基于人工神经网络的长江三峡年最大洪峰流量长期预报

将统计相关性分析与模糊方法相结合,识别出影响宜昌站年最大洪峰流量的前期流域降水、大气环流形势等预报因子。比较了LMBP算法、自适应BP算法、变步长BP算法、加动量BP算法等几种常见改进BP算法的优缺点,探讨了BP网络建模过程中存在的问题,建立了LMBP算法和自适应BP算法相耦合的中长期水文预报模型。预报实践表明,所建模型合理、预报效果好、精度高、具有较高的推广和应用价值。

Long-term Prediction of Annual Maximum Peak Discharge at Yangtze Three Gorges Based on Artificial Neural Network

An artificial neural network medium and long-term hydrological forecasting model combining with LM algorithm and self-adaptive algorithm was established in combining statistical analysis with fuzzy analysis,choosing predictors such as rainfall and atmospheric circulation in earlier stage which effect the annual maximum peak discharge at Yichang Station of Yangtze River,considering of comparing the advantage and disadvantage of several modified BP algorithms,discussing several problems in the modelling process.The result of calculation shows that tie models are highly effective and are worth popularization and application.