【摘要】 传统灰色GM(1,1)预测模型对初始条件的选择存在一定的不足。文章在分别以与作为预测模型初始条件研究的基础上,提出通过对任意两个数据之间进行加权生成,作为初始条件建立新的预测模型,并命名为FGM(1,1)模型。此模型改变了以往只利用原始数列的某一点作为初始条件的方式,也是传统模型与几种改进模型的一种全概括。经实例验证,新模型的拟合与预测精度不低于前几种模型,为最优。更多还原

【Abstract】 Traditional grey prediction model has some deficiencies in selecting initial conditions.On the basis of taking the first and the last vectors as initial condition,a new method with weighting initialization between any two data was put forward,and a model named FGM(1,1)(four variants grey model)was established.The new model changed the previous way of using only one point of the sequence as initial condition,and summarized all the traditonal models.Experimental results showed that the fitting and forecasting precision of new model was better than other models.更多还原