ABSTRACT
This chapter introduces a new class of adaptive controllers for uncertain robotic systems with actuator dynamics . In contrast with most adaptive robot controllers, the present strategies are developed using a mechatronic systems approach, in which the underlying electromechanical system structure is fully exploited in deriving the controllers. As a consequence, it is possible to design each scheme so that it possesses a simple and modular structure, is easy to implement, and requires virtually no information regarding either the mechanical or actuator models. Controllers are proposed which provide solutions to a number of important robotic system control problems, including system stabilization, trajectory tracking, impedance control, and position/force control. Additionally, it is shown how to extend the control system design methodology to account for actuator flexibility effects, which can be of importance in some applications . It is rigorously demonstrated that all of the controllers ensure uniform boundedness of all signals and provide arbitrarily accurate realization of the given control objectives. The efficacy of the proposed approach is illustrated through both computer simulations and hardware experiments with a variety of robotic manipulators.
