基于云GIS的地理国情统计分析方法研究

发布时间：2017-12-27 10:12

本文关键词：基于云GIS的地理国情统计分析方法研究　出处：《浙江大学》2017年硕士论文　论文类型：学位论文

【摘要】：地理国情统计分析是从海量地理国情普查数据中获取深层地理国情信息的必要手段,我国正处于社会经济结构升级的转折期,也是国情快速、复杂变化的关键时期,全面而现势的地理国情信息对科学决策的制定和科学管理的规划至关重要。本研究针对当前地理国情统计分析应用割离、服务单一,且存在单机数据存储与处理方式低效的局限性,结合网络服务和高性能计算技术,旨在建立面向地理国情统计分析的灵活、高效处理方法,构建云GIS环境下弹性可扩的统计分析服务体系。本文主要研究内容和创新点包括以下几个方面:(1)研制了地理国情统计分析多层次服务体系模型,通过多粒度服务的按需调用和随需组装,实现多种服务、资源的有效整合和弹性扩展,解决当前系统孤立、功能冗余、应用单一的问题,打破数据与信息壁垒,以高复用、高弹性、高可扩的服务支撑多样化的统计分析业务。(2)针对海量、多源的地理国情普查数据,构建了基于HDFS和HBase的分布式存储模式,实现普查大数据的云端管理和维护,解决了现有单机关系型数据库存储和管理瓶颈,为统计分析提供高效数据响应。(3)基于Hadoop云环境,设计了地理国情统计分析的并行化处理架构,利用MapReduce构建地理国情统计分析并行算法,解决普查数据的复杂空间分析与大规模要素空间统计的单机处理性能瓶颈,实现普查数据的高效处理与地理国情统计分析服务的准实时响应。(4)综合运用多层次服务体系、分布式数据库以及并行统计分析算法,搭建集多源海量地理国情普查数据统一管理、高效数据处理、统计、分析和业务于一体的地理国情统计分析服务平台,并进行了客户端应用实例验证。实验与实例效果证明,本文提出的方案能够满足地理国情普查数据的可扩存储和大批量普查数据统计分析任务的灵活、高效执行。
[Abstract]:Statistical analysis of geographical conditions is a necessary means to get the deep information of the geographical conditions from the census data in geographic conditions, our country is in the period of transition and upgrading of social economic structure, but also a critical period of rapid and complex changes in the situation, comprehensive and scientific management to develop and present the information of the geographical conditions of the scientific decision of planning is essential. Based on the current application of statistical analysis of geographical conditions, cut from a single service, and there are limitations of stand-alone data storage and processing inefficient, combined with network services and high performance computing technology, aims to establish a statistical analysis of the geographical conditions for flexible and efficient processing method, constructing the statistical elastic cloud environment GIS extendable analysis service system. In this paper, the main research contents and innovations include the following aspects: (1) developed a multi-level analysis service system model of geographical conditions, through multi granularity service on-demand call and on-demand assembly, effective integration and flexibility to achieve a variety of services, expansion of resources, solve the current system of isolation, functional redundancy, single application problems the data and information, to break the barriers to high, high elasticity, Gao Kekuo support services diversified statistical analysis business. (2) according to the census data massive, multi-source geographic conditions, constructed a distributed storage model of HDFS and HBase based on the realization of management and maintenance of the census data, solve the existing single relational database storage and management bottleneck, provide efficient number for statistical analysis according to the response. (3) Hadoop cloud based environment, parallel processing architecture analysis of geographical conditions statistical design, construction of geographic conditions, statistical analysis of parallel algorithm using MapReduce single bottleneck processing performance to solve complex spatial analysis and spatial statistics of large-scale census data, the quasi real time response and efficient processing of census geographic conditions, the statistical analysis of the data service the. (4) the comprehensive use of multi-level service system, distributed database, parallel algorithm of statistical analysis, statistical geographic conditions to build a set of multi-source geographic conditions survey data, unified management, data processing, statistics, analysis and business analysis in one of the service platform, and client application examples. Experiments and case studies show that the scheme proposed in this paper can satisfy the scalability and storage of census data and flexible and efficient implementation of mass census data analysis.
【学位授予单位】：浙江大学
【学位级别】：硕士
【学位授予年份】：2017
【分类号】：P208

【相似文献】