Local Embedded Sensing-Based Gear Fault Diagnosis Under Speed Varying Condition

The external signals of the gearbox sensed by the sensors are always affected by structural noise interference and path-transfer attenuation, and this would cause the fault characteristics cannot be effectively identified. Especially, those desired features would be further interfered by variable speed conditions. Focusing on these issues, this study introduced a diagnostic enhancement method via local embedded sensing (LES) system, where a vibration acceleration sensor is directly embedded in the rotational position to sense the internal excitation and response. Different from the conventional approach of external sensing system, two physical information, including vibration and speed, can be simultaneously sensed. Thereupon, a fault diagnosis method based on order analysis is later processed for the local embedded signal of gear. This diagnostic enhancement method for variable speed conditions mainly contains three steps. First, according to the characteristics of local embedded sensing signals, low-pass and high-pass digital filters are used to separate the low-frequency speed component and the high-frequency component. Then, the Hilbert transform is performed on the low-frequency speed component to get the phase curve, and then the least squares method is used to fit the phase curve to obtain the fitting phase curve. Finally, the angular domain of the high-frequency vibration signal is resampled by fitting the phase curve, and then the envelope order of the resampled angular domain signal is analyzed to extract the fault characteristics of the signal. The effectiveness of the proposed method has been verified by simulation analysis of local embedded signal of the gear.