近似非齐次指数序列的离散灰色模型特性研究

针对近似非齐次指数增长规律的数据序列,构建了近似非齐次指数序列的离散灰色模型,给出了该模型的参数求解公式和模型的递推函数。以演绎推理的方法对模型的仿射特性进行研究,分别从数乘变换和平行变换的角度分析了模型中仿射变换前后模型参数、模型模拟预测值的变化特征,在仿射变换条件下可以对原始数据序列进行简化建模。通过仿射变换,可以缩小数据的量级,简化建模过程,而不会改变模型的模拟和预测效果。

Research on the non-homogenous discrete grey model and its parameter's properties

To extend discrete grey model on the data sequences which satisfy to the non-homogenous exponential rule,a novel discrete grey model based on non-homogenous exponential sequence is proposed.The parameters equations and the recursive function of the model are constructed.And the affine properties of the model is studied with method theoretical deduction.The multiple and parallel transformations are the two embranchments of affine transformation.From the angle of this two transformations,the changes of the values of parameters,simulating and forecasting are discussed.The results indicate that the data level can be reduced through transforming,the course of constructing model can be simplified,but the simulative and predicative results cannt be changed.