条带状公路运营管理空间大数据降维组织及混合存储关键技术研究
发布时间：2018-08-09 09:58
【摘要】：公路以其灵活多变、运量大和速度快的特点在现代交通中占据重要地位。运营管理作为公路生命周期的重要一环,包括养护、交通、安全、服务等方面。空间信息技术是交通信息化建设的重要组成,贯穿在公路运营管理过程中。随着近年来,我国高分辨率对地观测系统和北斗卫星导航系统建设稳步推进,物联网、车联网技术不断推广,交通运营管理空间数据逐步呈现出体量大、种类多、速度快、价值高的大数据特征。传统存储管理方式不仅难以满足大数据的需求,更缺乏对条带状空间数据的针对性。因此,如何高效地存储管理公路运营管理空间大数据,是当下亟待解决的问题。本文基于NoSQL数据库和分布式云存储,围绕公路运营管理空间数据的特征和管理需求,提出公路空间大数据条带降维组织模型与公路空间大数据多态混合存储架构,解决海量多源异构公路运营管理空间大数据的存储检索效率问题。本文主要研究内容如下：(1)公路运营管理空间大数据特征分析：通过对不同数据来源进行分析,总结出其大数据特征和条带状空间分布特性。进而根据其特点对数据进行分类,并明确存储管理需求,使得数据组织与存储的设计更具针对性。(2)公路空间大数据条带降维组织模型：通过分析空间数据降维的理论和方法,指出划分空间格网是空间数据降维的有效方式,也是空间编码的基础。通过对Geohash格网与公路空间的尺度对比,提出公路空间格网划分方法。将基于Geohash格网的空间数据降维方法,与公路本身一维的线性参照系统相结合,提出公路空间大数据条带降维组织模型,并设计了点、线、面三类公路空间大数据的存储检索方式。(3)公路空间大数据多态混合存储架构：在深入分析不同空间大数据存储技术的基础上,将NoSQL数据库、分布式云存储及空间数据库引擎三者无缝结合,对动态与静态、结构化与非结构化、空间与非空间等多态的公路空间大数据进行混合存储。提出索引关联的混合存储协调管理引擎,通过空间信息降维进行索引,建立公路空间大数据之间的关联,实现了公路空间大数据的无缝集成和一体化存储。针对公路空间带状特点,采用影像预分块策略提高混合存储架构中影像数据按需获取的效率。(4)公路运营管理空间大数据管理原型系统：设计了原型系统的框架,实现了数据入库、数据检索、系统运行监控、数据可视化等核心功能,并介绍了原型系统在公路地质灾害应急处理中的应用。通过与传统的空间数据库引擎进行检索性能对比实验,论证了本文提出的公路运营管理空间大数据存储管理技术是可行的,并具有很好的性能。
[Abstract]:Highway plays an important role in modern traffic because of its flexibility, large volume and high speed. As an important part of road life cycle, operation management includes maintenance, transportation, safety, service and so on. Spatial information technology is an important component of traffic information construction, which runs through the course of highway operation and management. In recent years, the construction of high resolution Earth observation system and Beidou satellite navigation system has been advancing steadily, the technology of Internet of things and vehicle network has been popularized continuously, and the spatial data of traffic operation management have gradually presented large volume, many kinds and fast speed. Big data features of high value. The traditional storage management method is not only difficult to meet the needs of big data, but also lack the pertinence of strip spatial data. Therefore, how to store and manage highway management space big data efficiently is an urgent problem. Based on NoSQL database and distributed cloud storage, this paper proposes a hybrid storage architecture of highway spatial big data band reduction organization model and highway spatial big data polymorphic storage, focusing on the characteristics and management requirements of highway operation management spatial data. To solve the problem of storage and retrieval efficiency of massive multi-source heterogeneous highway operation management space big data. The main contents of this paper are as follows: (1) the big data characteristics of highway management space: through the analysis of different data sources, the characteristics of big data and strip spatial distribution are summarized. Then the data is classified according to its characteristics, and the requirement of storage management is clearly defined, which makes the design of data organization and storage more pertinence. (2) Highway spatial big data strip dimensionality reduction organization model: through analyzing the theory and method of spatial data dimensionality reduction, It is pointed out that dividing spatial grid is an effective way to reduce the dimension of spatial data, and it is also the basis of spatial coding. Based on the scale comparison between Geohash grid and highway space, a method of dividing highway space grid is proposed. The dimensionality reduction method of spatial data based on Geohash grid is combined with the one dimensional linear reference system of highway, and the dimensionality reduction organization model of big data strip in highway space is proposed, and the points and lines are designed. (3) Highway spatial big data polymorphic hybrid storage architecture: on the basis of in-depth analysis of different spatial big data storage technology, NoSQL database, The distributed cloud storage and spatial database engine are combined seamlessly to store the highway spatial big data which is dynamic and static, structured and unstructured, spatial and non-spatial. A hybrid storage coordination management engine for index association is proposed. By reducing the dimension of spatial information to index, the association between highway spatial big data is established, and the seamless integration and integrated storage of highway spatial big data are realized. According to the characteristics of highway spatial banding, image prepartitioning strategy is adopted to improve the efficiency of image data acquisition on demand in the hybrid storage architecture. (4) the big data management prototype system of highway operation management space: the framework of the prototype system is designed. The core functions of data storage, data retrieval, system operation monitoring, data visualization and so on are realized. The application of prototype system in highway geological disaster emergency treatment is introduced. By comparing the retrieval performance with the traditional spatial database engine, it is proved that the proposed big data storage and management technology of highway operation management space is feasible and has good performance.
