Load distribution algorithms for redundantly actuated manipulator resembling the human upper-extremity

This paper presents the optimization of muscle forces in the human upper extremity for the desired trajectory. The proposed kinematic model of the human upper extremity is a force redundant mechanism, implying that number of actuators are more than number of outputs. During the trajectory, the torque is generated for each joint by antagonistic activation of muscles. The biological actuators (muscles) can only generate forces in one direction (unidirectional). Based on this fact and considering the maximum forces limit for each muscle, we proposed a dynamic optimization technique in which only positive muscles forces not exceeding their maximum limits are generated to follow the desired trajectory.