气象卫星数据处理分层调度模型仿真  

A Hierarchical Meteorological Satellite Data Processing Scheduling Model Simulation

作　　者：赵现纲[1] 林曼筠[1] 冯小虎[1] 范存群[1] 

机构地区：[1]国家卫星气象中心,北京100081

出　　处：《计算机仿真》2016年第11期239-243,共5页Computer Simulation

基　　金：公益性行业(气象)科研专项(GYHY201006046);国家863计划项目(2011AA12A104)

摘　　要：由于风云卫星地面应用系统地面数据处理流程复杂,如何实现风云气象卫星复杂的地面数据快速处理和可靠调度是当前卫星地面系统所亟需解决的一个难题。针对上述问题,提出了一种基于簇的气象卫星数据处理分层调度模型。首先,以分层的思想设计了顶簇与分簇,顶簇协调分簇之间的协作,各分簇自主调度内部作业运行,尽可能地减少了各分簇之间的相互影响;然后,利用Petri网定义了顶簇和分簇的调度网和作业网结构;最后,设计了分簇调度故障处理算法。仿真结果表明改进模型可以很好地简化气象卫星数据处理调度和作业流程,缩小异常影响的范围,提高数据处理的可靠性。How to reliably schedule the complex data processing procedure of meteorological satellite ground system has always been a challenge. In view of this, this paper proposes a hierarchical meteorological satellite data processing scheduling model based on cluster. Firstly, according to this model, we decomposed the complex scheduling structure into distributed clustering structure which is relatively simple, independent and easy to be deployed and administrated. The top cluster coordinates the collaborations of sub-clusters and each sub-cluster schedules its internal tasks independently, thereby reducing the interactions between each sub-cluster as far as possible. Secondly, we defined the Petri net-based structures of scheduling nets and job nets for top clusters and sub-clusters respectively. Finally, we also designed the troubleshooting algorithm of the cluster scheduling model. Simulation results show the model can simplify meteorological satellite data processing and job scheduling process, narrow the scope of the abnormal effects, and improve the reliability of data processing.

关 键 词：气象卫星 调度模型 簇结构 佩特里网 异常处理 

分 类 号：TP391.9[自动化与计算机技术—计算机应用技术]