Field-Deployable Disturbance Rejection in Bipedal Robots via NMPC-WBC

Bipedal robots have high flexibility and mobility and are suitable for most human work scenarios, so their anti-disturbance performance is the main challenge. This paper proposes a design scheme for a bipedal robot with a single-leg 6-degree-of-freedom, in order to reduce the influence of the leg weight on the overall center of gravity of the robot, three motors for the hip joints are centrally distributed near the center of mass, and the ankle joints adopt a parallel actuator mechanism, which increases the robot's flexibility and further reduces the inertia of the moving leg. A controller based on nonlinear model predictive control (NMPC) and sequence-optimized whole body control (WBC) is designed to simplify the robot center of mass dynamics into a convex optimization problem, which is then transcribed into a nonlinear programming (NLP) problem to solve the optimized ground reaction force, and the simulation experiments prove that the robot is able to move stably at various speeds, and at the same time, it has a strong ability to resist interference.