Partial Discharge Pattern Classification Research Articles

Study on Deep-Learning Model for Phase Resolved Partial Discharge Pattern Classification Based on Convolutional Neural Network Algorithm

Study on Deep-Learning Model for Phase Resolved Partial Discharge Pattern Classification Based on Convolutional Neural Network Algorithm

Effectiveness of Wavelet Scalogram on Partial Discharge Pattern Classification of XLPE Cable Insulation

Effectiveness of Wavelet Scalogram on Partial Discharge Pattern Classification of XLPE Cable Insulation

Recognition of Fused Partial Discharge Patterns in High Voltage Insulation Systems: A Hybrid DCNN and SVM Based Approach

The performance assessment of insulation materials within the components of a high voltage power system depends upon effective condition monitoring techniques of the insulation. One of the important phenomena to be considered in the diagnostics and performance assessment of high voltage insulation materials is the Partial Discharge (PD) measurement. The improved Deep learning techniques help in the effective identification and classification of various sources of PD mechanism. In this work, Deep Convolution Neural Network (DCNN) technique is proposed for the fusion of several features that are extracted from the PD signals. Three-dimensional Phase Resolved patterns of PD images can be used for training the DNN. For the classification of PD patterns, linear and non-linear type of multi-class support vector machine has been proposed. The results obtained are compared with linear SVM and non-linear SVMs with the polynomial kernel, Radial Basis Function (RBF) kernel, and Sigmoidal kernel. The SVM type with a higher pattern recognition rate is identified to be effective in PD pattern recognition and classification. The results of the proposed work show that the fusion approach of PD patterns supports applying huge PD data sets as input, generated by multiple faults, for effective PD pattern recognition.

부분방전 데이터 처리 방법에 따른 CNN 기반 패턴 분류기의 비교 연구

This study is focused on solution to the problem that may occur due to noise signals when using the image obtained through the conventional partial discharge preprocessing methods as inputs to the CNN-based partial discharge pattern classifier. To solve such problem, a new data preprocessing method is proposed by considering a projection technique. In the data information obtained through the conventional partial discharge preprocessing methods, data information of useless noise signal leads to the problem of performance degradation when constructing highly qualified pattern classifier based on the data information of partial discharge signal. The proposed preprocessing method called as ‘Projection’ in this study is designed to solve this problem and improve a classification performance of CNN-based partial discharge pattern classifier. First of all, through GIS simulation, one-dimensional partial discharge data is obtained as five cases such as corona discharge, floating discharge, insulator discharge, free particle discharge, and steady state (noise) in an environment with a noise signal by using a UHF sensor. After that, by applying the two conventional partial discharge preprocessing methods such as PRPS(Phase Resolved Pulse Sequence), PRPD(Phase Resolved Partial Discharge) and the proposed partial discharge preprocessing method, one-dimensional partial discharge data is transformed into 3 data types such as image set of PRPS, image set of PRPD, and image set of Projection. Finally each image set is used as inputs to the designed CNN-based partial discharge pattern classifier. Through the comparative analysis of the feature maps of each layer in CNN as well as the performance accuracy of partial discharge pattern classification for each image set, the superiority of the proposed preprocessing method is demonstrated.

Development of Hypergraph Based Improved Random Forest Algorithm for Partial Discharge Pattern Classification

Precise partial discharge (PD) detection is a key factor in anticipating insulation failures. The continuous efforts of researchers have led to the design of a variety of algorithms focusing on PD pattern classification. However, the trade-off between features taken up for classification and the detection rate continues to pose considerable challenges in terms of feature selection from acquired data, increased computing time, and so on. In this article, a Hypergraph (HG) based improved Random Forest (RF) algorithm by employing the Recursive Feature Elimination (RFE) algorithm (HG-RF-RFE), has been developed for PD source classification. HG representation of data is considered for obtaining statistical features, which turn out to be a subset of a set of all hyper edges called Hyper statistical features (Helly, Non-Helly, and Isolated hyper edges). HG-RF-RFE takes hyper statistical features and hyper edges as features for classification. The algorithm’s efficiency is tested against noise-free PD data obtained from SASTRA High Voltage Laboratory, and large-sized noisy PD data obtained from High-Voltage Research and Test Laboratory at Universidad Tecnica Federico Santa Maria (LIDAT). The robustness of the proposed algorithm is tested with both time and phase domain PD features using the Mathews Correlation Coefficient (MCC), harmonic mean-based feature Score (F1 Score) as evaluation metrics, and by k-fold validation technique. The proposed HG-RF-RFE achieved 98.8% accuracy with minimal features and significantly reduces computation time without compromising accuracy. It is worth mentioning that the HG-RF-RFE technique is superior to many state of the art algorithms in terms of feature elimination and classification accuracy.

