坐标系转换--二维七参数大地坐标系转换模型变换关系理解

二维七参数大地坐标系转换模型变换关系理解

$$\begin{gathered} \left[\begin{array}{c}\Delta L\\ \Delta B\end{array}\right]=\left[\begin{array}{ccc}-\frac{\sin L}{N\cos B}\rho & \frac{\cos L}{N\cos B}\rho & 0\\ & & \\ -\frac{\sin B\cos L}{M}\rho & -\frac{\sin B\sin L}{M}\rho & \frac{\cos B}{M}\rho \end{array}\right]\left[\begin{array}{c}{T}_x\\ T_y\\ T_z\end{array}\right]+ \\ \left[\begin{array}{ccc}\tan B\cos L& \tan B\sin L& -1\\ & & \\ -\sin L& \cos L& 0\end{array}\right]\left[\begin{array}{c}{R}_x\\ R_y\\ R_z\end{array}\right]+ \\ \left[\begin{array}{c}0\\ \\ -\frac{N}{M}{e}^2\sin B\cos B\rho \end{array}\right]D+ \\ \left[\begin{array}{cc}0& 0\\ & \\ \frac{N}{Ma}{e}^2\sin B\cos B\rho & \frac{2-e^{2}si{n}^{2}B}{1-f}\sin B\cos B\rho \end{array}\right]\left[\begin{array}{c}\Delta a\\ \Delta f\end{array}\right]\text{\hspace{1em}(1)} \end{gathered}$$

$式中：$
$e^2：第一偏心率的平方，无量纲e^2=2f-f^2；$
$M：地球椭球子午圈曲率半径，单位为米(m)，M=\frac{a(1-e^2)}{(1-e^2{\sin }^{2}B{)}^{\frac{3}{2}}}；$
$N：地球椭球卯酉圈曲率半径，单位为米(m)，N=\frac{a}{(1-e^2{\sin }^{2}B{)}^{\frac{1}{2}}}；$
$B,L：点位纬度、经度，经纬度单位为弧度(rad)；$
$\Delta B,\Delta L：点位在两个坐标系下的纬度差、经度差，经纬度差值单位为角秒{(}^{\prime \prime })；$
$\rho ：角度与弧度间转换量，单位为角秒{(}^{\prime \prime })，\rho =180\times \frac{3600}{\pi }；$
$a,\Delta a：椭球长半轴和长半轴差，单位为米(m)；$
$f,\Delta f：椭球扁率和扁率差，无量纲；$
$T_x,T_y,T_z：平移参数，单位为米(m)；$
$R_x,R_y,R_z：旋转参数，单位为角秒{(}^{\prime \prime })；$
$D：尺度参数，无量纲。$

$令$
$$a_0=-\frac{\sin L}{N\cos B}\rho ，a_1=\frac{\cos L}{N\cos B}\rho ，$$
$$a_3=-\frac{\sin B\cos L}{M}\rho ，a_4=-\frac{\sin B\sin L}{M}\rho ，a_5=\frac{\cos B}{M}\rho ，$$
$$b_0=\tan B\cos L，b_1=\tan B\sin L，$$
$$b_3=-\sin L，b_4=\cos L，$$
$$c_1=-\frac{N}{M}{e}^2\sin B\cos B\rho$$
$$d_2=\frac{N}{Ma}{e}^2\sin B\cos B\rho ，d_3=\frac{2-e^{2}si{n}^{2}B}{1-f}\sin B\cos B\rho$$

$则(1)式变换为：$
$$\begin{gathered} \left[\begin{array}{c}\Delta L\\ \Delta B\end{array}\right]=\left[\begin{array}{ccc}{a}_0& a_1& 0\\ a_3& a_4& a_5\end{array}\right]\left[\begin{array}{c}{T}_x\\ T_y\\ T_z\end{array}\right]+\left[\begin{array}{ccc}{b}_0& b_1& -1\\ b_3& b_4& 0\end{array}\right]\left[\begin{array}{c}{R}_x\\ R_y\\ R_z\end{array}\right]+\left[\begin{array}{c}0\\ c_1\end{array}\right]D+\left[\begin{array}{cc}0& 0\\ d_2& d_3\end{array}\right]\left[\begin{array}{c}\Delta a\\ \Delta f\end{array}\right] \\ =\left[\begin{array}{c}{a}_0{T}_x+a_1{T}_y+0T_z+b_0{R}_x+b_1{R}_y-1R_z+0D\\ a_3{T}_x+a_4{T}_y+a_5{T}_z+b_3{R}_x+b_4{R}_y+0R_z+c_{1}D\end{array}\right]+\left[\begin{array}{c}0\\ d_2\Delta a+d_3\Delta f\end{array}\right]\text{\hspace{1em}(2)} \end{gathered}$$

