切削颤振信号的混沌时间序列识别方法研究

Research of Recognition Method on Chaos Time Series of Milling Chatter Signal

对切削颤振的在线监测识别是提高加工质量的关键技术之一。基于离散动力系统非线性混沌控制理论,以加工过程振动信号为非线性系统输入量,采用嵌入时间序列和关联积分的C-C算法,确定最佳嵌入维数m与时间延迟τ,进而对在线监测获得的离散振动信号进行相空间重构。实施铣削颤振试验,采用上述时间序列分析方法获得颤振产生前后的相图,分析表明相图表现出混沌特性,实现了切削颤振的识别。

The online detecting recognition of milling chatter is one of the key technologies to improve the machining quality. Based on the nonlinear chaos control theory in discrete dynamic system,the processing vibration signal was taken as system nonlinear input,the C-C algorithm of correlation integral and embedded time series were used to determine optimal embedding dimension m and time delay τ. Then the phase space was reconstructed by discrete vibration signal obtained online detecting. In milling chatters experiment,the time series analysis method was used to get the phase diagram before and after chatter. By analyzing the phase chart shows presence of the characteristics of chaos,and implementation of recognized the milling chatter.