Abstract：This study aimed to achieve high-density culture of Staphylococcus carnosus for the purpose of preparing highly active starter cultures. Tryptic soy broth was used as the basal culture medium. Optimization of medium components was carried out using one-factor-at-a-time method and Box-Behnken design combined with response surface methodology (RSM). Meanwhile, an artificial neural network-genetic algorithm (ANN-GA) model was developed. The results indicated that the nitrogen source was the most significant factor influencing the viable count of S. carnosus. Compared with the RSM model, the ANN-GA model provided more accurate predictions with smaller prediction errors and superior optimization results. The optimal medium determined by the ANN-GA method was composed of 3.21 g/L glucose, 20.17 g/L soy protein peptone, 20.17 g/L beef extract powder, 5.63 g/L dipotassium phosphate, 5.0 g/L sodium chloride, and 0.2 g/L magnesium sulfate heptahydrate. In a 5 L fermentor, the maximum viable bacterial count reached 1.67 × 1010 CFU/mL.
2025, Vol. 39 
