Human motion tracking control for humanoid robot based on the optimized motion retargeting

The main objective of this study is to provide a motion retargeting method for mapping the human motion to the humanoid robot. This paper describes the procedure to generate human-like upper limbs motion of a humanoid robot while maintaining balance. The human motions are acquired by the inertial measurement units (IMU). The motion retargeting method is implemented with four steps: kinematics modelling, inverse kinematics computation, enforcing constraints and difference optimization. Our method integrates multiply existing mature technologies to build a human-robot interaction system to imitate the human motion on a robot. Finally, the method is verified on a humanoid robot Nao to generate a human-like motion. Experimental results demonstrate that the method achieves a good kinematic match.