Control oriented uncertainty modelling using μ sensitivities and skewed μ analysis tools

In this paper, an approach for the computation of bounds on parametric model uncertainty for robust control design is proposed. The approach uses μ sensitivities to identify which uncertain parameters in the model are most critical in terms of closed-loop robustness objectives. Using skewed μ analysis tools, the maximum possible uncertainty bounds for model parameters which are most difficult or expensive to identify exactly are then computed. The application of the proposed approach is illustrated via a flight control law design example.