A systematic approach to SMB processes model identification from batch experiments

Prediction‐correction method for optimization of simulated moving bed chromatography

“…where c j i;out is the concentration of component i in the outlet connection of column j. We also use the simple relation c j i;out ðtÞ ¼ c j i ðL c ; tÞ (7) where c j i (L c , t) is the concentration of component i in column j at position z ¼ L c , and L c is the column length.…”

“…The discretization is done in gPROMS using a second-order central finite difference method with 100 finite elements in the axial domain of each column. The resulting system of differential algebraic equations (DAEs) consisting of (1)- (7) are solved by the method of lines 15 with initial conditions specified for the concentration of each component at each axial position in both liquid and adsorbed phases. For a simulation starting from clean beds the initial conditions are c j i (z, 0) ¼ 0 and q j i (z, 0) ¼ 0 for all i ¼ A, B and all j ¼ 1,2,…,ncol and all z ¼ (0, L c ].…”

“…2,3 Other researchers have studied ways to use batch experiments to estimate single-column model parameters to obtain a predictive process model for SMB. In Grosfils et al 7 there is a comparison between different single-column models and a study of the identifiability of model parameters using pulse injections and parameter estimation. They used design of experiments to reduce the effort required to obtain reliable process model parameters for SMB.…”

Combination of multi-model predictive control and the wave theory for the control of simulated moving bed plants

“…Several alternatives have been proposed for modeling mass transportation in the fluid phase and mass transfer between the liquid and solid phases in SMB plants. It is not the aim of this paper to review these several alternatives, the reader interested in these issues is referred to the literature (Guiochon et al, 1994;Grosfils et al, 2007b). Rather, a linear driving force (LDF) model is selected as a simulation model to test our nonlinear model predictive control (NMPC) approach.…”

“…Rather, a linear driving force (LDF) model is selected as a simulation model to test our nonlinear model predictive control (NMPC) approach. More details on the process can be found in Grosfils et al (2007b) and Ruthven and Ching (1989) and references therein.…”