基于DGSOM_A*的移动机器人地图创建和路径规划

Map Building and Path Planning for Mobile Robots Based on DGSOM_A*

针对移动机器人未知环境路径规划问题,基于动态自组织特征映射网络提出了一种自组织网络动态生成A*的算法(dynamic growing self-organizing map with A*,DGSOM_A*),并将其应用于移动机器人地图创建和路径规划.该方法利用Mobotsim二维仿真软件构造了环境模型,机器人通过无碰自由巡航获取环境信息,然后把上一步得到的环境信息作为DGSOM_A*算法样本通过SOM神经元自主生长进行地图创建,生成以少数SOM图神经元分布描述环境特征信息的拓扑地图,最后完成起始点到目标点的导航任务.实验结果表明,相比传统的SOM算法,基于DGSOM_A*算法机器人能有效地通过对环境地图的绘制熟悉复杂环境并能实现最优路径选取.

Aiming at the problems in unknown environment for mobile robots path planning, a dynamic growing self-organizing map with A ＊ （ DGSOM_A ＊ ） algorithm was proposed and used in the map building and path planning. Firstly, the method used a two-dimentional simulation software Mobotsim to obtain environmental information samples for DGSOM_A ＊ by moving the robot without collision, and then built the environment map through continuously increasing the new SOM neurons with the robot movement, thus generated the topology map to describe the environment information with a few SOM neurons. Finally, it completed navigation by finding an optimal path from the start point to the target. The experimental results show that the proposed DGSOM_A ＊ method detects the complex desktop environment effectively and correctly and the robot can realize environment mapping automatically by comparing to the SOM algorithm.