Tracking behaviours of cooperative robots within multi-agent domains
Fernando Ramos, Huberto Ayanegui · InTech eBooks · 2010
Autonomous Agents 46that it is able to discover satisfactory behaviors of strategic and tactical patterns as well as tracking the behaviors while the robots are evolving in a competitive complex task. Cooperative AgentsCooperative agents are focused on how a loosely-coupled network of problem solvers can work together to solve problems that are beyond their individual capabilities.Each problem-solving node in the network is capable of sophisticated problem-solving and can work independently, but the problems faced by the nodes cannot be completed without cooperation.Cooperation is necessary because no single node has sufficient expertise, resources, and information to solve a problem, and different nodes might have expertise for solving different parts of the problem (Durfee et al., 1989b).Multi-agent systems research is concerned with the wider problems of designing societies of autonomous agents, such as why and how agents cooperate (Wooldridge & Jennings, 1994); how agents can recognize and resolve conflicts (Adler et al., 1989; Galliers, 1988b;Galliers, 1990;Klein & Baskin, 1991;Lander et al, 1991); how agents can negotiate or compromise in situations where they are apparently at loggerheads (Ephrati & Rosenschein, 1993;Rosenschein and Zlotkin, 1994); and so on.An important concern is to design an appropriate organization of a multi-agent system for a particular domain and environment, such as described in the survey by (Horling & Lesser, 2004).In this work, advantages and disadvantages of these organizations are discussed.Such organizations can be hierarchies, holarchies, coalitions, teams, congregations, societies, federations, markets, and matrix organizations. Coordination of multi-agent systemsCoordination of tasks is an important aspect directly associated with the success of a plan that supports a strategy and/or tactic.In addition, the coordination plays an important role in the correct execution of cooperative actions, in such a way that conflictive situations could be avoided.In (Chernova & Veloso, 2008) a teacher instructs multiple robots to work together in coordination through a demonstration of the desired behavior, using a communication system and a sharing information system.Another distributed approach that share information, which help to facilitate the coordination and solve problems such as collisions is described in (Jansen & Sturtevant, 2008).One of the most important complex dynamic domains of application of multi-agent systems is the soccer robot systems.In (Candea et al., 2001), aspects of coordination of soccer robots within the framework of RoboCup are treated, doing emphasis in behavior based techniques.The communication and distributed coordination is addressed in this work.Multi-agent teamwork is critical in a large number of agent applications, including training, education, virtual enterprises and collective robotics (Nair et al., 2004).However, in multiagent domains, agent interactions become the domain highly complex for the analysis of agent-team behaviors, such is the case of robotic soccer.We consider "complex domains" to be those with enormous state action spaces, dynamic environment, competitive and real www.intechopen.