基于BP神经网络的马斯京根模型参数动态估计

Dynamic parameter estimation for Muskingum routing model based on BP artificial neural network

针对平均值法和流量分级法获取马斯京根模型参数导致实时洪水预报精度低的问题,本文提出基于BP神经网络的马斯京根模型参数动态估计方法。提取每场洪水的特征属性作为神经网络的输入,采用优化算法估计的每场洪水的参数作为神经网络的输出,对神经网络进行训练,并将此方法应用于实时洪水预报的参数估计,结果表明,此方法简单可行,精度较高,比较实用。

This paper purposes a dynamic parameter estimation method of Muskingum routing model for real-time flood forecasting based on BP artificial neural network to overcome the low accuracy problem of averaging and grading method in parameter calibration.In this new method,first flood characteristics are analyzed to obtain BP inputs and then an optimization algorithm is used for calibration of model parameters,i.e.BP outputs that are used for neutral network training.Application of this calibrated BP model to real-time flood forecast shows that the model is simple and more accurate.