论文部分内容阅读
东江流域为香港及广州等特大城市的重要水源地,水文过程模拟及预测对流域水资源开发与管理具有重要理论与现实意义.本研究针对小波神经网络中最佳母小波和最佳等级选择问题,在与传统模型比较基础上,研究得出东江流域径流模拟的最佳小波神经网络模型,并以此进行东江流域的径流预测分析.结果表明:1)选择恰当的母小波可以有效捕捉信号统计特征,该流域蒸发量、降水量、温度和湿度的最佳分解母小波分别为Db4,Sym2,Db9及Db4小波,其小波分解最佳等级为5;2)小波神经网络作为新型混合优化模型,在母小波选择和分解等级确定后,经东江博罗站径流模拟分析,确定为东江流域最佳小波神经网络模拟模型.该研究用于东江径流的预测,效果在满意范围内.
The Dongjiang River Basin is an important source of water for metropolitan areas such as Hong Kong and Guangzhou, and the simulation and prediction of hydrological processes have important theoretical and practical significance for the development and management of water resources in the basin.In this study, the optimal mother wavelet and optimal rank selection in wavelet neural networks Based on the comparison with the traditional model, the best wavelet neural network model of runoff simulation in Dongjiang River Basin is derived and runoff prediction analysis of Dongjiang River Basin is carried out. The results show that: 1) The selection of the proper mother wavelet can effectively capture the signal statistics The best decomposition mother wavelet of the basin evaporation, precipitation, temperature and humidity are respectively Db4, Sym2, Db9 and Db4, and the best wavelet decomposition is 5. 2) As a new hybrid optimization model, wavelet neural network After the mother wavelet selection and decomposition level were determined, the simulation model of the Dongjiang Boluo runoff was established as the best wavelet neural network simulation model for the Dongjiang River basin. This study was used to predict the runoff in the Dongjiang River with satisfactory results.