A New Hybrid Feature Extraction Method for Partial Discharge Signals Classification

• Autorzy: Liao, R. , Wang, K. , Yang, L. , Duan, L. , Li, J.

Warianty tytułu

PL

Hybrydowa metoda analizy obrazu do klasyfikacji wyładowań niezupełnych

• Języki publikacji: EN

Abstrakty

EN

In this paper, a new hybrid feature extraction method combining adaptive optimal radially Gaussian kernel (AORGK) time-frequency representation with two dimensional nonnegative matrix factorization (2DNMF) is proposed for partial discharge (PD) classification. Firstly, AORGK is applied to obtain the time-frequency matrices of PD ultra-high-frequency (UHF) signals. Then 2DNMF is employed to compress the AORGK amplitude (AORGKA) matrices to extract various feature vectors with different (d1, d2) combinations, i.e. (5, 5), (5, 10), (10, 5) and (10, 10). Finally, the extracted features are classified by fuzzy k nearest neighbor (FkNN) classifier and back propagation neural network (BPNN). 600 samples sam pled from four typical artificial defect models in Laboratory are adopting for testing of the proposed feature extraction algorithm. It is shown that the successful rate by FkNN and BPNN are all higher than 80%, and FkNN has superior classification accuracies than BPNN under four circumstances of (d1, d2) combinations. In addition, FkNN achieves the highest classification accuracy 93.73% with (10, 5) combination. The results demonstrate that it is feasible to apply the proposed algorithm to PD signal classification.

PL

W artykule przedstawiono nową hybrydową metodę klasyfikacji wyładowań niezupełnych (ang. Partial Discharge), wykorzystującą algorytm AORGK (ang. Adaptive Optimal Radially-Gaussian Kernel) o nieujemnej, matrycowej faktoryzacji dwuwymiarowej (ang. 2-Dimensional Nonnegative Matrix Factorization). W metodzie wykorzystano także algorytm k najbliższych sąsiadów oparty na teorii zbiorów rozmytych (ang. Fuzzy k Nearest Neighbour Classifier) oraz sieci neuronowe (ang. Back Propagation Neural Network).

Słowa kluczowe

PL

wyładowanie niezupełne; wybór cech obrazu; AORGK; 2DNMF; FkNNC

EN

partial discharge; feature extraction; adaptive optimal radially Gaussian kernel (AORGK); fuzzy k nearest neighbor classifier

• Czasopismo: Przegląd Elektrotechniczny
• Rocznik: 2012
• Tom: R. 88, nr 11a
• Strony: 191-195
• Opis fizyczny: Bibliogr. 23 poz., tab., rys.

Twórcy

autor

Liao, R.

autor

Wang, K.

autor

Yang, L.

autor

Duan, L.

autor

Li, J.

• State Key Laboratory of Power Transmission Equipment & System Security and New Technology, Chongqing University,

