Formulation and Process Optimization of Longjiang Brine: ANN-GA Optimization Based on the Color and Flavor of Brined Pork.

Abstract

Longjiang brine's complex composition challenges traditional empirical optimization methods. This study developed an artificial neural network coupled with multi-objective genetic algorithm (ANN-MOGA) to simultaneously optimize flavor compounds, color parameters, and brining process. Validation experiments were conducted to assess the application from the perspective of color, flavor, and taste of brined pork. The results showed that the single-objective optimization group used 2.24% dried galangal, 1.38% food caramel, and 4.22 h of brining time. After MOGA optimization, the parameters were adjusted to 2.57% dried galangal, 1.10% food caramel, and 4.41 h of brining time. The optimized brined pork exhibited significant improvements, including color score increased by 5.32%, flavor by 9.69%, taste by 4.87%, and overall acceptability by 9.10%. The results of the validation experiments indicated that the ANN-MOGA can be used as an effective tool to optimize the formulation and brining process of Longjiang brine, making it a useful tool for application in various complex brine systems.