Simulation, Evaluation And Optimization Of The Activated Sludge System Based On Computational Intelligence

The activated sludge process is being widely used for wastewater treatment.It is a very complicated and nonlinear process because of the variations of influent wastewater flow and compositions.Establishment of proper models is necessary for the design,control,operation and optimization of activated sludge systems.In this thesis,based on computational intelligence and combined with activated sludge model (ASM),simulation models were established for simulating activated sludge systems; evaluation models were developed for comparing the overall characteristics of aerobic granules and flocs;and optimization models were also constructed optimize the operating conditions of activated sludge system.Main contents and results are as follows:1.PHB-rich aerobic granules and AOB-rich nitrifying granules were cultivated successfully.Through controlling the COD and ammonium concentrations in the influent,the PHB-rich aerobic granular sludge with a good settling ability and a high PHB yield could be obtained.The development of AOB-rich nitrifying granules by inoculating aerobic granules could be accelerated.The cultivated granules were able to perform stable partial nitrification and accumulate nitrite in an SBR.2.A two-step nitrification model was established to determine the kinetic parameters of both AOB and NOB in nitrifying granules.In addition to nitrification reactions,the model also took into account biomass maintenance and mass transfer inside the granules.With batch esperiemental results,the kinetic parameters were obtained.The developed model performed well in simulating the oxygen uptake rate and nitrogen conversion kinetics of the AOB and NOB in the nitrifying granules and long-tcrm performance of the granule-based reactor.3.The fuzzy analytic hierarchy process and entropy weight approach were integrated to quantitatively evaluate the overall characteristics of aerobic granules and activated sludge flocs.The evaluation results with this integrated model show that the overall characteristics of aerobic granules were better than those of flocs.Due to significant influence of size on aerobic granules,the control of granule size was beneficial to the simultaneous nitrification,denireification and phosphate removal and essential for the long-term stability of granule-based reactors.4.The weighted nonlinear least-squares analysis and accelerating genetic algorithm were integrated to estimate the kinetic parameters of substrate consumption and storage product formation of activated sludge.The weighted least-squares analysis was employed to construct the objective function by integrating Monod equation,and the kinetic parameters were obtained by minimizing the objective function using accelerating genetic algorithm.The validity results suggest that this approach could evaluate the product formation kinetics of mixed cultures like activated sludge rapidly and accurately,and it could be applied for other biological wastewater treatment processes.5.An integrated dynamic model was developed through combining a mechanistic model,a neural network model and a genetic algorithm approach,in order to simulate the performance of a full-scale municipal wastewater treatment plant with substantial influent fluctuations.The simulation results matched the measured ones of the plant well,even under influent disturbance conditions.Compared with the mechanistic model and the neural network model,the integrated model was able to capture sufficient residual information to compensate for the inaccuracy of the mechanistic model and improve the extrapolative capability of the neural network model.6.An evaluation model was developed through projection pursuit method based on AGA to quantitatively evaluate the A~2O and reversed A~2O processes.The results show that the evaluation model could quantitatively compare the A~2O and reversed A~2O processes for the wastewater treatment plant.This evaluation model was demonstrated to be an effective and useful tool to quantitatively resolve the multi-factor evaluation problems.7.An integrated model through incorporating a modified version of ASM3 and the EAWAG Bio-P module,support vector machine and accelerating genetic algorithm was established to simulate and optimize the simultaneous nitrogen and phosphorus removal in an A~2O system in a wastewater treatment plant.With the integrated model,the optimal operating conditions were found after considering the effluent quality and cost saving.The integration of the mechanistic model,support vector machine model and accelerating genetic algorithm approach established in our work was found to be an effective and useful tool to optimize complex biological systems like the activated sludge process.