On-line fault diagnosis model for locomotive traction inverter based on wavelet transform and support vector machine
| Autor: | Sha Haoyuan, Pan Yi, Liu Ning, Zheng Jianyong, Mei Fei, Miao Huiyu |
|---|---|
| Rok vydání: | 2018 |
| Předmět: |
Computer science
medicine.medical_treatment 020208 electrical & electronic engineering Wavelet transform Hardware_PERFORMANCEANDRELIABILITY 02 engineering and technology Insulated-gate bipolar transistor Traction (orthopedics) Condensed Matter Physics Fault (power engineering) Atomic and Molecular Physics and Optics Fault detection and isolation Surfaces Coatings and Films Electronic Optical and Magnetic Materials Power (physics) 0202 electrical engineering electronic engineering information engineering medicine Electronic engineering Inverter 020201 artificial intelligence & image processing Electrical and Electronic Engineering Safety Risk Reliability and Quality Energy (signal processing) |
| Zdroj: | Microelectronics Reliability. :1274-1280 |
| ISSN: | 0026-2714 |
| DOI: | 10.1016/j.microrel.2018.06.069 |
| Popis: | A traction inverter is the power source for rail transit vehicles. An insulated-gate bipolar transistor (IGBT) is the primary component of a traction inverter. IGBT faults can cause serious problems in locomotive power supply systems. The disadvantage of traditional fault detection methods for IGBT modules is a lack of real-time processing and high efficiency. An on-line fault diagnosis method based on a wavelet transform and multi-classification support vector machine (multi-SVM) is proposed for IGBT faults. Wavelet decomposition is used to process fault current signals, and energy vectors are constructed. Multi-SVM is used to establish a fault recognition model. The validity of this method is verified by simulations with MATLAB/Simulink. |
| Databáze: | OpenAIRE |
| Externí odkaz: |
|
Full Text from ScienceDirect