Robust multiestimation based adaptive control of robotic manipulators applied to master-slave tandems

Abstract

The problem of controlling a tandem of robotic manipulators affected by structural uncertainties is addressed via the introduction of a robust multiestimation-based adaptive control law. Each robot is modeled by a multiestimation scheme composed by a finite set of parameter estimation algorithms including a relative dead-zone running in parallel and a higher order level supervisor which online selects one of them to parameterize an adaptive controller connected in feedback to the plant. The multiestimation scheme allows to improve the behavior of the closed loop response of each robot in practical applications in comparison to classical single model based adaptive controllers as it has been verified through simulation examples. Thus, the transient response of the tandem can be improved in the free motion case as well as in the constrained one when the slave robot interacts with the environment reflecting the interaction force to the master. In this case, the scheme reveals to be promising to improve the ability of the master robot to track the slave constrained trajectory.