Partial Discharge Recognition Research Articles

Partial Discharge Detection and Recognition in Random Matrix Theory Paradigm

The detection and recognition of partial discharge (PD) is an important topic in insulation tests and diagnoses. Take advantage of the affluent results from random matrix theory (RMT), such as eigenvalue analysis, M-P law, the ring law, and so on, a novel methodology in RMT paradigm is proposed for fast PD pulse detection in this paper. Furthermore, a scheme of time series modeling as random matrix is also proposed to extend RMT for applications with non-Gaussian noise context. Based on that, the eigenvalue distribution property is used for PD pattern recognition, which is completely new compared with traditional phase resolved PD and time-resolved PD methods. The simulation and experimental results show that the proposed methods are efficient, reliable, and feasible for PD detection and recognition especially for online applications.

Rough set theory applied to pattern recognition of Partial Discharge in noise affected cable data

This paper presents an effective, Rough Set (RS) based, pattern recognition method for rejecting interference signals and recognising Partial Discharge (PD) signals from different sources. Firstly, RS theory is presented in terms of Information System, Lower and Upper Approximation, Signal Discretisation, Attribute Reduction and a flowchart of the RS based pattern recognition method. Secondly, PD testing of five types of artificial defect in ethylene-propylene rubber (EPR) cable is carried out and data pre-processing and feature extraction are employed to separate PD and interference signals. Thirdly, the RS based PD signal recognition method is applied to 4000 samples and is proven to have 99% accuracy. Fourthly, the RS based PD recognition method is applied to signals from five different sources and an accuracy of more than 93% is attained when a combination of signal discretisation and attribute reduction methods are applied. Finally, Back-propagation Neural Network (BPNN) and Support Vector Machine (SVM) methods are studied and compared with the developed method. The proposed RS method is proven to have higher accuracy than SVM and BPNN and can be applied for on-line PD monitoring of cable systems after training with valid sample data.

Artificial Neural Network Application for Partial Discharge Recognition: Survey and Future Directions

In order to investigate how artificial neural networks (ANNs) have been applied for partial discharge (PD) pattern recognition, this paper reviews recent progress made on ANN development for PD classification by a literature survey. Contributions from several authors have been presented and discussed. High recognition rate has been recorded for several PD faults, but there are still many factors that hinder correct recognition of PD by the ANN, such as high-amplitude noise or wide spectral content typical from industrial environments, trial and error approaches in determining an optimum ANN, multiple PD sources acting simultaneously, lack of comprehensive and up to date databank of PD faults, and the appropriate selection of the characteristics that allow a correct recognition of the type of source which are currently being addressed by researchers. Several suggestions for improvement are proposed by the authors include: (1) determining the optimum weights in training the ANN; (2) using PD data captured over long stressing period in training the ANN; (3) ANN recognizing different PD degradation levels; (4) using the same resolution sizes of the PD patterns when training and testing the ANN with different PD dataset; (5) understanding the characteristics of multiple concurrent PD faults and effectively recognizing them; and (6) developing techniques in order to shorten the training time for the ANN as applied for PD recognition Finally, this paper critically assesses the suitability of ANNs for both online and offline PD detections outlining the advantages to the practitioners in the field. It is possible for the ANNs to determine the stage of degradation of the PD, thereby giving an indication of the seriousness of the fault.

Improving recognition accuracy of partial discharge patterns by image-oriented feature extraction and selection technique

