Simulated Moving Bed Numerical Solution And Optimal Control

Simulated Moving Bed chromatography separation process is a continuous and discretehybrid system. It is composed of chromatographic column or similar chromatographic columnfixed bed by the valve switching implementation components separation. Thus a lot of theSMB system parameters need to control so that it is easy to vulnerable to outside interference.In actual production process, tend to be on the safe side, system operation will deviate fromthe optimal operation conditions, so it is difficult to fully exert the SMB chromatographicseparation process of the separation ability. In order to make the SMB separation system playits potential, based on the results of the simulation model is particularly important. So in thispaper the following three aspects:First of all, Analyses the SMB mechanism model. The general rate model whichconsidered more comprehensive by using the finite element orthogonal collocation methodfor numerical. And analysis the influence of operation parameters on the purity of separatedcomponents, understanding some of the SMB system dynamic characteristics. For thesimplified model (the balance diffusion model) using the method of line and Crank-Nicolsonmethod to numerical solution. And analysis the system parameters effect on the purity ofcomponents. Compares the two methods and found that Crank-Nicolson method under thecondition of the guarantee accuracy, the solving efficiency has been greatly improved. It isprovided the basis for the research on the control and optimization of the model.In the control of SMB purity, Adopted based on the axial concentration adsorption ordesorption front position is studied and its essential has been analyzed. Because the numericalsolving method of spatial of the discrete points only100, the number is too few if only use the100discrete points to look for the point of the fronts. It is will result big error. So using aGaussian fit to search the inflection point position of the front. On the basis of the analysis ofthe relationship between the front point position and flow, proposed to the four front using PIcontrollers the purity of components. Through the simulation analysis, the scheme is feasible.Control of the model can solve the anti-interference ability of system whether the systemis running in the optimal operation point, still do not know, the online optimization wasachieved by PLS establish data model. dEWMA method is used to update the SMB model.Through the simulation analysis, the optimization process is slow and the volatility is big. Inthe actual product process is not conducive to production; need to improve in the futureresearch.