Standing balance control of a humanoid robot in the presence of continuous perturbations

For standing balance control of a humanoid robot, most previous work assumes that perturbations are impulsive. For standing balance control in the presence of continuous perturbations, the choice of the desired state receives little attention so far. Our study shows that the choice of the desired state has a large effect on the robustness of a standing balance controller. The present paper takes perturbations as continuous forces that last for a period of time and then proposes an optimization method to determine the desired state. Furthermore, it is applied to standing balance control using a trajectory library and makes the resultant controller more robust in the presence of continuous perturbations. The effectiveness of the proposed method is explored with simulation and experiment results.