微细铣削振动信号特征参数的主成分分析

铣削振动是描述微细铣削加工状态的重要特征参数. 利用微小型车铣加工中心、压电加速度计和多通道电荷放大器建立了微细铣削振动测试系统,分别提取了不同铣削方式和铣削转速条件下微细铣削振动信号的时域特征参数和频域特征参数. 针对特征参数数量繁多且变化趋势不一致的特点,引入主成分分析方法,利用主成分分别对时域和频域特征参数进行替换,定量描述出振动信号的能量和差异,以及主频带位置和能量分散程度之间的关系. 分析结果表明,通过时域与频域特征参数主成分的综合运用,应用较少的参数即可描述微细铣削振动信号的主要特征,显著降低了原始数据维数; 主成分分析结果可用于铣削方式和铣削转速等微细铣削加工参数的优化.

Principal Composition Analysis of the Characteristic Parameter of Vibration Signal in Micro Milling Process

Vibration is one of the most important parameters in micro milling. Based on the establishment of vibration signal measuring system with miniature turn-milling machine tool, tri-axial piezoelectricity accelerometer and multi-channel charge amplifier, vibration signal of three components as well as their time domain and frequency domain parameters were presented within different milling strategies and spindle speeds of micro milling. Principal composition analysis method was introduced considering that the machining condition parameters are various and their trends are significantly different. Time domain and frequency domain parameters were replaced by principal composition, then the relationship between energy and deviation in time domain, power spectrum frequency band and energy distribution in frequency domain were established quantificationally. The analysis results show that the characteristic of vibration signal in micro milling can be described within very limited parameters, which leads to the decrease of dimension of raw experimental data. As a result, the computational result of principal composition is very useful for the optimum selection of micro milling parameters such as milling strategy and spindle speed.