基于复合表的近似拓扑推理  

Approximate topological reasoning based on composition table

作　　者：南淑芹[1] 李三江[2] 

机构地区：[1]河北经贸大学数学与统计学学院,河北石家庄050091 [2]清华大学计算机科学与技术系智能技术与系统国家重点实验室,北京100084

出　　处：《陕西师范大学学报（自然科学版）》2004年第1期21-24,32,共5页Journal of Shaanxi Normal University：Natural Science Edition

基　　金：国家自然科学基金资助项目(60305005)

摘　　要：采用宽边界模型表示不确定区域,基于4 交模型对简单不确定区域间的拓扑关系进行了完备分类,得到在较粗粒度上的不确定区域间的11种完备拓扑关系,即disjoint,meet,α meet,β meet,γ meet,overlap,coveredBy,covers,inside,contains,equal.然后给出了这些基本拓扑关系之间的复合规则,并将这些规则总结在一个复合表中.关系复合表的建立为空间数据不确定性处理建立了推理基础,并为建立能处理不确定性的空间数据库和地理信息系统做了理论准备.With the broad boundary model to represent non-deterministic regions, and uses the (4-intersection) model to classify the topological relations among regions with broad boundary. The (11 jointly) exhaustive and pair-wise disjoint topological relations among these regions in lower resolution are obtained. These relations are disjoint, meet, α-meet, β-meet, γ-meet, overlap, coveredBy, covers, inside, contains, equal. The composition rules between these base relations and establish the corresponding composition table are given. This composition table provides a basis for reasoning about indeterminacy in spatial data and makes a preliminary theoretic attempt for the emergence of spatial databases and GISs capable of coping with indeterminacy.

关 键 词：复合表 近似拓扑推理 定性空间推理 拓扑关系 4-交模型 不确定性 宽边界 人工智能 

分 类 号：TP18[自动化与计算机技术—控制理论与控制工程]