A novel technique employing DWT-based envelope analysis for detection of power system transients

This study reports a novel technique for detection of power system transients based on discrete wavelet transform based envelope analysis. Existence of transients in power transmission and distribution networks are very common and need sophisticated detection techniques for proper monitoring of these events. In this contribution, at first multi-resolution analysis of two very frequently occurring power system transients namely oscillatory and impulsive transients are done using discrete wavelet transform (DWT). Next, the envelope spectrum of first four detail coefficients are obtained using Hilbert transform and several features are extracted from the selected envelope spectrums of both class of transient signals. Using ANOVA test, statistical analysis of the extracted features have been done to investigate the discrimination capability of the selected features which are finally used as inputs to a support vector machines (SVM) classifier for classification of power system transients. It has been observed that based on DWT envelope analysis and employing SVM classifier 100% classification accuracy is obtained in detection of different types of power system transients.

[1]  Tao Zhang,et al.  Classification of epilepsy EEG signals using DWT-based envelope analysis and neural network ensemble , 2017, Biomed. Signal Process. Control..

[2]  M. K. Khedkar,et al.  Feed-forward Artificial Neural Network–Discrete Wavelet Transform Approach to Classify Power System Transients , 2013 .

[3]  R. Sukanesh,et al.  Power quality disturbance classification using Hilbert transform and RBF networks , 2010, Neurocomputing.

[4]  B. Chatterjee,et al.  Identification of salt and salinity level of 11kV contaminated porcelian disc insulator using STD-MRA analysis of leakage current , 2015, 2015 International Conference on Condition Assessment Techniques in Electrical Systems (CATCON).

[5]  Amitava Chatterjee,et al.  A sparse representation based approach for recognition of power system transients , 2014, Eng. Appl. Artif. Intell..

[6]  Amitava Chatterjee,et al.  A dual-tree complex wavelet transform-based approach for recognition of power system transients , 2015, Expert Syst. J. Knowl. Eng..

[7]  Myeongsu Kang,et al.  Robust condition monitoring of rolling element bearings using de-noising and envelope analysis with signal decomposition techniques , 2015, Expert Syst. Appl..

[8]  Bhim Singh,et al.  Recognition of Single-stage and Multiple Power Quality Events Using Hilbert–Huang Transform and Probabilistic Neural Network , 2015 .

[9]  U. Rajendra Acharya,et al.  Classification of Normal, Neuropathic, and Myopathic Electromyograph Signals Using Nonlinear Dynamics Method , 2011 .

[10]  P. Pillay,et al.  Wavelet Analysis of Power Systems Transients Using Scalograms and Multiresolution Analysis , 1999 .