摘要
滚珠丝杠副的正常运行对于保持数控机床稳定性和加工性能具有重要意义。因此,滚珠丝杠副的性能衰退过程评估在健康管理工作中显得尤为必要。考虑到滚珠丝杠副运动的往复性,振动信号的非平稳性和非线性,常规特征提取方法难以直接提取其准确特征。研究了利用数据分段,模糊熵、典型时域特征流形距离进行滚珠丝杠副健康评估的方法。首先,针对原始振动信号进行数据分段处理,区分出滚珠丝杠副滑块正反向运行数据。其次,对分段后同一方向数据提取原始信号的模糊熵和典型时域特征进行多特征融合,构建特征空间。再次,将提取特征归一化计算正常数据与样本数据的流形距离。最后,将流形距离转换成置信值,从而得到滚珠丝杠副的健康程度。试验结果表明,所采用评估方法能够有效评估滚珠丝杠副的性能,为其视情维修提供依据。
The normal operation of ball screw pair is of great significance for maintaining the stability and processing performance of CNC machine tool.Therefore,performance degradation assessment of ball screw pair is a necessary part of health management.Considering the reciprocating motion of ball screw pair,the non-stationary and non-linear characteristics of the vibration signal,it is difficult to extract its accurate features based on conventional feature extraction methods directly.A ball screw pair performance assessment method based on data segmentation,fuzzy entropy and time-domain features extraction and manifold distance calculation was presented.Firstly,the original vibration signal was segmented,meanwhile,the forward and backward running data of the slide block of ball screw pair were picked out respectively.Secondly,the fuzzy entropy and typical time-domain features were extracted from the segmented data of the same direction,and multi-feature fusion was carried out to construct the feature space.Thirdly,these extracted features were normalized to calculate the manifold distance between baseline samples and test sample.Finally,the manifold distance was converted into a confidence value,and the health condition of the ball screw pair was obtained.The test results show that the presented method can effectively evaluate the performance of the ball screw pair and provide a basis for its condition-based maintenance.
作者
袁航
雷振兴
张会娟
刘建娟
YUAN Hang;LEI Zhen-xing;ZHANG Hui-juan;LIU Jian-juan(College of Electrical Engineering,Henan University of Technology,Zhengzhou 450001,China)
出处
《科学技术与工程》
北大核心
2023年第25期10808-10816,共9页
Science Technology and Engineering
基金
河南省科技攻关项目(212102210624)
河南工业大学校高层次人才基金(31401128)
河南工业大学河南省省属高校基本科研业务费专项资金资助(2018QNJH28)
河南工业大学青年骨干教师项目(21420120)
国家自然科学基金青年基金(51805148)。
关键词
滚珠丝杠副
健康评估
振动信号
数据分段
特征融合
流形距离
ball screw pair
health assessment
vibration signal
data segmentation
feature fusion
manifold distance
