地球信息科学学报 ›› 2017, Vol. 19 ›› Issue (9): 1217-1227.doi: 10.3724/SP.J.1047.2017.01217

• 全空间信息系统应用 • 上一篇    下一篇

全空间下并行矢量空间分析研究综述与展望

邱强1(), 秦承志2, 朱效民3, 赵晓芳1, 方金云1   

  1. 1. 中国科学院计算技术研究所 计算机应用研究中心, 北京 100190
    2. 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,北京100101
    3. 山东省计算中心,济南 250015
  • 收稿日期:2017-05-12 修回日期:2017-08-11 出版日期:2017-10-09 发布日期:2017-10-09
  • 作者简介:

    作者简介:邱 强(1987-),男,工程师,研究方向为地理信息系统并行计算及空间分析。E-mail: qiuqiang@ict.ac.cn

  • 基金资助:
    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)

Overview and Prospect on Spatial Analysis of Parallel Vectors in Pan-spatial Concept

QIU Qiang1,*(), QIN Chengzhi2, ZHU Xiaomin3, ZHAO Xiaofang1, FANG Jinyun1   

  1. 1. Institute of Computing Technology, Chinese Academy of Sciences, Beijing 100190, China
    2. State Key Laboratory of Resources and Environmental Information System, Institute of Geographic Sciences and Natural Resources Research, Chinese Academy of Sciences, Beijing 100101, China
    3. Computer Science Center of Shandong Province, Jinan 250015, China
  • Received:2017-05-12 Revised:2017-08-11 Online:2017-10-09 Published:2017-10-09
  • Contact: QIU Qiang E-mail:qiuqiang@ict.ac.cn

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。

关键词: 地理信息系统, 矢量空间分析, 并行计算, 全空间信息系统, 多粒度时空对象, 存算一体化

Abstract:

The new generation of parallel spatial analysis in pan-spatial information system is challenged by analysis of spatial big data and real-time spatial service. As one of the most important part of GIS, vector spatial analysis has some performance bottlenecks such as load unbalance, less ability of parallel expansion and low I/O efficiency. First, we review the history of the developing process of vector spatial analysis from application requirement and technical progress. Then, we expound the research findings of spatial analysis of parallel vector, summarize the algorithm features and technical bottlenecks, compare the different parallel programming model and present the parallel spatial algorithm of R&D processing. Finally, we predict the spatial data model in the future and the computing method based on spatio-temporal objects of multi-granularity in pan-spatial information system. Also, we present the new techniques which use memory computing to realize the storage-computing integration in vector spatial analysis.

Key words: geographic information system, vector spatial analysis, parallel computing, pan-spatial information system, spatio-temporal objects of multi-granularity, storage-computing integration