Research On Prediction Of Turning Tool Life Based On Workpiece Processing Qualit

With the arrival of the fourth industrial revolution,the manufacturing industry has ushered in a new round of industrial upgrading.How to use digital technology to empower the industry has increasingly become an important content of enterprises.There are some problems in turning,such as lathe failure caused by the end of turning tool life,low utilization rate of turning tool,low efficiency of turning tool management,decline of workpiece quality and so on.In this paper,the machining quality characteristics such as workpiece surface roughness and workpiece surface waviness are used to establish the turning tool wear model and life model,and a digital twin system is established to realize the turning tool wear prediction and life prediction,improve the production efficiency and workpiece machining quality,and reduce the production cost.The main research contents of this paper are as follows:(1)A new idea of predicting turning tool wear by using the characteristics of workpiece machining quality and a method of predicting turning tool life based on workpiece machining quality are proposed;The life prediction process of turning tool based on digital twin is designed.(2)The wear and life model of turning tool is constructed.According to the change law of turning tool wear,friction theory and integrated algorithm idea,the turning tool wear model based on logistic variant,the turning tool wear model based on hyperbolic sinusoidal function,the turning tool wear model based on polynomial regression,the turning tool wear model based on friction theory,multiple regression turning tool wear model based on integrated algorithm and turning tool wear model based on multiple regression are established;Three tool life models are constructed based on decision tree,random forest and logistic regression.(3)Turning experiments were carried out.The experimental scheme is designed,the experimental conditions are evaluated and the cutting scheme is designed;The variation laws of tool flank wear,workpiece relative diameter,workpiece surface roughness and workpiece surface waviness with workpiece processing time series are analyzed.(4)The parameter acquisition and precision comparative analysis of turning tool wear and turning tool life model are carried out.The tool wear model is trained by cutting experimental data,and six tool wear model parameters are obtained;By comparing the prediction accuracy and prediction effect of six turning tool wear models,the optimal turning tool wear model is selected as the multiple regression turning tool wear model based on integrated algorithm.The tool life model is trained by using the experimental data,and the prediction accuracy and prediction stability of the three tool life models are compared.Among them,the tool life model based on random forest has the highest prediction accuracy,and the average prediction accuracy is 98.41%,which can be applied to the prediction of turning tool life.(5)A digital twin system for tool life prediction is designed and developed.Firstly,the requirements of digital twin system for turning tool life prediction are analyzed,and the system workflow and system development process are designed;Secondly,the system data communication is realized based on TCP/IP and HTTP protocol,and the data processing is realized by writing specific programs;Finally,the tool wear prediction and tool life prediction are realized by using the digital twin system of tool life prediction.