Abstract
The isomerization of glucose to fructose is an important industrial process in obtaining high fructose syrup, a sweetener widely used in food industry. Traditional process in producing 55% high fructose syrup (HFS55) needs a large amount of desorbent. In this work, a hybrid simulated moving bed reactor (SMBR) system is optimized using experimentally verified dynamic SMB model to maximize the net productivity of HFS55 using minimum solvent. An adaptation of the state-of-the-art AI-based robust optimization technique, non-dominated sorting genetic algorithm with jumping genes is used in finding the Pareto (non-dominated) solutions for both the existing as well as SMBR system at the design stage. Finally, SMBR configuration was modified to further improve the system performance. Systematic multiobjective optimization resulted in significant performance improvement. Moreover, the new optimization technique gives much faster, smoother and larger spread of the Pareto-optimal solutions.
