概述
泰勒展开/泰勒级数表示用无限个项相加表示一个函数，这些相加的项由函数在某一点的导数求得
定义
对于一个以实数为变量的函数
f
(
x
)
f(x)
f(x)，如果它对变量
x
x
x无穷可导，则函数在实数
a
a
a领域上可以做如下泰勒展开：
f
(
x
)
=
∑
n
=
0
∞
f
(
n
)
(
a
)
n
!
(
x
−
a
)
n
=
f
(
a
)
+
(
x
−
a
)
f
1
(
a
)
+
(
x
−
a
)
2
f
2
(
a
)
⋯
(
x
−
a
)
n
f
n
(
a
)
∀
x
f(x)=\sum_{n=0}^{\infty} \frac{f^{(n)}(a)}{n !}(x-a)^{n}=f(a)+(x-a)f^{1}(a)+(x-a)^{2}f^{2}(a)\cdots(x-a)^{n}f^{n}(a)\quad \forall x
f(x)=n=0∑∞n!f(n)(a)(x−a)n=f(a)+(x−a)f1(a)+(x−a)2f2(a)⋯(x−a)nfn(a)∀x
其中
n
!
n!
n!是
n
n
n的阶乘，
f
(
n
)
(
a
)
{f^{(n)}(a)}
f(n)(a)表示函数
f
f
f在实数
a
a
a处的
n
n
n阶导数
SLAM中常用的泰勒级数
指数函数
e
x
=
∑
n
=
0
∞
x
n
n
!
=
1
+
x
+
x
2
2
!
+
x
3
3
!
+
⋯
+
x
n
n
!
+
⋯
∀
x
e^{x}=\sum_{n=0}^{\infty} \frac{x^{n}}{n !}=1+x+\frac{x^{2}}{2 !}+\frac{x^{3}}{3 !}+\cdots+\frac{x^{n}}{n !}+\cdots \quad \forall x
ex=n=0∑∞n!xn=1+x+2!x2+3!x3+⋯+n!xn+⋯∀x
三角函数
sin
x
=
∑
n
=
0
∞
(
−
1
)
n
(
2
n
+
1
)
!
x
2
n
+
1
=
x
−
x
3
3
!
+
x
5
5
!
−
⋯
∀
x
cos
x
=
∑
n
=
0
∞
(
−
1
)
n
(
2
n
)
!
x
2
n
=
1
−
x
2
2
!
+
x
4
4
!
−
⋯
∀
x
\begin{aligned} &\sin x=\sum_{n=0}^{\infty} \frac{(-1)^{n}}{(2 n+1) !} x^{2 n+1}=x-\frac{x^{3}}{3 !}+\frac{x^{5}}{5 !}-\cdots \quad \forall x\\ &\cos x=\sum_{n=0}^{\infty} \frac{(-1)^{n}}{(2 n) !} x^{2 n}=1-\frac{x^{2}}{2 !}+\frac{x^{4}}{4 !}-\cdots \quad \forall x \end{aligned}
sinx=n=0∑∞(2n+1)!(−1)nx2n+1=x−3!x3+5!x5−⋯∀xcosx=n=0∑∞(2n)!(−1)nx2n=1−2!x2+4!x4−⋯∀x
多元泰勒展开
雅可比矩阵：
假设某函数 f : R n → R m \mathbf{f}: \mathbb{R}^{n} \rightarrow \mathbb{R}^{m} f:Rn→Rm是 x ∈ R n \mathbf{x} \in \mathbb{R}^n x∈Rn映射到向量 f ( x ) ∈ R m \mathbf{f}(\mathbf{x}) \in \mathbb{R}^m f(x)∈Rm的函数，这个函数由m个上述的实函数组成：
R n \mathbb{R}^n Rn表示 n n n维的欧式空间
f = [ f 1 ( x 1 , ⋯ , x n ) f 2 ( x 1 , ⋯ , x n ) ⋮ y n ( x 1 , ⋯ , x n ) ] \mathbf{f}=\left[\begin{array}{c} f_{1}\left(x_{1}, \cdots, x_{n}\right) \\ f_{2}\left(x_{1}, \cdots, x_{n}\right) \\ \vdots \\ y_{n}\left(x_{1}, \cdots, x_{n}\right) \end{array}\right] f=⎣⎢⎢⎢⎡f1(x1,⋯,xn)f2(x1,⋯,xn)⋮yn(x1,⋯,xn)⎦⎥⎥⎥⎤