【学位授予单位】：浙江大学
【学位级别】：博士
【学位授予年份】：2015
【分类号】：U495
本文编号：2173720
[Abstract]:Highway plays an important role in modern traffic because of its flexibility, large volume and high speed. As an important part of road life cycle, operation management includes maintenance, transportation, safety, service and so on. Spatial information technology is an important component of traffic information construction, which runs through the course of highway operation and management. In recent years, the construction of high resolution Earth observation system and Beidou satellite navigation system has been advancing steadily, the technology of Internet of things and vehicle network has been popularized continuously, and the spatial data of traffic operation management have gradually presented large volume, many kinds and fast speed. Big data features of high value. The traditional storage management method is not only difficult to meet the needs of big data, but also lack the pertinence of strip spatial data. Therefore, how to store and manage highway management space big data efficiently is an urgent problem. Based on NoSQL database and distributed cloud storage, this paper proposes a hybrid storage architecture of highway spatial big data band reduction organization model and highway spatial big data polymorphic storage, focusing on the characteristics and management requirements of highway operation management spatial data. To solve the problem of storage and retrieval efficiency of massive multi-source heterogeneous highway operation management space big data. The main contents of this paper are as follows: (1) the big data characteristics of highway management space: through the analysis of different data sources, the characteristics of big data and strip spatial distribution are summarized. Then the data is classified according to its characteristics, and the requirement of storage management is clearly defined, which makes the design of data organization and storage more pertinence. (2) Highway spatial big data strip dimensionality reduction organization model: through analyzing the theory and method of spatial data dimensionality reduction, It is pointed out that dividing spatial grid is an effective way to reduce the dimension of spatial data, and it is also the basis of spatial coding. Based on the scale comparison between Geohash grid and highway space, a method of dividing highway space grid is proposed. The dimensionality reduction method of spatial data based on Geohash grid is combined with the one dimensional linear reference system of highway, and the dimensionality reduction organization model of big data strip in highway space is proposed, and the points and lines are designed. (3) Highway spatial big data polymorphic hybrid storage architecture: on the basis of in-depth analysis of different spatial big data storage technology, NoSQL database, The distributed cloud storage and spatial database engine are combined seamlessly to store the highway spatial big data which is dynamic and static, structured and unstructured, spatial and non-spatial. A hybrid storage coordination management engine for index association is proposed. By reducing the dimension of spatial information to index, the association between highway spatial big data is established, and the seamless integration and integrated storage of highway spatial big data are realized. According to the characteristics of highway spatial banding, image prepartitioning strategy is adopted to improve the efficiency of image data acquisition on demand in the hybrid storage architecture. (4) the big data management prototype system of highway operation management space: the framework of the prototype system is designed. The core functions of data storage, data retrieval, system operation monitoring, data visualization and so on are realized. The application of prototype system in highway geological disaster emergency treatment is introduced. By comparing the retrieval performance with the traditional spatial database engine, it is proved that the proposed big data storage and management technology of highway operation management space is feasible and has good performance.
【学位授予单位】：浙江大学
【学位级别】：博士
【学位授予年份】：2015
【分类号】：U495
【相似文献】
相关重要报纸文章 前5条
1 仝志辉 中国人民大学农业与农村发展学院副教授;金融服务农村集体“三资”管理空间广阔[N];中国城乡金融报;2013年
2 杨凯;拓管理空间，避免“以量换利”[N];中国煤炭报;2013年
3 记者 王延美 通讯员 董丽华;济钢一炼拓宽安全管理空间[N];中国冶金报;2000年
4 程志云;银行财富管理空间巨大[N];经济观察报;2010年
5 浙江省舟山市定海区白泉中心小学 顾恩良;校长应念好“开会经”[N];中国教育报;2008年
相关博士学位论文 前1条
1 侯志通;条带状公路运营管理空间大数据降维组织及混合存储关键技术研究[D];浙江大学;2015年
，本文编号：2173720
本文链接：https://www.wllwen.com/kejilunwen/daoluqiaoliang/2173720.html
