Multiple-Criteria Kinematic Optimization for the Design of Spherical Serial Mechanisms Using Genetic Algorithms

Abstract

A new kinematic design methodology is presented for optimization of spherical serial mechanisms. First, a new index, combining global manipulability and the uniformity of manipulability over the workspace, is presented to improve the synthesis results. This method integrates multiple criteria (workspace size, the new manipulability index, and mechanism size) linearly in one objective function. All these criteria are optimized simultaneously to lead to a solution with better performance. By changing the priorities of each criterion, different sets of desirable kinematic performance can be expressed. An adaptation of the method using a multiobjective Pareto front is also illustrated. The optimization result for a spherical bevel-geared mechanism using a genetic algorithm demonstrated that the proposed method effectively improves the quality of the optimum solution and provides insight into the workings of the mechanism. In addition, this flexible and adaptable methodology also presents a general optimization approach for linkage synthesis.