此函数
f
\mathbf{f}
f的雅可比矩阵
J
J
J 为
m
×
n
m×n
m×n 的矩阵，一般由以下方式定义：
J
=
[
∂
f
∂
x
1
⋯
∂
f
∂
x
n
]
=
[
∂
f
1
∂
x
1
⋯
∂
f
1
∂
x
n
⋮
⋱
⋮
∂
f
m
∂
x
1
⋯
∂
f
m
∂
x
n
]
J=\left[\begin{array}{ccc} \frac{\partial \mathbf{f}}{\partial x_{1}} & \cdots & \frac{\partial \mathbf{f}}{\partial x_{n}} \end{array}\right]=\left[\begin{array}{ccc} \frac{\partial f_{1}}{\partial x_{1}} & \cdots & \frac{\partial f_{1}}{\partial x_{n}} \\ \vdots & \ddots & \vdots \\ \frac{\partial f_{m}}{\partial x_{1}} & \cdots & \frac{\partial f_{m}}{\partial x_{n}} \end{array}\right]
J=[∂x1∂f⋯∂xn∂f]=⎣⎢⎡∂x1∂f1⋮∂x1∂fm⋯⋱⋯∂xn∂f1⋮∂xn∂fm⎦⎥⎤
由上可知，一个实函数
f
:
R
n
→
R
1
\mathbf{f}: \mathbb{R}^{n} \rightarrow \mathbb{R}^{1}
f:Rn→R1的雅可比矩阵是一个行向量，正如下面的二元函数示例所示，这也是SLAM优化问题中最常用的雅克比矩阵形式，后面的海塞矩阵也是采用的
f
:
R
n
→
R
1
\mathbf{f}: \mathbb{R}^{n} \rightarrow \mathbb{R}^{1}
f:Rn→R1函数。
海塞矩阵：
假设某函数
f
:
R
n
→
R
1
\mathbf{f}: \mathbb{R}^{n} \rightarrow \mathbb{R}^{1}
f:Rn→R1，也就是一个实函数
f
(
x
1
,
x
2
,
⋯
,
x
n
)
f\left(x_{1}, x_{2}, \cdots, x_{n}\right)
f(x1,x2,⋯,xn)，此函数的海塞矩阵
H
H
H
是
是
是n×n$的对称矩阵
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
]
H=\left[\begin{array}{cccc} \frac{\partial^{2} f}{\partial x_{1}^{2}} & \frac{\partial^{2} f}{\partial x_{1} \partial x_{2}} & \cdots & \frac{\partial^{2} f}{\partial x_{1} \partial x_{n}} \\ \frac{\partial^{2} f}{\partial x_{2} \partial x_{1}} & \frac{\partial^{2} f}{\partial x_{2}^{2}} & \cdots & \frac{\partial^{2} f}{\partial x_{2} \partial x_{n}} \\ \vdots & \vdots & \ddots & \vdots \\ \frac{\partial^{2} f}{\partial x_{n} \partial x_{1}} & \frac{\partial^{2} f}{\partial x_{n} \partial x_{2}} & \cdots & \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⎦⎥⎥⎥⎥⎥⎤
二元函数示例
例如二元函数
f
(
x
,
y
)
f(x,y)
f(x,y)是关于实数变量
x
x
x和
y
y
y的函数，函数在
(
x
k
,
y
k
)
(x_{k}, y_{k})
(xk,yk)处的领域可以做如下的泰勒展开：
f
(
x
,
y
)
=
f
(
x
k
,
y
k
)
+
(
x
−
x
k
)
f
x
1
(
x
k
,
y
k
)
+
(
y
−
y
k
)
f
y
1
(
x
k
,
y
k
)
+
1
2
!
