CONTROL OF FLEXIBLE JOINT ROBOT MANIPULATORS BY COMPENSATING FLEXIBILITY

A exible-joint robot manipulator is a complex system because it is nonlinear, multivariable, highly coupled along with joint exibility and un- certainty. To overcome exibility, several methods have been proposed based on exible model. This paper presents a novel method for controlling exible- joint robot manipulators. A novel control law is presented by compensating exibility to form a rigid robot and then control methods for rigid robots are applied. Feedback linearization and direct adaptive fuzzy control, based on rigid model, are designed with torque control strategy. A decentralized adaptive fuzzy controller is designed because of simplicity and ease of imple- mentation. Eectiveness of the proposed control approach is demonstrated by simulations, using a three-joint articulated exible-joint robot, driven by permanent magnet dc motors.