摘要
通过对发电机转子线圈的电磁特性分析,得出转子绕组主磁场变化特性,以及转子匝间短路在定子绕组并联支路出现的电势差和环流。采用频谱特征向量作为学习样本,建立径向基神经网络的频谱特征和故障类型的映射关系。应用表明,采用该方法能有效地诊断出发电机转子系统的故障。
Through the analysis of the electromagnetic properties for the rotor coil of the generator,the main magnetic field variation characteristics for the rotor winding and the short circuit fault between rotor inter-turns causing the potential difference and circulation between the parallel branches was obtained.The spectral feature vector was treated as learning sample,the Radial Basis Function neural network can reflect the mapping relations between spectral features and fault types.The practical application showed that the spectral analysis method and the RBF neural network can effectively improve the diagnostic accuracy and efficiency.
出处
《电子测量技术》
2011年第2期80-83,共4页
Electronic Measurement Technology
基金
上海市教委重点学科建设项目资助基金(J51801)
关键词
发电机
监测与诊断
频谱分析
RBF神经网络
generator
detection and diagnosis
spectrum analysis
RBF neural network