This paper aims to improve recognition accuracies of single sources of partial discharge (PD) by employing image-oriented feature extraction and selection algorithms. 419 diversified samples of PD data acquired from typically artificial defect models, where the defect size, applied voltage and insulation aging are taken into account, are employed for algorithms validation. In the proposed method, the PD gray images are decomposed into various vectors by two-dimensional principal component analysis (2DPCA) on horizontal and vertical directions, respectively, in which 9 representative parameters are extracted from each image vector. Based on the 419 PD samples, the proposed image features are confirmed to achieve better PD recognition performances than typical image compression methods relevant to 2DPCA. To further improve the recognition performance and reduce the feature dimension, the feature selection technique based on non-dominated sorting genetic algorithm II (NSGA-II) is adopted to search the optimal 2DPCA features. By using the proposed image-oriented feature extraction and selection algorithm, fuzzy k-nearest neighbor classifier (FkNNC), back-propagation neural network (BPNN) and support vector machine (SVM) with 16-dimensional 2DPCA features can now achieve 96.4%, 94.4% and 97.5% accuracies, respectively. The increment of average recognition accuracies of 5%–7% are obtained by different classifiers compared with phase-resolved partial discharge (PRPD) statistical features in previous studies. The investigations on the recognition results relevant to image distortion, cavity discharge change with time and PD contaminated by random noises with different level confirm the robustness of the 16-dimensional 2DPCA features, which signify a great potential in applying the proposed method to field PD pattern recognition in the future.

Partial Discharge Recognition Reliability Considering the Influence of Multi-factors Based on the Two-directional Fuzzy-weighted Two-dimensional Principal Component Analysis Algorithm

—In the current work, a new image-oriented feature extraction algorithm is proposed to improve partial discharge recognition accuracy when the multi-factor influences of insulation aging, defect size, and applied voltage are taken into consideration. A fuzzy-weighted method is designed to modify two-dimensional principal component analysis, producing the fuzzy-weighted two-dimensional principal component analysis algorithm, which incorporates samples distribution to the extracted partial discharge features. By synchronously implementing horizontal and vertical fuzzy-weighted two-dimensional principal component analysis, the proposed two-directional fuzzy-weighted two-dimensional principal component analysis algorithm is developed to extract the partial discharge image features. For algorithm testing, 419 diversified partial discharge samples acquired from typically artificial defect models are employed, in which the multi-factor influences of insulation aging, defect size, and applied voltage are taken into account. It is shown that the optimally successful clustering rate of 91.41% is obtained by fuzzy C-means clustering with the variation of three algorithm parameters. The comparisons with phase-resolved partial discharge statistical features and other popular image compression methods based on the support vector machine also confirms the improvement of partial discharge recognition accuracy using the proposed two-directional fuzzy-weighted two-dimensional principal component analysis algorithm.

Diagnosis of High Voltage Insulators Made of Ceramic Using Spectrophotometry

The paper presents results of comparative analysis of optical signals emitted by partial discharges occurring on three types of high voltage insulators made of porcelain. The research work consisted of diagnosis of the following devices: a long rod insulator, a cap insulator, and an insulating cylinder. For optical signal registration a spectrophotometer was applied. All measurements were performed under laboratory conditions by changing the value of partial discharges generation voltage. For the cylindrical insulator also the distance between high voltage and ground electrodes was subjected for investigation as a factor having influence on partial discharges. The main contribution which resulted from the studies is statement that application of spectrophotometer enables faster recognition of partial discharges, as compared to standard methods.

Automatic selection of frequency bands for the power ratios separation technique in partial discharge measurements: part II, PD source recognition and applications

Partial discharge (PD) measurement is a widely extended technique for the diagnosis of electrical machines and power cables. Since PD and noise recognition is really important in industrial environments, the authors proposed the use of the spectral Power Ratios (PR) of the detected high-frequency (HF) signals, which show promising results. However, the frequency bands selected for their calculation have a clear influence on the position of the clusters in the classification maps. Following this research trend, a new algorithm for the automatic selection of the frequency bands which gives a proper separation between the clusters, has been proposed in part I of this paper. The first results presented in this paper revealed a good behavior of the technique for PD and noise separation in simple test objects. In this second part of the paper, more results from the application of this system to PD source recognition in simple test objects and also in real equipment are developed.

