Title :
Optimal Design of Parallel Robots for the Prescribed Regular Dexterous Workspace
Author :
Wang, Zhongfei ; Ji, Shiming ; Wan, Yuehua ; Ou, Changjin ; Sun, Jianhui ; Wang, Guan
Author_Institution :
Zhejiang Univ. of Technol., Hangzhou
Abstract :
This paper presents a method for determining design parameters of parallel robots (PRs) in its optimal kinematic design. To produce a more compact and economical PR, the determination of an optimal workspace is important. This includes defining a workspace that encases the prescribed regular dexterous workspace (PRDW). Two requirements should be simultaneously satisfied: (i) the local or global performance index; (ii) a regular shape. In this paper, the optimal problem is firstly discussed and translated into two linear problems. Then, an optimal design procedure which adopts the generalized pattern search algorithm in the genetic algorithm and direct search toolbox of Matlab is proposed to solve these problems. Taken as an example, a linear Delta robot is designed for the four different PRDWs, and its optimal result provides some useful information in measuring the distance between a PRDW and the maximal sized normalized robot. On the basis of the value of distance, some wrong solutions of PR are clarified. It is concluded that the optimal design method is useful for the design and comparison of a general PR.
Keywords :
dexterous manipulators; genetic algorithms; manipulator kinematics; search problems; direct search toolbox; generalized pattern search algorithm; genetic algorithm; global performance index; optimal kinematic design; parallel robot optimal design; prescribed regular dexterous workspace; Algorithm design and analysis; Design automation; Design methodology; Design optimization; Genetic algorithms; Kinematics; Logistics; Manufacturing automation; Parallel robots; Robotics and automation; Generalized Pattern Search; Optimal Design; Parallel Robot; Prescribed Regular Dexterous Workspace;
Conference_Titel :
Automation and Logistics, 2007 IEEE International Conference on
Conference_Location :
Jinan
Print_ISBN :
978-1-4244-1531-1
DOI :
10.1109/ICAL.2007.4338628