学术
摘 要:
提出了一种基于GIS与小波神经网络方法相结合构建而成的水库日径流预测模型（GWNNR）,通过模糊C均值聚类分析将水库历史径流数据分成4类,并分别建立相应的小波神经网络预测模型,应用遗传算法（Genetic algorithm）和误差反传递（Back-propagation）算法对模型的参数进行优化,对某水库2005年日平均来流进行分类预测,结果表明,该方法具有较好的训练速度和较高的预测精度。[著者文摘]
文章出处:
《水力发电学报》-2008年27卷1期 -5-10页
分 类 号:
文献标识码:
A
Research on runoff forecasting of reservoir by using fuzzy clustering analysis and wavelet neural networks
SONG Hailiang, SONG Yang, ZHANG Denghua, ZHONG Wei （ School of Civil Engineering, Tianjin University, Tianjin 300072）
Abstract:
This paper presents a daily runoff forecasting method by using wavelet neural networks, the fuzzy Cmeans clustering analysis and the wavelet neural network. The historical runoff data are divided into four categories by using the fuzzy C-means clustering. The corresponding wavelet neural network is built, the parameters of model are optimized by genetic algorithm and the back-propagation algorithm. The daily mean runoff of reservoir in 2005 is forecasted under categories by the corresponding wavelet neural network. The forecasting result shows that the proposed method possesses the faster training speed and the greater forecasting accuracy.[著者文摘]
Key words:
engineering management; runoff forecasting; wavelet neural networks; fuzzy C-means clustering; genetic algorithm;back-propagation algorithm
基金资助:
国家杰出青年科学基金（50525927）资助项目
更多>>免费文章
cqvip.com