Study on partial discharge pattern classification of GIS by adaptive neuro-fuzzy inference system

In this study, a PD defect identification system based on linear discriminant a nalysis and adaptive neural fuzzy inference system is developed. in this study, the par tial discharge signs of typical defects of the GIS measurement were measured with a measuring device. the corresponding characteristic parameters extracted from the GIS measurement were filtered according to the optimization algorithm and stored in the form of three-dimensional discharge mode (q-φ-t). On this basis, the 3d discharge mode is converted into statistical parameters. statistical pd features can be used for defect classification purposes, but it is clear that for this goal, the treatment of sovolume data base will be inefficient. therefore, it would be very helpful to select those features t hat are more effective and more efficient to identify and distinguish potential pd defects through their relevant statistical parameters. Select feature extraction techniques for dimensionality reduction. In solving pattern classification problems, feature extraction methods are usually used as preprocessing techniques, which cannot only reduce the computational complexity, but also achieve better classification performance by reducing irrelevant and redundant information in the data.

Fractal‐based autonomous partial discharge pattern recognition method for MV motors

On-line partial discharge (PD) monitoring is being increasingly adopted to improve the asset management and maintenance of medium-voltage (MV) motors. This study presents a novel method for autonomous analysis and classification of motor PD patterns in situations where a phase-reference voltage waveform is not available. The main contributions include a polar PD (PPD) pattern and a fractal theory-based autonomous PD recognition method. PPD pattern that is applied to convert the traditional phase-resolved PD pattern into a circular form addresses the lack of phase information in on-line PD monitoring system. The fractal theory is then presented in detail to address the task of discrimination of 6 kinds of single source and 15 kinds of multi-source PD patterns related to motors, as outlined in IEC 60034. The classification of known and unknown defects is calculated by a method known as centre score. Validation of the proposed method is demonstrated using data from laboratory experiments on three typical PD geometries. This study also discusses the application of the proposed techniques with 24 sets of on-site PD measurement data from 4 motors in 2 nuclear power stations. The results show that the proposed method performs effectively in recognising not only the single-source PD but also multi-source PDs.

SVM-based Partial Discharge Pattern Classification for GIS

Partial discharges (PD) occur when there are localized dielectric breakdowns in small regions of gas insulated substations (GIS). It is of high importance to recognize the PD patterns, through which we can diagnose the defects caused by different sources so that predictive maintenance can be conducted to prevent from unplanned power outage. In this paper, we propose an approach to perform partial discharge pattern classification. It first recovers the PRPD matrices from the PRPD2D images; then statistical features are extracted from the recovered PRPD matrix and fed into SVM for classification. Experiments conducted on a dataset containing thousands of images demonstrates the high effectiveness of the method.

High noise tolerance feature extraction for partial discharge classification in XLPE cable joints

Cable joints are the weakest point in cross-linked polyethylene (XLPE) cables and are susceptible to insulation failures. Partial discharge (PD) analysis is a vital tool for assessing the insulation quality in cable joints. Although many works have been done on PD pattern classification, it is usually performed in a noise-free environment. Also, works on PD pattern classification are mostly done on lab fabricated insulators, where works on actual cable joint defects are less likely to be found in literature. Therefore, in this work, classifications of cable joint defect types from partial discharge measurement under noisy environment were performed. Five XLPE cable joints with artificially created defects were prepared based on the defects commonly encountered on-site. A high noise tolerance principal component analysis (PCA)-based feature extraction was proposed and compared against conventional input features such as statistical and fractal features. These input features were used to train the classifiers to classify each defect type in the cable joint samples. Classifications were performed using Artificial Neural Networks (ANN) and Support Vector Machine (SVM). It was found that the proposed PCA features displayed the highest noise tolerance with the least performance degradation compared to other input features under noisy environment.

