基于蚁群算法的移动机器人多目标路径规划

The Multi-objective Path Planning for Mobile Robot Based on Ant Colony Algorithm

迫于工作空间的限制以及对绿色生产理念的追求,在智能制造等领域人们通常需要机器人并行地执行多个任务,因此研究机器人的多目标路径规划更加符合实际需求。针对栅格模型中四、八邻域搜索方向较少的问题,提出了改进的十六邻域搜索方法;同时通过删除冗余转折点对路径进行了平滑处理,改善了路径存在的锯齿效果;结合蚁群优化算法与Dijkstra路径搜索算法,提出了一种多目标路径规划方法。在几种障碍环境中进行了测试,结果表明,上述算法能较好地适应各种不同的地图,即使是复杂度较高的地图,所提算法也能有效地找到一条较优的路径。

Due to the limitation of space and the pursuit of green production concept,robot is often used to perform multiple tasks in intelligent manufacturing.Therefore,the study on multi-objective path planning of robots is more in line with the actual needs.For this research,firstly,an improved sixteen-neighborhood searching method was proposed to solve the problem of less searching directions in the raster map,meanwhile,the path was smoothed by deleting redundant turning points,which weakens the sawtooth effect.Then,a multi-objective PP algorithm was proposed by combining ant colony algorithm with Dijkstra algorithm.Finally,the multi-objective PP algorithm was tested on several maps.As a result,this method can well adapt to various maps and effectively find an approximate optimal path even the map with a high complexity.