Please wait a minute...
Shopping Cart Email List Chinese
Advanced Search Fig/Tab
Frontiers of Electrical and Electronic Engineering

ISSN 2095-2732

ISSN 2095-2740(Online)

CN 10-1028/TM

Featured Articles  |  Most Downloaded  |  Most Reading
Online Submission Subscribe to Our RSS Content Feed
Front Elect Electr Eng    2012, Vol. 7 Issue (4) : 391-398    https://doi.org/10.1007/s11460-012-0208-9
RESEARCH ARTICLE
The fusion of classifier outputs to improve partial discharge classification
R. AMBIKAIRAJAH(), B. T. PHUNG, J. RAVISHANKAR
School of Electrical Engineering and Telecommunications, University of New South Wales, Sydney, NSW 2052, Australia
 Download: PDF(292 KB)   HTML
 Export: BibTeX | EndNote | Reference Manager | ProCite | RefWorks
Abstract

The detection of partial discharge signals and classifying its patterns is an area of interest in the analysis of defects in high voltage cables. This paper investigates a filter-bank based approach to extract frequency domain based features to represent partial discharge signals. By applying the fast Fourier transform, the sampled partial discharge data are mapped into equivalent discrete frequency bins, which are then grouped into N equal sub-bands and also octave sub-bands, each providing N-dimensional features for partial discharge pattern classification. Two classifiers, namely, the support vector machine and the sparse representation classifier, are implemented and their outputs are fused, in order to improve the accuracy of classifying partial discharge. Classification accuracy is also compared with wavelet domain based octave frequency sub-band features.
Keywords partial discharge      features      fusion      classification     
Corresponding Author(s): AMBIKAIRAJAH R.,Email:r.ambikairajah@student.unsw.edu.au   
Issue Date: 05 December 2012
 Cite this article:   
B. T. PHUNG,J. RAVISHANKAR,R. AMBIKAIRAJAH. The fusion of classifier outputs to improve partial discharge classification[J]. Front Elect Electr Eng, 2012, 7(4): 391-398.
 URL:  
https://academic.hep.com.cn/fee/EN/10.1007/s11460-012-0208-9
https://academic.hep.com.cn/fee/EN/Y2012/V7/I4/391
Fig.1  De-noising PD signals in the frequency domain
Fig.1  De-noising PD signals in the frequency domain
Fig.1  De-noising PD signals in the frequency domain
Fig.1  De-noising PD signals in the frequency domain
Fig.1  De-noising PD signals in the frequency domain
Fig.2  De-noising of PD data using signal boosting technique
Fig.2  De-noising of PD data using signal boosting technique
Fig.2  De-noising of PD data using signal boosting technique
Fig.2  De-noising of PD data using signal boosting technique
Fig.2  De-noising of PD data using signal boosting technique
Fig.3  (a) FFT-based feature extraction process (equal sub-bands); (b) DWT-based feature extraction process (octave frequency sub-bands)
Fig.3  (a) FFT-based feature extraction process (equal sub-bands); (b) DWT-based feature extraction process (octave frequency sub-bands)
Fig.3  (a) FFT-based feature extraction process (equal sub-bands); (b) DWT-based feature extraction process (octave frequency sub-bands)
Fig.3  (a) FFT-based feature extraction process (equal sub-bands); (b) DWT-based feature extraction process (octave frequency sub-bands)
Fig.3  (a) FFT-based feature extraction process (equal sub-bands); (b) DWT-based feature extraction process (octave frequency sub-bands)
Fig.4  (a) De-noising of PD data using signal boosting technique; (b) de-noising of PD data using thresholding technique
Fig.4  (a) De-noising of PD data using signal boosting technique; (b) de-noising of PD data using thresholding technique
Fig.4  (a) De-noising of PD data using signal boosting technique; (b) de-noising of PD data using thresholding technique
Fig.4  (a) De-noising of PD data using signal boosting technique; (b) de-noising of PD data using thresholding technique
Fig.4  (a) De-noising of PD data using signal boosting technique; (b) de-noising of PD data using thresholding technique
Fig.5  A PD classification system
Fig.5  A PD classification system
Fig.5  A PD classification system
Fig.5  A PD classification system
Fig.5  A PD classification system
