仿真模型的智能化验证方法  

Intelligent validation method of simulation model

作　　者：闵飞炎[1,2] 杨明[2] 王子才[2] 

机构地区：[1]中国船舶重工集团公司第七○七研究所,江西九江332007 [2]哈尔滨工业大学控制与仿真中心,哈尔滨150001

出　　处：《大连海事大学学报》2010年第1期59-64,共6页Journal of Dalian Maritime University

基　　金：国家自然科学基金重点资助项目(60434010);黑龙江杰出青年基金资助项目(JC200602)

摘　　要：为实现仿真模型可信度验证过程的自动化,并解决验证技术在应用中过度依赖领域专家经验等问题,提出基于知识的模型验证方法.首先,对模型验证的客观和主观方法中的数据和信息转化过程,以及用于模型验证中真实系统测试数据和领域专家的经验等参考信息进行抽象和归纳.其次,在知识抽象的基础上,分别建立领域知识模型、验证技术模型和验证任务模型.最后,在知识模型的基础上设计和实现智能化验证系统.对电磁轨道炮仿真模型的验证表明,该方法可有效对仿真模型进行验证,且主观和客观的验证方法均可以集成到该系统中,提高了模型验证效率.To automate the validation of simulation models and settle the problems of excessive dependence on domain expert experience,knowledge-based model validation method was proposed.First,the knowledge involved in subjective and objective validation method was induced.Second,the domain knowledge model,validation technique model and validation task model were established respectively based on knowledge abstraction.And then intelligent validation system was designed and implemented based on knowledge model.Validation test of the electromagnetic rail gun simulation system shows that the proposed method can automate validation procedure effectively,and both subjective and objective methods are integrated,which can improves validation process of simulation system.

关 键 词：仿真模型 智能化验证 基于知识的方法 

分 类 号：TP391.9[自动化与计算机技术—计算机应用技术]