摘要
随着电磁仿真技术的发展和仿真需求的增加,仿真验证得到了国内外相关研究领域的广泛重视。近年来,定量评估电磁建模数据差异的特征选择验证(FSV)方法迅速成为研究热点。本文针对FSV方法对含噪声数据失效的问题,引入数据的平滑处理,在保留原始特征的同时,减少噪声的影响。从幅度、趋势、特征点、误差四个方面对平滑后的数据进行性能参数验证和均方根误差计算,该方法在一定程度上改善了FSV方法,并对电磁仿真计算的可信度评估具有一定的参考意义。
With the development of electromagnetic simulation technology and the increase of simulation demand,the validation of computational electromagnetic simulations has attracted broad attention in the domestic and international research fields.In recent years,the feature selection verification(FSV)method for quantitative evaluation of electromagnetic modeling data differences has become a research hotspot rapidly.In this paper,for the certain problem of FSV method’s failure to noisy data,we introduce data smoothing to reduce the impact of noise while retaining the original features and reducing the impact of noise.Perform performance parameter verification and root mean square error calculation on the smoothed data from four aspects:amplitude,trend,feature point,and error.This method improves the FSV method to a certain extent,and has certain reference significance for the credibility evaluation of electromagnetic simulation calculation.
作者
赵颖
王楠
赵勋旺
林中朝
ZHAO Ying;WANG Nan;ZHAO Xun-wang;LIN Zhong-chao(Shaanxi Key Lab.of Large Scale Electromagnetic Computing,Xidian University,Xi’an 710071,China)
出处
《微波学报》
CSCD
北大核心
2020年第S01期41-44,共4页
Journal of Microwaves
基金
国家重点研发计划(2017YFB0202102)
国家自然科学基金(61901323)
中央高校基本科研业务费专项资金(XJS190210)
关键词
特征选择验证
性能参数验证
平滑处理
误差
可信度评估
Feature Selective Validation
performance parameters verification
smoothing
error
confidence evaluation
