Analysis of Effects of BP Artificial Neural Network Structures on Precision of Flow Forecasting for Manwan Reservoir

A three-tiered artificial neural network (ANN) model with a feed-forward, back-propagation network structure was developed to forecast river flow in the Manwan Reservoir. Based on historical data, various ANN models with different structure were analyzed and tested in order to find a satisfied one. The results of application have showed that the used ANN model is successful and effective in the hydrologic forecasting of the Manwan Reservoir.