Adaptive passification of nonlinear systems

Abstract

In this paper we consider feedback passive systems, defined as nonlinear systems that can be rendered (strictly) passive by means of state feedback. For feedback passive systems in a parametric normal form, we present an adaptive controller that effectively 'passifies' the system. The proposed control uses techniques of Speed-Gradient (S-G) methodology from the Russian literature. Stability results can be given which do not require the plant to be linear in the parameters. Moreover, examples indicate that the S-G approach has several good features such as avoiding overparametrization, preserving physical structure and improving the speed of convergence.