Archives of Acoustics,
36, 1, pp. 65–76, 2011
Active Noise-Vibration Control using the Filtered-Reference LMS Algorithm with Compensation of Vibrating Plate Temperature Variation
Vibrating plates have been recently used for a number of active noise control
applications. They are resistant to difficult environmental conditions including dust,
humidity, and even precipitation. However, their properties significantly depend
on temperature. The plate temperature changes, caused by ambient temperature
changes or plate heating due to internal friction, result in varying response of the
plate, and may make it significantly different than response of a fixed model. Such
mismatch may deteriorate performance of an active noise control system or even
lead to divergence of a model-based adaptation algorithm.
In this paper effects of vibrating plate temperature variation on a feedforward
adaptive active noise reduction system with the multichannel Filtered-reference LMS
algorithm are examined. For that purpose, a thin aluminum plate is excited with
multiple Macro-Fiber Composite actuators. The plate temperature is forced by a set
of Peltier cells, what allows for both cooling and heating the plate. The noise is
generated at one side of the plate, and a major part of it is transmitted through the
plate. The goal of the control system is to reduce sound pressure level at a specified
area on the other side of the plate.
To guarantee successful operation of the control system in face of plate temperature
variation, a gain-scheduling scheme is proposed to support the Filteredreference
LMS algorithm.
applications. They are resistant to difficult environmental conditions including dust,
humidity, and even precipitation. However, their properties significantly depend
on temperature. The plate temperature changes, caused by ambient temperature
changes or plate heating due to internal friction, result in varying response of the
plate, and may make it significantly different than response of a fixed model. Such
mismatch may deteriorate performance of an active noise control system or even
lead to divergence of a model-based adaptation algorithm.
In this paper effects of vibrating plate temperature variation on a feedforward
adaptive active noise reduction system with the multichannel Filtered-reference LMS
algorithm are examined. For that purpose, a thin aluminum plate is excited with
multiple Macro-Fiber Composite actuators. The plate temperature is forced by a set
of Peltier cells, what allows for both cooling and heating the plate. The noise is
generated at one side of the plate, and a major part of it is transmitted through the
plate. The goal of the control system is to reduce sound pressure level at a specified
area on the other side of the plate.
To guarantee successful operation of the control system in face of plate temperature
variation, a gain-scheduling scheme is proposed to support the Filteredreference
LMS algorithm.
Keywords:
active noise-vibration control; active structural acoustic control; adaptive control; gain scheduling; temperature influence
Full Text:
PDF
Copyright © Polish Academy of Sciences & Institute of Fundamental Technological Research (IPPT PAN).