Adaptive control of robot manipulators with only position measurements

Abstract

This paper proposes a globally asymptotically stable adaptive control strategy for trajectory tracking of robot manipulators with uncertainties by using only position measurements. Two general nonlinear time-varying filters are presented to eliminate the velocity measurements, which are quite different from the existing linear high-pass filters. In addition, it is also proved that the parameter estimates will converge to the true values if the persistent excitation condition is met. Simulation tests are included, with reference to a robot having two degrees of freedom.