Reference State Trajectory Generation for Output Tracking with Constraints using Search Trees

Abstract

This paper generates a reference state trajectory from a given reference output trajectory for redundant nonlinear affine dynamic systems subject to input, state and output constraints. Extending upon the expansive search tree (EST) methodology, a new probabilistic sampling method is proposed to address time dependence and output mappings. It is shown that the methodology is able to find a feasible solution given constraints, if such a solution exists. The use of the proposed methodology is illustrated by considering the end effector tracking of a multi-link cable-driven parallel robot.