Fault Diagnosis Method for Rolling Bearings in High-Noise Environments Based on COA-FMD-ITEO

To address the challenges in extracting features from rolling bearing vibration signals in high- noise environments, a novel fault diagnosis method combining feature modal decomposition based on Cheetah Optimization Algorithm (COA-FMD) and Improved Teager Energy Operator (ITEO) is proposed. First, the Gini coefficient of square envelope spectrum (GISES) is utilized as the fitness function to adaptively optimize the key parameters of FMD through COA. Subsequently, the optimal modal components are selected from the decomposed modes by integrating the feature indicators of the envelope spectrum. Finally, the ITEO energy operator is employed to demodulate the selected modal components, using its effectiveness to enhance impact features to precisely identify the characteristic frequencies of bearing faults via the ITEO energy spectrum. Experimental findings indicate that the proposed methodology is effective in extracting fault signal characteristics in high-noise environments and accurately identifying the type of fault present.