Multivariable Fuzzy Controller For The Activated Sludge Process

Abstract

Hybrid neuro-fuzzy collect the strengths of neural networks and fuzzy systems and minimize the drawbacks of the individual approaches. The most common hybrid neuro-fuzzy system uses a neural network to tune the parameters of a fuzzy system. This paper presents the design of a multivariable fuzzy controller for the wastewater treatment process. For simulation we have used Matlab 6.5\Simulink and Fuzzy Logic Toolbox. At first, we have manually changed the type, shape or place of membership functions, we have abed membership functions or another rules or we have changed their importance. Finally, we have Automatically tuned these parameters by using neuro-adaptive learning techniques incorporated into anfis function in the Fuzzy Logic Toolbox , for creating , training and testing a Takagi-Sugeno fuzzy controller.