A generalized trajectory tracking controller for robot manipulators with bounded inputs

Abstract

A generalized controller based on stability theory of singularly perturbed systems is proposed, to deal with the problem of bounded actuator inputs in robot trajectory tracking control. The saturation function with error-gain matrix is applied in the torque control law, which ensures the upper bound of torque inputs in any given limited range. Through appropriately setting the entries of the error-gain matrix, the tracking performance can be improved. Moreover, a pseudo signal is generated from a linear filter to substitute for the actual velocity error, eliminating the need for velocity measurements. Finally, to verify the effectiveness of the generalized controller, a new saturated controller with error-gain-contained arc tangent function is designed. Comparison experiments show that the proposed controller can strictly guarantee the bound of the torque inputs in situations with non-zero initial tracking errors, and gives a better tracking result than other controllers.