Extraction of wave parameter for classification of partial discharge using Pearson's correlation method

Partial discharge (PD) behaviour in electrical apparatus leads to the identification of the degradation of insulating materials. The major problem here is identifying the different types of PD discharges that affect the insulation system. In general the characteristics of PD pulse will be in the form of internal, corona and surface discharges (both in air and oil conditions). Thus the classification of PD patterns aims to recognise the discharges from unknown source. In this paper a new enhanced technique is used to identify the PD discharges namely Pearson's correlation coefficient techniques. Starting with PD data on different families of specimen, extract φ-q-n PD patterns (3D patterns) and apply to the enhanced techniques, then segmented the 3D patterns in to number of samples. Thereafter, by evaluating statistical operators like skewness, kurtosis and Pearson's correlated values from the PD signals and recognise the different PD patterns using artificial neural networks and adaptive neuro fuzzy interference system.

Extraction of Wave Parameter for Classification of Partial Discharge using Pearson’s Correlation Method

Partial discharge (PD) behaviour in electrical apparatus leads to the identification of the degradation of insulating materials. The major problem here is identifying the different types of PD discharges that affect the insulation system. In general the characteristics of PD pulse will be in the form of internal, corona and surface discharges (both in air and oil conditions). Thus the classification of PD patterns aims to recognise the discharges from unknown source. In this paper a new enhanced technique is used to identify the PD discharges namely Pearson’s correlation coefficient techniques. Starting with PD data on different families of specimen, extract Φ-q-n PD patterns (3D patterns) and apply to the enhanced techniques, then segmented the 3D patterns in to number of samples. Thereafter, by evaluating statistical operators like skewness, kurtosis and Pearson’ss correlated values from the PD signals and recognise the different PD patterns using artificial neural networks and adaptive neuro fuzzy interference system.

Applying WPD and SVD to Classification of EM Wave Induced by Partial Discharge in Power Transformer

Partial discharge (PD) current is an impulse signal at nanosecond level, which can generate electromagnetic (EM) wave containing broadband frequency information. The frequency band of EM signal is from MHz up to GHz. Due to different PD patterns, impulse currents with different shapes induce different EM waves containing different frequency information. Therefore, using the features extracted from frequency domain of EM signals, the classification of PD patterns can be effectively got. It is good to use wavelet or wavelet packet decomposition to select features. However, if the decomposition level is too shallow to find enough effective features, it cannot group the EM signals to the right pattern. On the contrary, although it is easier to find features to distinguish the PD pattern if the decomposition level is deep, there will be a lot of redundancy variables and it is hard to select features among so many variables. In this paper, a method is presented, which selected features in the whole decomposition tree instead of selecting among the leaf node of the tree, because more potential features can be found in the whole tree. With the present method, it is possible not only to get enough features, but also to eliminate the redundancy variables effectively. In order to validate the method, large EM signals from four PD patterns in a power transformer are acquired as the training data and testing data for feature selection and classification, and three common classification methods are introduced to classify the PD patterns using the features selected by the method. Most of the classification results are satisfactory indicating that the proposed method is effective.

Investigation on Using Fractal Geometry for Classification of Partial Discharge Patterns

Classification of Partial discharge (PD) patterns is an important tool to determine the sources of the PD. Also the features that specify of the patterns can be used monitor the condition of the electrical insulation. A PD patterns classification approach of artificial partial discharge sources by using neural networks is proposed in this paper. The classification process was based on features generated from three- dimension PD patterns. These features were obtained with the aid of fractal geometry. The box counting technique is used to generate the fractal features (fractal dimension and lacunarity). Each PD pattern has a unique Fractal dimension and several values for lacunarity depends upon the size of the box. Therefore each PD pattern is characterized by two features but with several quantities. The number of input features and their contents of information determine the performance of neural networks based classification system. In this paper an attempt to improve the performance of PD classification systems by minimizing the number of the input features was introduced. This was done by investigating the ability the different lacunarity values to classify different PD sources. A wide range of box size was investigated. The obtained results show that the lacunarity values generated from a very narrow range of box sizes has the maximum ability for PD classification.