(
x
−
x
k
)
2
f
x
x
2
(
x
k
,
y
k
)
+
1
2
!
(
x
−
x
k
)
(
y
−
y
k
)
f
x
y
2
(
x
k
,
y
k
)
+
1
2
!
(
x
−
x
k
)
(
y
−
y
k
)
f
y
x
2
(
x
k
,
y
k
)
+
1
2
!
(
y
−
y
k
)
2
f
y
y
2
(
x
k
,
y
k
)
+
o
n
\begin{gathered} f(x, y)=f\left(x_{k}, y_{k}\right)+\left(x-x_{k}\right) f_{x}^{1}\left(x_{k}, y_{k}\right)+\left(y-y_{k}\right) f_{y}^{1}\left(x_{k}, y_{k}\right) \\ +\frac{1}{2 !}\left(x-x_{k}\right)^{2} f_{x x}^{2}\left(x_{k}, y_{k}\right)+\frac{1}{2 !}\left(x-x_{k}\right)\left(y-y_{k}\right) f_{x y}^{2}\left(x_{k}, y_{k}\right) \\ +\frac{1}{2 !}\left(x-x_{k}\right)\left(y-y_{k}\right) f_{y x}^{2}\left(x_{k}, y_{k}\right)+\frac{1}{2 !}\left(y-y_{k}\right)^{2} f_{y y}^{2}\left(x_{k}, y_{k}\right) \\ +o^{n} \end{gathered}
f(x,y)=f(xk,yk)+(x−xk)fx1(xk,yk)+(y−yk)fy1(xk,yk)+2!1(x−xk)2fxx2(xk,yk)+2!1(x−xk)(y−yk)fxy2(xk,yk)+2!1(x−xk)(y−yk)fyx2(xk,yk)+2!1(y−yk)2fyy2(xk,yk)+on
其中：
- f x 1 ( x k , y k ) f_{x}^{1}\left(x_{k}, y_{k}\right) fx1(xk,yk)表示 f f f在 ( x k , y k ) (x_{k}, y_{k}) (xk,yk)处关于 x x x的一阶导数，即 f x 1 ( x k , y k ) = ∂ f ∂ x f_{x}^{1}\left(x_{k}, y_{k}\right)=\frac{\partial f}{\partial x} fx1(xk,yk)=∂x∂f
- f y 1 ( x k , y k ) f_{y}^{1}\left(x_{k}, y_{k}\right) fy1(xk,yk)表示 f f f在 ( x k , y k ) (x_{k}, y_{k}) (xk,yk)处关于 y y y的一阶导数，即 f y 1 ( x k , y k ) = ∂ f ∂ y f_{y}^{1}\left(x_{k}, y_{k}\right)=\frac{\partial f}{\partial y} fy1(xk,yk)=∂y∂f
- f x x 2 ( x k , y k ) f_{x x}^{2}\left(x_{k}, y_{k}\right) fxx2(xk,yk)表示 f f f在 ( x k , y k ) (x_{k}, y_{k}) (xk,yk)处关于 x x x的二阶导数，即 f x x 2 ( x k , y k ) = ∂ 2 f ∂ x 2 f_{x x}^{2}\left(x_{k}, y_{k}\right)=\frac{\partial^{2} f}{\partial x^{2}} fxx2(xk,yk)=∂x2∂2f
- f x y 2 ( x k , y k ) f_{x y}^{2}\left(x_{k}, y_{k}\right) fxy2(xk,yk)表示 f f f在 ( x k , y k ) (x_{k}, y_{k}) (xk,yk)处分别关于 x x x和 y y y的二阶导数，即 f x y 2 ( x k , y k ) = ∂ 2 f ∂ x ∂ y f_{x y}^{2}\left(x_{k}, y_{k}\right)=\frac{\partial^{2} f}{\partial x \partial y} fxy2(xk,yk)=∂x∂y∂2f
