Generated-X LMS (GXLMS)

  • The quality of the reference signal is essential for the adaptation process of an LMS or one of its derivatives. The reference signal affects the stability, the convergence rate and the maximum achievable attenuation. Since the error signal and the control signal are available as numerical values in the algorithm for the LMS, the reference signal can be calculated from both signals. The error signal is the interference between the control signal and the reference signal. This interference of the control signal and the reference signal can be noted mathematically as a simple addition. It is therefore possible to deduce the reference signal from a known error signal and control signal. This approach is the basis of the generated-x LMS (GxLMS) developed by us. It calculates the reference signal itself without having to rely on an externally supplied reference signal. The advantages of the GxLMS are primarily in fields where the reference signal is difficult or impossible to detect. For example, the detection of the reference signal canThe quality of the reference signal is essential for the adaptation process of an LMS or one of its derivatives. The reference signal affects the stability, the convergence rate and the maximum achievable attenuation. Since the error signal and the control signal are available as numerical values in the algorithm for the LMS, the reference signal can be calculated from both signals. The error signal is the interference between the control signal and the reference signal. This interference of the control signal and the reference signal can be noted mathematically as a simple addition. It is therefore possible to deduce the reference signal from a known error signal and control signal. This approach is the basis of the generated-x LMS (GxLMS) developed by us. It calculates the reference signal itself without having to rely on an externally supplied reference signal. The advantages of the GxLMS are primarily in fields where the reference signal is difficult or impossible to detect. For example, the detection of the reference signal can be problematic due to design reasons or measurement technology. For example, flow noise could have a negative effect on an acoustic detection of the reference signal. However, the calculation of the reference signal in the GxLMS represents a further feedback signal path, which affects the stability of the algorithm as a whole. Based on the theoretical principles mathematically sufficient convergence conditions can be formulated taking into account the delays existing in the signal paths. The experimental testing took place on an acoustic duct with monofrequency disturb signals. Since the use of an efficient design of experiments (DoE) could be excluded, the measurement was designed as parameter variation (one factor at time) and therefore very time-consuming. The theoretical background of the GxLMS as well as the results from the experiments are presented.show moreshow less

Export metadata

Additional Services

Share in Twitter Search Google Scholar Statistics
Metadaten
Author details:Markus GluglaGND, Ralf Kiran SchulzGND
URL:https://iiav.org/content/archives_icsv_last/2021_icsv27/content/papers/papers/full_paper_524_20210503203519677.pdf
ISBN:978-83-7880-799-5
Title of parent work (English):Proceedings of the 27th International Congress on Sound and Vibration
Subtitle (English):A modified LMS Algorithm without external reference signal
Publisher:Silesian University Press
Place of publishing:Prague
Publication type:Article
Language:English
Year of first publication:2021
Publication year:2021
Release date:2022/07/29
Tag:Active noise control; Convergence; GxLMS algorithm; LMS without reference signal
Number of pages:8
First page:1
Last Page:8
Organizational units:Extern
Wirtschafts- und Sozialwissenschaftliche Fakultät / Wirtschaftswissenschaften / Funktionsstellen / Lehreinheit für Wirtschafts-Arbeit-Technik
DDC classification:3 Sozialwissenschaften / 33 Wirtschaft / 330 Wirtschaft
3 Sozialwissenschaften / 37 Bildung und Erziehung / 370 Bildung und Erziehung
6 Technik, Medizin, angewandte Wissenschaften / 60 Technik / 600 Technik, Technologie
Peer review:Referiert
Einverstanden ✔
Diese Webseite verwendet technisch erforderliche Session-Cookies. Durch die weitere Nutzung der Webseite stimmen Sie diesem zu. Unsere Datenschutzerklärung finden Sie hier.