Hierarchical Cooperative Locomotion Control of Human and Quadruped Robot Based on Interactive Force Guidance

Abstract

In this article, we propose a new cooperative motion mode of the human and quadruped robot similar to walking the dog or camels carrying goods. This mode allows quadruped robots to follow human wishes, such as guiding quadruped robots to transport goods. The article first establishes the complex hybrid model of cooperative motion between humans and quadruped robots, with a focus on considering the impact of interaction forces on the quadruped robots' motion. Then, to make the motion mode more stable, a cooperative motion control method based on hierarchical control is proposed. At the higher level, the model predictive control method is used to solve the optimal control problem of hybrid models. At the lower level, a nonlinear controller is used to track the trajectory generated by the higher-level control to ensure that the result is more stable. Meanwhile, according to the magnitude and direction of the interaction force, the foothold is quickly adjusted based on the divergent component of motion to realize the stable motion of the quadruped robot. Finally, we carry out simulation and experiments in multiple scenarios, which proves the rationality and effectiveness of the proposed method.