基于市场法的多机器人协作未知环境探索  

Unknown Environment Exploration of Multi-robot Based on Market

作　　者：赵慧潮 石朝侠[1] 

机构地区：[1]南京理工大学计算机科学与工程学院,南京210094

出　　处：《计算机与数字工程》2017年第11期2085-2089,共5页Computer & Digital Engineering

基　　金：国家自然科学基金项目"基于XX环境搜索面上研究"(编号:61371040)资助

摘　　要：多任务分配是多机器人未知环境协作探索的关键问题。传统市场法由于追求单体机器人最优化,从而牺牲了整体最优化,而基于蚁群算法的多任务分配方法虽然实现了整体最优化,但是不适合在未知环境探索中使用。论文在市场法的基础上提出了一种改进方法,该方法将机器人到任务点所经过路径上的排斥素的多少作为拍卖获胜的条件。该方法提高了多机器人系统完成环境探索的效率。通过实验验证该算法的有效性。Multi-task assignment is a key problem in multi-robot..s collaborative exploration of unknown environment. Tradi.tional market methods sacrifice the overall optimization because of the pursuit of single-robot..s optimization. However,the methodbasing on the partition exploration can achieve the overall optimization,which is not suitable for exploration in the unknown environ.ment. This paper put forward a market approach based on improved the traditional market approach,this method took the number ofrejection pheromones in the path through the robot as the condition of winning the auction. The method improves the efficiency of themulti-robot system to complete the environment exploration. What..s more,the effectiveness of the algorithm has been verified by ex.periments.

关 键 词：多机器人 协作探索 市场法 排斥素 

分 类 号：TP242.6[自动化与计算机技术—检测技术与自动化装置]