# 李群和李代数

## 问题的引入

[ R , t ] ⟶ Δ R , Δ t [ R ′ , t ′ ] t ′ = t + Δ t R ′ = R + Δ R [R, t] \stackrel{\Delta R, \Delta t}{\longrightarrow} [R', t'] \\ t' = t + \Delta t \\ R' = R + \Delta R

U ( R ) = d U d R = lim ⁡ Δ R → 0 U ( R + Δ R ) − U ( R ) Δ R U(R) = \frac{dU}{dR} = \lim\limits_{\Delta R \to 0}\frac{U(R + \Delta R) - U(R)}{\Delta R}

## 预备知识

### 矩阵指数

x ⋅ = a x ( t ) \stackrel{·}{x} = ax(t)

x ( t ) = e a t x 0 x(t) = e^{at}x_0

e a t = 1 + a t + ( a t ) 2 2 ! + ( a t ) 3 3 ! + ⋅ ⋅ ⋅ e^{at} = 1 + at + \frac{(at)^2}{2!} + \frac{(at)^3}{3!} + ···

x ⋅ ( t ) = A x ( t ) \stackrel{·}{x}(t) = Ax(t)

x ( t ) = e A t x 0 \\ x(t) = e^{At}x_0

e A t = I + A t + ( A t ) 2 2 ! + ( A t ) 3 3 ! + ⋅ ⋅ ⋅ \\ e^{At} = I + At + \frac{(At)^2}{2!} + \frac{(At)^3}{3!} + ···

### 矩阵指数与矩阵旋转的关系

p ⋅ = ω ^ × p \stackrel{·}{p} = \hat{\omega} \times p

p ⋅ = [ ω ^ ] p \stackrel{·}{p} = [\hat{\omega}]p

[ ω ] = [ 0 − ω 3 ω 2 ω 3 0 − ω 1 − ω 2 ω 1 0 ] [\omega] = \begin{bmatrix} 0 & -\omega_{3} & \omega_{2} \\ \omega_{3} & 0 & -\omega_{1} \\ -\omega_{2} & \omega_{1} & 0 \end{bmatrix}

p ( t ) = e [ ω ^ ] t p ( 0 ) p(t) = e^{[\hat\omega]t}p(0)

p ( t ) = e [ ω ^ ] θ p ( 0 ) p(t) = e^{[\hat\omega]\theta}p(0)

e [ ω ^ ] θ = I + [ ω ^ ] θ + [ ω ^ ] 2 θ 2 2 ! + [ ω ^ ] 3 θ 3 3 ! + ⋅ ⋅ ⋅ = I + ( θ − θ 3 3 ! + θ 5 5 ! − ⋅ ⋅ ⋅ ) [ ω ^ ] + （ θ 2 2 ! − θ 4 4 ! + θ 6 6 ! − ⋅ ⋅ ⋅ ） [ ω ^ ] 2 = I + sin ⁡ θ [ ω ^ ] + ( a − cos ⁡ θ ) [ ω ^ ] 2 e^{[\hat\omega]\theta} = I + [\hat\omega]\theta + [\hat\omega]^2\frac{\theta ^ 2}{2!} + [\hat\omega]^3\frac{\theta ^ 3}{3!} + ··· \\ = I + (\theta - \frac{\theta ^ 3}{3!} + \frac{\theta ^ 5}{5!} - ···)[\hat\omega] + （\frac{\theta ^ 2}{2!} - \frac{\theta ^ 4}{4!} + \frac{\theta ^ 6}{6!}-···）[\hat\omega] ^ 2 \\ = I + \sin{\theta}[\hat\omega] + (a-\cos{\theta})[\hat\omega] ^ 2

