将个性行为绑定到信息Agent上,提出了一种基于个性的协同强化学习算法,让不同个性的Agent充当合适的角色.基于共同的目标和相同的信息,Agent可能产生共同的意图,规划出共同的行为,使得MAS可以适用于动态、实时、有干扰、对抗的环境中.
The paper applys personality and behavior to the Information Agent, introduces the algorithm of cooperative reinforcement learning based on personality and makes different personality Agent as suitable role based on common objectives and the same information. Agent may have a common intention to develop a common Code of conduct, making MAS can be applied to a dynamic, real-time, interferential and confrontational environment.