正实数阶随机弱化缓冲序列多变量灰色预测模型  

Multivariable Grey Prediction Model of Positive Real Order Stochastic Weakening Buffer Sequence

作　　者：刘基伟 闵素芹[1] 金梦迪 Liu Jiwei;Min Suqin;Jin Mengdi(College of Science,Communication University of China,Beijing 100024,China)

机构地区：[1]中国传媒大学理学院

出　　处：《统计与决策》2019年第20期77-81,共5页Statistics & Decision

基　　金：中央高校基本科研业务费专项资金资助项目（3132018XNG1830)

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。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,文章在此基础上提出一种正实数阶随机弱化缓冲序列。通过对不同位置上的观测数据进行不同阶数的弱化缓冲,提高权重分配的精细程度,满足现实系统的预测需求。通过计算显示,使用正实数阶随机弱化缓冲序列可以获得更精准的预测结果。The grey system realizes the assignment of different schemes of the observation data’s weights by constructing the weakening buffer operators of different orders, thus changing the effect of observation data on parameter estimation. The original weakening buffer sequence based on fractional order and positive real order is to weaken buffer of the same order for observation data in different positions. On this basis, the paper proposes a sequence of positive real order stochastic weakening buffers. And then, the observation data at different positions are attenuated by different orders to improve the precision of weight distribution and meet the prediction needs of the real system. The calculation shows that more accurate prediction results can be obtained by using the stochastic weakening buffer sequence of positive real order.

关 键 词：随机阶弱化缓冲 正实数阶 多变量预测 

分 类 号：N941.5[自然科学总论—系统科学]