基于初值改进的多变量MGM(1,m)模型研究

The Study on the Improved Multiple Variable MGM(1,m) Model Based on Improving the Initial Value

传统多变量MGM(1,m)模型在求解灰色微分方程组时以序列矩阵X^((1))中各序列的第一个分量组成的向量作为初始条件进行建模,没有充分利用新信息。基于新信息优先原理,在传统多变量MGM(1,m)模型建模的基础上,本文提出把序列矩阵X^((1))中各序列的第n个分量组成的向量作为灰色微分模型的初始条件,对传统多变量MGM(1,m)模型进行改进,以提高模型的预测精度。最后,通过一个算例验证模型的改进效果。

The traditional multiple variable MGM(1,m) model takes the vector composed of the first component of each sequence of sequence matrix X(1) as the initial condition to conduct modeling.However,the impact of new information on the prediction precision isn' t taken into account.Based on the new information priority theory in grey theory,the traditional multiple variable MGM(1,m) model is improved.That is,the vector composed of the nth component of each sequence in sequence matrix X is taken as the initial condition of grey differential equations so as to improve the traditional multiple variable MGM(1,m) model to enhance the prediction precision of model well.Finally,an example is illustrated to verify the improved effect of model,which can be concluded that improving the initial value of traditional MGM(1,m) model is able to enhance the simulation accuracy and prediction accuracy.