Varying Parameters of the Object Model on Fuzzy Neural Network Controller for Activated Sludge Method of the Sewage Treatment System

Abstract

In this paper, we take characteristics of wastewater treatment and process technology, drawing on the effectiveness of thetraditional PID control and on the basis of its lack, with the key steps in the sewage treatment process - Aeration control of part of the process parameters, Fuzzy neural network control of dissolved oxygen concentration (DO) to achieve negative feedback control loop,design a model-based closed-loop cascade control system. Fuzzy systems, membership function, the structure of the network topology and algorithms are based on the actual issues identified in the fuzzy variables. Aiming at the four parts of the fuzzy control, adopting four fuzzy neural network based on the standard model - the input layer, Fuzzy layer,Inference layer,Clear layer are corresponding with it. Standing on two points: the dissolved oxygen concentration control and the rate of change from the error ,then design the Fuzzy neural network controller. Then the fuzzy neural network control technology could be used in wastewater treatment on the specific application of process control.