基于数字孪生的刀具磨损监测和预测方法研究

Research on Tool Wear Monitoring and Prediction Method Based on Digital Twins

利用数字孪生方法建立了刀具磨损监测和预测模型,并利用实际测量得到的数据对算法进行验证。首先,在理论方面归纳总结了数字孪生模型的基本理论和实现方法,主要包括几何、物理、行为以及规则4种子模型。其次,以随机森林算法为基础,实现了刀具磨损与预测数学模型的搭建,主要改进点为利用特征向量的表示优化随机森林算法。最后,利用实验数据验证该算法。结果显示,该算法提升了刀具磨损预测的准确率。

A tool wear monitoring and prediction model is established by using digital twin method,and the algorithm is verified by the actual measurement data.In this paper,the basic theory and implementation method of digital twin model are summarized theoretically,which mainly includes geometric,physical,behavioral and rule 4 seed models.Secondly,the mathematical model of tool wear and prediction is built based on the random forest algorithm.The main improvement point is to optimize the random forest algorithm by using the representation of eigenvector.Finally,experimental data are used to verify the algorithm.The results show that the algorithm improves the accuracy of tool wear prediction.