Title | ||
---|---|---|
Computationally Efficient Multiwavelets Construction Method with New Signal-Dependent-Multiplicity Determination Scheme |
Abstract | ||
---|---|---|
Multiwavelets, which possess more favorable mathematical properties than scalar-wavelets, are deemed a promising means for non-stationary signal analysis. Due to high computational-complexity incurred in the construction of multiwavelets with arbitrary multiplicities (MWAM), the multiplicity of the multiwavelets currently utilized in practice is usually restricted to two, which limits the signal-processing performance. In this paper, a new fast MWAM construction method based on the finite-element analysis is introduced. The four orthogonality equations for solving the filter-coefficient matrices in the construction of multiwavelet functions are simplified to two equations mathematically, so the computational-complexity of the MWAM construction is significantly reduced. Meanwhile, a new optimal multiplicity selection strategy is also proposed based on the comparison between the time-frequency bandwidth-products of the multiscaling function and the probed signal. Finally, our proposed new fast MWAM-construction method in tandem with the new multiplicity-determination scheme is tested by an application, namely mechanical-bearing fault-diagnosis. The corresponding performance is quite promising. Our proposed new MWAM construction method along with the associated multiplicity-determination scheme can be applicable for many time-frequency signal-approximation and feature-extraction applications. |
Year | DOI | Venue |
---|---|---|
2022 | 10.1007/s00034-022-02022-6 | Circuits, Systems, and Signal Processing |
Keywords | DocType | Volume |
Multiwavelets with arbitrary multiplicities (MWAM), Multiscaling function, Multiwavelet function, Multiwavelet construction, Multiplicity selection | Journal | 41 |
Issue | ISSN | Citations |
9 | 0278-081X | 0 |
PageRank | References | Authors |
0.34 | 10 | 5 |
Name | Order | Citations | PageRank |
---|---|---|---|
Xiao Yan | 1 | 7 | 4.88 |
Hsiao-chun Wu | 2 | 959 | 97.99 |
Wang, Qian | 3 | 29 | 16.68 |
Yin, Chunyu | 4 | 0 | 0.34 |
Li, Pengwei | 5 | 0 | 0.34 |