断层(地质)
小波
方位(导航)
信号(编程语言)
振动
模式识别(心理学)
特征提取
频域
算法
小波变换
计算机科学
工程类
人工智能
声学
计算机视觉
物理
地质学
地震学
程序设计语言
作者
Yanjiang Yu,Xuezhi Zhao
标识
DOI:10.1016/j.ymssp.2024.111118
摘要
The flexible thin-wall bearing, characterized by different kinematic attributes and fault characteristic frequency in comparison to rolling bearings, introduces a great challenge in the domain of fault diagnosis. Although wavelet transform is capable of describing the time–frequency characteristics of signals, the accuracy of its results is highly dependent on the wavelet parameters. This paper proposes a novel approach by introducing the concepts of energy concentration and reconstruction precision to establish an objective function for parameter selection. In addition, the African vultures optimization algorithm is introduced in this process. Moreover, a fault feature extraction and reconstruction method guided by the correlated Gini index is designed to capture the fault-related frequency slice and reconstruct the fault-related information. The experiment results demonstrate the proposed parameter selection method enables time–frequency analysis techniques to better represent the vibration characteristics of signals. Furthermore, the proposed extraction and reconstruction method successfully separates fault characteristic impulses hidden in the original signal, providing evidence for the time-varying nature of fault characteristic frequencies in flexible thin-wall bearing.
科研通智能强力驱动
Strongly Powered by AbleSci AI