摘要
针对磨床信号非线性、多维性和非平稳性的特点,提出一种基于二维经验模态分解(BEMD)和控制图理论的磨床磨削颤振预测方法.通过BEMD筛选出有效的二维固有模态函数(BIMF)分量,计算出各分量的实时方差,合成总的方差特征量,并通过控制图对颤振进行预测.结果表明,基于BEMD和控制图的颤振预测方法能够提供颤振阈值并有效地对颤振进行预测.
In view of the non-linearity, multidimensional and non--stationary of grinder signal,a grinding chatter pre- diction method based on Bivariate Empirical Mode Decompo- sition(BEMD) and Control Chart is proposed in this paper. Firstly,the effective BIMFs are sifted by using BEMD and criterion of the correlation coefficient. Then the real-time va- riance of each component is calculated, and the synthetic real- time variance feature vector is composed. Finally, grinder chatter prediction is achieved by using Control Chart. The results show that the pro-posed method based on BEMD and Control Chart could provide chatter threshold and predict grinder chatter effectively and efficiently.
出处
《成组技术与生产现代化》
2017年第4期13-18,共6页
Group Technology & Production Modernization
基金
国家自然科学基金重点资助项目(E050302)
浙江省重点科技创新团队资助项目(2010R50005)
关键词
磨削颤振预测
固有模态函数
控制图
二维经验模态分解
grinding chatter prediction
intrinsic modefunction
control chart
bivariate empirical mode decomposition