Guaranteed robustness/performance adaptive control with limited torque for robot manipulators

A novel guaranteed robustness/performance adaptive algorithm with limited torque is proposed for tracking control of robot manipulators in this paper. In framework of the adaptive controller, a continuous differentiable increasing function is introduced into the parameter estimator of controller so as to restrain from parameter drifting. Furthermore, both robustness and performance are guaranteed without any persistent excitation (PE) condition: (i) asymptotic stability is achieved in absence of disturbance when the estimation of parameter-range is available; (ii) the close-loop system is Lp input/output stable and its solution is uniformly ultimately bounded regardless whether the true parameters are located within the estimated parameter-range or not. Finally, taking some practical complexity into account, such as disturbances, dead-zone as well as dynamics of brushed direct-current motor, simulation results via three-dof robot manipulator show the significant improvement in terms of robustness and performance in comparison with the conventional adaptive and projection algorithm.