A Novel Approach for Estimation of Sediment Load in Dam Reservoir With Hybrid Intelligent Algorithms

Predicting the amount of sediment in water resource projects is one most important measures to be taken, while sediments have an unknown nature their behavior. In this research, using data recorded at Mazrae station between 2002 and 2013, catchment area Maku Dam has been predicted different models intelligent algorithms. Recorded including river flow (m 3 /s), concentration (mg/L), temperature (°C) were considered input data, load (ton/day) was output data. Initially, correlation test, relationship each with considered. The results show high low these order find best combination for prediction, single, binary, triple sensitivity analysis. achieve purpose study, first classical adaptive neuro-fuzzy inference system (ANFIS), predicted, then evolutionary algorithms ANFIS training, performance examined. used study ant colony optimization extended continuous domain, particle swarm optimization, differential evolution, genetic algorithm. showed that system–ant system–particle system–genetic algorithm, system–differential had predicting load. meantime, it observed coefficient determination, root mean square error, scatter index test mode domain algorithm prediction dataset (sediment + flow) are equal 0.991, 13.001, (ton/day), 0.112, those weakest (temperature 0.490, 107.383 0.929, respectively. present use training able improve its Dam.