Research On Control And Optimization Method For Separation Of Simulated Moving Bed

Simulated moving bed is a device that can realize continuous separation of chemical and other products.Because of the complexity of the model,it is difficult to simulate the simulation optimization and control based on the conventional algorithm,so the research on the control and optimization of the simulated moving bed chromatography using intelligent algorithm has become one of the hot spots.The purpose of this paper is to solve the problem of Simulated moving bed control and optimization with a variety of intelligent algorithms,so that Simulated moving bed device can be more effectively applied to the continuous separation of chemical,pharmaceutical and other industries.Firstly,the development history and related theory of Simulated moving bed separation technology are introduced,and the model of the column is analyzed.the different models are simulated by MATLAB software.then,the model is solved by using the on-line solution method for Simulated moving bed rate model,and the concentration change of the two components under different bed layers is obtained by simulation analysis.With the separation of Simulated moving bed,the program is first designed,and then based on the triangle theory,the linear adsorption isotherms,standard Langmuir adsorption isotherms,anti-Langmuir adsorption isotherms,m1-Langmuir adsorption isotherms,m2-Langmuir adsorption isotherms are developed,and the genetic algorithm is used to determine the parameters of the different adsorption isotherms,so that they could be applied to the operation separation process.Finally,the separation of toluene and xylene mixture is carried out by the operation separation procedure combined with KNAUER company's Simulated moving bed device,and the separation results are satisfactory by using shimadzu high performance liquid chromatography.For Simulated moving bed control,for the simulation mobile bed rate model under the nonlinear adsorption isotherm,the flow ratio of the triangle theory,m2?m3 as the input,the purity of the extraction solution and the purity of the raffinate as the output,the model is identified by the prediction error method and the subspace identification method,and then the model is controlled by the predictive control algorithm,and the control effect is analyzed by simulation.For Simulated moving bed purification process,only the purity of the extract is considered,the model identification method was used to identify the two-input and one-output transfer function model,and then PID algorithm,fuzzy PID algorithm,state feedback control algorithm,model predictive control algorithm are used to design the controller respectively,and the simulation results are obtained by MATLAB simulation results.the model predictive control algorithm is compared with other algorithms,and it is further verified that the model predictive control algorithm can control the simulated moving bed separation process and purification process.For Simulated moving bed optimization,the research is carried out from two aspects of single objective optimization and multi-objective optimization.For single objective optimization,on the one hand,based on Simulated moving bed mechanism model,genetic algorithm(GA)and particle swarm algorithm(PSO)are used to optimize Simulated moving bed sheet targets,and the genetic algorithm and particle swarm optimization algorithm to join quantum behavior are compared.The simulation results show that the genetic algorithm and particle swarm algorithm join quantum behavior faster than the basic genetic algorithm and particle swarm algorithm.On the other hand,using BP and RBF neural network algorithm to fit the Simulated moving bed data to get the neural network model,using Artificial Fish Swarm Algorithm,bacterial foraging algorithm,quantum genetic algorithm to optimize the parameters of the neural network model,the simulation results show that the RBF neural network combined with Artificial Fish Swarm Algorithm algorithm can complete the simulation mobile bed target optimization task with the small amount of data.For multi-objective optimization,the NSGAII algorithm and MOPSO algorithm are used to optimize the performance index(purity of raffinate solution purity,extraction purity,total productivity,recovery maximum,desorption agent total consumption).the efficiency of the algorithm is studied,and the concentration is set as the constraint condition by the actual industrial demand,with the minimum energy consumption(recovery maximum,the desorption agent flow minimum,the maximum productivity)as the optimization goal,MOPSO algorithm is used to optimize the optimization,and the Pareto optimal solution set is obtained,which provides guidance for industrial production.Finally,the research of the control and optimization of Simulated moving bed chromatography separation technology is summarized,and the limitations of the research and the future research direction are further analyzed.