Research On Adaptive Control Strategy Of Continuous Ethanol Fermentation Process Based On Coupled Membrane Separatio

Biofuels have great potential on ensuring global energy supply and reduce greenhouse gas emissions.In the continuous ethanol fermentation process,it is generally believed that the co-expression of yeast metabolic cycle under high ethanol content leads to the emergence of self-oscillations,which inhibits the normal operation of the fermentation process.In situ product removal(ISPR)technology as an effective method to solve end-product inhibition,has gained considerable attentions.However,the high operating cost limits the development of its real-time application.In this work,Adaptive control strategies are developed to reduce the operating frequence of ISPR while achieveing continuous production.For this purpose,research on perspectives of operation and control are conducted.The contents of this dissertation mainly include the following aspects.(1)The Andronov-Hopf bifurcation analysis of the structured fermentation reactor model was carried out,and the evolution to self-oscillations are detailed;then,ISPR is introduced to eliminate product inhibition and to stabilize the process;afterwards,dynamic behavior of external signal exerting on the system is analyzed.Because of the complexity of fermentation process caused by production inhibition,an ethanol selective membrane based on ISPR was constructed and the stabilizing effect was verified.To further modeling the process through data-driven methods,angle measurement is introduced on the just-in time learning(JITL)algorithm,and the distance measurement and angle measurement are used to evaluate the similarity between the data,and the parameter stability constraint is added to solve the stability problem of the local model.The purge steam is analyzed and the optimal working state is determined by analyzing the impact of different states on the ethanol fermentation process.(2)In view of the strong nonlinear and time-varying characteristics of the continuous ethanol fermentation process,adaptive model predictive control(AMPC)method is adopted to control the nonlinear process,in which the Kalman filter algorithm is used to estimate the parameters of the ARX model online,and the optimal control concentration of ethanol is obtained through analysis,and the control effects of adaptive-MPC and MPC are compared.Based on the ethanol fermentation system,a washout filter controller is built.Through the washout filter,the nonlinear dynamic system is stabilized.(3)Aiming at the problem that the fermentation process is slow and the purging process is fast,which might cause mismatch between fermentation and product separation,periodic external signal is introduced to analysis and realize periodic opeartion.The performance of membrane reactor with periodic external disturbance was evaluated by Laplace-Borel transformation,and it was found that the forced frequency of steam had no significant effect on the increase of ethanol production,while the forced amplitude showed a positive increase.Therefore,the membrane reactor system adopts the on/off operation,and periodic reference on ethanol concentration by adaptive MPC could well guide the purging steam to be inpulsive;On this basis,a decision algorithm is introduced into the adaptive MPC control strategy to constrain the optimal solution of the rolling optimization,and the tracking error is combined to operate it to make the purging steam in the control process more regular.This result is helpful to the periodic operation and control of high concentration ethanol fermentation.