Optimization pollutants removals from wastewater treatment plant using artificial neural networks

Abstract

Water quality, treatment plant management, and environmental concerns all affect how well a sewage treatment plant performs. Due to the high degree of nonlinearity in the plant as well as the nonuniformity and unpredictability of the influent amount, quality parameters, and operational conditions, modelling the sludge capacity index of the activated sludge method in municipal wastewater treatment plants is a challenging mission. To assess the effectiveness of the al-diwaniyah wastewater treatment plant (WWTP) operation and to estimate quality parameters, the study’s first goal is to improve the WWTP by using artificial neural networks (ANNs). Second, increasing the efficiency of the ANNs model to determine the best WWTP procedure. ANNS were created to predict the sludge volume index (SVI) using the al-diwaniyah WWTP operational and influent quality characteristics. The neural network’s best model for predicting SVI consists of an input node with six input variables, a hidden layer with five nodes, and an output layer with one variable, with an R2 value of 0.965. The outcomes show how effective the right neural network models are at predicting SVI. This is a highly helpful tool that WWTP operators may use in their daily management to improve the effectiveness of the treatment process and the dependability of the WWTP.