Research On Modeling, Optimization And Control Of Activated Sludge Process In Wastewater Biological Treatment

Activated sludge process based wastewater biological treatment is an effective method that makes the use of microorganisms to remove organic pollutants in wastewater, and it is the main way of current organic wastewater treatment, especially in urban wastewater treatment. The nonlinear process is a complex dynamics with the characters of strong outside interference, large time-varying and strong coupling. The laws of survival and the dependents on the conditions of the microorganisms in wastewater environment have not been fully aware. Because of the field test takes a long cycle and a higher cost, so the computer simulation based on some complex mechanism models is particularly important in the research of wastewater treatment.ASM series models of the ASP processes has made great strides in recent decades. Although the process was complete description in these models, but its too complicated to be directly applied in the actual wastewater treatment plant for process simulation; The stability of DO control is still not enough, which makes the water quality can not be guaranteed, but also caused a lot of aeration costs wasted for wastewater treatment enterprises. The WWTPs are non-profit, high energy consumption and supported by government finances. Large processing costs make them ends meet and survival worrying, and it also does not meet the national saving society development ideas. In this thesis, a study to address the problem in the ASP model, process control and control practice in WWTP is done and it has achieved the following contributions.In the researches of the process model and simulation software, according to the status for complex model can not be applied directly to the actual WWTPs, a simplified organic carbon remove model is by exploded and simplified from the ASM1. The model is simple and reliable to handle a specific process, which is for the purpose of carbon remove operation of the wastewater treatment. Also, a complete ASP process simulation package is developed by using Matlab/Simulink in this thesis, and it is applied to the simulation the ASP process to verification, and achieved good results. Reasonable classification of the running data from the simulation software packages through improved clustering method results to the establishment of the multi-model representation of the process. The simulation results show that the multi-modeling methods improved the accuracy of the process modeling, and it can be the model support in the subsequent control problem researches.In the researches of the DO concentration control and the other relational problems in the ASP process, some new strategies are proposed based on the MPC, optimized NN based GMC. The optimized NN based GMC control strategy is used to solve the control problem of DO concentration in the organic carbon remove process. The strategy combined with the characteristics that the neural network can infinite approximation to the nonlinear process, and use the NN to represent the unknown parts of the process will make more accurate simulation results. It is more efficient and reliable performances than others. And it also reduces the processing costs and saving the operating costs.In the researches of the ASP process optimization, a new strategy is proposed to optimize the ASP process combing with the two outstanding problems together, which are high effluent organic concentration and high processing costs of the ASP process. It takes the flowrate of wastage and the concentration of dissolved oxygen (DO) as control parameters and regards effluent quality as restriction factor and operation cost as performance index. And the clonal selection algorithm is first time used in the multivariable optimal control in ASP process, which performance good in the comparing simulation study.