高级检索
    陶海军 王亚东 郭茂祖 王翰伦. 基于熟人联盟及扩充合同网协议的多智能体协商模型[J]. 计算机研究与发展, 2006, 43(7): 1155-1160.
    引用本文: 陶海军 王亚东 郭茂祖 王翰伦. 基于熟人联盟及扩充合同网协议的多智能体协商模型[J]. 计算机研究与发展, 2006, 43(7): 1155-1160.
    Tao Haijun, Wang Yadong, Guo Maozu, and Wang Hanlun. A Multi-Agent Negotiation Model Based on Acquaintance Coalition and Extended Contract Net Protocol[J]. Journal of Computer Research and Development, 2006, 43(7): 1155-1160.
    Citation: Tao Haijun, Wang Yadong, Guo Maozu, and Wang Hanlun. A Multi-Agent Negotiation Model Based on Acquaintance Coalition and Extended Contract Net Protocol[J]. Journal of Computer Research and Development, 2006, 43(7): 1155-1160.

    基于熟人联盟及扩充合同网协议的多智能体协商模型

    A Multi-Agent Negotiation Model Based on Acquaintance Coalition and Extended Contract Net Protocol

    • 摘要: 合同网协议可扩充性好,处理动态环境能力强,在多智能体系统协商中应用广泛.在分析了经典合同网协议的优缺点后,提出了基于熟人联盟及扩充合同网协议的多智能体系统协商模型.根据模型设计了适合扩充合同网协议的系统结构,引入了熟人联盟以及信任度参数,提出熟人联盟生成方法及信任度更新规则并构造了基于经典合同网协议的扩充合同网协议.最后通过对一个导弹防御例子的测试及分析,证实了该模型在保证协商质量的基础上,有效地降低了协商代价.

       

      Abstract: Contract net protocol is widely used in multi-agent systems because of its highextensibility and flexibility in a dynamic environment. However, many shortages exist while applying the traditional contract net protocol in multi-agent systems, such as rapidly increasing of the negotiation cost along with the system's dimension, no guarantee of the negotiation quality and difficulties in adapting the dynamic environment, etc. A multi-agent negotiation model based on the acquaintance coalition and the extended contract net protocol are presented to improve the efficiency of negotiation. The structure of the multi-agent negotiation model is given to support the extended contract net protocol. The acquaintance coalition policy, the trust degree parameter and the adjustment rule are introduced and used to construct the extended contract net protocol. Finally, through an example and analysis of a missile defense system which uses the model, improvement of the negotiation efficiency is proven.

       

    /

    返回文章
    返回