Chen Zhixiang  Yin Shuyan  Lu Zhenyang 

(College of Mechanical Engineering & Applied Electronics Technology, Beijing University of Technology, Beijing 100022)

Abstract: The coordinated path planning is lucubrated. At first, the welding path is represented by an array of quintuple. Being the foundation of welding path planning, welding position function, bead forming quality function, joint configuration function and smooth motion function are designed carefully as the objective functions for evaluating the path of the robot welding. Then, the weighted sum approach is used conveniently as the base idea to meet the multi-objective optimization. A genetic simulated annealing algorithm (GSA), which combined the advantages of the genetic algorithms and the simulated annealing algorithm, is employed to plan the coordinated path of robot and positioner for arc welding. Through coordinated path planning, robot can weld the workpiece in the best welding position and the best torch orientation, with smooth motion. Therefore, the coordinated path planning is able to improve the quality and the efficiency of robot arc welding.

Key words: Arc welding robot  Coordinated path planning  Genetic simulated annealing

CLC No: TG409

北京市科技基金 (H010310050111)和北京工业大学博士科研启动基金(200400080)资助项目. Received 20040602, received in revised form 20020041130