A control law for robotic manipulators based on a filtered signal to generate PD action and velocity estimates

This paper deals with an adaptive control law for robotic manipulators based on a filtered signal to generate both the PD action and velocity estimates of the joints, suitable for trajectory tracking tasks, with the particular aim of reducing the harmonic content of the mechanical torques developed at the joints and thus avoiding excitation of unmodelled dynamics and instability. The practical aspects relative to the implementation of the control law are considered as relevant and, consequently, are detailed. In particular, several methods suitable to compute velocity estimates are discussed and compared with the method described in the paper. All of the above methods are illustrated by means of digital simulation and experimental findings carried out on a 2-DOF SCARA-type manipulator. The tests carried out show the validity of the discussed approach.