Can artificial neural networks (ANN) accurately predict the performance of wastewater treatment processes? This study explores the application of ANNs as a function approximation tool to predict the performance of a trickling filter treatment process at a municipal wastewater treatment plant in Solon, Ohio, USA. The goal is to determine if ANNs can provide more accurate predictions compared to traditional methods. The treatment plant uses a trickling filter followed by an activated sludge process, with the ANN model being evaluated based on its ability to predict effluent BOD and TSS. The best ANN model achieved prediction errors of 31.45% for effluent BOD and 32.54% for TSS. The research revealed that the number of input variables and nodes in the hidden layer did not have a definite effect on the prediction error for the ANN model. Nevertheless, the prediction errors obtained with ANN models were lower than those obtained by multiple regression analysis, suggesting that ANNs are a more effective tool for predicting effluent quality in this specific wastewater treatment plant. This study demonstrates the potential of ANNs to improve the efficiency and reliability of wastewater treatment processes, highlighting their advantage over traditional methods for predicting effluent quality. These findings have important implications for optimizing treatment plant operations and ensuring environmental compliance.
This paper, featured in Environmental Management and Health, aligns with the journal's emphasis on environmental technologies and the management of environmental resources. By examining the use of artificial neural networks in predicting wastewater treatment performance, the study contributes to discussions on innovative approaches for improving environmental management and health outcomes.