摘要
利用样条插值函数对原始数据列进行插值,提取原始数据列所含的信息,然后把插值后得到的函数用于对单变量系统的微分预测模型的参数辨识,从而建立趋势外推预测模型。经检验,此法预测精度较高,且能较好地反映数据的摆动性。
To withdraw information from original data,the spline interpolation function is used. Then,the function obtained through interpolation is used to one variable differential prediction model to inversely resolve the parameters of the model.By these,an tendency reckoning prediction model can be established.Calculation examples show that this prediction method has high precision and can reflect the oscillation of original data.
出处
《河海大学学报（自然科学版）》
CAS
CSCD
1992年第4期47-53,共7页
Journal of Hohai University(Natural Sciences)
关键词
单变量系统
样条插值
微分模型
one variable system
spline interpolation
differential prediction model
