Designing a perishable closed-loop poultry supply chain: metaheuristic approaches and model evaluation

In this study, a cutting-edge approach focusing on pivotal innovation to design a closed-loop poultry supply chain network is developed. As the poultry sector faces complex challenges regarding cost management and customer service enhancement, we propose a novel mixed-integer linear programming framework specifically designed for designing closed-loop poultry supply chains. The proposed framework introduces flexibility in the case of a closed-loop poultry supply chain, while most of the perishable products outside the poultry industry, like red meat, dairy products, seafood, and fresh produce, can be helped through visibility and optimization strategies that can help cross-industrial boundaries. In this regard, the objective is to minimize the supply chain cost. The model is widely evaluated using simulated data with twenty-four test problems of varying scales. Recognizing the NP-hard characteristics of the developed model, which make it impractical for large-scale problems, this limitation is addressed by incorporating four established meta-heuristic algorithms, namely Differential Evolution, Simulated Annealing, and Genetic Algorithm. Additionally, we explore two hybrid combinations of these algorithms, namely Hybrid Genetic Algorithm with Simulated Annealing and Hybrid Differential Evolution and Simulated Annealing Algorithm. The results of comprehensive test problems and sensitivity analysis bring out the remarkable efficacy of our proposed network design model, pointing out its prospects for optimizing closed-loop poultry supply chain network operations. Specifically, Hybrid Genetic with Simulated Annealing Algorithm in small cases, and Differential Evolution in medium and large-size cases emerge as the best-performing algorithms that perform superiorly in optimizing supply chain networks across different problem sizes.