一种基于模糊集的混合空间推理方法

A hybrid spatial reasoning approach based on fuzzy sets

基于模糊集合理论,提出一种将定性空间推理与定量空间推理相结合进行混合空间推理的方法.将对象间的空间关系解释为表达有关对象空间信息语言变量的限制,引入语言变量表达定性空间关系;把语言变量与模糊集合关联起来,获得定量空间信息;通过将语言值映射为模糊集合,以及将模糊集合的元素映射为语言限制词,实现了定性空间信息和定量空间信息间的转换.此外,还给出了与Allen提出的二元基本关系关联的模糊关系,并得到了基于模糊关系的Allen推理算法,从而实现了空间关系间的推理.

Based on fuzzy set theory, a hybrid spatial reasoning approach is proposed that combines qualitative spatial reasoning with quantitative one. The spatial relations among objects are interpreted as restrictions on linguistic variables that represent spatial information about the objects, and the linguistic variables are provided to express qualitative spatial relations, and then these variables are associated with fuzzy sets for quantitative spatial information. By mapping linguistic values to fuzzy sets and elements of fuzzy sets to linguistic hedges, it is possible to switch back and forth between qualitative spatial information and quantitative one. (Besides), we realized the reasoning of these spatial relations by showing that each Allen's basic binary relation can be associated with a fuzzy relation and obtaining an Allen's algorithm based on these fuzzy relations.