## 李群和李代数

### 群

• 封闭性： ∀ a 1 , a 2 ∈ A , a 1 ⋅ a 2 ∈ A \forall a_1,a_2 \in \mathbf{A},a_1 \cdot a_2 \in \mathbf{A}
• 结合律： ∀ a 1 , a 2 , a 3 ∈ A , ( a 1 ⋅ a 2 ) ⋅ a 3 = a 1 ⋅ ( a 2 ⋅ a 3 ) \forall a_1,a_2,a_3 \in \mathbf{A}, (a_1 \cdot a_2) \cdot a_3 = a_1\cdot (a_2 \cdot a_3)
• 幺元： ∃ a 0 ∈ A , s . t . ∀ a ∈ A , a 0 ⋅ a = a ⋅ a 0 \exists a_0 \in \mathbf{A}, s.t. \forall a \in \mathbf{A}, a_0 \cdot a = a \cdot a_0
• 逆元： ∀ a ∈ A , ∃ a − 1 ∈ A , s . t .   a ⋅ a − 1 = a 0 \forall a \in \mathbf{A},\exists a^{-1} \in A, \quad s.t. \space a \cdot a^{-1}=a_0

S O ( 3 ) = { R ∈ R 3 × 3 ∣ R R T = I , d e t ( R ) = 1 } SO(3) = \{R \in \mathbb{R}^{3 \times 3} | RR^T = I, det(R) = 1\}

S E ( 3 ) = { T = [ R t 0 T 1 ] ∈ R 4 × 4 ∣ R ∈ S O ( 3 ) , t ∈ R 3 × 3 } SE(3) = \{ T = \begin{bmatrix} R & t \\ {0^T} & 1 \\ \end{bmatrix} \in \mathbb{R}^{4\times 4} | R \in SO(3), t \in \mathbb{R}^{3\times 3}\}

### 李群

R ⋅ ( t ) R ( t ) T + R ( t ) R ⋅ ( t ) T = 0 \stackrel{·}{R}(t)R(t)^T + R(t)\stackrel{·}{R}(t)^T = 0

R ⋅ ( t ) R ( t ) T = − ( R ( t ) R ⋅ ( t ) T ) T \stackrel{·}{R}(t)R(t)^T = -{( R(t)\stackrel{·}{R}(t)^T )}^T

a ∧ = A = [ 0 − a 3 a 2 a 3 0 − a 1 − a 2 a 1 0 ] , A ∨ = a = { a 1 , a 2 , a 3 } . a^{\wedge} = A = \begin{bmatrix} 0 & -a_3 & a_2 \\ a_3 & 0 & -a_1 \\ -a_2 & a_1 & 0 \\ \end{bmatrix}, \quad A^{\vee} = a = \{a_1, a_2, a_3\}.

R ⋅ ( t ) = ϕ ∧ ( t ) R ( t ) \stackrel{·}{R}(t) = \phi^{\wedge}(t)R(t)

R ( t ) = e ϕ ∧ t R(t) = e^{\phi ^ \wedge t}

### 李代数

• 封闭性： ∀ X , Y ∈ V ,   [ X , Y ] ∈ V \forall \mathbf{X}, \mathbf{Y} \in V,\ [\mathbf{X}, \mathbf{Y}] \in V
• 双线性： ∀ X , Y , Z ∈ V ,   a , b ∈ F \forall \mathbf{X}, \mathbf{Y}, \mathbf{Z} \in V,\ a, b \in F , 有 [ a X + b Y ,   Z ] = a [ X , Z ] + [ Y , Z ] ,   [ Z , a X + b Y ] = a [ Z , X ] + b [ Z , Y ] [a\mathbf{X} +b\mathbf{Y},\ \mathbf{Z}]=a[\mathbf{X}, \mathbf{Z}] + [\mathbf{Y}, \mathbf{Z}],\ [\mathbf{Z},a\mathbf{X} +b\mathbf{Y}]=a[\mathbf{Z},\mathbf{X}] +b[\mathbf{Z},\mathbf{Y}]
• 自反性： ∀ X ∈ V ,   [ X , X ] = 0 \forall \mathbf{X} \in V,\ [\mathbf{X}, \mathbf{X}]=\mathbf{0}
• 雅可比等价： ∀ X , Y , Z ∈ V ,   [ X , [ Y , Z ] ] + [ Z , [ X , Y ] ] + [ Y , [ Z , X ] ] = 0 \forall \mathbf{X},\mathbf{Y},\mathbf{Z} \in V,\ [\mathbf{X},[\mathbf{Y},\mathbf{Z}]] + [\mathbf{Z},[\mathbf{X},\mathbf{Y}]] +[\mathbf{Y},[\mathbf{Z},\mathbf{X}]]=\mathbf{0}

