基于“虚拟范围”的多机器人围捕算法

 讨论了基于SQL Server数据库的栅格地图建立方法。介绍了一种建立在势场法和栅格地图上的室外多机器人路径规划方法，分析了在已知地图中应用势场法进行路径规划时的路径死点和规划失败问题，并给出了解决方法。以势场栅格法为基础探讨了多机器人协作围捕动态目标的策略。在围捕算法中，根据室外机器人工作环境的特点，提出了“虚拟范围”的概念，以减少动态规划次数，提高围捕速度。介绍了机器人以“虚拟范围”为基准，在各种状态之间的转换机制。利用基于无线局域网的室外多机器人系统进行了试验，试验表明引入“虚拟范围”之后能有效缩短围捕时间，并且存在最短时“虚拟范围”。


Hunting Algorithm for Multi-mobile Robot System Based on Virtual Range

At first,the method of constructing a grid map is discussed by using SQL server database.Secondly,the path planning based on potential grid method is presented for a multi-mobile robot system.The deadpoint and planning failure problems are analyzed when the potential grid method is used in a known map,and the answer is presented.Furthermore,according to the potential grid method,the strategy of cooperatively hunting a dynamic object is discussed for the multi-mobile robot system.The concept of Virtual Range are presented to reduce the times of dynamic path planning and increase the hunting speed,which is suitable for robots moving in the outdoor environment.The Virtual Range is used as a criterion to trigger the change of the robot states which are respectively introduced.Furthermore,a real mult-mobile robot system based on wireless LAN is used to validate the algorithm.The test result shows that a reasonable Virtual Range can shorten the surrounding time,and that there is a best Virtual Rang for a certain multi-mobile robot system.