Dynamic Modeling and Parameter Identification of a SCARA Robot Including Nonlinear Friction and Ball Screw Spline Coupling

摘要

Abstract This paper presents a comprehensive approach to dynamic modeling and parameter identification for SCARA robots. The focus is on accurately capturing nonlinear friction and coupling effects, which are frequently inadequately addressed in standard modeling and identification techniques. A novel two-step identification procedure is proposed, where friction parameters are identified independently from inertial parameters. This decoupled approach allows for a more accurate representation of friction behavior, especially at low velocities. Additionally, the influence of ball screw spline coupling between joints 3 and 4 is investigated, revealing its impact on the robot’s dynamics, particularly under payload conditions. The identified dynamic model is validated through both standard torque prediction and model-based control. Experimental results demonstrate the superiority of the proposed approach over traditional methods, leading to significantly improved torque prediction accuracy and enhanced tracking performance in model-based control.