A Novel Adaptive Multiperiod Differential Means for Rotating Machinery Condition Monitoring With Optical Encoder Signal Under Low- and Variable-Speed Conditions

Condition monitoring of rotating machinery under low- and variable-speed conditions is a hot and difficult issue. Compared with vibration signals, optical encoder signals have the advantages of no external installation, direct correlation with power, and no time-varying transmission path, which has been a widespread concern in recent years. However, the encoder installation errors, sampling errors, speed trend components, and so on greatly interfere with the fault detection of rotating machinery. To address this issue, based on the accumulative characteristic of the multiperiod mean, the differential tool without amplitude interference, and the energy ratio diagnostic feature (ERDF) indicator, an optical encoder-based feature enhancement scheme is proposed, the so-called adaptive multiperiod differential means (AMPDM). Besides, the instantaneous angular speed (IAS)-based average downsampling estimation scheme based on the optical encoder estimation characteristics is introduced for suppressing high-order cyclostationary components. Furthermore, the simulation models of bearings fault and gears fault are proposed. Finally, it is verified from aspects of demodulation analysis and shock enhancement that conventional methods may not be able to extract fault features from encoder signals, and simulation and experiment results show that the AMPDM algorithm can enhance the features related to faulty gears and bearings under low- and variable-speed conditions.