电动机转子故障在线检测误判率高的原因及其改进方法

Reasons of high misdiagnosis rate of online motor rotor fault detection and its improvement

长期以来,对异步电动机稳态运行下转子断条故障在线检测的准确度不高。虽然转子断条故障的特征频率分量与工频分量非常接近,且幅值也远远小于工频分量的幅值是影响断条故障检测的一个重要因素,但也还存在其他各种原因,一些检测手段或分析方法本身也存在各种缺陷。通过对造成电动机断条故障检测准确度不高的各种原因进行深入的分析,针对这些原因提出了具体的改进方法,并提出了一种基于频距比较的故障检测新方法。实例分析表明,提出的改进措施及检测方法,能将断条特征频率分量与其他频率分量有效地区分开,降低了转子断条故障的误判率,从而大大提高了其故障检测的准确度。

The accuracy of broken rotor detection for the motors under steady operation is not high for long time .An important reason is that the CFCFR(Characteristic Frequency Component of Faulted Rotor)is very close to the power frequency component but far lower in its amplitude.How-ever,there are still other reasons,such as the defects in some detection or analysis methods.The reasons of low detection accuracy are analyzed in detail and the improved methods are proposed.A fault detection approach based on frequency distance comparison is presented.It is proved by actual examples that the improved methods and the proposed approach can very effectively distin-guish the CFCFR from other frequency components,which reduces the misdiagnosis rate of rotor fault detection and enhances the fault detection accuracy greatly.