Surface vibration source number estimation of the 220kV running power transformer based on the Wavelet-SVD-Clustering

The existing blind source separation algorithm has two problems for the separation of the vibration signal of the transformer. One is the uncertain number of the transformer vibration source. The other is the vibration measuring point, which is not sure after the signal separation. Aiming at these problems, this paper apply the Wavelet-SVD-Clustering analysis method on determinating the number of sources by using one single sensor acquisition of transformer tank vibration signal. The effectiveness of the method is verified by simulation and experimental data analysis. The method of this paper provides guarantee for the correct analysis and separation of the transformer vibration signal.

[1]  Guillaume Gelle,et al.  BLIND SOURCE SEPARATION: A TOOL FOR ROTATING MACHINE MONITORING BY VIBRATIONS ANALYSIS? , 2001 .

[3]  Shixi Yang.  INDEPENDENT COMPONENT ANALYSIS BASED NETWORKS FOR FAULT FEATURES EXTRACTION AND CLASSIFICATION OF RATATING MACHINES , 2004 .

[4]  Evon M. O. Abu-Taieh,et al.  Comparative Study , 2020, Definitions.

[5]  Hsien-Tsai Wu,et al.  Source number estimators using transformed Gerschgorin radii , 1995, IEEE Trans. Signal Process..

[6]  Scott C. Douglas,et al.  Convolutive blind separation of speech mixtures using the natural gradient , 2003, Speech Commun..

[7]  Abdelhak M. Zoubir,et al.  A comparative study on source number detection , 2003, Seventh International Symposium on Signal Processing and Its Applications, 2003. Proceedings..

[8]  Takeshi Yamada,et al.  Estimation of the number of sound sources using support vector machines and its application to sound source separation , 2003, 2003 IEEE International Conference on Acoustics, Speech, and Signal Processing, 2003. Proceedings. (ICASSP '03)..

[9]  Tom Minka,et al.  Automatic Choice of Dimensionality for PCA , 2000, NIPS.

[10]  Lucas C. Parra,et al.  Convolutive Blind Source Separation Methods , 2008 .

[11]  Boualem Boashash,et al.  Separating More Sources Than Sensors Using Time-Frequency Distributions , 2005, EURASIP J. Adv. Signal Process..

[12]  Christine Servière,et al.  Blind source separation: a new pre-processing tool for rotating machines monitoring? , 2003, IEEE Trans. Instrum. Meas..