多元抽象复合函数求二阶偏导
一、多元复合函数
- 形如 f ( x + y + z , x y z ) f(x+y+z,xyz) f(x+y+z,xyz) 的函数
二、求导策略
- 一阶偏导求解的核心策略是:链式法则(最好用!)
- 二阶偏导求解的核心策略是:矩阵公式法
三、雅可比矩阵与海森矩阵
01 雅可比矩阵与海森矩阵的概念
现有多元函数组 f 1 ( x 1 , x 2 , ⋯ , x n ) , f 2 ( x 1 , x 2 , ⋯ , x n ) , ⋯ , f m ( x 1 , x 2 , ⋯ , x n ) \mathrm{f_1(x_1,x_2,\cdots,x_n)\ ,\ f_2(x_1,x_2,\cdots,x_n)\ ,\cdots\ ,\ f_m(x_1,x_2,\cdots,x_n)} f1(x1,x2,⋯,xn) , f2(x1,x2,⋯,xn) ,⋯ , fm(x1,x2,⋯,xn),
那么其雅可比矩阵如下所示,它储存了该函数组所有一阶偏导数的信息
J
=
[
∂
f
1
∂
x
1
∂
f
1
∂
x
2
⋯
∂
f
1
∂
x
n
∂
f
2
∂
x
1
∂
f
2
∂
x
2
⋯
∂
f
2
∂
x
n
⋮
⋮
⋱
⋮
∂
f
m
∂
x
1
∂
f
m
∂
x
2
⋯
∂
f
m
∂
x
n
]
\mathrm{ J= \left[\begin{array}{cccc} \displaystyle{ \frac{\partial f_{1}}{\partial x_{1}} } & \displaystyle{ \frac{\partial f_{1}}{\partial x_{2}} } & \cdots & \displaystyle{ \frac{\partial f_{1}}{\partial x_{n}} } \\ \displaystyle{ \frac{\partial f_{2}}{\partial x_{1}} } & \displaystyle{ \frac{\partial f_{2}}{\partial x_{2}} } & \cdots & \displaystyle{ \frac{\partial f_{2}}{\partial x_{n}} } \\ \vdots & \vdots & \ddots & \vdots \\ \displaystyle{ \frac{\partial f_{m}}{\partial x_{1}} } & \displaystyle{ \frac{\partial f_{m}}{\partial x_{2}} } & \cdots & \displaystyle{ \frac{\partial f_{m}}{\partial x_{n}} } \\ \end{array}\right] }
J=
∂x1∂f1∂x1∂f2⋮∂x1∂fm∂x2∂f1∂x2∂f2⋮∂x2∂fm⋯⋯⋱⋯∂xn∂f1∂xn∂f2⋮∂xn∂fm
我们可以观察到,它满足同一行(hang)同一函数(han),同一列(lie)同一自变量(liang)的规律
当多元函数组退化为一个多元函数时,其表示成
J
=
[
∂
f
∂
x
1
,
∂
f
∂
x
2
,
⋯
,
∂
f
∂
x
n
]
\mathrm{ J= \left[\begin{array}{cccc} \displaystyle{ \frac{\partial f}{\partial x_{1}} } \ ,\ \displaystyle{ \frac{\partial f}{\partial x_{2}} } \ ,\ \cdots \ ,\ \displaystyle{ \frac{\partial f}{\partial x_{n}} } \end{array}\right] }
J=[∂x1∂f , ∂x2∂f , ⋯ , ∂xn∂f]
此时这个多元函数的海森矩阵如下所示,它储存了这个多元函数所有二阶偏导数的信息
H
=
[
∂
2
f
∂
x
1
2
∂
2
f
∂
x
1
∂
x
2
⋯
∂
2
f
∂
x
1
∂
x
n
∂
2
f
∂
x
2
∂
x
1
∂
2
f
∂
x
2
2
⋯
∂
2
f
∂
x
2
∂
x
n
⋮
⋮
⋱
⋮
∂
2
f
∂
x
n
∂
x
1
∂
2
f
∂
x
n
∂
x
2
⋯
∂
2
f
∂
x
n
2
]
\mathrm{ H= \left[\begin{array}{cccc} \displaystyle{ \frac{\partial^2 f}{\partial x_1^2} } & \displaystyle{ \frac{\partial^2 f}{\partial x_1\partial x_2} } & \cdots & \displaystyle{ \frac{\partial^2 f}{\partial x_1\partial x_n} } \\ \displaystyle{ \frac{\partial^2 f}{\partial x_2\partial x_1} } & \displaystyle{ \frac{\partial^2 f}{\partial x_2^2} } & \cdots & \displaystyle{ \frac{\partial^2 f}{\partial x_2\partial x_n} } \\ \vdots & \vdots & \ddots & \vdots \\ \displaystyle{ \frac{\partial^2 f}{\partial x_n\partial x_1} } & \displaystyle{ \frac{\partial^2 f}{\partial x_n\partial x_2} } & \cdots & \displaystyle{ \frac{\partial^2 f}{\partial x_n^2} } \\ \end{array}\right] }
