Sadaf Haghiri - PhD Student of Environmental Engineering Faculty of Middle East Technical University, Turkey
Sina Moharramzadeh - M.Sc. Student of Environmental Engineering Faculty of Environmental Engineering University of Tehran, Iran
Ali Nahvi - B.Sc. Student of Engineering Science Faculty of Engineering Science University of Tehran
Amin Daghighi - B.Sc. Student of Civil Engineering Faculty of Civil Engineering University of Tehran, Iran

Nowadays the proper utilization of water treatment plants and optimizing their use is of particular importance. Coagulation and flocculation in water treatment are among the common ways in which the use of coagulants leads to instability of particles and the formation of larger and heavier particles, resulting in the improvement of sedimentation and filtration processes. Determination of the optimum dose of this coagulant is of particular significance. High dose, in addition to adding costs, causes the sediment to remaining the filtrate which would be dangerous according to the standards. Furthermore, sub-adequate doses of coagulants will result in the reduction of the required quality and acceptable performance in the coagulation process. Traditionally, jar tests are used for this case. However, this experiment is faced with many constraints in evaluating the results for the sudden changes in the parameter of input water because of the large costs, required relatively long time, and complex relationships between the many factors that influence the efficiency of coagulant and test results (Turbidity, temperature, pH, alkalinity and etc.). Modeling can be used to overcome these limitations. In this research, artificial neural network MLP With one hidden layer is used for the modeling of the Jar test to determine the dosage of used coagulant in water treatment processes. The data contained in this research are related to the Drinking Water Treatment Plant located in the Ardabil province. To evaluate the performance of the model, the parameters of MSE and the Correlation coefficient R^2 are used. The obtained values are within the acceptable range which shows high accuracy of the models in the estimation of water quality characteristics and the optimal dose of coagulants. Therefore, using these models will allow operators not only to reduce the costs and time taken to perform experimental jar tests, but also to predict a proper dose for the coagulant amounts in real variable conditions and to project the quality of the output water.

Modeling, Artificial Neural Networks, Water Treatment, Testing, Current Testing