中国机械工程 ›› 2025, Vol. 36 ›› Issue (8): 1774-1783.DOI: 10.3969/j.issn.1004-132X.2025.08.013

• 智能制造 • 上一篇    

基于SABO-VMD的数控机床元动作单元故障可诊断性评价

葛红玉(), 赵展, 郭安祥, 孙佳瑞   

  1. 西安科技大学机械工程学院, 西安, 710100
  • 收稿日期:2024-07-17 出版日期:2025-08-25 发布日期:2025-09-18
  • 作者简介:葛红玉*,女,1982年生,副教授、博士。研究方向为智能制造装备可靠性。E-mail: gxy-xkd@xust.edu.cn
  • 基金资助:
    国家自然科学基金(51705417);陕西省自然科学基础研究计划(2019JQ-086)

Fault Diagnosability Evaluation of Meta Actuation Units Based on SABO-VMD

Hongyu GE(), Zhan ZHAO, Anxiang GUO, Jiarui SUN   

  1. College of Mechanical Engineering,Xi’an University of Science and Technology,Xi’an,710100
  • Received:2024-07-17 Online:2025-08-25 Published:2025-09-18

摘要:

为了判断并量化元动作单元故障诊断的难度,提出一种元动作单元的故障可诊断性评价方法。利用减法平均优化算法(SABO)优化的变分模态分解(VMD)对元动作单元的故障信号进行分解,利用峭度准则筛选IMF分量,构建基于包络熵的元动作单元特征向量;以余弦距离作为相似性度量指标,将故障可诊断性定量评价问题转换为不同故障模式下振动信号特征向量的相似性度量问题;构建元动作单元故障可诊断性评价矩阵,从而建立元动作单元的故障可诊断性评价指标。最后以蜗轮元动作单元为例进行实验验证分析,结果表明所提方法能够实现元动作单元不同故障模式的可诊断性的定量评价。

关键词: 元动作单元, 故障可诊断性, 减法平均优化算法, 变分模态分解

Abstract:

This paper introduced an evaluative approach to gauge the complexity of fault diagnosis within meta actuation units. The methodology commences with the decomposition of fault signals from these units, utilizing a VMD technique refined by SABO. The processes included the applications of a Kurtosis-based criterion to select pertinent intrinsic mode functions (IMFs), culminating in the creation of a feature vector grounded in envelope entropy. The evaluative task then pivoted on employing Cosine distance as a measure of similarity, recasting the fault diagnosability problems into one of assessing the likeness of vibration signal feature vectors across varying fault conditions. A diagnosability evaluation matrix for the meta actuation units was formulated, which layed the foundation for a diagnostic index. It is concluded with an empirical validation using a worm gear-based meta actuation unit; the findings confirm the method’s efficacy in quantitatively gauging the diagnosability of diverse fault patterns.

Key words: meta actuation unit, fault diagnosability, subtraction-average-based optimizer (SABO), variational mode decomposition (VMD)

中图分类号: