Raimond, Kumudha (PhD)Asnake, Nardos2018-06-292023-11-042018-06-292023-11-042005-03http://etd.aau.edu.et/handle/123456789/4754An agent oriented software engineering paradigm has been applied in solving engineering problems. In this work, agent architecture is developed that is suitable to solve a given problem. The system is developed progressively beginning from simple expert system to a system where agents learn to solve a given task based on some performance measure and agent architecture. Self-organization, emergence of global behavior and learning at population are considered while designing the agent architecture. Different ways of interaction and cooperation between agents are used to realize the collaboration among them. The transfer of problem solving task from human being to agents is clearly elaborated in the test scenario chosen for the worenOptimizationIncrementally Autonomous Light Weight Agent Architectures for Optimization TaskThesis