Robust Composite Adaptive Control of Linearisable Systems With Improved Performance

Abstract

Robust composite adaptive control designs for linearisable systems with known/unknown input gain functions are presented. They mainly consist of a standard adaptive linearizing controller and an immersion and invariance (I&I) based composite update algorithm, with the inclusion of a Lyapunov-based canceling term and a $\sigma $ -modification term for ensuring the practical stability and preventing the parameter-drift phenomenon at the same time. Next, the adding an integrator and the dynamic surface control (DSC) techniques are incorporated for preventing the algebraic-loop problem in the latter case. In particular, the proposed designs ensure the convergence of prediction errors to an order of the exogenous disturbance without relying on persistent excitation (PE), which in turn improves the tracking performance significantly.

Keywords

• Adding an integrator
• composite adaptive control
• dynamic surface control
• immersion and invariance
• smooth switching