Robust model predictive control and observer for direct driveapplications

摘要

In this paper, robust position control of a direct drive using anstate space model predictive control (MPC) algorithm is presented. Thenproposed controller consists of a state feedback regulator and anfeedforward controller. Their gains are obtained by minimizing a costnfunction that is a sum of the position tracking errors and the controlncost over some user defined time horizons. The effects of the controllernparameters on the dynamic performance and the robustness of the directndrive are investigated. To provide good estimates of the state variablesnin the presence of load disturbance, a new observer based on thenreceding horizon concept is also formulated. Experimental results arenpresented to demonstrate the effectiveness of the approach