Defects In Gas Insulated Switchgear Research Articles

Semi-supervised learning for gas insulated switchgear partial discharge pattern recognition in the case of limited labeled data

Semi-supervised learning has better and more efficient performance than supervised algorithms in the case of limited labeled data. Existing methods for diagnosing partial discharge (PD) insulation defects in gas-insulated switchgear (GIS) equipment can only be effective if there is sufficient labeled data. However, in the actual working conditions of GIS equipment, insulation defect data is very scarce, where labeled data is more expensive to obtain and most of the data is unlabeled. In the case of limited PD labeled data, it is still a serious challenge to achieve higher classification accuracy of GIS PD pattern recognition. Therefore, we propose a semi-supervised self-training algorithm based on density peaks of local neighbor Information. Firstly, an improved density peak clustering algorithm based on local neighbor information is proposed, which no longer depends on the truncation distance, and considers the local information to better reflect the local density. Secondly, using local neighbor Information of labeled data, the criterion of confidence of unlabeled data is improved. Then, the PD unlabeled data with pseudo-labels are used to build a strong classifier for GIS PD pattern recognition. The experimental results show that the proposed algorithm has higher classification accuracy than other semi-supervised algorithms. When the proportion of labeled data is 10 %, the recognition accuracy can reach 65.98 %, which is the highest in the comparison algorithm. The proposed algorithm provides a feasible solution for GIS PD pattern recognition in the case of limited labeled data.

Meta-autoencoder-based zero-shot learning for insulation defect diagnosis in gas-insulated switchgear

Data-driven methods are the primary methods of training models for the diagnosis of insulation defects in gas-insulated switchgear (GIS). Due to complicated operating environments, target samples are not available for training sometimes, leading to insufficient feature learning. Therefore, a meta-autoencoder-based zero-shot learning (MAZL) method is proposed for the diagnosis of GIS insulation defects. First, the visual features of insulation defects’ signals are extracted by a convolutional neural network. Next, the mapping between visual and semantic spaces is learned by semantic autoencoder (SAE). For SAE, the triplet loss is adapted to align semantic information, and unknown class semantic constraints are used to alleviate the domain offset. Finally, a nearest-neighbor algorithm is used to achieve zero-sample diagnosis. In model training, episode training is introduced to make the MAZL model learn a more discriminative semantic prototype, which alleviates classification deviations. The MAZL achieved a diagnostic accuracy for single and multi-source GIS insulation defects of 96.215% and 90.41% without using test classes for training, respectively. This finding provides ideas for the diagnosis of new GIS insulation defects.

SF6 decomposition behaviour under partial discharge of typical insulation defects in gas‐solid insulated valve‐side bushing of converter transformer

AbstractIn this paper we investigated SF6 decomposition behaviour of typical insulation defects in gas‐solid insulated converter transformer valve‐side bushing under partial discharge (PD). Additionally, we compared the results with that of gas insulated switchgear (GIS). The results indicated that SF6 decomposition behaviour varies with the insulation defects type. For metal protrusion defects, the proportion ranking of CO2 and SO2 can be related to whether the needle tip is wrapped with epoxy resin‐impregnated paper (ERIP). The proportion of carbon‐containing products is much higher under contamination defect than under other defects. Moreover, the proportion of carbon‐containing components is higher under flashover condition than under PD condition and is accompanied by the appearance of H2S and CS2. However, the yield of SO2F2 is always low. Compared to negative DC voltages, positive DC voltages have a higher proportion of carbon‐containing components, mainly in the form of more CO2 and less SOF2. Compared to the PD decomposition behaviour of typical insulation defects in GIS, when PD involves ERIP, the proportion of carbon‐containing products, especially CO2, is significant higher, while the proportion of sulphur containing products, especially SO2F2, is lower.

