Speed estimation in planetary gearboxes: A method for reducing impulsive noise. (October 2021)

• Record Type: Journal Article
• Authors: Peng, Dikang; Smith, Wade A.; Randall, Robert B.; Peng, Zhongxiao; Mechefske, Chris K.
• Is Part Of: Mechanical systems and signal processing. Volume 159(2021)
• Publication Date: 2021-10
• Subjects: Structural dynamics -- Periodicals; Vibration -- Periodicals; Constructions -- Dynamique -- Périodiques; Vibration -- Périodiques; Structural dynamics; Vibration; Periodicals; 621
• DOI: 10.1016/j.ymssp.2021.107786
• Languages: English
• ISSNs: 0888-3270

Abstract

Highlights: Source of impulsive noise in IAS estimates of planetary gearboxes explained. Systematic errors identified in IAS estimates of sequentially phased gearboxes. Sequentially phased gearboxes also more prone to random noise in IAS estimates. Removal of noise via compensation from other harmonics proposed. Method validated using data from laboratory test rig and industrial gearbox. Abstract: Wind-turbine gearboxes mostly work under varying speed conditions. With the increased need for cost-saving and installation difficulties in many applications, vibration-based speed estimation approaches have attracted more attention recently. However, the authors have observed that instantaneous speed or phase estimates from planetary gearbox vibration signals are often more vulnerable to impulsive noise than those IAS estimated by shaft rotation The reason why it occurs, especially frequently in planetary gearboxes, is still not fully understood. Therefore, the generation of this impulsive noise and how it affects the order tracking result is fully investigated in this study. A signal processing framework is proposed to minimize this noise to reduce the overfitting issue and increase the accuracy of the vibration-based speed estimation result. The reason why the noise occurs was validated using signal model and two datasets. The method to improve the accuracy of speed estimation was compared with existing speed estimation methods using both industry and laboratory data to show its benefits compared with existing phase-demodulation based speed estimation methods. The technique is most suitable for speed estimation in cases of mixed amplitude and phase modulation.