Multi-period supply chain optimization with contango and backwardation effects using an improved hybrid genetic algorithm

In real-world markets, supply chain costs often fluctuate over time due to the contango and backwardation effects, making multi-period supply chain planning complex and critical. This paper presents a multi-period supply chain optimization model that explicitly incorporates these effects into cost forecasting and decision-making. A multi-period supply chain model is developed, considering the cost uncertainty introduced by contango and backwardation. An integrated polynomial regression fuzzy method is proposed to address this problem by predicting future fluctuations in purchasing, ordering, and logistics costs. A mixed-integer linear programming (MILP) model is formulated to minimize the total supply chain cost across multiple periods. Moreover, improving the hybrid genetic algorithm (IHGA) is proposed to solve this problem. The performance of the proposed IHGA is triggered by integrating trust region, quasi-Newton, and pattern search methods. Response Surface Methodology (RSM) determines the optimal parameter settings and hybridization structure. A real-world case study involving surgical instrument manufacturing companies validates the proposed approach. The results highlight optimal supplier selection and order allocations for each period, and performance comparisons reveal that the IHGA outperforms traditional algorithms in terms of cost efficiency, computational time, and convergence behavior.