Study on Intermittent Discharge Characteristics of Typical Solid Insulation Defects in Gas Insulated Switchgear

Study on Intermittent Discharge Characteristics of Typical Solid Insulation Defects in Gas Insulated Switchgear

MCTSR: A super‐resolution method for defects in gas‐insulated switchgear X‐ray images based on multi‐scale contextual transformers

AbstractAs the core equipment of transmission and distribution hubs, the operational status of gas‐insulated switchgear (GIS) is closely linked to the safety of the power system. Recently, X‐ray digital imaging technology has been extensively used in GIS equipment fault detection. However, the X‐ray image of GIS is blurred, which is not conducive to the detection of tiny defects. Thus, a super‐resolution method for GIS X‐ray images based on multi‐scale context transformers is proposed in this study, namely MCTSR. Firstly, a second‐order image degradation model is introduced to generate GIS X‐ray low‐resolution images that more closely resemble the real world. Secondly, a contextual transformer gate module is constructed to improve attention to tiny defects in GIS X‐ray images. Thirdly, a U‐Net discriminator network based on multi‐scale contextual transformers is intended to enrich the information of the generated images. Finally, the proposed discriminator is combined with the existing generator to compose a super‐resolution method applicable to GIS X‐ray images. The experimental results demonstrate that the method outperforms other methods in peak signal‐to‐noise ratio and structural similarity on the constructed GIS X‐ray image dataset. In addition, the output image of the proposed method facilitates the subsequent defect detection.

Adsorption properties of Ni cluster modified TiO 2 (1 0 1) towards SF 6 decomposition gases

Online detection of insulation defects the decomposition products of SF6 plays a key role in ensuring the safe operation of the gas-insulated switchgear (GIS) equipment. The pristine TiO2 (1 0 1) modified by Ni cluster (1–3 Ni atoms) is used as a new gas-sensing material to study its gas-sensing characteristics to the characteristic decomposition products of SF6: SO2, SOF2, and SO2F2. Through density functional theory (DFT) calculation, we found that the modification of Ni clusters significantly improved the conductivity of pristine TiO2 (1 0 1) and the gas sensitivity of SF6 decomposition products. More importantly, the change of conductivity after gas adsorption lays a theoretical foundation for identifying the type and concentration of SF6 characteristic decomposition products and further judging the type of defects in GIS.

A novel partial discharge pattern recognition for GIS with unbalanced sample based on conditional variational autoencoder

Deep learning has been widely used in gas insulated switchgear (GIS) partial discharge (PD) pattern recognition with its powerful ability to automatically extract features. As a typical data-driven model, the diagnostic performance of deep learning methods will decrease with the reduction of the training sample size, resulting in lower recognition rates for minority class samples. In actual situation, the probability of failure of various insulation defects in GIS is not the same, so that the samples of various types of defects obtained in practical applications cannot maintain a balanced distribution. To solve the problem of class imbalance among training samples, this paper proposes an improved variational autoencoder (VAE) to augment the minority class fault samples. The PD signal is first converted into a time-spectrogram with high time-frequency resolution by S transform. Then use conditional AE to perform directed augmentation on unbalanced class samples. Finally, the obtained balanced data set is input into convolutional neural network to complete the fault diagnosis of PD. Experiments show that the proposed method achieves excellent recognition results under various unbalanced conditions, indicating that it has strong tolerance and high generalization ability for unbalanced samples.

Exploring the Impact of Parallel Architecture on Improving Adaptable Neuro-Fuzzy Inference Systems for Gas-Insulated Switch Defect Recognition

Gas-insulated switchgear malfunctions during power system operation may occur due to electrical, thermal, or human errors in the manufacturing process. The leading causes of insulation deterioration of gas-insulated switchgear are discharging along the surface caused by dirt on the insulating material, internal discharge caused by impurities and cavities in the insulating material, corona discharge caused by poor assembly or construction at the site, and electric tree channel discharge caused by the intense internal discharge. Since different defects produce different partial discharge characteristics, the operating power equipment can be analyzed using measurement instruments to detect partial discharge for preventive equipment fault diagnosis, avoiding unnecessary power outages and losses; therefore, evaluating the defects in gas-insulated switchgear is essential. In this study, three gas-insulated switchgears were prefabricated with different defects before encapsulation, and the partial discharge data of each defect were measured by applying different test voltages. The adaptive neuro-fuzzy inference system (ANFIS) input data were used to evaluate the recognition effect, showing that the average recognition rate of the core for all defects was over 90%. The proposed system architecture can continuously accumulate the defect measurement database of gas-insulated switchgear and be used as a reference for constructing electrical equipment defect recognition systems.

