高数 02.01导数的概念

$\color{blue}{第二章 一元函数微分学}$

微积分创始人：英国数据家牛顿，德国数学家莱布尼兹

$微分学\left \{ \begin{array}{l} 导数--描述函数变化的快慢 \\ 微分 -- 描述函数变化的程度 \end{array} \right.$

$\color{blue}{第二章 第一节 导数的概念}$

一、引例
二、导数的定义
三、导数的几何意义
四、函数的可导性与连续性的关系
五、单侧导数

$\color{blue}{一、引例}$

1.变速直线运动的速度
$设描述质点运动位置的函数为 s = f(t)，$
$则t_0到t的平均速度为：$
$\bar{v} = \dfrac{f(t) - f(t_0)}{t - t_0}$
$而在t_0时刻的瞬时速度为$
$v = \lim_{t \rightarrow t_0}{\dfrac{f(t) - f(t_0)}{t - t_0}}$

2.曲线的切线斜率
$曲线C: y = f(x)在M点处的切线$
$--割线MN的极限位置MT$
$\qquad \qquad (当\varphi \rightarrow \alpha时)$
$切线MT的斜率 k = \tan{\alpha} = \lim_{\varphi \rightarrow \alpha}{\tan{\varphi}}$
$割线MN的斜率 \tan{\varphi} = \dfrac{f(x) - f(x_0)}{x - x_0}$

$k = \lim_{x \rightarrow x_0}{\dfrac{f(x) - f(x_0)}{x - x_0}}$

两个问题的共性：
所求量为函数增量与自变量增量之比的极限.
类似问题还有：
加速度是速度增量与时间增量之比的极限
角速度是转角增量与时间增量之比的极限
线密度是质量增量与长度增量之比的极限
电流强度是电量增量与时间增量之比的极限

$\color{blue}{二、导数的定义}$

$定义1.设函数 y = f(x) 在点x_0的某邻域内有定义，若\\ \lim_{x \rightarrow x_0}{\dfrac{f(x) - f(x_0)}{x - x_0}} =\lim_{\Delta{x} \rightarrow 0}{\dfrac{\Delta{y}}{\Delta{x}} }存在，则\\ 称函数f(x)在点x_0处可导，并称此极限为y = f(x)在点x_0的导数.记作：$
$\left. y^\prime \right|_{x = x_0}; f^{\prime}(x_0); \left. \dfrac{dy}{dx}\right|_{x = x_0};\left. \dfrac{df(x)}{dx} \right|_{x = x_0}$
$即 \left. y^{\prime} \right|_{x = x_0} = f^{\prime}(x_0) = \lim_{\Delta{x} \rightarrow 0}{\dfrac{\Delta{y}}{\Delta{x}}}$
$= \lim_{\Delta{x} \rightarrow 0}{\dfrac{f(x_0 + \Delta{x}) - f(x_0)}{\Delta{x}}}$
$= \lim_{h \rightarrow 0}{\dfrac{f(x_0 +h) - f(x_0)}{h}}$

$运动质点的位置函数s = f(t)$
$在t_0时刻的瞬间速度$
$v = \lim_{t \rightarrow t_0}{\dfrac{f(t) - f(t_0)}{t - t_0}}$

$曲线C:y = f(x)在M点的切线斜率$
$k = \lim_{x \rightarrow x_0}{\dfrac{f(x) - f(x_0)}{x - x_0}} = f^{\prime}(x_0)$

说明：在经济学中，边际成本率，边际劳动生产率和边际税率等从数学角度看就是导数。

$\lim_{x \rightarrow x_0}{\dfrac{f(x) - f(x_0)}{x - x_0}} = \lim_{\Delta{x} \rightarrow 0}{\dfrac{\Delta{y}}{\Delta{x}}}$
$若上述极限不存在，就说函数在点x_0处不可导.$
$若\lim_{\Delta{x} \rightarrow 0}{\dfrac{\Delta{y}}{\Delta{x}}} = \infty,也称f(x)在x_0的导数为无穷大.$
$若函数在开区间I内每点都可导，就称函数在I内可导.$
$此时导数值构成的新函数称为导函数.$
$记作：y^{\prime};f^{\prime}(x);\dfrac{dy}{dx};\dfrac{df(x)}{dx}$
$注意：f^{\prime}(x_0) = \left. f^{\prime}(x) \right|_{x = x_0} \neq \dfrac{df(x_0)}{dx}$