Bibliografia

• [1] Stone G.C., Partial discharge diagnostics and electrical equipment insulation condition assessment, IEEE Trans. Diel. Electr. Insul.,12 (2005), No. 5, 891-904
• [2] Wang M. H., Partial discharge pattern recognition of current transformers using an ENN, IEEE Trans. Power Del. 20 (2005), No. 3, 1984-1990
• [3] Mazzetti C., Mascioli F.M.F. , Baldini M. , Panella M. , Risica R. , Bartnikas R. , Partial discharge pattern recognition by neuro-fuzzy networks in heat-shrinkable joints and terminations of XLPE insulated distribution cables, IEEE Trans. Power Del. 13 (2006), No. 4, 1035-1044
• [4] Gu F.C., Chang H.C., Chen F.H., Kuo C.C. Partial discharge pattern recognition of power cable joints using extension method with fractal feature enhancement, Expert Syst. Appl., 39 (2012), No. 3, 2804-2812
• [5] Abdel -Galil T.K., Hegazy Y.G., Salama M.M.A., Bartnikas R., Fast match-based vector quantization parial discharge pulse pattern recognition, IEEE Trans. Instru. Measur., 54 (2005), No. 1, 3-9
• [6] Chang C.S., Jin J., Chang C.,Hoshino T., Hanai M., Kobayashi N., Online source recognition of partial discharge for gas insulated substations using independent component analysis, IEEE Trans. Diel. Electr. Insul., 13 (2005), No. 4, 892-902
• [7] Jiang T.Y., L i J., Zheng Y.B., Sun C.X., Improved bagging algorithm for pattern recognition in UHF signals of partial discharges, Energies, 4 (2011), No. 7, 1087-1101
• [8] Pinpart T., Judd M.D., Differentiating between partial discharge sources using envelope comparison, IET Scie. Meas.Technol., 4 (2010), No. 5, 256-267
• [9] Abdel -Gal i l T.K., E I -Hag A.H., Gaouda A.M., Salama M.M.A., Bartnikas R., De-noising of partial discharge signal using eigen-decomposition technique, IEEE Trans. Diel. Electr. Insul., 15 (2008), No. 6, 1657-1662
• [10] Cifrek M., Medved V., Tonkovic S., Ostojic S., Surface EMG based muscle fatigue evaluation in biomechanics, Clin. Biomech., 24 (2009), No. 4, 327-340
• [11] Shi G.M., Chen X.Y., Song X.X., Qi F., Ding A.L., Signal matching wavelet for ultrasonic flaw detection in high background noise, IEEE Trans. Ultrason. FERR., 58 (2011), No. 4, 776-787
• [12] Roshan-Ghias A., Shamsollahi M.B., Mobed M., Behzad M., Estimation of modal parameters using bilinear joint time-frequency distributions, Mech. Syst. Signal Proces., 21 (2007), No.5, 2125-2136
• [13] Jones D.L., Baraniuk R.G., An adaptive optimal-kernel time-frequency representation, IEEE Trans. Signal Proces., 43 (1995), No. 10, 2361-2371
• [14] Baraniuk R.G., Jones D.L., A radially-Gaussian, signal-dependent time-frequency representation, ICASSP-91, 5 (1991), 3181-3184
• [15] Lee D.D., Seung H.S., Learning the parts of objects by non-negative matrix factorization, Nature, 401 (1999), No. 6755, 788-791
• [16] Lin C.J., Projected gradient methods for nonnegative matrix factorization, Neural Comput., 19 (2007), No. 10, 2756-2779
• [17] Paatero P., Tapper U., Positive matrix factorization: a non-negative factor model with optimal utilization of error estimates of data values, Envirometrics, 5 (1994), No. 2, 111-126
• [18] Li B., Zhang P.L., Liu D.S., Mi S.S., Ren G.Q., Tian H., Feature extraction for rolling element bearing fault diagnosis utilizing generalized S transform and two-dimensional non-negative matrix factorization, J. Sound Vib., 330 (2011), No. 10, 2388-2399
• [19] Keller J.M., Gray M.R., Givens J.A., A fuzzy k-nearest neighbor algorithm, IEEE Trans. Syst. Man Cyber., 15 (1985), No. 4, 580-585
• [20] Li J., Sun C.X., Gryzbowski S., Taylor C.D., Partial discharge image recognition using a new group of features, IEEE Trans. Diel. Electr. Insul., 13 (2006), No. 6, 1245-1253
• [21] Guo Z.H., Wu J., Lu H.Y., Wang J.Z., A case study on a hybrid wind speed forecasting method using BP neural network, Knowl.-based Syst., 24 (2011), No. 7, 1048-1056
• [22] Demuth H., Beale M., Neural network toolbox for use with MATLAB, The MathWorks, Inc. 2001
• [23] Feng C.X.J., Gowrisankar A.C., Smith A.E., Yu Z.G.S., Practical guidelines for developing BP neural network models of measurement uncertainty data, J. Manuf. Syst., 25 (2006), No. 4, 239-250