改进小波-ARMA分频模型在径流预测中的应用  

Application of the improved wavelet-ARMA model in runoff prediction

作　　者：何一韬[1] 朱永忠[1] 

机构地区：[1]河海大学理学院,江苏南京211100

出　　处：《云南民族大学学报（自然科学版）》2016年第2期145-151,共7页Journal of Yunnan Minzu University:Natural Sciences Edition

基　　金：国家自然科学基金(50979029);河海大学自然科学基金(2008431111)

摘　　要：在简要分析径流影响因素的基础上,将改进的小波-ARMA分频模型应用到年径流预测中.讨论不同小波分解方式和拓延方式下沱沱河站年径流量预测,并将预测效果与其它3种模型预测进行比较.经过实证研究结果表明:整体分解方式下模型预测精度高,分步分解方式下模型实用性强.该方法拓宽了小波分析理论在影响因素不稳定条件下年径流量研究的应用范围,为径流量的短期科学预测提供了一种新方法.With a brief analysis of the influencing factors of runoff, this paper intends to apply the improved wavelet - ARMA frequency - division model to the annual runoff prediction. It discusses the annual runoff prediction of the Tuotuo River with a consideration of the different ways of wavelet decomposition and future extension, and compares the prediction result with the results by three other models. It concludes that the prediction accuracy of this model with integral decomposition is very good, while the practicability of this model with multi - step decomposition is good. This method expands the application of the theory of wavelet analysis to the annual runoff research with unstable influencing factors, and provides a new approach to the short - term runoff prediction in a scientific way.

关 键 词：小波分析 ARMA模型 整体分解 分步分解 径流预测 

分 类 号：O29[理学—应用数学]