2022

abstract

english

EN The paper presents research on the possibility of using the combination of short-time Fourier transform (STFT) and cepstrum assumptions for acoustic signal analysis in the aspect of rail vehicle drive components diagnostics. A passive experiment was conducted in the form of a pass-by test of the sound generated during the passage of a Solaris S105p tram. In cooperation with the operator, on the basis of a service inspection after the test, the acoustic signals recorded for the passage of undamaged trams and those with damaged traction gearboxes with a similar passing speed during the test were selected for further comparative analysis. As a result of a combination of STFT and Cepstrum assumptions, three short-time cepstrum (STC) methods were developed and described, using complex, real, and power cepstrums. This resulted in a time-quefrency domain representation of the signal, which is an alternative to the time-frequency domain representation. Damage to the gears in the traction gearbox was found to result in the appearance of quefrency components corresponding to the characteristic frequencies of the mesh with increased amplitude of the sound pressure. It has been shown that STC based on the real and power cepstrums allows, for the case studied, to determine the occurrence of damage more precisely, compared to STC based on the complex cepstrum. Furthermore, it was found that the STC based on the cepstrum allows for a more accurate analysis of the signal components compared to the STFT with the same window parameters. The results were validated on an independent pair of trams of the same model with identical damage, showing that the described methods can be successfully repeated for different vehicles.

26 - 26

7th International Congress on Technical Diagnostics, September 14-16, 2022, Radom, Poland : book of abstracts

7th International Congress on Technical Diagnostics, ICTD 2022, 14-16.09.2022, Radom, Poland