The fusion of classifier outputs to improve partial discharge classification

The detection of partial discharge signals and classifying its patterns is an area of interest in the analysis of defects in high voltage cables. This paper investigates a filter-bank based approach to extract frequency domain based features to represent partial discharge signals. By applying the fast Fourier transform, the sampled partial discharge data are mapped into equivalent discrete frequency bins, which are then grouped into N equal sub-bands and also octave sub-bands, each providing N-dimensional features for partial discharge pattern classification. Two classifiers, namely, the support vector machine and the sparse representation classifier, are implemented and their outputs are fused, in order to improve the accuracy of classifying partial discharge. Classification accuracy is also compared with wavelet domain based octave frequency sub-band features.

Identification of Insulation Defects in Cryogenic Dielectric Materials for the HTS Power Applications

Recently, various high temperature superconducting (HTS) power applications have been developed and prepared for field tests and commercial applications. Comparing to conventional power applications, it could offer several advantages such as reduced size and weight, high efficiency, decreased losses, no oil, nonflammable and decrease of CO2 emissions. Besides overload operation is possible with no loss of lifetime. For HTS power applications in low temperature and high voltage environments, partial discharge (PD) measurements in cryogenic dielectric materials of HTS power applications are very important because PD was regarded as primary source for ageing and breakdown of cryogenic materials. One of the diagnostic methods for safety of the power components, the detection of PD taking place inside the apparatus has been widely investigated. The first method, phase resolved partial discharge (PRPD) Analysis was developed in the early 1970s taking the phase information of the applied AC voltage into account. We also proposed a pattern analysis method named chaotic analysis of PD(CAPD) for PDs occurred in liquid nitrogen, considering three normalized parameters obtained from the values between two consecutive PD pulses: amplitude difference (Pt), occurring time difference (Tt) and correlation between Tt and Pt. This pattern analysis method can identify the type of defects by means of PD pattern classification without employing the phase information of the applied voltage signal. For the experimental investigation, three artificial defects have been fabricated considering possible defects formed during the manufacturing process of HTS power applications: turn to turn insulation, floating particle and protrusion. And PD signals originated from these artificial defects are measured and analyzed by means of CAPD. Throughout this work, it seems that the correlation between the consecutive PD pulses, depending on the nature PDs, could be clarified by CAPD.

Effectiveness of Partition and Graph Theoretic Clustering Algorithms for Multiple Source Partial Discharge Pattern Classification Using Probabilistic Neural Network and Its Adaptive Version: A Critique Based on Experimental Studies

Partial discharge (PD) is a major cause of failure of power apparatus and hence its measurement and analysis have emerged as a vital field in assessing the condition of the insulation system. Several efforts have been undertaken by researchers to classify PD pulses utilizing artificial intelligence techniques. Recently, the focus has shifted to the identification of multiple sources of PD since it is often encountered in real‐time measurements. Studies have indicated that classification of multi‐source PD becomes difficult with the degree of overlap and that several techniques such as mixed Weibull functions, neural networks, and wavelet transformation have been attempted with limited success. Since digital PD acquisition systems record data for a substantial period, the database becomes large, posing considerable difficulties during classification. This research work aims firstly at analyzing aspects concerning classification capability during the discrimination of multisource PD patterns. Secondly, it attempts at extending the previous work of the authors in utilizing the novel approach of probabilistic neural network versions for classifying moderate sets of PD sources to that of large sets. The third focus is on comparing the ability of partition‐based algorithms, namely, the labelled (learning vector quantization) and unlabelled (K‐means) versions, with that of a novel hypergraph‐based clustering method in providing parsimonious sets of centers during classification.

Robust Heteroscedastic Probabilistic Neural Network for multiple source partial discharge pattern recognition – Significance of outliers on classification capability

