Online Centroidal Angular Momentum Reference Generation and Motion Optimization for Humanoid Push Recovery

摘要

This letter presents a new push recovery algorithm for humanoid robots in balancing scenarios by exploiting the system's rotational dynamics. The proposed framework actively generates centroidal angular momentum (CAM) references based on the force magnitude and direction of the push to counteract the disturbance and maintain its balance. Since a humanoid robot can only store a limited amount of angular momentum, the CAM reference is generated in three consecutive phases: 1) CAM generation phase to counteract the push; 2) CAM reduction phase to bring the robot to a halt; and 3) posture recovery phase to converge to the reference pose again. A subsequent whole-body motion optimizer, formulated as a constrained quadratic optimization problem, generates kinematically feasible whole-body trajectories based on the CAM reference. The proposed framework is validated through experiments with the humanoid robot TORO.