圆锥曲线轨道
轨道六根数指:
半长轴(a)、偏心率(e),轨道倾角(i),升交点赤经(
Ω
\Omega
Ω)、近地点幅角(
ω
\omega
ω)、真近点角(
ϕ
\phi
ϕ)。
其它参数:
角动量
h
⃗
\vec{h}
h、半通径
p
p
p、远拱点
r
a
r_a
ra、周期
T
T
T。
E
=
−
μ
2
a
=
v
2
2
−
μ
r
=
v
i
n
f
2
2
h
⃗
=
r
⃗
×
v
⃗
∣
h
⃗
∣
=
∣
r
⃗
∣
∣
v
⃗
∣
sin
θ
a
=
p
1
−
e
2
=
−
μ
2
E
=
μ
r
2
μ
−
r
v
2
p
=
h
2
μ
e
⃗
=
v
⃗
×
h
⃗
μ
−
r
⃗
∣
r
⃗
∣
=
(
v
2
μ
−
1
r
)
r
⃗
−
v
⃗
⋅
r
⃗
μ
v
⃗
e
=
1
−
h
2
a
μ
r
=
p
1
+
e
cos
f
f
=
arccos
(
1
e
(
h
2
μ
r
−
1
)
)
r
a
=
a
(
1
+
e
)
T
=
2
π
a
3
μ
\begin{aligned} \mathcal{E} =& -\frac{\mu}{2a}=\frac{v^2}{2}-\frac{\mu}{r} = \frac{v_{inf}^2}{2} \\ \vec{h} =& \vec{r}\times\vec{v} \\ |\vec{h}| =& |\vec{r}||\vec{v}|\sin\theta \\ a =& \frac{p}{1-e^2}=-\frac{\mu}{2\mathcal{E}} = \frac{\mu r}{2\mu-rv^2} \\ p =& \frac{h^2}{\mu} \\ \vec{e} =& \frac{\vec{v}\times\vec{h}}{\mu} - \frac{\vec{r}}{|\vec{r}|} = (\frac{v^2}{\mu}-\frac{1}{r})\vec{r} - \frac{\vec{v}\cdot\vec{r}}{\mu}\vec{v} \\ e =& \sqrt{1-\frac{h^2}{a\mu}} \\ r =& \frac{p}{1+e\cos f} \\ f =& \arccos(\frac{1}{e}(\frac{h^2}{\mu r}-1)) \\ r_a =& a(1+e) \\ T =& 2\pi\sqrt{\frac{a^3}{\mu}} \end{aligned}
E=h=∣h∣=a=p=e=e=r=f=ra=T=−2aμ=2v2−rμ=2vinf2r×v∣r∣∣v∣sinθ1−e2p=−2Eμ=2μ−rv2μrμh2μv×h−∣r∣r=(μv2−r1)r−μv⋅rv1−aμh21+ecosfparccos(e1(μrh2−1))a(1+e)2πμa3
由轨道元素计算位置和速度向量
R
=
[
c
Ω
c
ω
−
s
Ω
c
i
s
ω
−
c
Ω
s
ω
−
s
Ω
c
i
c
ω
s
Ω
s
i
s
Ω
c
ω
+
c
Ω
c
i
s
ω
−
s
Ω
s
ω
+
c
Ω
c
i
c
ω
−
c
Ω
s
i
s
i
s
ω
s
i
c
ω
c
i
]
r
⃗
=
R
[
a
(
1
−
e
2
)
1
+
e
cos
θ
cos
θ
a
(
1
−
e
2
)
1
+
e
cos
θ
sin
θ
0
]
v
⃗
=
μ
a
(
1
−
e
2
)
R
[
−
sin
θ
e
+
cos
θ
0
]
\begin{aligned} R =& \begin{bmatrix} c_\Omega c_\omega-s_\Omega c_i s_\omega & -c_\Omega s_\omega-s_\Omega c_i c_\omega & s_\Omega s_i \\ s_\Omega c_\omega+c_\Omega c_i s_\omega & -s_\Omega s_\omega+c_\Omega c_i c_\omega & -c_\Omega s_i \\ s_i s_\omega & s_i c_\omega & c_i \end{bmatrix} \\ \vec{r} =& R\begin{bmatrix} \displaystyle\frac{a(1-e^2)}{1+e\cos\theta}\cos\theta \\ \displaystyle\frac{a(1-e^2)}{1+e\cos\theta}\sin\theta \\ 0 \end{bmatrix} \\ \vec{v} =& \sqrt{\frac{\mu}{a(1-e^2)}}R\begin{bmatrix} -\sin\theta \\ e+\cos\theta \\ 0 \end{bmatrix} \end{aligned}