### 李代数 s o ( 3 ) so(3)

[ ϕ 1 , ϕ 2 ] = ( Φ 1 Φ 2 − Φ 2 Φ 1 ) ∨ [\phi_1, \phi_2] = (\mathbf{\Phi}_1\mathbf{\Phi}_2 - \mathbf{\Phi}_2\mathbf{\Phi}_1)^{\vee}

[ ϕ 1 , ϕ 2 ] = ϕ 1 × ϕ 2 [\phi_1, \phi_2] = \phi_1 \times \phi_2

### 李代数 s e ( 3 ) se(3)

s o ( 3 ) so(3) 不同的是， s e ( 3 ) se(3) 定义在 R 6 \mathbb{R}^6 空间中。

s e ( 3 ) = { ξ = [ ρ ϕ ] ∈ R 6 , ρ ∈ R 3 , ϕ ∈ s o ( 3 ) , ξ ∧ = [ ϕ ∧ ρ 0 T 0 ] ∈ R 4 × 4 } se(3)=\begin{Bmatrix} \mathbf{\xi}=\begin{bmatrix} \mathbf{\rho} \\ \mathbf{\phi} \end{bmatrix} \in R^6,\mathbf{\rho} \in R^3, \mathbf{\phi} \in so(3), \mathbf{\xi}^\wedge=\begin{bmatrix}\phi^\wedge & \mathbf{\rho} \\ \mathbf{0}^T & 0 \end{bmatrix} \in R^{4 \times 4} \end{Bmatrix}

[ ξ 1 , ξ 2 ] = ( ξ 1 ∧ ξ 2 ∧ − ξ 2 ∧ ξ 1 ∧ ) ∨ [\xi_1, \xi_2]=(\xi_1^\wedge\xi_2^\wedge-\xi_2^\wedge\xi_1^\wedge)^\vee

[ ξ 1 , ξ 2 ] = { Φ 1 ∧ ρ 2 − Φ 2 ∧ ρ 1 0 3 × 3 } [\xi_1, \xi_2] = \begin{Bmatrix} \Phi_1^{\wedge}\rho_2 - \Phi_2^{\wedge}\rho_1 \\ \mathbf{0}_{3\times 3} \end{Bmatrix}

## 指数映射和对数映射

### S O ( 3 ) SO(3) 上的指数映射

R ⋅ ( t ) = ϕ ∧ ( t ) R ( t ) \stackrel{·}{R}(t) = \phi^{\wedge}(t)R(t)

e x p ( ϕ ∧ ) = ∑ n = 0 ∞ 1 n ! ( ϕ ∧ ) n exp(\phi^{\wedge}) = \sum\limits_{n=0}^\infty\frac{1}{n!}(\phi^\wedge)^n

a ∧ a ∧ = a a T − I a ∧ a ∧ a ∧ = − a ∧ \mathbf{a}^\wedge \mathbf{a}^\wedge=\mathbf{a}\mathbf{a}^T-\mathbf{I} \\ \mathbf{a}^\wedge \mathbf{a}^\wedge \mathbf{a}^\wedge=-\mathbf{a}^\wedge

exp ⁡ ϕ ∧ = exp ⁡ ( θ a ) = ∑ n = 0 ∞ 1 n ! ( θ a ∧ ) n = cos ⁡ θ I + ( 1 − cos ⁡ θ ) a a T + sin ⁡ θ a T \exp{\phi^\wedge}=\exp(\theta \mathbf{a})=\sum\limits_{n=0}^\infty \frac{1}{n!}(\theta \mathbf{a}^\wedge)^n \\ =\cos\theta\mathbf{I} + (1-\cos\theta)\mathbf{a}\mathbf{a}^T + \sin\theta\mathbf{a}^T

ϕ = ln ⁡ ( R ) ∨ = ( ∑ n = 0 ∞ ( − 1 ) n n + 1 ( R − I ) n + 1 ) ∨ \phi=\ln(R)^\vee=(\sum\limits_{n=0}^\infty\frac{(-1)^n}{n+1}(\mathbf{R}-\mathbf{I})^{n+1})^\vee