Partial discharge recognition in gas insulated switchgear based on multi-information fusion

Partial discharge (PD) recognition is an important tool for online monitoring of gas insulated switchgear (GIS) and diagnosing existing defects. At present, there are two different types of data patterns used for analysis and evaluation of PD signals: phase resolved partial discharge (PRPD) mode and time resolved partial discharge (TRPD) mode. Using different types of data patterns separately can lead to inconsistent or even conflicted recognition results, but the two types of data patterns having complementary information between each other can be used together for data fusion. Dempster-Shafer (DS) evidence theory is introduced to address the problem of evidence conflict and low fusion efficiency. First, two sub-networks for PD recognition are established and compared employing the back propagation neural network (BPNN) learning algorithm under the two modes respectively. Secondly, to minimize (with possible elimination of) the possibility of misclassification, a new fusion decision-making system for PD recognition is proposed on the basis of fusing the results from the sub-networks. Finally, extensive field experiments on two artificial defects are conducted in order to evaluate the performance of the proposed method in this paper. The classification results reveal that the proposed method significantly outperforms the ways of using only single type of data patterns. The method proposed in this paper reduces the uncertainty and effectively improves the credibility of diagnosis.

A new image-oriented feature extraction method for partial discharges

Partial discharge (PD) measurement and recognition is of great importance to assess the health condition of power transformers. However, the variation of defect size, applied voltage as well as insulation aging gives rise to the dispersion and crossover of PD features, which would influence PD recognition reliability of transformers. An image-oriented two-directional modified fuzzy-weighted two-dimensional linear discriminant analysis (TD-MFW-2DLDA) for feature extraction of PD gray images, aiming at solving the above problem, is proposed in this paper. Two classification models including fuzzy C-means (FCM) clustering and support vector machine with genetic algorithm (GA-SVM) are designed for features evaluation and PD pattern recognition. 419 PD samples of four typically artificial defect models are measured in laboratory, in which the multi-factors of defect size, applied voltage and insulation aging are taken into account, and further adopted for algorithms testing. It is shown that TD-MFW-2DLDA achieves optimal successful FCM clustering rate of about 93% and GA-SVM recognition accuracy of about 96%, which are better than that of typical PD features influenced by the multi-factors. Additionally, FCM clustering validity measures provide better compactness within class and separability between classes of TD-MFW-2DLDA which is suitable for on-site PD diagnostic applications.

Partial Discharge Pattern Recognition for Cast Resin Current Transformer Using Fuzzy Neural Network

This paper proposes a pattern recognition approach based on the fuzzy neural network for identifying insulation defects of cast resin current transformer (CRCT) arising from partial discharge (PD). Pattern recognition of PD is used for identifying defects causing the PD, such as internal discharge, external discharge, corona, etc. This information is vital for estimating the harmfulness of the discharge in the insulation. The PD patterns are collected by a PD detecting system in the laboratory. Several statistical methods are used on the phase related distributions in this paper to extract the features for recognition. A set of features, used as operators, for each PD pattern is extracted through statistical methods. To verify the proposed approach, experiments were conducted to demonstrate the field-test PD pattern recognition of CRCT models with artificial defects are purposely created to produce the common PD activities of insulators by using feature vectors of field-test PD patterns. The experimental data are found to be in close agreement with the recognized data. The test results show that the proposed approach may achieve quite satisfactory recognition of PD patterns.

Application of Back Propagation Neural Network for Partial Discharge Pattern Recognition

This paper proposes a pattern recognition approach based on the back propagation (BP) neural network for identifying insulation defects of high-voltage electrical apparatus arising from partial discharge (PD). Pattern recognition of PD is used for identifying defects causing the PD, such as internal discharge, external discharge, corona, etc. This information is vital for estimating the harmfulness of the discharge in the insulation. Since an insulation defect, such as one resulting from PD, would have a corresponding particular pattern, pattern recognition of PD is significant means to discriminate insulation conditions of high-voltage electrical apparatus. To verify the proposed approach, experiments were conducted to demonstrate the field-test PD pattern recognition of model insulators with artificial defects are purposely created to produce the common PD activities of insulators by using feature vectors of field-test PD patterns. The experimental data are found to be in close agreement with the recognized data. The experimental results show that the proposed approach is very effective for recognizing the defects of high-voltage electrical apparatus.

