基于背景值改进的GM(1,1)预测模型的研究及其应用

Modeling Research of Model GM(1,1) Background Value-Building Based on Gauss-Legendre Quadrature and Its Application

从GM(1,1)预测模型的建模机制出发,分析出GM(1,1)背景值对预测模型的精度有着重要的作用,提出了利用线性函数插值和高斯公式来构造模型中的背景值,利用新的背景值计算公式能够有效的提高模型的预测精度。最后用本文提出的方法对我国人均发电量建立了灰色模型,结果表明了本文提出方法的有效性。

In this paper, the authors proposes that the building approach method of background value in grey mode GM（1,1） is important and that rationally constructing background value can improves precision and adaptability of the grey model. The method of background value reconstruction in grey GM（1, 1） model based on Gauss-Legendre formula of numerical value integral is pointed out. Simulation examples show the effectiveness of the proposed approach.