基于双重逆极限的地貌信息挖掘方法研究

发布时间：2018-06-04 13:43

本文选题：数据挖掘 + 地貌信息系统　；参考：《燕山大学》2013年硕士论文

【摘要】：庞杂的地貌信息作为一种亟待人们探索和研究的空间数据信息对地球物理学的研究有着不可估量的意义。近年来，随着信息的高速发展，地貌信息特征提取要求不断增高，数据挖掘高效率的、准确的特点逐渐被地理信息科研工作者所认可，地貌信息数据挖掘方法进入了快速发展阶段。其中，近年来，地貌信息数据挖掘方法在工程建设、农业生产、矿产勘探、自然灾害防治和环境保护等领域中得到了广泛的应用。本文针对传统方法中仅局限于对离散的地貌信息处理的缺点，基于双重逆极限理论，利用拓扑传递性对繁杂的地貌特征信息进行挖掘，针对此问题从以下几个方面对地貌信息挖掘方法进行研究首先，本文基于双重逆极限空间理论，根据不完备信息四元组的特点，将地貌信息分为地貌单元对象集合、地貌单元对象属性集合和地貌单元属性值域集合三部分，从拓扑学的角度建立了适用于不完备地貌信息的拓扑空间模型，并结合三维地表形态函数证明了其有效性。然后，基于信息熵及概率相似度的思想，针对地貌信息，以映射诱导法建立关于地貌信息的双重逆极限空间到地貌信息双重空间的逆映射，并证明原信息空间与目标信息空间之间存在拓扑同构，进而利用拓扑传递性分析原地貌信息空间与目标信息空间的拓扑相似度。再次，本文根据地貌信息的四元组特点，建立地貌信息系统模型，并对建立起的地貌信息系统的相轨迹、Lyapunov指数及分岔等系统的动力学行为进行仿真和分析，验证本方法在地貌信息挖掘过程中的稳定性和准确性。最后，利用MAPGIS软件针对地貌高程信息、坡度信息和坡向信息对软件随机模拟的地貌进行验证和详细的分析，，以此为基础实现对地貌信息数据挖掘结果的准确解释和综合评价，为建立地形数据库信息、局部适配区域评估等统计参考提供方法依据。
[Abstract]:As a kind of spatial data that needs to be explored and studied urgently, the complex geomorphological information is of inestimable significance to the study of geophysics. In recent years, with the rapid development of information, the requirements for feature extraction of geomorphological information are increasing, and the characteristics of high efficiency and accuracy of data mining are gradually recognized by geographic information researchers. Geomorphological information data mining method has entered a rapid development stage. In recent years, geomorphological information mining methods have been widely used in engineering construction, agricultural production, mineral exploration, natural disaster prevention and environmental protection. In this paper, aiming at the shortcoming of the traditional method, which is limited to the discrete geomorphologic information processing, based on the double inverse limit theory, the complicated geomorphological characteristic information is excavated by using topological transitivity. Aiming at this problem, the methods of geomorphological information mining are studied from the following aspects Firstly, based on the theory of double inverse limit space, according to the characteristics of incomplete information quaternion, the geomorphological information is divided into three parts: geomorphological unit object set, geomorphologic unit object attribute set and geomorphologic unit attribute value domain set. A topological space model suitable for incomplete geomorphological information is established from the perspective of topology, and its validity is proved by combining with 3D surface morphology function. Then, based on the idea of information entropy and probability similarity, the double inverse mapping of geomorphological information from double inverse limit space to double inverse space of geomorphological information is established by mapping induction method. The topological isomorphism between the original information space and the target information space is proved, and the topological similarity between the original geomorphological information space and the target information space is analyzed by topological transitivity. Thirdly, according to the quaternion characteristics of geomorphological information, the geomorphological information system model is established, and the dynamic behavior of the geomorphologic information system, such as phase trajectory Lyapunov exponent and bifurcation, is simulated and analyzed. The stability and accuracy of this method in geomorphological information mining are verified. Finally, MAPGIS software is used to verify and analyze the geomorphology of random simulation based on the geomorphological elevation information, slope information and slope direction information, based on which the accurate interpretation and comprehensive evaluation of geomorphological information mining results are realized. It provides the method basis for the establishment of topographic database information, local adaptation area evaluation and other statistical references.
【学位授予单位】：燕山大学
【学位级别】：硕士
【学位授予年份】：2013
【分类号】：P208;TP311.13

【参考文献】