基于MAS的多机器人系统及其协作机制研究

【作者】 吴旭华；

【导师】 王润孝；

【作者基本信息】 西北工业大学 ， 机械电子工程， 2004， 硕士

【摘要】 20世纪80年代以来，随着计算机、控制、传感等技术和人工智能理论的发展，机器人技术得到日益广泛的应用与深入研究，机器人的应用方式开始从单体式应用转向系统式应用。在实际生产工程系统中，作为群体工作的机器人群已出现在汽车工业、工程机械、建筑等行业和部门中，多机器人系统也逐渐成为机器人技术应用与研究的热点和趋势。 本文在深入分析多机器人系统的发展前景及其研究现状的基础上，提出可以充分借鉴和利用当前人工智能领域的研究成果，特别是分布式人工智能领域中的MAS(Multi-Agent System，多Agent系统)理论，进行多机器人系统的研究，并从此认识出发开展了基于MAS理论的多机器人系统的研究。论文得到西北工业大学研究生创业种子基金(编号：Z20030058P)的资助，具体完成了如下工作： 1．论文总结了当前有关多机器人系统的研究，指出存在的不足，并根据多机器人技术领域的研究与发展趋势，针对多机器人系统的分布式问题求解特点，深入分析和研究了分布式人工智能领域中的MAS理论。 2．论文在对MAS理论深入研究的基础上，完成了基于MAS理论的多机器人系统结构分析，构建了一种针对复杂不确定环境下复杂动态任务的多机器人系统结构框架。提出了一种两级系统层次结构模型，并建立了一种混合式机器人Agent结构模型。 3．论文对当前关于多机器人系统的关键问题——协调协作机制、通信以及学习问题的研究进行了深入分析与总结，明确了各种方法的应用特点和优势，提出可以充分借鉴MAS理论中的相关研究内容和研究成果，实现多机器人系统中同类问题的解决。 4．论文深入研究了多机器人系统的协作机制，并将系统内机器人间的协作问题转化为系统优化和自适应问题，同时在分析作为当前应用较多的优化与自适应问题求解方法——遗传算法的基础上，将进化计算方法应用到系统的协作机制中，具体解决了协作中的任务分配与调度问题，实现了算法设计与仿真验证。 本论文在研究内容和研究方法上具有一定的创新性和开拓性，对多机器人技术的深入研究有较高的参考价值，并为多机器人系统的深入研究和未来的推广应用做了基础工作。更多还原

【Abstract】 With the development of computer technology, controlling technology, sensing technology, artificial intelligence theory, and etc., the robot technology is getting the development advancing by leaps and bounds. Since the 1980’s, with the development and wide application of the robot technology, the application of robots begins to change from the single style to the system style. The groups of robots which work together as a production engineering system have appeared in auto industry, engineering machinery, building trade, etc. And multi-robot systems are also becoming the focus and trend of the study and application of the robot technology.Based on analyzing the development prospect and current research of the multi-robot system in depth, this thesis demonstrates that it’s reasonable and effective to make full use of research results in present artificial intelligence field, especially the theory of MAS (Multi-agent system) in the distributed artificial intelligence field for the research of the multi-robot system. In terms of this opinion, the thesis carries on the research of the multi-robot system based on the theory of MAS. This thesis is supported by the Innovation Seed Fund for Graduate Student of Northwest Polytechnical University (NO.Z20030058P).The detailed content in the thesis is described in follows:1. The thesis summarizes the present research on the multi-robot system and presents the deficiency in the research. According to the trend of the research and development in the multi-robot technology field, it also further investigates the theory of MAS in the distributed artificial intelligence field in terms of the characteristics of distributed problems in the multi-robot system.1. Based on the further investigation of the MAS theory, the thesis completes the research on the systematic structure of the multi-robot system, and designs a system structure model of the multi-robot system in the dynamic and complicated uncertain environment, and put forward a systematic hierarchical structure model and a kind of composite robot Agent structure.2. The thesis analyses further the present research on the Key problems of multi-robot systems, such as coordination&cooperation, communication and study problem, and defines the application characteristics and advantages of various methods of solving these problems. It also presents an opinion of making full use ofrelevant research contents and results in the theory of MAS to solve those similar problems in multi-robot systems.3. The thesis researches further the cooperation mechanisms of the multi-robot system, and turns the systematic cooperation problem into a system optimization and self-adaptive problem. On the basis of analyzing the theory of GA (Genetic Algorithms) which is a popular method solving optimal and adaptive problems, it also applies the EC (Evolutionary Computation) to the systemic cooperation mechanism, and complete the design and simulation of the algorithm.It is of innovation and pioneering on the research contents and research approach, and there is higher reference value on further research of multi-robot systems. The thesis also completes some fundamental work to the further research and popular application of multi-robot systems in the future.更多还原