Trajectory Generation for Wheel-Legged Robot Jumping Based on Maximum Torque Minimization

This paper proposes a method for optimizing the jump trajectory of a robot, aiming to minimize the maximum torque. By utilizing the dynamics of the entire machine and iteratively refining the range of joint torques, the maximum torque during the jump is effectively reduced. Validation was conducted on a self-developed wheeled-legged robot, and the results show that the generated trajectory is effective and reliable. This method significantly improves the robot’s jumping performance and has important implications for the motion planning of wheeled-legged robots.