金属切削过程刀具磨损信号的混沌特征

摘要

针对刀具磨损过程中声发射信号非线性特征，提出基于混沌理论的信号分析及特征提取方法。采用延迟时间法对去噪后的时间序列进行相空间重构，分析延迟时间及嵌入维数随刀具磨损的变化规律；用关联维数、最大 Lya-punov 指数及 Kolmogorov 熵三种混沌特征参数定量分析刀具在不同切削条件下随磨损量增大所呈现的变化规律。研究结果表明，刀具磨损声发射信号具有明显的混沌特征，三种混沌特征参数、延迟时间及嵌入维数与刀具磨损状态具有明显的对应关系，可用作刀具磨损状态监测、磨损量预测的特征参数。%Aiming at the nonlinear characteristics of acoustic emission signals from tool wear,a method of signal analysis and feature extracting based on the chaos theory was proposed.Here,the phase space reconstruction of denoised time series with the time delay method and the analysis of the variation laws of time delay and embedded dimension versus tool wear were performed firstly.Then,the variation laws of tools with increase in the amount of tool wear under different cutting conditions were analyzed using three chaotic characteristic parameters including correlative dimension,the maximum Lyapunov exponent and Kolmogorov entropy.The results showed that the acoustic emission signals from tool wear have obvious chaos characters,moreover the above three chaotic characteristic parameters,time delay and embedded dimension have significant corresponding relationships with the states of tool wear so that they can be used as parameters for condition monitoring of tool wear states and prediction of the amount of tool wear.