Nonlinear feedback stabilization revisited

From MaRDI portal





This paper surveys some recent results by the author and coworkers about stabilization of finite-dimensional, continuous-time systems of the form \[ \dot x=f(x,u). \] The first part of the paper aims to show that discontinuous feedback are unavoidable especially when the control \(u\) appears in a nonlinear way. Continuity at the origin is a particularly critical issue. The topological restrictions imposed by the existence of feedback laws which are smooth for \(x\neq 0\) and continuous at \(x=0\) are illustrated by the classical Brockett's necessary conditions. The main purpose of the paper is to describe a special use of feedback and a corresponding definition of solution which allow to prove that a system is stabilizable if and only if it is asymptotically controllable. The system is treated as a discontinuous one: indeed, the action of the feedback is determined by a sampling schedule of appropriate frequency. The proof involves methods from proximal analysis. The last part of the paper addresses the problem of robustness of this approach with respect to measurement errors and actuator errors. For the entire collection see [Zbl 0911.00015].




Cited in
(22)








This page was built for publication: Nonlinear feedback stabilization revisited

Report a bug (only for logged in users!)Click here to report a bug for this page (MaRDI item Q4240473)