基于优化变分模态分解和多尺度熵特征的铣削颤振监测研究  

Research on Milling Chatter Detection Based on Optimized Variational Modal Decomposition and Multiscale Entropy Features

作　　者：朱晓慧 刘长福 于新丽[2] 刘博 Zhu Xiaohui;Liu Changfu;Yu Xinli;Liu Bo

机构地区：[1]沈阳工学院机械工程与自动化学院 [2]辽宁石油化工大学机械工程学院

出　　处：《工具技术》2024年第8期140-149,共10页Tool Engineering

摘　　要：在加工低刚度工件时,颤振是影响表面质量、加工效率和刀具寿命等方面的重要因素。为更准确地识别铣削加工状态,提出了一种新的颤振特征提取方法。利用梳状滤波器和经验模态分解对原始信号进行预处理;基于皮尔逊相关系数对本征模态函数(IMF)进行重构,利用灰狼优化算法(GWO)获得变分模态分解(VMD)的最优参数;计算各子频率带的能量熵特征,获得具有丰富颤振信息的频带;应用多尺度排列熵(MPE)和多尺度模糊熵(MFE)作为颤振特征,基于熵的取值范围,在各加工状态下选择最佳尺度特征。结果表明,该方法基于最佳尺度可以有效提取敏感颤振特征。When machining low stiffness workpieces,chatter is an important factor affecting surface quality,machining efficiency and tool life.A novel method for extracting chatter features has been proposed to more accurately identify milling machining status.The raw signal is preprocessed using comb filters and empirical mode decomposition.The intrinsic mode function(IMF)is reconstructed based on the Pearson correlation coefficient,and the optimal parameters of variational mode decomposition(VMD)are obtained using the grey wolf optimization algorithm(GWO).To obtain frequency bands with rich chatter information,the energy entropy characteristics of each sub frequency band are calculated.Multi-scale permutation entropy(MPE)and multi-scale fuzzy entropy(MFE)are calculated as chatter characteristics.Based on the range of entropy values,the optimal scale features in each processing state is selected.The analysis results indicate that the proposed method can effectively extract sensitive chatter features based on the optimal scale.

关 键 词：颤振监测 灰狼优化 VMD 多尺度排列熵 多尺度模糊熵 

分 类 号：TG54[金属学及工艺—金属切削加工及机床] TH161[机械工程—机械制造及自动化]