Random Convolutional Kernel Transform with Empirical Mode Decomposition for Medium Voltage Insulator Classification based on Ultrasound Sensor

Published: Jan. 22, 2024

The electrical energy supply relies on the satisfactory operation of insulators. ultrasound recorded from insulators in different conditions has a time series output, which can be used to classify faulty random convolutional kernel transform (Rocket) algorithms use filters extract various features data. This paper proposes combination Rocket algorithms, machine learning classifiers, and empirical mode decomposition (EMD) methods, such as complete ensemble with adaptive noise (CEEMDAN), wavelet (EWT), variational (VMD). results show that EMD methods combined MiniRocket, significantly improve accuracy logistic regression insulator fault diagnosis. proposed strategy achieves respectively an 0.992 using CEEMDAN, 0.995 EWT, 0.980 VMD. These highlight potential incorporating failure detection models enhance safety dependability power systems.

Language: Английский

---

Research on the prediction method of porcelain insulator pollution degree based on electric field monitoring

Electric Power Systems Research, Journal Year: 2024, Volume and Issue: 233, P. 110488 - 110488

Published: May 22, 2024

Language: Английский

---

A Failure Diagnosis Approach for Firefighting Pump Based Deep Learning GRU Methodology

Published: June 26, 2024

Language: Английский

---

Failure Classified Method for Diesel Generators Based Long Short-Term Memory Approach

Published: June 26, 2024

Language: Английский

---

Failure evaluation of outdoor polymeric insulators based on statistical condition-assessment methods

Electric Power Systems Research, Journal Year: 2024, Volume and Issue: 238, P. 111134 - 111134

Published: Oct. 9, 2024

Language: Английский

---

Ensemble Pretrained Convolutional Neural Networks for the Classification of Insulator Surface Conditions

Energies, Journal Year: 2024, Volume and Issue: 17(22), P. 5595 - 5595

Published: Nov. 8, 2024

Overhead transmission line insulators are non-conductive materials that separate conductors from grounded towers. Once in operation, they frequently experience environmental pollution and electrical or mechanical stress. Since adverse operational conditions can lead to insulation failure, regular inspections essential prevent power outages. To this end, paper proposes a novel technique based on deep convolutional neural networks (CNNs) classify high-voltage insulator surface their image. Successful applications of CNNs computer vision have led several pretrained architectures for image classification. use these models, practitioner typically fine-tunes selects one final model via selection stage discards all other models. In contrast with many existing studies such “winner-takes-all” approach, here, we identify the best subset seven popular CNN combined by soft voting form an ensemble classifier. From machine learning (ML) perspective, focus is warranted because base each architecture operates as feature extractor them works combination various extraction rules. Our numerical experiments demonstrate advantage identified individual architectures.

Language: Английский

---

Dielectric property prediction of coated high voltage glass insulators based on experimental analysis

Progress in Organic Coatings, Journal Year: 2024, Volume and Issue: 200, P. 109003 - 109003

Published: Dec. 18, 2024

Language: Английский

---

Evaluating Insulator Contamination Using Dielectric Loss Factor: A Comparative Study of Silicone Rubber and Eva Composites

Published: Jan. 1, 2024

Language: Английский

---