Learning control for nonlinear systems in output feedback form

The class of single-input, single-output, minimum phase, nonlinear, time-invariant systems with unknown output-dependent nonlinearities, unknown parameters and known relative degree ρ is considered. The output regulation problem is addressed and solved in the presence of unknown periodic reference and/or disturbance signals of known common period. A simple learning control algorithm is designed which guarantees asymptotic output tracking for any initial condition belonging to any given connected compact set. It can be interpreted as a generalization of the classical PID ρ − 1 control which solves the regulator problem when reference and disturbance signals are constant. As far as linear systems are concerned, global results are achieved.