Abstract
This article discusses the development and application of a digital twin (DT) for intelligent engineering of a specialized meatbased product. The focus is on the use of advanced intelligent optimization methods to create a virtual model capable of adaptively real time managing the formula and process parameters, in accordance with the product’s final purpose. To solve a food engineering problem (optimization of product formula and production modes), three metaheuristic intelligent algorithms were applied and compared: genetic algorithm (GA), particle swarm optimization (PSO), and sparrow search algorithm (SSA). The results of the comparative analysis demonstrated that SSA algorithm provided the best accuracy and convergence in solving the assigned optimi zation problems, making it the preferred tool for integration into the digital twin system. SSA algorithm demonstrates the ability to escape local optima, thereby avoiding the problem of premature convergence typical for some metaheuristic methods, such as GA and PSO algorithms. Based on an optimized digital twin, this approach enables dynamic prediction of physicochemical pa rameters, their compliance with the established medical and biological requirements for the final product, promptly adjusting its formula to meet the intended purpose of providing complete nutrition and enhancing the rehabilitation of patients with traumatic brain injuries (TBI), and ensures flexibility when scaling the technology or transferring production to another facility. Digital twins, integrating data from smart sensors and biosensors based on big data analysis, are the foundation for creating a safe, trace able, and adaptive food system. The study demonstrates that intelligent engineering based on a digital twin is a key technology for creating personalized, safe, and effective specialized food products within the framework of modern manufacturing principles and One Health concept.

This publication has 33 references indexed in Scilit: