Parallel algorithms for adaptive control: Robust stability

From MaRDI portal





The authors describe a class of adaptive control systems which can be constructed modularly, with an identifier, an internal regulator, and a tuner as the primary components. The identifier and the regulator make up a parameterized controller. This controller can be implemented as a number of linear dynamical systems running in parallel, and can be designed using standard linear system design techniques. The tuner or supervisor consists of a (real-time) switching algorithm in charge of selecting the most appropriate signal to be fed back into the process according to observations of suitably defined identification errors. The authors' method preserves the robustness properties of the adaptive linear control design. The authors suggest that more focused attention to the interplay between computational capacity and the formulation of control algorithms could lead to better performing control systems. For the entire collection see [Zbl 0864.00080].











This page was built for publication: Parallel algorithms for adaptive control: Robust stability

Report a bug (only for logged in users!)Click here to report a bug for this page (MaRDI item Q4353670)