同时改进背景值和初始条件的一种GM（1，1）预测方法

摘要

Traditional GM (1,1) model has some deficiencies in selecting background value and original conditionals .In order to im-prove the prediction precision and fitting precision .In this paper , a new model is proposed that the background value and the original condition of the model are simultaneous optimization .The new model changed the previous way of using only background value or orig -inal condition, integrating two impact conditions .Finally, comparison of the fitting data and forecast data of the model with several models through an example , indicate that the prediction precision of the new model are better than those of the previous model .%传统GM（1，1）模型在背景值和初始条件的选取方面存在不足，为了提高灰色GM（1，1）模型的预测精度和拟合精度，本文提出了一种同时改进背景值和初始条件选取的新模型，此模型对以往单独对背景值或初始条件的优化进行了改进，将两个影响条件整合到一起。最后通过实例将该新模型与几种模型的拟合和预测数据比较，显示出了新模型的预测精度高。