Answers are grounded in this paper and cite their sources.
Variational mode decomposition (VMD) is a signal processing technique used to analyze bearing faults by breaking down signals into simpler components. While effective, traditional VMD requires users to manually set key parameters—such as the initial center frequency <tex-math notation="LaTeX">ω ¹</tex-math>, the penalty parameter <tex-math notation="LaTeX">α</tex-math>, and the number of modes <tex-math notation="LaTeX">K</tex-math>—which, if improperly selected, can obscure fault features or reduce analytical clarity. To overcome these limitations, we propose a parametric self-contained VMD (PSVMD) method. Specifically, <tex-math notation="LaTeX">ω ¹</tex-math> is automatically determined based on the narrowband features of the signal, removing the need for manual selection. Furthermore, adaptive strategies are introduced to estimate both <tex-math notation="LaTeX">α</tex-math> and <tex-math notation="LaTeX">K</tex-math> based on the intrinsic properties of the decomposed modes, thereby enhancing accuracy and robustness. Building on PSVMD, we also propose a novel correlation-kurtosis index to effectively distinguish fault-related features from noise. Simulation results and real-world bearing fault data validate the effectiveness and superiority of the proposed method.
No takes yet. Share an insight, caveat, or question.
Lang et al. (2025) studied this question.
Synapse has enriched 10 related papers. Consider them for comparative context: