Gogolewski Damian, Zmarzły Paweł, Kozior Tomasz, Mathia Thomas G
Department of Mechanical Engineering and Metrology, Kielce University of Technology, al. Tysiąclecia Państwa Polskiego 7, 25-314 Kielce, Poland.
Laboratoire de Tribologie et Dynamique des Systemes (LTDS), Ecole Centrale de Lyon, Centre National de la Recherche Scientifique, 69134 Lyon, France.
Materials (Basel). 2023 Jan 31;16(3):1228. doi: 10.3390/ma16031228.
The article presents research results related to assessing the possibilities of applying modern filtration methods to diagnosing measurement signals. The Fourier transformation does not always provide full information about the signal. It is, therefore, appropriate to complement the methodology with a modern multiscale method: the wavelet transformation. A hybrid combination of two algorithms results in revealing additional signal components, which are invisible in the spectrum in the case of using only the harmonic analysis. The tests performed using both simulated signals and the measured roundness profiles of rollers in rolling bearings proved the advantages of using a complex approach. A combination of the Fourier and wavelet transformations resulted in the possibility to identify the components of the signal, which directly translates into better diagnostics. The tests fill a research gap in terms of complex diagnostics and assessment of profiles, which is very important from the standpoint of the precision industry.
本文介绍了有关评估应用现代滤波方法诊断测量信号可能性的研究结果。傅里叶变换并不总是能提供关于信号的完整信息。因此，用现代多尺度方法——小波变换来补充该方法是合适的。两种算法的混合组合能够揭示出额外的信号成分，而这些成分在仅使用谐波分析时在频谱中是不可见的。使用模拟信号以及滚动轴承中滚子的实测圆度轮廓进行的测试证明了采用复杂方法的优势。傅里叶变换和小波变换的结合使得识别信号成分成为可能，这直接转化为更好的诊断效果。这些测试填补了复杂诊断和轮廓评估方面的研究空白，从精密工业的角度来看这非常重要。
