Neuro-fuzzy adaptive control for flexible-link robots including motor dynamics

A stable neuro-fuzzy (NF) adaptive controller that incorporates the robot dynamics as well as the joint motor dynamics is proposed in this paper for the trajectory tracking of a flexible-link robot with poorly known dynamics. Above all, based on singular perturbation and time-scale decomposition method, dynamics of a flexible-link robot is reduced into a slow subsystem of an equivalent rigid-link arm, a fast subsystem of flexible vibration. Then a NF adaptive controller and a fuzzy proportional-derivative (PD) controller are designed for the slow and fast subsystems separately, and a parameter adaptive controller method for joint motors is proposed. The system stability and tracking error convergence are proved using Lyapunov stability theorem, learning algorithms for free parameters of NF systems and joint motor dynamics are obtained thereby. Finally, simulation studies are carried out to show the viability and effectiveness of the proposed control approach.