灰色区间预测的新方法  

A New Method for Grey Interval Forecasting

作　　者：宋中民[1] 

机构地区：[1]华中科技大学自控系,武汉430070

出　　处：《武汉理工大学学报（交通科学与工程版）》2002年第6期796-799,共4页Journal of Wuhan University of Technology（Transportation Science & Engineering）

摘　　要：针对具有离散程序大且总体趋势增长的一类序列 ,给出了一种灰色区间预测的新方法 .该方法是先讨论序列的级比 ,对级比选取一些不同的阈值 ,根据不同的阈值对原始数据列进行分组 ,然后对各个分组序列分别建立非等时距的灰色模型 。Being aimed at a kind of series that the discrete degree is large and the whole tendency is increasing, a new method for grey interval forecasting has given by the grey system theory in this paper. the method firstly calculates the class ratio of series, chooses some clique values for class ratio and divides original dates into groups by different clique value, then, establishes non equigap GM(1,1) model for different group and searches forecasting area of original dates by the solution of model.

关 键 词：级比 非等时距 GM(1 1)模型 

分 类 号：N941.5[自然科学总论—系统科学]