日流量预测的小波网络模型初探  

Study on wavelet network model and its application to the prediction of daily discharge

作　　者：王文圣[1] 熊华康[2] 丁晶[1] 

机构地区：[1]四川大学水利水电学院 [2]湖北清江水电开发有限责任公司

出　　处：《水科学进展》2004年第3期382-386,共5页Advances in Water Science

基　　金：国家自然科学基金资助项目(50279023);四川大学高速水力学国家重点实验室开放基金资助~~

摘　　要：针对日流量时间序列的非线性和多时间尺度特性,提出了将小波分析与人工神经网络相结合进行日流量预测的新方法———小波网络模型。该模型吸取了小波分析的多分辨功能和人工神经网络的非线性逼近能力。以长江寸滩站日流量预测为例,研究表明,所构造的模型各预见期的拟合、检验精度较高。小波网络模型延长了预见期,提高了预报精度,具有广阔的应用前景。Based on the multi-time scale and the nonlinear character of the daily discharge time series, the wavelet network model a hybrid model between wavelet and artificial neural network (ANN), is presented. The suggested model has super advantage with its absorbing some merits of wavelet and artificial neural network. The predicted accuracy has been risen and the length of predicted time has been lengthened. The prediction of daily discharge at Cuntan station of the Yangtze River in China is researched. The results show that the presented model is satisfactory.

关 键 词：日流量预测 小波分析 人工神经网络 小波网络模型 非线性 

分 类 号：P332.4[天文地球—水文科学]