Pub Date : 2025-07-30DOI: 10.1109/TDEI.2025.3588640
{"title":"IEEE Dielectrics and Electrical Insulation Society Information","authors":"","doi":"10.1109/TDEI.2025.3588640","DOIUrl":"https://doi.org/10.1109/TDEI.2025.3588640","url":null,"abstract":"","PeriodicalId":13247,"journal":{"name":"IEEE Transactions on Dielectrics and Electrical Insulation","volume":"32 4","pages":"C3-C3"},"PeriodicalIF":3.1,"publicationDate":"2025-07-30","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://ieeexplore.ieee.org/stamp/stamp.jsp?tp=&arnumber=11104953","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"144739785","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":3,"RegionCategory":"工程技术","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2025-07-30DOI: 10.1109/TDEI.2025.3588642
{"title":"IEEE Transactions on Dielectrics and Electrical Insulation Publication Information","authors":"","doi":"10.1109/TDEI.2025.3588642","DOIUrl":"https://doi.org/10.1109/TDEI.2025.3588642","url":null,"abstract":"","PeriodicalId":13247,"journal":{"name":"IEEE Transactions on Dielectrics and Electrical Insulation","volume":"32 4","pages":"C2-C2"},"PeriodicalIF":3.1,"publicationDate":"2025-07-30","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://ieeexplore.ieee.org/stamp/stamp.jsp?tp=&arnumber=11104951","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"144739978","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":3,"RegionCategory":"工程技术","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
The phenomenon of surface charging caused by electrical discharges has attracted significant attention because of its adverse effects on electrical systems and its industrial applications. Since the surface charging and discharge current are both influenced by the charges generated during the discharge, it is important to study the correlation quantitatively. The primary goal of the study is to analyze the relationship between the discharge current pulse and the corresponding surface charge deposition for positive and negative excitations. To establish the relation, variations were introduced in two key parameters influencing the discharge process: the discharge medium (N2, CO2, and dry air) and the pressure (100, 90, 80, 70, and 60 kPa) in each medium. The excitation voltage waveform is chosen to ensure the generation of only a single current pulse during the discharge. The positive excitation resulted in a higher pulse magnitude for N2 and dry air, whereas CO2 exhibited an opposite trend. The change in the current pulse is found to be directly proportional to the variation in charge deposition. The derived empirical formulas establish a linear correlation between the total charge computed from the current pulse and the deposited surface charge, verified by Pearson’s correlation coefficient, which suggests a good correlation strength. Of the three gaseous media, CO2 has shown a lower margin of error and consistent discharge results.
{"title":"Correlating Discharge Current Pulse With Surface Charge Deposition in Diverse Gaseous Environment","authors":"Shelly Saini;Shakthi Prasad D;Thami Zeghloul;Lucian Dascalescu","doi":"10.1109/TDEI.2025.3591391","DOIUrl":"https://doi.org/10.1109/TDEI.2025.3591391","url":null,"abstract":"The phenomenon of surface charging caused by electrical discharges has attracted significant attention because of its adverse effects on electrical systems and its industrial applications. Since the surface charging and discharge current are both influenced by the charges generated during the discharge, it is important to study the correlation quantitatively. The primary goal of the study is to analyze the relationship between the discharge current pulse and the corresponding surface charge deposition for positive and negative excitations. To establish the relation, variations were introduced in two key parameters influencing the discharge process: the discharge medium (N2, CO2, and dry air) and the pressure (100, 90, 80, 70, and 60 kPa) in each medium. The excitation voltage waveform is chosen to ensure the generation of only a single current pulse during the discharge. The positive excitation resulted in a higher pulse magnitude for N2 and dry air, whereas CO2 exhibited an opposite trend. The change in the current pulse is found to be directly proportional to the variation in charge deposition. The derived empirical formulas establish a linear correlation between the total charge computed from the current pulse and the deposited surface charge, verified by Pearson’s correlation coefficient, which suggests a good correlation strength. Of the three gaseous media, CO2 has shown a lower margin of error and consistent discharge results.","PeriodicalId":13247,"journal":{"name":"IEEE Transactions on Dielectrics and Electrical Insulation","volume":"32 5","pages":"2756-2764"},"PeriodicalIF":3.1,"publicationDate":"2025-07-22","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145190288","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":3,"RegionCategory":"工程技术","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
The inverter-driven motors are increasingly used in industrial and mobility applications, driving the demand for greater performance and reliability. Recent advances in power electronics have raised inverter output frequencies and slew rates, increasing the risk of discharge and insulation failure. A better understanding of discharge phenomena is thus essential. The authors are developing numerical simulations of dielectric barrier discharge (DBD) in twisted pairs of enameled wire. This study investigates the estimation and applicability of the secondary electron emission (SEE) coefficient ($gamma $ ) to the DBD simulations, addressing the lack of empirical data. As a result, fitting was found effective for estimating $gamma $ from discharge voltage measurements. The estimated values were $4.7times 10^{text {-3}}$ for polyimide (PI) and $7.5times 10^{text {-3}}$ for polyethylene (PE). Applying these values in DBD simulations suggests the potential to estimate discharge voltages under various pressure conditions. These findings imply that DBD simulations can enhance the accuracy of predictions of discharge phenomena in twisted pairs of enameled wire.
