| Autor: |
Glukhova, N. V., Kovalenko, I. V. |
| Předmět: |
|
| Zdroj: |
Scientific Bulletin of National Mining University; 2012, Issue 1, p121-126, 6p |
| Abstrakt: |
Background. The most common scheme for investigation of properties the some measuring transducers or measuring channels in general is the spectral analysis based on Fourier transform. This mathematical apparatus is inconvenient for study of non-stationary and pulsed signals. For study of dynamic properties of measuring transducers it is urgent to use alternative mathematical tools such as wavelet analysis and the window Fourier transform in addition to classic methods. Methods. For study of information-processing signals it is necessary to input data into computer. Data capture microsystem m-DAQ with USB interface was used for this purpose. It was analyzed various frequency-modulated signals in the data collection channel and frequencymodulated signal with peaks and their power spectrums. Spectrograms of window Fourier transform for each signal were done. The periodic signal, non-stationary aperiodic signal and the signal with breaks were analyzed by means of wavelet transform. Result. It is determined that signal frequency sequence change and the presence of sharp pulsed peaks does not lead to changes in the power spectrum for the fast Fourier transform. For non-stationary aperiodic signal a wavelet spectrum can be made for both small time values and high-frequency noise components in different time length. Wavelet spectrum gives the opportunity observe local features of the signal with breaks. Conclusion. Research results showed that there is no single universal approach for any case. Even a new and powerful tool as wavelet analysis cannot differ the effects of amplitude and frequency modulation. That is why preprocessing is required for determination of type of the signal. [ABSTRACT FROM AUTHOR] |
| Databáze: |
Complementary Index |
| Externí odkaz: |
|
