Original Article

Order component extraction technology for predictive maintenance system in rotary machine

¹ School of Electrical Engineering, ShangHai DianJi University, ShangHai, China
² School of Electrical Engineering, ShangHai Maritime University, ShangHai, China
³ General Manager Department, Shang hai Tengtec Electronics CO, Ltd, Shang Hai, China
⁴ General Manager Department, Li Yang HongDa Motor CO, Ltd, Chang Zhou, China

^* e-mail: luyan@sdju.edu.cn

Received: 13 December 2024
Accepted: 19 February 2025

Abstract

The most obvious difference between the recent smart factory and the traditional automation factory is that the techniques about Predictive Maintenance (PdM) are introduced, PdM is also one of the key enabling technologies in Industry 4.0. In general, the smart factory that employs PdM intelligently ensures efficient and reliable industrial operations. The intelligent maintenance and fault diagnosis of rotating machinery, a core component of smart factories, is crucial. Due to the large speed fluctuation of manufacturing equipment in smart factory, its condition signal often presents multi-component property combination with fast-varying instantaneous frequency. However not much has been done in terms of PdM for smart factory and very few works tries to deal with time-varying multiple components extraction. Different failures for smart factory are attributable to the lack of research on PdM under large speed fluctuation. This work details a an order component extraction model according to Synchronous Extraction Transform (SET) combination with Vold-Kalman Filtering (VKF), The model extracts instantaneous frequency based on the time-frequency distribution, effectively avoiding the problem of spectral blurring. Additionally, by combining VKF technology, it accurately extracts the order components of condition signal. Finally, this paper develops an order component extraction system, it mainly consists of a signal acquisition module, and data processing module with good application prospect and promotion value in smart factory.