本文在实验的基础上探讨了金属切削过程中切削参数对声发射(Acoustic Emission)的影响,分析了产生断续切屑、连续切屑及切屑缠绕时声发射信号在时域、幅域和频域的特性,提出在信号包络的幅域和频域中选择切屑形态的模式特征,并用基于Bayes准则的LMSE迭代算法确定的二维线性判别函数对切屑形态进行分类,得到比较满意的结果。 Based on the experiments, the influence of cutting parameters on acoustic emission during metal cutting was analyzed. The effects of cutting noise, continuous chip and chip wrapping on the acoustic emission in the time, amplitude and frequency domains We propose to select the pattern features of the chip shape in the amplitude and frequency domain of the signal envelope and classify the chip morphology using the two-dimensional linear discriminant function determined by the LMSE iterative algorithm based on the Bayes criterion. The satisfactory results are obtained.