- f y x 2 ( x k , y k ) f_{y x}^{2}\left(x_{k}, y_{k}\right) fyx2(xk,yk)表示 f f f在 ( x k , y k ) (x_{k}, y_{k}) (xk,yk)处分别关于 y y y和 x x x的二阶导数，即 f y x 2 ( x k , y k ) = ∂ 2 f ∂ y ∂ x f_{y x}^{2}\left(x_{k}, y_{k}\right)=\frac{\partial^{2} f}{\partial y \partial x} fyx2(xk,yk)=∂y∂x∂2f
- f y y 2 ( x k , y k ) f_{y y}^{2}\left(x_{k}, y_{k}\right) fyy2(xk,yk)表示 f f f在 ( x k , y k ) (x_{k}, y_{k}) (xk,yk)处关于 y y y的二阶导数，即 f y y 2 ( x k , y k ) = ∂ 2 f ∂ y 2 f_{y y}^{2}\left(x_{k}, y_{k}\right)=\frac{\partial^{2} f}{\partial y^{2}} fyy2(xk,yk)=∂y2∂2f
- o n o^{n} on表示三次项往上的余项，对于多元展开一般最多展开到二次项。
令 Δ x 1 = x − x k , Δ x 2 = y − y k , Δ x = [ Δ x 1 Δ x 2 ] \Delta x_{1}=x-x_{k}, \Delta x_{2}=y-y_{k},\Delta x=\left[\begin{array}{l} \Delta x_{1} \\ \Delta x_{2} \end{array}\right] Δx1=x−xk,Δx2=y−yk,Δx=[Δx1Δx2]
将上式写为矩阵的形式：
f
(
x
)
=
f
(
x
k
,
y
k
)
+
∇
f
(
x
k
,
y
k
)
T
Δ
x
+
1
2
Δ
x
T
H
(
x
k
,
y
k
)
Δ
x
+
⋯
f(x)=f\left(x_{k}, y_{k}\right)+\nabla f\left(x_{k}, y_{k}\right)^{\mathrm{T}} \Delta x+\frac{1}{2} \Delta x^{\mathrm{T}} H\left(x_{k}, y_{k}\right) \Delta x+\cdots
f(x)=f(xk,yk)+∇f(xk,yk)TΔx+21ΔxTH(xk,yk)Δx+⋯
其中
∇
f
(
x
k
,
y
k
)
\nabla f\left(x_{k}, y_{k}\right)
∇f(xk,yk)是函数
f
(
x
,
y
)
f(x,y)
f(x,y)在
(
x
k
,
y
k
)
(x_{k}, y_{k})
(xk,yk)处的梯度：
∇
f
(
x
k
,
y
k
)
=
[
∂
f
∂
x
∂
f
∂
y
]
(
x
k
,
y
k
)
\nabla f\left(x_{k}, y_{k}\right)=\left[\begin{array}{l} \frac{\partial f}{\partial x} \\ \frac{\partial f}{\partial y} \end{array}\right]_{(x_{k}, y_{k})}
∇f(xk,yk)=[∂x∂f∂y∂f](xk,yk)
梯度的转置其实就是雅可比矩阵， ∇ f ( x k , y k ) T = J ( x k , y k ) \nabla f\left(x_{k}, y_{k}\right)^{\mathrm{T}}=J\left(x_{k}, y_{k}\right) ∇f(xk,yk)T=J(xk,yk)。
另外，
H
(
x
k
,
y
k
)
H\left(x_{k}, y_{k}\right)
H(xk,yk)
H
(
x
k
,
y
k
)
=
[
∂
2
f
∂
x
2
∂
2
f
∂
x
∂
y
∂
2
f
∂
y
∂
x
∂
2
f
∂
y
2
]
(
x
k
,
y
k
)
H\left(x_{k}, y_{k}\right)=\left[\begin{array}{cc} \frac{\partial^{2} f}{\partial x^{2}} & \frac{\partial^{2} f}{\partial x \partial y} \\ \frac{\partial^{2} f}{\partial y \partial x} & \frac{\partial^{2} f}{\partial y^{2}} \end{array}\right]_{(x_{k}, y_{k})}
H(xk,yk)=[∂x2∂2f∂y∂x∂2f∂x∂y∂2f∂y2∂2f](xk,yk)
也就是海塞矩阵。
总结：
综上所述，对于一个多元函数
f
(
x
)
,
x
=
[
