Bayesian Learning-Based Model-Predictive Vibration Control for Thin-Walled Workpiece Machining Processes

Thin-walled flexible workpieces are known to be the most commonly used flexible elements in mechanical structures and machines in the industries of aerospace, national defense, integrated circuit production, and so on. The bottleneck of improving machining quality of thin-walled flexible workpieces lies in the workpiece deformation during machining processes. We hereby establish a machining platform for thin-walled flexible workpieces and develop a purely data-driven Sparse Bayesian learning-based method to predict the future deformation merely by using historical displacement information. Accordingly, a dual-mode predictive controller is developed to mitigate the machining vibrations, and the quality of the workpiece surface has been thus substantially improved. Finally, the superiority and effectiveness of our proposed method are demonstrated through extensive machining experiments of thin-walled flexible workpieces.