Comparing Three Main Methods of Artificial Intelligence in Flood Estimation in Yalphan Catchment

Estimation of discharge as one of the major issues in water resource management and flood control has a key role in the success of water construction design and efficiency of Bio-Mechanical proceeding in catchments. In this research, discharge Peak of Yalphan River has been simulated using three main methods of artificial intelligence (MLP neural network model, subtractive clustering and ANFIS model, and the combination of neural network and genetic algorithm). For this purpose, 8 parameters have been prepared as input data (2001-2012) including precipitation in the event day, precipitations during 5 days before the event day, base flow in the event day and CN map. Peak of flow has been considered as output data of models. RSME, MAE and NSE indicators has been used to assess the artificial intelligence models. Output data of neural network model have been imported to the combined model of neural network and genetic algorithm. Also, output data of subtractive clustering model have been imported to ANFIS model. Finally three models have been assessed using the mentioned indicators. The results showed that the combined model of neural network and genetic algorithm is better than the other models in Yalphan Catchment.