Real-time detection of bearing faults through a hybrid WTMP analysis of frequency-related states

The paper’s primary focus is on the monitoring of vibration signals and introduces an innovative method for the detection of bearing faults in electric machines WTMP. While conventional techniques based on vibration signals are popular in identifying the characteristic frequencies associated with faults, they encounter difficulties when dealing with signals that vary over time (non-stationary signals). To tackle this challenge, the proposed approach combines three distinct techniques: Continuous Wavelet Transform (CWT), Wavelet Packet Transform (WPT), and Matrix Pencil (MP). This hybrid method has several objectives: It aims to reconstruct signals that exhibit non-stationary behavior, emphasize the frequency related to bearing faults, and ultimately enhance the accuracy of fault detection. By harnessing the unique strengths of CWT, WPT, and MP, this proposed approach significantly improves the effectiveness of condition monitoring in electric machines, particularly in the context of detecting bearing faults. To validate the method’s performance, an experimental setup has been established. This setup allows for testing under various load conditions, offering a comprehensive assessment of the capabilities of the proposed technique. This rigorous experimental testing ensures the method's reliability and practical applicability in real-world scenarios.