Scheduling of arc welding robot considering heat-caused distortion

Abstract

In this thesis, we focus arc welding robot-scheduling problem considering heat-caused distortion with the objective of minimizing total cycle time. Since this problem is known as a special case of (with additional constraints) traveling salesman problem, it is difficult to solve within a polynomial time. Therefore, we develop a heuristic algorithm for finding the schedule-information in which total cycle time is minimized within a polynomial time. The heuristic is a two-phase method which combines tour construction phase and tour improvement phase. To evaluate its performance, we compare the proposed heuristic with the existing heuristics after modification that is necessary in the problem. Computational results based on randomly generated problems show that the proposed heuristic dominates all those existing heuristics.