Dynamic control of bipeds using postural adjustment strategy

In this paper, we propose a new way to solve the dynamic control of a humanoid biped robot in an unknown environment, taking into account the postural adjustment. For that a multichain mechanical model of a human body (called BIPMAN) is proposed with a general control architecture. We use the postural adjustment notion (physiological notion) to ensure the biped dynamic equilibrium. The principle is based on behavioral rules using limiting conditions and “equilibrium classes”. This notion is introduced at the “coordinator level” of our architecture and more especially in RTCA technique development (real-time contraints adaptation). This technique is based on a general criterion which is able to minimize roll and pitch angles in function of the task constraints (bipodal or unipodal posture). Finally, we illustrate the whole strategy through examples