θ = arccos ⁡ t r ( R ) − 1 2 R n = n \theta = \arccos{\frac{tr(\mathbf{R}) - 1}{2}} \\ \mathbf{R}n = n

### S E ( 3 ) SE(3) 上的指数映射

exp ⁡ ( ξ ∧ ) = [ ∑ n = 0 ∞ 1 n ! ( ϕ ∧ ) n ∑ n = 0 ∞ 1 ( n + 1 ) ! ( ϕ ∧ ) n ρ 0 T 1 ] = [ R J ρ 0 T 1 ] = T \exp(\xi^\wedge)=\begin{bmatrix} \sum\limits_{n=0}^\infty\frac{1}{n!}(\phi^\wedge)^n & \sum\limits_{n=0}^\infty\frac{1}{(n+1)!}(\phi^\wedge)^n\rho \\ \mathbf{0}^T & 1 \\ \end{bmatrix}= \begin{bmatrix} \mathbf{R} & \mathbf{J}\rho \\ \mathbf{0}^T & 1 \\ \end{bmatrix}=\mathbf{T}

J = sin ⁡ θ θ I + ( 1 − sin ⁡ θ θ ) a a T + 1 − cos ⁡ θ θ a ∧ \mathbf{J} = \frac{\sin\theta}{\theta}\mathbf{I}+(1-\frac{\sin\theta}{\theta})\mathbf{a}\mathbf{a}^T+\frac{1-\cos\theta}{\theta}\mathbf{a}^\wedge

## 李代数求导与扰动模型

### 李代数与李群的计算对应关系

e x p ( ϕ 1 ∧ ) e x p ( ϕ 2 ∧ ) = e x p ( ( ϕ 1 + ϕ 2 ) ∧ ) exp(\phi_1^{\wedge})exp(\phi_2^{\wedge}) = exp({(\phi_1 + \phi_2)}^{\wedge})

ln ⁡ ( exp ⁡ ( A ) exp ⁡ ( B ) ) = A + B + 1 12 [ A , B ] + 1 12 [ A , [ A , B ] ] − 1 12 [ B , [ A , B ] ] + … \ln(\exp(\mathbf{A})\exp(\mathbf{B}))=\mathbf{A}+\mathbf{B}+\frac{1}{12}[\mathbf{A},\mathbf{B}]+\frac{1}{12}[\mathbf{A},[\mathbf{A},\mathbf{B}]]-\frac{1}{12}[\mathbf{B},[\mathbf{A},\mathbf{B}]]+\ldots

ln ⁡ ( exp ⁡ ( ϕ 1 ∧ ) exp ⁡ ( ϕ 2 ∧ ) ) ∨ ≈ { J l ( ϕ 2 ) − 1 ϕ 1 + ϕ 2 ,   当 ϕ 1 为 小 量 J r ( ϕ 1 ) − 1 ϕ 2 + ϕ 1 ,   当 ϕ 2 为 小 量 \ln(\exp(\phi^\wedge_1)\exp(\phi^\wedge_2))^\vee \approx \begin{cases} \mathbf{J}_l(\phi_2)^{-1}\phi_1 + \phi2,\ 当\phi_1 为小量\\ \mathbf{J}_r(\phi_1)^{-1}\phi_2 + \phi_1,\ 当\phi_2 为小量\\ \end{cases}

J l = J = sin ⁡ θ θ I + ( 1 − sin ⁡ θ θ ) a a T + 1 − cos ⁡ θ θ a ∧ \mathbf{J}_l = \mathbf{J} = \frac{\sin\theta}{\theta}\mathbf{I}+(1-\frac{\sin\theta}{\theta})\mathbf{a}\mathbf{a}^T+\frac{1-\cos\theta}{\theta}\mathbf{a}^\wedge

J l − 1 = θ 2 cot ⁡ θ 2 I + ( 1 − θ 2 cot ⁡ θ 2 ) a a T − θ 2 a ∧ \mathbf{J}_l^{-1}=\frac{\theta}{2}\cot\frac{\theta}{2}\mathbf{I}+(1-\frac{\theta}{2}\cot\frac{\theta}{2})\mathbf{a}\mathbf{a}^T-\frac{\theta}{2}\mathbf{a}^\wedge

