作者投稿和查稿 主编审稿 专家审稿 编委审稿 远程编辑

计算机工程 ›› 2010, Vol. 36 ›› Issue (24): 116-118. doi: 10.3969/j.issn.1000-3428.2010.24.041

• 安全技术 • 上一篇    下一篇

多Agent合作中的信任模型研究

隋 新,蔡国永,史 磊,罗柏发   

  1. (桂林电子科技大学计算机科学与工程学院,广西 桂林 541004)
  • 出版日期:2010-12-20 发布日期:2010-12-14
  • 作者简介:隋 新(1983-),男,硕士研究生,主研方向:分布式人工智能,多智能体系统,信任模型;蔡国永,教授、博士;史 磊、罗柏发,硕士研究生
  • 基金资助:
    广西自然科学基金资助项目(0728089);广西研究生教育创新计划基金资助项目(2009105950812M18)

Research on Trust Model in Multi-Agent Cooperation

SUI Xin, CAI Guo-yong, SHI Lei, LUO Bai-fa   

  1. (School of Computer and Control, Guilin University of Electronic Technology, Guiling 541004, China)
  • Online:2010-12-20 Published:2010-12-14

摘要: 传统的信任模型研究通常建立在具有完全历史合作信息的基础上,难以满足动态演化环境中的合作需求。针对缺乏直接合作历史信息的情况,借鉴现代电子商务中第三方管理思想,分析基于第三方的间接信任模式,结合贝叶斯信任模型,提出DRI信任模型。通过与传统信任模型的对比实验结果表明,基于DRI信任模型的合作能提高合作对象选取的可靠性,提高平均合作成功率。

关键词: 多Agent, 合作, 信任模型, 第三方

Abstract: Traditional research on trust model is based on environment with history cooperative information, it can hardly meet the needs of dynamic evolution of the cooperative environment. Aiming at situations which are without direct history information, indirect trust model based on the third-party organization is analyzed by referring to the third-party management method in modern E-commerce, and a DRI(Direct, Recommending and Indirect) trust model is proposed by combining with Bayesian trust model. Experimental result by comparing with traditional trust model shows that the cooperation based on DRI trust model can effectively improve the reliability of cooperative object selection, and improve the average successful rates of cooperation.

Key words: multi-Agent, cooperation, trust model, third-party

中图分类号: