Point-Mass Satellite Moving in a Plane (1)

Model description: 

The point-mass satellite moving in a plane, subject to an inverse-square law force field

$$\begin{align*} \dot{r} &=v \\ \dot{v} &=r\omega^{2}-a_{1}\frac{1}{r^2}+a_{2}u_{1} \\ \dot{\phi} &=\omega \\ \dot{\omega} &=-2\frac{v\omega}{r}+a_{2}\frac{1}{r}u_2, \end{align*}$$

where the states are: the satellite's radial position $r < 0$, its radial velocity $v$, its angular position $\phi$ and its angular velocity $\omega$. The satellite is equipped with a radial thruster $u_1$ and a tangential thruster $u_2$, and the known constant model parameters $a_1$ and $a_2$ are related to the strength of the force field and the mass of the satellite. If the outputs are taken as $y_1=\sin{\phi}$ and $y_2=\cos{\phi}$, then the transformation $z_{1,1}=r\sin{\phi}$, $z_{1,2}=v\sin{\phi}+r\omega\sin{\phi}$, $z_{2,1} = r\cos{\phi}$ and $z_{2,2}=v\cos{\phi} - r\omega\cos{\phi}$, brings the system into the NOF

$$\begin{align*} \dot{z}_{1,1} &=z_{1,2} \\ \dot{z}_{1,2} &=-a_{1}\frac{y_1}{\left(y_{1}^{2}+y_{2}^{2}\right)^{3/2}}+\bar{u}_{1} \\ \dot{z}_{2,1} &=z_{2,2} \\ \dot{z}_{2,2} &=-a_{1}\frac{y_2}{\left(y_{1}^{2}+y_{2}^{2}\right)^{3/2}}+\bar{u}_{2}, \end{align*}$$

where $\bar{u}_1 = a_2 ((u_1y_1 + u_2y_2)/(y_1^2 + y_2^2)^{1/2})$ and $\bar{u}_2 = a_2 ((u_1y_2 + u_2y_1)/(y_1^2 + y_2^2)^{1/2})$.

Type: 

Form: 

Model order: 

4

Time domain: 

Linearity: 

Publication details: 

TitleGlobal Tracking via Output Feedback for Nonlinear MIMO Systems
Publication TypeJournal Article
Year of Publication2011
AuthorsKvaternik, K., and Lynch A.F.
JournalIEEE Transactions on Automatic Control
Volume56
Start Page2179
Issue9
Pagination2179-2184
Date Published05/2011
ISSN0018-9286
Accession Number12216413
Keywordscontrol system synthesis, feedback, MIMO systems, nonlinear control systems, observers, tracking
AbstractIn this note we present a constructive method for the design of global asymptotic tracking control for a class of MIMO nonlinear systems by output feedback. The class of systems considered is a special case of those in nonlinear observer form and coincides with the Output Feedback Form when there is only one input and one output. This approach generalizes a SISO method which uses filtered transformations and backstepping. The technique presented here may be useful in accommodating subsystem coupling in other MIMO design contexts. We demonstrate our method by example and observe several interesting features that distinguish it from the SISO case.
DOI10.1109/TAC.2011.2158134