基于数字孪生的铣刀磨损状态识别方法研究

Research on Recognition Method of Milling Cutter Wear State Based on Digital Twin

实时精准地监测铣刀磨损状态对于提高加工质量与加工效率具有重要意义,提出一种基于数字孪生的铣刀磨损状态识别方法,该方法通过结合VMD-MPE特征提取方法和GA-SVM状态识别模型构建数字孪生体对铣刀磨损状态进行实时监测。首先,利用变分模态分解算法(VMD)分解铣刀振动信号得到包含磨损状态信息的模态分量;其次,引入多尺度排列熵(MPE)从包含磨损状态信息的模态分量中提取铣刀的非线性动力学特征,并取各有效模态分量的多尺度排列熵平均值作为特征矩阵;最后,通过遗传算法(GA)优化支持向量机(SVM)构建铣刀磨损状态识别模型。实验结果表明,所构建的数字孪生体具有良好识别效果,其识别精度可达97.33%。

Real-time and accurate monitoring of the wear state of milling cutter is of great significance to improve the machining quality and efficiency.In this paper,a milling cutter wear state recognition method based on digital twin is proposed.This method combines VMD-MPE feature extraction method and GA-SVM state recognition model to construct digital twins to monitor the wear state of milling cutter in real time.Firstly,the variational mode decomposition(VMD)algorithm is used to decompose the vibration signal of the milling cutter to obtain the modal component containing the wear state information.Secondly,the multi-scale permutation entropy(MPE)is introduced to extract the nonlinear dynamic characteristics of the milling cutter from the modal components containing the wear state information,and the average value of the multi-scale permutation entropy of each effective modal component is taken as the characteristic matrix.Finally,the genetic algorithm(GA)is used to optimize the support vector machine(SVM)to construct the milling cutter wear state recognition model.The experimental results show that the digital twin constructed in this paper has a good recognition effect,and its recognition accuracy can reach 97.33%.