作者: ZhiBo Yang , XueFeng Chen , Maciej Radzienski , Pawel Kudela , Wieslaw Ostachowicz
DOI: 10.1007/S11431-016-9036-7
关键词: Artificial intelligence 、 Second-generation wavelet transform 、 Wavelet packet decomposition 、 Computer vision 、 Laser scanning vibrometry 、 Energy operator 、 Mathematics 、 Singularity 、 Wavelet 、 Vibration 、 Algorithm 、 Continuous wavelet transform
摘要: The continuous wavelet transform (CWT) is one of the crucial damage identification tools in vibration-based assessment. Because vanishing moment property, CWT method capable featuring singularity higher scales, and separating global trends noise progressively. In classical investigations about this issue, localization property usually deemed as most critical point. abundant information provided by scale-domain corresponding effectiveness are, however, neglected to some extent. Ultimately, neglect restricts sufficient application localization, especially noisy conditions. order address problem, correlation operator introduced into detection a post-processing. By means correlations among different proposed suppresses noise, cancels trends, intensifies features for various mode shapes. demonstrated numerically with emphasis on characterizing environments, where scale Teager-Kaiser energy taken benchmark comparison. Experimental validations are conducted based data from composite beam specimens measured scanning laser vibrometer. Numerical experimental validations/comparisons present that introduction effective