PREDICTIVE OPTIMIZING REFERENCE GOVERNOR FOR CONSTRAINED 2 DOF's ROBOT WITH ABRUPT SET-POINT TRAJECTORIES

Abstract

This paper discusses application of the predictive optimizing reference governor (multi-layer control strategy) to a process operates under basic feedback (unconstrained) controllers. The process is two degrees of freedom IMI robot equipped with PD-controllers. The PD-controllers have been considered as a direct (basic) control layer in the inner feedback loop of the hierarchical control scheme. The direct controllers receive their reference trajectory values (optimum set-point values) from a nonlinear constrained predictive optimizing governor (outer loop). The system objective is to fulfil both constraints and position tracking performance. The IMI robot is direct driven (DDA), has nonlinear dynamics of high coupling. These dynamics are linearized, at each sampling time, about the generated optimum values from application of Taylor's series method. The Matlap code simulation results prove the advantageous of the applied technique.