Suspended Sediment Load Simulation for Delaware River Using ANNs and Sediment Rating Curves

Predictions of sediment load are required in a wide variety of water resources engineering problems. In the presented study artificial neural network (ANN) was used to model the daily sediment load in the Delaware River, USA. The feed-forward back-propagation ANNs with one, two and three hidden layers were employed for this purpose. A sediment rating curve also was used to compare the results. Compared with classical sediment rating curves, ANN can generate a better fit under the same data requirement.

کلیدواژه ها:

Artificial neural network ، Suspended sediment load ، Feed-forward back-propagation ، Sediment rating curve