{"title":"Estimation of Secondary Electron Emission Coefficients for Dielectric Barrier Discharge Simulations","authors":"Yoshitaka Miyaji;Hirotaku Ishikawa;Yasutomo Otake;Fuma Yamada;Yusuke Kikuchi","doi":"10.1109/TDEI.2025.3589990","DOIUrl":"https://doi.org/10.1109/TDEI.2025.3589990","url":null,"abstract":"The inverter-driven motors are increasingly used in industrial and mobility applications, driving the demand for greater performance and reliability. Recent advances in power electronics have raised inverter output frequencies and slew rates, increasing the risk of discharge and insulation failure. A better understanding of discharge phenomena is thus essential. The authors are developing numerical simulations of dielectric barrier discharge (DBD) in twisted pairs of enameled wire. This study investigates the estimation and applicability of the secondary electron emission (SEE) coefficient (<inline-formula> <tex-math>$gamma $ </tex-math></inline-formula>) to the DBD simulations, addressing the lack of empirical data. As a result, fitting was found effective for estimating <inline-formula> <tex-math>$gamma $ </tex-math></inline-formula> from discharge voltage measurements. The estimated values were <inline-formula> <tex-math>$4.7times 10^{text {-3}}$ </tex-math></inline-formula> for polyimide (PI) and <inline-formula> <tex-math>$7.5times 10^{text {-3}}$ </tex-math></inline-formula> for polyethylene (PE). Applying these values in DBD simulations suggests the potential to estimate discharge voltages under various pressure conditions. These findings imply that DBD simulations can enhance the accuracy of predictions of discharge phenomena in twisted pairs of enameled wire.","PeriodicalId":13247,"journal":{"name":"IEEE Transactions on Dielectrics and Electrical Insulation","volume":"32 5","pages":"3117-3119"},"PeriodicalIF":3.1,"publicationDate":"2025-07-16","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145189980","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":3,"RegionCategory":"工程技术","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2025-07-16DOI: 10.1109/TDEI.2025.3589978
Yuwei Fu;Peng Ji;Shuai Wen;Rong Liang
Magnetron sputtering is widely used in thin film fabrication and surface modification of materials. During the sputtering process, the spatial species distribution significantly impacts the deposited film’s properties. However, there are still some difficulties in understanding the spatial species distribution, transport and energy control, resulting in uneven coating and low target utilization. In this article, we utilized a 2-D magnetron sputtering plasma model to further investigate the species distribution of the plasma under different excitation voltage sources and consequently obtain the Ar+ sputtering energy distribution. The erosion phenomenon was studied in the transport of ions in matter (TRIM) software, and the particle energy and angle obtained from the plasma simulation were used as input to study the incident distribution and sputtering yield. The results show significant differences in the distribution, density, and sputtering energy of plasma under dc, radio frequency (RF) (13.56 MHz) and high-power pulse (HPP) excitation voltage sources. Under dc, the electron distribution is more uniform than other excitation sources, covering 40%–50% of the target surface area. The initial sputtering energy distribution ranges from 0 to 400 eV with an erosion depth of 20Å, and the sputtering yield is approximately proportional to the voltage. The sputtering yield increases slower under RF when the voltage reaches 1000 V. Under RF, the electric field distribution is uniform at 800 V, but Ar+ is concentrated covering only 15% of the target surface. Under HPP, the electron and Ar+ densities reach $10^{{17}}$ –$10^{{18}}$ m${}^{-{3}}$ , with the highest electron current density reaching $5times 10^{{3}}$ A/m2. The sputtering depth is 30Å. This research has significant importance in optimizing the process parameters of magnetron sputtering and improving film performance. It provides strong support for the development and application of magnetron sputtering processes.
