Genetic algorithms to optimal design of a 3 DOF parallel robot

Abstract

The paper presents a multi-objective optimum design procedure to a 3 degrees of freedom (DOF) parallel robot with regards to four optimality criteria: workspace boundary, transmission quality index, stiffness and numerical aspects. Additional objective functions can be used to extend the proposed design procedure to more general but specific design problems. The main aspects of their structure, kinematics, workspace, parameter design and optimization indices are introduced. A kinematic optimization was performed to maximize the workspace of the parallel robot. In order to perform an optimal design of 3 DOF parallel robots, an objective function was developed first, and then Genetic Algorithms applied for optimization. The experimental results demonstrate the advantages of the presented optimal design method in design of a 3 DOF parallel robot. The performances of the 3 DOF translation parallel robot provide high potential and good prospect for their practical implementation in manufacturing industry.