Control strategies for robotic manipulators

Abstract

This survey is aimed at presenting the major robust control strategies for rigid robot manipulators. The techniques discussed are feedback linearization/Computed torque control, Variable structure compensator, Passivity based approach and Disturbance observer based control. The first one is based on complete dynamic model of a robot. It results in simple linear control which offers guaranteed stability. Variable structure compensator uses a switching/relay action to overcome dynamic uncertainties and disturbances. Passivity based controller make use of passive structure of a robot. If passivity of a feedback system is proved, nonlinearities and uncertainties will not affect the stability. Disturbance observer based controllers estimate disturbances, which can be cancelled out to achieve a nominal model, for which a simple controller can then be designed. This paper, after explaining each control strategy in detail, finally compares these strategies for their pros and cons. Possible solutions to cope with the drawbacks have also been presented in tabular form.