嘉陵江流域北碚站年输沙量的变化规律及预测研究

Study on the changing characteristics and the prediction of annual sediment transport at Beibei station of Jialin River

首先利用小波变换对北碚站年输沙量的变化规律进行了研究,结果表明:年输沙量呈现出明显的减少趋势;其次,将小波变换结合BP神经网络建立小波网络模型,并利用该模型对北碚站的年输沙量进行预测,同时将预测结果与BP神经网络模型的预测结果进行了比较。认为在缺乏其它相关资料的情况下,单从输沙量和径流量资料出发,小波网络模型的预测效果明显优于BP神经网络模型。由此表明,小波网络模型不仅能对年输沙量的趋势进行预测,还能对年输沙量的大小进行较为准确的预测,从而为在资料较少的情况下进行输沙量的定量分析提供了一种新的方法。

The changing characteristics of annual sediment transport at Beibei station is firstly researched by using wavelet transform in this paper. The results show that the annual sediment transport decreased clearly with time. Based on Back Propagation network and wavelet transform, a wavelet network model is then developed to predict the annual sediment transport at Beibei station. The forecast result by this model is compared with the result by BP network model. The comparison demonstrates that the predicted annual sediment by the wavelet network model is better than that by BP network model if only annual runoff and sediment data are available. So the wavelet network model can forecast not only the tendency of the annual sediment transport, but also the quantity. The new model provides a new way to predict the annual sediment transport under the condition of not having sufficient data.