📝 题目
＊13．求下列函数的 $\frac{\partial^{2} z}{\partial x^{2}}, \frac{\partial^{2} z}{\partial x \partial y}, \frac{\partial^{2} z}{\partial y^{2}}$（其中 $f$ 具有二阶连续偏导数）： （1）$z=f(x y, y)$ ； （2）$z=f\left(x, \frac{x}{y}\right)$ ； （3）$z=f\left(x y^{2}, x^{2} y\right)$ ； （4）$z=f\left(\sin x, \cos y, \mathrm{e}^{x+y}\right)$ ．
💡 答案与解析
[AI解答]
[AI解答]
我们逐小题计算二阶偏导数，假设 $f$ 具有二阶连续偏导数，因此混合偏导可交换次序。
---
### （1）$z = f(xy, y)$
令 $u = xy$，$v = y$，则 $$ z = f(u, v) $$
一阶偏导： $$ \frac{\partial z}{\partial x} = f_u \cdot y + f_v \cdot 0 = y f_u $$ $$ \frac{\partial z}{\partial y} = f_u \cdot x + f_v \cdot 1 = x f_u + f_v $$
二阶偏导：
对 $x$ 再求导： $$ \frac{\partial^2 z}{\partial x^2} = \frac{\partial}{\partial x}(y f_u) = y \left( f_{uu} \cdot y + f_{uv} \cdot 0 \right) = y^2 f_{uu} $$
先对 $x$ 后对 $y$： $$ \frac{\partial^2 z}{\partial x \partial y} = \frac{\partial}{\partial y}(y f_u) = f_u + y \left( f_{uu} \cdot x + f_{uv} \cdot 1 \right) = f_u + xy f_{uu} + y f_{uv} $$
对 $y$ 再求导： $$ \frac{\partial^2 z}{\partial y^2} = \frac{\partial}{\partial y}(x f_u + f_v) = x \left( f_{uu} \cdot x + f_{uv} \cdot 1 \right) + \left( f_{vu} \cdot x + f_{vv} \cdot 1 \right) $$ 由连续性 $f_{uv}=f_{vu}$，得： $$ = x^2 f_{uu} + x f_{uv} + x f_{uv} + f_{vv} = x^2 f_{uu} + 2x f_{uv} + f_{vv} $$
---
### （2）$z = f\left(x, \frac{x}{y}\right)$
令 $u = x$，$v = \frac{x}{y}$，则 $$ z = f(u, v) $$
一阶偏导： $$ \frac{\partial z}{\partial x} = f_u \cdot 1 + f_v \cdot \frac{1}{y} = f_u + \frac{1}{y} f_v $$ $$ \frac{\partial z}{\partial y} = f_u \cdot 0 + f_v \cdot \left(-\frac{x}{y^2}\right) = -\frac{x}{y^2} f_v $$
二阶偏导：
对 $x$ 再求导： $$ \frac{\partial^2 z}{\partial x^2} = \frac{\partial}{\partial x}\left( f_u + \frac{1}{y} f_v \right) = \left( f_{uu} \cdot 1 + f_{uv} \cdot \frac{1}{y} \right) + \frac{1}{y} \left( f_{vu} \cdot 1 + f_{vv} \cdot \frac{1}{y} \right) $$ $$ = f_{uu} + \frac{2}{y} f_{uv} + \frac{1}{y^2} f_{vv} $$
先对 $x$ 后对 $y$： $$ \frac{\partial^2 z}{\partial x \partial y} = \frac{\partial}{\partial y}\left( f_u + \frac{1}{y} f_v \right) = \left( f_{uu} \cdot 0 + f_{uv} \cdot \left(-\frac{x}{y^2}\right) \right) + \left( -\frac{1}{y^2} f_v + \frac{1}{y} \left[ f_{vu} \cdot 0 + f_{vv} \cdot \left(-\frac{x}{y^2}\right) \right] \right) $$ $$ = -\frac{x}{y^2} f_{uv} - \frac{1}{y^2} f_v - \frac{x}{y^3} f_{vv} $$
