摘要
针对BDI模型在不确定领域应用的薄弱,提出了一种改进的BDI(Beliefs—Desires—Intentions)模型。以BDI体系结构为基础,结合贝叶斯网络,建立了智能体对可能世界的认知表示;通过贝叶斯网络推理,实现了对达成目标状态的期望估计;在意图中,通过引入规划因子,完成行为决策。用Java语言在eclipse平台上对农作物栽培实例进行仿真,验证了模型能够在不确定环境中,进行理性的认知、推理、规划。
As the BDI(Beliefs-Desires-Intentions)model is weak in uncertain field, a kind of improved BDI model is proposed.Based on BDI system structure, combined with the Bayesian network, agent established cognitive representation ofpossibleworld; using the Bayesian network inference, realize the target state estimation to reach expectations; In intentions,it introduces the planning factor to complete the behavior decision. The crop cultivation examples emulate on the Eclipse platformusing Java language, the model is verfied to be able to cognize, reason and plan rationally in an uncertain environment.
作者
胡晓辉
刘全
王元鹏
HU Xiaohui;LIU Quan;WANG Yuanpeng(College of Electronic and Information, Lanzhou Jiaotong University, Lanzhou 730070, China)
出处
《计算机工程与应用》
CSCD
北大核心
2016年第16期73-78,共6页
Computer Engineering and Applications
