基于背景值和初始值改进的灰色预测模型  

Improved Gray Prediction Model Based on Background Value and Initial Value

作　　者：刘俊峰[1] 夏爱生[1] 夏军剑[1] 

机构地区：[1]军事交通学院基础部,天津300161

出　　处：《军事交通学院学报》2015年第6期75-78,共4页Journal of Military Transportation University

摘　　要：基于GM(1,1)背景值的构造方法,提出相邻数据加权平均取值的优化模型,并给出具体的推导过程,同时对初始值进行改进,根据最小二乘原理确定时间响应函数中的参数C,使得预测的平均相对误差最小。结合实例验证改进模型具有较高的预测精度。The paper raises an optimization model of adjacent data weighted average with the construction method based on background value, and shows the specific derivation process. In the meanwhile, it improves the initial value, and minimi- zes the average relative error of prediction according to the least square principle to determine the time response of the C pa- rameter in the function. The example proves that the improved model has higher prediction accuracy.

关 键 词：GM(1 1)模型 背景值 初始值 

分 类 号：O29[理学—应用数学]