Nonregular feedback linearization for a class of second-order nonlinear systems

From MaRDI portal





This paper is concerned with the linearization of affine control systems by nonregular static state feedback, i.e. a feedback that may not be invertible with respect to control. More specifically, it has been shown that certain multi-input systems can be transformed by nonregular feedback to single-input systems that are feedback linearizable in the classical sense. This result resembles somehow the well-known Heymann lemma in feedback stabilization of linear systems. The main result of this paper assserts that smooth systems of the form NEWLINE\[NEWLINE\ddot x= f(x,\dot x)+ p(x,\dot x)y+ g(x,\dot x)u,\quad \ddot y= u,NEWLINE\]NEWLINE where \(x,y,u\in \mathbb{R}^n\), the matrix \(p(x,\dot x)\) is upper triangular with non-zero diagonal elements, and the matrix \(g(x,\dot x)\) is strictly upper triangular with non-zero elements over the diagonal, subject to a feedback can be made single-input feedback linearizable. This result has been applied to the full model of the dynamics of a flexible joint manipulator with constant inertia submatrix reflecting links-actuators coupling. This model is known to be dynamic, not static feedback linearizable. A linearization strategy proposed in the paper leads to an elaboration of a nonregular stabilizing static state feedback. Although inspiring, this paper leaves two things to be desired: first, a simulation example would help to assess the quality of the proposed stabilization algorithm for the flexible joint manipulators. Second, when reviewing the literature on dynamic feedback linearization the authors have ignored completely the fundamental results of the French school. This means that the editors of Automatica should be more demanding.











This page was built for publication: Nonregular feedback linearization for a class of second-order nonlinear systems

Report a bug (only for logged in users!)Click here to report a bug for this page (MaRDI item Q5950256)