摘要
RoboCup机器人足球比赛是当前人工智能领域的一个研究热点,其目的在于为多智能体系统提供一个标准的研究平台。为了让RoboCup仿真比赛中球员智能体实时地作出合理的动作决策,提出一种基于动作序列模型的决策机制,通过对球员智能体的动作空间分解、建立动作序列并对动作序列进行评价,而让智能体选择出当前最优的动作执行。仿真结果表明应用这种决策机制提高了智能体对环境的适应性。
Robocup is a research hotspot in artificial intelligence field, which aims to provide a standard research platform for multiagent system. An action sequence model based decision mechanism is proposed in this paper for agents making reasonable action decision. By decomposing the action space, establishing action sequence and evaluating action sequence, current optimal action can be chosen to execute. Simulation results indicate that with application of applying this decision mechanism, environment adaptability of agents is enhanced.
出处
《微型机与应用》
2013年第7期89-91,94,共4页
Microcomputer & Its Applications
基金
扬州职业大学教研项目(编号:07202)