Run-and-jump Planning and Control of a Compact Two-wheeled Legged Robot

Yunsong Hao,Biao Lu,Haixin Cao,Jinyang Dong,Rui Liu
DOI: https://doi.org/10.1109/ACIRS55390.2022.9845589
2022-01-01
Abstract:With high energy efficiency and strong terrain adaptability, the two-wheeled legged robot combines the advantages of legged robots and wheeled robots. The compact mechanism enables the robot to run and jump only with four motors. However, the robot’s underactuation and instability make it difficult to achieve simultaneous run-and-jump motion. In order to solve this problem, a motion planning and control scheme is proposed. First, the dynamic model of the compact two-wheeled legged robot is established. Subsequently, the run-and-jump trajectory is planned by using nonlinear programming. In order to accurately execute the planned trajectories, two controllers are respectively designed for the jumping motion and running motion. Specifically, an impedance tracking controller is designed to achieve smooth and compliant jumping behavior under different practical conditions. On the other hand, an LQR (Linear Quadratic Regulator) controller is constructed to realize the stable running motion. The overall planning and control scheme is validated in Simscape. With the peak torque of 35 Nm for the hip motor, the run-and-jump motion can reach 20 cm jump height and 40 cm horizontal distance.
What problem does this paper attempt to address?