中国机械工程 ›› 2026, Vol. 37 ›› Issue (9): 2133-2140.DOI: 10.3969/j.issn.1004-132X.2026.09.006

• 智能诊断与运维专栏 • 上一篇    

基于复合多尺度加权反向模糊散布熵和增强t分布随机邻域嵌入的滚动轴承退化评估

王晓龙(), 伏元雷, 罗文宇, 宫兆岩, 刘徐祎昊, 焦健, 郭睿   

  1. 华北电力大学机械工程系, 保定, 071000
  • 收稿日期:2025-10-27 出版日期:2026-09-25 发布日期:2026-10-21
  • 通讯作者: 王晓龙
  • 作者简介:王晓龙*(通信作者),男,1989年生,博士、副教授。研究方向为风电机组状态监测及故障诊断。E-mail: wangxiaolong0312@126.com。
  • 基金资助:
    国家自然科学基金(52005180);河北省自然科学基金(E2025502001);中央高校基本科研业务费专项资金(2025MS133);国家级大学生创新创业训练计划(202510079022)

Degradation Assessment of Rolling Bearings Based on CMWRFDE and ATSNE

WANG Xiaolong(), FU Yuanlei, LUO Wenyu, GONG Zhaoyan, LIU Xuyihao, JIAO Jian, GUO Rui   

  1. Department of Mechanical Engineering,North China Electric Power University,Baoding,Hebei,071000
  • Received:2025-10-27 Online:2026-09-25 Published:2026-10-21
  • Contact: WANG Xiaolong

摘要:

针对滚动轴承早期退化时刻难以准确检测的问题,提出了一种基于复合多尺度加权反向模糊散布熵(CMWRFDE)和增强t分布随机邻域嵌入(ATSNE)的退化状态评估模型。对采集到的振动信号利用CMWRFDE算法提取多尺度退化特征,然后利用ATSNE算法对提取到的多尺度特征进行压缩降维,构造出能准确表示滚动轴承性能变化的综合特性指标,最后通过切比雪夫不等式健康阈值策略得到滚动轴承的性能退化评估结果。结果表明,该模型能够准确描述滚动轴承性能退化规律,及时捕捉到初始退化时刻,从而实现对滚动轴承性能退化状态的准确评估。

关键词: 退化状态评估, 复合多尺度加权反向模糊散布熵, 增强t分布随机邻域嵌入

Abstract:

To address the problem of accurately detecting the early degradation moment of rolling bearings, a degradation state assessment model was proposed based on CMWRFDE and ATSNE. First, the CMWRFDE algorithm was utilized to extract multi-scale degradation features from the collected vibration signals. Then, the ATSNE algorithm was employed to compress and reduce the dimensionality of the extracted multi-scale features, constructing a comprehensive characteristic indicator that accurately represents the performance changes of the rolling bearing. Finally, the performance degradation assessment result of the rolling bearing was obtained through a health threshold strategy based on Chebyshev's inequality. The results demonstrate that the proposed model accurately characterizes the performance degradation trend of rolling bearings, promptly detects the initial degradation moment,and enables precise assessment of the bearing's degradation state.

Key words: degradation state assessment, composite multi-scale weighted reverse fuzzy dispersion entropy (CMWRFDE), augmented t-distributed stochastic neighborhood embedding(ATSNE)

中图分类号: