摘要
首先提出了一种变换,该变换既可以提高序列的光滑度,而且逆变换后相对误差保持不变,然后针对模拟残差时正时负的现象,又给出了一种变换,该变换可以将任意摆动序列变换为单调递增序列,最后给出了基于残差修正的GM(1,1)模型.将模型应用于实际,预测效果比较满意.
The transformation is given out and it not only improves the smooth of align- ment but also keeps relative error of inversion of transformation steady. Simulation error is sometimes positive and sometimes negative to give anther transformation, which can change arbitrary awing alignment into monotone in creasing alignment and at last gets GM（1,1） model based on residual error correction. The model has been applied to practice and the result of prediction is more satisfactory.
出处
《数学的实践与认识》
CSCD
北大核心
2013年第9期121-125,共5页
Mathematics in Practice and Theory
关键词
数据变换
GM(1
1)模型
交通量
预测
data transformation
GM（.1,1）model
volume of traffic
prediction
