Partial Discharge Data Research Articles

Development of Hankel‐SVD hybrid technique for multiple noise removal from PD signature

Detection and measurement of partial discharge (PD) phenomena combined with the separation and identification of PD sources is the way to achieve effective insulation integrity assessment. However, during measurement, PD signals are coupled with interferences (discrete spectral, pulsive, and white noises). Recovering PD signals from such interferences would improve PD source separation (thus identification), but still remains a challenging task. Several denoising methods have been proposed to suppress interferences. However, using a universal method to achieve interference removal is probably impossible, as the characteristics of the interferences are distinct. This study proposes a novel low-rank H-Matrix-based singular value decomposition (SVD) filter (H-SVD) that removes different types of interferences. Denoising is done by projecting the measured pulse in a lower dimensional signal space. To assess the effectiveness of the proposed method, H-SVD filter is first applied to simulated PD data and later on real-time PD data with the introduction of three different types of synthetic noises. The results of the evaluation metrics confirm that H-SVD has significant performance improvements compared to existing state-of-the-art PD denoising methods.

Partial Discharge Data Matching Method for GIS Case-Based Reasoning

With the accumulation of partial discharge (PD) detection data from substation, case-based reasoning (CBR), which computes the match degree between detected PD data and historical case data provides new ideas for the interpretation and evaluation of partial discharge data. Aiming at the problem of partial discharge data matching, this paper proposes a data matching method based on a variational autoencoder (VAE). A VAE network model for partial discharge data is constructed to extract the deep eigenvalues. Cosine distance is then used to calculate the match degree between different partial discharge data. To verify the advantages of the proposed method, a partial discharge dataset was established through a partial discharge experiment and live detections on substation site. The proposed method was compared with other feature extraction methods and matching methods including statistical features, deep belief networks (DBN), deep convolutional neural networks (CNN), Euclidean distances, and correlation coefficients. The experimental results show that the cosine distance match degree based on the VAE feature vector can effectively detect similar partial discharge data compared with other data matching methods.

Big data storage optimization and parallel processing technology for power equipment surveillance under cloud platform

This paper describes how to strengthen the management capacity of power equipment surveillance system for big data. Here is the ODPS list storage for partial discharge data of transformer proposed against the limitation of self-built Hadoop storage system. According to analysis of results, it is co nfirmed that this method can make a lot of storage space available and improve system performance. In relation to Hadoop distributed file system, HBase, ODPS, it has been proven to be highly reliable.

HANKEL‐EM‐SVD: a hybrid data dropout estimation technique for high voltage partial discharge signals

Partial discharge (PD) measurement and characterisation is the most effective method to assess the insulation conditions of equipment for its diagnosis. A significant challenge in this area is PD denoising. Despite recent advancements in denoising algorithms, data drops out because of inappropriate threshold selection during denoising. Processing PD signals with missing data leads to an inaccurate assessment of the insulation conditions, indicating recovering such missed data is a potential research scope under the moonlight. This study addresses the issues of recovering such data dropouts by considering the PD data vector matrix as a low-rank matrix that needs completion by Hankel-matrix decomposition methods. After finding the probabilistic estimate of the missing values using an expectation maximisation algorithm, singular-value decomposition is used to find the actual missing values by soft-thresholding the singular values of the matrix. After testing this technique on simulated PD data, it is implemented to test SASTRA-High-Voltage Laboratory data, showing root mean square error (RMSE) 0.77% and mean absolute error 0.84%. The efficiency of this technique is confirmed when tested with large-sized noise free PD data of 36 samples from the Laboratory of BOLOGNA University (with RMSE varying from 0.15 to 0.31% and mean absolute error varying from 0.29 to 0.6%).

Diagnosis platform development of partial discharge data based on CMMI model

Capability Maturity Model Integration (CMMI) model is an international standard in software development process and quality management. It integrates the idea of project management body of knowledge and total quality management. Through continuous process improvement for CMMI model, a flexible, regular and progressive capability model framework is established to increase the transparency and controllability of project management and improve the overall effectiveness of software development, which not only ensures software quality, but also solves the bottleneck of traditional project management applied in information technology (IT) industry. This paper presents the analysis and application of CMMI in partial discharge data diagnosis platform, which illustrates its outstanding capability of IT project management. CMMI model make the process manageable and measurable so that the main factors that lead to poor quality and low productivity can be controlled or eliminated.