x
1
x
2
⋮
x
n
]
f(\mathbf{x}),\mathbf{x}=\left[\begin{array}{c} x_{1} \\ x_{2} \\ \vdots \\ x_{n} \end{array}\right]
f(x),x=⎣⎢⎢⎢⎡x1x2⋮xn⎦⎥⎥⎥⎤，在
x
0
=
=
[
x
10
x
20
⋮
x
n
0
]
\mathbf{x_0}==\left[\begin{array}{c} x_{10} \\ x_{20} \\ \vdots \\ x_{n0} \end{array}\right]
x0==⎣⎢⎢⎢⎡x10x20⋮xn0⎦⎥⎥⎥⎤处的泰勒展开式为：
f
(
x
)
,
=
f
(
x
0
)
+
J
(
x
0
)
Δ
x
+
1
2
Δ
x
T
H
(
x
0
)
Δ
x
+
⋯
f(\mathbf{x}),=f\left(\mathbf{x_0}\right)+J(\mathbf{x_0}) \Delta \mathbf{x}+\frac{1}{2} \Delta \mathbf{x}^{\mathrm{T}} H\left(\mathbf{x_0}\right) \Delta \mathbf{x}+\cdots
f(x),=f(x0)+J(x0)Δx+21ΔxTH(x0)Δx+⋯
其中：
Δ x = [ Δ x 1 Δ x 2 ⋮ Δ x n ] = [ x 1 − x 10 x 2 − x 20 ⋮ x n − x n 0 ] \Delta\mathbf{x}=\left[\begin{array}{c} \Delta x_{1} \\ \Delta x_{2} \\ \vdots \\ \Delta x_{n} \end{array}\right]=\left[\begin{array}{c} x_{1}-x_{10} \\ x_{2}-x_{20} \\ \vdots \\ x_{n}-x_{n0} \end{array}\right] Δx=⎣⎢⎢⎢⎡Δx1Δx2⋮Δxn⎦⎥⎥⎥⎤=⎣⎢⎢⎢⎡x1−x10x2−x20⋮xn−xn0⎦⎥⎥⎥⎤， J ( x 0 ) = [ ∂ f ∂ x 1 ⋯ ∂ f ∂ x n ] x 0 J(\mathbf{x_0})=\left[\begin{array}{ccc} \frac{\partial f}{\partial x_{1}} & \cdots & \frac{\partial f}{\partial x_{n}} \end{array}\right]_\mathbf{x_0} J(x0)=[∂x1∂f⋯∂xn∂f]x0， H ( x 0 ) = [ ∂ 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 ] x 0 H(\mathbf{x_0})=\left[\begin{array}{cccc} \frac{\partial^{2} f}{\partial x_{1}^{2}} & \frac{\partial^{2} f}{\partial x_{1} \partial x_{2}} & \cdots & \frac{\partial^{2} f}{\partial x_{1} \partial x_{n}} \\ \frac{\partial^{2} f}{\partial x_{2} \partial x_{1}} & \frac{\partial^{2} f}{\partial x_{2}^{2}} & \cdots & \frac{\partial^{2} f}{\partial x_{2} \partial x_{n}} \\ \vdots & \vdots & \ddots & \vdots \\ \frac{\partial^{2} f}{\partial x_{n} \partial x_{1}} & \frac{\partial^{2} f}{\partial x_{n} \partial x_{2}} & \cdots & \frac{\partial^{2} f}{\partial x_{n}^{2}} \end{array}\right]_{\mathbf{x_0}} H(x0)=⎣⎢⎢⎢⎢⎢⎡∂x12∂2f∂x2∂x1∂2f⋮∂xn∂x1∂2f∂x1∂x2∂2f∂x22∂2f⋮∂xn∂x2∂2f⋯⋯⋱⋯∂x1∂xn∂2f∂x2∂xn∂2f⋮∂xn2∂2f⎦⎥⎥⎥⎥⎥⎤x0
泰勒展开在SLAM中的应用
推导罗德里格斯公式
任意矩阵
A
\boldsymbol{A}
A的指数映射也可以像指数函数的泰勒级数一样进行泰勒展开：
exp
(
A
)
=
∑
n
=
0
∞
1
n
!
