基于分布式工作流的地理国情普查统计计算执行研究

Statistical Perform Model for First Nationwide General Survey of Geographic Conditions Based on Distributed Workflow

地理国情普查用于查清我国自然和人文地理要素现状和空间分布情况,统计分析是国情普查的关键环节,基于分布式工作流的数据统计是一种先进高效数据计算技术,符合大数据时代海量数据处理及数据分布式、任务协同式、流程自动化的要求。在地理国情普查统计分析中采用任务驱动的规则定义TCA(Task Condition Action)相对于事件驱动ECA(Event Condition Action)更简洁;针对数据统计的流程服务,分析了一个最小构成群內节点上的工作流引擎计算过程的任务分布及流转模型,并通过实验验证系统计算效率与节点数量的对应关系。

The first nationwide general survey of geographic conditions aims to ascertain the status and the spatial distribution of Chinese natural and human geography elements.Statistical analysis is one of the key steps.Statistics based on distributed workflow is an advanced statistical technique for big data,and it meets the requirements of distributed data, collaborative task and automatic process in the era of big data.Task Condition Action(TCA)based on Task Engine is simpler and more effective than Event Condition Action(ECA)in the first nationwide general survey of geographic conditions.The task schedule model was analyzed in a micro computing group.The corresponding relationship between computing nodes and statistics time was analyzed by experiment.