摘要
为了达到更有效率的全域覆盖路径规划,研究一种基于云计算的多机器人控制方法应用于垃圾搜集任务;先由服务器端口处理收集任务相关资讯并计算路径,使用类神经网络ANN来产生每台机器人的路径;采用多机器人同时进行巡航,以节省全域路径规划的时间,在机器人的控制与协调方面则采用云计算的概念,用以指挥机器人巡航的任务;模拟实验测试了1~4台机器人在一个无障碍区域的任务规划,机器人每一步伐以4个方格为单位,IPSO导航时则视情况以1~7个方格为单位;实验发现每增加一台机器人所需的单位移动时间,将会以接近1/N(N为机器人数量)的方式缩短;将实际效能与期望效能做比较结果也如此;证明适当数量的多机器人的执行效率优于单机器人。
In order to achieve more efficient global coverage path planning, research a kind of multiple robot control method based on cloud computing is applied to garbage collection. First by the server port processing tasks related to information and calculate the paths, the neural network ANN is used to produce each robot path. Multi robot cruise at the same time, in order to save the global path planning time. In the aspect of robot control and coordination, the use of the cloud computing concept, to command the robot cruise task. Simula- tion test for 1--4 robot in an accessible area of mission planning, the robot each pace is an unit with four squares, IPSO navigation depends on the situation in 1 to 7 squares. It was found that each additional unit of moving the time required for a robot, will close to 1 / N (N num- ber of robots) to shorten. The actual effectiveness and expected performance comparison results. Proved that the proper number of multi-- robot execution efficiency is better than that of single robot.
出处
《计算机测量与控制》
北大核心
2014年第4期1130-1133,共4页
Computer Measurement &Control
关键词
云计算
多机器人控制
类神经网络
路径规划
cloud computing
robot control
artificial neural network
the path planning