This paper proposes the fault diagnosis method in 6 high cold rolling mill which consist of 5 stand to assess the normal and fault conditions. The proposed method concerns with the strip rupture fault diagnosis based on transient current signal. Firstly, the signal smoothing technique is performed initially to highlight the fundamental of transient signal at normal and fault condition. Then the smoothed signal is subtracted from the original signal in order to transform the original data become useful data that used for further analysis. Next, discrete wavelet transform (DWT) method is performed to present the detail signal. Moreover, features are calculated from detail signal of DWT and then extracted using principal component analysis (PCA) and kernel principal component analysis (KPCA) for dimensionality reduction purpose. Finally, using support vector machine (SVM) for classification, the results of stand 5 shows more clear classified compare with other stands.
CITATION STYLE
Yang, S. W., Widodo, A., Caesarendra, W., Oh, J. S., Shim, M. C., Kim, S. J., … Lee, W. H. (2009). Support vector machine and discrete wavelet transform method for strip rupture detection based on transient current signal. In Engineering Asset Lifecycle Management - Proceedings of the 4th World Congress on Engineering Asset Management, WCEAM 2009 (pp. 671–678). https://doi.org/10.1007/978-0-85729-320-6_78
Mendeley helps you to discover research relevant for your work.