Home   >   CSC-OpenAccess Library   >    Manuscript Information
Vehicle Noise Pattern Recognition by Self-Organizing Maps
Stefan Twieg, Holger Opfer, Helmut Beikirch
Pages - 180 - 190     |    Revised - 30-12-2009     |    Published - 31-01-2010
Volume - 3   Issue - 6    |    Publication Date - January 2010  Table of Contents
MORE INFORMATION
KEYWORDS
signal classification, SOM, Noise Vibration Harshness (in car), pattern recognition
ABSTRACT
Interior vehicle acoustics are in close connection with our quality opinion. The noise in vehicle interior is complex and can be considered as a sum of various sound emission sources. A nice sounding vehicle is objective of the development departments for car acoustics. In the process of manufacturing the targets for a qualitatively high-valuable sound must be maintained. However, it is possible that production errors lead to a deviation from the wanted vehicle interior sound. This will result in customer complaints where for example a rattling or squeak refers to a worn-out or defective component. Also in this case, of course, the vehicle interior noise does not fulfill the targets of the process of development. For both cases there is currently no possibility for automated analysis of the vehicle interior noise. In this paper an approach for automated analysis of vehicle interior noise by means of neural algorithms is investigated. The presented system analyses microphone signals from car interior measured at real environmental conditions. This is in contrast to well known techniques, as e.g. acoustic engine test bench. Self-Organizing Maps combined with feature selection algorithms are used for acoustic pattern recognition. The presented system can be used in production process as well as a standalone ECU in car.
CITED BY (2)  
1 Pietila, G. M. (2013). Intelligent Systems Approaches to Product Sound Quality Analysis (Doctoral dissertation, University of Cincinnati).
2 Pietila, G., & Lim, T. C. (2012). Intelligent systems approaches to product sound quality evaluations–A review. Applied Acoustics, 73(10), 987-1002.
1 Google Scholar 
2 ScientificCommons 
3 Academic Index 
4 CiteSeerX 
5 refSeek 
6 iSEEK 
7 Scribd 
8 SlideShare 
9 PDFCAST 
10 PdfSR 
A. Asuncion, D. Newman, UCI Machine Learning Repository, University of California, Irvine, School of Information and Computer Sciences, 2007
A.-W. Moore, Information Gain. Carnegie Mellon University Pittsburgh, 2003
B. Orend, I. Meyer, Schadensfrüherkennung mittels Körperschall. MTZ, 2009
J. Biesiada, W. Duch, Feature Selection for High-Dimensional Data: A Kolmogorov-Smirnov Correlation-Based Filter, Advances in Soft Computing, Springer Verlag, pp. 95-104, 2005
K. Stepper, Ein Beitrag zur akustischen Güteprüfung von Komponenten der Kfz.- und Automatisierungstechnik. TU Berlin, 1999
K.-M. Würzner, Textkategorisierung: Entscheidungsbäume. Universität Bielefeld, 2003
L. Yu, H. Liu, Feature Selection for High-Dimensional Data: A Fast Correlation-Based Filter Solution. ICML, 2003
L. Yu, H. Liu, I. Guyon, Efficient Feature Selection via Analysis of Relevance and Redundancy. Journal of Machine Learning Research, pp. 1205-1224, 2004
M. A. Hall, L. A. Smith, Feature subset selection: a correlation based filter approach. International Conference on Neural Information Processing and Intelligent Information Systems. © Springer, pp. 855-858, 1997
Niemann, H. (1988). Klassifikation von Mustern. Springer Verlag, Berlin, Heidelberg, New York, Tokyo, 1988
O. Chapelle, Feature selection for Support Vector Machines, Max Planck Institute for Biological Cybernetics, Tübingen, 2005
R. Jahnke, Beitrag zur akustischen Qualitätsprüfung im Bereich industrieller Fertigung. TU Berlin, 2001
S. Twieg, H. Opfer, H. Beikirch, Correlation based method for acoustic condition recognition. 1st IEEE Germany Student Conference, Hamburg University of Technology, 2009
T. Fincke, V. Lobo, F. Baco, Visualizing self-organizing maps with GIS, GI Days 2008 Münster, 2008
T. Kohonen, Self-organizing maps, Springer Verlag, 2001
Zhipeng Li & Fuqiang Liu, Cellular Automaton Based Simulation for Signalized Street. International Journal of Engineering and Interdisciplinary Mathematics, vol. 1, no. 1, pp. 11-22, 2009
Mr. Stefan Twieg
Volkswagen Group Research - Germany
stefan.twieg@volkswagen.de
Dr. Holger Opfer
- Japan
Professor Helmut Beikirch
- Germany