Optimization of the Simulated Moving Bed Start-Up Process Based on Continuous Time Scale Adaptation

Abstract Simulated moving-bed (SMB) chromatography is a highly efficient novel separation technique. However, the establishment period from start-up in an empty column to cyclic steady state is extremely long. Traditional fixed-parameter start-up modes are not only time-consuming but also result in significant eluent consumption and severe waste of transition products. Current start-up optimization studies often employ staged parameter optimization and rely on preset ideal reference trajectories, neglecting separation performance under steady-state conditions. Therefore, a continuous-time-scale adaptive startup optimization control strategy is proposed. First, the optimal steady-state operating parameters under the current equipment conditions are determined using a computational model. Then, a continuous-time scaling factor and a Hermitian interpolation method are introduced to achieve parameter smoothing and scaling of the optimization time. Finally, a composite objective function is constructed and solved by utilizing a two-layer collaborative optimization architecture. Comparative experiments demonstrate that this strategy significantly reduces start-up time and eluent consumption: it requires only 9.56 h to reach steady state, achieving a 57.3% reduction compared to the 22.42 h required by the traditional fixed-parameter strategy. Furthermore, eluent consumption during the start-up period is drastically decreased from 56 mL in the traditional strategy to just 16 mL. Simultaneously, under the strict premise that the product purity is greater than 99%, the proposed strategy increases the final steady-state productivity by 12.8% compared to the traditional approach. This novel strategy successfully achieves the dual objectives of shortening the start-up time and securing superior separation performance, thereby greatly enhancing overall economic efficiency.