The more work was to study learning and rule discovery theory of agent by applying the improved michigan method , and to analyze the method of agents how to gain the knowledge belief in the multi - agent environment in virtue of the multi - person game theory , as well as study the game tree structure of the agent ' s belief - desire - intention ( bdi ) model and agent how to decide 应用改进的密歇根法研究了主体的学习和规则发现的理论与应用，同时借助多人博弈结构理论，分析了主体在多主体环境下知识信念的获取方法，以及主体bdi模型的博弈树结构和主体决策机制。
