ALTERNATIVE MODELS FOR PREDICTION OF AERATION EFFICIENCY IN STEPPED CASCADES

Abstract

Water quality and its enhancement have a close connection with the presence of dissolved oxygen concentration. A higher dissolved oxygen level indicates better water quality. Aeration is used in water treatment to increase the concentration of dissolved oxygen. Aeration is the process by which the area of contact between water and air is increased, either by natural methods or by hydraulic structures. Aeration enhancement by macro-roughness is well-known in water treatment and one form is the stepped cascade. The macro-roughness of the steps significantly reduces flow velocities and leads to flow aeration along the stepped cascade. In this study, models based on LS-SVM, ANN, and ANFIS methods were developed to predict aeration efficiency in stepped cascades using channel slope, energy-loss ratio and flow regime parameters. Experimental results were compared with the results of these artificial intelligence methods. Good agreements between the experimental results and predicted results prove the validity of the models. The best result was obtained with model based on LS-SVM. The obtained R-2 value for this model was 0.9728.