Encoder Signal Analysis and its Application in Gear Fault Detection

Gear fault detection based on encoder signals has attracted attention in recent years, in which it is important to capture the instantaneous angular speed (IAS) jitters caused by the gear faults. The central difference method (CDM) is widely used to calculate the IAS. However, due to the quantization error of the encoder and the measurement noise in practice, it is often hard to accurately estimate the jitters of the IAS caused by the gear fault using the CDM directly. To address this issue, the scheme of encoder signal reconstruction and the synchronous average merging is proposed to improve the estimation accuracy of the IAS. First, the encoder signal is reconstructed according to the square wavenumber of the encoder corresponding to each tooth of the gear. Then, the time-synchronous averaging (TSA) process is performed on the reconstructed signal. Third, the IAS signal is estimated by the CDM. Finally, the blind deconvolution based on the cyclostationarity maximization (CYCBD) method is used to process the IAS signal to enhance the jitters caused by the gear fault. The simulation and experimental results support the proposed method.