Study On Hybrid Intelligent Control For Variable Rate Anoxic Biological Filter Wastewater Treatment

Intelligent control is recently developed from conventional control theory. It consists ofseveral control theories, such as fuzzy control (FC), artificial neural network control (ANNC),expert control, etc. The intelligent control of wastewater treatment system is a focus inwastewater treatment research field. Base on the all-around review and analysis of theprogress of wastewater treatment study, a hybrid-intelligent control method is presented in linewith fuzzy theory in which FC, PID and ANN are all taken into consideration. Theintelligent-intelligent control is used in wastewater treatment of variable rate bio-filter techniqueand several valuable conclusions are reached. The intelligent-intelligent control of Variable Rate Anoxic Biological Filter(VRABF)Wastewater Treatment is studied especially for Chongqing Science and Technology KeyProject—The kernel techniques and Engineering Example Project of Small and medium cityWastewater Treatment. The intelligent-intelligent control system is developed and will beputted into operation months later. Some experiments are also conducted. On all accounts, thedissertation details the innovate contribution on following aspects. Based on the features analysis of VRABF wastewater treatment, the general framework ofthe hybrid-intelligent control is constructed, the design method and control structure are givenwith the consideration of Fuzzy-PID IGA-Fuzzy(improved genetic algorithm) andERFNNC(Error Reinforcements based Fuzzy-neural Network Controller). A short-term forecasting method is proposed based on the distinguishing of fuzzy sampledistribution, which reduce the forecasting error and the stand division. The method is useful tosolve the influence of influent quantity and enhance the stability of control system. The modelconstruction techniques of ANN and fuzzy mathematics are combined to describe thewastewater biological treatment process, then a soft-computation model is built for waterquality prediction and the influent water indexes are evaluated. For this reason, the lagproblem of measuring system is solved satisfactorily. Obviously, it is very difficult to construct the fuzzy control rules for wastewaterbio-filter treatment control system, and the parameters of fuzzy controller is hard to adjust aswell for VRABF. The dissertation presents a fast systemic fuzzy self-adaptive method for fuzzycontroller design which combines the fuzzy theory and PID control theory organically. In thisway, the parameters can be revised automatically with the variation of wastewater treatmentoperating conditions. Improved genetic algorithm is employed to optimize the fuzzy control III重庆大学博士学位论文rules and coefficients and helps to avoid the aimlessness and difficulty by the way ofexperiential decision. A novel fuzzy-neural network controller is designed which can construct the control rulesautomatically making use of the nonlinear self-training capability of ANN. It can delete theinefficient rules, and settle the control rules explosion problem effectively. The robust fuzzy control idea is invented for uncertain parameters activated sludgesystem. Considering the characteristics of uncertain parameters activated sludge system, andthe effluent water quality should be treated according to its following-up techniques. Takingaccount of these factors, the control model is set up based on the improved Riccati foractivated sludge system. The methods presented above are used in the control system design of wastewatertreatment by VRABF technique. The COD removed ratio is adopted to control the intensityand time of reversing flush. The hybrid-intelligent control method is simulated on Matlab12 anddemonstrate its effectiveness and practical utility in the control of wastewater treatment.