Study on Multi-RBF-SVM for Transformer Fault Diagnosis
| Autor: | Li-ping Qu, Chong-jie Liu, Zhao Lu, Hao-han Zhou |
|---|---|
| Rok vydání: | 2018 |
| Předmět: |
Normal type
Computer science 020209 energy ComputingMethodologies_IMAGEPROCESSINGANDCOMPUTERVISION 02 engineering and technology law.invention Support vector machine Prediction algorithms ComputingMethodologies_PATTERNRECOGNITION law Kernel (statistics) 0202 electrical engineering electronic engineering information engineering Transformer Algorithm |
| Zdroj: | 2018 17th International Symposium on Distributed Computing and Applications for Business Engineering and Science (DCABES). |
| DOI: | 10.1109/dcabes.2018.00056 |
| Popis: | Two improved fault diagnosis algorithms were proposed in this paper. One is New Three-Ratio (NTR) algorithm which introduces the code "000"as normal type code into the Transformer Three-Ratio(TTR). The other is Multi-Radial Basis Function-Support Vector Machine (M-RBF-SVM) algorithm which introduces the multi-RBF kernel function. And, the representative Tradition Three-Ratio algorithm is selected as the simulation comparison object. The results indicate that M-RBF-SVM can achieve higher diagnosis accuracy and excellently generalization ability. |
| Databáze: | OpenAIRE |
| Externí odkaz: |
|