On Quantifying Tolerable Uncertainty for a Specified Level Closed-Loop Performance
Abstract
Given a plant and a feedback controller it is natural to ask:How much uncertainty can be tolerated by the closed-loop,while achieving a specified level of performance? Here, acharacterisation of this question is formulated in terms of anoptimisation problem with a cost that reflects the size of weightsused to quantify system uncertainty and a structured singularvalue constraint, which captures the specified level of robustperformance. In the case of unstructured uncertainty the problemcan be solved as a family of convex problems pointwise infrequency. An iterative algorithm is developed for the case ofstructured uncertainty.