Mathematical Modeling And Applications Of Dynamic Mechanical Models For Biological Nitrogen Removal Process

Domestic water pollution still grievously exists and corresponding controlenterprises keep high energy consumption and low efficiency, although wastewatertreatment related industry has zoomed and expanded for almost ten years. Currently,desirable target of energy saving and emission reduction can be realized only whenmore economic, efficient and stable technologies of water treatment are developedand implemented. Based on the experience of many developed counties, a modeling&simulation workbench’s construction is one of the core approaches to update thewastewater treatment trade. Such the workbench should couple fluid flow simulationwith dynamic process reaction modeling based on fundamental mechanismunderstanding of the activated sludge process. Furthermore, such the workbench needco-work with developing construction of the system of automation, intelligentizationand informationization for optimization of the wastewater treatment trade, andprofessional consultation and education mechanism also acts as a critical part.Considering the fact that current mathematical modeling of the biologicalnitrogen removal process has been far behind the corresponding theories andtechnologies’ development, this study mainly aims at building a systematical methodconstituted of model construction （directed by application goals, emphasizing theessential processes and cutting down the redundant ones）, calibration （importingcurrently existed modeling methods out of idiographic need） and application. Such themethod based on developing understanding of the activated sludge process, would actas an optional fundermental theory and technology for related process research andengineering optimization. Concretely, this study mainly organized the followingresearch aspects:（1） Ressearch on the Model Construction Method two suits of nitrification anddenitrification machnical reaction flows including a two-step form and a multiple-stepform are built based on distinct application goals and process machnisms. Afterdetailed analyzing the stoichiometric relation and corresponding kinetic equation ofeach sub-processes, relevant two-step and multiple-step nitrification anddenitrification models are constructed.（2） Ressearch on the Theoretic Modeling Method A novel―NumericalOptimal Approach Procedure‖（NOAP） for calibration of ASMs is designed whichimports proper numerical global parameter sensitivity and correlation analysisalgorithm to support parameter identifiability analysis and estimate selectedparameters by hybrid optimization algorithm to offer feedback criterion. The proposedNOAP is an integrated combination of experience-based approaches and systems analysis approaches, which can robustly find and tail the values of the best-fitparameter point and adapt to all the initial ordinary differential equations model asASMs.（3） Research on the Implementation Method of the Theoretic ModelingMethod Based on the MATLAB/SIMULINK workbench, biological nitrogen removalmodels constructed in this study are programmed in M-file and C-MEX S-functionform, and programmed models are solved by ode45and ode15s solver respectively.Genetic Algorithm is imported as the global optimization code to realizing parameterestimation. Numerical implementation procedure of the NOAP method is expanded,using data from a CSTR reactor which receive real sewage wastewater.（4） Research on the Application of the Models （Process Analysis andOptimization Suggestion） Using the NOAP method, the constructed two-step andmultiple-step nitrification and denitrification models are applied to analyze a SBRreactor which receive real sewage wastewater. From which, laws of the total nitrogenloss （including the N₂O emission） during aerobic period and N₂O accumulationduring anoxic period are discussed detailedly. An easy implementary way to reachhigher extent of Simultaneous Nitrification and Denitrification （SND） and loweraccumulation of N₂O is developed.（5） Research on the Application of the Models （Process Control and SoftSensing） Dissolved Oxygen （DO） setpoint control is implemented in simulation SBRcase using a technology of Nonlinear Model Predictive Control based on a simplifiedtwo-step nitrification model. An process stage prediction method is designed based ondynamic process aeration rate when the SBR system is under DO setpoint control.Simultaneously, a simple soft sensing method based on the DO simulation results isintroduced in constantly aerated SBR for substrates （readily biodegradable COD,ammonium, nitrite and nitrate） as well as Oxygen Uptake Rate.（6） Research on the Application of the Models （Simulator Development）worked on MATLAB/SIMULINK workbench, using its GUI developmentenvironment to design the graphical interfaces,―the Dynamic Simulator of BiologicalNitrogen Removal Process for Conventional SBR System‖is developed. Such thesimulator can model the processes of nitrification and denitrification separately andmodel calibration, aeration related intelligent control design and nitrification softsensing are also integrated in it.