| 引用本文: | 赵舒娅,郝亮亮,蔡宇昂,段贤稳,李华忠.基于改进SVDD的多相无刷励磁系统旋转整流器故障诊断[J].电力自动化设备,2025,45(7):105-113 |
| ZHAO Shuya,HAO Liangliang,CAI Yuang,DUAN Xianwen,LI Huazhong.Fault diagnosis of rotating rectifier in multiphase brushless excitation system based on improved SVDD[J].Electric Power Automation Equipment,2025,45(7):105-113 |
|
| 摘要: |
| 多相环形无刷励磁系统已在大容量核电机组中广泛应用，旋转整流器故障的准确高效诊断对提升励磁系统的运行可靠性具有重要意义。然而，通过分析励磁电流谐波特征的传统诊断方法难以准确区分旋转整流器的所有故障类型。为此，提出基于改进支持向量域描述(SVDD)的多相无刷励磁系统旋转整流器故障诊断方法。从故障诊断模式和需求出发，分析了励磁电流波形所反映出的各类故障特征，建立了分类诊断的理论基础。采用改进SVDD模型，通过核聚类的方式捕捉样本的内在结构，能够反映励磁电流中的微弱故障特征，从而建立相应的故障诊断规则。利用11相无刷励磁系统的动模实验验证所提方法的有效性。结果表明，所提方法能够准确区分旋转整流器的不同故障模式，且具有一定的未知故障检测能力。 |
| 关键词: 无刷励磁系统 旋转整流器 故障诊断 支持向量域描述 未知故障检测 |
| DOI：10.16081/j.epae.202502006 |
| 分类号:TM461;TL4 |
| 基金项目:中央高校基本科研业务费专项资金资助项目(2023YJS162,2020JBM070)；中广核集团公司科技资金资助项目(3100077013) |
|
| Fault diagnosis of rotating rectifier in multiphase brushless excitation system based on improved SVDD |
|
ZHAO Shuya1, HAO Liangliang1, CAI Yuang1, DUAN Xianwen2, LI Huazhong3
|
|
1.School of Electrical Engineering, Beijing Jiaotong University, Beijing 100044, China;2.China Nuclear Power Operations Co.,Ltd.,Shenzhen 518172, China;3.NR Electric Co.,Ltd.,Nanjing 211102, China
|
| Abstract: |
| The multiphase ring brushless excitation system has been widely used in large capacity nuclear power units. Accurate and efficient fault diagnosis of rotating rectifier is of great significance for improving the operational reliability of the excitation system. However, traditional diagnostic methods based on harmonic analysis of field current are difficult to accurately distinguish all types of faults in rotating rectifiers. Therefore, a fault diagnosis method for rotating rectifiers in multiphase brushless excitation system based on improved support vector domain description(SVDD) is proposed. Starting from the fault diagnosis modes and requirements, the various fault characteristics reflected by the field current waveforms are analyzed and the theoretical basis for classification diagnosis is established. Then, the improved SVDD model is adopted to capture the intrinsic structure of the samples through kernel clustering, which can reflect the weak fault characteristics in field current, so corresponding fault diagnosis rules are established. The effectiveness of the proposed method is validated through the dynamic simulation experiments of an 11-phase brushless excitation system. The results show that the proposed method can distinguish different fault modes of rotating rectifiers accurately and has a certain ability to detect unknown faults. |
| Key words: brushless excitation system rotating rectifier fault diagnosis support vector domain description unknown fault detection |
