2019/04/04 by Faeza Hafiz, Hafiz, Faeza, Celia Shahnaz +1
Engineering · Materials Science · #FOS: Electrical engineering #Machine Fault Diagnosis Techniques #Magnetic Properties and Applications #Power Quality and Harmonics #Power Transformer Diagnostics and Insulation #Signal Processing (eess.SP) #electronic engineering #information engineering
paper · pdf · doi:10.48550/arxiv.1904.02836
openalex publication_date 2019/04/04 · openalex created_date 2022/07/29 · openalex updated_date 2026/07/28
The aim of this paper is to propose a new approach for the pattern\nrecognition of power quality (PQ) disturbances based on Empirical mode\ndecomposition (EMD) and k Nearest Neighbor (k-NN) classifier. Since EMD\ndecomposes a signal into intrinsic mode functions (IMF) in time-domain with\nsame length of the original signal, it preserves the information that is hidden\nin Fourier domain or in wavelet coefficients. In this proposed method, power\nsignals are decomposed into IMFs in EMD domain. Due to the presence of\nnon-linearity and noise on the original signal, it is hard to analyze them by\nsecond order statistics. Thus, an effective feature set is developed\nconsidering higher order statistics (HOS) like variance, skewness, and kurtosis\nfrom the decomposed first three IMFs. This feature vector is fed into different\nclassifiers like k-NN, probabilistic neural network (PNN), and radial basis\nfunction (RBF). Among all the classifiers, k-NN showed higher classification\naccuracy and robustness both in training and testing to detect the PQ\ndisturbance events. Simulation results evaluated that the proposed HOS-EMD\nbased method along with k-NN classifier outperformed in terms of\nclassification accuracy and computational efficiency in comparison to the other\nstate-of-art methods both in clean and noisy environment.\n