基于逐步回归的机床温度测点优化及热误差建模技术

Temperature measurement point optimization and thermal error modeling of CNC machine tools based on stepwise regression

通过对机床温度测点进行优化,建立其与机床热误差之间的数学模型,对机床热误差进行实时预测与补偿控制,是提高数控机床加工精度的重要途径。为解决现有机床热误差模型预测精度低、鲁棒性差的问题,提出一种基于逐步回归的数控机床温度测点优化方法。通过偏F统计量的检验,在初步建立的回归模型中逐个引入新变量,剔除不显著的老变量,实现温度测点的优化布置,获得数控机床热误差的最优回归模型。将该方法应用于某数控机床,结果表明,基于逐步回归的机床热误差模型,所用温度变量最少,且预测精度最高。

Through temperature measurement point optimization of CNC machine tools,thermal error model is presented,and which is the prerequisite of real- time prediciton and compensation. In order to solve the problem of lower prediciton accuracy and poor robustness of present thermal error models,a new temperature measurement point optimization method based on stepwise regression is proposed. According to the calculation results of partial- F statistic,new variable is introduced into the regression model one by one,inapparent variables are deleted,and optimization layout of temperature measurement point is achieved,the optimized regression model for machine tool thermal error is presented.The new regression modeling method is applied on a CNC machine tool,the result shows that thermal error model based on stepwise regression has least temperature variables and highest prediction accuracy.