Bearing Fault Diagnosis Based on CEEMDAN and CMDE Using PSO-ELM

Fault diagnosis is crucial for ensuring the stable operation of key components such as rotating machinery and rolling bearings in industrial equipment maintenance. To enhance the accuracy of fault diagnosis, this paper proposes a novel method that combines Complete Ensemble Empirical Mode Decomposition with Adaptive Noise (CEEMDAN) and Composite Multiscale Dispersion Entropy (CMDE). The method analyzes the vibration signal characteristics under different bearing conditions and performs classification using a Particle Swarm Optimization-based Extreme Learning Machine (PSO-ELM). First, CEEMDAN is employed to decompose the vibration signals and extract a series of Intrinsic Mode Functions (IMFs). Then, CMDE is used to extract feature vectors from these IMFs. Finally, PSO-ELM is utilized for fault classification. This study evaluates the impact of the optimized approach on diagnostic performance to demonstrate its effectiveness. Experiments were conducted using the Case Western Reserve University (CWRU) bearing dataset, and the results showed that the proposed method achieved a classification accuracy of 97.5%, demonstrating superior diagnostic performance compared to previous methods.