摘要
针对GM(1,N)模型在模拟与预测方面的不足,提出了GM(1,N)模型的一种优化组合方式.第一步是在原GM(1,N)模型灰微分方程上添加一个扰动因素,然后利用优化的背景值确定相应的新参数;第二步利用"最小二乘法"得到模型白化方程近似解中新的初始条件,进而得到一种新的GM(1,N)模型的模拟表达式.实例验证表明,新GM(1,N)模型的适用范围明显拓宽,而且模拟和预测精度均大大提高.
In order to make up the defect of GM( 1, N), optimizing GM( 1, N) model by three steps. The first step is adding disturbance to grey differentia| equation, using a kind of optimized background value to get new parameters, the second step is using "least squares method" to get the initial value of the solution of white differential equation. Through accumulating example, we can see that lhe optimized GM( 1, N) model has higher simulation and forecasting precision obviously.
出处
《四川文理学院学报》
2013年第5期7-10,共4页
Sichuan University of Arts and Science Journal
基金
四川省教育厅科研项目"灰色系统模型的优化及数据预处理应用研究"(13ZB0013)
关键词
GM(1
N)
优化
预测
GM ( 1, N )
optimization
prediction