面向分布协同控制研究的仿真环境设计与实现  

Design and Implementation of Simulation Environment for Distributed Cooperation Control

作　　者：吴军[1] 徐昕[1] 宋金泽[1] 刘春明[1] 贺汉根[1] 

机构地区：[1]国防科技大学机电工程与自动化学院,长沙410073

出　　处：《系统仿真学报》2009年第19期6039-6042,共4页Journal of System Simulation

基　　金：霍英东青年教师基金优选资助课题(114005);湖南省自然科学基金(07JJ3122);国家自然科学基金(60774076)

摘　　要：针对分布式系统,尤其是多机器人系统中协同控制问题的研究需要,提出了一种具有分层嵌套结构的多Agent仿真环境设计方案,以克服现有仿真环境存在的兼容性差和算法移植困难的问题。设计方案在系统总体结构中引入了层次多Agent的设计思想;在子系统中采用Acromovi结构框架设计;在个体Agent中采用了混合型的内部结构设计。具体应用于多机器人仿真环境设计中,通过不同的实验配置方案,逐步实现多机器人系统的虚拟仿真和半实物仿真。最终的多机器人编队控制实验表明了设计方案的有效性。A layered and nested multi-agent simulation environment design was proposed for cooperation control study on distributed system, especially on a multi-robot system, in order to overcome the incompatibility and algorithm transplant problem in current simulation environment. In the design, a layered method was adopted in system framework, then an Acromovi framework was introduced in the subsystem design, and a hybrid construction was adopted in standalone agent design. By using the above scheme and different experimental configurations, the virtual simulation and HIL simulation of a multi-robot system were realized step-by-step. Finally, the feasibility of the scheme is validated by results of multi-robot formation control trials.

关 键 词：多智能体 分布协同控制 Acromovi 多机器人 半实物仿真 

分 类 号：TP391.9[自动化与计算机技术—计算机应用技术]