Fault Classification Using Multi-Layer Perceptrons and Support Vector Machines

Tshilidzi Marwalaa, Snehashish Chakraverty and Unathi Mahola
School of Electrical and Information Engineering
University of the Witwatersrand Private Bag x 3 Wits 2050 South Africa
e-mail: t.marwala@ee.wits.ac.za
Central Building Research Institute
Roorkee-247 667, U.A. India
e-mail: sne_chak@yahoo.com


This paper introduces support vector machines (SVM) to classify faults in a population of cylindrical shells.  The proposed procedure is tested on a population of 20 cylindrical shells and its performance is compared to the procedure, which uses multi-layer
perceptrons (MLP).  The modal properties extracted from vibration data are used to train the SVM and MLP. It is observed that the SVM produces 94% classification accuracy while the MLP produces 88% classification rates.

To read the full article,
please log in:
If you haven't registered already,
you can do so for free:
Copyright 2000-2006 Professor F.R. Hall & Dr I. Oraifige, University of Wolverhampton.
close this window