函数变换对灰色建模精度的影响研究  

Influence of function transformation on grey modeling accuracy

作　　者：陈心愿 陈友军[1] CHEN Xinyuan;CHEN Youjun(School of Mathmatics and Information,China West Normal University,Nanchong Sichuan 637002,China)

机构地区：[1]西华师范大学数学与信息学院,四川南充637002

出　　处：《周口师范学院学报》2022年第5期25-31,共7页Journal of Zhoukou Normal University

基　　金：西华师范大学英才科研基金项目(17YC370);西华师范大学教学改革与研究重点项目(jgxmzd 1824)。

摘　　要：探究了函数变换影响灰色建模精度的机理,指出目前函数变换研究中的两个有误的结论:一是通过单一函数变换一定能提高建模精度;二是组合变换一定能提高建模精度。给出了一种解决函数变换还原后相对误差扩大的方法,最后给出了根据相应的先验条件能较大程度上提高建模精度的函数变换建模方法,进一步拓展了灰预测建模研究。Explore the mechanism of function transformation that affects the accuracy of grey modeling,pointing out that two wrong conclusions in the current function transformation research:The first is to improve the modeling accuracy through a single function transformation;the second is that the combination transformation will definitely improve the modeling accuracy.Gives a method to solve the relative error enlargement of function transformation.Finally,the function transformation modeling method which can greatly improve the modeling accuracy according to the corresponding priority conditions is given,which further expands the research of grey prediction modeling.

关 键 词：函数变换 光滑度 级比偏差 凹凸性 建模精度 

分 类 号：N941.5[自然科学总论—系统科学]