全空间下并行矢量空间分析研究综述与展望

发布时间：2018-04-22 00:32

本文选题：地理信息系统 + 矢量空间分析　；参考：《地球信息科学学报》2017年09期

【摘要】：新一代并行空间分析将面临空间大数据分析和实时空间分析服务的挑战。矢量空间计算作为GIS系统中的重要组成部分,在并行化算法设计中存在负载不均,并行扩展性差,IO性能低等技术瓶颈。本文首先从应用需求和技术发展的演变历史回顾了矢量空间分析算法发展过程;然后,从研究现状的角度详细阐述了并行矢量空间分析计算的研究成果,总结了并行空间分析算法的算法特征和技术瓶颈,对不同并行编程模型进行了对比,并提出了并行空间分析算法的研发流程;最后,从发展前景的角度预测了全空间信息系统中基于多粒度时空对象的空间数据模型和计算方法的发展趋势,提出了以内存计算等技术实现存算一体化的新型空间数据模型和分析方法的技术趋势。
[Abstract]:The new generation of parallel spatial analysis will face the challenge of large spatial data analysis and real-time spatial analysis services. Vector space computing is an important part of the GIS system. In the design of parallel algorithms, there are many technical bottlenecks in the design of parallel algorithms, such as load uneven, poor parallel scalability and low performance of IO. The development process of vector space analysis algorithm is reviewed. Then, the research results of parallel vector space analysis are described in detail from the perspective of research. The algorithm characteristics and technical bottlenecks of parallel spatial analysis algorithm are summarized, and the different parallel programming models are compared, and the research and development process of parallel spatial analysis algorithm is put forward. Finally, the development trend of spatial data model and calculation method based on multi granularity spatiotemporal objects in the whole space information system is predicted from the perspective of development prospects, and the technical trend of a new spatial data model and analysis method based on memory calculation and other technologies is proposed.

【作者单位】：中国科学院计算技术研究所计算机应用研究中心;中国科学院地理科学与资源研究所资源与环境信息系统国家重点实验室;山东省计算中心;
【基金】：国家重点研发计划项目“全空间信息系统与智能设施管理”(2016YFB0502300)子课题“多粒度时空对象组织与管理”(2016YFB0502302)
【分类号】：P208

【相似文献】