Push Recovery of a Quadrupedal Robot in the Flight Phase of a Long Jump

Abstract

Legged robots are well-suited for operation in challenging natural environments, such as steep obstacles or vast gaps in the ground. Aside from difficult terrain, robots may also encounter unanticipated impact forces while performing jumping gaits. When performing their gaits, legged robots should be able to maintain and regain their stability in the face of external perturbations. External disturbances should be detected, and necessary actions should be taken to maintain the robot's balance in order to ensure optimum landing conditions. This paper considers flight phase disturbances in the form of a push on the robot body and introduces a novel push recovery algorithm that uses angular momentum to generate reference trajectories for a quadrupedal robot with waist joints during the flight phase of a long jump. This method creates joint position reference trajectories for the quadrupedal robot's waist and rear hip joints in order to achieve the required orientation of the robot in the air. In order to track reference trajectories, PID joint control is utilized. The robot model employed for the computations is comprehensive because components of the robot body - the leg links and three torso sections - are represented with independent mass values. The proposed push recovery trajectory generation approach is computationally efficient and hence suitable to be employed in real-time applications. The suggested method is used to simulate a quadrupedal robot to test the push recovery algorithm following external disturbances in the flight phase of a long jump. The results demonstrate that the suggested approach performs well in terms of angular position and angular velocity accuracy and it can achieve a posture suitable for landing.