Biologically-inspired postural and reaching control of a multi-segment humanoid robot

This article presents a biologically-inspired framework for postural and reaching control of a multi-segment humanoid robot. It is based mainly on the underpinning principles of human postural control as identified in a previous work by the author and tested experimentally on a one-segment special-purpose robot. Thus the goal of this article is to extend the control and estimation architecture to cover multi-segment humanoids. In addition to the inherent instability of humanoids, multi-segment motion introduces destabilizing dynamic coupling effects that requires heterogeneous control in the form of full state feedback. Thus a robust multi-input-multi-output tracking control architecture with feedback, feedforward, and integral control parts is implemented. This article addresses as well a variety of external disturbances that might act on the humanoid such as supporting surface translational and rotational motion as well as external pull/push forces acting on the body of the humanoid. A central extended disturbance observer is used to estimate the external disturbances is employed. Simulation experiments demonstrate good performance for this method and highlight the merits of the central extended observer which is based on a full internal model that takes into account the kinematic and kinetics of the humanoid.