Prediction of water quality parameters using evolutionary computing-based formulations

Rivers, as one of the crucial components of water resources, play a substantial role in human life. Due to a wide range of water consumptions, water quality is emphatically recommended assigning at a permissible level. Enormous field investigations have been carried out to quantify indicators of water quality parameters which have a great impact on the water quality. In this way, there are not explicit equations to predict WQPs with an allowable degree of accuracy. In the present work, gene expression programming (GEP), evolutionary polynomial regression (EPR), and model tree (MT) have been employed to estimate three indices including biochemical oxygen demand (BOD), dissolved oxygen, and chemical oxygen demand (COD). To develop the proposed models, nine input parameters Ca2+, Na+, Mg2+, NO2-, NO3-, PO43-, EC, PH, and turbidity were selected as effective variables. Results of training and testing stages for these approaches have been investigated. Performance of the models indicated that the relative superiority of the EPR approach compared to the GEP and MT models. Gamma test was applied to determine important parameters for predicting BOD, COD, and COD indices. It was found that PO43-, Ca2+, and pH have the most significant effect on BOD, COD, and COD, respectively.