim x^n\\ \end{array} \right\} \Rightarrow \int_0^{g\left( x \right)}{f\left( t \right)}\text{d}t\sim x^{\left( m+1 \right) n}$
极限计算
等价无穷小替换
- $x\sim\sin x\sim\tan x\\ \ \sim\arcsin x\sim\arctan x\\ \ \sim\text e^x-1\sim\ln(1+x)$
- $1-\cos x\sim \dfrac12x^2$
- $1-\cos ^\alpha x\sim \dfrac{\alpha}{2}x^2$
- $(1+x)^a-1\sim ax$
- $(1+f(x))^{g(x)}-1\sim f(x)g(x)$
- $a^x-1\sim x\ln a$
- $x-\ln(1+x)\sim \dfrac12x^2$
- $\tan x-\sin x\sim\dfrac12x^3$
- $\tan x-x\sim x-\arctan x\sim\dfrac13 x^3$
- $x-\sin x\sim \arcsin x-x\sim\dfrac 16x^3$
- $x+\sin x \sim 2x$
- $\text e^x-x-1$ ~ $\dfrac{1}{2}x^2$
- $\ln(x+\sqrt{1+x^2})\sim x$
- ${\lim}\dfrac{\beta}{\alpha}=0,$则$\alpha+\beta\sim\alpha$
- $\ln \dfrac{\sin x}{x} \sim -\dfrac{1}{6}x^2$
- $1-\sqrt{\cos x}\sim \dfrac{x^2}4$
泰勒公式
$$\text e^x=1+x+\dfrac{x^2}{2!}+…+\dfrac{x^n}{n!}$$
$$\ln (1+x)=x-\dfrac{1}{2}x^2+…+(-1)^n\dfrac{x^n}{n}$$
$$(1+x)^{\alpha}=1+\alpha x+\dfrac{\alpha(\alpha-1)}{2}x^2+o(x^2)$$
$$\sqrt{1+x}=1+\dfrac x2-\dfrac{x^2}8+\dfrac{x^3}{16}+o(x^3)$$
$$\sin x=x-\dfrac{1}{3!}x^3+…+(-1)^n \dfrac{1}{(2n+1)!}x^{2n+1}$$
$$\cos x=1-\dfrac{1}{2!}x^2+…+(-1)^{n}\dfrac{1}{(2n)!}x^{2n}$$
$$\tan x=x+\dfrac{1}{3}x^3+o(x^3)$$
$$\arcsin x=x+\dfrac{1}{6}x^3+o(x^3)$$
$$\arctan x=x-\dfrac{1}{3}x^3+o(x^3)$$
$$\dfrac{1}{1-x}=1+x+x^2+…+x^n$$
$$\dfrac{1}{1+x}=1-x+x^2+…+(-1)^nx^n$$
$$(1+x)^{\frac1x}=\text e(1-\dfrac12x+\dfrac{11}{24}x^2-\dfrac{7}{16}x^3+\text o)$$
常用基本极限
- $\underset{x\rightarrow0}{\lim}{\dfrac{\text{sin}x}{x}} =1$
- $\underset{x\rightarrow {0}}{\lim}{(1+f(x))^{g(x)}} =e^{\underset{x\rightarrow {0}}\lim f(x)g(x)}$
- $\underset{n\rightarrow ∞}{\lim}\sqrt[n]{n}=1$
- $\underset{n\rightarrow ∞}{\lim}\sqrt[n]{a}=1$
- $ \underset{x\rightarrow \textit{∞ }}{\lim}\frac{b_nx^n+...+b_0}{a_mx^m+...+a_0}=\left\{ \begin{array}{l}0,\ \ \ n < m\\ \frac{b_n}{a_m},\ n=m\\ \textit{∞ },\ \ n>m\\ \end{array} \right.$
- $\underset{x\rightarrow 0}{\lim}\left( \dfrac{a_{1}^{x}+a_{2}^{x}\cdots +a_{n}^{x}}{n} \right) ^{\frac{1}{x}}= \left( a_1\cdot a_2\cdots a_n \right) ^{\frac{1}{n}}$
- $\underset{x\rightarrow 0^+}{\lim}x\ln x=0$
- $\underset{x\rightarrow 0^+}{\lim}x^x=1$
- $\underset{x\rightarrow +\infty}{\lim}x^{\frac{1}{x}}=1$
- $\underset{x\rightarrow 0^+}{\lim}\left( 1+x \right) ^{\frac{1}{x}}=\text{e}-\dfrac{\text{e}}{2}x+\dfrac{11\text{e}}{24}x^2$
- $\underset{n\rightarrow∞}{\lim}\sqrt[n]{a_1^n+a_2^n+...+a_m^n}=\max(a_i)$
不定型极限
$\dfrac{0}{0}$
- 等价无穷小代换
- 洛必塔
- 泰勒公式
- Δ→∇
- 分子√相减,有理化
$\dfrac{∞}{∞}$
- $ \underset{x\rightarrow \textit{∞ }}{\lim}\dfrac{b_nx^n+...+b_0}{a_mx^m+...+a_0}=\left\{ \begin{array}{l}
0,\ \ \ n
m\\ \end{array} \right.$ - 洛必塔
- 转化为$\dfrac{0}{0}$
$1^\infty$
- 凑$(1+x)^{\frac{1}{x}}$
- $\text{lim}(1+\alpha(x))^{\beta(x)}=\text e^{\alpha(x)\beta(x)}$
$\infty^0$
- $f(x)^{g(x)}=\text e^{g(x)\ln f(x)}$
$∞-∞$
- 有分母,通分
- 无分母,分子有理化或换元人为制造分母
- 当分子形似函数值相减时,中值定理
$0-0$
- $\Rightarrow \dfrac{0}{0}$
- $\Rightarrow \dfrac{∞}{∞}$
$0 \times ∞$
- $\Rightarrow \dfrac{0}{0}$
- $\Rightarrow \dfrac{∞}{∞}$
n项和求极限
先求和再求极限
夹逼定理
定积分定义
$$\underset{n\rightarrow \textit{∞ }}{\lim} \dfrac{1}{n} \sum_{i=n}^n{f(\dfrac{i}{n})}=\int_0^1{f\left( x \right) \text dx}$$
$\int_0^a f(x)\text d x=\dfrac an\underset{n\rightarrow \infty}{\lim}\sum_{i=1}^nf(\dfrac an i)$ $\int_0^1f(x)\text d x$$=\dfrac 1n \underset{n\rightarrow \infty}{\lim}\sum_{i=1}^nf(\dfrac in )\\=\dfrac 1n \underset{n\rightarrow \infty}{\lim}\sum_{i=1}^nf(\dfrac {i-1}n )$ $\underset{n\rightarrow \infty}{\lim}\sum f(a+\dfrac{b-a}n i)\dfrac{b-a}n\\=\underset{n\rightarrow \infty}{\lim}\sum f(a+\dfrac{b-a}n (i-1))\dfrac{b-a}n\\=\int_a^bf(x)\text dx$
渐近线
铅锤渐近线:$\lim f(x_0)=∞$
水平渐近线:$\lim f(x_0)=C$
斜渐近线:
$\underset{x \rightarrow ∞}{\lim}\dfrac{f(x)}{x}=k \\\underset{x \rightarrow ∞}{\lim}[f(x)-kx]=b$斜渐近线:$y=kx+b$
有水平渐近线没有斜渐近线
分离将$\dfrac{\arctan }{P_m(x)}$为$x(\arctan +a)$
连续与间断
连续
$f(x)在x=a$连续$\leftrightarrow$$\underset {x\rightarrow a}{\lim}{d(x)=f(a)}$
$f(x)$在$[a,b]$连续$\leftrightarrow \left\{ \begin{array}{l} f(x)\in \text C_{(a,b)} \\ f(a)= f(a+0)\\f(b)=f(b-0) \\ \end{array} \right. $
$f’(x)$在$(a,b)$有界$\Rightarrow$$f(x)$在$(a,b)$有界
间断
第一类
左右极限$f(a-0)f(a+0)$存在
可去间隔点
$f(a-0)=f(a+0)$
跳跃间断点
$f(a-0)≠f(a+0)$
**第二类 **
$f(a-0)、f(a+0)$至少一个不存在
无穷间断点
$\lim f(a)=∞$
振荡间断点
$\lim \sin{\dfrac{1}{x}}$
第一类间断点最好说清,第二类除了无穷与振荡还有其他的不必
判断间断点
$\underset{n\rightarrow \infty}{\lim}x^n=\left\{ \begin{array}{l} \infty \ \ \ \ |x|>1\\ 0\ \ \ \ \ \ |x|<1\\ 1\ \ \ \ \ \ \ x\ =1\\ \nexists \ \ \ \ \ \ \ \text{}x\ =-1\\ \end{array} \right. $$\underset{x\rightarrow \infty}{\lim}\text{e}^{x}=\left\{ \begin{array}{l} +\infty \ \ \ \ x>0\\ 0\ \ \ \ \ \ \ x<0\\ 1\ \ \ \ \ \ \ x=0\\ \end{array} \right. $
$\underset{x\rightarrow 0}{\lim}\text{e}^{\frac1x}=\left\{ \begin{array}{l} +\infty \ \ \ \ x\rightarrow0^+\\0\ \ \ \ \ \ \ \ \ x\rightarrow0^- \end{array} \right. $
$\underset{x\rightarrow 0}{\lim}a^{\frac{1}{x}}\left( a>1 \right) =\left\{ \begin{array}{l} +\infty \ \ x\rightarrow 0^+\\ 0\ \ \ \ \ \ \ x\rightarrow 0^-\\ \end{array} \right. $
$\underset{x\rightarrow 0}{\lim}\arctan \dfrac{1}{x}=\left\{ \begin{array}{l} \ \ \ \dfrac{\pi}{2} \ \ \ \ \ x\rightarrow 0^+\\ -\dfrac{\pi}{2}\ \ \ \ \ x\rightarrow 0^-\\ \end{array} \right. $
$\underset{x\rightarrow 0}{\lim}\text{arccot} \dfrac{1}{x}=\left\{ \begin{array}{l} 0\ \ x\rightarrow 0^+\\ \pi \ \ x\rightarrow 0^-\\ \end{array} \right. $
$\underset{x\rightarrow 0}{\lim}\left[ x \right] =\left\{ \begin{array}{l} 0\ \ \ \ \ x\rightarrow 0^+\\ -1\ \ x\rightarrow 0^-\\ \end{array} \right. $
$\underset{x\rightarrow k\pi}{\lim}\dfrac{f\left( x \right) g\left( x \right)}{\sin x}=\left\{ \begin{array}{l} g\left( x \right) \ \ \ \ f\left( x \right) =0\\ \infty \ \ \ \ \ \ \ \ \ f\left( x \right) \ne 0\\ \end{array} \right. $