A Novel Federated Transfer Learning Framework for Intelligent Diagnosis of Insulation Defects in Gas-Insulated Switchgear

Deep learning driven diagnostic methods have shown significant ability to diagnose certain types of insulation defects. However, for these methods to achieve excellent results, it is necessary for the clients to obtain a large amount of data, which is not realistic for a typical client. For this reason, there is an incentive for different clients to collaboratively develop effective diagnostic models, but due to conflicts of interest and privacy protection issues, data sharing has become a challenge. In addition, due to domain shift, it is difficult to apply the model developed by some clients to others. In response to this, we propose a novel federated transfer learning framework. Specifically, we developed a federated adversarial learning that achieves domain adaptation while protecting data privacy, and we introduced a federated minimax algorithm for global model aggregation, which not only solves the problem of gradient drift caused by sample imbalance but also improves the accuracy of the global model. We verified that our method can achieve high-precision and robust diagnosis of gas-insulated switchgear insulation defects with experiments composed of laboratory and field clients. The superior performance for unbalanced small samples on-site shows that the application of the proposed federated transfer learning is promising.

Experimental research on the influence of temperature on the discharge signal of void defects in GIS

Temperature is an important environmental factor during the operation of gas-insulated switchgear (GIS), affecting the evaluation results of the GIS equipment to increase the risk of the power system. However, the influence of temperature on the partial discharge detection signal of GIS is still unclear. Aimed at the common void defects in GIS, the law of change on the number of ultrahigh-frequency (UHF) pulses, the UHF amplitude, the characteristic value of the UHF map, and the maximum apparent charge of a single pulse with temperature are obtained using the UHF method and IEC60270, and corresponding theoretical analysis is carried out. The results show that an increase in temperature leads to a decrease in the void discharge delay time, causing an increase in UHF pulses and a decrease in the apparent charge of a single discharge pulse in the experiment. The increase of temperature makes the void discharge current rise quickly so that the induced UHF amplitude increases. In the range of 40–70°C, the maximum pulse amplitude increases by approximately 30% for every 10°C increase, and the average pulse amplitude increases by approximately 12%. The result of UHF signals affected by temperature obtained in this study has research significance for the realisation of a comprehensive evaluation of the insulation state of GIS equipment considering temperature.

GIS Partial Discharge Pattern Recognition Based on a Novel Convolutional Neural Networks and Long Short-Term Memory.

Distinguishing the types of partial discharge (PD) caused by different insulation defects in gas-insulated switchgear (GIS) is a great challenge in the power industry, and improving the recognition accuracy of the relevant models is one of the key problems. In this paper, a convolutional neural network and long short-term memory (CNN-LSTM) model is proposed, which can effectively extract and utilize the spatiotemporal characteristics of PD input signals. First, the spatial characteristics of higher-level PD signals can be obtained through the CNN network, but because CNN is a deep feedforward neural network, it does not have the ability to process time-series data. The PD voltage signal is related to the time dimension, so LSTM saves and analyzes the previous voltage signal information, realizes the modeling of the time dependence of the data, and improves the accuracy of the PD signal pattern recognition. Finally, the pattern recognition results based on CNN-LSTM are given and compared with those based on other traditional analysis methods. The results show that the pattern recognition rate of this method is the highest, with an average of 97.9%, and its overall accuracy is better than that of other traditional analysis methods. The CNN-LSTM model provides a reliable reference for GIS PD diagnosis.

Towards Secured Online Monitoring for Digitalized GIS Against Cyber-Attacks Based on IoT and Machine Learning

Recently, the Internet of Things (IoT) has an important role in the growth and development of digitalized electric power stations while offering ambitious opportunities, specifically real-time monitoring and cybersecurity. In this regard, this paper introduces a novel IoT architecture for the online monitoring of the gas-insulated switchgear (GIS) status instead of the traditional observation methods. The proposed IoT architecture is derived from the concept of the cyber-physic system (CPS) in Industry 4.0. However, the cyber-attacks and the classification of the GIS insulation defects represent the main challenges against the implementation of IoT topology for the online monitoring and tracking of the GIS status. For this purpose, advanced machine learning techniques are utilized to detect cyber-attacks to conduct the paradigm and verification. Different test scenarios on various defects in GIS are performed to demonstrate the effectiveness of the proposed IoT architecture. Partial discharge pulse sequence features are extracted for each defect to represent the inputs for IoT architecture. The results confirm that the proposed IoT architecture based on the machine learning technique, that is the extreme gradient boosting (XGBoost), can visualize all defects in the GIS with different alarms, besides showing the cyber-attacks on the networks effectively. Furthermore, the defects of GIS and the fake data due to the cyber-attacks are recognized and presented on the dashboard of the proposed IoT platform with high accuracy and more clarified visualization to enhance the decision-making about the GIS status.