{"title":"The Distribution and Erosion Characteristics of Plasma Particles in Magnetron Sputtering Under Different Excitation Voltage Sources","authors":"Yuwei Fu;Peng Ji;Shuai Wen;Rong Liang","doi":"10.1109/TDEI.2025.3589978","DOIUrl":"https://doi.org/10.1109/TDEI.2025.3589978","url":null,"abstract":"Magnetron sputtering is widely used in thin film fabrication and surface modification of materials. During the sputtering process, the spatial species distribution significantly impacts the deposited film’s properties. However, there are still some difficulties in understanding the spatial species distribution, transport and energy control, resulting in uneven coating and low target utilization. In this article, we utilized a 2-D magnetron sputtering plasma model to further investigate the species distribution of the plasma under different excitation voltage sources and consequently obtain the Ar+ sputtering energy distribution. The erosion phenomenon was studied in the transport of ions in matter (TRIM) software, and the particle energy and angle obtained from the plasma simulation were used as input to study the incident distribution and sputtering yield. The results show significant differences in the distribution, density, and sputtering energy of plasma under dc, radio frequency (RF) (13.56 MHz) and high-power pulse (HPP) excitation voltage sources. Under dc, the electron distribution is more uniform than other excitation sources, covering 40%–50% of the target surface area. The initial sputtering energy distribution ranges from 0 to 400 eV with an erosion depth of 20Å, and the sputtering yield is approximately proportional to the voltage. The sputtering yield increases slower under RF when the voltage reaches 1000 V. Under RF, the electric field distribution is uniform at 800 V, but Ar+ is concentrated covering only 15% of the target surface. Under HPP, the electron and Ar+ densities reach <inline-formula> <tex-math>$10^{{17}}$ </tex-math></inline-formula>–<inline-formula> <tex-math>$10^{{18}}$ </tex-math></inline-formula> m<inline-formula> <tex-math>${}^{-{3}}$ </tex-math></inline-formula>, with the highest electron current density reaching <inline-formula> <tex-math>$5times 10^{{3}}$ </tex-math></inline-formula> A/m2. The sputtering depth is 30Å. This research has significant importance in optimizing the process parameters of magnetron sputtering and improving film performance. It provides strong support for the development and application of magnetron sputtering processes.","PeriodicalId":13247,"journal":{"name":"IEEE Transactions on Dielectrics and Electrical Insulation","volume":"32 5","pages":"2730-2737"},"PeriodicalIF":3.1,"publicationDate":"2025-07-16","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145190289","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":3,"RegionCategory":"工程技术","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
This article presents a novel method for the direct observation of electric field profiles in solid dielectrics based on the step electroacoustic method. In the proposed approach, long pressure waves generated by a step voltage excitation are detected using a piezoelectric sensor that is significantly thicker than the test specimen. The transducer acts as a high-pass filter, allowing a step-function-like pressure signal to be observed for each space charge. As a result, the oscillogram directly shows the electric field profile in the specimen. Using an 80-$mu $ m-thick PVDF sensor and a 50-$mu $ m-thick PET film as the test specimen, we experimentally demonstrate that the spatial resolution of the profile after deconvolution can be less than $10~mu $ m.
