Multi-objective optimization of industrial batch Balhimycin antibiotic and fed-batch Lysine biochemical processes

Abstract

Lysine is an essential amino acid, and its market value is USD 6.8 billion in 2023, and it is expected to double by 2033 worldwide. Balhimycin belongs to an important group of antibiotics with highly complex structures. With the utilization of advanced computational knowledge and new technologies, even a small improvement in productivity, yield, process time, or product quality can bring noticeable improvement in the production technology of biochemical processes. This study proposes a jumping gene adaptation of the elitist multi-objective differential evolution (Elitist MODE) algorithm. It is tested and applied successfully on the five benchmark test problems and two real-world problems, batch Balhimycin and fed-batch Lysine fermentation. Different constraint-handling techniques are used to test the performance of the proposed algorithm. The result obtained shows that the proposed algorithm can give a widespread Pareto front for both the test and industrial optimization problems. In batch Balhimycin fermentation, the optimal amount of glycerol required is not exceeding 59.84 g/L, and the optimal uptake of phosphate is also very low and doesn’t exceed 0.5 g/L, upon which product inhibition occurs. The uncertainty-specific study reveals that, with variation in upper bound of AMS concentration, product concentration varies significantly from 0.61 to 0.75 g/L.