Control con rechazo activo de perturbaciones para el equilibrio de robots humanoides

This paper presents a control strategy for the balance of humanoid robots while they are standing without walking, using the active disturbance rejection control approach. Such control technique regulates the position of the ZMP of the humanoid in order to keep it within the support polygon by means of a discrete-time linear quadratic regulator enhanced with on-line disturbance detection and correction using an extended-state observer. The proposed control strategy was implemented on a Nao humanoid robot using Python and showing successful results by keeping the robot standing despite of external forces that affect the position of its ZMP. The results encourage the use of this controller for disturbance rejection meanwhile other complex activities are performed by humanoids, such as walking, or its application in the control of different kinds of robots.