基于RELAX频谱分析方法的鼠笼式异步电动机转子故障诊断

Fault Diagnosis Way Based on RELAX Spectrum Analysis in Squirrel Cage Induction Motors

在基于电流信号分析的异步电机故障诊断方法中,故障特征成分(尤其是断条特征)往往容易被基波分量和噪声信号中所淹没。因此,有效地克服基波和噪声的影响是诊断过程的关键。RELAX是一种对加性噪声以及系统误差假设可松弛的算法,提出了一种基于该算法的鼠笼式异步电动机转子故障监测与诊断方法。文中从理论上推导了鼠笼式异步电动机转子故障时的电流信号模型,将RELAX算法应用于电源基波特征参数提取,并从噪声和杂波的连续谱中估计出特定故障的离散谱参数,从而达到消除电源和噪声影响、突出故障特征的目的。通过对样机实测信号进行了分析处理,实验结果验证了RELAX算法的有效性和优越性。

Among the methods of fault diagnosis based on the current analysis for Squirrel Cage Induction Motors （SCIM）, the faults＇ characteristic components are often submerged by fundamental component and noises, especially for the broken rotor bars, and how to effectively eliminate the influences of the fundamental component and noises is essential for the diagnosis system. The relax algorithms is a kind of relaxant data processing algorithms for additive noises and system errors. A new approach based on relax algorithms for fault diagnosis and monitoring in SCIM is proposed. First, the current signal model of SCIM under fault condition is deduced theoretically. Aimed at the problem that the characteristic components being submerged by the fundamental one in the spectrum of the single phase current, the relax algorithms is used to extract the characteristic components for the fundamental one of power, and estimate the discrete-time spectrum parameter of particular fault from the continuous spectrum of the yawp and mixed waves, which can eliminate the effective of power and yawp and highlight the fault characteristic components. At last experimental results have demonstrated the effectiveness and advantage of the proposed technique.