MIMO H2 optimal feedback controller estimation for a vi-brating plate using Subspace Model Identification
conference paper
Adaptive Active Control algorithms, such as the well known Filtered-X LMS and Filtered-U LMS algorithms, often do not yield optimal performance in practise, due to finite length impulse response of the controller (Filtered-X) or convergence to a local minimum (Filtered-U). In addition, especially for the MIMO case, these adaptive controllers suffer from slow convergence rate. We have proposed a method to estimate a state-space realization of the H2 optimal controller by Subspace Model Identification (SMI) using the IMC approach and the Causal Wiener filter. This paper presents the MIMO H2 optimal controller estimation by SMI for a vibrating plate in feedback configuration. We show that this method estimates the H2 optimal controller by solving a control-relevant identification problem. Experiments on a 4 input/output vibrating plate illustrate the method.
Topics
TNO Identifier
236816
ISBN
9789073802797
Source title
Proceedings of the 2002 International Conference on Noise and Vibration Engineering, ISMA, 16-18 September 2002, Leuven, Belgium
Editor(s)
Sas, P.
Hal, B.
Hal, B.
Pages
201-209
Files
To receive the publication files, please send an e-mail request to TNO Repository.