Inductive Sensor Performance in Partial Discharges and Noise Separation by Means of Spectral Power Ratios

Partial discharge (PD) detection is a standardized technique to qualify electrical insulation in machines and power cables. Several techniques that analyze the waveform of the pulses have been proposed to discriminate noise from PD activity. Among them, spectral power ratio representation shows great flexibility in the separation of the sources of PD. Mapping spectral power ratios in two-dimensional plots leads to clusters of points which group pulses with similar characteristics. The position in the map depends on the nature of the partial discharge, the setup and the frequency response of the sensors. If these clusters are clearly separated, the subsequent task of identifying the source of the discharge is straightforward so the distance between clusters can be a figure of merit to suggest the best option for PD recognition. In this paper, two inductive sensors with different frequency responses to pulsed signals, a high frequency current transformer and an inductive loop sensor, are analyzed to test their performance in detecting and separating the sources of partial discharges.

Partial Discharge Pattern Recognition of Cast Resin Current Transformers Using Radial Basis Function Neural Network

This paper proposes a novel pattern recognition approach based on the radial basis function (RBF) neural network for identifying insulation defects of high-voltage electrical apparatus arising from partial discharge (PD). Pattern recognition of PD is used for identifying defects causing the PD, such as internal discharge, external discharge, corona, etc. This information is vital for estimating the harmfulness of the discharge in the insulation. Since an insulation defect, such as one resulting from PD, would have a corresponding particular pattern, pattern recognition of PD is significant means to discriminate insulation conditions of high-voltage electrical apparatus. To verify the proposed approach, experiments were conducted to demonstrate the field-test PD pattern recognition of cast resin current transformer (CRCT) models. These tests used artificial defects created in order to produce the common PD activities of CRCTs by using feature vectors of field-test PD patterns. The significant features are extracted by using nonlinear principal component analysis (NLPCA) method. The experimental data are found to be in close agreement with the recognized data. The test results show that the proposed approach is efficient and reliable.

Non linear support vector machine based partial discharge patterns recognition using fractal features

Partial discharge (PD) is an important tool for assessing the quality of the insulation system in High Voltage (HV) power apparatus. In this work, four different PD sources namely corona and surface discharges in both air and oil are measured in the HV laboratory. Initially 3-D (Φ-q-n) PD patterns are extracted from the PD data. Then it is subjected to two different fractal image compression techniques namely box counting method and semi variance method. For box counting method, the fractal dimensions like fractal dimension average, standard deviation and lacunarity are evaluated. For semi variance method, horizontal and vertical fractal dimension averages are evaluated. The extracted fractal features from 3-D PD patterns are used as input parameters for non linear Support Vector Machine (SVM) for PD recognition. The performance of non linear SVM is compared with Artificial Neural Network (ANN) and linear SVM classifiers. The non linear SVM with semi variance method provides outer performance as compared with other methods due to its gain flexibility and good out-of-sample generalization.

Propagation Characteristics of Partial Discharge Pulses in the Cable

In order to determine the type and location of partial discharge in cable, the effect of partial discharge (PD) pulse propagation in the cable is studied. Firstly, pulses are injected to cables of different lengths so that input and output signal can be measured at both ends of each cable. Then the transfer function of pulse propagation path can be defined. Secondly, high-voltage test is done in the cable joint with man-made defects, and typical PD waveforms are gotten. Seven parameters of waveform characteristics are calculated, including edge times, waveform shape and statistical characteristics. They are used to distinguish different types of PD or distances of the pulse propagation. Thus the efficiency of PD recognition in cable can be improved.

The Software Design of Intelligent Diagnosis System for Partial Discharge Fault of GIS

The correct identification of GIS partial discharge type is essential for assessing the insulation status of GIS and developing a reasonable maintenance strategy. After the analysis of the intelligent diagnosis system design, the software architecture of host and remote computer is given by this article. Site host computer completes the data preprocessing and the remote host computer completes the function of real-time observation, data storage and inquiry, spectrum analysis and pattern recognition. The test shows that the software can complete the intelligent recognition of partial discharge.

