摘要
灰色系统通过构建不同阶数的弱化缓冲算子实现观测数据权重的不同方案的分配,从而改变观测数据对参数估计的影响。由于原有基于分数阶、正实数阶的弱化缓冲序列是对不同位置上的观测数据进行同一阶数的弱化缓冲,因此文章提出随机弱化缓冲序列,进行不同阶数的弱化缓冲,提高权重分配的精细程度,并且提出弱化缓冲阶数的确定方法。通过计算显示,使用正实数阶随机弱化缓冲序列的GM(1,1)模型可以获得更精准的预测结果。
The grey system realizes different schemes of weight assignment of the observation data by constructing weak buffer operator of different order so as to change the influence of the observation data on parameter estimation. Since the original weakening buffer sequence based on fractional order and positive real order is the weakening buffer of the same order for the observation data at different positions, this paper advances a random weakened buffer sequence to perform weakening buffer with different order and improve the precision of weight allocation. The paper also proposes a method to determine the order of weakening buffer. The calculation shows that the GM(1,1) model with positive real order random weakening buffer sequence can obtain more accurate prediction results.
作者
刘基伟
闵素芹
金梦迪
Liu Jiwei;Min Suqin;Jin Mengdi(College of Data Science and Intelligent Media,Communication University of China,Beijing 100024,China)
出处
《统计与决策》
CSSCI
北大核心
2020年第11期37-40,共4页
Statistics & Decision
关键词
随机阶弱化缓冲
正实数阶
GM(1
1)
阶数的确定方法
random order weakening buffer
positive real number order
GM(1
1)
order determination method