R=r=v=
cΩcω−sΩcisωsΩcω+cΩcisωsisω−cΩsω−sΩcicω−sΩsω+cΩcicωsicωsΩsi−cΩsici
R
1+ecosθa(1−e2)cosθ1+ecosθa(1−e2)sinθ0
a(1−e2)μR
−sinθe+cosθ0
平近点角(mean anomaly,
M
M
M)
真近点角(true anomaly,
θ
\theta
θ)
偏近点角(eccentric anomaly,
E
E
E)
E
−
e
sin
E
=
M
tan
E
2
tan
θ
2
=
1
−
e
1
+
e
\begin{aligned} & E-e\sin E=M \\ & \frac{\tan\frac{E}{2}}{\tan\frac{\theta}{2}} = \sqrt{\frac{1-e}{1+e}} \end{aligned}
E−esinE=Mtan2θtan2E=1+e1−e
根据平近点角计算真近点角
θ
=
M
+
(
2
e
−
1
4
e
3
)
sin
M
+
5
4
e
2
sin
2
M
+
13
12
e
3
sin
3
M
+
O
(
e
4
)
\begin{aligned} \theta =& M+\left(2e-{\frac {1}{4}}e^{3}\right)\sin {M} + {\frac {5}{4}}e^{2}\sin {2M} \\ &+ {\frac {13}{12}}e^{3}\sin {3M}+O(e^{4}) \end{aligned}
θ=M+(2e−41e3)sinM+45e2sin2M+1213e3sin3M+O(e4)
已知
μ
\mu
μ、初始位置
r
⃗
\vec{r}
r 和速度
v
⃗
\vec{v}
v 计算经过指定时间
T
0
T_0
T0 后的位置和速度
- 根据初始位置速度计算轨道能量 E \mathcal{E} E 和偏心率 e ⃗ \vec{e} e
- 根据 μ \mu μ 和 E \mathcal{E} E 计算半长轴 a a a 进而计算轨道周期 T T T
- 根据 T 0 T_0 T0 和 T T T 计算平近点角进而计算偏近点角 E E E
- 根据 e ⃗ \vec{e} e 计算半短轴 b b b 和焦距 c c c,由椭圆参数方程得到 r ⃗ \vec{r} r
- 由 E \mathcal{E} E 计算速度大小 ∣ v ⃗ ∣ |\vec{v}| ∣v∣
- 由偏心率公式计算速度 v ⃗ \vec{v} v
[ e 0 0 ] = ( v ⃗ × h ⃗ μ − r ⃗ ∣ r ⃗ ∣ ) [ r x r y 0 ] − v x r x + v y r y μ [ v x v y 0 ] k = v ⃗ × h ⃗ μ − r ⃗ ∣ r ⃗ ∣ r x v x 2 + r y v x v y = μ ( k r x − e ) r y v y 2 + r x v x v y = μ k r y r x 2 v x 2 + r x r y v x v y = μ ( k r x − e ) r x r y 2 v y 2 + r x r y v x v y = μ k r y r y r x 2 v x 2 − r y 2 v y 2 = μ ( k r x − e ) r x − μ k r y r y = k 1 v x 2 + v y 2 = v 2 v x 2 = r y 2 v 2 + k 1 r 2 v y 2 = r x 2 v 2 − k 1 r 2 k 1 = μ k ( r x 2 − r y 2 ) − e r x = ( v 2 − μ r ) ( r x 2 − r y 2 ) − μ e r x \begin{aligned} &\left[\begin{matrix} e \\ 0 \\ 0 \end{matrix}\right] = (\frac{\vec{v}\times\vec{h}}{\mu} - \frac{\vec{r}}{|\vec{r}|}) \left[\begin{matrix} r_x \\ r_y \\ 0 \end{matrix}\right] - \frac{v_xr_x+v_yr_y}{\mu} \left[\begin{matrix} v_x \\ v_y \\ 0 \end{matrix}\right] \\ & k = \frac{\vec{v}\times\vec{h}}{\mu} - \frac{\vec{r}}{|\vec{r}|} \\ & r_xv_x^2+r_yv_xv_y = \mu(kr_x-e) \\ & r_yv_y^2+r_xv_xv_y = \mu kr_y \\ & r_x^2v_x^2+r_xr_yv_xv_y = \mu(kr_x-e)r_x \\ & r_y^2v_y^2+r_xr_yv_xv_y = \mu kr_yr_y \\ & r_x^2v_x^2-r_y^2v_y^2 = \mu(kr_x-e)r_x - \mu kr_yr_y = k_1 \\ & v_x^2+v_y^2 = v^2 \\ & v_x^2 = \frac{r_y^2v^2 + k_1}{r^2} \\ & v_y^2 = \frac{r_x^2v^2 - k_1}{r^2} \\ &k_1 = \mu k(r_x^2-r_y^2) - er_x = (v^2 - \frac{\mu}{r})(r_x^2-r_y^2) - \mu er_x \end{aligned} e00 =(μv×h−∣r∣r) rxry0 −μvxrx+vyry vxvy0 k=μv×h−∣r∣rrxvx2+ryvxvy=μ(krx−e)ryvy2+rxvxvy=μkryrx2vx2+rxryvxvy=μ(krx−e)rxry2vy2+rxryvxvy=μkryryrx2vx2−ry2vy2=μ(krx−e)rx−μkryry=k1vx2+vy2=v2vx2=r2ry2v2+k1vy2=r2rx2v2−k1k1=μk(rx2−ry2)−erx=(v2−rμ)(rx2−ry2)−μerx
四元数、欧拉角、旋转矩阵
旋转坐标系下的速度和加速度
QUEST 算法
QUEST(QUaternion ESTimator)已知多个参考向量
y
k
y_k
yk 和对应的多个观测向量
x
k
x_k
xk,求旋转矩阵
A
A
A 使得
y
=
R
x
y=Rx
y=Rx 的误差最小,即最小化指标函数
J
0
=
1
2
∑
k
=
1
n
α
k
∣
y
k
−
A
x
k
∣
2
J_0=\frac{1}{2}\sum_{k=1}^n\alpha_k|y_k-Ax_k|^2
J0=21k=1∑nαk∣yk−Axk∣2
或者最大化指标函数
J
=
∑
k
=
1
n
α
k
y
k
T
A
x
k
J=\sum_{k=1}^n\alpha_ky_k^\text{T}Ax_k
J=k=1∑nαkykTAxk
其中
y
k
y_k
yk 和
x
k
x_k
xk 均为单位向量,
α
k
≥
0
\alpha_k\ge 0
αk≥0 为加权因子,
∑
α
k
=
1
\sum\alpha_k=1
∑αk=1,并且
J
=
1
−
J
0
J=1-J_0
J=1−J0。
一组数据的情况
首先研究一个向量
y
=
A
x
y=Ax
y=Ax,设
B
=
y
x
T
B=yx^\text{T}
B=yxT,利用关于矩阵的迹的两个公式
x
T
y
=
tr
(
y
x
T
)
=
tr
(
x
y
T
)
tr
(
A
B
)
=
tr
(
B
A
)
x^\text{T}y=\text{tr}(yx^\text{T})=\text{tr}(xy^\text{T}) \\ \text{tr}(AB)=\text{tr}(BA)
xTy=tr(yxT)=tr(xyT)tr(AB)=tr(BA)
则
J
=
y
T
A
x
=
tr
(
y
(
A
x
)
T
)
=
tr
(
y
x
T
A
T
)
=
tr
(
B
A
T
)
=
tr
(
A
B
T
)
\begin{aligned} J=y^\text{T}Ax=\text{tr}(y(Ax)^\text{T})=\text{tr}(yx^\text{T}A^\text{T}) =\text{tr}(BA^\text{T})=\text{tr}(AB^\text{T}) \end{aligned}
J=yTAx=tr(y(Ax)T)=tr(yxTAT)=tr(BAT)=tr(ABT)
将四元数与旋转矩阵的关系 (四元数、欧拉角、旋转矩阵笔记)
R
=
(
w
2
−
q
v
T
q
v
)
I
+
2
q
v
q
v
T
−
2
w
q
v
∧
R=(w^2-q_v^\text{T}q_v)I+2q_vq_v^\text{T}-2wq_v^{\wedge}
R=(w2−qvTqv)I+2qvqvT−2wqv∧
代入
J
J
J 得
J
=
tr
(
[
(
w
2
−
q
v
T
q
v
)
I
+
2
q
v
q
v
T
−
2
w
q
v
∧
]
B
T
)
=
(
w
2
−
q
v
T
q
v
)
tr
(
B
)
+
2
tr
(
q
v
q
v
T
B
T
)
−
2
w
tr
(
q
v
∧
B
T
)
\begin{aligned} J =& \text{tr}([(w^2-q_v^\text{T}q_v)I +2q_vq_v^\text{T}-2wq_v^{\wedge}]B^\text{T}) \\ =& (w^2-q_v^\text{T}q_v)\text{tr}(B)+2\text{tr}(q_vq_v^\text{T}B^\text{T}) -2w\text{tr}(q_v^{\wedge}B^\text{T}) \end{aligned}
J==tr([(w2−qvTqv)I+2qvqvT−2wqv∧]BT)(w2−qvTqv)tr(B)+2tr(qvqvTBT)−2wtr(qv∧BT)
记
σ
=
tr
(
B
)
S
=
B
+
B
T
=
y
x
T
+
x
y
T
Z
=
y
×
x
\begin{aligned} \sigma =& \text{tr}(B) \\ S =& B+B^\text{T} = yx^\text{T}+xy^\text{T} \\ Z =& y\times x \end{aligned}
σ=S=Z=tr(B)B+BT=yxT+xyTy×x
则
J
=
(
w
2
−
q
v
T
q
v
)
σ
+
2
tr
(
q
v
q
v
T
x
y
T
)
−
2
w
tr
(
q
v
∧
x
y
T
)
=
w
2
σ
−
q
v
T
q
v
σ
+
tr
(
x
y
T
q
v
q
v
T
)
+
q
v
T
x
tr
(
q
v
y
T
)
−
2
w
tr
(
q
v
∧
x
y
T
)
=
w
2
σ
−
q
v
T
q
v
σ
+
y
T
q
v
tr
(
x
q
v
T
)
+
q
v
T
x
tr
(
y
T
q
v
)
−
2
w
y
T
q
v
∧
x
=
w
2
σ
−
σ
q
v
T
q
v
+
q
v
T
y
x
T
q
v
+
q
v
T
x
y
T
q
v
+
2
w
q
v
T
y
∧
x
=
w
2
σ
−
σ
q
v
T
q
v
+
q
v
T
S
q
v
+
2
w
q
v
T
(
y
×
x
)
=
w
2
σ
+
q
v
T
Z
w
+
w
Z
T
q
v
+
q
v
T
S
q
v
−
σ
q
v
T
q
v
=
(
w
σ
+
q
v
T
Z
)
w
+
(
w
Z
T
+
q
v
T
(
S
−
σ
I
)
)
q
v
=
[
w
q
v
T
]
[
σ
Z
T
Z
S
−
σ
I
]
[
w
q
v
]
=
q
T
K
q
