Tip trajectory tracking of a flexible-joint robot using stable inversion

Abstract

Design of output tracking controllers for nonlinear nonminimum phase systems is challenging. Among existing methods the regulation approach usually leads to large transient errors while the classical inversion approach results in unbounded internal dynamics for nonminimum phase systems. In this paper, stable inversion is applied to the design of tip trajectory tracking for a single-link flexible-joint robot mounted on a wobbly platform. This new type of tracking controller achieves remarkably accurate output tracking without any transient or steady-state errors together with guaranteed stability of both external and internal signals. After development of system dynamics for the robot system, this paper defines the stable inversion problem for such a system which is followed by construction of the unique inverse solution to this problem. Then, this stable inversion approach is applied to the design of a tip trajectory tracking controller with only partial state measurements. Simulation study demonstrates the effectiveness of this approach in output tracking.