模糊相似性和空间数据建模及其在多维尺度矢量地图空间中的应用研究

发布时间：2018-01-25 20:37

本文关键词： 空间数据库自动泛化相似变换模糊空间关系多维缩放地图隶属函数利用地理信息系统模糊集理论模糊逻辑　出处：《北京理工大学》2015年博士论文　论文类型：学位论文

【摘要】：与离散化表示和模糊空间目标相结合的自动控制在多维矢量地图空间目标上已经出现了矛盾和不一致性。许多处理空间对象的空间数据库,由于其本质上的不确定性和模糊的本质,而不能充分地用确定性、明确的空间概念描述。地理信息系统和空间数据库系统目前无法精确地应对这种空间对象。是建设空间数据库的必要工具。建设空间数据库是空间数据基础设施的关键,它为社会中像经济、国防、教育、交通、PD、环境保护、科研、电信等众多领域提供了地理空间定位基地。然而,这仍然是一个理想,因为特征空间中的对象映射的概括是并不是没有参数的,因此需要人为的干扰。其中一个重要的原因是,自动泛化空间相似变换在多维尺度矢量地图空间是模糊性的。并且,没有发现一种理论支持这种类型的转换。本文侧重于开发基于多维尺度矢量地图空间的模糊空间相似关系的理论,针对提出的定义、模式和方法,可以在多维尺度矢量地图空间中选择自动参数泛化算法。系统的回顾现有的成果,包括相似性的定义,评估各领域相似的特点,分类系统解决相似的特点和从数学、心理学、信息科学和地理领域测量工具的相似性关系,以及基于大量的栅格模型来衡量图像之间的相似度。本文取得了如下创新性的贡献。首先实现了模糊集理论的效用和模糊逻辑的相似变换用于多维尺度矢量地图的发展空间推广过程中,首次在文学理论开发了模糊空间相似关系的理论的实用程序。在自动综合的上下文中,没有报道过这项工作。第二,在地理空间和多维尺度向量图上,模糊空间相似关系是基于模糊集理论范式,而不是脆的空间对象的集合理论的范式,这增强了建模空间对象的能力。第三是提出了基于模糊关系的模糊空间相似关系自动综合的特点和模糊逻辑,以加强在地理空间和多维尺度矢量地图理论的逻辑关系。第四是模糊因素,其识别和模糊性在空间对象的相似变换地理空间和多维尺度向量地图空间中的贡献。第五是提出模糊化程序,以获取隶属函数的模糊度的全貌,使我们在参数选择的空间对象的相似变换中能有一个非常清楚和最优决策的完整画面。虽然这项研究侧重于自动泛化的模糊空间相似关系,特别是在地形图,但这种理论的应用超过了GIS的领域。我们称之为空间对象的空间数据处理的科学与工程的任何分支将受益于模糊空间相似性关系。它能被利用于如空间逻辑,空间推理领域,从而最终将其利益扩大到测绘工程、医学、健康、基因工程、计算机辅助设计、计算机辅助制造和机器人技术领域等这些有助于国家经济发展的领域。
[Abstract]:The automatic control combined with discrete representation and fuzzy spatial object has already appeared contradiction and inconsistency on multi-dimensional vector map spatial target. Many spatial databases dealing with spatial objects. Because of its inherent uncertainty and fuzzy nature, it can not fully use certainty. A clear description of spatial concepts. Geographic information systems and spatial database systems are currently unable to respond precisely to such spatial objects. They are essential tools for the construction of spatial databases, which are the spatial data infrastructure. The key. It provides geo-spatial positioning bases in many areas of society, such as economy, national defense, education, transportation PDs, environmental protection, scientific research, telecommunications, etc. However, this is still an ideal. Because the generalization of object mapping in a feature space is not parameterless, it requires human interference. One of the important reasons is that. Automatic generalization spatial similarity transformation is fuzzy in multi-dimensional vector map space. No theory has been found to support this type of transformation. This paper focuses on the development of fuzzy spatial similarity relations based on multidimensional scale vector map space, aiming at the proposed definitions, patterns and methods. Automatic parameter generalization algorithm can be selected in multidimensional scale vector map space. The system reviews the existing results including the definition of similarity and evaluates the characteristics of similarity in various fields. The classification system addresses the similarity between similar features and measurement tools from mathematics, psychology, information science and geography. And based on a large number of raster models to measure the similarity between images. This paper has made the following innovative contributions. Firstly, the utility of fuzzy set theory and the similarity transformation of fuzzy logic are realized for multidimensional scale vector map. In the process of developing and popularizing. For the first time, a practical program for the theory of similarity in fuzzy space has been developed in literary theory. This work has not been reported in the context of automatic synthesis. Secondly, in geographical space and multidimensional scale vector graph. The similarity relationship in fuzzy space is based on the paradigm of fuzzy set theory, not the paradigm of set theory of brittle spatial objects. This enhances the ability of modeling spatial objects. Thirdly, the characteristics and fuzzy logic of automatic synthesis of fuzzy spatial similarity relations based on fuzzy relations are proposed. In order to strengthen the logical relationship of vector map theory in geographic space and multidimensional scale. 4th is the fuzzy factor. The contribution of its recognition and fuzziness to the similar transformation of spatial objects in geographic space and multidimensional scale vector map space. 5th is a fuzzification program proposed to obtain a complete picture of the fuzziness of membership function. It enables us to have a very clear and complete picture of optimal decision in the similarity transformation of spatial objects with parameter selection, although this study focuses on the automatic generalization of fuzzy spatial similarity relations, especially in topographic maps. But the application of this theory extends beyond the domain of GIS. Any branch of science and engineering that we call spatial data processing will benefit from fuzzy spatial similarity relations. It can be used for example spatial logic. . Spatial reasoning will ultimately extend its benefits to mapping engineering, medicine, health, genetic engineering, and computer-aided design. Areas such as computer-aided manufacturing and robotics are areas that contribute to the country's economic development.
【学位授予单位】：北京理工大学
【学位级别】：博士
【学位授予年份】：2015
【分类号】：P208

【参考文献】