社会计算实验动态仿真引擎的研究与实现  

Research and Implement on Dynamic Simulation Engine for Society Computing Experiments

作　　者：周云[1] 乔海泉 邱晓刚[1] 黄柯棣[1] 胡德文[1] 

机构地区：[1]国防科技大学机电工程与自动化学院,湖南长沙410073 [2]北京清河大楼子6,北京100085

出　　处：《国防科技大学学报》2011年第4期163-167,共5页Journal of National University of Defense Technology

基　　金：国家自然科学基金资助项目(61074108,91024030/G30)

摘　　要：应对新形势下非常规突发事件的严峻挑战是各国政府亟待解决的问题,综合人工社会-计算实验-平行执行的ACP方法,是解决非常规突发事件应急管理问题的有效方法。本文引入ACP方法,阐述构建高性能社会计算实验动态仿真引擎的作用;按照层次化模块化思想设计动态仿真引擎D-PARSE的体系结构,详细描述各组成部分的具体功能,并基于并行仿真引擎μsik实现所设计的动态仿真引擎D-PARSE;测试了不同事件发送策略对仿真系统性能的影响。Responding to the serious challenge of tmeonventional emergencies under the new situation is the urgent problem of governments. Integrating Artificial Societies---Computational Experiments--Parallel Execution, ACP is an effective method to solve the unconventional emergency management. Firstly, the ACP method was introduced, which was used as the dynamic simulation engine for constructing high performance society eornpufing experiments. Then, the architeeture of the dynamic simulation engine D-PARSE was designed following the concept of hierarchy and modularity. The specific function of each component was described in detail. D-PARSE was implemented based on the （sik parellel simulation engine. Finally, the influence of different event-send-strategy on simulation system performance was tested.

关 键 词：动态仿真引擎 通信库 微内核 微进程 

分 类 号：TP391.19[自动化与计算机技术—计算机应用技术]