灰色GM(1,1)模型建模的理论探讨

Discussion on the Grey Modeling of GM(1,1)

在分析GM(1,1)建模中灰导数及其白化背景值对模型精度与适应性的影响的基础上,从灰导数、灰导数的白化背景值的构造证明了原始时间序列数据变化越平缓,发展系数的绝对值越小,GM(1,1)模型的拟合与预测精度越高,模型的适应性越强,同时提出了原始序列数据的一些处理方法。

In this paper, we mainly discuss the influence of grey derivative and white background value to the GM（1, 1 ） model＇s precision. Investigation shows that the GM（1,1） model is very precisive when the original series change slowly and evolution parameter is tiny. So we can deal with the original series by some methods, such as logarithmatics etc.