基于改进蚁群算法激光导航轮式机器人路径规划

针对激光导航轮式机器人在复杂环境中路径规划原始算法存在路径较长和收敛速度较慢的问题，提出了一种改进蚁群算法。在实际算法中，先利用MAKLINK图论建立AGV运行环境的空间模型，接着用Dijkstra算法搜索优化路径；然后，在Dijkstra算法的基础上采用蚁群算法搜索最优路径；紧接着，在改进蚁群算法中，优先选择搜索前后两节点同起点到终点夹角一致或相差不大的后一个搜索节点，获取新的信息素更新策略，并进行角度的初始化和信息素计算；最后，在Matlab上完成算法的编写并得到仿真结果。结果表明，改进蚁群算法路径优化性能更好，对实际环境中机器人的路径规划具有指导意义。

Path Planning for Laser Navigation Wheeled Robots Based on Improved Ant Colony Algorithm

This article puts forward an improved ant colony algorithm to deal with the problems of long planning path and low convergence rate present in the primal algorithm for path planning for the laser navigation wheeled robots in a complex environment. In actual algorithms, the graph theory of MAKLINK is used to establish the space model of the operating environment of AGV. Then the Dijkstra algorithm is used to search the optimal path, and the ant colony algorithm is used to search the optimal path based on Dijkstra. In the improved ant colony algorithm, by judging whether angle between previous and latter node is same or little different to angle between original and terminal node, the next node will be the latter one, thus obtaining new strategy of updated pheromone and allowing initialization of angle and calculation of pheromone. Finally, the whole algorithm is written on Matlab and simulated results are obtained. Results show that the path optimization performance in the improved ant colony algorithm is better, and has guiding significance in path planning for laser navigation wheeled robots in the real environment.