The Development of a Neuro-Fuzzy Expert System for Wastewater pH Control

Abstract

The paper presents the development of an experimental system for the control of wastewater pH (SENFpHCTRL) in Petrobrazi Refinery, namely an expert system based on an artificial intelligence technique, the so called adaptive neuro-fuzzy inference systems. In order to test the software component of the system, i.e. the developed neuro-fuzzy controller, in MatLab/Simulink there was built an automatic system for pH control named ANFISpHControl using the flows and volumes from Petrobrazi and a mathematical model from literature for the process of wastewater pH neutralization. The chosen model was also validated for Petrobrazi Refinery through a set of simulations and experiments presented in this paper.