Random Forest Based Optimal Feature Selection for Partial Discharge Pattern Recognition in HV Cables

Optimal selection of features of Partial Discharge (PD) signals recorded from defects in High Voltage (HV) cables will contribute not only to the improvement of PD pattern recognition accuracy and efficiency but also to PD parameter visualization in HV cable condition monitoring and diagnostics. This paper presents a novel Random Forest (RF)-based feature-selection algorithm for PD pattern recognition of HV cables. The algorithm is applied to feature selection of both PD signals and interference signals with the aim of obtaining the optimal features for data processing. First, the experimental data acquisition and feature extraction processes are introduced. PD signals were captured from faults created in a cable to obtain the raw PD data, and then, a set of 3500 transient PD pulses and a set of 3500 typical interference pulses were extracted, based on which 34 PD features were extracted for further processing. Furthermore, 119 two-dimensional features and 1082 three-dimensional features were generated. The paper then discusses the basic principle of the RF algorithm. Finally, RF-based feature selection was implemented to determine the optimal features for PD pattern recognition. The results were obtained and evaluated with the Back Propagation Neural Network and Support Vector Machine. Results show that the proposed RF-based method is effective for PD feature selection of HV cables with the potential for application to additional HV power apparatus.

A Novel Application of Deep Belief Networks in Learning Partial Discharge Patterns for Classifying Corona, Surface, and Internal Discharges

This paper introduces a new application of deep belief network (DBN) as an emerging artificial neural network for recognizing and classifying different partial discharge (PD) patterns. Phase resolved PD (PRPD) technique with different window intervals is used to manipulate three PD types, including corona, surface, and internal discharges measured in a high-voltage lab. Four approaches are proposed for extracting features from the raw measured data. In the first approach, the DBN is used as both a feature extractor and a PD classifier. The other three approaches extract discriminatory features using statistical and vector-norm-based operators to train a DBN classifier. The impact of the various phase windows through the PRPD method on the performance of the trained classifier is evaluated to obtain the best window interval. It is shown that the deep architectures are capable of learning important distinguishable features from PD data without any data preprocessing. This eliminates time-consuming feature extraction processes that produce the handcrafted features. Based on a comparison analysis, when the input data are corrupted by noise levels or no feature extraction technique is used to preprocess the data, the proposed approach outperforms other techniques, such as artificial neural networks, fuzzy logic classifiers, and support vector machines.

Separating Multi-Source Partial Discharge Signals Using Linear Prediction Analysis and Isolation Forest Algorithm

Partial discharge (PD) detection is an effective way to find defects and diagnose the insulation condition of power equipment. During manufacturing and operating, there may exist multi-source PD signals in the equipment, which could seriously affect the accuracy of subsequent defect analyses. In this paper, a universal separation methodology using linear prediction analysis (LPA) and the isolation forest algorithm (IFA) is introduced. By approximating the present waveform point by a linear combination of several points in the past, a 12-D linear prediction cepstrum coefficient (LPCC) feature space can be established which can accurately characterize PD waveforms. In this paper, we applied principal component analysis (PCA) to reduce the feature space to 2-D space. The IFA was adopted to separate multi-source PD signals which quantified the degree of clustering and added one more parameter based on the original features. Thus, the proposed method can separate multiple PD sources effectively even if the difference in the features of different types of PDs is small. In addition, the proposed algorithm easily removed the noise points during the separation process. The algorithm was applied to different types of PD signals on 35-kV transformers and a three-source PD data set on a 252-kV gas-insulated switchgear (GIS) platform. The separation results confirm that the proposed algorithm can effectively separate and distinguish various PD signals.

A New Denoising Method for UHF PD Signals Using Adaptive VMD and SSA-Based Shrinkage Method.

