Experimental application of individual column state and parameter estimation in SMB processes to an amino acid separation

摘要

The simulated moving bed (SMB) process is a highly efficient continuous chromatographic separation process. Due to its hybrid process dynamics that lead to discontinuities and sharp fronts on the state trajectories, optimal SMB process operation is challenging. Process performance can be improved by applying model-based optimizing control methods. For this, online information about states and individual column parameters are required. The strategy for simultaneous state and parameter estimation used here exploits the switching nature of the SMB process. The successful experimental application of the strategy is demonstrated for the continuous separation of two amino acids on an SMB pilot plant where extra-column equipment effects need to be considered.