【摘要】 针对Agent个体学习与群体学习各自的不足,探讨了多智能体系统(MAS)中的合作与协商及Agent学习技术,提出了一种新的混合智能学习算法.将个体学习与群体学习有效结合起来,提高了Agent的个体性能及系统整体的智能水平.在足球机器人仿真系统中进行了实验,结果表明了算法的可行性与有效性.更多还原

【Abstract】 Multiagent System is a mayor branch of Distributed Artificial Intelligence (DAI) and one of its research focuses is agent learning algorithm. To solve more complex problems appeared nowadays, cooperation & negotiation and agent learning techniques in MAS are studied, a new hybrid intelligent learning algorithm is proposed to combine individual learning & group learning effectively and improve agent’s individual ability and the system’s intelligence level. The specific application and experimental results in robot soccer prove the feasibility and validity of this algorithm.更多还原