集群并行技术优化泥石流模拟数据处理研究  

Research of data-processing optimization based on cluster parallel computing in debris flow simulation

作　　者：何光军[1] 管群[1] 

机构地区：[1]四川大学计算机学院,四川成都610064

出　　处：《计算机工程与设计》2010年第20期4484-4487,4491,共5页Computer Engineering and Design

基　　金：国家自然科学基金项目(40771026);NSFC-RFBR项目(40911120089)

摘　　要：为解决泥石流等灾害模拟过程中海量数据处理的耗时问题,提出了采用集群并行方法来提升模拟过程中的数据处理效率,以达到灾害流模拟的高效性。集群计算技术已在大量的科学计算中得到广泛应用并且效果明显,但在工程模拟方面还未得到系统的实践,构建算法也没有相应标准参考。为泥石流模拟计算过程设计并行算法、构建集群平台,测试结果表明,集群并行处理方法明显提高了泥石流模拟数据处理的效率。To improve the processing efficiency of mass date in debris flow simulation and other disaster simulation,the cluster parallel processing method is presented.Cluster computing technology is used widely in a large number of scientific computing,and the effectiveness is obviously,but in the engineering simulation area there is no reference algorithms and standard.Parallel algorithm is designed,and the cluster system is built.Results indicated that the method of cluster parallel processing of debris flow simulation for the efficiency of data processing has improved significantly.

关 键 词：灾害数据模拟 泥石流 数据处理 并行处理 集群 

分 类 号：TP338[自动化与计算机技术—计算机系统结构]