Dissolved oxygen estimation using artificial neural network for water quality control

Abstract

Dissolved oxygen (DO) is one of the key parameters when analyzing river water quality. Correct estimation of DO being carried by a river is very important for water quality control. DO is affected by lots of variables such as decomposition, nitrification, reaeration, sedimentation, photosynthesis, water discharge and temperature for that reason it is hard to solve such a complex problem. The methods available in the literature for DO estimation are complicated, time consuming and necessitate numbersome parameter estimation procedures. Artificial Neural Networks (ANNs) are simply mathematical representations of the functioning of the human brain. This paper examines the potential of ANN in estimating the DO from limited data (NO2-N, NO3-N, BOD, water discharge and temperature). This study employed feed forward (FF) type ANN for computing monthly values of DO. The results of the study clearly demonstrate that the ANN results are very close to the observed values of DO.