This paper presents a real-time velocity optimization method for arbitrary robot trajectories. Based on time scaling, The method optimizes trajectory velocity by replanning while ensuring kinematic and dynamic constraints. It is independent from the original trajectory generation process, allowing application to any predefined trajectory. Operating in joint space, the method simultaneously adjusts position and orientation velocities of the robot's end effector. Through interpolation of the original trajectory, it preserves geometric characteristics and multi-joint synchronization. The method efficiently applies kinematic and dynamic constraints and achieves real-time optimization by using trajectory generation methods with low computational complexity for velocity replanning, Experimental results confirm the method's effectiveness.
Yu et al. (Tue,) studied this question.