并基于Matlab程序对不同类型的数据建立灰色GM(1,1)模型
从理论上分析灰色GM(1,1)模型预测指数序列产所生的相对误差特性
并基于Matlab程序进行实验做出相对误差特性曲线
而实际中需要预测的数据并不可能完全服从指数分布
从相对误差曲线特征入手建立误差修正模型
因此建立了随机值概念来表示实际数据对指数分布的偏离
然后基于指数序列建模
做出相对误差特性曲线和相对误差的一次函数拟合曲线
） 摘要： 首先介绍了灰色GM(1,1)建模机理
） Abstract： First the mechanism of grey model GM(1,1) is introduced, then the feature of the relative error of gray model GM(1,1) using exponential series is analyzed theoretically
The relative error feature diagram of the model is presented based on the experiment using Matlab
It is found that the relative error is approximatively linear
But actually the data which needs to forecast does not completely obey index distribution
So random value is used to express the departure of the actual data from exponential distribution
The grey model GM(1,1) is built by using different types of data based on Matlab
Relative error feature diagrams and its linear function fitting curves are presented
For the feature of the relative error curves, the error correction model is built
The practical application results show that the error correction model can improve the forecasting precision
Keywords： GM(1,1);experimental study;relative error;error correction model 下载PDF阅读器 PDF全文下载： 初稿 （ 308 ） 作者简介： 通信联系人： 【收录情况】 中国科技论文在线： 陈鹏宇
灰色GM(1,1)模型误差分析及误差修正模型研究[EB/OL]
