Modeling of Tehran South Water Treatment Plant Using Neural Network and Fuzzy Logic Considering Effluent and Sludge Parameters

The disposal of sewage with acceptable qualitative characteristics to different acceptor resources is an environmental issue that today's societies face (with). Using the MatLab software, a neural network model, and an adaptive neuro-fuzzy inference system (ANFIS), this study has predicted the qualitative parameters (COD, BOD5, and TSS of the wastewater, along with TS, VS, and SOUR of the sludge) for the south Tehran sewage treatment plant and finally chosen the best models by validating the model and using the defined criteria. Moreover, using these developed models and comparing their results with the available standard values provides a suitable classification to reuse the wastewater and sludge of the south Tehran wastewater treatment plant. The results indicated acceptable errors of both systems, the adaptive neuro-fuzzy inference system and the artificial neural network, in predicting the qualitative characteristics of the sludge and wastewater of the south Tehran sewage treatment plant and the priority of the adaptive neuro-fuzzy inference system over the artificial neural network in estimating the quality of the treated wastewater and sludge.