云计算环境下的电力GIS数据管理和分析研究

发布时间：2018-10-12 07:18

【摘要】：近年来,电力GIS凭借其在空间信息展现能力以及管理分析能力上的优势,有效地提高了电网设备的可视化水平,增强了电网的实时监管能力,已经成为电力信息化建设中不可或缺的一部分。随着智能电网建设的迅速推进以及电网规模的不断扩大,电力行业正面临一系列新的挑战,如电网信息更加庞大、数据分布更加广泛、电气接线日益复杂、电力分析难度不断加大等,传统的存储策略与计算模式越来越难以适应电力GIS新的需求,现有的GIS服务器在计算资源、数据处理与响应速度等方面的局限性也逐渐体现出来,如何高效地存储与管理海量数据,是电力GIS急需解决的问题。云计算作为一种具有良好扩展性和可用性的分布式计算架构,为海量电力GIS数据的管理提供了新的解决思路。本文选用Hadoop开源云平台,对云计算在电力GIS领域的应用进行了积极的探索与研究。在对电力GIS各类数据进行归纳和分析的基础上,充分考虑关系型数据库和非关系型数据库的优势,给出了电力GIS数据存储策略以及基于Hadoop的数据管理架构。设计了遵循OGC标准的空间数据模型、基于横纵表结构的运行数据模型以及其它核心对象模型,并利用MapReduce实现了电力GIS数据并行处理的相关技术,包括瓦片金字塔的并行构建技术、空间索引的并行生成技术、空间数据的并行分析技术和运行数据的并行查询技术。为了对所提出的电力GIS数据处理方法进行验证,本文在传统单机环境和Hadoop集群环境下进行了一系列对比实验,实验结果表明,基于MapReduce的数据并行处理方法效率较高,扩展性较好,在数据量达到一定规模时,瓦片金字塔构建时间、索引生成时间、数据分析与查询的平均时间大幅度缩短,能够很好的满足海量电力GIS数据存储与管理的需求。
[Abstract]:In recent years, with its advantages in spatial information presentation and management analysis, power GIS has effectively improved the visualization level of power grid equipment and enhanced the real-time supervision ability of power grid. It has become an indispensable part of electric power information construction. With the rapid development of smart grid construction and the continuous expansion of power grid scale, the electric power industry is facing a series of new challenges, such as the information of power grid is larger, the data distribution is more extensive, and the electrical connection is increasingly complex. With the increasing difficulty of power analysis, the traditional storage strategy and computing mode are becoming more and more difficult to adapt to the new demands of power GIS, and the limitations of existing GIS servers in computing resources, data processing and response speed are gradually reflected. How to store and manage massive data efficiently is an urgent problem for GIS. Cloud computing, as a distributed computing architecture with good scalability and availability, provides a new solution for the management of massive power GIS data. In this paper, Hadoop open source cloud platform is used to explore and research the application of cloud computing in the field of electric power GIS. On the basis of summarizing and analyzing all kinds of data of power GIS, considering the advantages of relational database and non-relational database, the data storage strategy of power GIS and the data management architecture based on Hadoop are given. The spatial data model following OGC standard, the running data model based on the structure of horizontal and vertical table and other core object models are designed, and the related technology of parallel processing of power GIS data is realized by using MapReduce. It includes the parallel construction technology of tile pyramid, the parallel generation technology of spatial index, the parallel analysis technology of spatial data and the parallel query technology of running data. In order to verify the proposed method of power GIS data processing, a series of comparative experiments are carried out in the traditional single-machine environment and Hadoop cluster environment. The experimental results show that the parallel data processing method based on MapReduce is more efficient. When the amount of data reaches a certain scale, the building time of tile pyramid, the time of index generation, the average time of data analysis and query are greatly shortened, which can meet the demand of massive power GIS data storage and management.
【学位授予单位】：华北电力大学
【学位级别】：硕士
【学位授予年份】：2015
【分类号】：TM769;TP311.13

【参考文献】