Effectiveness Comparison of Neuro-Fuzzy and Neural Network methods Artificial in Estimation of Suspended Load Hazard of Rivers(Case Study: Taleghan Basin)

Lots of discharge sediment of cross-cutting are formed Suspended load in many rivers. Sediment delivery and transition load by river flow make many problems for under stream that are included: sedimentation in reserve dam and decrease of impressive volume dams, variation of river direction due to sedimentation in Bed River, decrease of transit capacity of channels and powerhouse transition, variation of quality water from potable and agriculture. Estimation of suspended load is one of important problems for design of reserves, transition volume of sediment, and estimation lakes pollution. In other hand are required for determinate of damages due to sedimentations in environment and determined effecting on the watersheds. There are many methods for estimating load suspended, one of this methods that solved different problems of discharge sediment in flow and can be predict it, used of Nero fuzzy or ANFIS (Adaptive Network Fuzzy Inference System) method, and NNA (Artificial Neural Network) method. These can be related a function between sediment and simultaneous discharge by used of MATLAB and Nero Solution Software's and modeling relations from among variants. The goal of this research is comparison effectiveness Nero fuzzy, neural network artificial and statistical methods for estimating suspended load river in Glinak station of Taleghan Basin. Results this research showed that estimations from Nero fuzzy method by MAE 1006 ton/day, and correlation efficiency (R) 77% and RMSE 2621 ton/day more than Neural Network Artificial are significantly. In contract, fuzzy laws can be illustrated better than methods of neural networks artificial, variation of rivers sediment load. Other profit this method hasn't sensitively to existence some errors in early data. Also are showed that imputed Nero fuzzy method proper reply by increasing train data rather than test data.

Keywords:

Nero fuzzy , Neural Network Artificial , Hazard , Suspended load river , Taleghan