The Research On Self-organized Sliding Control Methods For Municipal Wastewater Treatment

The process control of wastewater treatment processes(WWTPs)is an important mean that ensure the effluent quality reaching the standard and improve the treatment efficiency.Its basic idea is: By using the online data,the process control system regulates blowers,reflux pumps,dosing pumps,etc.,to achieve the desired operational results.However,since WWTPs includes physical,biological and chemical processes,the achievement of process control suffers from enormous challenges,which reflect on: 1)There are many coupling factors affecting the state of WWTPs.The mechanism of WWTPs is extremely complex,and is also lack of precise model.2)It is difficult to detect the variables of WWTPs in real-time,such as the key water quality parameters such as total phosphorus(TP)and ammonia nitrogen(NH4-N).3)The inflow of WWTPs can only be passively accepted,and the inflow flow,compositions,pollutant types and organic matter concentration fluctuate greatly.Therefore,it is nonlinear dynamic system with serious disturbance and always works in a non-stationary state.Then,how to accurately identify the real state of WWTPs;how to use effective control methods to achieve stable and accurate control.Those are still hot and issues in the current study of WWTPs.In view of the challenges faced by the above-mentioned process control of WWTPs,this thesis aims at achieving the stable effluent quality to design a suitable control framework based on the process technologies and principles of WWTPs;analyse the biochemical reaction mechanism of the process;extract the water qualities which are difficult to be detected online;establishe an intelligent detection model.Then,typical water qualities,such as effluent total nitrogen(TP)and ammonia nitrogen(NH4-N),are measured online.Afterwards,typical control processes and their operating conditions are listed and studied,and related factors affecting process control performance are summarized.A self-organizing sliding mode controller with strong adaptive ability and robustness is designed to realize stable and high-precision tracking control for typical process variables such as dissolved oxygen and internal reflux of WWTPs.On the basis of the system architecture,and the functions of detection and control,a self-organizing sliding mode control system for WWTPs is constructed and tested in the real wastewater treatment test platform.The main research work and innovations are as follows:(1)Design of Intelligent Detection Model(IDM)for Key Water QualitiesTo solve the problem that the key water quality parameters such as effluent TP and NH4-N are difficult to accurately detect online,an IDM is designed based on self-organizing fuzzy neural network(SOFNN).Firstly,the relevant variables of key water quality parameters are extracted by least squares method to determine the input and output variables of IDM.Secondly,the model structure is optimized based on the structural risk evaluation index.Meanwhile,the structure and parameters of the model are updated to improve the accuracy of IDM.Finally,the convergence conditions of IDM are analyzed and the structural optimization algorithm is presented.The criterion of parameters selection under convergence condition are given to ensure the successful application.The practical results show that the proposed IDM has better real-time performance and accuracy than other methods.(2)Design of Robust Sliding Mode Controller for WWTPsTo deal with the serious external and internal disturbances in WWTPs,a robust sliding mode control method(RSMC)is proposed in this thesis.Firstly,an external robust controller(ERC)is designed to accelerate the response of the sliding mode control(SMC)based on the adaptive reaching law,which achieves the fast convergence and the suppression of external disturbances.Secondly,an internal robust controller(IRC)is designed to mimic the dynamic characteristics of internal disturbances based on FNN,which forms an IRC control law and ensures the robustness of the global system despite the internal disturbances.By constructing Lyapunov functions,the performance analysis of RSMC is completed.The application results of typical non-linear systems and nitrate nitrogen control in municipal wastewater treatment process show that RSMC can not only maintain stable control performance under internal and external disturbances,but also has the advantage of fast convergence.(3)Design of single-variable self-organizing sliding mode controller(SOSMC)for WWTPsTo cope with serious nonlinearity and uncertainty in WWTPs,a single-variable SOSMC is proposed in this thesis,which realizes the precise control of dissolved oxygen concentration.Firstly,using the tracking error and structural risk of SOSMC,the performance evaluation index of the controller is designed.Then,the structure of SOSMC is optimized according to the evaluation results to alleviate the chattering phenomenon of SMC.Secondly,the parameters of the controller are updated by the adaptive projection algorithm to improve the response speed of the control.Finally,the stability and convergence of the controller are discussed with the fixed and dynamic structures of SOSMC.Meanwhile,the parameter initialization method is given to ensure the successful application of SOSMC.The application results of nonlinear systems and dissolved oxygen control in WWTPs show that the single-variable SOSMC can not only maintain stable and accurate tracking control performance,but also have fast control response to track the dynamic set-points.(4)Design of multi-variable SOSMC for WWTPsTo solve the problem that the whole process of WWTPs is difficult to operate smoothly,a multi-variable SOSMC is proposed in this thesis.Firstly,a suitable multi-variable SOSMC architecture for the whole process of WWTPs is designed,which includes multi-input multi-output sliding mode control,multi-input multi-output self-organizing fuzzy neural network compensation control and adaptive strategy.Secondly,based on the control law and system output,an adaptive control gain optimization strategy(ACGOC)is proposed.By adjusting the controller gain and smoothing the output of the controller,the problem of the control performance degradation caused by the large fluctuation of the control signal is solved.Finally,in order to ensure the stability of the controller,the design principle and execution steps of the parameters of the multi-variable SOSMC are given to ensure the practicability of the controller.The application of multi-variable SOSMC in WWTPs shows that the controller not only control the aeration and internal reflux stably and accurately,but also maintain the output of control system continuously and smoothly.(5)SOSMC System for WWTPsIn order to realize the verification and application of SOSMC in WWTPs,a SOSMC system is developed in this thesis.Firstly,with the aspect of control system architecture,encapsulated software and hardware can be embedded,including data acquisition module,key water quality parameter detection module and SOSMC module,to ensure compatibility with other control systems and realize the function of SOSMC;Secondly,with the aspect of control system operation,the interface of implementation system is used to monitor the operation status.The design of the signal read-write port,the initiaition of the control module and the execution of the signal input are used to realize the automatic control of arbitrary control variables in the typical control process.Finally,the effectiveness of the system was tested by using the aeration control and the total phosphorus control in the pilot-scale platform of WWTPs.The results show that the system is easily operated,and achieves the expected control targets according to the real requirements.