Font Size: a A A

Hybrid Modeling And Control Of Urban Wastewater Treatment Process

Posted on:2023-02-28Degree:MasterType:Thesis
Country:ChinaCandidate:Z X ChenFull Text:PDF
GTID:2532306812475334Subject:Control theory and control engineering
Abstract/Summary:PDF Full Text Request
The wastewater treatment process has the characteristics of strong nonlinearity and time delay.Therefore,the real-time monitoring and quality control of the wastewater treatment process has certain theoretical and practical significance,but also has certain difficulties.In the actual wastewater treatment process,the nitrate nitrogen concentration in the second biochemical reaction tank and the dissolved oxygen concentration in the fifth biochemical reaction tank are the key control indicators,and the quality of their control affects directly the quality of the effluent.For wastewater treatment processes,it is important to establish an accurate and reliable prediction model of nitrate nitrogen and dissolved oxygen.In this thesis,the prediction model of nitrate nitrogen and dissolved oxygen concentration in wastewater treatment process is established by parallel hybrid modeling,and a predictive control algorithm is proposed based on the hybrid prediction model.The details of the thesis research are as follows:(1)Firstly,the technology and reaction mechanism of the activated sludge process in the wastewater treatment process were analyzed deeply.According to the reaction kinetics,a BSM1(Benchmark Simulation Model 1)model was established to simulate the actual process.Secondly,the mechanism model of nitrate nitrogen concentration in the second biochemical reaction tank and dissolved oxygen concentration in the fifth biochemical reaction tank is established,and the relationship between input and output is simulated and analyzed to lay the foundation for the subsequent hybrid model.(2)A parallel hybrid modeling method is proposed to establish the prediction model of nitrate nitrogen concentration in the second biochemical reaction tank and dissolved oxygen concentration in the fifth biochemical reaction tank.Firstly,the mechanism model of wastewater treatment process is established,because there is a certain error between the actual value and the output value of the mechanism model.Secondly,the BP(Back Propagation)error compensation model is established.Finally,the mechanism model and BP error compensation model are combined in parallel to form a hybrid prediction model.(3)Taking the concentration of nitrate nitrogen in the second biochemical reaction tank and the concentration of dissolved oxygen in the fifth biochemical reaction tank as the controlled quantities,the closed-loop simulation of the hybrid model is carried out through the PI(Proportion Integral)controller.Through the closed-loop simulation results under different weather conditions,it is found that the PI controller has a good control effect on the concentration of dissolved oxygen in the fifth biochemical reaction tank,but the control accuracy of nitrate nitrogen concentration in the second biochemical reaction tank is not high.(4)After in-depth study and analysis of the principle and characteristics of predictive control,this thesis proposes a predictive control algorithm based on hybrid model to predict and control the concentration of nitrate nitrogen in the second biochemical reaction tank and dissolved oxygen in the fifth biochemical reaction tank.The Harris Hawks optimization controller optimizes the current optimal oxygen transfer coefficient and internal reflux flow,which is compared with the simulation results of PI control,It is found that the predictive control algorithm based on hybrid model has a good tracking control effect on the concentration of nitrate nitrogen and dissolved oxygen.
Keywords/Search Tags:Wastewater treatment, Mechanism models, BP neural network models, Hybrid models, Predictive control
PDF Full Text Request
Related items