Intelligent Diagnosis of Rolling Bearings Based on Parameter-Optimized Variational Mode Decomposition and BiLSTM

Rolling bearings are critical components in modern mechanical systems, and their failures can result in significant downtime and safety risks. This study proposes a diagnostic framework for bearing fault detection by integrating a Black-winged Kite Algorithm (BKA)-optimized Variational Mode Decomposition (VMD) with a Bidirectional Long Short-Term Memory network (BiLSTM). In the proposed approach, BKA is employed to fine-tune key VMD parameters, enabling the effective decomposition of vibration signals into distinct intrinsic mode functions and facilitating the extraction of fault-relevant features. The BiLSTM model subsequently captures temporal dependencies for robust classification. Experimental validation on the CWRU dataset demonstrates that the proposed framework outperforms a baseline VMD-BiLSTM method in terms of classification accuracy and stability, underscoring its potential to improve predictive maintenance practices in industrial applications.