| 河网非恒定流的克隆仿真方法研究 |
| |
| 引用本文： | 张国义,黄刘生,陈国良. 河网非恒定流的克隆仿真方法研究[J]. 中国科学技术大学学报, 2007, 37(9): 1088-1095 |
| |
| 作者姓名： | 张国义 黄刘生 陈国良 |
| |
| 作者单位： | 1. 中国科学技术大学计算机科学技术系,安徽合肥,230027
2. 国家高性能计算中心(合肥),安徽合肥,230027 |
| |
| 基金项目： | 国家高技术研究发展计划(863计划);国家自然科学基金 |
| |
| 摘 要： | 针对河网非恒定流设定情景批量仿真中存在大量的重复计算,提出了一种基于仿真克隆技术的并行累进增量克隆仿真方法.基本思想是:合并相互关联的初始设定情景,并依据关联点对合并后的设定情景进行并行增量克隆仿真.理论分析与实验结果均表明,该方法具有较好的计算性能,在同等时间复杂度下,其计算效率比传统的并行与分布仿真方法提高了两倍以上.
|
| 关 键 词： | 仿真克隆 河网非恒定流 洪水控制调度 并行与分布式计算 |
| 文章编号： | 0253-2778（2007）09-1088-08 |
| 修稿时间： | 2006-06-21 |
Algorithms for cloning simulation of unsteady flow in river networks |
| |
| ZHANG Guo-yi,HUANG Liu-sheng,CHEN Guo-liang. Algorithms for cloning simulation of unsteady flow in river networks[J]. Journal of University of Science and Technology of China, 2007, 37(9): 1088-1095 |
| |
| Authors: | ZHANG Guo-yi HUANG Liu-sheng CHEN Guo-liang |
| |
| Affiliation: | 1.Department of Computer Science and Technology, USTC, He fei 230027, China;2.National High Performance Computing Center at Hefei, Hefei 230027, China |
| |
| Abstract: | Based on the technology of simulation cloning,PPISC(parallel and progressive incremental simulation cloning) approach was proposed to concurrently analyze alternative scenarios of the numerical simulation of unsteady flow in river networks for flood mitigation operation,which was used to optimize the simulation execution by avoiding unnecessary repeated computation among independent scenarios.The basic idea of PPISC was: merging associated scenarios into a compound one and performing the parallel incremental simulation cloning for each compound scenario according to the time sequence of its joint-points.The results of theoretical analysis and numerical test both manifest that the proposed approach has higher computational performance and twice more the parallel efficiency than traditional parallel and distributed simulation methods under the same time complexity. |
| |
| Keywords: | simulation cloning unsteady flow in river network flood mitigation operation parallel and distributed computing |
| 本文献已被 CNKI 维普 万方数据 等数据库收录！ |