$在导数的定义中，虽然x可以取区间I内的任意数值，\\ 但在极限的过程中，x是常数，\Delta{x}或h是变量。$
$函数f(x)在点x_0处的导数f^{\prime}(x_0)就是导函数f^{\prime}(x)\\ 在点x = x_0处的函数值，即 f^{\prime}(x_0) =\left. f^{\prime}(x) = \right|_{x = x_0}$

$例1.求函数f(x) = C(C为常数)的导数.$
$解: y^\prime= \lim_{\Delta{x} \rightarrow 0}{\dfrac{f(x + \Delta{x}) - f(x)}{\Delta{x}}} \\ = \lim_{\Delta{x} \rightarrow 0}{\dfrac{C - C}{\Delta{x}}} = 0$
$即(C)^{\prime} = 0$

$例2.求函数f(x) = a^x(a > 0, a \neq 1)的导数.$
$解：f^{\prime}(x) = \lim_{h \rightarrow 0}{\dfrac{f(x + h) - f(x)}{h}} \\ = \lim_{h \rightarrow 0}{\dfrac{a^{x + h} - a^x} {h}} \\ = \lim_{h \rightarrow 0}{\dfrac{a^x(a^h - 1)} {h}} \\ = a^x\ln{a}$
$这就是指数函数的导数公式.特殊地，当a = e时,因\ln{e} = 1,故有(e^x)^{\prime} = e^x$

$例3.求函数f(x) = x^n(n \in N^+)$
$解：f^{\prime}(a) = \lim_{x \rightarrow a}{\dfrac{f(x) - f(a)}{x - a}} \\ = \lim_{x \rightarrow a}{\dfrac{x^n - a^n}{x - a}} \\ = \lim_{x \rightarrow a}{(x^{n - 1} + ax^{n - 2} + a^2x^{n - 3} + \cdots + a^{n - 1})} \\ = na^{n - 1}$
$说明：对一般幂函数y = x^{\mu}({\mu}为常数) \\ (x^{\mu})^{\prime} = {\mu}x^{{\mu} - 1}$
$\mu$
$例如，(\sqrt{x})^{\prime} = (x^\frac{1}{2})^{\prime} = {\dfrac{1}{2}}x^{-\frac{1}{2}} = \dfrac{1}{2\sqrt{x}}$

$(\dfrac{1}{x})^{\prime} = (x^{-1})^{\prime} = -x^{-1 - 1} = - \dfrac{1}{x^2}$

$(\dfrac{1}{\sqrt{x \sqrt{x}}})^{\prime} = (x^{-\frac{3}{4}})^{\prime} = -\dfrac{3}{4}x^{\frac{7}{4}}$

$例4.求函数f(x) = \sin{x}的导数.$
$解：令h = \Delta{x},则 \\ f^{\prime}(x) = \lim_{h \rightarrow 0}{\dfrac{f(x _ h) - f(x)}{h}} \\ = \lim_{h \rightarrow 0}{\dfrac{\sin(x + h) - \sin{x}}{h}} \\ = \lim_{h \rightarrow 0}{\dfrac{2\cos(x + \dfrac{h}{2})\sin{\dfrac{h}{2}}}{h}} \\ = \lim_{h \rightarrow 0}{{\cos(x + \dfrac{h}{2}) \cdot \dfrac{\sin{\dfrac{h}{2}}}{\dfrac{h}{2}}}} \\ = \cos{x} \\ 即 (\sin{x})^{\prime} = \cos{x}$
$类似可证得 (\cos{x})^{\prime} = - \sin{x}$

