Feedforward tracking control for lower-limb robot with antagonistic driven mechanism

This paper addresses a control issue for lower-limb robots with antagonistic actuated joints. The conventional control method has previously been used to compensate for the steady-state error with respect to a reference trajectory. However, the question of how to improve the tracking performance with respect to a reference trajectory remains a problem that needs to be solved. This paper proposes two feedforward (FF) control methods: based on a phase-lead filter and based on the inverse dynamics of the plant. The two proposed FF control methods significantly improve the tracking performance with respect to a reference trajectory. The effectiveness of the proposed control methods was confirmed by numerical simulations of a lower-limb robot with antagonistic actuated joints.