图形工作流驱动的空间信息服务链研究  

Spatial Information Services Chaining Based on Graphic-Workflow

作　　者：张建博[1,2] 刘纪平[2] 王蓓[3] 

机构地区：[1]武汉大学资源环境与科学学院,武汉430079 [2]中国测绘科学研究院,北京100830 [3]中国科学院地理科学与资源研究所,北京100101

出　　处：《计算机研究与发展》2012年第6期1357-1362,共6页Journal of Computer Research and Development

基　　金：国家自然科学基金项目(40901195)

摘　　要：空间信息服务链由于其空间数据操作对时间的要求以及其应用于工作流对接口要求的特殊性,使得传统的工作流很难调度空间信息服务链,主要的问题在于接口不匹配以及运行效率低下.在传统WPS服务链的基础上,针对其不足提出了基于图形工作流的空间信息服务链聚合模型.通过分析工作流调度空间信息服务链存在的主要问题,提出相应的接口改进方法,以及基于地址引用、压缩GML的数据调度策略.并结合Kepler工作流引擎,给出了空间信息服务链聚合的具体过程.经过实验证明:基于图形工作流环境,不但突破了以往空间信息服务应用于工作流的接口瓶颈,而且改进的数据流调度策略有助于空间信息服务链执行效率的提高.Traditional workflow technology can＇t meet the requierments of spatial＇information services chaining aggregation because of the particularity of the long time cost of spatial data processing and the interface that workflow calls spatial information services chaining. The main problems lie in the mismatch of interface, and low efficiency. Thence, the model of spatial information service chaining aggregation is established, expending the model of traditional WPS chaining whose insufficient is considered. By analyzing the main problems between the interface and spatial data processing about spatial information services during the workflow scheduling, we I^resent corresponding methods of the interface transformation to make the workflow call spatial information services chaining well, and adopt a strategy of GML-compressed scheduling to guarantee the spatial data flow be lowly transmitted on the network. Finally, the concrete solutions of spatial information services chaining processing are presented in the paper combined Kepler workflow engine. The experimental results show that this scheduling method not only breakthroughs interface bottlenecks through the interface transformation used in the workflow before, but also improves spatial information service chain execution efficiency through the strategy transformation of spatial data flow.

关 键 词：空间信息服务 服务链 聚合 GML调度 图形工作流 KEPLER 

分 类 号：TP311[自动化与计算机技术—计算机软件与理论]