机构地区: 中南大学信息科学与工程学院
出 处: 《中南大学学报(自然科学版)》 2006年第2期336-340,共5页
摘 要: 提出一种基于Morlet组合小波的梳状滤波和包络检波方法。梳状滤波在间隔的几个故障调制频率下进行窄带滤波,抑制噪声的能力强;Morlet组合小波构造简单,计算量小,仅改变组合小波的部分参数就可获得良好的滤波效果,方便灵活、适应能力强;Morlet组合小波的实部和虚部分别相当于一个零相移滤波器和一个90°相移滤波器,信号经小波变换后直接满足包络检波条件。工程应用结果表明:该方法对机械振动信号的滤波和包络检波一步到位,与故障特征频率相对应的包络谱谱峰清晰可见,适用于旋转式或往复式运动机械系统的故障诊断。 Based on combined Morlet wavelets, a comb-filter and envelope-demodulation method was presented. Comb-filter belongs to narrowband-pass filters which only pass fault modulating signal. It is more useful to suppress the noise than a broadband filter. And the calculation of wavelet transform is simple. By adjusting the parameters of wavelet, an ideal filter can be attained which can adapt to diverse detecting object. As real parts and imaginary parts of Morlet combined wavelet are 0° phase-shifting filter and 90° phase-shifting filter, the fault characteristic information can be demodulated directly from the output signal of wavelet transform. The feasibility and validity of the method have been validated by the engineering application.