基于ICA的镗削过程颤振征兆信号分离方法研究  

ICA based separation of chatter symptom signals for precision hole boring processing

作　　者：李欣[1] 梅德庆[1] 陈子辰[1] 

机构地区：[1]浙江大学机械工程学系,杭州310027

出　　处：《振动与冲击》2013年第9期5-9,共5页Journal of Vibration and Shock

基　　金：国家自然科学基金资助项目(50405036);浙江省自然科学基金资助项目(Y104462);国家教育部留学回国人员科研启动基金项目(教外司留[2009]1590号)

摘　　要：针对精密孔镗削加工过程中易出现颤振、导致精密孔表面质量下降,如何能快速、准确识别出颤振征兆发生问题,提出基于独立分量分析(ICA)的镗削振动信号信噪分离方法,以实现对镗削颤振征兆信号的快速分离。该方法据颤振信号的时频特点,利用经验模态分解(EMD)对镗削振动信号进行分解;对EMD分解所得各本征模态分量(IMF)构造出的虚拟通道进行ICA分析,分离出包含颤振发生征兆的信号。实验结果表明,利用EMD和ICA对镗削振动信号进行分解处理,可快速分离出镗削颤振征兆信号,为后续颤振识别预报及抑制环节提供基础,从而有效提高精密孔的表面加工质量。Recognizing chatter symptom rapidly and exactly is the most important premise for chatter prediction and suppression. In order to accomplish boring chatter symptom separation from the vibration signal of the intelligent boring bar during boring process, an ICA based method was proposed. According to the characteristics of chatter signals, EMD was introduced to decompose the vibration signal of the intelligent boring bar. Then, the IMFs virtual channels were constructed, the signal-noise separation was accomplished by ICA, and the chatter symptom signal was gained. The experimental results show that the EMD-ICA based vibration signal processing can separate the chatter symptom signal rapidly and effectively, and this can provide a foundation for follow-up of the study on chatter prediction and suppression, and improve the quality of the workpiece surfaces.

关 键 词：精密孔镗削 颤振 经验模态分析 独立分量分析 信噪分离 

分 类 号：TG502.96[金属学及工艺—金属切削加工及机床]