Numerical Optimal Control of a Spherical Parallel Manipulator Based on Unique Kinematic Solutions

This paper presents a framework for generating optimal motor trajectories for a spherical parallel manipulator (SPM) with revolute joints, actuated by servomotors with default internal position control settings. The proposed framework consists of three phases. First, an approach to obtain unique forward kinematics is introduced, in order to relate the angular positions of the servomotors to the orientation of the SPM top mobile platform. Then, a configuration space for the SPM is defined by using numerical procedures, in order to guarantee the absence of singularities and of collisions between links during the motion of the manipulator. Finally, reference trajectories of the servomotors are defined via convex optimization. These trajectories determine an optimal evolution of the SPM motion, based on configuration space and original servomotor dynamics. The proposed strategy is experimentally demonstrated on a prototype of Agile Wrist SPM with three servomotors.