基于VMD的铣刀破损检测

针对铣削过程中的切削振动信号具有非平稳性的特点,提出了一种基于变分模态分解(VMD)的铣刀破损检测方法。该方法通过VMD将切削振动信号分解成若干个模态分量,由于铣刀发生破损后,不同模态分量的频带分布会发生变化,因此提取各模态分量的中心频率和能量组成特征向量;对特征向量进行归一化处理,最终输入到支持向量机(SVM)进行铣刀破损检测。在多种切削参数下进行铣削加工实验,结果表明该方法比基于EMD的铣刀破损检测方法能抑制模态混叠的发生且具有更高的检测精度。

Milling cutter breakage detection based on VMD

Aiming at the non-stationary characteristics of the cutting vibration signal in the end milling process, a milling cutter breakage detection method based on variational mode decomposition (VMD) was proposed.The method decomposes the cutting vibration signal into several modal components by VMD.After the milling cutter is broken, the frequency band distribution of different modal components will change, and the center frequency and energy of each modal component are extracted to construct a feature vector.The feature vector was normalized and input to the support vector machine (SVM) for milling cutter breakage detection.Milling experiments under various cutting parameters show that the method can suppress modal mixture and has higher detection accuracy than the EMD-based milling cutter damage detection method.