Spatial localization of intermittent noise sources by acoustic antennae

Patent metadata

Field	Value
Publication number	US-9423490-B2
Application number	US-201414761788-A
Country	US
Kind code	B2
Filing date	Jan 18, 2014
Priority date	Jan 18, 2013
Publication date	Aug 23, 2016
Grant date	Aug 23, 2016

Abstract

A noise source localization process that uses as few as four microphones in an antenna array. The process combines different averaging to the information from the various signals with non-linear (wavelet-based) filtering, cross-correlation and triangulation to more particularly locate a noise source relative to a target point. The process calculates a propagation time based on the distance from a target to each microphones, delays each of the signals according to the propagation time from the target, calculates a continuous wavelet transform for each of the signals to band-pass filter each of the signals according to a predetermined frequency of interest, determines the product of each pair of signals to produce correlation fringes, and locates the noise source based on the product of the pairs of signals.

First claim

What is claimed is:

1. A method of localizing a noise source, comprising the steps of:
obtaining at least four acoustic signals, each of which is provided by a separate microphone;
selecting a target and calculating the distance from said target to each of said microphones;
delaying each of said signals according to a propagation time delay from said target based on said distance;
calculating a continuous wavelet transform for each of said signal to band-pass filter each of said signals according to a predetermined frequency of interest;
determining the product of each pair of said signals to produce correlation fringes;
calculating the local average of said fringes;
concluding that said noise source is at said target if the peak of the average correlation occurs for zero additional delay relative to the propagation time delay.

2. The method of claim 1 , wherein said step of concluding that said noise source is at said target comprises locating said noise source at said target when said product has a positive value at the matching maxima and minima of said signals.

3. The method of claim 2 , wherein the distance between said microphones is equal to or larger than said distance to said target.

4. The method of claim 3 , wherein the step of calculating a continuous wavelet transform for each of said signals comprises using a Mexican hat wavelet.

5. The method of claim 3 , wherein the step of calculating a continuous wavelet transform for each of said signals comprises using a real or imaginary part of a Morlet transform.

6. A system for localizing a noise source, comprising:
at least four microphones, each of which is configured to output a corresponding acoustic signals, each of which is provided by a separate microphone;
a processor interconnected to each of said microphones and programmed to digitally sample said acoustic signals;
wherein said processor is programmed to calculate a propagation time based on the distance from a target to each of said microphones, to delay each of said signals according to the propagation time from said target, to calculate a continuous wavelet transform for each of said signals, to band-pass filter each of said signals according to a predetermined frequency of interest, to determine the product of each pair of said signals to produce correlation fringes for several delays larger and shorter than said delays according to the propagation time, to perform a local averaging of said correlation fringes, and to conclude that the source is at said target if the maximum of said average occurs for said delays associated with said target.

7. The system of claim 6 , wherein said processor is programmed to locate said noise source based on said product of said pairs of signals by locating said noise source at said target when said product has a positive value at the matching maxima and minima of said signals.

8. The system of claim 7 , wherein the distance between said microphones is equal to or larger than said distance to said target.

9. The system of claim 8 , wherein said processor is programmed to calculate a continuous wavelet transform for each of said signals using a Mexican hat wavelet.

10. The system of claim 8 , wherein said processor is programmed to calculate a continuous wavelet transform for each of said signals using a real or imaginary part of a Morlet transform.

Assignees

• Lewalle Jacques
• Univ Syracuse

Inventors

• Lewalle Jacques

Classifications

• G01S5/30 Primary

  Determining absolute distances from a plurality of spaced points of known location · CPC title