Robust control from data via uncertainty model sets identification

In this paper an integrated robust identification and control design procedure is proposed. The plant to be controlled is supposed to be linear, time invariant, stable, possibly infinite dimensional and a set of noise-corrupted input–output measurements is supposed to be available. The emphasis is placed on the design of controllers guaranteeing robust stability and robust performances, and on the trade-off between controller complexity and achievable robust performances. First, uncertainty models are identified, consisting of parametric models of different order and tight frequency bounds on the magnitude of the unmodelled dynamics. Second, Internal Model Controllers, guaranteeing robust closed-loop stability and best approximating the ‘perfect control’ ideal target, are designed using H_infinity / mu-synthesis techniques. Then, the robust performances of the designed controllers are computed, allowing one to determine the level of model/controller complexity needed to guarantee desired closed-loop performances.