H=
∂x12∂2f∂x2∂x1∂2f⋮∂xn∂x1∂2f∂x1∂x2∂2f∂x22∂2f⋮∂xn∂x2∂2f⋯⋯⋱⋯∂x1∂xn∂2f∂x2∂xn∂2f⋮∂xn2∂2f
显然,当多元函数
f
\mathrm{f}
f 二阶偏导数连续时,海森矩阵为实对称矩阵
02 不同情况下雅可比行列式的形式
当多元函数组由两个二元函数
{
u
(
x
,
y
)
=
0
v
(
x
,
y
)
=
0
\mathrm{\begin{cases}\ u(x,y)=0 \\ \ v(x,y)=0 \end{cases}}
{ u(x,y)=0 v(x,y)=0 组成时,我们可以把它写成
J
=
[
∂
u
∂
x
∂
u
∂
y
∂
v
∂
x
∂
v
∂
y
]
\mathrm{ J= \left[\begin{array}{cccc} \displaystyle{ \frac{\partial u}{\partial x} } & \displaystyle{ \frac{\partial u}{\partial y} } \\ \displaystyle{ \frac{\partial v}{\partial x} } & \displaystyle{ \frac{\partial v}{\partial y} } \end{array}\right] }
J=
∂x∂u∂x∂v∂y∂u∂y∂v
当多元函数组由三个二元函数
{
u
(
x
,
y
)
=
0
v
(
x
,
y
)
=
0
w
(
x
,
y
)
=
0
\mathrm{\begin{cases}\ u(x,y)=0 \\ \ v(x,y)=0\\ \ w(x,y)=0 \end{cases}}
⎩
⎨
⎧ u(x,y)=0 v(x,y)=0 w(x,y)=0 组成时,我们可以把它写成
J
=
[
∂
u
∂
x
∂
u
∂
y
∂
v
∂
x
∂
v
∂
y
∂
w
∂
x
∂
w
∂
y
]
\mathrm{ J= \left[\begin{array}{cccc} \displaystyle{ \frac{\partial u}{\partial x} } & \displaystyle{ \frac{\partial u}{\partial y} } \\ \displaystyle{ \frac{\partial v}{\partial x} } & \displaystyle{ \frac{\partial v}{\partial y} } \\ \displaystyle{ \frac{\partial w}{\partial x} } & \displaystyle{ \frac{\partial w}{\partial y} } \end{array}\right] }
J=
∂x∂u∂x∂v∂x∂w∂y∂u∂y∂v∂y∂w
03 不同情况下海森矩阵的形式
当函数为
f
(
x
,
y
)
=
0
f(x,y)=0
f(x,y)=0 时,海森矩阵表达为:
H
=
[
f
x
x
′
′
f
x
y
′
′
f
y
x
′
′
f
y
y
′
′
]
\mathrm{ H= \left[\begin{array}{cccc} f''_{xx} & f''_{xy}\\ f''_{yx} & f''_{yy} \end{array}\right] }
H=[fxx′′fyx′′fxy′′fyy′′]
当函数为
F
(
x
,
y
,
z
)
=
0
F(x,y,z)=0
F(x,y,z)=0 时,海森矩阵表达为:
H
=
[
F
x
x
′
′
F
x
y
′
′
F
x
z
′
′
F
y
x
′
′
F
y
y
′
′
F
y
z
′
′
F
z
x
′
′
F
z
y
′
′
F
z
z
′
′
]
\mathrm{ H= \left[\begin{array}{cccc} F''_{xx} & F''_{xy} & F''_{xz}\\ F''_{yx} & F''_{yy} & F''_{yz}\\ F''_{zx} & F''_{zy} & F''_{zz} \end{array}\right] }
H=
Fxx′′Fyx′′Fzx′′Fxy′′Fyy′′Fzy′′Fxz′′Fyz′′Fzz′′
四、矩阵公式法
四、矩阵公式法
01 定理内容
(1) 一般的偏导数求解
- 若函数 u = f ( x 1 , x 2 , ⋯ , x n ) u=f\left(x_{1}, x_{2}, \cdots, x_{n}\right) u=f(x1,x2,⋯,xn) 在 ( x 1 , x 2 , ⋯ , x n ) \left(x_{1}, x_{2}, \cdots, x_{n}\right) (x1,x2,⋯,xn) 可微,而 x i = φ i ( t 1 , t 2 , ⋯ , t m ) x_{i}=\varphi_{i}\left(t_{1}, t_{2}, \cdots, t_{m}\right) xi=φi(t1,t2,⋯,tm) 在点 ( t 1 , t 2 \left(t_{1}, t_{2}\right. (t1,t2, ⋯ , t m ) \left.\cdots, t_{m}\right) ⋯,tm) 处
