Manipulator Trajectory Planning Based on Clustering Curve Discretization and B-Spline

Abstract

In addressing the welding trajectory planning problem for a six-axis robot under visual guidance, a method based on clustering and isometric image trajectory curve discretization is proposed to enhance welding trajectory precision and improve welding quality. Firstly, the kinematic model of the six-axis robot is established using the D-H parameter method. Subsequently, the discrete processing of the welding trajectory map recognized by an industrial camera and chord error calculation are performed. The discrete trajectory is then interpolated using fifth-order non-uniform B-spline interpolation, with welding speed planning based on an S-shaped speed planner. Finally, the planned trajectory is verified by welding simulation in MATLAB and V-REP simulation. The simulation results show that the end trajectory of the manipulator is smooth and stable. Compared with the equal arc length discretization method, the discretization accuracy of the proposed method is improved by 19% (plane curve) and 10.7% (space curve), indicating that the proposed discretization method improves the end tracking accuracy of the six-axis robot.