In this thesis, motion control of robot manipulator with joint torque measurements is considered. Control schemes for both rigid and flexible joint robots are developed.

An adaptive controller with positive joint torque feedback is proposed for the motion control of rigid robots. The positive joint torque feedback cancels out the uncertainties in the link dynamics;​ the adaptive controller copes with the uncertainties of the drive systems and compensates for the effects of nonlinear friction. It is proved that, with the proposed control scheme, the motion tracking errors converge asymptotically to zero. Robustness to uncertainties and high performance are achieved.

The proposed control scheme is extended to controlling flexible joint robots. The effects of joint flexibility on the system performance are investigated. A desired trajectory adjustment is proposed to ensure that the joint angles track the desired positions precisely.

Control schemes with negative joint torque feedback are also studied for the motion control of flexible joint robots. A centralized joint torque controller is developed. The convergence analysis of the control system shows that the system tracks asymptotically the desired trajectory, and the joint torques track asymptotically the desired joint torques. Futhermore, a decentralized control scheme with negative and positive joint torque feedback is designed. The resultant system requires limited amount of computation, and is robust to uncertainties in link dynamics and drive systems.

Simulation results verify the effectiveness of the proposed control methods.