-
摘要: 本文提出了一个基于传统BDI结构的新型Agent模型,即:BDIM模型。通过在传统的BDI结构模型基础上追加激励算子(motivation)试图解决传统的BDI结构在描述自学习机制和知识动态维护方面的局限性。本文在Rao和Georgeff基于扩展计算树逻辑对BDI结构形式化建模的基础上进行了扩展,把激励相关算子无缝地嵌入到一个已经比较完善的形式化体系结构中.Abstract: In this paper, we brought forward a novel modeling architecture of Agent after a careful analysis of the limitations of traditional BDI(Belief, Desire, Intention) architecture and of its dynamic features. We imported the motivation as another factor with other mind properties to describe the Agent. The new model of Agent is superior to describe the self-learning and evolution mechanism of Agent compared with the BDI architecture. As to the theoretical basis of BDIM, we inherited and expanded the CTL* theory on BDI architecture by Rao and Georgeff.
-
Key words:
- Agent, motivation /
- BDI /
- computation tree logic /
计量
- 文章访问数: 2390
- HTML全文浏览量: 78
- PDF下载量: 2626
- 被引次数: 0