Input-Output LPV Models: Colsed-loop Identification and IQC based Controller Synthesis

Beschreibung

This work concerns identification and control of Linear Parameter Varying (LPV) systems in the input-output setting. A method to identify unstable LPV models in input-output form is proposed where indirect identification is performed via an instrumental variable (IV) method. Another method to identify LPV models in input-output form is proposed where it is assumed that the controller is unknown. A method for synthesizing a controller directly in the input-output form is proposed. Synthesis conditions presented are essentially an extension of Integral Quadratic Constraints (IQC) based input-output analysis conditions for LTI systems and are based on the well known Finsler's lemma. Based on the modelling of the parameter rate of variation, a stability condition for input-output LPV systems is proposed that explicitly utilises the knowledge of parameter rate of variation. Another condition based on IQC is proposed in the same setting that also incorporates the performance.

Autor*in

Ahsan Ali

Stimmen zu »Input-Output LPV Models: Colsed-loop Identification and IQC based Controller Synthesis«

Details

Link teilen