adaptive feedback linearization of non-linear systems with significant actuator dynamics

When applying feedback linearization techniques actuator dynamics is usually neglected. This is crucial for actuators with large time constants. A large class of systems, e.g. chemical processes, belong to this category. In such cases conventional feedback linearization techniques based on adaptive control or sliding control perform poorly. This paper discusses an approach which allows the inclusion of the actuator dynamics in the controller design. The design methodology has validation both for adaptive feedback linearization as well as sliding control. An augmented Lyapunov-like function candidate including the actuator states was used to prove global stability. This is done by applying Barbalat's Lyapunov-like lemma for non-autonomous systems. The control law is simulated for both the nominal and adaptive cases. In the adaptive case both the plant parameters and the actuator time constant are assumed to be unknown.