{"title":"Direct Observation of Electric Field in Solid Dielectrics Using SEA Method With Piezoelectric Sensor Thicker Than Test Specimen","authors":"Kazunori Kadowaki;Shinji Yudate;Ryotaro Ozaki;Masumi Fukuma","doi":"10.1109/TDEI.2025.3589336","DOIUrl":"https://doi.org/10.1109/TDEI.2025.3589336","url":null,"abstract":"This article presents a novel method for the direct observation of electric field profiles in solid dielectrics based on the step electroacoustic method. In the proposed approach, long pressure waves generated by a step voltage excitation are detected using a piezoelectric sensor that is significantly thicker than the test specimen. The transducer acts as a high-pass filter, allowing a step-function-like pressure signal to be observed for each space charge. As a result, the oscillogram directly shows the electric field profile in the specimen. Using an 80-<inline-formula> <tex-math>$mu $ </tex-math></inline-formula>m-thick PVDF sensor and a 50-<inline-formula> <tex-math>$mu $ </tex-math></inline-formula>m-thick PET film as the test specimen, we experimentally demonstrate that the spatial resolution of the profile after deconvolution can be less than <inline-formula> <tex-math>$10~mu $ </tex-math></inline-formula>m.","PeriodicalId":13247,"journal":{"name":"IEEE Transactions on Dielectrics and Electrical Insulation","volume":"32 5","pages":"3114-3116"},"PeriodicalIF":3.1,"publicationDate":"2025-07-15","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145210079","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":3,"RegionCategory":"工程技术","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2025-07-14DOI: 10.1109/TDEI.2025.3588794
Giacomo Galli;Michael J. Kirkpatrick;Emmanuel Odic
Experiments are carried out to determine if the changes in the level of ionizing radiation provoke detectable changes in the inception voltage of electrical discharges. The inception voltage for dc corona discharge from a sharp cathode is measured for four cases: 1) ambient conditions; 2) shielded conditions; 3) in the presence of a beta source; and 4) in the presence of a gamma source. The results indicate a decrease of a few hundred volts in the inception voltage when the sources of gamma or beta radiation are placed in the proximity of the discharge zone. No difference in the extinction voltage is detected between the different cases. Additionally, variability in the inception voltage is observed to be notably reduced in the presence of ionizing radiation.
{"title":"Effects of Ionizing Radiation on the Inception Voltage of Electrical Discharges","authors":"Giacomo Galli;Michael J. Kirkpatrick;Emmanuel Odic","doi":"10.1109/TDEI.2025.3588794","DOIUrl":"https://doi.org/10.1109/TDEI.2025.3588794","url":null,"abstract":"Experiments are carried out to determine if the changes in the level of ionizing radiation provoke detectable changes in the inception voltage of electrical discharges. The inception voltage for dc corona discharge from a sharp cathode is measured for four cases: 1) ambient conditions; 2) shielded conditions; 3) in the presence of a beta source; and 4) in the presence of a gamma source. The results indicate a decrease of a few hundred volts in the inception voltage when the sources of gamma or beta radiation are placed in the proximity of the discharge zone. No difference in the extinction voltage is detected between the different cases. Additionally, variability in the inception voltage is observed to be notably reduced in the presence of ionizing radiation.","PeriodicalId":13247,"journal":{"name":"IEEE Transactions on Dielectrics and Electrical Insulation","volume":"32 5","pages":"2822-2829"},"PeriodicalIF":3.1,"publicationDate":"2025-07-14","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145189972","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":3,"RegionCategory":"工程技术","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Breakdown voltage prediction for air gaps is a crucial topic in high-voltage insulation research. Building upon existing studies and fundamental theories, the authors identify the gap configuration as the root cause influencing breakdown voltage. In light of this, the authors propose a novel approach for mapping the gap configuration to the breakdown voltage. First, the data structuring method is proposed to abstract the air-gap configuration into a mathematical point set. Second, inspired by the strengths of graph neural networks (GNNs) and Transformers, the authors introduce a hybrid model, the point graph Transformer (PGT), which enables the prediction of the breakdown voltage through the point set. This approach is applicable to different gap configurations and influencing factors. To improve prediction accuracy and facilitate comparative studies, the authors present tailored data augmentation and padding strategies. Specifically, focusing on air gaps with 2-D axisymmetric structures, comprehensive breakdown voltage prediction research and comparative studies are conducted. Furthermore, an ablation study is conducted and analyzed. The results demonstrate that the method delivers accurate breakdown voltage predictions, outperforming other methods. Notably, the evaluation metrics for the test set achieve a mean relative error (MRE) of 7.25%, a maximum relative error (MaxRE) of 14.83%, and an R-square of 96.86%.
