| In petrochemical industry, one of the commonly used equipment for the production sector is reciprocating compressors, equipment failure has a serious securiyt risk, in addition to affect production efficiency and reduce economic efficiency.Therefore, when a compressor failure occurs,a timely and effective fault problem solving is particularly important. However, due to the structure of reciprocating compressor is relatively complex, the vibration generating the excitation source, the fault diagnosis of reciprocating compressor are of great difficulty.In reciprocating compressor,Valve plates which are key components are the most prone to failure, and in the vibration signals are rich of frequency modulation signals hiddened. Because of traditional time-frequency analysis method is difficult to effectively extract the valve FM fault signal, this paper propose a method based on Fractional Fourier transform of the reciprocating compressor fault diagnosis methods research.Firstly, through the study of traditional time-frequency method, found that although the Wigner distribution has certain energy is focused, but, in multi-component signal analysis, is difficult to solve the problem of cross terms. Then, Fractional Fourier transform is applied to the simulation analysis on the standard sine signal, single component LFM signal, multi-component LFM signals and LFM signal containing noise, to verify the exactness and effectiveness of the discrete Fractional Fourier transform. The analysis results show that the Fractional Fourier transform can not only effectively extract the frequency characteristics of LFM signal with high time-frequency focusing, but also can effectively suppress noise interference. Finally, it is applied to fault diagnosis of valve plates, and experimental results show that different valve faults correspond to different frequency characteristics, so as to realize the quantitative diagnosis of reciprocating compressor valve failure. |
PDF Full Text Request