A Dynamic Linearization Modeling of Thermally Induced Error Based on Data-Driven Control for CNC Machine Tools

This paper proposes a novel dynamic linearization modeling method for machine tool thermal errors based on data-driven control theory, with improved accuracy and robustness under various practical working conditions of machine tool. The nonlinear, quasi-static and pseudo-hysteric characteristics of the machine tool temperature field are identified as the main causes for poor robustness in conventional thermal error mathematical models. The theoretical and practical difficulties in applying conventional modeling approaches based on the model-based control theory are demonstrated using two types of CNC machine tools as examples. The data-driven control theory is applied to dynamic linearization modeling and the developed data model has shown significant improvement over the general dynamic model in terms of model accuracy and robustness. The feasibility and effectiveness of the proposed dynamic linearization modeling method has been verified using two experiments, demonstrating excellent robustness and ability to adapt to various machining conditions and to improve machining accuracy.