Research on the Jumping Control Methods of a Quadruped Robot That Imitates Animals

At present, most quadruped robots can move quickly and steadily on both flat and undulating ground; however, natural environments are complex and changeable, so it is important for a quadruped robot to be able to jump over obstacles immediately. Inspired by the jumping movement of quadruped animals, we present aerial body posture adjustment laws and generate animal-like jumping trajectories for a quadruped robot. Then, the bionic reference trajectories are optimized to build a trajectory library of a variety of jumping motions based on the kinematic and dynamic constraints of the quadruped robot. The model predictive control (MPC) method is employed by the quadruped robot to track the optimized trajectory to achieve jumping behavior. The simulations show that the quadruped robot can jump over an obstacle of 40 cm in height. The effectiveness of the animal-like jump control method is verified.

Automated summary

Currently, most quadruped robots can move swiftly and steadily on flat and uneven surfaces. However, the ability to immediately jump over obstacles is crucial due to the complex and changing natural environments. Inspired by quadruped animals' jumping movement, this study presents laws for aerial body posture adjustment and generates animal-like jumping trajectories for a quadruped robot. The generated bionic reference trajectories are optimized to create a trajectory library of various jumping motions, adhering to the kinematic and dynamic constraints of the robot. The quadruped robot utilizes model predictive control (MPC) to track the optimized trajectory, successfully achieving jumping behavior. Simulation results demonstrate the robot's capability to jump over a 40 cm obstacle, validating the effectiveness of the animal-like jump control method.