Comparison of metaheuristic optimization algorithms for dimensional synthesis of a spherical parallel manipulator

It is well known that the performance of parallel manipulators is highly dependent on their geometry, therefore the dimensional synthesis plays an important role in the design of a manipulator. In this work, the dimensional synthesis of a spherical parallel manipulator is carried out by means of a multi-objective optimization. This manipulator is planned to be a prosthetic wrist. The considered objective functions were the dexterity, the variability of the dexterity, the maximum required torque and the stiffness. The optimization was performed using three different metaheuristic algorithms: NSGA-II, OMOPSO, MOEA/D. After performing a set of simulations, the results show a clearly better performance of the MOEA/D algorithm. The optimized geometry of the manipulator allowed to design a virtual prototype of a prosthetic wrist.