基于HHT谱的齿轮状态监测与识别的方法

Approach to Condition Monitoring and Identification of Gearbox Based on HHT Spectrum

由于Wigner-Ville分布(Wigner-Ville Distribution,WVD)存在干扰项,导致结果中出现虚假信号、有时频率分辨率不够高等缺点,提出了一种"信号提纯重构———HHT谱分析"进行机械设备状态监测与识别的方法。对比分析信号包括:谐波、调幅、调频的单个信号以及由这些信号叠加的混合信号,表明HHT谱比WVD有更高的频率分辨率,并且它克服了后者的干扰项的影响,但因HHT对噪声很敏感,文章提出先对被分析信号进行提纯重构再进行HHT谱分析。最后对实际的故障齿轮的信号用这种方法进行了分析,指出能为故障诊断和状态监测提供一种新的有效的方法。

As Wigner-Ville Distribution （WVD） has some disturbance term which will results in a false signal and not high frequency resolution at some time, a novel method on ＂ signal purification and reconstruction--Hilbert-Huang Transform （HHT） spectrum analyses＂ was proposed to carry on condition monitoring and state recognition of mechanical equipment. The compared signals included： single signal and composite signal of harmonic wave, amplitude modulation and frequency modulation. The study shows that： HHT Spectrum has higher frequency resolution than WVD, and it overcomes the latter defects. But the HHT spectrum is very sensitive to noise, so the analyzed signal should be extracted and remodeled, and then analysed by HHT spectrum. Real signals of fault gear were also analyzed by this method, and it indicated that the method can be a new and effective way for fault diagnosis and condition monitoring.