Noise suppression is one of the key issues for the partial discharge (PD) ultra-high frequency (UHF) method to detect and diagnose the insulation defect of high voltage electrical equipment. However, most existing denoising algorithms are unable to reduce various noises simultaneously. Meanwhile, these methods pay little attention to the feature preservation. To solve this problem, a new denoising method for UHF PD signals is proposed. Firstly, an automatic selection method of mode number for the variational mode decomposition (VMD) is designed to decompose the original signal into a series of band limited intrinsic mode functions (BLIMFs). Then, a kurtosis-based judgement rule is employed to select the effective BLIMFs (eBLIMFs). Next, a singular spectrum analysis (SSA)-based thresholding technique is presented to suppress the residual white noise in each eBLIMF, and the final denoised signal is synthesized by these denoised eBLIMFs. To verify the performance of our method, UHF PD data are collected from the computer simulation, laboratory experiment and a field test, respectively. Particularly, two new evaluation indices are designed for the laboratorial and field data, which consider both the noise suppression and feature preservation. The effectiveness of the proposed approach and its superiority over some traditional methods is demonstrated through these case studies.

Partial Discharges Pattern Recognition of Transformer Defect Model by LBP & HOG Features

Partial discharge (PD) measurement and identification have great importance to condition monitoring of power transformers. In this paper, a new method for recognition of single and multi-source of PD based on extraction of high level image features has been introduced. A database, involving 365 samples of phase-resolved PD (PRPD) data, is developed by measurement carried out on transformer artificial defect models (having different sizes of defect) under a specific applied voltage, to be used for proposed algorithm validation. In the first step, each set of PRPD data is converted into grayscale images to represent different PD defects. Two “image feature extraction” methods, the Local Binary Pattern (LBP), and the Histogram of Oriented Gradient (HOG), are employed to extract features from the obtained gray scale images. Different variants of Support Vector Machine (SVM) are adjusted for optimal classification of PD sources in this process. Impact of the employed parameters in the image processing such as image resolution, random noise, and phase shift, on identification accuracy is investigated and addressed. It is shown that by using HOG-SVM method 99.3% accuracy can be achieved. This is hardly affected by various external factors. Two case studies have been conducted on multi-source PD for evaluating the performance of the proposed algorithm. A void defect is implemented into the transformer model and the resultant recorded signal is used for the study. The DBSCAN algorithm is used as the mean of PD source clustering and sub-PRPD pattern development. It is shown that HOG-SVM method has superior performance in identifying active sources, under sub-PRPD pattern application.

A Convolutional Neural Network-Based Deep Learning Methodology for Recognition of Partial Discharge Patterns from High-Voltage Cables

It is a great challenge to differentiate partial discharge (PD) induced by different types of insulation defects in high-voltage cables. Some types of PD signals have very similar characteristics and are specifically difficult to differentiate, even for the most experienced specialists. To overcome the challenge, a convolutional neural network (CNN)-based deep learning methodology for PD pattern recognition is presented in this paper. First, PD testing for five types of artificial defects in ethylene-propylene-rubber cables is carried out in high voltage laboratory to generate signals containing PD data. Second, 3500 sets of PD transient pulses are extracted, and then 33 kinds of PD features are established. The third stage applies a CNN to the data; typical CNN architecture and the key factors which affect the CNN-based pattern recognition accuracy are described. Factors discussed include the number of the network layers, convolutional kernel size, activation function, and pooling method. This paper presents a flowchart of the CNN-based PD pattern recognition method and an evaluation with 3500 sets of PD samples. Finally, the CNN-based pattern recognition results are shown and the proposed method is compared with two more traditional analysis methods, i.e., support vector machine (SVM) and back propagation neural network (BPNN). The results show that the proposed CNN method has higher pattern recognition accuracy than SVM and BPNN, and that the novel method is especially effective for PD type recognition in cases of signals of high similarity, which is applicable for industrial applications.