Study of the Ultrasonic Characteristics of Typical Partial Discharge on GIS

This Paper studied the typical PD in GIS by ultrasonic method. A GIS PD simulating and testing system is established, including a PD simulating system, an ultrasonic detecting system and a conventional pulse current method measurement system. The thesis focuses on four typical kinds of PD signals in GIS, including metal particles, bad contacts on high and grounding potential conductor, the protrusion on the high potential conductor. The statistics characteristics are obtained which include N-Φ spectrogram (relationship between discharge frequency and discharge phase Spectrogram) and Umax-Φ spectrogram (relationship between discharge amplitude and discharge phase Spectrogram). Experimental results show that low-frequency ultrasonic sensors can detect this four PD signals， signal spectra analysis shows that ultrasonic signals of different statistical model are with different characteristics. The results provide test data for the pattern recognition of Partial Discharge in GIS.

Gas-insulated switchgear PD signal analysis based on Hilbert-Huang transform with fractal parameters enhancement

This study proposes a novel method of partial discharge (PD) electrical signal analysis based on the Hilbert-Huang transform (HHT) with fractal feature enhancement. Firstly, this study establishes four defect types of 15 kV gas-insulated switchgear (GIS) and uses a commercial high-frequency current transformer (HFCT) to measure the electrical signals caused by the PD phenomenon. Second, the authors applied HHT for the PD electrical signal process. The HHT can represent instantaneous frequency components through empirical mode decomposition (EMD) and then transform to a 3D Hilbert energy spectrum. Finally, this study extracts the fractal parameters from the 3D energy spectrum and uses a neural network (NN) for PD recognition. To demonstrate the effectiveness of the proposed method, this study uses 160 sets of field-tested PD patterns generated by GIS, and then compares the recognition rate of the signal with and without the EMD process. The result shows that the proposed method can easily separate various defect types. The method can also be employed by the construction unit to verify the GIS quality and determine the GIS insulation status.

Pattern Recognition for Partial Discharge in Power Transformer Based on Probabilistic Neural Networks

Partial discharge diagnosis plays an important role in condition monitoring of power transformer. Different discharge types cause varying degrees of effects on insulation degradation. Therefore pattern recognition for partial discharge is necessary for power transformer fault diagnosis. Probabilistic Neural Networks (PNN) is based on the theory of Bayesian minimum risk which perfectly fits for classification. In this paper PNN is employed for pattern recognition of partial discharge. 18 characteristic parameters are extracted from UHF partial discharge signals as the inputs of the PNN classifier. Results show that the PNN method gives fast and accurate classification performance.

Shielding effect of power transformers tanks in the ultra-high-frequency detection of partial discharges

Partial discharges (PD) measurement is a well-known technique for judgment of the condition of power transformer offline. Phase-resolved PD measurement is a normalized technique for PD tests, but noise, especially by PD-measurements onsite is sometimes difficult to reject, and a post-processing of a pattern is necessary to separate different sources. Higher-frequency-detection devices and processing strategies are useful for PD classification and identification. Recently, ultra-high-frequency (UHF) detection by means of antennas has been shown to be a promising detection system for both offline and online PD-measurements. However, it is necessary to assess whether PD source separation and classification is possible by means of pulse analysis because some additional UHF noise sources are coupled to the detected signals, and it is not clear whether different sources produce different pulses. In this study, the attenuation effect of the metallic tank of a power transformer on the inner PD activity when measured outside the tank was studied. Additionally, experiments were conducted to detect and characterize two different PD sources (internal and external discharges) using two antennas measuring the same phenomenon inside and outside of a transformer. It will also be shown that broadband UHF signals are useful for PD recognition and that a deep study of frequencies below 500 MHz can separate PD occurring inside from those occurring outside when measured with an antenna outside the tank.