基于HHT方法的径向基神经网络年径流预测模型  

Yearly Runoff Prediction Model of RBFNN Based on HHT

作　　者：石教智[1] 林汝颜[2] 

机构地区：[1]广东省水利厅,广东广州510635 [2]广东省水利电力勘测设计研究院,广东广州510170

出　　处：《广东水利电力职业技术学院学报》2007年第2期34-38,共5页Journal of Guangdong Polytechnic of Water Resources and Electric Engineering

摘　　要：在华南地区径流演化研究中基于新的预测理念,首次将HHT方法的自适应多分辨分析与人工神经网络的强有力逼近功能结合起来,建立基于HHT方法的径向基神经网络预测模型,并以东江流域博罗站近四十五年的年径流序列为例对模型进行验证,得到较理想的结果,从而丰富了径流预测的理论和方法。On the basis of new predicting theory,this paper firstly couples HHT with ANN and builds a new model in the research on variation of basin runoff in South China——a yearly runoff prediction model of RBFNN based on HHT.Having been used to testify the yearly runoff dataset of Boluo Station,East River Valley for nearly forty-five years,the model proves to be workable and it enriches the theory of runoff prediction for a moderate & long period.

关 键 词：HILBERT-HUANG变换 径向基神经网络 年径流 预测 

分 类 号：P333.9[天文地球—水文科学]