基于背景值和边值修正的GM(1,1)模型优化

GM(1,1) model optimization based on the background value and boundary value correction

针对灰色GM(1,1)预测模型提高精度的问题,提出了新的背景值优化公式代替传统的背景值优化公式,再进行边值修正的方法.该方法采用新的背景值优化公式求出紧邻均值生成序列,并使用均方误差和最小准则,针对原始序列和生成序列进行边值的修正.通过对优化后的模型实证测算,验证了修正后的模型在提高预测精度上的可行性和有效性.

To improve the accuracy of GM（1,1） prediction model, a new optimization model based on the combination of optimized background value and the boundary value was proposed. This method used a new optimal background value formula generates the mean sequence, then by minimizing the sum of the squared error, gave calculation method of determine the boundary value correction. The simulation result verifies that the proposed method is feasible and efficient.