$又\because$
$$\left[\begin{array}{c}{a}_0{T}_x+a_1{T}_y+0T_z+b_0{R}_x+b_1{R}_y-1R_z+0D\\ a_3{T}_x+a_4{T}_y+a_5{T}_z+b_3{R}_x+b_4{R}_y+0R_z+c_{1}D\end{array}\right]=\left[\begin{array}{ccccccc}{P}_1& P_2& P_3& P_4& P_5& P_6& P_7\\ P_8& P_9& P_{10}& P_{11}& P_{12}& P_{13}& P_{14}\end{array}\right]\left[\begin{array}{c}{T}_x\\ T_y\\ T_z\\ R_x\\ R_y\\ R_z\\ D\end{array}\right]$$
$$\Rightarrow$$
$$\begin{gathered} P_1=a_0，P_2=a_1，P_3=0，P_4=b_0，P_5=b_1，P_6=-1，P_7=0， \\ {P}_8=a_3，P_9=a_4，P_{10}=a_5，P_{11}=b_3，P_{12}=b_4，P_{13}=0，P_{14}=c_1 \end{gathered}$$
$\therefore (2)式变换为：$
$$\left[\begin{array}{c}\Delta L\\ \Delta B\end{array}\right]=\left[\begin{array}{ccccccc}{a}_0& a_1& 0& b_0& b_1& -1& 0\\ a_3& a_4& a_5& b_3& b_4& 0& c_1\end{array}\right]\left[\begin{array}{c}{T}_x\\ T_y\\ T_z\\ R_x\\ R_y\\ R_z\\ D\end{array}\right]+\left[\begin{array}{c}0\\ d_2\Delta a+d_3\Delta f\end{array}\right]\text{\hspace{1em}(3)}$$

基于最小二乘与多对同名点对计算参数

$设存在n对同名点对：(L_a,B_a{)}_1\to (L_b,B_b{)}_1，\cdots ，(L_a,B_a{)}_n\to (L_b,B_b{)}_n.$
$令$
$$\theta =\left[\begin{array}{c}{T}_x\\ T_y\\ T_z\\ R_x\\ R_y\\ R_z\\ D\end{array}\right]$$
$$\Delta L_i=(L_b-L_a{)}_i，\Delta B_i=(B_b-B_a{)}_i，\Delta H_i=(H_b-H_a{)}_i$$
$$v_i=(\Delta L-0,\Delta B-(d_2\Delta a+d_3\Delta f){)}_i^T，$$
$$P_i={\left[\begin{array}{ccccccc}{a}_0& a_1& 0& b_0& b_1& -1& 0\\ a_3& a_4& a_5& b_3& b_4& 0& c_1\end{array}\right]}_i，$$
$$i=1,\cdots ,n$$

$根据式(3)，代入样本值得到方程组如下：$
$$\{\begin{array}{l}{P}_1\theta =v_1\\ P_2\theta =v_2\\ ⋮\\ P_n\theta =v_n\end{array}$$
$则变换为矩阵方程为：$
$$v=P\theta$$
$$P=\left[\begin{array}{c}{P}_1\\ P_2\\ ⋮\\ P_n\end{array}\right]，v=\left[\begin{array}{c}{v}_1\\ v_2\\ ⋮\\ v_n\end{array}\right]$$

$考虑v=P\theta 无解，需要从P的列空间中找出最接近v的向量u（u可以理解为v在P的列空间中的投影，理解如下图所示：）$

$如上图所示，p是b在\left[\begin{array}{cc}{a}_1& a_2\end{array}\right]列空间中的投影。$
$令e=v-u，最小二乘就是找到\parallel e{\parallel }^2最小的点，最小二乘就是指向量长度的最小平方。$

$由上可知，u位于P的列空间中，即u是P的各列的线性组合：$
$$令P的列空间为P=\left[\begin{array}{cccc}{C}_1& C_2& \cdots & C_m\end{array}\right]$$
$$故存在u=C_1\widetilde{{\theta }_1}+C_2\widetilde{{\theta }_2}+\cdots +C_m\widetilde{{\theta }_m}$$
$即P\tilde{\theta }=u有解。$

$$e=v-u=v-P\tilde{\theta }$$
$e正交于P的列空间，存在：$
$$e\perp C_1,e\perp C_2,\cdots ,e\perp C_m$$

$由向量点积关系式可得：$

$$\Rightarrow \{\begin{array}{l}{C}_1^T(v-P\tilde{\theta })=0\\ C_2^T(v-P\tilde{\theta })=0\\ ⋮\\ C_m^T(v-P\tilde{\theta })=0\end{array}$$

$$\Rightarrow \left[\begin{array}{c}{C}_1^T\\ C_2^T\\ C_3^T\\ ⋮\\ C_m^T\end{array}\right](v-P\tilde{\theta })=\left[\begin{array}{c}0\\ 0\\ 0\\ ⋮\\ 0\end{array}\right]$$

$$\because P=\left[\begin{array}{cccc}{C}_1& C_2& \cdots & C_m\end{array}\right]$$
$$\therefore P^T=\left[\begin{array}{c}{C}_1^T\\ C_2^T\\ ⋮\\ C_m^T\end{array}\right]$$

$$\Rightarrow P^T(v-P\tilde{\theta })=0$$
$$\Rightarrow P^{T}P\tilde{\theta }=P^{T}v$$
$$\Rightarrow \tilde{\theta }=(P^{T}P{)}^{-1}{P}^{T}v$$

$即\tilde{\theta }=(P^{T}P{)}^{-1}{P}^{T}v为基于最小二乘计算出来的最接近实际参数的转换值$