On structure of local models for hybrid controllers

MaRDI QIDQ4546623zbMATH OpenFDO

Authors Tatiana Valentine Guy, Miroslav Kárný

Publication date 4 December 2002

zbMATH Keywords

adaptive controlBayesian methodologycomplexityfilteringmultiresolution approximationregression vectorsamplingsystem structure simplification

Mathematics Subject Classification ID

System structure simplification (93B11) Adaptive control/observation systems (93C40) Sampled-data control/observation systems (93C57) Filtering in stochastic control theory (93E11)

Review text

The basic problem in adaptive control is to have a short sampling period and to prevent an increase in complexity of the model. The paper proposes an approach to an optimal trade-off between these requirements, which exhibits the following properties: filtering of measurements to generate controller inputs, decrease of the regression vector length, and Bayesian methodology for optimizing the structure of the model. The basic equations are derived, and approximate solutions are obtained by using the multiresolution approximation of the input-output signals. For the entire collection see [Zbl 0969.00072].

Recommendations

• Hybrid adaptive control
• scientific article; zbMATH DE number 4141906
• Multiple model adaptive control. Part 1: Finite controller coverings
• scientific article; zbMATH DE number 4174785
• Multirate hybrid adaptive control

Cited in

(4)

• scientific article; zbMATH DE number 3972750 (Why is no real title available?)
• On the performance and complexity of a class of hybrid controller switching policies
• Low-Complexity Method for Hybrid MPC with Local Guarantees
• Local model networks based mixed-sensitivity H-infinity control of CE-150 helicopters

This page was built for publication: On structure of local models for hybrid controllers

Report a bug (only for logged in users!)Click here to report a bug for this page (MaRDI item Q4546623)