Zhengzheng Wan
Cooperative multi-agent robotic systems are ones within which several homogeneous or heterogeneous robots,through their cooperation, jointly solve problems, attain to common or individual goals or maximize their utilities. Modern distributed artificial intelligent robotic systems is rapidly developing with increased research efforts putting into approaches on more cooperative, better-organized, strong unknown environment adaptive multi-agents systems, However, problems caused by dramatically rising of complexity and much pre-examination appeared. In this talk, I will present a hierarchical organized multi-agent robotic system. Distribute each behavior-based robot to a specific colony to create a parent-children relationship between cooperative group members and their leader, All the robot sub-colonies are well managed through communication among each group member and the group supervisor. Motivation behind the research on hierarchical self-organization comes from number of advantages it possesses over traditional distributed cooperative system. Once agents are released into an unknown environment, by grouping or self-grouping agents hierarchically, elements can cooperate with each other to complete some specified tasks for human being in different portions of a very large unknown environment parallelly, therefore, large communication complexity will be highly reduced, cooperation in more complicated circumstances will be efficiently done. Additionally, due to individual subsumption architectured control systems, also named neuron network are plugged in each agent, intelligent behaviors can be processed simultaneously, emergency handling and artificial learning can be done in an effective way. A few examples will be given during the presentation.