부산대학교 도서관

In the welding process of large and complex components, the welding gun must go through all the welds to complete the work, and the working path of the welding gun is often not unique. Finding a path that satisfies the welding process constraints and is as short as possible in these paths can significantly improve the welding efficiency. In view of the traditional ant colony algorithm in path planning problems, due to factors such as terrain environment, the planned path is not practical, and when faced with large-scale problems, the traditional ant colony algorithm often reflects the slow convergence speed, easy to fall into local optimum, and it is difficult to jump out of local optimum. The shortcomings of the proposed welding robot path planning based on improved ant colony algorithm. In this paper, the grid method is used to segment the environment and identify the obstacles in the program in the form of 0 / 1 matrix, so that the ant colony in the algorithm can have the ability to avoid obstacles. By improving the roulette method in the traditional ant colony algorithm, the next movement scheme searched by the roulette method is not limited to the range of nine grids, and the geometric judgment of obstacles, target points and starting points makes the path searched by the algorithm more direct. Through Matlab simulation experiments, the improved ant colony algorithm can plan a better global path more quickly. It can avoid the robot falling into local optimum, and ensure that the robot avoids obstacles while planning the shortest path.