De-noising for vibration signal of rotating machinery based on generalized mathematical morphological filter
SHEN Lu ZHOU Xiaojun ZHANG Wenbin ZHANG Zhigang
(Zhejiang Province Key Laboratory of Advanced Manufacturing Technology,Institute of Modern Manufacturing Engineering, Zhejiang University, Hangzhou 310027,China)
Abstract：Statistic bias occurs because same structure element was used in traditional morphological filter. Based on basic morphological transforms and their combination modes, a generalized morphological filter cascaded by different structure element was proposed to reduce noise of rotating machinery vibration signal. De-noising performance of generalized morphological filter on impulse noise and random noise were investigated. By comparing performance of generalized morphological filter and traditional morphological filter on simulation and practical signal, it is demonstrated that generalized morphology filter could reduce noise more effectively than traditional morphology filter. Generalized morphological filter algorithm includes addition, subtraction, maximum and minimum operations, and does not involve multiplication and division. Therefore, it has high calculation speed and better application value.
2009, Vol. 28 
