The University of Southampton
University of Southampton Institutional Repository

Reconstruction of acoustic source strength distributions and their interactions by inverse techniques

Reconstruction of acoustic source strength distributions and their interactions by inverse techniques
Reconstruction of acoustic source strength distributions and their interactions by inverse techniques

This thesis is devoted to the study of a discrete inverse problem in acoustics. In particular, this work is aimed at the reconstruction of the complex strengths of a number of acoustic sources when given the acoustic pressures measured at a number of points in the acoustic field. Since in the discrete inverse problem the condition number of the transfer function matrix to be inverted plays a crucial role in producing a satisfactory reconstruction, this is investigated in detail both analytically and empirically. In formulating our problem, we employ the least squares method with minimal norm. This approach reconstructs acoustic source strength distributions and their interactions with extremely good agreement with the true values in the ideal situation in which no measurement noise and no error involved in the transfer functions are assumed. In practice, however, their effect is unavoidable and in such cases the accuracy of the reconstruction is determined by the conditioning of the matrix to be inverted and the amount of measurement noise and error contained in the transfer functions.

Accordingly, regularisation methods such as Tikhonov regularisation, singular value discarding and singular value decomposition based regularisation are used to filter our the influence of the noise by improving the conditioning. In order to choose appropriately the regularisation parameter and the singular values to be discarded, we suggest two main methods. One method requires prior knowledge of either the acoustic source or the noise. The other method does not demand such prior knowledge and the techniques of ordinary cross-validation and generalised cross-validation are investigated. In our problem it is necessary to measure the acoustic pressure auto- and cross-spectra at a number of positions in the acoustic field. Direct measurement of these quantities, however, requires considerable experimental effort, especially when the number of field points to be measured is large. For this reason, a technique using a few reference microphones is proposed which enables the construction of the full cross-spectral matrix of acoustic pressures with the minimum number of measurements.

University of Southampton
Yoon, Seongho
Yoon, Seongho

Yoon, Seongho (1998) Reconstruction of acoustic source strength distributions and their interactions by inverse techniques. University of Southampton, Doctoral Thesis.

Record type: Thesis (Doctoral)

Abstract

This thesis is devoted to the study of a discrete inverse problem in acoustics. In particular, this work is aimed at the reconstruction of the complex strengths of a number of acoustic sources when given the acoustic pressures measured at a number of points in the acoustic field. Since in the discrete inverse problem the condition number of the transfer function matrix to be inverted plays a crucial role in producing a satisfactory reconstruction, this is investigated in detail both analytically and empirically. In formulating our problem, we employ the least squares method with minimal norm. This approach reconstructs acoustic source strength distributions and their interactions with extremely good agreement with the true values in the ideal situation in which no measurement noise and no error involved in the transfer functions are assumed. In practice, however, their effect is unavoidable and in such cases the accuracy of the reconstruction is determined by the conditioning of the matrix to be inverted and the amount of measurement noise and error contained in the transfer functions.

Accordingly, regularisation methods such as Tikhonov regularisation, singular value discarding and singular value decomposition based regularisation are used to filter our the influence of the noise by improving the conditioning. In order to choose appropriately the regularisation parameter and the singular values to be discarded, we suggest two main methods. One method requires prior knowledge of either the acoustic source or the noise. The other method does not demand such prior knowledge and the techniques of ordinary cross-validation and generalised cross-validation are investigated. In our problem it is necessary to measure the acoustic pressure auto- and cross-spectra at a number of positions in the acoustic field. Direct measurement of these quantities, however, requires considerable experimental effort, especially when the number of field points to be measured is large. For this reason, a technique using a few reference microphones is proposed which enables the construction of the full cross-spectral matrix of acoustic pressures with the minimum number of measurements.

This record has no associated files available for download.

More information

Published date: 1998

Identifiers

Local EPrints ID: 463359
URI: http://eprints.soton.ac.uk/id/eprint/463359
PURE UUID: 9c09bec3-463a-44a3-b1b7-1cd3edefa33f

Catalogue record

Date deposited: 04 Jul 2022 20:50
Last modified: 04 Jul 2022 20:50

Export record

Contributors

Author: Seongho Yoon

Download statistics

Downloads from ePrints over the past year. Other digital versions may also be available to download e.g. from the publisher's website.

View more statistics

Atom RSS 1.0 RSS 2.0

Contact ePrints Soton: eprints@soton.ac.uk

ePrints Soton supports OAI 2.0 with a base URL of http://eprints.soton.ac.uk/cgi/oai2

This repository has been built using EPrints software, developed at the University of Southampton, but available to everyone to use.

We use cookies to ensure that we give you the best experience on our website. If you continue without changing your settings, we will assume that you are happy to receive cookies on the University of Southampton website.

×