Fault Diagnosis Based on Feature Mode Decomposition of Whale Optimization Algorithm

For the increasingly complex problems of mechanical fault signals, a recently introduced adaptive signal processing technique called feature mode decomposition (FMD) has been effectively utilized to diagnose mechanical equipment faults. Since FMD requires prior knowledge, its parameters need to be input manually, and its parameter adaptability directly affects the decomposition performance of FMD. Based on this, this work suggests a method for choosing FMD parameters by the Whale optimization Algorithm (WOA) and uses it to identify mechanical equipment bearing faults. First, the WOA is utilized to search for the best combination of FMD’s mode number (K) and filter length (L) after utilizing the sample entropy as the fitness function. The modal components are then obtained through signal decomposition, and the mode component with the highest kurtosis index is chosen. To identify the issue, resonance demodulation analysis is carried out for the target mode, and fault characteristic data is collected. The research on the Xi’an Jiaotong University dataset and MFPT dataset has demonstrated that this technique, which is quicker and more accurate than variational mode decomposition (VMD), can extract fault signals, identify fault types, and be applied to bearing fault detection.