Skip to Main Content
As an important cooperation mechanism in multi-agent system (MAS), agent coalition has become a key topic in multi-agent research area. However, for a given task, it is difficult to search for the optimal agent coalition as the search space is exponentially dependent on the number of the agents in the system. This paper presents an ant colony optimization based algorithm which aims at finding the optimal, task-oriented agent coalition in MAS. A simple pheromone strategy and a type of sophisticated heuristic information are integrated in the algorithm. We evaluate the proposed algorithm through experimental study. The computational results show that the algorithm is efficient.