基于高分辨率谱估计的早期转子断条故障诊断

发布时间：2018-06-01 10:55

本文选题：鼠笼电机 + Hilbert变换　；参考：《仪器仪表学报》2017年02期

【摘要】：以快速傅里叶变换(FFT)为基础的电机电流信号特征分析(MCSA)具有频率分辨率低的固有缺陷,从而严重影响了鼠笼电机早期转子断条故障的诊断性能。为解决这一问题,提出基于高分辨率谱估计的早期转子断条故障诊断方法。首先利用Hilbert变换和离散小波变换对单相定子电流信号预处理,然后采用扩展Prony算法对预处理后的信号进行定性/定量分析。运用该方法对不同故障严重程度、不同负载条件下的3 k W电机稳态定子电流信号进行分析,并与FFT分析结果做对比。实验结果表明,即使在短时数据条件下所提方法仍然能够准确诊断出早期转子断条故障,验证了该方法的有效性和优越性。
[Abstract]:The motor current signal characteristic analysis (MCSA) based on fast Fourier transform (FFT) has the inherent defect of low frequency resolution, which seriously affects the diagnosis performance of the early rotor break fault of the squirrel cage motor. In order to solve this problem, an early fault diagnosis method based on high resolution spectrometer is proposed. First, the Hi is used for the fault diagnosis of the rotor broken bar. Lbert transform and discrete wavelet transform preprocess the single-phase stator current signal, and then use the extended Prony algorithm to analyze the pre processed signal qualitatively / quantitatively. The method is used to analyze the steady-state stator current signal of 3 K W motor with different fault severity and different load conditions, and compare with the FFT analysis results. The experimental results show that the proposed method can accurately diagnose the fault of the early rotor bar even in the short time data, which proves the effectiveness and superiority of the method.
【作者单位】：沈阳工业大学信息科学与工程学院;辽宁科技大学应用技术学院;辽宁科技大学电子与信息工程学院;中国科学院沈阳自动化研究所;
【基金】：中国科学院重点部署项目(KGZD-EW-302) 辽宁省科学技术计划(2015020140)项目资助
【分类号】：TM307.1

【相似文献】