Fig.6  (a) = , where there are coefficients and equations; (b) >, where there are a larger number of equations compared to the number of coefficients; (c) < where there are a larger number of coefficients compared to the number of equations
Fig.6  (a) = , where there are coefficients and equations; (b) >, where there are a larger number of equations compared to the number of coefficients; (c) < where there are a larger number of coefficients compared to the number of equations
Fig.6  (a) = , where there are coefficients and equations; (b) >, where there are a larger number of equations compared to the number of coefficients; (c) < where there are a larger number of coefficients compared to the number of equations
Fig.6  (a) = , where there are coefficients and equations; (b) >, where there are a larger number of equations compared to the number of coefficients; (c) < where there are a larger number of coefficients compared to the number of equations
Fig.6  (a) = , where there are coefficients and equations; (b) >, where there are a larger number of equations compared to the number of coefficients; (c) < where there are a larger number of coefficients compared to the number of equations
Fig.7  (a) Coefficients of for a particular Class 1 test vector; (b) residual error, , calculated for each class
Fig.7  (a) Coefficients of for a particular Class 1 test vector; (b) residual error, , calculated for each class
Fig.7  (a) Coefficients of for a particular Class 1 test vector; (b) residual error, , calculated for each class
Fig.7  (a) Coefficients of for a particular Class 1 test vector; (b) residual error, , calculated for each class
Fig.7  (a) Coefficients of for a particular Class 1 test vector; (b) residual error, , calculated for each class
Fig.8  Classifier fusion process using SRC and SVM classifiers, e.g.,
Fig.8  Classifier fusion process using SRC and SVM classifiers, e.g.,
Fig.8  Classifier fusion process using SRC and SVM classifiers, e.g.,
Fig.8  Classifier fusion process using SRC and SVM classifiers, e.g.,
Fig.8  Classifier fusion process using SRC and SVM classifiers, e.g.,
class typetraining datatest data
Class 1: surface discharge4422
Class 2: corona discharge4225
Class 3: internal discharge3421
Class 4: corona discharge with polarity reversal238
Class 5: surface discharge with polarity reversal4322
Tab.1  Test and training data for five classes
Fig.9  Waveforms of the five different classes of partial discharge used in this paper
Fig.9  Waveforms of the five different classes of partial discharge used in this paper
Fig.9  Waveforms of the five different classes of partial discharge used in this paper
Fig.9  Waveforms of the five different classes of partial discharge used in this paper
Fig.9  Waveforms of the five different classes of partial discharge used in this paper
classifierfeatures (C0, C1, C2, …, C6)
FFT (7 equal bands)DWT (7 octave bands)
support vector machine (SVM)96.9%90.8%
sparse representation classifier (SRC)82.7%79.6%
fused classifier outputs97.9%96.9%
Tab.2  Classification accuracy for SVM, SRC, and the fused classifier outputs
classesFFT (7 equal bands) with SVM classifierFFT (7 equal bands) with SRC classifierfused classifier outputs
123451234512345
1221000225420221000
2024010020510024000
3002100001200002100
4000600005000070
5000122000022000122
accuracy96.9%82.7%97.9%
Tab.3  Confusion matrix for FFT feature with two classifiers and the fused outputs
1 Hao L, Lewin P L. Partial discharge source discrimination using a support vector machine. IEEE Transactions on Dielectrics and Electrical Insulation , 2010, 17(1): 189-197
doi: 10.1109/TDEI.2010.5412017
2 Sriram S, Nitin S, Prabhu K M M, Bastiaans M J. Signal denoising techniques for partial discharge measurements. IEEE Transactions on Dielectrics and Electrical Insulation , 2005, 12(6): 1182-1191
3 Kyprianou A, Lewin P L, Efthymiou V, Stavrou A, Georghiou G E. Wavelet packet de-noising for online partial discharge detection in cables and its application to experimental field results. Measurement Science and Technology , 2006, 17(9): 2367-2379
4 Ambikairajah R, Phung B T, Ravishankar J, Blackburn T R, Liu Z. Smart sensors and online condition monitoring of high voltage cables for the smart grid. In: Proceedings of the Fourteenth International Middle East Power Systems Conference (MEPCON) . 2010, 807-811
