Initial experiment of practical algorithm for precision motion control of human-like tendon-driven joint mechanism

This paper presents a new precision control algorithm for the antagonistic tendon-driven joint (TDJ) which has human-like structure. The antagonistic TDJ mechanism has the flexibility of joints by properties of the tendon and mechanical uncertainties by its complex structure, which renders it difficult to control the TDJ mechanism. In order to deal with this difficulty in control, we perform modeling of the TDJ and propose a new control method which enables high precision operation by compensating for mechanical uncertainties. We verify the effectiveness of the proposed control method by extensive experimental tests to our developed TDJ manipulator system.