- 存在偏导数 ( i = 1 , 2 , ⋯ , n ) (i=1,2, \cdots, n) (i=1,2,⋯,n),则 u = f ( φ 1 ( t 1 , t 2 , ⋯ , t m ) , ⋯ , φ n ( t 1 , t 2 , ⋯ , t m ) ) u=f\left(\varphi_{1}\left(t_{1}, t_{2}, \cdots, t_{m}\right), \cdots, \varphi_{n}\left(t_{1}, t_{2}, \cdots, t_{m}\right)\right) u=f(φ1(t1,t2,⋯,tm),⋯,φn(t1,t2,⋯,tm)) 在 ( t 1 , t 2 , ⋯ , t m ) \left(t_{1}, t_{2}, \cdots, t_{m}\right) (t1,t2,⋯,tm) 处存在偏导数,
- 且有: ∂ u ∂ t j = ∂ u ∂ x 1 ∂ φ 1 ∂ t j + ∂ u ∂ x 2 ∂ φ 2 ∂ t j + … + ∂ u ∂ x n ∂ φ n ∂ t j , ( j = 1 , ⋯ , m ) . \displaystyle{ \frac{\partial u}{\partial t_{j}}=\frac{\partial u}{\partial x_{1}} \frac{\partial \varphi_{1}}{\partial t_{j}}+\frac{\partial u}{\partial x_{2}} \frac{\partial \varphi_{2}}{\partial t_{j}}+\ldots+\frac{\partial u}{\partial x_{n}} \frac{\partial \varphi_{n}}{\partial t_{j}},(j=1, \cdots, m) . } ∂tj∂u=∂x1∂u∂tj∂φ1+∂x2∂u∂tj∂φ2+…+∂xn∂u∂tj∂φn,(j=1,⋯,m).
(2) 矩阵公式定理
-
设函数 u = f ( x 1 , x 2 , ⋯ , x n ) , ∂ u ∂ x 1 , ∂ u ∂ x 2 , ⋯ , ∂ u ∂ x n \displaystyle{ u=f\left(x_{1}, x_{2}, \cdots, x_{n}\right)\ ,\ \frac{\partial u}{\partial x_{1}}, \frac{\partial u}{\partial x_{2}}, \cdots, \frac{\partial u}{\partial x_{n}} } u=f(x1,x2,⋯,xn) , ∂x1∂u,∂x2∂u,⋯,∂xn∂u 在 ( x 1 , x 2 , ⋯ , x n ) \left(x_{1}, x_{2}, \cdots, x_{n}\right) (x1,x2,⋯,xn) 处可微,
-
函数 x i = φ i ( t 1 , t 2 x_{i}=\varphi_{i}\left(t_{1}, t_{2}\right. xi=φi(t1,t2, ⋯ , t m ) \left.\cdots, t_{m}\right) ⋯,tm) 在点 ( t 1 , t 2 , ⋯ , t m ) \left(t_{1}, t_{2}, \cdots, t_{m}\right) (t1,t2,⋯,tm) 有二阶偏导数,
-
则复合函数 u = f ( φ 1 ( t 1 , t 2 , ⋯ , t m ) , ⋯ , φ n ( t 1 , t 2 , ⋯ , t m ) ) u=f\left(\varphi_{1}\left(t_{1}, t_{2}, \cdots, t_{m}\right), \cdots, \varphi_{n}\left(t_{1}, t_{2}, \cdots, t_{m}\right)\right) u=f(φ1(t1,t2,⋯,tm),⋯,φn(t1,t2,⋯,tm)) 在点 ( t 1 , t 2 , ⋯ , t m ) \left(t_{1}, t_{2}, \cdots, t_{m}\right) (t1,t2,⋯,tm) 处存在二阶偏导数,
-
且有公式( i ∈ { 1 , 2 , ⋯ , m } , j = 1 , 2 , ⋯ , m i\in\{1,2, \cdots, m\}\ , \ j=1,2, \cdots, m i∈{1,2,⋯,m} , j=1,2,⋯,m ):
02 常见形式
在解决实际问题的时候,我们通常接触的是,一阶、二阶、三阶矩阵的形式,我们下面分别给出:
(1) 一阶形式
(2) 二阶形式
(3) 三阶形式
(4) 统一形式
显然,我们可以统一写成如下形式:
(4) 线性复合
接下来根据上述结论我们讨论一种常见的类型,线性复合。
形如 z = f ( u , v ) , u = a x + b y + e , v = c x + d y + f z=f(u,v)\ , \ u=ax+by+e\ , \ v=cx+dy+f z=f(u,v) , u=ax+by+e , v=cx+dy+f 的二元抽象线性复合函数,代入公式可以得到: