Investigating the Performance of New Smart Techniques in Simulation of Groundwater Nitrate.
Subject Areas : Water Resources
1 - Department of Water Engineering, Shoushtar Branch, Islamic Azad University, Shoushtar, Iran
Keywords: Nitrate, Simulation, Artificial Neural Network, Support Vector Machine, neuro-fuzzy.,
Abstract :
Today, due to recent drought, one of the main sources of drinking water in the country is underground resources, and also nitrate is one of the most important pollutants of groundwater resources, which has adverse effects on people's health. The present study seeks to compare and provide an efficient and innovative technique for simulating and predicting nitrate in these resources. Therefore, three artificial neural network techniques (ANN) , neuro-fuzzy inference system (ANFIS) and support vector machine(SVM) are compared in simulation as a data-driven tool. Simulation based on observation samples from wells in the aquifer under study for 13 years and the modeling period has been selected monthly. Estimates of model simulation include magnesium (Mg), bicarbonate (Hco3), calcium (Ca), sodium (Na). First the simulation of each technique is done separately from the various arrangements. Based on the results of the evaluation of the neuro-fuzziness system with the coefficient The correlation between R2 = 9978 and R2 = 0/0002 MSE = are more suitable.
