Development of Model-Based Robot Polishing System: Measurement and Compensation for Robot Stiffness Distribution

Abstract

As the working-age population continues to decline, the use of robots in polishing processes is increasing. However, the integration of robots into these processes presents challenges, necessitating a model-based approach to understand and accommodate their characteristics. Although previous studies have developed robotic polishing systems, most have relied on imitating skilled technicians, with limited success. Additionally, attempts to implement such systems using feedback control have faced difficulties owing to response delays and system complexity. To address these challenges, this study developed an innovative robotic polishing system that operates without feedback control, aiming to reduce response-related limitations and simplify the system architecture. A stiffness distribution of the robot was measured, and a model-based framework was established. Polishing characteristics were investigated through experiments on a large-coated film plate. The results demonstrated that, by compensating for variations in polishing pressure due to the robot’s stiffness distribution, polishing in accordance with Preston’s law can be achieved.