Robust H.INF. control of a Parallel Link Robot Manipulator.

摘要

This paper proposes a linear robust control scheme for robotic trajectory tracking based on the H∞ control theory. The controller consists of a model-based linear robust controller, and it does not employ inverse dynamic computation, hence once the controller has been obtained, computational burden for H∞ controller is relatively small, because the H∞ controller is linear and time invariant and the computational burden is only matrix vector multiplications. Further, the control input is continuous. Therefore, the H∞ controller is suitable for realization of real-time model-based robust control.At first, our experimental manipulator is introduced. Then, the dynamics and model uncertainties of the robot system including actuator are discussed. Dynamic couplings and the gravity forces are treated as the real structured uncertainties. Then, setting up the control objectives in H∞ synthesis framework, an H∞ controller is designed with the static constant scaling matrix D to achieve robust performance of the closed-loop system by small gain theorem. Finally, we show the effectiveness of proposed linear robust H∞ control scheme for a robot manipulator with experimental results.