2025 International Conference on Equipment Intelligent Operation and Maintenance (ICEIOM2025)

243 / 2025-06-15 16:54:59

Optimization of early fault identification method for rotating machinery based on feature mode decomposition

Rotating machinery; Eigenmode decomposition; Early fault identification; Harris Hawks Optimization

Aiming at the influence of filter length L and the number of mode decompositions n on the decomposition effect of Feature Mode Decomposition (FMD), the Harris Hawks Optimization (HHO) algorithm is used to optimize the two preset parameters of FMD, and an early fault identification method is proposed with kurtosis, envelope spectrum peak factor, and Pearson correlation coefficient as the comprehensive objective functions. By comparing the comprehensive objective functions of various component signals with different preset parameters, selecting L and n corresponding to the minimum value as the final FMD parameters, and identifying the fault situation of rotating machinery by calculating the envelope spectrum characteristics of the components. After verification with the full life cycle data of bearings from XJTU-SY and diesel engine test bench data, this method has good early fault identification ability.