Multi-agent coordination method based on fuzzy Q-learning

Traditional reinforcement learning algorithm can only solve the learning problem of the intelligent agent with discrete state space and discrete action space. This paper studies the coordination of multiple intelligent agents in a complicated dynamic environment with uncertainty. A coordination model based on the fuzzy Q-learning technique is suggested. This model uses fuzzy logic to generalize the agentpsilas continuous state space. Every agent, when making decisions on its actions, needs to consider the influences of other agents to the environment. The agent first evaluates the actions they select, then, uses the fuzzy Q-learning to learn their action strategy. In the process, the action keeps improving and the conflicts among agents can be resolved. This model was used in the RoboCup soccer simulation game and the simulation results showed that the performance of attacking is obviously improved.