关键词: 再制造, 物流网络, 模型, 混合遗传算法, 多周期优化设计

Abstract: To improve logistics management performance of manufacturing/remanufacturing systems, an appropriate design approach for closed-loop logistics network structrues was proposed. By considering design parameters'differences between inter-related time periods, capital's time value and logistics facility construction by stages, a Mixed Integer Non-Linear Programming (MINLP) model was developed for the optimal design of the logistics network considering expandable facility capacities. With the help of the model, the following items in the hybrid systems and for different periods could be determined, such as building or rebuilding locations and numbers for various sort of facilities, the extending standard cell numbers at rebuilding locations, the opening standard cell numbers for capacity enlarging at operation facilities and the logistics quantities between the corresponding facilities, also the minimum total fees of the logistics systems in planning horizon could be obtained simultaneously. To acquire design results rapidly and effectively, a hybrid genetic algorithm for MINLP model was presented, and its realization steps were introduced. Finally, effectiveness of MINLP model and its algorithm was verified by an example.

Key words: remanufacturing, logistics network, models, hybrid genetic algorithm, multi-period optimal design