An Analysis on the Impact of Approximation of Scaling and Manipulability Approaches in Dimensional Synthesis of Robots

Dimensional synthesis of robots is often performed with approximated performance indices due to the large computation involved. These approximations often undermine the performance of the robot and hence a study on the impact of these approximations gives the designer a better insight on the trade-off between the robot’s performance and the computational investment. Motivated by this, the current study focuses on the impact of scaling of Jacobian and manipulability index by comparing the ellipsoid-based manipulability index with polytope-based manipulability index and the three-point scaling of Jacobian with linear-angular-separated conditioning of Jacobian. Optimisation problems are formulated for performance around a single point for four manipulators, and the results are presented. The results show that there is a significant impact on the approximation, which suggests the designers to adopt more accurate performance parameters for dimensional synthesis.