Among various insulation diagnostic techniques utilized by researchers and personnel handling power equipment, partial discharge (PD) recognition and analysis has emerged as a vital methodology since it is inherently a non-intrusive testing strategy. Of late, the focus of researchers has shifted to the identification and classification of multiple sources of PD since it is most often encountered in practical insulation systems of power apparatus. Researchers have carried out studies to recognize multi-source PD and expounded the difficulties experienced in discriminating such discharge patterns. It has also been observed that identification of such patterns becomes increasingly difficult with the degree of overlap. Review of recent research studies indicates that classification of fully overlapped patterns is yet an unresolved issue and that techniques such as Mixed Weibull Functions, neural networks (NN) and Wavelet Transformation have been attempted with reasonable degree of success for single source and partially overlapped PD patterns only. This research study focuses on extending the previous work attempted by the authors in utilizing the novel approach of Heteroscedastic Probabilistic Neural Network (HRPNN) for classification of single source PD patterns to that of multiple PD sources also. Further, a Robust Heteroscedastic Probabilistic Neural Network (RHRPNN) is implemented for the classification of multi-source PD patterns. The RHRPNN utilizes the jackknife procedure for handling problems associated with training the neural network due to the presence of outliers, thus providing a compact yet effective set of centers in terms of probability density functions. In addition to the previously utilized traditional statistical operators in the pre-processing phase, a Two Pass Split Window (TPSW) scheme has also been developed to study and compare the classification capability of the RHRPNN with that of HRPNN. Detailed analysis of the performance of RHRPNN is carried out to ascertain the influence of the smoothing parameter in classifying PD patterns, to determine the role played by the pre-processing techniques during classification and to find the significance of the parsimonious set of centers in eliminating the effect of outliers during classification. Finally, the ability of the HRPNN and the RHRPNN in classifying large dataset multiple source PD patterns obtained from varying applied voltages is analyzed for its further applicability in real-time PD pattern recognition studies.

Identification of insulation defects in gas-insulated switchgear by chaotic analysis of partial discharge

The recent increase in the number of failures during the installation and insulation of gas-insulated switchgear (GIS) has led to a need for a reliable risk assessment technique. Accordingly, in this study, a fundamental database of UHF partial discharge (PD) patterns corresponding to different types of defects is presented for risk assessment and for the observation and diagnosis of the state of insulation of GIS at field sites; the patterns have been obtained by modelling a number of defects that have been reported to be the most critical in GIS. For the realisation of a system that can simultaneously detect and analyse UHF PD (it is internationally accepted that the use of such a system is the most effective method to diagnose GIS insulation), a wideband UHF sensor and amplifier are designed and fabricated. The system operation is then investigated. In addition, a chaotic analysis of partial discharge (CAPD) is proposed. This analysis can identify the type of defect by means of PD pattern classification without employing the phase information of the applied voltage signal. The proposed CAPD can replace the conventional phase-resolved partial discharge (PRPD) analysis and can be employed at field sites when the phase information is unavailable. Especially, the PD patterns of free-moving conducting particles, known to be very important in GIS, can be distinguished from those of other defects when using the CAPD method, which is not the case when the PRPD analysis is being performed.

Digital Image Processing for On-line Classification of Partial Discharge Patterns in Electrical Equipment

This paper presents a new method to analyze partial discharge (PD) digital patterns obtained from power equipment measurements to automatically classify PD patterns than can issue alarms in continuous monitoring systems. The method has two steps, in the firs one, a preclassification analyses the information contained in an *.dat type file and it generates a JPG image of the PD pattern. Then a color to grey reduction is performed to account for random variations and sporadic pulses are removed to discriminate sporadic activity and background interference. The second step involves the use of graphical filters and morphological operations like image opening and closing to segment and define regions of interest. Finally, a contour following technique is applied with Gaussian regression to identify corona from partial discharges. Obtained results include arrow pointing images with text labels that identify different PD types. This feature will greatly improve on-line monitoring system diagnosis.

Partial discharge pattern classification using composite versions of probabilistic neural network inference engine

A major requirement of any power apparatus is the reliable performance of its insulation. The incidence of minor flaws and irregularities such as voids, surface imperfections, in the electrical insulation is however inevitable and lead to partial discharge (PD). Since each defect has a unique degradation mechanism, it is imperative to ascertain the correlation between the discharge patterns and the type of defect in order to evaluate the quality of the insulation. Efforts to correlate discharge patterns with the type of defects have been undertaken by several researchers. Though encouraging attempts to recognize and classify simple PD defect sources have been reported, misclassifications still occur, which affect the assessment of the index of the insulating degradation. A Composite Probabilistic Neural Network Inference System has been devised and elucidated in this research using two versions of Probabilistic Neural Network. The inference is obtained based on the outcome to innovatively conceived fourteen unique characteristic vector inputs to enable an accurate and reliable decision in the classification of complex stochastic PD patterns thus obviating the necessity of skilled operators. Validation of the fingerprints of PD patterns has also been carried out using well-established techniques.