HLA/RTI网格化与以模型为中心的分布仿真互联模式研究  

Research on HLA/RTI gridding and model-centered distributed simulation connecting mode

作　　者：邸彦强[1] 柴旭东[2] 李伯虎[1] 唐震[1] 余海燕[3] 

机构地区：[1]北京航空航天大学自动化学院,北京100083 [2]航天科工集团二院,北京100854 [3]中国科学院计算技术研究所,北京100080

出　　处：《计算机集成制造系统》2006年第4期504-510,共7页Computer Integrated Manufacturing Systems

基　　金：国家863/CIMS主题资助项目(2004AA104330)。~~

摘　　要：应用网格技术对分布仿真高层体系结构和运行时间框架软件的共享进行研究,实现“以模型为中心”的大规模、大范围分布仿真的分布互联模式。首先,设计运行时间框架守卫模块,实现对运行时间框架进程运行状态的监控、高层体系结构多联邦信息的管理和监视。在此基础上,按照Globus Toolkit 3网格服务规范对运时时间框架守卫模块服务化,并基于具有我国自主知识产权的织女星网格操作系统,进行同类资源的整合。此外,基于VEGA GOS的服务作业处理和批作业处理,开发了运行时间框架网格服务的客户端,实现对共享运行时间框架资源的异地访问。最后,简要介绍了该方法在863计划仿真网格项目中的实现和应用情况。实验表明,该方法可以实现对运行时间框架资源的共享和有效监控,并且在满足一定带宽的网络上实现基于共享运行时间框架的仿真联邦运行。Grid technology was applied to study the sharing of distributed simulation High Level Architecture （HLA） /Run-Time Infrastructure （RTI） software, which has realized model-centered organization mode for large scale distributed simulation system. Firstly, RTI daemon module was designed to monitor RTI process running state so as to manage all the executions of HLA simulation federations. According to the grid service specification of Globus Toolkit 3, RTI daemon grid service was developed and deployed into service container Tomcat, thus the grid service was managed by VEGA Grid Operating System（GOS）,which is Chinese product with proprietary intellectual property. In addition, a RTI grid client was designed with GOS＇s job and batch job process service, through which the shared RTI resource could be accessed. The research result was applied in Simulation Grid Project of China National High-Tech. Program,which showed that the method could effectively share and monitor RTI resource, and some simulation federations could execute on the shared RTI based on the network with certain band width.

关 键 词：分布式仿真 网格 高层体系结构 运行时间框架 网格操作系统 

分 类 号：TP391.9[自动化与计算机技术—计算机应用技术]