Modelling waste water treatment performance using artifical neural networks

This study aimed modelling of waste water treatment performance using artificial neural networks. In this study, MATLAB R2008a was used as a modelling tool. The data used in the study were provided from a vegetated submerged bed system (Marahatta, 2004). In this study, some different input parameters were used to determine the treatment performance based on various output parameters. These parameters were, CODinf, CODeff, Total Solidinf, Total Solideff, Volatile Suspended Solidsinf, Volatile Suspended Solidseff and Temperature respectively. According to this model approach, the parameters demonstrated the highest effect on treatment plant performance were COD, TS, VSS, and temperature. Treatment plant data model estimated %98 accuracy. According to the literature, with other kinetic and mathematical models ANN is a very useful tool for modeling full-scale wastewater treatment plants.Anahtar kelimeler: Artificial Neural Networks, Modelling, Waste Water Treatment Plant, Performance, Sequencing Batch Reactor