Density-based clustering methods for unsupervised separation of partial discharge sources

The recognition of partial discharge (PD) sources is an important task of the monitoring and diagnostics of high-voltage components. Nowadays, digital PD measuring systems have the capability of extracting features and form scatter plots with such data sets. Part of an unsupervised PD analysis system is to discover clusters within the data sets and link them to particular PD sources. Due to the nature of PD data sets, clusters may appear very close to each other or even merged hindering the separation of sources. Clustering methods based on spatial density such as the density peak clustering (DPC) method and DBSCAN are suitable approaches to discover clusters within PD data sets. However, their accuracy can be reduced due to the proximity among clusters. In this paper, a new method is presented to improve the accuracy of the DPC method. Our method proposes to partition the data set and later pass the resulting subsets to the DPC method. The partitioning is based on the spatial density of data computed by a smoothed density method (SD). SD has the advantage of being fast and not requiring high computational power. As a final step, a routine is applied to group the sub clusters as per the DPC method having a threshold for the data contour distance as a criterion. This method proved higher accuracy to discover clusters in actual PD data sets. However, the threshold for the data contour distance still needs further research.

Online condition monitoring of MV cable feeders using Rogowski coil sensors for PD measurements

Condition monitoring is a highly effective prognostic tool for incipient insulation degradation to avoid sudden failures of electrical components and to keep the power network in operation. Improved operational performance of the sensors and effective measurement techniques could enable the development of a robust monitoring system. This paper addresses two main aspects of condition monitoring: an enhanced design of an induction sensor that has the capability of measuring partial discharge (PD) signals emerging simultaneously from medium voltage cables and transformers, and an integrated monitoring system that enables the monitoring of a wider part of the cable feeder. Having described the conventional practices along with the authors’ own experiences and research on non-intrusive solutions, this paper proposes an optimum design of a Rogowski coil that can measure the PD signals from medium voltage cables, its accessories, and the distribution transformers. The proposed PD monitoring scheme is implemented using the directional sensitivity capability of Rogowski coils and a suitable sensor installation scheme that leads to the development of an integrated monitoring model for the components of a MV cable feeder. Furthermore, the paper presents forethought regarding huge amount of PD data from various sensors using a simplified and practical approach. In the perspective of today’s changing grid, the presented idea of integrated monitoring practices provide a concept towards automated condition monitoring.

Application of improved Hilbert-Huang transform to partial discharge signal analysis

As a key concern in a power system, a deteriorated insulation will cause a partial discharge phenomenon and hence degrades the power supply quality. Thus, a partial discharge test has been turned into an approach of significance to protect a power system from an unexpected fault. As the first step in this work, a defect cast resin transformer is treated as a test object, and the detected partial discharge data are then transformed into a time-frequency-energy distribution through the Hilbert-Huang-Transform (HTT). The distribution is capable of providing both time-domain and frequency-domain information. It is a highly promising approach to pattern identification of a partial discharge and fault diagnosis. There is an excellent adaptability when applied to a nonlinear as well as nonstationary signal analysis for HHT, but with two major concerns, that is, the determination of optimal shifting number and the identification of unintended illusive components. For this sake, a combination of Kolmogorov-Smirnov (K-S test) and a sorting by signal energy ratio is proposed in the determination of the optimal shifting number. Subsequently, illusive components are ridded through a cumulative K-S test, such that an intrinsic mode function (IMF) is precisely extracted, intrinsic physical meaning contained over each spectral band is well preserved, and a mode confusion problem is removed as well.

GIS partial discharge pattern recognition via deep convolutional neural network under complex data source

This paper aims to improve recognition accuracies of partial discharge (PD) of complex data sources by employing deep convolutional neural network (DCNN). First, a dataset with complex data sources is established, which contains PD experiments data, substation live detection data and inference data. During the PD experiments, data are acquired from five types of artificial defect models on a real gas insulated switchgear (GIS) platform, using different PD detection instruments. Substation live detection data are collected from the running GIS in more than 30 substations, using two types of portable PD detection devices. Typical inference data in PD detection are also employed for algorithm validation. Second, a DCNN based PD pattern recognition method is presented. In the proposed method, all of the PD data are normalized into a uniform format of phase resolved pulse sequence (PRPS). A DCNN model is employed to automatically extract the features of a complex data set. The results are obtained by a Softmax classifier. Third, the DCNN based PD pattern recognition method is applied to the dataset with complex data sources and achieves 89.8% accuracy. The back-propagation neural network (BPNN) and support vector machine (SVM) methods with traditional statistical features are compared with the developed method. The result shows that accuracy is improved by the method proposed in this paper. With the enlargement of the data set and a more complex data sample, the improved value will further increase, thus the proposed method is more suitable for the engineering application of a big data platform.

