Initialization parameters guided adaptive chirp mode decomposition for damage characteristic extraction
Bearings are critical mechanical components in industrial equipment, and damage detection is of great significance for effectively preventing prolonged downtime and catastrophic accidents. Nevertheless, the extraction of weak damage characteristics at the early stage of bearing degradation remains challenging under harsh operating conditions. To address this issue, an initialization parameters guided adaptive chirp mode decomposition (IPGACMD) method is proposed. First, the envelope spectrum crest factor diagram (ESCFgram) is developed to guide the selection of the optimal frequency band, and a bandwidth empirical expression is further introduced to guide penalty parameter selection. IPGACMD is constructed by integrating these two schemes, which can extract bearing damage characteristics without requiring any prior knowledge or subjective experience. The feasibility and superiority of IPGACMD are verified using experimental bearing damage signals. Comparative analysis results demonstrate that IPGACMD outperforms other methods in bearing damage characteristic extraction. Furthermore, IPGACMD is also applied to bearing condition assessment, achieving accurate extraction of degradation characteristics. Combined with the Chebyshev inequality threshold strategy, it provides early warning of the initial degradation moment of bearings.
Authors
- Dao Gong (ORCID: https://orcid.org/0000-0001-6678-9399)
- Kai Zhou (ORCID: https://orcid.org/0000-0001-6689-2193)
- Xiaolong Wang (ORCID: https://orcid.org/0000-0002-5061-2529)
- Hui Zheng (ORCID: https://orcid.org/0000-0002-7217-5032)
- Jinsong Zhou (ORCID: https://orcid.org/0000-0001-7688-4373)
- Hanwei Jin (ORCID: https://orcid.org/0009-0009-2891-0245)
Institutions
- Tongji University (CN)
- North China Electric Power University (CN)
- Fuyao University of Science and Technology (CN)
Publication Details
- Journal
- Journal of Vibration and Control
- Published
- 2026-11-12
- DOI
- https://doi.org/10.1177/10775463261500188
- Primary Topic
- Machine Fault Diagnosis Techniques
- Type
- article
- Field-Weighted Citation Impact
- 0.00
Funders
- Natural Science Foundation of Hebei Province