Research on effectiveness of lightning impulses with different parameters for detecting protrusion defects in GIS

The focus of this paper is on the effectiveness of lightning impulse (LI) with different parameters for detecting protrusion defects in gas insulated switchgear (GIS). The study examines the 50% (U 50 ) breakdown probability for typical protrusion defects on the conductor of a 363 kV class GIS. A quantitative method is developed to predict breakdown voltage, calculate minimum detectable length of defect, and express the effectiveness of the LI test. The results show that there is little difference of U 50 between aperiodic and oscillating LI. Corresponding detecting effectiveness decreases for LI with long wavefront, which is expressed quantitatively by detectable area. The critical defect length detected increases with increased wavefront. This study provides a general method to express the effectiveness of LI with different parameters for detecting protrusion defects and it support the LI test for on-site testing of GIS.

A Classification System for Insulation Defect Identification of Gas-Insulated Switchgear (GIS), Based on Voiceprint Recognition Technology

Insulation defects that occur in gas-insulated switchgear (GIS), which is one of the most important types of equipment in the power grid, can lead to serious accidents. The ultrasonic detection method is commonly used to detect partial discharge (PD) signals in power equipment to discover defects. However, the traditional method to diagnose defects in GIS with ultrasonic PD signals is still based on the experience of testers. In this study, a classification system was proposed to identify insulation defects of GIS, based on voiceprint recognition technology. Twelve coefficients from mel frequency cepstral coefficient (MFCC) and 24 delta MFCC features were extracted as the acoustic features of the system. A support vector machine (SVM) multi-classifier was constructed to perform the classification and the sequential minimal optimization (SMO) algorithm was used to optimize the computational efficiency of the SVM. The experiments were conducted on a 110 kV GIS with different kinds of insulation defects. The results verified that the classification system with SMO-SVM achieved better identification accuracy and efficiency than the system with SVM. Therefore, it reveals the feasibility of the system to realize identification of insulation defects in GIS automatically and accurately.

On-line Monitoring of GIS Using Fourier Transform Infrared Spectroscopy

Different decomposition component gases caused by different insulation defects in gas-insulated switchgear will accelerate insulation degradation, which is a major threat to human and equipment safety. In this paper, we proposed online monitoring of GIS based on Fourier transform infrared spectroscopy. Firstly, we analysed the absorption characteristics of infrared light in SO2 and moisture. Secondly, we introduce the design principle and signal stabilization method of the laser. Thirdly, we propose the methods for measuring the concentration of SO2 and moisture. Finally, experiments with various concentration of SO2 and moisture are deployed, which proves that the GIS on-line monitoring method for decomposed gas has small measurement error and can be applied in practical applications.

Identification of Partial Discharge Defects in Gas-Insulated Switchgears by Using a Deep Learning Method

In this study, a novel method based on deep learning was developed for partial discharge (PD) pattern recognition. Traditional PD recognition methods are crucial for extracting features from PD patterns. The method of extracting crucial features is the key to PD pattern identification. The fractal theory is commonly used to determine the features of discharge patterns. The feature distribution of different defect types can be determined according to the fractal dimensions and lacunarity. However, finding fractal features is a complicated process. In this study, a PD image was entered as an input into a deep learning system to reduce the complexity of finding features. First, four defect type of gas-insulated switchgear (GIS) experimental models are established. Then, an LDP-5 inductive sensor (L-sensor) was used to measure the ground line signals caused by PD phenomenon. Second, these electrical signals were transformed into the most representative 3D ($n$ -$q$ -$\varphi$ ) PD patterns. Finally, a convolutional neural network was employed for PD image pattern recognition. A total of 160 sets of PD patterns were measured using a 15-kV GIS. The results obtained with the proposed method were compared with those obtained with the fractal method. The results revealed that the proposed method is easy to use and can easily distinguish various defect types. The proposed approach can determine the GIS insulation status and provide information to construction units for maintenance.

Insulation characteristics of GIS bus with conductive protrusion attached to the high-voltage electrode under lightning impulses with different wavefront times

The wavefront time T f of lightning impulse plays an important role in the detection of insulation defects in metal enclosed gas insulated switchgear (GIS). Usually, a rod-plane electrode system with different geometric parameters is adopted to simulate local field enhancements in GIS. But, for GIS, the defects may occur on the high-voltage electrode which can not be properly simulated by a rod-plane electrode. To research the influence of waveform parameters on actual defects in GIS, cylinder-plane, coaxial cylinder, and basin-type insulator with a conductive protrusion, were used in this study. A generating system of double-exponential impulses with wavefront times in the range of 0.08 to 15 μs and wave tail times of approximately 50 μs was used to simulate different impulse test waveforms based on a fully enclosed oil-immersed Marx generator. The results show that for the cylinder-plane gap with a protrusion, when the gap distance of the needle tip to the plane is constant, increasing the length of the needle decreases the 50% breakdown voltage. Comparing with the insulation characteristics of rod-plane gap, the U 50 %-T f curves of the cylinder-plane electrode with needle defects thereafter no longer show a U shape as the rod-plane electrode, but increase with an increase in the wavefront time. The difference can be explained by the shielding effect of the cylinder electrode on the defect, which changes the electric field distribution of the gap. The U 50 %-T f curves of GIS bus with different voltage levels and basin-type insulator with surface conductive defect also show the similar increase trend. For V-t characteristics of GIS bus with different voltage levels, we find that the higher the voltage level, the more obvious the trend of the V-t characteristic bathtub curve.

Pattern Recognition of Partial Discharge by Using Scale parameters-Energy Entropy Characteristic Pairs

In this paper, the complex wavelet transform (CWT) was used to process the ultra-high frequency partial discharge (UHF PD) signal in gas insulated switchgear (GIS) at different scales. The trend curves of complex wavelet transform energy entropy (CWT-EE) under different decomposition scale were analyzed, and it was found that the PD feature information mainly distributed in the scales, in which the gradient of CWT-EE is big. Besides, The CWT-EE characteristics and their scales were extracted to the structure characteristic pairs for PD type identification. The recognition results show that the characteristic pair could effectively identify four typical defects in GIS and obviously reduce the feature dimension.

Decomposition characteristics of SF<inf>6</inf> under overheating conditions

Overheating is one of the severe defects in gas-insulated switchgear (GIS). However, researches on decomposition characteristics of SF 6 under overheating conditions are relatively limited. In this paper, SF 6 decomposition products are studied under three overheating conditions which are overheating of solid insulation, overheating of metal, and overheating of metal adjacent to insulating material. Influence of overheating temperature, gas pressure, initial moisture content and adsorbent on the decomposition characteristics of SF 6 is discussed. The results show that typical decomposition products of SF 6 under overheating of solid insulation are H 2 , CO, CO 2 , CH 4 , SOF 2 , SO 2 , H 2 S, COS and CF4, and typical decomposition products of SF 6 under overheating of metal are SOF 2 , SO 2 , CO and CO 2 . Besides, typical decomposition products of SF 6 under overheating of metal adjacent to insulating material are CO, CO 2 , SOF 2 , SO 2 , CF 4 , CS 2 and COS. Types of overheating can be recognized by characteristic ratio of (SOF 2 +SO 2 )/(CO+CO 2 ): for overheating of solid insulation, the ratio is 0 to 0.25, and for overheating involving metal, it is greater than 1. Two types of overheating involving metal can be distinguished by their characteristic products such as CS 2 , COS and CF 4 .

Partial discharge detection and analysis of needle-plane defect in SF6 under negative oscillating lightning impulse voltage based on UHF method

Gas insulated switchgear (GIS) is used in power system more and more widely. With the increased voltage level of power system, field impulse voltage withstand test becomes increasingly important. It would be beneficial for a comprehensive insulation evaluation to carry out the detection of partial discharge (PD) under impulse voltage withstand test in field. Metal protrusions or particles on HV conductor are typical defects in GIS and are critical at lightning impulse voltages. This paper therefore aims at PD characteristics study on needle-plane defect under oscillating lightning impulse (OLI). PD detection system was established based on ultra-high frequency (UHF) and pulse current detection. Then, the PD of a needle-plane defect under negative OLI was measured. The relation between statistical time lag, PD amplitude, PD times and the applied voltage was analyzed respectively based on UHF detection. The results show that UHF detection has high sensitivity and it is suitable in field impulse voltage withstand test. PDs occur around each wave peak of the applied voltage. With the increase of peak voltage, PD magnitude and PD times in each oscillation cycle both increase linearly. Further analysis indicates that the occurrence and development of PD under negative OLI can be interpreted by electron emission from metal surface, field detachment from negative ions and corona stabilization effect. This approach is valuable for PD measurement and analysis in field impulse voltage withstand test for GIS.