摘要
针对传统时空推理研究大多局限于静态空间对象以及Rupam方位关系模型ROR复合表和概念邻域图所存在的问题,提出了一种动态方位关系推理方法。首先对ROR关系模型中复合表及概念邻域图进行完善:提出了ROR关系静态推理算法,并给出了复合表;提出了一种ROR概念邻域关系自动生成算法并构建了概念邻域图;将ROR关系复合表和概念邻域图与Rupam方法进行比较,结果表明,本文方法不仅能给出Rupam所给的正确结果,还能修正Rupam方位关系模型不存在的复合结果,且复合结果和概念邻域图更完整。然后基于ROR关系复合表和概念邻域图给出了运动对象间方位关系推理方法DRA。最后通过智能家居中的应用实例,说明了DRA方法的有效性。
To solve the problem of limitation of traditional spatio-temporal reasoning in static space, and the problems in Rupam's orientation relation composition table and conceptual neighborhood graphs, a dynamic reasoning method based on Rupam's orientation relations model was proposed. First, the ROR base relations' composition table and conceptual neighborhood graphs are improved; a static reasoning algorithm is proposed and a ROR based relations'composition table is created; an auto-generated algorithm to produce conceptual neighborhood graphs of orientation relations is put forward. A comparison of the composition table and conceptual neighborhood graphs with those of Rupam was conducted and results show that the composition table and conceptual neighborhood graphs are more complete than Rupam's. Then a dynamic reasoning method (DRA) is proposed. To demonstrate the effectiveness of DRA, an application of smart home was illustrated based on the composition table and neighborhood graphs.
出处
《吉林大学学报(工学版)》
EI
CAS
CSCD
北大核心
2014年第1期117-123,共7页
Journal of Jilin University:Engineering and Technology Edition
基金
国家自然科学基金项目(61170092
61133011
61272208
61103091
61202308)
关键词
人工智能
方位关系复合表
概念邻域图
区间关系
运动对象
artificient intelligence
orientation relations" composition table
conceptual neighborhood graphs
interval relations
object in motion