ارزيابي روش شبكه عصبي مصنوعي در برآورد رسوب معلق روزانه (مطالعه موردي: حوزه آبخيز خارستان)

Evaluation of Artificial Neural Networks Method for Daily Suspended Sediment Yield Estimation (Case Study: Kharestan Watershed)

Estimation of soil erosion and sediment yield in a river is difficult and several methods have been suggested for its estimation. One the new methods in river engineering and suspended sediment estimation is application of artificial neural networks which uses the same algorithm of human brain to find the internal relation between data based on the training process. The objective of current study is to find out the capability of artificial neural networks method for estimation of daily suspended sediment in Kharestan watershed located in the northwest of Fars province, Iran. The study of efficiency is based on the comparison of neural network with regression models. For this purpose, 22 years of water and sediment discharge of Shoor Kharestan River were considered and tested for outliers. Then the estimation was done based on neural networks and linear regression method (sediment rating curve) and were compared based on RMSe, MAE and R2. The results showed that estimation of neural network is more accurate than linear regression (sediment rating curve). The estimations of RMSe, MAE and R2 for neural networks method was 19.27, 12.14 and 0.98 respectively while these values for linear regression were 36.84, 20.75 and 0.74 which showed the lower errors of neural networks method compared with linear regression.