基于CEEMDAN-S的数控铣床刀具磨损故障特征提取

The Fault Feature Extraction of CNC Milling Machine Based on CEEMDAN-S

【目的】解决因刀具失效而造成的工件报废和关键部件损坏等问题。【方法】以VDM850E型立式加工中心为试验平台,提出一种基于CEEMDAN-S的数控铣床刀具磨损故障特征提取法。先用CEEMDAN算法对采集到的振动信号数据进行分解,得到多个IMF分量,再根据相关系数的选取准则,从中挑选出有效的IMF分量值,以消除噪声等干扰因素,保留有效特征,从而实现对原始数据的降噪处理。对重构信号进行S变换,分析其时频特征。【结果】将重构信号与原始信号的S变换进行对比,能明显看出故障特征。【结论】研究结果表明,对刀具的磨损故障特征进行提取,提出的CEEMDAN-S算法识别精度高,故障特征明显,并优于其他特征提取方法。

[Purposes]This paper aims to address issues such as workpiece scrapping and key component damage caused by tool failure.[Methods]This paper proposed a method of cutting tool wear fault feature extraction based on CEEMDAN-S.First,CEEMDAN algorithm is used to decompose the collected vibration signal data to obtain multiple IMF components,and then according to the selection criteria of correlation coefficient,the effective IMF component value is selected to eliminate noise and other interference factors and retain effective characteristics,so as to realize the noise reduction processing of the original data.S transform of the reconstructed signal to analyze the time frequency characteristics.[Findings]By comparing the S transformation of the reconstructed signal with the original signal,fault feature can be obviously found.[Conclusions]The results show that for the extraction of fault feature of the milling machine the proposed CEEMDAN-S algorithm has high identification accuracy,obvious fault features,and is better than other feature extraction methods.