$例5.求函数f(x) = \ln{x}的导数.$
$解：f^{\prime}(x) = \lim_{h \rightarrow 0}{\dfrac{f(x + h) - f(x)}{h}} \\ = \lim_{h \rightarrow 0}{\dfrac{\ln(x + h) - \ln(x)}{h}} \\ = \lim_{h \rightarrow 0}{\dfrac{1}{h} \cdot \ln(1 + \dfrac{h}{x})} \\ = \lim_{h \rightarrow 0}{[\ln(1 + \dfrac{h}{x})^{\frac{x}{h}}]^{\frac{1}{x}}} \\ = \dfrac{1}{x} \cdot \lim_{h \rightarrow 0}{\ln(1 + \dfrac{h}{x})^{\frac{x}{h}}} \\ = \dfrac{1}{x} \ln{e} \\ = \dfrac{1}{x} \\ 即 (ln{x})^{\prime} = \dfrac{1}{x}$

$例6.证明函数f(x) = |x|在x = 0不可导.$
$证：\because \dfrac{f(0 + h) - f(0)}{h} = \dfrac{|h|}{h} = \left \{ \begin{array}{l} 1, h > 0 \\ -1, h < 0 \end{array} \right.$
$\therefore \lim_{h \rightarrow 0}{\dfrac{f(0 + h) - f(0)}{h}}不存在,即|x|在x = 0处不可导.$

$例7.设f^{\prime}(x_0)存在，求极限\lim_{h \rightarrow 0}{\dfrac{f(x_0 + h) - f(x_0 -h)}{2h}}$
$解：$
$\lim_{h \rightarrow 0}{\dfrac{f(x_0 + h) - f(x_0 -h)}{2h}} \\ = \lim_{h \rightarrow 0}{[\dfrac{f(x_0 + h) - f(x_0)}{2h} + \dfrac{f(x_0) - f(x_0 -h)}{2h}]} \\ = \dfrac{1}{2} \cdot [\lim_{h \rightarrow 0}{\dfrac{f(x_0 + h) - f(x_0)}{h}} + \lim_{h \rightarrow 0}{\dfrac{f(x_0 + (-h)) - f(x_0)}{(-h)}}] \\ = \dfrac{1}{2}[f^{\prime}(x_0) + f^{\prime}(x_0)] \\ = f^{\prime}(x_0)$

$\color{blue}{三、导数的几何意义}$

$曲线y = f(x)在点(x_0, y_0)的切线斜率为\tan{\alpha} = f^{\prime}(x_0) \\ 若f^{\prime}(x_0) > 0,曲线过(x_0, y_0)上升;\\ 若f^{\prime}(x_0) < 0,曲线过(x_0, y_0)下降; \\ 若f^{\prime}(x_0) = 0,切线与x轴平行，x_0称为驻点;\\ 若f^{\prime}(x_0) = \infty,切线与x轴垂直.$
$f^{\prime}(x_0) \neq \infty时,曲线在点(x_0, y_0)处的 \\ \color{blue}{切线方程: y - y_0 = f^{\prime}(x_0)(x - x_0)}$
$\color{blue}{法线方程(与切线垂直)：y - y_0 = \dfrac{1}{f^{\prime}(x_0)}(x - x_0)}$

$例8.问曲线y = \sqrt[3]{x}哪一点有垂直切线？哪一点处的切线与\\ 直线y = \dfrac{1}{3}x - 1 平行？写出其切线方程.$
$\color{blue}{解: \because y^{\prime} = (\sqrt[3]{x})^{\prime} = \dfrac{1}{3} x^{-\frac{2}{3}} = \dfrac{1}{3} \dfrac{1}{\sqrt[3]{x^2}} \\ \therefore \left. y^{\prime}\right|_{x = 0} = \infty，故在原点(0, 0)有垂直切线x = 0\\ 直线平行，斜率相同。\\ \dfrac{1}{3} \dfrac{1}{\sqrt[3]{x^2}} = \dfrac{1}{3}，得 x = \pm 1, 对应y = \pm 1,\\ 则在点(1, 1), (-1, -1)处与直线y = \frac{1}{3}x - 1平行的切线方程分别为：\\ y - 1 = \frac{1}{3}(x - 1), y + 1 = \frac{1}{3}(x + 1) \\ 即 \quad x - 3y \pm 2 = 0 }$

