本文已被:浏览 688次 下载 1136次
中文摘要: 本文使用常见的15类函数,采取粗选与精选两步,选择非线性预报因子。根据预报对象与因子时空尺度相匹配的原则,确定待选因子,建立非线性回归预报方程。试验表明,预报因子出现的形态高达390类,且非线性方程的效果优于线性方程。
中文关键词:
Abstract:In this paper, fifteen kinds of ordinary functions were applied to select the nonlinear predictors. Coarse and fine selections were accepted. A nonlinear forecasting equation has been built. The results showed that the predictors had 390 kinds of formation and the effects of the nonlinear forecasting equation in this method were better than those of the linear equations.
keywords:
文章编号: 中图分类号: 文献标志码:
基金项目:
作者 | 单位 |
姜子俊 | 内蒙古自治区气象学校 |
乔志敏 | 内蒙农牧业环保站 |
Author Name | Affiliation |
引用文本:
姜子俊,乔志敏,1991.非线性回归方程的探索[J].气象,17(12):13-16.
,1991.Probe of nonlinear regression equation[J].Meteor Mon,17(12):13-16.
姜子俊,乔志敏,1991.非线性回归方程的探索[J].气象,17(12):13-16.
,1991.Probe of nonlinear regression equation[J].Meteor Mon,17(12):13-16.