基于铣削加工仿真和现场数据的切削力系数挖掘方法

A method for mining cutting force coefficient based on milling simulation and real-time data

为提高切削力系数的辨识精度,提出了一种基于铣削加工仿真和现场数据的切削力系数挖掘方法。分别利用实验法和仿真分析法获得切削加工过程数据集。建立了基于刀具旋转周期的切削力系数挖掘方法,并将仿真数据作为数据挖掘的约束条件,分析了切削力系数的分布特性。研究结果表明:周期相关性切削力系数呈正态分布,可通过正态分布估计获得其准确数值。该方法可以有效地挖掘切削力系数,实现切削数据的准确转换。数据转换平均误差仅为3%,达到了实验标定效果,有效地提高了切削力系数辨识效率。

Cutting force is one of the important parameters for monitoring the machining process and quality,and cutting force coefficient directly affects the prediction accuracy of cutting force.This paper proposes a cutting force coefficient mining method based on milling simulation and real-time data to improve the identification accuracy of cutting force coefficients.First,the cutting process dataset is obtained by experimental and simulation analysis methods.Second,a cutting force coefficient mining method based on tool rotation period is built,and the simulation data is used as a constraint condition for data mining to analyze the distribution characteristics of cutting force coefficients.Our research results indicate the distribution of periodic correlated cutting force coefficients exhibits a normal distribution form,and its accurate value is estimated through the normal distribution.This method effectively mines the cutting force coefficient and achieves accurate conversion of cutting data.The average error of data conversion is only 3%,achieving the same effect as experimental calibration and effectively improving the efficiency of cutting force coefficient identification.