ARTIFICIAL INTELLIGENCE TOOLS FOR DISCRETE MULTIESTIMATION ADAPTIVE CONTROL SCHEME WITH MODEL REDUCTION ISSUES

摘要

A multiestimation-based adaptive control scheme is presented for a plant with known poles and unknown zeros. The plant is decomposed in several first order filters with unknown scalar numerators. A set of different reduced plant models is obtained by containing a distinct number of the first order filters in order to consider different approximated models. The scheme chooses in real time the estimator/controller possessing the best performance according to an identification performance index by implementing a switching rule between estimators. The switching rule is subject to a minimum residence time at each identifier/adaptive controller parameterization for closed-loop stabilization purposes. A higher supervision algorithm is used in order to find the best updated online value for a weighting factor.