基于改进蚁群算法的机械臂焊接路径规划

Welding Path Planning of Manipulator Based on Improved Ant Colony Algorithm

针对传统蚁群算法在求解机械臂焊接路径过程中容易陷入局部最优,导致机械臂移动路径长、效率低等问题,提出一种改进蚁群算法。将焊点布局图进行转换并引入粒子群算法,通过编码的方法进行全局焊点路径寻优,引入全局最优路径片段及局部最优路径片段,改变信息素的更新方式。确定机械臂可视化范围并通过视觉反馈系统反馈障碍坐标,利用三维蚁群算法求解不同焊点间的路径,避免障碍发生。仿真结果表明,改进前算法及改进后算法规划路径长度分别为16379.5 mm和16298.6 mm,相比于改进前算法,改进后算法的路径长度缩短了0.47%,迭代次数减少了11.7%,同时改进算法可以在软件中应用且具有稳定性。

Aiming at the problems that the traditional ant colony algorithm is easy to fall into local optimization in solving the welding path of the manipulator,resulting in the long moving path and low efficiency of the manipulator,an improved ant colony algorithm is proposed.The solder joint layout was transformed and particle swarm optimization algorithm was introduced to optimize the global solder joint path by coding method,and the global optimal path segment and local optimal path segment were introduced to change the update mode of pheromone.Determining the visual range of the manipulator and feeding back the obstacle coordinates through the visual feedback system,the path between different solder joints was solved by three-dimensional ant colony algorithm to avoid interference.The simulation results showed that the planned path lengths of the unimproved algorithm and the improved algorithm were 16379.5 mm and 16298.6 mm respectively.Compared with the unimproved algorithm,the path length of the improved algorithm was shortened by 0.47%and the number of iterations was reduced by 11.7%.At the same time,the improved algorithm could be applied in software and had stability.