Improvement of motion accuracy and energy consumption of a mechanical feed drive system using a Fourier series-based nonlinear friction model

Abstract

Friction occurring in all mechanical systems, such as computer numerical controlled (CNC) machine tools, is an important issue in achieving the high accurate performance. Friction adversely affects not only motion accuracy of drive axes but also excessively consumes energy. Feed drives of CNC machines normally operate all day and night around the world, and therefore consumed energy reduction is highly expected. The motivation behind this work is to construct a novel friction model that can comprise many unknown friction sources in both low and high velocity regions and enable a friction compensator to precisely describe actual frictional behavior. A sliding mode control (SMC) is designed to verify the effectives of the proposed friction model in a biaxial feed drive system. Experimental results confirm that a combination of SMC and the proposed friction can effectively improve tracking accuracy and further achieve significant reduction of consumed energy compared to combining with the conventional model. Results show that the proposed approach can largely decrease the mean tracking error to less than 5 µm for each axis. The new friction also achieved effective reduction of control variance by 7.62%. Consequently, consumed energy of feed drives was significantly improved by 12.83% compared to using the conventional model.