基于HHT的仿真模型动态一致性方差分析方法  

Variance Analysis Method for Simulation Model Dynamic Consistency Base on Hilbert-Huang Transform

作　　者：陈永兴[1] 吴晓燕[1] 曾华[1] 

机构地区：[1]空军工程大学防空反导学院,西安710051

出　　处：《电子科技大学学报》2014年第4期568-574,共7页Journal of University of Electronic Science and Technology of China

基　　金：陕西省自然科学基础研究计划(2012JM8020)

摘　　要：仿真模型动态数据常具有非线性和非平稳性,如何有效提取动态数据的验证指标和利用动态数据指导模型修改是仿真模型动态验证需要解决的重要问题。该文提出了基于希尔伯特?黄变换(HHT)仿真模型动态一致性方差分析方法,将HHT得到的动态数据时?频分布作为验证特征,建立了基于方差分析的仿真动态一致性检验模型,实现了仿真模型的定量动态验证,解决了差异源的定位问题,为提高模型可信性提供了指导。仿真和实例应用表明,该方法能较为精确地检验动态数据的一致性和实现误差定位,具有一定的可行性和实用性。With respect to the nonlinear and nonstationary of dynamic simulation model data, the methods that can extract validation metric and guide model modification are the challenges faced by simulation dynamic validation. The variance analysis method for simulation model dynamic consistency based on Hilbert-Huang transform （HHT） is proposed in this paper. Firstly, taking the time-frequency distribution of dynamic data obtained by HI-IT as validation metric, the dynamic consistency test model based on variance analysis is built up. And then, the quantitative validation of simulation model and the location of error source are achieved by variance analysis. Finally, simulation and case study show the practicability and feasibility of this method in the validation of dynamic consistency and the localization of error source.

关 键 词：动态验证 希尔伯特-黄变换 仿真模型 方差分析 

分 类 号：TP391.9[自动化与计算机技术—计算机应用技术]