Intelligent Control Strategy Based on Back-Propagation Neural Network with Adaptive Genetic Algorithm for Lincomycin Fermentation Process

Microbial fermentation feed production is challenged by a low degree of automation and poor feed volume accuracy. To address this, we propose an intelligent feed control strategy based on a back-propagation neural network (BPNN) with adaptive genetic algorithm (AGA). Based on the technological characteristics of microbial fermentation processes, BPNN is used to establish a dynamic mathematical model of the fermentation process, and AGA is introduced to optimize the parameters of the dynamic BPNN feeding control model. Cell concentration is selected as the dominant variable in the process of lincomycin fermentation. Finally, the established AGA-BPNN intelligent feed control model is embedded in the ARM11 controller, the optimal feeding quantity is calculated, and the feeding execution is remotely controlled through the wireless transmission module. The simulation results show that the cell concentration under the control of AGA-BP feeding is increased by 8.62% compared with manual feeding control, and 4.55% higher than that of traditional BPNN model. Therefore, the AGA-BPNN-based methodology has higher feeding control accuracy, which helps fermentation achieve automated production, improves fermentation efficiency, and reduces production costs.