{"title":"From Gap Configuration to Breakdown Voltage: A Novel Data-Driven Method for Predicting Breakdown Voltage","authors":"Shaocheng Wu;Linong Wang;Zisheng Zeng;Shengxuan Zu;Bin Song;Jiachen Gao","doi":"10.1109/TDEI.2025.3586605","DOIUrl":"https://doi.org/10.1109/TDEI.2025.3586605","url":null,"abstract":"Breakdown voltage prediction for air gaps is a crucial topic in high-voltage insulation research. Building upon existing studies and fundamental theories, the authors identify the gap configuration as the root cause influencing breakdown voltage. In light of this, the authors propose a novel approach for mapping the gap configuration to the breakdown voltage. First, the data structuring method is proposed to abstract the air-gap configuration into a mathematical point set. Second, inspired by the strengths of graph neural networks (GNNs) and Transformers, the authors introduce a hybrid model, the point graph Transformer (PGT), which enables the prediction of the breakdown voltage through the point set. This approach is applicable to different gap configurations and influencing factors. To improve prediction accuracy and facilitate comparative studies, the authors present tailored data augmentation and padding strategies. Specifically, focusing on air gaps with 2-D axisymmetric structures, comprehensive breakdown voltage prediction research and comparative studies are conducted. Furthermore, an ablation study is conducted and analyzed. The results demonstrate that the method delivers accurate breakdown voltage predictions, outperforming other methods. Notably, the evaluation metrics for the test set achieve a mean relative error (MRE) of 7.25%, a maximum relative error (MaxRE) of 14.83%, and an R-square of 96.86%.","PeriodicalId":13247,"journal":{"name":"IEEE Transactions on Dielectrics and Electrical Insulation","volume":"33 1","pages":"671-681"},"PeriodicalIF":3.1,"publicationDate":"2025-07-07","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"146102956","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":3,"RegionCategory":"工程技术","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2025-07-03DOI: 10.1109/TDEI.2025.3585848
Changyun Li;Jun Shao;Fengtian Sun;Yongjin Yu
With the global energy structure evolving, high-capacity, long-distance power transmission will increasingly employ dc systems. Accurate modeling and analysis of the dc cable terminal (CT)—a power device with insulating dielectrics in solid, liquid, and gas states—are crucial for preventing fault formation. Based on the thermal-assisted/variable-range hopping conductance model of cross-linked polyethylene (XLPE), this study identifies the electric field reversal phenomenon in the radial section of XLPE under various conditions. By using different position parameters and operating parameters as independent variables for electric field intensity fitting, we enhance the efficiency of accurately obtaining the internal electric field intensity during the dc CT operation. The analytical results show that the internal electric field is most uniform when an appropriate load is applied to the conductor, representing the healthiest operating mode for the dc CT. This study provides a reference for determining the internal insulation state of the CT under different operating conditions, enabling early detection of potential issues and the adoption of corresponding countermeasures.
{"title":"XLPE Electric Field Reversal Caused by Temperature Rise of Oil-Filled DC Cable Terminal","authors":"Changyun Li;Jun Shao;Fengtian Sun;Yongjin Yu","doi":"10.1109/TDEI.2025.3585848","DOIUrl":"https://doi.org/10.1109/TDEI.2025.3585848","url":null,"abstract":"With the global energy structure evolving, high-capacity, long-distance power transmission will increasingly employ dc systems. Accurate modeling and analysis of the dc cable terminal (CT)—a power device with insulating dielectrics in solid, liquid, and gas states—are crucial for preventing fault formation. Based on the thermal-assisted/variable-range hopping conductance model of cross-linked polyethylene (XLPE), this study identifies the electric field reversal phenomenon in the radial section of XLPE under various conditions. By using different position parameters and operating parameters as independent variables for electric field intensity fitting, we enhance the efficiency of accurately obtaining the internal electric field intensity during the dc CT operation. The analytical results show that the internal electric field is most uniform when an appropriate load is applied to the conductor, representing the healthiest operating mode for the dc CT. This study provides a reference for determining the internal insulation state of the CT under different operating conditions, enabling early detection of potential issues and the adoption of corresponding countermeasures.","PeriodicalId":13247,"journal":{"name":"IEEE Transactions on Dielectrics and Electrical Insulation","volume":"32 4","pages":"2366-2374"},"PeriodicalIF":3.1,"publicationDate":"2025-07-03","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"144739799","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":3,"RegionCategory":"工程技术","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2025-07-03DOI: 10.1109/TDEI.2025.3585844
Saurabh Dutta;Shiyu Chen;Hazlee Azil Illias
Partial discharge (PD) is a prevalent phenomenon in high-voltage (HV) equipment, and its accurate classification is crucial for ensuring the reliability of power systems. For in situ systems, different types of faults, such as corona, floating electrode, surface, and void discharge, exhibit varying occurrences, posing challenges to accurate classification. This research addresses the inherent issues of classification accuracy caused by unbalanced fault data. Employing z-score normalization and combined synthetic data generation using a random undersampling and synthetic minority oversampling technique (SMOTE) ensures a fair representation of different fault types, leading to more accurate classification results. Further, after applying grid-search to optimize the hyperparameters, k-nearest neighbor (KNN), random forest (RF), and gradient boosting (GB) have achieved accuracies of 98.43%, 95.29%, and 88.54% for balanced denoised, unbalanced denoised, and unbalanced noisy datasets, respectively. The presented results also demonstrate a significant statistical difference in classifier accuracies between the three datasets, as confirmed by the analysis of variance (ANOVA) test. This emphasizes the efficacy of balancing the denoised signal features for improved classification performance. The findings of this work contribute valuable insights into the optimization of PD classification models, paving the way for more reliable fault detection and classification in HV equipment.
{"title":"Enhancing Partial Discharge Classification Through Augmented Fault Data Balancing","authors":"Saurabh Dutta;Shiyu Chen;Hazlee Azil Illias","doi":"10.1109/TDEI.2025.3585844","DOIUrl":"https://doi.org/10.1109/TDEI.2025.3585844","url":null,"abstract":"Partial discharge (PD) is a prevalent phenomenon in high-voltage (HV) equipment, and its accurate classification is crucial for ensuring the reliability of power systems. For in situ systems, different types of faults, such as corona, floating electrode, surface, and void discharge, exhibit varying occurrences, posing challenges to accurate classification. This research addresses the inherent issues of classification accuracy caused by unbalanced fault data. Employing z-score normalization and combined synthetic data generation using a random undersampling and synthetic minority oversampling technique (SMOTE) ensures a fair representation of different fault types, leading to more accurate classification results. Further, after applying grid-search to optimize the hyperparameters, k-nearest neighbor (KNN), random forest (RF), and gradient boosting (GB) have achieved accuracies of 98.43%, 95.29%, and 88.54% for balanced denoised, unbalanced denoised, and unbalanced noisy datasets, respectively. The presented results also demonstrate a significant statistical difference in classifier accuracies between the three datasets, as confirmed by the analysis of variance (ANOVA) test. This emphasizes the efficacy of balancing the denoised signal features for improved classification performance. The findings of this work contribute valuable insights into the optimization of PD classification models, paving the way for more reliable fault detection and classification in HV equipment.","PeriodicalId":13247,"journal":{"name":"IEEE Transactions on Dielectrics and Electrical Insulation","volume":"32 5","pages":"2948-2957"},"PeriodicalIF":3.1,"publicationDate":"2025-07-03","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145189967","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":3,"RegionCategory":"工程技术","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
京公网安备 11010802042870号
京ICP备2023020795号-1
扫码关注我们
求助内容：
应助结果提醒方式：