Data-mechanism fusion modeling and compensation for the spindle thermal error of machining center based on digital twin

Current methods for measuring thermal errors due to spindle operation often capture data from other components, complicating the measurement process. Furthermore, data-driven modeling struggles to integrate structural thermal deformation mechanisms, resulting in poor model generalization. To address these challenges, the data-mechanism fusion digital twin (DT) system for spindle thermal errors modeling and compensation is established, which encompasses the physical entity layer (PEL), DT prediction layer (DT-PL), and DT interaction service layer (DT-ISL). In the PEL, information from the machine tool is collected. In the DT-PL, the thermal error experiment is designed to identify the spindle thermal errors, and the multi-channel ensemble algorithm leveraging the physical mechanism (MCEA-PM) is proposed to calculate the spindle thermal deformation. The DT-ISL handles thermal error calculation, data visualization, and interaction with machine tools. The effectiveness of the proposed system was evaluated, achieving over 90 % prediction accuracy and a 72 % increase in machining accuracy during processing.