A
n
\exp (\boldsymbol{A})=\sum_{n=0}^{\infty} \frac{1}{n !} \boldsymbol{A}^{n}
exp(A)=n=0∑∞n!1An
所以SO(3)上的指数映射可以写为：
R
=
exp
(
ϕ
∧
)
=
exp
(
θ
a
∧
)
=
∑
n
=
0
∞
1
n
!
(
θ
a
∧
)
n
R=\exp \left(\phi^{\wedge}\right)=\exp \left(\theta \boldsymbol{a}^{\wedge}\right)=\sum_{n=0}^{\infty} \frac{1}{n !}\left(\theta \boldsymbol{a}^{\wedge}\right)^{n}
R=exp(ϕ∧)=exp(θa∧)=n=0∑∞n!1(θa∧)n
对于单位向量的反对称矩阵有如下性质：
a
∧
a
∧
=
[
−
a
2
2
−
a
3
2
a
1
a
2
a
1
a
3
a
1
a
2
−
a
1
2
−
a
3
2
a
2
a
3
a
1
a
3
a
2
a
3
−
a
1
2
−
a
2
2
]
=
a
a
T
−
I
\boldsymbol{a}^{\wedge} \boldsymbol{a}^{\wedge}=\left[\begin{array}{ccc} -a_{2}^{2}-a_{3}^{2} & a_{1} a_{2} & a_{1} a_{3} \\ a_{1} a_{2} & -a_{1}^{2}-a_{3}^{2} & a_{2} a_{3} \\ a_{1} a_{3} & a_{2} a_{3} & -a_{1}^{2}-a_{2}^{2} \end{array}\right]=\boldsymbol{a} \boldsymbol{a}^{\mathrm{T}}-\boldsymbol{I}
a∧a∧=⎣⎡−a22−a32a1a2a1a3a1a2−a12−a32a2a3a1a3a2a3−a12−a22⎦⎤=aaT−I
a ∧ a ∧ a ∧ = a ∧ ( a a T − I ) = − a ∧ \boldsymbol{a}^{\wedge} \boldsymbol{a}^{\wedge} \boldsymbol{a}^{\wedge}=\boldsymbol{a}^{\wedge}\left(\boldsymbol{a} \boldsymbol{a}^{\mathrm{T}}-\boldsymbol{I}\right)=-\boldsymbol{a}^{\wedge} a∧a∧a∧=a∧(aaT−I)=−a∧
所以
R
=
∑
n
=
0
∞
1
n
!
(
θ
a
∧
)
n
=
I
+
θ
a
∧
+
1
2
!
θ
2
a
∧
a
∧
+
1
3
!
θ
3
a
∧
a
∧
a
∧
+
1
4
!
θ
4
(
a
∧
)
4
+
⋯
=
a
a
T
−
a
∧
a
∧
+
θ
a
∧
+
1
2
!
θ
2
a
∧
a
∧
−
1
3
!
θ
3
a
∧
−
1
4
!
θ
4
(
a
∧
)
2
+
⋯
=
a
T
+
(
θ
−
1
3
!
θ
3
+
1
5
!
θ
5
−
⋯
)
⏟
sin
θ
a
∧
−
(
1
−
1
2
!
θ
2
+
1
4
!
