摘要
地震勘探资料数据处理是地震勘探重要而不可缺少的组成部分,当前地震勘探资料数据处理具有处理的数据量大、处理过程多以及资料的重新处理,反复处理等典型特征。因传统的地震勘探资料处理系统中各个处理模块之间采用临时文件的方式来传递数据,这些特征反映到传统的地震勘探资料处理系统中将面临的一个核心问题:巨大的磁盘I/O访问量。这个磁盘I/O访问量不仅跟地震勘探数据的大小有关,而且跟处理流程中的处理模块数有关。为此,我们设计数据流软件flobus来减少地震勘探资料处理流程中对磁盘I/O的访问量。在采用数据流flobus的地震勘探资料数据处理系统中,处理流程中总的磁盘I/O访问量只跟输入数据大小有关,而跟处理流程中的处理模块数无关,从而大大减少了磁盘I/O的访问量。数据流软件flobus充分考虑到提供多模块处理复杂系统的数据流共享机制,其主要特征包括:1)提供多任务的数据流控制,实现数据驱动的并发机制;2)提供多模块的数据聚合机制;3)支持同一级别模块间的数据依赖关系;4)支持任务流中模块的动态加载。
Seismic measurement data processing is important and indispensible part of seismic exploration. Huge amount data, multiple processing modules, reprocessing are most typical characteristic in current seismic measurement data processing. In traditiona seismic exploration processing system, data are transfered through temporary files among the processing modules. A core problem coming in tradition seismic exploration processing system is the huge disk I/O access, which is not just related to the amount of the seismic measurement data, but also relate to the count of modules in the processing flow. So we designed a data flow middle-ware, the flobus which aims to decrease the amount access of disk I/O in seismic measurement data processing flow. In this case, the amount of disk I/O just relates to the size of seismic measurement data, but not to the count of modules in the flow. The flobus provides a data share mechanic for multiple modules processing system. The main characteristics are: 1) implements data flow control for data drive concurrent, 2) mplements data cluster, 3) supports data depend relation in same level, 4) supports dynamically load modules in processing flow
出处
《科研信息化技术与应用》
2011年第6期83-91,共9页
E-science Technology & Application
关键词
地震勘探
地震勘探资料数据
数据流
数据驱动
并发
Seismic exploration technology
Seismic measurement data processing
Data flowData drive
Concurrent