CMM-Based Cooperative Control Strategy of Supernumerary Robotic Limbs for Human Motion.

The Supernumerary Robotic Limbs (SRL) have been proven to cooperate with the wearer to perform tasks in manufacturing domains such as aircraft assembly, mainly by providing support to the wearer through physical contact with the environment, thus reducing human energy consumption. However, current SRL control strategies cannot provide synchronous motion with the wearer’s motion, which causes interference and requires the wearer to expend more energy to maintain the human-SRL system balance. Therefore, in order to meet the demand for synchronized human and SRL motion, this study proposes a centroid-momentum-matrix (CMM)-based SRL control strategy to effectively reduce the human energy consumption caused by the non-cooperative motion of the SRL with the human motion. The method indirectly controls the interaction forces and moments between the SRL and the wearer by solving a quadratic programming problem to control the rate of change of the linear and angular momentum of the SRL. Finally, experiments on the bending motion of the human trunk were conducted to validate the proposed control method by measuring the interaction forces and moments as well as the surface electromyography of the erector spinae muscles, which demonstrating the effectiveness of the proposed control method.