θ
4
−
⋯
)
⏟
cos
θ
a
∧
a
∧
=
a
∧
a
∧
+
I
+
sin
θ
a
∧
−
cos
θ
a
∧
a
∧
=
(
1
−
cos
θ
)
a
∧
a
∧
+
I
+
sin
θ
a
∧
=
cos
θ
I
+
(
1
−
cos
θ
)
a
a
T
+
sin
θ
a
∧
.
\begin{aligned} &R=\sum_{n=0}^{\infty} \frac{1}{n !}\left(\theta \boldsymbol{a}^{\wedge}\right)^{n}\\ &=\boldsymbol{I}+\theta \boldsymbol{a}^{\wedge}+\frac{1}{2 !} \theta^{2} \boldsymbol{a}^{\wedge} \boldsymbol{a}^{\wedge}+\frac{1}{3 !} \theta^{3} \boldsymbol{a}^{\wedge} \boldsymbol{a}^{\wedge} \boldsymbol{a}^{\wedge}+\frac{1}{4 !} \theta^{4}\left(\boldsymbol{a}^{\wedge}\right)^{4}+\cdots \\ &=\boldsymbol{a} \boldsymbol{a}^{\mathrm{T}}-\boldsymbol{a}^{\wedge} \boldsymbol{a}^{\wedge}+\theta \boldsymbol{a}^{\wedge}+\frac{1}{2 !} \theta^{2} \boldsymbol{a}^{\wedge} \boldsymbol{a}^{\wedge}-\frac{1}{3 !} \theta^{3} \boldsymbol{a}^{\wedge}-\frac{1}{4 !} \theta^{4}\left(\boldsymbol{a}^{\wedge}\right)^{2}+\cdots \\ &=\boldsymbol{a}^{\mathrm{T}}+\underbrace{\left(\theta-\frac{1}{3 !} \theta^{3}+\frac{1}{5 !} \theta^{5}-\cdots\right)}_{\sin \theta} \boldsymbol{a}^{\wedge}-\underbrace{\left(1-\frac{1}{2 !} \theta^{2}+\frac{1}{4 !} \theta^{4}-\cdots\right)}_{\cos \theta} \boldsymbol{a}^{\wedge} \boldsymbol{a}^{\wedge} \\ &=\boldsymbol{a}^{\wedge} \boldsymbol{a}^{\wedge}+\boldsymbol{I}+\sin \theta \boldsymbol{a}^{\wedge}-\cos \theta \boldsymbol{a}^{\wedge} \boldsymbol{a}^{\wedge} \\ &=(1-\cos \theta) \boldsymbol{a}^{\wedge} \boldsymbol{a}^{\wedge}+\boldsymbol{I}+\sin \theta \boldsymbol{a}^{\wedge} \\ &=\cos \theta \boldsymbol{I}+(1-\cos \theta) \boldsymbol{a} \boldsymbol{a}^{\mathrm{T}}+\sin \theta \boldsymbol{a}^{\wedge} . \end{aligned}
R=n=0∑∞n!1(θa∧)n=I+θa∧+2!1θ2a∧a∧+3!1θ3a∧a∧a∧+4!1θ4(a∧)4+⋯=aaT−a∧a∧+θa∧+2!1θ2a∧a∧−3!1θ3a∧−4!1θ4(a∧)2+⋯=aT+sinθ
(θ−3!1θ3+5!1θ5−⋯)a∧−cosθ
(1−2!1θ2+4!1θ4−⋯)a∧a∧=a∧a∧+I+sinθa∧−cosθa∧a∧=(1−cosθ)a∧a∧+I+sinθa∧=cosθI+(1−cosθ)aaT+sinθa∧.
罗德里格斯公式得证。
约等式 exp ( A ) ≈ I + A \exp (\boldsymbol{A})\approx I+\boldsymbol{A} exp(A)≈I+A以及 exp ( − A ) ≈ I − A \exp (-\boldsymbol{A})\approx I-\boldsymbol{A} exp(−A)≈I−A
这个约等式是李群李代数中经常用到的约等式，只不过李群李代数中的
A
\boldsymbol{A}
A通常是反对称矩阵，这个约等式就是是上面矩阵指数的截断泰勒展开
exp
(
A
)
=
∑
n
=
0
∞
1
n
!
A
n
≈
I
+
A
\exp (\boldsymbol{A})=\sum_{n=0}^{\infty} \frac{1}{n !} \boldsymbol{A}^{n}\approx I+\boldsymbol{A}
exp(A)=n=0∑∞n!1An≈I+A
扫一扫
854
被折叠的 条评论
为什么被折叠?
到【灌水乐园】发言