5 Ambikairajah R, Phung B T, Ravishankar J, Blackburn T R, Liu Z. Novel frequency domain features for the pattern classification of partial discharge signals. In: Proceedings of the XVII International Symposium on High Voltage Engineering . 2011, Paper F-044
6 Kuncheva L, Bezdek J C, Duin R P W. Decision templates for multiple classifier fusion: an experimental comparison. Pattern Recognition , 2001, 34(2): 299-314
doi: 10.1016/S0031-3203(99)00223-X
7 Evagorou D, Kyprianou A, Lewin P L, Stavrou A, Efthymiou V, Metaxas A C, Georghiou G E. Feature extraction of partial discharge signals using the wavelet packet transform and classification with a probabilistic neural network. IET Science, Measurement and Technology , 2010, 4(3): 177-192
8 Ma X, Zhou C, Kemp I J. Interpretation of wavelet analysis and its application in partial discharge detection. IEEE Transactions on Dielectrics and Electrical Insulation , 2002, 9(3): 446-457
doi: 10.1109/TDEI.2002.1007709
9 Wright J, Yang A Y, Ganesh A, Sastry S S, Ma Y. Robust face recognition via sparse representation. IEEE Transactions on Pattern Analysis and Machine Intelligence , 2009, 31(2): 210-227
doi: 10.1109/TPAMI.2008.79 pmid:19110489
10 Elad M. Sparse and Redundant Representations. New York, NY: Springer, 2009
11 Mohimani G H, Babaie-Zadeh M, Jutten C. Fast sparse representation based on smoothed L0 norm. In: Proceedings of the Seventh International Conference on Independent Component Analysis and Signal Separation . 2007, 389-396
12 Kanevsky D, Sainath T N, Ramabhadran B, Nahamoo D. An analysis of sparseness and regularization in exemplar-based methods for speech classification. In: Proceedings of Interspeech 2010 . 2010, 2842 -2845
[1] Xiaojun SUN, Guangming YAN. Time-varying optimal distributed fusion white noise deconvolution estimator[J]. Front Elect Electr Eng, 2012, 7(3): 318-325.
[2] Zhiyuan LIU, Maosong SUN. Can prior knowledge help graph-based methods for keyword extraction?[J]. Front Elect Electr Eng, 2012, 7(2): 242-253.
[3] Junxia GU, Xiaoqing DING, Shenjing WANG. Action recognition from arbitrary views using 3D-key-pose set[J]. Front Elect Electr Eng, 2012, 7(2): 224-241.
[4] Hongxia ZHANG, Yanning ZHANG, Zhe GUO, Zenggang LIN, Chao ZHANG. 3D face recognition based on principal axes registration and fusing features[J]. Front Elect Electr Eng Chin, 2011, 6(2): 347-352.
[5] Xudong ZHAO, Peng LIU, Jiafeng LIU, Xianglong TANG. Feature extraction for classification of different weather conditions[J]. Front Elect Electr Eng Chin, 2011, 6(2): 339-346.
[6] Guizhu FENG, Wei CHEN, Zhigang CAO. An efficient cooperative sensing scheme with bandwidth-limited control channel[J]. Front Elect Electr Eng Chin, 2010, 5(4): 456-463.
[7] Deqiang HAN, Chongzhao HAN, Yi YANG, . Modified center-based feature line classification approach[J]. Front. Electr. Electron. Eng., 2010, 5(2): 173-178.
[8] Xiaojun SUN, Zili DENG. Distributed fusion white noise deconvolution estimators[J]. Front Elect Electr Eng Chin, 2009, 4(3): 270-277.
[9] Wei DONG, Jiandong LI, Zhuo LU, Linjing ZHAO. Parameter estimation for MIMO system based on MUSIC and ML methods[J]. Front Elect Electr Eng Chin, 2009, 4(2): 161-165.
[10] Yun ZHANG, Boqin FENG, Shouqiang MA, Lianmeng LIU. Text clustering based on fusion of ant colony and genetic algorithms[J]. Front Elect Electr Eng Chin, 2009, 4(1): 15-19.
[11] SUN Xiaojun, ZHANG Peng, DENG Zili. Self-tuning decoupled fusion Kalman filter based on the Riccati equation[J]. Front. Electr. Electron. Eng., 2008, 3(4): 459-464.
[12] WANG Jiangtao, YANG Jingyu. Relative discriminant coefficient based multi-cue fusion for robust object tracking[J]. Front. Electr. Electron. Eng., 2008, 3(3): 274-282.
[13] LI Min, FENG Xiangchu. Image restoration using total variation and anisotropic diffusion equation[J]. Front. Electr. Electron. Eng., 2007, 2(4): 400-403.
[14] HU Zhilan, LIN Xinggang, YAN Hong. Document image retrieval based on multi-density features[J]. Front. Electr. Electron. Eng., 2007, 2(2): 172-175.
[15] LIU Jun-min, ZHANG Chen, LIU Shi. A TDOA location algorithm based on data fusion[J]. Front. Electr. Electron. Eng., 2006, 1(3): 330-333.
Viewed
Full text


Abstract

Cited
  Shared   
  Discussed