Research on the signal feature extraction method in Steel cord conveyor belt with metal magnetic memory testing

摘要

The detection accuracy level of the metal magnetic memory (MMM) technology is decided by the extraction of weak fault signal.In this paper,the fault characteristic signal of the MMM was extracted by the method of combining wavelet analysis and singular value decomposition (SVD).The fault signal was decomposed into different scale components with the wavelet analysis.And the SVD was applied to decompose characteristic matrix,which was composed with the different scale components,into different subspace.With the reconstruction of the singular vectors,the signal features were extracted effectively.The experiments show that the method is effective.