Research on optimization of temperature sensitive points of machine tool thermal error based on independent variable selection criteria

In order to improve the prediction accuracy and robustness of the machine tool thermal error model, a method to select temperature sensitive points of the machine tool thermal error model based on independent variable selection criteria is proposed. Firstly, the positioning error of the machine tool is divided into geometric positioning error and thermal positioning error, and polynomial is utilized to fit the geometric positioning error. Secondly, based on the independent variable selection criteria, the temperature sensitive points are screened out, and a long short-term memory network (LSTM) is constructed between the temperature sensitive points and the thermal positioning error. Finally, the above two error terms are superimposed to establish a comprehensive positioning error model. The experimental results of a three-axis vertical machine tool showed that the prediction accuracy of the thermal error model constructed by the proposed method is very refined, which indicates that this method is feasible and can provide a theoretical basis for the selection of temperature sensitive points in the thermal error empirical modeling method.