Data Driven Advanced Control Method Of Simulated Moving Bed Chromatography Separation Process

Simulated moving bed(SMB)chromatographic separation technology is a new separation technology,which is developed on the basis of traditional fixed bed adsorption operation and real moving bed(TMB)chromatographic separation technology.It is widely used in petrochemical,biopharmaceutical and food chemical industries,because SMB can efficiently separate the mixture.SMB chromatographic separation process has nonlinear,strong coupling and non-equilibrium characteristics.Thus,it has important theoretical significance and practical value for the study on the control methods on SMB chromatography separation process.In this thesis,the data driven advanced control method of SMB chromatography separation process was studied in details.The specific research contents are described as follows.Firstly,the input-output data matrix can be constructed by utilizing the SMB chromatographic separation process data.Multi-variable output-error state space(MOESP)and numerical algorithms for subspace state space system identification(N4SID)were used to model the SMB chromatographic separation process,and the3-order and 4-order state space yield models of the SMB chromatographic separation systems based two subspace system identification algorithms were obtained.In order to verify the matching degree of the models,the output values of the 3-order and4-order yield models were compared with the real values.Finally,the reliability of the subspace yield models were verified.Secondly,based on the yield models of SMB chromatographic separation process,the impact of different control indicators on the system response performance for the SMB yield model under predictive control method are discussed.The output response curves of the yield model were obtained by changing each parameter,and the yield model prediction index was optimized.Then,the predictive controller under the yield models of SMB chromatographic separation process ware designed,and a predictive control method based on SMB chromatographic separation process yield model was proposed.The simulation results show that the control and prediction of the yield models can be realized based on the given yield prediction range.Finally,by utilizing the piecewise affine system,modeling method and predictive control algorithm,the SMB yield model based on the piecewise affine system was established.By comparing the subspace yield model and the piecewise affine yield model,the prediction control effects under different conditions are compared.The simulation results show that the model identification effect of the piecewise affine system is better than that of the subspace system identification,and it can moreaccurately fit the given target output.