[1]严修红,许伦辉,董世畅.基于数据预处理灰色神经网络组合和集成预测[J].智能系统学报,2007,2(04):58-62.
 YAN Xiu-hong,XU Lun-hui,DONG Shi-chang.Grey neural network and integrated forecasting based on preprocessed data[J].CAAI Transactions on Intelligent Systems,2007,2(04):58-62.
点击复制

基于数据预处理灰色神经网络组合和集成预测(/HTML)
分享到:

《智能系统学报》[ISSN:1673-4785/CN:23-1538/TP]

卷:
第2卷
期数:
2007年04期
页码:
58-62
栏目:
学术论文—智能系统
出版日期:
2007-08-25

文章信息/Info

Title:
Grey neural network and integrated forecasting based on preprocessed data
文章编号:
1673-4785(2007)04-0058-05
作者:
严修红12许伦辉2董世畅1
1.顺德区容山中学,广东顺德528303;
2.江西理工大学机电工程学院,江西赣州341000
Author(s):
YAN Xiu-hong12XU Lun-hui2DONG Shi-chang1
1.Rongshan Middle School of Shunde County,Shunde 528303, China;
2.Institute of Electromechanical Engineering,Jiangxi University of Scie nce and Technology,Ganzhou 341000,China
关键词:
时间序列预测灰色神经网络组合预测
Keywords:
time series forecasting grey neural networkcombined forecasting
分类号:
U491.14
文献标志码:
A
摘要:
当研究的系统扰动因素过大或系统行为在某个时间点发生突变,出现严重扰动系统的异常数据时,提出不应直接按原始数据建模预测,而应根椐实际情况适当地对数据预处理. 提出了基于数据修正的改进型灰色神经网络组合和集成预测,并根据南昌火车站旅客发送量时间序列建立了多个模型,从模型预测效果对比中说明数据修正﹑改进型灰色模型和改进型灰色神经网络﹑灰色神经网络组合和集成确实能提高预测精度.另外,修正数据要把握一个度,不能修正全部数据,只能修正较异常的数据,要在数据的趋势性和预测的灵敏性间取得平衡.
Abstract:
When a system disturbance is too great or a sudden change occurs, the resulting abnormal data can severely disturb the forecasting system. In this sit uation,running a forecasting model before abnormalities in the original data ar e identified produces misleading results. In this paper, an improved grey neural network forecasting model and integrated forecasting method are proposed on the basis of data modification. Several forecasting models were tested based on tim e sequences of passenger volume in Nanchang Railway Station. After comparing mod el predictions with real data, it became clear that prediction accuracy is consi derably improved with revised data, or an improved grey model, or a combined gre y neural network. But the data modification must be done properly. Not all data should be modified, it is only necessary to modify abnormal data in order to mai ntain balance between the data tendency and forecasting sensitivity.

参考文献/References:

[1]陈泽淮,张 尧,武志刚.RBF神经网络在中长期负荷预测中的应用[J].电力系统及其自动化学报,2006,18(1): 15-19.
 CHEN Zehuai,ZHAGN Yao,WU Zhigang.Application of RBF neural network in mediu m and longterm load forecasting[J]. Proceedings of the Chinese Society of U n iversities ,2006,18(1): 15-19.
[2]中国国家统计局.中国统计年鉴2005[M].北京:中国统计出版社,2005.、
[3]邓聚龙.灰色系统理论教程[M].武汉:华中理工大学出版社,1990
[4]王翠茹,孙辰军,杨 静,冯海迅.改进残差灰色预测模型在负荷预测中的应用[ J].电力系统及其自动化学报,2006,18(1):86-89.
 WANG Cuiru,SUN Chenjun,YANG Jing,FENG Haixun. Application of modified resid ual error gray prediction model in power load forecasting[J]. Proceedings of t he Chinese Society of Universities ,2006,18(1):86-89.
[5]祖恩三.云南GDP的灰色预测和分析[J].经济师,2006(6):272-274.
ZU Ensan. A grey forecasting model and its application in GDP forecasting of yu nnan[J].China Economist,2006(6):272-274.
[6]张大海,江世芳,史开泉.灰色预测公式的理论缺陷及改进[J].系统工程理论与实践,2002(8):140-142.
 ZHANG Dahai,JIANG Shifang,SHI Kaiquan. Theoretical defect of grey prediction for mula and its improvement[J]. Systems EngineeringTheory & Practice,2002(8):140-142.

备注/Memo

备注/Memo:
收稿日期:2006-09-30.
基金项目:
国家自然科学基金资助项目(60664001);
 江西省自然科学基金资助项目(0511030)
作者简介:
严修红,男,1974年生,一级教师,主要研究方向为智能预测.
 E-mail:yxh3o9@163.com.
许伦辉,男,1965年生,教授,主要研究方向为智能交通系统、交通环境与交通安全、交通系统建模与仿真,发表论文50多篇.
董世畅,男,1968年生,一级教师,主要研究方向为中学物理.
更新日期/Last Update: 2009-05-07