基于励磁电流时序特征的核电多相旋转整流器二极管开路故障诊断  

Diode Open-circuit Fault Diagnosis of Nuclear Multi-phase Rotating Rectifier Based on Timing Features of Field Current

作　　者：胡启昊 郝亮亮[1] 周艳真 蔡宇昂 梁郑秋 段贤稳 王光[4] HU Qihao;HAO Liangliang;ZHOU Yanzhen;CAI Yu’ang;LIANG Zhengqiu;DUAN Xianwen;WANG Guang(School of Electrical Engineering,Beijing Jiaotong University,Haidian District,Beijing 100044,China;Department of Electrical Engineering,Tsinghua University,Haidian District,Beijing 100084,China;China Nuclear Power Operations Co.,Ltd.,Shenzhen 518172,Guangdong Province,China;NR Electric Co.,Ltd.,Nanjing 211102,Jiangsu Province,China)

机构地区：[1]北京交通大学电气工程学院,北京市海淀区100044 [2]清华大学电机工程与应用电子技术系,北京市海淀区100084 [3]中广核核电运营有限公司,广东省深圳市518172 [4]南京南瑞继保电气有限公司,江苏省南京市211102

出　　处：《中国电机工程学报》2023年第20期8082-8093,共12页Proceedings of the CSEE

摘　　要：实现多相旋转整流器的故障诊断对保障核电系统安全稳定运行至关重要。然而,现有方法难以准确识别旋转整流器的二极管开路故障模式。为此,该文首先分析将励磁电流作为诊断变量的理论依据;然后,引入形状子特征(Shapelet)变换算法,充分利用其在可解释性和局部差异识别准确性的优势,提取和挖掘不同故障模式下励磁电流最具有代表性和辨识度的特征子序列;最后,结合机器学习分类器得到故障模式分类结果,形成完整的故障诊断方案。通过11相无刷励磁系统仿真模型和动模样机实验进行验证。结果表明,所提方法能够直观并精细表征不同类别故障电流的局部特征,变换后不同故障数据分布的差异性明显提升,仅需结合简单分类器即可实现快速、准确的故障诊断,具有实际应用潜力。It is of great significance to realize the fault diagnosis of multi-phase rotating rectifier for ensuring the safe and stable operation of nuclear power system.However,it is almost impossible to accurately identify the diode open-circuit fault modes of rotating rectifier by existing methods.First,the theoretical basis of taking field current as diagnostic variable is analyzed.Then,the Shapelet transform algorithm is introduced to extract and mine the most representative and identifiable sub-sequence of field current under different fault modes based on its advantages in interpretability and accuracy of local difference identification.Finally,combined with the machine learning classifier,the fault mode classification results are obtained and a complete fault diagnosis scheme is proposed.Through simulation model and model machine of 11-phase brushless excitation system to validate the proposed method,the results show that the method can directly and precisely characterize the local characteristics of different types of fault currents,and the difference of distribution of fault data is significantly improved after transformation.Simple classifiers can be combined to achieve fast and accurate fault diagnosis,which has practical application potential.

关 键 词：无刷励磁机 多相环形 旋转整流器 故障诊断 励磁电流 形状子特征变换 

分 类 号：TP183[自动化与计算机技术—控制理论与控制工程] TM461[自动化与计算机技术—控制科学与工程]