这篇个人笔记详细解释了Nabla操作符，包括索引符号、定义、导数等概念。Nabla用于表示梯度、散度和旋度，并通过矩阵形式展示向量函数的雅可比矩阵。此外，还介绍了标量函数和向量函数的导数表达式。
This is a personal note with personal understanding.
Notations
Index Notation is included.
Scalars are not bold ( a , b , c ⋯ a,b,c\cdots a,b,c⋯), while vectors are bold ( a , b , c ⋯ \boldsymbol{a,b,c\cdots} a,b,c⋯)
Sometimes for convenience,
∂
x
=
∂
∂
x
∂
y
=
∂
∂
y
∂
z
=
∂
∂
z
\partial_x = \frac{\partial}{\partial x} \qquad\partial_y = \frac{\partial}{\partial y} \qquad\partial_z = \frac{\partial}{\partial z}
∂x=∂x∂∂y=∂y∂∂z=∂z∂
Row vector inner product a ⋅ b = a b T \newcommand{\V}[1]{\boldsymbol{#1}}\V{a}\cdot\V{b} = \V{a}\V{b}^T a⋅b=abT
Definitions
Operators and functions can be expressed as a row vector.
∇
=
(
∂
∂
x
,
∂
∂
y
,
∂
∂
z
)
\nabla = \Big(\frac{\partial}{\partial x}, \frac{\partial}{\partial y}, \frac{\partial}{\partial z}\Big)
∇=(∂x∂,∂y∂,∂z∂)
Gradient of a scalar function is a vector
grad
f
=
∇
f
=
(
∂
∂
x
,
∂
∂
y
,
∂
∂
z
)
f
=
(
∂
f
∂
x
,
∂
f
∂
y
,
∂
f
∂
z
)
\text{grad} f = \nabla f = \Big(\frac{\partial}{\partial x}, \frac{\partial}{\partial y}, \frac{\partial}{\partial z}\Big)f = \Big(\frac{\partial f}{\partial x}, \frac{\partial f}{\partial y}, \frac{\partial f}{\partial z}\Big)
gradf=∇f=(∂x∂,∂y∂,∂z∂)f=(∂x∂f,∂y∂f,∂z∂f)
Divergence of a vector function is a scalar
div
F
=
∇
⋅
F
=
∂
F
i
∂
x
i
=
∂
F
x
∂
x
+
∂
F
y
∂
y
+
∂
F
z
∂
z
\newcommand{\V}[1]{\boldsymbol{#1}} \text{div}\V{F} = \nabla\cdot\V{F} = \frac{\partial F_i}{\partial x_i} = \frac{\partial F_x}{\partial x} + \frac{\partial F_y}{\partial y} + \frac{\partial F_z}{\partial z}
divF=∇⋅F=∂xi∂Fi=∂x∂Fx+∂y∂Fy+∂z∂Fz
Curl of a vector function is a vector
curl
F
=
∇
×
F
=
det
∣
e
^
1
e
^
2
e
^
3
∂
x
∂
y
∂
z
F
x
F
y
F
z
∣
\newcommand{\V}[1]{\boldsymbol{#1}} \text{curl}\V{F} = \nabla\times\V{F} = \det\left|\begin{matrix}\hat{e}_1 & \hat{e}_2 & \hat{e}_3 \\ \partial_x & \partial_y & \partial_z \\ F_x & F_y & F_z \end{matrix}\right|
curlF=∇×F=det∣∣∣∣∣∣e^1∂xFxe^2∂yFye^3∂zFz∣∣∣∣∣∣
Jacobian Matrix of a vector function is a 2d matrix
J
F
=
D
F
=
∇
F
=
(
∂
x
F
x
∂
y
F
x
∂
z
F
x
∂
x
F
y
∂
y
F
y
∂
z
F
y
∂
x
F
z
∂
y
F
z
∂
z
F
z
)
\newcommand{\V}[1]{\boldsymbol{#1}} \V{J}_\V{F} = D\V{F} = \nabla\V{F} = \left(\begin{matrix}\partial_x F_x & \partial_y F_x & \partial_z F_x \\ \partial_x F_y & \partial_y F_y & \partial_z F_y\\ \partial_x F_z & \partial_y F_z & \partial_z F_z \end{matrix}\right)
JF=DF=∇F=⎝⎛∂xFx∂xFy∂xFz∂yFx∂yFy∂yFz∂zFx∂zFy∂zFz⎠⎞
Derivative
Infinitesimal changes form column vectors.
For a scalar function
f
f
f, total derivative is
d
f
=
(
∇
f
)
d
r
=
(
∂
f
∂
x
,
∂
f
∂
y
,
∂
f
∂
z
)
[
d
x
d
y
d
z
]
\newcommand{\V}[1]{\boldsymbol{#1}} df = (\nabla f)d\V{r} = \Big(\frac{\partial f}{\partial x}, \frac{\partial f}{\partial y}, \frac{\partial f}{\partial z}\Big)\left[\begin{aligned}dx \\ dy\\ dz\end{aligned}\right]
df=(∇f)dr=(∂x∂f,∂y∂f,∂z∂f)⎣⎢⎡dxdydz⎦⎥⎤
For a vector function
F
\boldsymbol{F}
F
d
F
=
(
∇
F
)
d
r
=
(
∂
x
F
x
∂
y
F
x
∂
z
F
x
∂
x
F
y
∂
y
F
y
∂
z
F
y
∂
x
F
z
∂
y
F
z
∂
z
F
z
)
[
d
x
d
y
d
z
]
\newcommand{\V}[1]{\boldsymbol{#1}} d\V{F} = (\nabla \V{F})d\V{r} = \left(\begin{matrix}\partial_x F_x & \partial_y F_x & \partial_z F_x \\ \partial_x F_y & \partial_y F_y & \partial_z F_y\\ \partial_x F_z & \partial_y F_z & \partial_z F_z \end{matrix}\right)\left[\begin{aligned}dx \\ dy\\ dz\end{aligned}\right]
dF=(∇F)dr=⎝⎛∂xFx∂xFy∂xFz∂yFx∂yFy∂yFz∂zFx∂zFy∂zFz⎠⎞⎣⎢⎡dxdydz⎦⎥⎤
This can be applied to gradient of dot product.
∇
(
A
⋅
B
)
=
A
⋅
(
∇
B
)
+
B
⋅
(
∇
A
)
\newcommand{\V}[1]{\boldsymbol{#1}} \nabla(\V{A}\cdot\V{B}) = \V{A}\cdot(\nabla\V{B}) + \V{B}\cdot(\nabla\V{A})
∇(A⋅B)=A⋅(∇B)+B⋅(∇A)
扫一扫
被折叠的 条评论
为什么被折叠?
到【灌水乐园】发言
