Research On Optimal Control Of Anaerobic Digestion Process Based On Extreme Value Search

In recent years,China's economy has developed rapidly,but China's energy composition is still dominated by fossil energy such as coal and petroleum,and a large amount of organic waste has not been used reasonably.It has brought about increasingly prominent problems of energy shortage and environmental pollution,which have seriously affected China's sustainable development.The anaerobic digestion technology,as a green and efficient organic waste treatment technology,uses anaerobic microorganisms to degrade organic waste and generate green energy-hydrogen and methane,which can save thousands of tons of oil and effectively improve energy shortages and environmental pollution at this stage.Aiming at the existing anaerobic digestion process with low gas production rate and instability,this paper focuses on the research of anaerobic digestion process with the production of hydrogen and methane,and uses the extremum seeking algorithm based on the fusion of data and prior knowledge to study the optimization control.The specific work is as follows:1)Firstly,in view of the fact that the existing anaerobic digestion model is too simple and only considers the generation of methane,the anaerobic digestion process that can produce hydrogen and methane at the same time is mathematically modeled.Based on the study of the anaerobic digestion process and the related mechanism of the anaerobic digestion process with the production of hydrogen of insoluble substrate,the mathematical models of hydrogen production bioreactor BR1 and methanogenesis bioreactor BR2 are established using the conservation of material quality theory.In proposed model,the conversion processes of intermediate products such as acetic acid,propionic acid,and butyric acid were added.In order to achieve better model performance,the values of the model correlation coefficients are studied and analyzed.In addition,Matlab/Sumlink's open-loop simulation experiments and static characteristics verify the availability of the mathematical model and the existence of extreme points of gas production.2)Secondly,considering that the anaerobic digestion process has the characteristics of complex mechanism,high degree of non-linearity and difficult to obtain accurate mathematical models,this paper proposes an extremum seeking control with a simple structure and without model information.First,based on the study of the Kalman filter algorithm,a gradient estimator that can obtain accurate gradient estimates is designed,and then it is introduced into the perturbation extremum seeking control to speed up the convergence speed.At the same time,the perturbation signal uses a more general stochastic signal,and the principle and convergence of the control algorithm are given.Finally,the simulation comparison with the perturbation extremum seeking control validates the superiority of KF based extremum seeking control3)Thirdly,in order to eliminate the steady-state oscillation in the extremum seeking control and further improve the control performance,a Netwon extremum seeking control without steady-state oscillation is proposed.In this control algorithm,a stochastic perturbation signal whose amplitude is directly proportional to the extreme value estimation error and a KF estimator that simultaneously estimates the gradient and Hessian information are designed.Next,the principle of the control algorithm is introduced and the convergence of the perturbation signal amplitude is analyzed.In addition,simulation experiments verify that the proposed control algorithm can not only accelerate the convergence speed,but also eliminate the unfavorable factor of steady state oscillation.4)Finally,in order to improve the robustness of the control algorithm,a self-optimizing extremum seeking control is proposed to solve the nonlinear system optimization problem of disturbances affecting the changes of extreme points.The control algorithm adopts a layered structure of the optimization layer and the control layer.The extremum seeking controller is located in the optimization layer,which provides accurate setting values for the self-optimizing controller in the lower control layer,and the self-optimizing controller quickly processes the disturbances according to the new setting values.This control algorithm can combine the model-free characteristics of the extremum seeking controller with the anti-disturbance characteristics of the self-optimizing controller,and can quickly and effectively handle the expected and unmodeled disturbances.The comparative simulations verify the effectiveness of the self-optimized extremum seeking control.