Research On Modeling And Optimization Of Simulated Moving Bed Separation Process

Simulated moving bed (SMB) technology developed by UOP in earliest 1960s has emerged as a powerful technique for continuous countercurrent adsorption separation. Originally, the SMB technology was used for the large-scale fractionation of sugars and xlynene isomers. In the last decade, the SMB applications have expanded greatly in the area of biotechnology, fine chemistry and pharmaceutical industry. How to improve the separation performance of SMB based on the mathematical model is the major problem in the research field of SMB.Particle Swarm optimization (PSO) is a recently invented high-performance optimizer that is very easy to understand and implement. It is similar in some ways to genetic algorithms or evolutionary algorithms, but requires less computational bookkeeping and generally only a few lines of codes. NSGA-â…¡ has several advantages over the current available multi-objective optimization algorithms and has been applied recently to optimize the SMB process. The main research work and contributions of this dissertation are as followings:(1) SMB's background, basic concept, merits and industrial application are introduced. A survey of SMB research work about the modeling and optimization of SMB is given.(2) Modeling strategies SMB are developed based on TMB model and SMB model. The prediction of these two models is compared in terms of steady-state performance. The influence of operation parameters in the SMB model prediction is also analyzed. The simulation relusts show thar the TMB stategy requires less time comsuption than the SMB strategy.(3) There has been a growing interest in solving multi-objective operation optimization of SMB problems using evolutionary approaches since the mid-1990. Several optimal operation strategies based on TMB model are proposed to optimize the separation performance of SMB. NSGA-â…¡ algorithms and MOCLPSO algorithm is respectively applied to solve the optimal operation strategies and a comparison study of results obtained by the tow evolutionary algorithms is presented. The results of multiobjective optimization show that MOCLPSO algorithm is suited to the single objective optimization of operation condition while NSGA-â…¡ is suied to the multiobjective optimizatio of operation condition, which suggests that there may be optimizationscenarios better suited to one mothed versus the other.(4) A SMB model using driving force is proposed for describing the Eluxyl SMB process. The axial dispersion and film mass transfer is calculated by the correlation respectively. Calculated results conform well to actual measurement data.Finally some conclusions and future researches are drawn in this dissertation.