Characterization of acoustic signals due to surface discharges on H.V. glass insulators using wavelet radial basis function neural networks

• Published in: Applied soft computing 2012-04, Vol.12 (4), p.1239-1246
• Main Authors: Al-geelani, Nasir A., Piah, M. Afendi M., Shaddad, Redhwan Q.
• Format: Article
• Language: English
• Subjects: Acoustic signal; Dry bands; Glass insulator; RBF-NN; Surface discharge; Wavelet transform
• ISSN: 1568-4946; 1872-9681
• DOI: 10.1016/j.asoc.2011.12.018

[Display omitted] ► We presented a methodology for a detection system which incorporates advanced digital signal processing. ► The high rate of classification is due to the preprocessing of data by the wavelet transform. ► We design an intelligent model that can track, identify, characterize and diagnose the surface discharge patterns. ► Increasing the triggering sensitivity of the cleaning operation. A hybrid model incorporating wavelet and radial basis function neural network is presented which is used to detect, identify and characterize the acoustic signals due to surface discharge activity and hence differentiate abnormal operating conditions from the normal ones. The tests were carried out on cleaned and polluted high voltage glass insulators by using surface tracking and erosion test procedure of international electrotechnical commission 60587. A laboratory experiment was conducted by preparing the prototypes of the discharges. This study suggests a feature extraction and classification algorithm for surface discharge classification, which when combined together reduced the dimensionality of the feature space to a manageable dimension, by “marrying” the wavelet to radial basis function neural network very high levels of classification are achieved. Wavelet signal treatment toolbox is used to recover the surface discharge acoustic signals by eliminating the noisy portion and to reduce the dimension of the feature input vector. A radial basis function neural network classifier was used to classify the surface discharge and assess the suitability of this feature vector in classification. This learning method is proved to be effective by applying the wavelet radial basis function neural network in the classification of surface discharge fault data set. The test results show that the proposed approach is efficient and reliable.