A Signal Processing Algorithm Based Upon the Weighted Morphological Undecimated Wavelet Decomposition

In this paper, a novel method based on weighted morphological un-decimated wavelet decomposition is proposed. Firstly, the method is presented under the scheme of morphological un-decimated wavelet decomposition. It combines decomposition layers according to their feature contribution weights. The weighted result is taken as the basis of feature extraction. On this basement, the feature operator is built so as to choose the number of decomposition layers and the initial value structure elements. Finally, the validity of the method is testified by the real vibration signal of swash plate axial piston pump.