A logic-based benders decomposition method for refrigerated container stacking and relocation problem
研究了冷藏集装箱在混合存储模式下的堆存与翻箱问题，提出基于逻辑Benders分解的精确算法，在大型实例中求解效率提升36.7%和52.3%，为港口设施升级时机提供决策支持。
Efficient management of refrigerated container (RC) storage plays a critical role in maintaining the quality of cold-chain cargoes and enhancing the efficiency of cold chain logistics. This study addresses the Refrigerated Container Stacking and Relocation Problem with Duplicate Priorities and Discrete Stacking Mode (RCSRP-DPDSM), which arises in hybrid storage modes at container terminals. A mixed-integer programming model is formulated to optimise the storage area assignment integrated with stacking and relocation scheduling, considering the unique characteristics of the duplicate priorities and discrete stacking mode. The integrated model is decomposed, and an exact algorithm based on logic-based Benders decomposition (LBBD) is designed to solve it. Several master problem strengthening strategies are proposed to obtain tighter lower bounds. To further accelerate convergence, unique features of RCSRP-DPDSM are explored and an enhanced minimal suboptimal set searching heuristic is tailored. Compared with the existing Benders cut strengthening strategy and the developed heuristic method, the tailored LBBD-based approach not only obtains the optimal solution but also improves solution efficiency by 36.7% and 52.3%, respectively, under large-scale instances. Further sensitivity analyses on handling capacity configurations are conducted to provide practical insights for port managers regarding the appropriate timing and approach to facility upgrades.