Active noise control using a modified filtered-error algorithm with frequency domain decoupling

A. P. Berkhoff, R. Haasjes

Research output: Chapter in Book/Report/Conference proceedingConference contributionAcademicpeer-review

8 Downloads (Pure)

Abstract

This paper presents an algorithm for broadband active noise control, using a modified version of the filtered-error algorithm with frequency domain decoupling and preconditioning. The decoupling is done efficiently in the frequency domain without regard to causality, but the decoupling transfer functions can be made causal by adding a delay. The influence of this delay on the speed of convergence is eliminated by using the modified version of the filtered-error scheme. The control effort is reduced by adding a regularization part which is consistent with the decoupling approach. The system can be used for multiple error signals, multiple secondary sources, and multiple references. This paper presents simulations of the behavior of the algorithm based on acoustic transfer functions and signals, showing that the frequency domain decoupling approach using the delay compensation technique increases the speed of convergence.

Original languageEnglish
Title of host publicationProceedings of ISMA 2022 - International Conference on Noise and Vibration Engineering and USD 2022 - International Conference on Uncertainty in Structural Dynamics
EditorsW. Desmet, B. Pluymers, D. Moens, S. Neeckx
PublisherKatholieke Universiteit Leuven
Pages155-161
Number of pages7
ISBN (Electronic)9789082893151
Publication statusPublished - 2022
Event30th International Conference on Noise and Vibration Engineering, ISMA 2022 - Leuven, Belgium
Duration: 12 Sept 202214 Sept 2022
Conference number: 30

Conference

Conference30th International Conference on Noise and Vibration Engineering, ISMA 2022
Abbreviated titleISMA 2022
Country/TerritoryBelgium
CityLeuven
Period12/09/2214/09/22
OtherOrganised in conjunction with the 9th International Conference on Uncertainty in Structural Dynamics (USD2022)

Fingerprint

Dive into the research topics of 'Active noise control using a modified filtered-error algorithm with frequency domain decoupling'. Together they form a unique fingerprint.

Cite this