对 $y$ 再求导： $$ \frac{\partial^2 z}{\partial y^2} = \frac{\partial}{\partial y}\left( -\frac{x}{y^2} f_v \right) = \frac{2x}{y^3} f_v - \frac{x}{y^2} \left( f_{vu} \cdot 0 + f_{vv} \cdot \left(-\frac{x}{y^2}\right) \right) = \frac{2x}{y^3} f_v + \frac{x^2}{y^4} f_{vv} $$
---
### （3）$z = f(x y^2, x^2 y)$
令 $u = x y^2$，$v = x^2 y$，则 $$ z = f(u, v) $$
一阶偏导： $$ \frac{\partial z}{\partial x} = f_u \cdot y^2 + f_v \cdot 2xy = y^2 f_u + 2xy f_v $$ $$ \frac{\partial z}{\partial y} = f_u \cdot 2xy + f_v \cdot x^2 = 2xy f_u + x^2 f_v $$
二阶偏导：
对 $x$ 再求导： $$ \frac{\partial^2 z}{\partial x^2} = \frac{\partial}{\partial x}(y^2 f_u + 2xy f_v) = y^2 ( f_{uu} \cdot y^2 + f_{uv} \cdot 2xy ) + 2y f_v + 2xy ( f_{vu} \cdot y^2 + f_{vv} \cdot 2xy ) $$ $$ = y^4 f_{uu} + 2x y^3 f_{uv} + 2y f_v + 2x y^3 f_{uv} + 4x^2 y^2 f_{vv} $$ $$ = y^4 f_{uu} + 4x y^3 f_{uv} + 4x^2 y^2 f_{vv} + 2y f_v $$
先对 $x$ 后对 $y$： $$ \frac{\partial^2 z}{\partial x \partial y} = \frac{\partial}{\partial y}(y^2 f_u + 2xy f_v) = 2y f_u + y^2 ( f_{uu} \cdot 2xy + f_{uv} \cdot x^2 ) + 2x f_v + 2xy ( f_{vu} \cdot 2xy + f_{vv} \cdot x^2 ) $$ $$ = 2y f_u + 2x y^3 f_{uu} + x^2 y^2 f_{uv} + 2x f_v + 4x^2 y^2 f_{uv} + 2x^3 y f_{vv} $$ $$ = 2y f_u + 2x f_v + 2x y^3 f_{uu} + (x^2 y^2 + 4x^2 y^2) f_{uv} + 2x^3 y f_{vv} $$ $$ = 2y f_u + 2x f_v + 2x y^3 f_{uu} + 5x^2 y^2 f_{uv} + 2x^3 y f_{vv} $$
对 $y$ 再求导： $$ \frac{\partial^2 z}{\partial y^2} = \frac{\partial}{\partial y}(2xy f_u + x^2 f_v) = 2x f_u + 2xy ( f_{uu} \cdot 2xy + f_{uv} \cdot x^2 ) + x^2 ( f_{vu} \cdot 2xy + f_{vv} \cdot x^2 ) $$ $$ = 2x f_u + 4x^2 y^2 f_{uu} + 2x^3 y f_{uv} + 2x^3 y f_{uv} + x^4 f_{vv} $$ $$ = 2x f_u + 4x^2 y^2 f_{uu} + 4x^3 y f_{uv} + x^4 f_{vv} $$
---
### （4）$z = f(\sin x, \cos y, e^{x+y})$
令 $u = \sin x$，$v = \cos y$，$w = e^{x+y}$，则 $$ z = f(u, v, w) $$
一阶偏导： $$ \frac{\partial z}{\partial x} = f_u \cos x + f_w e^{x+y} $$ $$ \frac{\partial z}{\partial y} = f_v (-\sin y) + f_w e^{