Observer-controller design for global tracking of nonholonomic systems

From MaRDI portal





The authors consider the system \[ \begin{cases} \dot x_1= u_1,\;\dot x_2= u_2,\\ \dot x_k= x_{k- 1}u_1,\;k= 3,\dots,n,\\ y= (x_1,x_n)^T\end{cases} \] whose control input \(u_1\) is subject to the restriction \(|u_1(t)|\leq u_{1\max}\). Given a desired reference path \(x_d(t)\) generated by \[ \begin{cases} \dot x_{1d}= u_{1d},\;\dot x_{2d}= u_{2d},\\ \dot x_{k,d}= x_{k-1,d} u_{1d},\;k= 3,\dots,n\end{cases} \] with \(u_d= (u_{1d}, u_{2d})\), as a bounded and continuous time-varying reference signal, it is required to find a dynamic feedback law \[ \begin{cases} u=\mu(t,y,\chi),\\ \dot\chi= \nu(t,y,\chi)\end{cases} \] which forces \(x(t)\) to track asymptotically \(x_d(t): \lim_{t\to\infty}|x(t)- x_d(t)|= 0\). The problem is tackled under the assumption that \(u_{1\max}> \sup_{t>0}|u_{1d}(t)|\). The approach is based on suitable observer design facilitated by the saturation condition, the backstepping method to design the output feedback \(u_2\) and the Jurdjević and Quinn method [see \textit{V. Jurdjević} and \textit{J. P. Quinn}, J. Differ. Equations 28, 381-389 (1978; Zbl 0417.93012)] to obtain \(u_1\). For the entire collection see [Zbl 0915.00063].











This page was built for publication: Observer-controller design for global tracking of nonholonomic systems

Report a bug (only for logged in users!)Click here to report a bug for this page (MaRDI item Q4256884)