Adaptive Neuro-fuzzy inference system (ANFIS) model for Management of an Activated Sludge Process in Wastewater Treatment Project

Abstract

Abstract: The activated sludge process is the most versatile technological method used for urban wastewater treatment system. However, complex, non-linear and uncertain nature of the process made building its model quite challenging. The existing models are highly complex to use for optimization, estimation and control purposes. Therefore, there is an urgent need for suitable, efficient, simple and easy to use process model. An adaptive neuro-fuzzy inference system (ANFIS) model is going to be presented in this paper for an activated sludge process. High flexibility, less computational cost, accuracy, robustness, fast learning and adaptation capabilities are some of the enticing features taken into account for choosing ANFIS. For comparison, the standard international accepted benchmark simulation model no.1 (BSM1) was used. The promising results obtained are in conformity with the BSM1 results. This revealed that the proposed or suggested model is an efficient powerful tool to describe and predict an activated sludge process.