Fault Detection of Water Hydraulic Motor by Demodulated Vibration Signal Analysis with the Hilbert Transform and Genetic Algorithm

The vibration signals analysis from the water hydraulic motor is the most modern technique for fault detection of water hydraulic motor. The machine's residual life is estimated by the detection of the fault presence and type at an earlier stage for the adequate plan of maintenance. The most important components in the impulsive vibration signal spectra are the rotational frequency and its harmonics as well as sidebands due to modulation phenomena. This paper studied the fault detection of fluid machines by the demodulated frequency components of the periodic impulse vibration signal. We apply the Hilbert transform to demodulate the characteristic envelope of the periodic impulsive signal to show the fundamental frequencies. The adaptive spectrogram is optimized to get the optimal parameters to show the characteristic frequencies of the Hilbert transform-based envelope of the vibration signals. The results show that it is applicable and effective to fault detection of water hydraulic motor by analyzing the demodulated frequency components of the periodic impulsive signal.