摘要
不确定性知识处理是人工智能研究中经常遇到的问题,基于定性映射的属性Petri网模型在动态表示认知思维中不确定性知识与逻辑推理方面具有优势。在属性拓扑空间中,给出了属性粒的基本定义和基本推理,在属性Petri网中对不确定性知识进行表达,并基于属性Petri网给出归结推理的基本形式和基本算法。结果显示,这种方法可以使定性映射和Petri网更易于动态和显式地表达认知不确定性知识,可为进一步研究Petri网在认知模型中的作用提供参考。
To deal with the uncertain knowledge is often encountered in artificial intelligence research. Attribute Petri net model based on qualitative mapping has the advantages of a dynamic representation of uncertainty knowledge and logical reasoning in cognitive thinking. The basic definition and basic reasoning of attribute granular were given in the property topological space in this paper. Uncertainty knowledge can be expressed with the attribute granular in attribute Petri net. Finally, the basic form and basic algorithm of resolution reasoning were given in attribute Petri net. The re- suits show that this method can make qualitative mapping and Petri net to more dynamically explicit expression of the cognitive uncertainty knowledge,and it can also provide a reference for further study of Petri net in the cognitive model.
出处
《计算机科学》
CSCD
北大核心
2014年第8期101-105,共5页
Computer Science
基金
国家自然科学基金项目(60075016)
广东省科技计划项目(2012B010100049)资助
关键词
定性映射
PETRI网
粒计算
知识表示
归结推理
Qualitative mapping, Petri net, Granular computing, Knowledge representation, Resolution reasoning