Transformer Fault Diagnosis Based on Multi-Strategy Enhanced Dung Beetle Algorithm and Optimized SVM
Accurate fault diagnosis of transformers is crucial for preventing power system failures and ensuring the continued reliability of electrical grids. To address the challenge of low accuracy in transformer fault diagnosis using support vector machines (SVMs), an enhanced fault diagnosis model is prop...
Saved in:
| Main Authors: | , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
MDPI AG
2024-12-01
|
| Series: | Energies |
| Subjects: | |
| Online Access: | https://www.mdpi.com/1996-1073/17/24/6296 |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| _version_ | 1850049815349559296 |
|---|---|
| author | Shuming Zhang Hong Zhou |
| author_facet | Shuming Zhang Hong Zhou |
| author_sort | Shuming Zhang |
| collection | DOAJ |
| description | Accurate fault diagnosis of transformers is crucial for preventing power system failures and ensuring the continued reliability of electrical grids. To address the challenge of low accuracy in transformer fault diagnosis using support vector machines (SVMs), an enhanced fault diagnosis model is proposed, which utilizes an improved dung beetle optimization algorithm (IDBO) to optimize an SVM. First, based on dissolved gas analysis (DGA), five characteristic quantities are selected as input features. Second, improvements to the DBO algorithm are made by incorporating Chebyshev chaotic mapping, a golden sine strategy, and dynamic weight coefficients for position updates. The performance of the IDBO is validated using four benchmark test functions, demonstrating faster convergence. Subsequently, the IDBO optimizes the SVM’s penalty factor <i>C</i> and kernel function parameter <i>g</i>, which are then input into the SVM for training, yielding an efficient fault diagnosis model. Finally, comparisons with other methods confirm the usefulness of the proposed model. Experimental results demonstrate that the IDBO–SVM model attains accuracy improvements of 1.69%, 8.47%, and 10.17% over dung beetle optimization–SVM (DBO–SVM), sparrow search algorithm–SVM (SSA–SVM), and grey wolf optimization–SVM (GWO–SVM) models, respectively. In addition to higher accuracy, the IDBO–SVM model also delivers a faster runtime, further highlighting its superior performance in transformer fault diagnosis. The proposed model has practical significance for enhancing the stability of transformer operation. |
| format | Article |
| id | doaj-art-1f2b2c48a5cf408fbb22cea2c3cf5550 |
| institution | DOAJ |
| issn | 1996-1073 |
| language | English |
| publishDate | 2024-12-01 |
| publisher | MDPI AG |
| record_format | Article |
| series | Energies |
| spelling | doaj-art-1f2b2c48a5cf408fbb22cea2c3cf55502025-08-20T02:53:38ZengMDPI AGEnergies1996-10732024-12-011724629610.3390/en17246296Transformer Fault Diagnosis Based on Multi-Strategy Enhanced Dung Beetle Algorithm and Optimized SVMShuming Zhang0Hong Zhou1East China Electric Power Test and Research Institute, China Datang Corporation Science and Technology General Research Institute Co., Ltd., Hefei 230088, ChinaSchool of Electrical Engineering and Automation, Wuhan University, Wuhan 430072, ChinaAccurate fault diagnosis of transformers is crucial for preventing power system failures and ensuring the continued reliability of electrical grids. To address the challenge of low accuracy in transformer fault diagnosis using support vector machines (SVMs), an enhanced fault diagnosis model is proposed, which utilizes an improved dung beetle optimization algorithm (IDBO) to optimize an SVM. First, based on dissolved gas analysis (DGA), five characteristic quantities are selected as input features. Second, improvements to the DBO algorithm are made by incorporating Chebyshev chaotic mapping, a golden sine strategy, and dynamic weight coefficients for position updates. The performance of the IDBO is validated using four benchmark test functions, demonstrating faster convergence. Subsequently, the IDBO optimizes the SVM’s penalty factor <i>C</i> and kernel function parameter <i>g</i>, which are then input into the SVM for training, yielding an efficient fault diagnosis model. Finally, comparisons with other methods confirm the usefulness of the proposed model. Experimental results demonstrate that the IDBO–SVM model attains accuracy improvements of 1.69%, 8.47%, and 10.17% over dung beetle optimization–SVM (DBO–SVM), sparrow search algorithm–SVM (SSA–SVM), and grey wolf optimization–SVM (GWO–SVM) models, respectively. In addition to higher accuracy, the IDBO–SVM model also delivers a faster runtime, further highlighting its superior performance in transformer fault diagnosis. The proposed model has practical significance for enhancing the stability of transformer operation.https://www.mdpi.com/1996-1073/17/24/6296transformer fault diagnosisdung beetle optimization algorithmsupport vector machinedissolved gas analysis |
| spellingShingle | Shuming Zhang Hong Zhou Transformer Fault Diagnosis Based on Multi-Strategy Enhanced Dung Beetle Algorithm and Optimized SVM Energies transformer fault diagnosis dung beetle optimization algorithm support vector machine dissolved gas analysis |
| title | Transformer Fault Diagnosis Based on Multi-Strategy Enhanced Dung Beetle Algorithm and Optimized SVM |
| title_full | Transformer Fault Diagnosis Based on Multi-Strategy Enhanced Dung Beetle Algorithm and Optimized SVM |
| title_fullStr | Transformer Fault Diagnosis Based on Multi-Strategy Enhanced Dung Beetle Algorithm and Optimized SVM |
| title_full_unstemmed | Transformer Fault Diagnosis Based on Multi-Strategy Enhanced Dung Beetle Algorithm and Optimized SVM |
| title_short | Transformer Fault Diagnosis Based on Multi-Strategy Enhanced Dung Beetle Algorithm and Optimized SVM |
| title_sort | transformer fault diagnosis based on multi strategy enhanced dung beetle algorithm and optimized svm |
| topic | transformer fault diagnosis dung beetle optimization algorithm support vector machine dissolved gas analysis |
| url | https://www.mdpi.com/1996-1073/17/24/6296 |
| work_keys_str_mv | AT shumingzhang transformerfaultdiagnosisbasedonmultistrategyenhanceddungbeetlealgorithmandoptimizedsvm AT hongzhou transformerfaultdiagnosisbasedonmultistrategyenhanceddungbeetlealgorithmandoptimizedsvm |