Application Of Intelligent Control In Sewage Plant

This thesis mainly focuses on the research how to decide optimal quantity of coagulant dosage according to changes of water quality in sewage plant. Based on theories of neural network and adaptive neural fuzzy inference system (ANFIS), and also with the study of property of mixing process and control status quo, control systems for coagulant dosage process have been designed in this thesis.This thesis firstly gives a brief introduction on sewage plant process and evolution of coagulant dosage control. In Section 2, a detailed introduction of BP neural network, fussy logic and ANFIS including system structure, algorithm are presented. Through analyzing the characteristic of coagulant dosage process and mostly used control method, a control scheme for the process has been proposed which include a BP neural network feed forward controller and an ANFIS feed forward controller. Simulation based on laboratory data also has been done to show the different effects of the controllers and conventional feed-forward controller. Finally, a preliminary proposal for the engineering of this control system is provided.