Optimization of the behavior system for controlling bipedal locomotion against external disturbances

Humans show remarkable skills in reactive balancing control on unknown disturbances while standing and walking.Though current bipedal robots can walk, run and step obstacles, they normally perform in a well-controlled environment. Still unexpected perturbations can cause a fall of bipedal robots, therefore advanced capabilities for rejecting unknown disturbances are needed. Studies from neurology and psychophysics of human stance attempt to examine how the human deals with external disturbances. In the following, a short overview of how some of those findings are interpreted and integrated into a behaviour based control architecture for bipedal locomotion is given. Firstly a short introduction to the control architecture is given. Secondly a method for integrating a method for disturbance estimation and compensation (DEC) while standing is illustrated. In the third section an approach to classify, detect and react to external pushes during walking is proposed.