\begin{aligned} J=& (w^2-q_v^\text{T}q_v)\sigma+2\text{tr}(q_vq_v^\text{T}xy^\text{T}) -2w\text{tr}(q_v^{\wedge}xy^\text{T}) \\ =& w^2\sigma-q_v^\text{T}q_v\sigma+\text{tr}(xy^\text{T}q_vq_v^\text{T}) +q_v^\text{T}x\text{tr}(q_vy^\text{T}) -2w\text{tr}(q_v^{\wedge}xy^\text{T}) \\ =& w^2\sigma-q_v^\text{T}q_v\sigma+y^\text{T}q_v\text{tr}(xq_v^\text{T}) +q_v^\text{T}x\text{tr}(y^\text{T}q_v) -2wy^\text{T}q_v^{\wedge}x \\ =& w^2\sigma-\sigma q_v^\text{T}q_v+q_v^\text{T}yx^\text{T}q_v +q_v^\text{T}xy^\text{T}q_v+2wq_v^\text{T}y^{\wedge}x \\ =& w^2\sigma-\sigma q_v^\text{T}q_v+q_v^\text{T}Sq_v+2wq_v^\text{T}(y\times x) \\ =& w^2\sigma+q_v^\text{T}Zw+wZ^\text{T}q_v +q_v^\text{T}Sq_v-\sigma q_v^\text{T}q_v \\ =& (w\sigma+q_v^\text{T}Z)w+(wZ^\text{T}+q_v^\text{T}(S-\sigma I))q_v \\ =& \begin{bmatrix} w & q_v^\text{T} \end{bmatrix} \begin{bmatrix} \sigma & Z^\text{T} \\ Z & S-\sigma I \\ \end{bmatrix} \begin{bmatrix} w \\ q_v \end{bmatrix} \\ =& q^\text{T}Kq \end{aligned}
J=========(w2−qvTqv)σ+2tr(qvqvTxyT)−2wtr(qv∧xyT)w2σ−qvTqvσ+tr(xyTqvqvT)+qvTxtr(qvyT)−2wtr(qv∧xyT)w2σ−qvTqvσ+yTqvtr(xqvT)+qvTxtr(yTqv)−2wyTqv∧xw2σ−σqvTqv+qvTyxTqv+qvTxyTqv+2wqvTy∧xw2σ−σqvTqv+qvTSqv+2wqvT(y×x)w2σ+qvTZw+wZTqv+qvTSqv−σqvTqv(wσ+qvTZ)w+(wZT+qvT(S−σI))qv[wqvT][σZZTS−σI][wqv]qTKq
其中第3行到第4行
y
T
q
v
∧
x
=
−
q
v
T
y
∧
x
y^\text{T}q_v^{\wedge}x = -q_v^\text{T}y^{\wedge}x
yTqv∧x=−qvTy∧x 实际上等价于
a
⋅
(
b
×
c
)
=
b
⋅
(
c
×
a
)
=
c
⋅
(
a
×
b
)
a\cdot (b\times c) = b\cdot (c\times a) = c\cdot (a\times b)
a⋅(b×c)=b⋅(c×a)=c⋅(a×b)
考虑到单位四元数
q
q
q 满足
∣
q
∣
=
1
|q|=1
∣q∣=1,使用拉格朗日乘子法,
L
=
q
T
K
q
+
λ
(
q
T
q
−
1
)
∂
L
∂
q
=
2
K
q
−
2
λ
q
=
0
\begin{aligned} L =& q^\text{T}Kq+\lambda(q^\text{T}q-1) \\ \frac{\partial L}{\partial q} =& 2Kq-2\lambda q=0 \end{aligned}
L=∂q∂L=qTKq+λ(qTq−1)2Kq−2λq=0
进一步扩展到多组数据并加入权重,取
B
=
∑
k
=
1
n
α
k
y
k
x
k
T
σ
=
tr
(
B
)
S
=
B
+
B
T
Z
=
∑
k
=
1
n
α
k
y
k
×
x
k
\begin{aligned} B =& \sum_{k=1}^n \alpha_k y_k x_k^\text{T} \\ \sigma =& \text{tr}(B) \\ S =& B+B^\text{T} \\ Z =& \sum_{k=1}^n \alpha_k y_k \times x_k \end{aligned}
B=σ=S=Z=k=1∑nαkykxkTtr(B)B+BTk=1∑nαkyk×xk