Fault Features Diagnosis Method of Rolling Bearing via Optimized S Synchroextracting Transform
Huan Yang, Kun Zhang, Zuhua Jiang, Xiangfeng Zhang, Yonggang Xu
Abstract
Synchroextracting transform (SET) is a time-frequency (TF) post-processing algorithm that can significantly improve the readability and energy aggregation of TFR, but it has the drawbacks of uncertainty in window width selection and fixed TF resolution. Based on this, this paper proposes a new time-frequency analysis (TFA) method named Optimized S synchroextracting transform (OSSET). This algorithm is a TFA method with multi-time-frequency resolution and adaptive selection of optimal parameter set, which can not only overcome the uncertainty of the window width selection and the fixed TF resolution of short time Fourier transform, but also overcome the shortcoming of the local frequency dependence of S transform. Moreover, the improved synchronous extraction operator is used to further improve the energy concentration and enhance the TF readability. The analysis results of simulation signal show that OSSET can more clearly and accurately represent the time-varying characteristics of strong frequency modulation signal with high resolution, and has higher noise robustness than SET and other post-processing algorithms. Finally, the superiority and practicability of the proposed algorithm are further demonstrated in the applications of rolling bearing fault experimental data.