Optimized Disturbance Weighting for Robust System Design under Parameter Uncertainties

Abstract:
We propose a control system design optimization that is robust with respect to product manufacturing tolerances. To achieve an optimal performance and robust controller design, we model plant parameter tolerances as structured uncertainty and compute the worst case perturbed plant by solving a constrained nonlinear optimization problem. The mixed sensitivity robust design optimization is improved by fitting the weighting transfer function to the worst case perturbed response. A DC motor control is utilized to compare and evaluate the system performance and robustness of several design approaches. Furthermore, we defined a reliable index for robustness evaluation of control system designs.