Efficient Model Reduction of SMB Chromatography by Krylov-subspace Method with Application to Uncertainty Quantification

We address the model reduction of the high-dimensional model for the simulated moving bed process by a Krylov-subspace method. Full-update and partial-update schemes are proposed to derive the reduced-order models. The performance of each scheme for the calculation of the cyclic steady state solution is evaluated using a glucose-fructose separation example. The simulation and uncertainty quantification studies demonstrate that both schemes share the same advantage of high accuracy. The full-update scheme results in reduced models of a significantly lower order, while the partial-update scheme is computationally more efficient.