Combined diagnosis of partial discharge based on the multi-dimensional characteristic parameters

This paper presents a comprehensive multi-parameter diagnosis method based on multiple partial discharge signals include high-frequency current, ultrasound, ultrahigh frequency (UHF) etc. First, acquire the high-frequency current, ultrasound, UHF partial discharge data under various types of defects, and extract the characteristic values, including nine basic characteristic parameters, eight phase characteristic parameters and the like. Diagnose signals respectively, with the method based on information fusion and semi-supervised learning for high-frequency current PD data, the method based on adaptive mutation parameters of particle entropy for ultrasonic signals, the method based on IIA-ART2A neural network for UHF signals. Then integrate the diagnostic results, which is the probability of fault of various defects and matrix, of different PD diagnosis signals, and analysis with the multiple classifier based on multi-parameter fuzzy integral to get the final diagnosis.

UHF PD measurements on power transformers-advantages and limitations

Recently, Ultra-High Frequency (UHF) Partial Discharge (PD) measurements on power transformers has attracted a lot of attention. Despite numerous publications, there are still missing information about considerations and limitations of this method in regards to antenna characteristics, relation between IEC 60270 and UHF detection methods and robustness against external noises. In this paper, each of the mentioned sections is investigated in detail by simulation analysis or laboratory measurements. Applying the results obtained in each section leads to a more precise measurement, more accurate analysis of UHF PD data and solving some of the problems associated with it.

Identification of single and multiple partial discharge sources by optical method using mathematical morphology aided sparse representation classifier

Partial Discharge (PD) is one of the most critical electrical phenomena affecting the life of electrical equipment. Repetitive PD leading to arcing faults is a recurrent problem in many air insulated systems, such as high voltage air-insulated switchgears. Hence, detection and localization of PD inside such electrical equipment is necessary for early prevention of impending failure. Keeping this in mind, this paper proposes a methodology to localize single and multiple PD sources employing two recent developments in signal processing and machine learning techniques. Optical sensors have been used to record the PD data inside a cubical steel box. A cylindrical barrier has been inserted inside this box to emulate geometrical structures inside Switchgears. Mathematical Morphology aided feature extraction has been employed to extract important features from the PD signal. Sparse Representation classification has been employed to classify the extracted features and to identify the type and location of PD source. The results show that this methodology gives very high classification accuracy corresponding to both single and multiple PD sources.

Automated Partial Discharge Data Evaluation During Monitoring of Stator Winding

Automated Partial Discharge Data Evaluation During Monitoring of Stator Winding

Assessment of PD severity in gas‐insulated switchgear with an SSAE

Scientific partial discharge (PD) severity evaluation is highly important to the safe operation of gas-insulated switchgear. However, describing PD severity with only a few statistical features such as discharge time and discharge amplitude is unreliable. Hence, a deep-learning neural network model called stacked sparse auto-encoder (SSAE) is proposed to realise feature extraction from the middle layer with a small number of nodes. The output feature that is almost similar to the input PD information is produced in the model. The features extracted from PD data are then fed into a soft-max classifier to be classified into one of four defined PD severity states. In addition, unsupervised greedy layer-wise pre-training and supervised fine-tuning are utilised to train the SSAE network during evaluation. Results of testing and simulation analysis show that the features extracted by the SSAE model effectively characterise PD severity. The performance of the SSAE model, which possesses an average assessment accuracy of up to 92.2%, is better than that of the support vector machine algorithm based on statistical features. According to the tested number of SSAE layers and features and the training sample size, the SSAE model possesses good expansibility and can be useful in practical applications.