Simultaneous adaptation to rough terrain and unknown external forces for biped humanoids

This paper reports on the applicability of our passivity-based contact force control framework for humanoids. We present its adaptation to unknown rough terrain. The adaptation to uneven ground is achieved by an optimally-distributed anti-gravitational forces applied to preset contact points in a feed-forward manner even without explicitly measuring the external forces or the terrain shape. The adaptation to unknown inclination is also possible by combining an active balancing controller based on the measurement of the CoM with respect to the inertial frame. Furthermore, it is shown that a simple impedance controller for the supporting feet or hands allows the robot to adapt to a low friction ground without prior knowledge of the ground friction. The proposed adaptive mechanism is experimentally validated on a full-size biped humanoid robot balancing on uneven ground or time-varying incline.