A kind of biomimetic control method to anthropomorphize a redundant manipulator for complex tasks

It is an urgent problem for robots to operate complex tasks with some unknown motion mechanisms caused by the strong coupling of force and motion. However, humans can perform complex tasks well due to their natural evolution and postnatal training. A novel biomimetic control method based on a human motion mechanism with high movement adaptability is proposed in this paper. The core is to present a novel variable-parameter compliance controller based on human operation mechanisms with an action-planning method derived from optimization by human motion, and the main contribution is to change the parameters of compliance controller according to human operating intention synchronized with humanoid motion; this change could establish a humanoid map between the force and motion for a seven degree-of-freedom redundant manipulator to deal with the unknown motion mechanism in complex tasks, so the redundant manipulator can operate complex tasks with high performance. Sufficient experiments were performed, and the results validated the effectiveness of the proposed algorithm.