Application of Artificial Neural Networks to estimate pollutants and minerals transferring below the root zone via irrigation with wastewater
2007
Hasan Oqli, A`li Reza | Shafi`ei, Parvin | Liyaqat, A`bd Ol-Majid
In recent years, using Artificial Neural Network was attended as method that can luodel namral pheuomenas. Neural network is similar to human neuron reticulation, that can es1Jmate output parameters from input parameters after training. In this study, perceptorn structure was need to estimnate wastewater pollutants and minerals transferring below the root zone via irrigation. To design the ANN model, datas obtained from lysimetric researches on irrigation of edible vegetables (tomato. carrot and parsley) with treated wastewater were used (Nitrogen and COD or chemical oxygen demand). These datas perpared in another research which was carried out during past years. By using of 110 irrigation water sample measures (effluent of Ekbatan treatment plant) and the same number of drainage water samples (drained from lysimeters), 80% of them (90 samples) were used for training of model and die retnain measures (22 samples) tlsed for test of developed ANN models.The results indicated tlmt ANN model can estinmte Nitrogen and COD leaching with good correlati01l coefficients (R2= 0.98 and 0.81 respectively). Also we can use these trained models to estimate the critical levels of wastewater pollutants in irrigation with treated domestic wastewater to prevent the environmental impacts.
显示更多 [+] 显示较少 [-]