Thermal Error Modeling for Heavy Duty CNC Machine Tool Based on Convolution Neural Network

Thermal error is one of the main factors that affect the machining precision of heavy CNC machine tools. Through the establishment of thermal error model based on temperature field monitoring, the compensation of thermal error is an effective method to improve the machining precision. Because the mechanism of the thermal error is complex, now a data driven modeling method is usually used to establish a thermal error model. However, traditional data driven modeling methods need to choose different key temperature points according to different working conditions to model in order to improve prediction accuracy and generalization. Therefore, it is difficult to get the thermal error model with strong generalization in different working conditions. Convolution neural network is a kind of deep neural network that can automatically extract feature of input data without artificial construction. A thermal error prediction model based on convolution neural network is proposed and the generalization is improved. The experimental results show that this method can accurately predict the thermal error of the Z axis in different working conditions, and validate the effectiveness of the modeling method.