We found a match
Your institution may have access to this item. Find your institution then sign in to continue.
- Title
Fault diagnosis of rotating machines based on modified hierarchical fluctuation dispersion entropy and multi-cluster feature selection.
- Authors
Li, Baoyue; Yu, Yonghua; Hu, Jia; Cao, Bingxin; Yao, Yangfeng; Xu, Defeng
- Abstract
The vibration signal of rotating mechanical equipment contains a large amount of information that can be used for the fault diagnosis of rotating mechanical equipment. However, the vibration information is distributed in multiple dimensions, and a single-scale analysis cannot effectively reflect its damage characteristics, reducing the accuracy of fault diagnosis. Accordingly, an improved hierarchical fluctuation dispersion entropy (MHFDE) method based on the improved hierarchical processing is proposed. MHFDE can simultaneously mine low- and high-frequency features in the time series, avoiding information omission. Comparison results of the simulated signals show that the proposed method has the advantages of high stability and accurate measurement of complexity. In combination with the multi-cluster feature selection (MCFS) and kernel limit learning machine (KELM) optimized by whale optimization algorithm (WOA), a rotating machinery damage recognition method based on MHFDE-MCFS and WOA-KEM was proposed. Three sets of typical rotating machinery datasets are used to verify the effectiveness of the proposed method. The results show that this method can not only accurately and stably identify the damage types of the three selected machinery but also have a higher accuracy of damage recognition compared with the existing feature extraction methods.
- Subjects
FAULT diagnosis; ROTATING machinery; METAHEURISTIC algorithms; ROLLER bearings; ENTROPY; MACHINE learning; FEATURE selection; FEATURE extraction
- Publication
Journal of Mechanical Science & Technology, 2023, Vol 37, Issue 12, p6343
- ISSN
1738-494X
- Publication type
Article
- DOI
10.1007/s12206-023-1110-5