$\color{blue}{四、函数的可导性与连续性的关系}$

$定理1.f(x)在点x处可导 \Longrightarrow f(x)在点x处连续$
$证:设y = f(x)在点x处可导，即\lim_{\Delta{x} \rightarrow 0}{\dfrac{\Delta{y}}{\Delta{x}}} = f^{\prime}(x)存在，因此必有\\ \dfrac{\Delta{y}}{\Delta{x}} = f^{\prime}(x) + \alpha,其中\lim_{\Delta{x} \rightarrow 0}{\alpha} = 0 \\ 故 \quad \Delta{y} = f^{\prime}(x) \Delta(x) + \alpha \Delta{x} \stackrel{\Delta{x} \rightarrow 0}{\longrightarrow} 0 \\ 所以函数y = f(x)在点x连续.$
$注意：函数在点x连续未必可导.$
$反例：y = |x|在x = 0处连续，但不可导$

$\color{blue}{五、单侧导数}$

$定义2.设函数y = f(x) 在点x_0的某个右{\color{green}{(左)}}邻域内有定义，若极限\\ \lim_{\stackrel{\Delta{x} \rightarrow 0^+}{\color{green}{(\Delta{x} \rightarrow 0^-)}}}{\dfrac{\Delta{y}}{\Delta{x}}}= \lim_{\stackrel{\Delta{x} \rightarrow 0^+}{\color{green}{(\Delta{x} \rightarrow 0^-)}}}{\dfrac{f(x_0 + \Delta{x}) - f(x_0)}{\Delta{x}}} \\ 存在，则称此极限值为f(x)在x_0处的右{\color{green}{(左)}}导数,记作f^{\prime}_{\stackrel{+}{\color{green}{(-)}}}(x_0)$
$即 \quad {f^{\prime}}_{\stackrel{+}{\color{green}{(-)}}}{(x_0)} = \lim_{\stackrel{\Delta{x} \rightarrow 0^+}{\color{green}{(\Delta{x} \rightarrow 0^-)}}}{\dfrac{f(x_0 + \Delta{x}) - f(x_0)}{\Delta{x}}}$
$例如，f(x) = |x|在x = 0处有$
$f^{\prime}_{+}(0) = 1,$ $f^{\prime }_{ - }(0) = -1$

$定理2.函数 y = f(x) 在点 x_0 可导的充分必要条件是f^{\prime}_{+}{(x_0)}与f^{\prime}_{-}{(x_0)}存在，且f^{\prime}_{+}{(x_0)} = f^{\prime}_{-}{(x_0)}.$
$简写为: \quad f^{\prime}{(x_0)}存在 \Longleftrightarrow f^{\prime}_{+}{(x_0)} = f^{\prime}_{-}{(x_0)}$

$定理3.函数f(x)在点x_0处右(左)导数存在 \Longrightarrow f(x)在点x_0必右(左)连续.$
$若函数f(x)在开区间(a, b)内可导，且f^{\prime}_{+}{(a)}与f^{\prime}_{-}{(b)}都存在，则称f(x)在闭区间[a, b]上可导.$
$显然：f(x)在闭区间[a, b]上可导 \Longrightarrow f(x) \in C[a, b]$

内容小结
$1.导数的实质：增量比的极限；$
$2.f^{\prime}{(x_0)} = a \Longleftrightarrow f^{\prime}_{+}{(x_0)} = f^{\prime}_{-}{(x_0)} = a$
3.导数的几何意义：切线的斜率;
4.可导比连续，但连续不一定可导;
5.已学过的导数公式
$(C)^{\prime} = 0; (x^{\mu})^{\prime} = \mu x^{\mu -1}; (\ln x)^{\prime} = \dfrac{1}{x}$
$(\sin x)^{\prime} = \cos x; (\cos x)^{\prime} = - \sin x;$
$(a^x)^{\prime} = a^x \ln a; (e^x)^{\prime} = e^x$

$6.判断可导性 \left \{ \begin{array}{l}不连续，一定不可导. \\ 直接用导数定义; \\ 看左右导数是否存在且相等. \end{array} \right.$