一种非线性组合预测方法在径流预测中的应用

Application of a Kind of Nonlinear Combination Forecast Method in Runoff Prediction

将基于人工神经网络的非线性组合预测方法应用于径流预测中 ,以 1996年太阳沱的洪水资料为例 ,对已经建立的 2个模型 ,采用 3层的BP网络进行组合模拟预测 ,从多方面分析比较 ,证明用该种方法能够提高预测精度 。

Nonlinear combination forecast method based on artificial neural network is used in runoff prediction. Taking the flood data of Taiyangtuo in 1996, three levels of BP network is used to simulate and predict by two models established. The analyses from different aspects show that this method can improve prediction precision, and obtain satisfying results.