群體機器人之自主導航研究

• Main Authors: Hsieh Chung Hung, 謝忠宏
• Other Authors: 唐天生
• Format: Others
• Language: zh-TW
• Published: 2006
• Online Access: http://ndltd.ncl.edu.tw/handle/90287225437667413113

碩士 === 國防大學中正理工學院 === 兵器系統工程研究所 === 94 === ABSTRACT This research is aimed at autonomous robot agent navigation; we adopt the design of distributed artificial intelligence to consider three kinds of behaviors of the autonomous robot agent by software agent techniques. We firstly consider that the single-agent system (SAS) living in artificial life environment independently. We extend SAS to multi-agent system (MAS) designed that operates together with their own decision mechanisms. So we use different algorithms of performing three kinds of decision-making mechanism for agent behaviors. From simple reactive behavior performed by natural network (using winner-take-all learning algorithm) to the complex behavior involves the influence of each other of the behavior among multiagent, which perform proactive and social behavior using fuzzy logic for their decision-making mechanism. The decision-making process uses fuzzy rules for coordination and action selection can enable the agents to survive in the unknown environment and extent the technology to research the navigation of the autonomous robot agents.