笼型异步电动机转子断条故障诊断方法

发布时间：2018-04-16 03:35

本文选题：笼型异步电动机 + 断条　；参考：《电机与控制学报》2015年06期

【摘要】：针对笼型异步电动机发生转子断条故障时,用于判定故障类型及其严重程度的定子电流信号中的边频信号容易被主频信号所淹没的问题,研究了一种基于Hilbert变换和支持向量机理论的笼型异步电动机断条故障诊断方法。首先进行了详细的理论推导,为该方法在断条故障诊断中的应用奠定了基础。然后设计并完成了一系列断条故障试验,取得了真实有效的故障数据。最后,将该方法应用于试验数据的分析与处理,结果表明Hilbert变换能有效提取到断条故障时定子电流信号中的故障特征量,而采用这些特征量训练得到的支持向量机分类模型则能在故障样本有限的前提下实现最优分类,将二者结合起来用于断条故障诊断的准确率高达98%。
[Abstract]:In view of the problem that the side frequency signal of stator current signal used to determine the type of fault and its severity is easily submerged by the main frequency signal when the rotor broken bar fault occurs in the cage asynchronous motor.A fault diagnosis method for cage asynchronous motor based on Hilbert transform and support vector machine theory is studied.Firstly, the detailed theoretical derivation is carried out, which lays a foundation for the application of this method in broken bar fault diagnosis.Then a series of broken bar fault tests are designed and completed, and real and effective fault data are obtained.Finally, the method is applied to the analysis and processing of test data. The results show that the Hilbert transform can effectively extract the fault characteristics of stator current signal when the bar is broken.The SVM classification model trained by these features can achieve the optimal classification under the premise of limited fault samples, and the accuracy of combining the two methods for fault diagnosis of broken bars is as high as 98%.
【作者单位】：太原理工大学煤矿装备与安全控制山西省重点实验室;山西昌生电磁线有限公司;
【基金】：高等学校博士学科点专项科研基金资助项目(20111402110010) 国家自然科学基金面上项目(51377113)
【分类号】：TM343

【参考文献】