高数 07.03 全微分

$\color{blue}{\S 第七章 第三节 全微分}$

$一元函数 y = f(x)的微分\\ \Delta{y} = \underline{A \Delta{x}} + o(\Delta{x}) \\ \qquad \quad \downarrow \\ dy = f^{\prime}(x)\Delta{x}$

$\color{blue}{一、全微分的定义}$

$定义:如果函数z = f(x, y)在定义域D的内点(x, y)处{\color{blue}{全增量}}\\ \Delta{z} = f(x + \Delta{x}, y + \Delta{y}) - f(x, y)可表示成\\ \Delta{z} = A\Delta{x} + B\Delta{y} + o(\rho), \rho = \sqrt{(\Delta)^2 + (\Delta{y})^2} \\ 其中,A,B不依赖与\Delta{x},\Delta{y},仅与x,y有关,则称函数f(x, y)\\ 在点(x, y){\color{blue}{可微}},A\Delta{x} + B\Delta{y}称为函数f(x, y)在点(x, y)的\\ {\color{blue}{全微分}},记作\\ {\color{blue}{dz = df = A\Delta{x} + B\Delta{y} }} \\ 若函数在域D内各点都可微,则称此{\color{blue}{函数在D内可微}}.$

由微分定义:
$\lim _ {\Delta{x} \rightarrow 0, \Delta{y} \rightarrow = 0}{\Delta{z}} = \lim _ {\rho \rightarrow 0}{[(A\Delta{x} + B\Delta{y}) + o(\rho)]} = 0$
$得\lim_{\Delta{x} \rightarrow 0, \Delta{y} \rightarrow = 0}{f(x + \Delta{x}, y + \Delta{y}} = f(x, y)$
$\Delta{z} = f(x + \Delta{x}, y + \Delta{y}) - f(x, y)$
$即 函数z = f(x, y)在点(x, y)可微 \rightarrow 函数在该点连续$
下面两个定理给出了可微与偏导数的关系：
$(1)函数可微 \longrightarrow 偏导数存在$
$(2)偏导数连续 \longrightarrow 函数可微$

$定理1.(必要条件)若函数z = f(x, y)在点(x, y)可微,\\ 则该函数在该点偏导数\dfrac{\partial z}{\partial x}, \dfrac{\partial z}{\partial y}必存在,且有\\ dz = \dfrac{\partial z}{\partial x}\Delta{x} + \dfrac{\partial z}{\partial y} \Delta{y}$
$证:由全增量公式\Delta{z} = A\Delta{x} + B\Delta{y} + o(\rho),\\ 令\Delta{y} = 0,得到对x的偏增量\\ \Delta_x{z} = f(x + \Delta{x}, y) - f(x, y) = A\Delta{x} + o(|\Delta{x}|) \\ \therefore \dfrac{\partial z}{\partial x} = \lim_{\Delta{x} \rightarrow 0}{\dfrac{\Delta_{x}{z} } {\Delta{x}} } = A \\ 同样可证 \dfrac{\partial z}{\partial y} = B,因此有dz = \dfrac{\partial z}{\partial x} \Delta{x} + \dfrac{\partial z}{\partial y} \Delta{y}$

$定理2(充分条件)若函数z = f(x, y)的偏导数\dfrac{\partial z}{\partial x}, \dfrac{\partial z}{\partial y}在点(x, y)连续,则函数在该点可微分.$
$证:\Delta{z} = f(x + \Delta{x}, y + \Delta{y}) - f(x, y)\\ = [f(x + \Delta{x}, y + \Delta{y}) - f(x, y + \Delta{y})] + [f(x, y + \Delta{y}) - f(x, y)] \\ 根据拉格朗日中值定理\\ = f_x(x + \theta_1 \Delta{x}, y + \Delta{y}) \Delta{x} + f_y(x, y + \theta_2 \Delta{y}) \Delta{y} \quad (0 < \theta_1, \theta_2 < 1) \\ 因为f(x, y)在点(x, y)连续,\\ = [f_x(x, y) + \alpha]\Delta{x} + [f_y(x, y) + \beta]\Delta{y}$

$例1.计算函数z = e^{xy}在点(2, 1)处的全微分.$
$解:\\ \dfrac{\partial z}{\partial x} = ye^{xy} \\ \dfrac{\partial z}{\partial y} = xe^{xy} \\ \left. \dfrac{\partial z}{\partial x} \right| _ {(2, 1)} = e^2 \\ \left. \dfrac{\partial z}{\partial y} \right| _ {(2, 1)} = 2e^2 \\ \left. dz \right| _ {(2, 1)} = e^2dx + 2e^2 dy$

$例2.计算函数u = x + \sin{\dfrac{y}{2}} + e^{yz}的全微分$
$解:\\ du = \dfrac{\partial u}{\partial x} dx + \dfrac{\partial u}{\partial y} dy + \dfrac{\partial u}{\partial z} dz \\ = dx + (\dfrac{1}{2} \cos{\dfrac{y}{2}} +ze^{yz})dy + ye^{yz}dz$

内容小结
$1.微分定义：(z = f(x, y))$
$\Delta{z} = f_x(x, y) \Delta{x} + f_y(x, y)\Delta{y} + o(\rho) \quad \rho = \sqrt{(\Delta{x})^2 + (\Delta{y})^2}$

$2.重要关系：\\ [偏导数连续] \longrightarrow [函数可微]\\ [函数可微] \longrightarrow [函数连续] \\ [函数可微] \longrightarrow [函数可导] \\ [函数连续] 和 [函数可导] 没有 充要关系$

3.微分应用
近似计算
$\Delta{z} \approx f_x(x, y)\Delta{x} + f_y(x, y)\Delta{y}$
$f(x + \Delta{x}, y + \Delta{y}) \approx f(x, y) + f_x(x, y)\Delta{x} + f_y(x, y)\Delta{y}$
估计误差
$绝对误差 \delta_z = |f_x(x, y)|\delta_x + |f_y(x, y)|\delta_y$
$相对误差 \dfrac{\delta_z}{|z|} = \left|\dfrac{f_x(x, y)}{f(x, y)} \right|\delta_x + \left|\dfrac{f_y(x, y)}{f(x, y)} \right|\delta_y$

练习
$1. 函数z = f(x, y)在(x_0, y_0)可微的充分条件是(\ \ {\color{blue}{D} }\ \ )$
$(A) f(x, y)在(x_0, y_0)连续; \\ (B) f_x^{\prime}(x, y), f_y^{\prime}(x, y)在(x_0, y_0)的某领域内存在;\\ (C) \Delta{z} - f_x^{\prime}(x, y)\Delta{x} - f_y^{\prime}(x,y)\Delta{y},当\sqrt{(\Delta{x})^2 + (\Delta{y})^2} \rightarrow 0时是无穷小量; \\ (D) \dfrac{\Delta{z} - f_x^{\prime}(x, y)\Delta{x} - f_y^{\prime}(x, y)\Delta{y}}{\sqrt{(\Delta{x})^2 + (\Delta{y})^2}}, 当\sqrt{(\Delta{x})^2 + (\Delta{y})^2} \rightarrow 0时是无穷小量.$

$2.已知z = \arctan{\dfrac{x}{y}},求dz$
$解：\\ dz = \dfrac{\partial z}{\partial x} dx + \dfrac{\partial z}{\partial y} dy \\ = \dfrac{1}{1 + (\frac{x}{y})^2} \cdot \dfrac{1}{y} dx + \dfrac{1}{1 + (\frac{x}{y})^2} \cdot \dfrac{-x}{y^2} dy \\ = \dfrac{ydx -xdy}{x^2 + y^2}$