并行分布仿真中前瞻量问题的研究  

Research on Lookahead Problem in Parallel and Distributed Simulation

作　　者：王学慧[1] 李革[1] 邱晓刚[1] 黄柯棣[1] 

机构地区：[1]国防科技大学机电工程与自动化学院军仿332教研室,湖南长沙410073

出　　处：《系统仿真学报》2006年第3期623-625,632,共4页Journal of System Simulation

基　　金：重点国防预研项目(413040402)

摘　　要：并行分布式仿真对复杂大规模动态系统的研究,以及探索其长远的应用空间提供了便利,近年来日益成为研究的热点。时间管理是影响并行分布仿真系统高效运行的重要因素之一。首先介绍了前瞻量的概念与设置方法,其次提供了前瞻量的形式化描述,然后深入分析了由于引入零前瞻量而可能导致的回退循环问题,紧接着针对这一问题提出了解决方案及优化策略。最后,分析了前瞻量变动对系统的影响,并提出了一种在RTI中如何实现动态更新前瞻量的算法。Recently, there has been a great deal of interest in parallel and distributed simulation, which runs on multiple processors to accelerate simulation. It facilitates the research of large-scale complex dynamic systems, even to explore far-ranging application spaces. However, time synchronization is one of the important factors that affect the PDS running performance. The conception of the lookahead and the method of its selection were introduced, and then the formalized description of the lookahead was provided. Moreover, the rollback recurrence problem was researched that is likely to result from zero lookahead, and also the solution and the optimum strategy were put forward. Finally, on the basis of discussing the effect of the lookahead alteration, t a novel lookahead dynamic update algorithm in RTI was designed.

关 键 词：并行分布仿真 前瞻量 形式化描述 回退循环 动态更新算法 

分 类 号：TP391.9[自动化与计算机技术—计算机应用技术]