Seamless Reaction Strategy for Bipedal Locomotion Exploiting Real-Time Nonlinear Model Predictive Control

This paper presents a reactive locomotion method \nfor bipedal robots enhancing robustness and external distur- \nbance rejection performance by seamlessly rendering several \nwalking strategies of the ankle, hip, and footstep adjustment. \nThe Nonlinear Model Predictive Control (NMPC) is formulated \nto take into account nonlinear Divergent Component of Motion \n(DCM) error dynamics that predicts the future states of the robot \nin response to the walking strategies. This formulated NMPC \nenables the seamless application of these strategies improving \npush disturbance rejection performance. The proposed controller \nis validated in simulation and through an experiment on a bipedal \nrobot platform, Gazelle, which confirms its effectiveness in real-time.