Answers are grounded in this paper and cite their sources.
To address the challenge of extracting compound fault features from rolling bearings under variable-speed and strong-noise conditions, a fault feature extraction method integrating the improved hippopotamus optimization algorithm (IHOA) and enhanced angle-domain feature mode decomposition (EAFMD) is proposed. First, cepstrum pre-whitening (CPW) is introduced to preprocess the raw vibro-acoustic signals, effectively suppressing rotational frequency harmonics and electromagnetic interference. Second, the enhanced angle-domain feature mode decomposition (EAFMD) is constructed by employing the average Hoyer index (AHI) as the objective function to successfully separate the compound fault modes from the intense noise background. Subsequently, to adaptively optimize the core parameters of EAFMD, an improved hippopotamus optimization algorithm (IHOA) is formulated by integrating Latin hypercube sampling (LHS) and progressive refractive mirror learning (PRML). This effectively overcomes the reliance of traditional AFMD on empirical parameter selection and ensures the robustness of the algorithm. Finally, an adaptive mode screening mechanism is designed to achieve the separation of optimal mode components for each fault type, thereby accomplishing the accurate extraction of compound fault features. Simulation and experimental results indicate that the proposed method can more accurately and comprehensively extract compound fault features under variable speed and strong noise conditions and exhibits favorable performance in terms of noise robustness and feature extraction stability.
No takes yet. Share an insight, caveat, or question.
Qu et al. (2026) studied this question.
Synapse has enriched 10 related papers. Consider them for comparative context: