一种SCNGO⁃MMPE⁃VMD的滚动轴承故障诊断方法

打开文本图片集

关键词： 正余弦算法； 滚动轴承； 故障诊断； 改进北方苍鹰优化算法； 多尺度均值排列熵； 变分模态分解中图分类号：TN911.23⁃34；TH17 文献标识码： A 文章编号：1004⁃373X（2025）12⁃0054⁃07

Abstract：In allusion to the problems of uneven population distribution and premature convergence of algorithm in the rolling bearing fault diagnosis， a method of SCNGO⁃MMPE⁃VMD fault diagnosis for rolling bearings is proposed. The refraction reverse learning method is used to initialize the individual， and generate the reverse solution to expand the search scope. The sine⁃ cosine algorithm （SCA） strategy is introduced into the survey phase of the northern goshawk optimization （NGO） algorithm， dynamically adjusting the step size search factor by means of the nonlinear weighting coefficient ω ， reducing the dependence of individual position updates on local information， and significantly improving the convergence speed and accuracy of the algorithm. The comprehensive index of multi⁃scale mean permutation entropy （MMPE） and kurtosis fusion is taken as the fitness function to enhance the sensitivity of the fault feature. The different actual measured signals were tested. The results show that， under the strong noise interference， in comparison with the traditional methods， this method can identify fault characteristics 300 min in advance （initial fault） and 700min in advance （weak fault）， verifying its engineering practicality.

Keywords：sine⁃cosine strategy; rolling bearing; fault diagnosis; improved northern goshawk optimization algorithm; multi⁃ scale mean arrangement entropy; variational mode decomposition

0 引 言

滚动轴承是许多工业领域中至关重要的组件，在旋转机械中发挥着关键作用。（剩余8064字）