Radial basis function-neural network model for channel flood routing

The radial basis functionneural network method is applied to channel flood routing. In combination with the specific modes of channel flood routing, radial basis functionartificial neural network models, based on the Muskingum method and flood routing method with a forecast period, are developed, and the least square method is used to determine the parameters of the models. The models are applied to flood routing for two natural river channels, and the results show that the models have the characteristics of fast calculation, high precision, and high value of application.