J r ( ϕ ) = J l ( ϕ ) \mathbf{J}_r(\phi) = \mathbf{J}_l(\phi)

exp ⁡ ( ϕ ∧ ) exp ⁡ ( Δ ϕ ∧ ) = exp ⁡ ( ( ϕ + J l ( ϕ ) − 1 Δ ϕ ) ∧ ) \exp(\phi^\wedge)\exp(\Delta\phi^\wedge)=\exp((\phi+\mathbf{J}_l(\phi)^{-1}\Delta\phi)^\wedge)

e x p ( ( ϕ + Δ ϕ ) ∧ ) = e x p ( ( J l Δ ϕ ) ∧ ) e x p ( ϕ ∧ ) = e x p ( ϕ ∧ ) e x p ( ( J r Δ ϕ ) ∧ ) exp((\phi + \Delta \phi)^\wedge) = exp((\mathbf{J_l}\Delta\phi)^\wedge)exp(\phi^\wedge) = exp(\phi^\wedge)exp((\mathbf{J_r}\Delta\phi)^\wedge)

1. 利用李代数表示位子，然后根据李代数的加法对李代数进行求导。（导数模型）
2. 对李群来左乘或右乘微小扰动，并对这个扰动的李代数求导。（扰动模型）

### 求导模型

∂ ( R p ) ∂ R \frac{\partial(\mathbf{R}\mathbf{p})}{\partial\mathbf{R}}

R \mathbf{R} 对应的李代数为 ϕ \phi ，就可以写成：

∂ ( R p ) ∂ R = ∂ ( exp ⁡ ( ϕ ∧ ) p ) ∂ ϕ \frac{\partial(\mathbf{R}\mathbf{p})}{\partial\mathbf{R}} = \frac{\partial(\exp(\phi^\wedge)\mathbf{p})}{\partial\phi}

∂ ( exp ⁡ ( ϕ ∧ ) p ) ∂ ϕ = lim ⁡ δ ϕ → 0 exp ⁡ ( ( ϕ + δ ϕ ) ∧ ) p − exp ⁡ ( ϕ ∧ ) p δ ϕ = − ( R p ) ∧ J l \frac{\partial(\exp(\phi^\wedge)\mathbf{p})}{\partial\phi}=\lim_{\delta\phi\to0}\frac{\exp((\phi+\delta\phi)^\wedge)\mathbf{p}-\exp(\phi^\wedge)\mathbf{p}}{\delta\phi}=-(\mathbf{R}\mathbf{p})^\wedge\mathbf{J}_l

### 扰动模型

∂ ( T p ) ∂ δ ξ = lim ⁡ δ ξ → 0 exp ⁡ ( δ ξ ∧ ) exp ⁡ ( ξ ∧ ) p − exp ⁡ ( ξ ∧ ) p δ ξ = [ I 3 × 3 − ( R p + t ) 3 × 3 ∧ 0 1 × 3 T 0 1 × 3 T ] = ( T p ) ⨀ \frac{\partial(\mathbf{T}\mathbf{p})}{\partial\delta\xi}=\lim_{\delta\xi\to0}\frac{\exp(\delta\xi^\wedge)\exp(\xi^\wedge)\mathbf{p}-\exp(\xi^\wedge)\mathbf{p}}{\delta\xi}= \begin{bmatrix} \mathbf{I}_{3\times3} & -(\mathbf{R}\mathbf{p} + \mathbf{t})^\wedge_{3\times3} \\ \mathbf{0}^T_{1\times3} & \mathbf{0}^T_{1\times3}\\ \end{bmatrix}=(\mathbf{T}\mathbf{p})^{\bigodot}

• 点赞
• 评论
• 分享
x

海报分享

扫一扫，分享海报

• 收藏
• 手机看

分享到微信朋友圈

x

扫一扫，手机阅读

• 打赏

打赏

Cpp快餐包

我不恰饭，你们多赞

C币 余额
2C币 4C币 6C币 10C币 20C币 50C币
• 一键三连

点赞Mark关注该博主, 随时了解TA的最新博文

04-20 5442
10-16 4568
07-03 1007
05-17 1696
09-09 46