Partial Discharge Pattern Recognition of Transformers Based on the Gray-Level Co-Occurrence Matrix of Optimal Parameters

Abstract

The partial discharge (PD) is the most common fault of transformers, which is the main factor affecting the stable operation of transformers. Therefore, the PD should be monitored and identified timely to improve the reliability of the transformers. In this paper, a transformer PD pattern recognition algorithm based on the gray-level co-occurrence matrix of optimal parameters and support vector machine (GLCMOP-SVM) is proposed. Firstly, the GLCM of optimal parameters (GLCMOP) is proposed to be determined by calculating the proportion of the off-diagonal elements (PODE) in GLCM. The GLCMOP has the advantage of avoiding the subjectivity of parameter selection and simplifying the calculation process. Then, the phase-resolved partial discharge (PRPD) maps are used as the PD samples and are converted into the GLCMOP to extract the PD features. Moreover, the feature space of the GLCMOP is dimensionally reduced by screening out the features with high distinguishability, which can improve the generalization ability and recognition speed of the classifier. Finally, the SVM classifier is trained to sort the PD samples and recognize the PD types, which include the tip discharge, surface discharge, and air discharge PD types. Lab tests are performed to verify the accuracy and validity of the proposed methodology. Compared with the traditional algorithms based on GLCM, XGBoost (eXtreme Gradient Boosting) and artificial neural network (ANN), the performance of GLCMOP-SVM is better. The GLCMOP-SVM has less memory consumption and faster recognition speed, so it is very suitable for the online and real-time monitoring of PD occurred in the transformers.

A Comprehensive Review of Signal Processing and Machine Learning Technologies for UHF PD Detection and Diagnosis (II): Pattern Recognition Approaches

“…Let (W , b en ) and (W T , b de ) be the parameters of encoder and decoder, respectively, then the code vector H in latent space can be calculated by f (X), and the output X of AE is obtained by g(H). The goal of AE is to minimize the difference between X and X, thus the objective function can be written as: Cross−entropy loss function (19) By stacking multiple AE models sequentially, the stacked AEs network can be obtained, which is a typical DAN model. Besides, there are many other improved versions of DAN models [182], such as the SAE, DAE, CAE, VAE, etc.…”

“…The results showed that compared with existing methods using statistical features, the LBP or HOG features extracted in that paper can increase the average accuracy rate by more than 10%. In [19], Sun et al proposed an improved texture feature extraction method namely GLCMOP and applied it to the 2-D PRPD grayscale images. To reduce the influence of noise on the PRPS graphs, Li et al introduced the SURF algorithm to extract the feature points and feature descriptors of the PRPS grayscale images [20].…”

“…SVM has long been used for type recognition of PD faults. For example, authors of [19], [27], [35], [36], and [47] introduced SVM to classify PD data for different power equipment, and results of these studies are very competitive (all higher than 94%). However, the performance of SVM is highly depend on the setting of hyperparameters like the kernel function and the penalty coefficient.…”

Local energy-based image region segmentation using Legendre polynomials

“…The above matrix specifies the spatial relationship between the intensity of the image pixels. x(n) is the gray level co-occurrence matrix component in the image 23 fw′(x)=∑n=−∞∞|x(n)|2.…”

“…The above matrix specifies the spatial relationship between the intensity of the image pixels. xðnÞ is the gray level co-occurrence matrix component in the image 23 E Q -T A R G E T ; t e m p : i n t r a l i n k -; e 0 0 6 ; 1 1 7 ; 5 3 2…”