The Selection and Modeling of Temperature Variables for the Thermal Error Compensation on Machine Tools

A method based on the multivariable statistics analysis was described of the thermal error compensation process of CNC machine tools.The measurement points are selected based on the cluster analysis theory and the optimized regression model is given by stepwise regression.The compensation model is made for machine tools.The result shows that the methods not only avoid the correlation of the measurement points and thus ensure the accuracy of the model but also save the time and cost.