仿真数据质量及VV&C研究  

Research of simulation data quality and VV&C

作　　者：许国珍[1] 焦鹏[2] 査亚兵 

机构地区：[1]国防科学技术大学机电工程与自动化学院,湖南长沙410073 [2]国防科学技术大学信息系统与管理学院,湖南长沙410073

出　　处：《计算机工程与科学》2013年第4期168-172,共5页Computer Engineering & Science

基　　金：国家自然科学基金资助项目(60904055);国家部委基金资助项目(9140A04010110KG0110)

摘　　要：仿真系统由模型和仿真数据组成,仿真系统可信性取决于模型可信性和仿真数据的质量,在仿真系统过程开发中,仿真数据质量主要依靠数据校核、验证与认证(VV&C)来保证。缺乏适用的VV&C实施策略和工具,是目前数据VV&C面临的重要问题。本文首先介绍仿真数据质量和VV&C的基本概念,并从数据生产者和数据使用者两个角度分析VV&C在数据生产、使用周期中的作用;然后针对不同的数据参与者,建立VV&C实施策略:数据生产者通过质量元数据模板记录数据质量,数据使用者通过VV&C过程模型开展VV&C活动;最后基于VV&C过程模型,借鉴工作流的管理思想,开发出VV&C管理工具PVV&CM,实现对VV&C活动流程化管理,保证VV&C工作的规范开展。Simulation system is composed of models and simulation data. The credibility of simulation system depends on the credibility of models and the quality of data. The simulation data quality relies on data verification, validation and certification when simulation system is developed. The lack of implementation strategies of VVC and tools is a big problem. Firstly, the concept of data quality and VV＆C was introduced in this paper. Secondly, the effect of VV^C was analyzed in the circle of data manufacture and usage from the views of data producer and user. Thirdly, the implementation strategies of VV＆C for different data participants were employed： data quality can be recorded by data producer with data quality metadata, and VV＆C;can be carried out by data user with VV＆C process model. Finally, based on VV＆C process model, PVV＆CM for VV＆C management was developed by using management workflow, and the VV＆C process can be managed as a flow and be carried out with standard.

关 键 词：数据质量 校核 验证与认证 元数据 过程模型 

分 类 号：TP391.9[自动化与计算机技术—计算机应用技术]