Denoising of data recorded in a multichannel way with use of a wavelet transform
Referat wygłoszony
| Status: | |
| Autorzy: | Jedliński Łukasz |
| Rok wydania: | 2011 |
| Język: | angielski |
| Źródło: | 10th International Technical Systems Degradation Conference, 27-30.04.2011, Liptovský Mikuláš, Slovakia |
| Państwo wystąpienia: | POLSKA |
| Efekt badań statutowych | NIE |
| Abstrakty: | angielski |
| In order to gather information about a phenomenon or an object one is interested in the most frequent way is to record selected measurement signals obtained through sensors. Unfortunately, apart from desirable information, the received signals may also contain noise caused by, among other things, measurement path properties and processes accompanying the object operation. In many cases one needs to measure the same value in different spots of the object and/or different directions. Therefore devices to improve the ratio of signal to noise of the signals recorded in a multichannel way are required. A wavelet transform is a relatively new method of data processing which has been used in different fields like technology and physics. When talking about signals it can be used for denoising, compression, or detecting trend or signal discontinuity. In this paper a wavelet transform was used for denoising of the signal oscillation recorded from two triaxial sensors. The object of the research was a spiral bevel gear. Denoising of the signals was aimed at improving the effectiveness of diagnosing a gear damage. |
