采用栅格理论的多机器人系统任务元管理方法  

Multi-robot mission cell management based on lattice theory

作　　者：李霞[1] 谢军[1] 

机构地区：[1]空军工程大学工程学院,西安710038

出　　处：《电光与控制》2008年第9期69-72,共4页Electronics Optics & Control

摘　　要：以目标为研究对象以及传统的传感器管理方法已难以满足多机器人系统中的多传感器管理研究,但如果将目标与系统任务状态结合为目标任务状态,以其对应的任务元作为研究对象会使复杂系统简单化,而基于栅格理论的管理网络也可以有效地管理不同层次不同类别的任务元,并且可以实时地监测多机器人协同工作系统任务元权重的变化情况。首先将栅格网络中每一层的目标元素对应为多机器人系统不同级别的工作任务;其次从顶层开始,逐一确定上层任务包含哪些下层任务;然后采用基于离差最大化的多属性决策方法确定任务元的权重概率,最后以一简单算例验证该管理方法的简单有效。Traditional sensor management method and taking the target as the research object can not satisfy the needs of multi-sensor management in multi-robot coordinate work system. Combining the target with function as the target mission condition and taking the corresponding mission cell as the research object will simplify a complex system. A mission cell management network based on lattice theory can effectively manage the mission cells of different levels and different types, and also can monitor the change of mission cell＇ s weights of the multi-robot system in real time. First, the target element in each layer of the lattice network is made corresponding to the mission of multi-robot system at different level. Then, the inclusion relation between the upper and the lower layer is conf＇m-ned, one layer after another beginning from the top layer. Finally, the mission cell＇ s weight probability is determined by using multi-property decislon-making based on maximum dispersion. A simple example demonstrated that this management method is efficient and simple.

关 键 词：多机器人协同工作系统 任务元管理 任务元权重 栅格理论 多属性决策 

分 类 号：V271.4[航空宇航科学与技术—飞行器设计] TP24[自动化与计算机技术—检测技术与自动化装置]