Abstract: The research insists on the combination of the Logistics Distribution Center（LDC） location with the long-term development of enterprises, and fully considers the costs of construction, operation and long-term distribution. The location model, which gets comprehensive evaluation values of all qualitative elements using rough set and group eigenvalue theory, and constructs the decision-making model combining qualitative elements and quantitative elements, and searches the best decision-making plan by Genetic Algorithms（GA）. The model fully considers all elements of LDC location, which include qualitative factors and quantitative factors and increases the year parameter [λ] of in-kind distribution, which enhances the practical value of it. The simulation analysis by example has been sufficiently validated the validity and practicability of the model.

Key words: multi-logistics distribution center, optimal location, rough set theory, group eigenvalue, genetic algorithms

摘要： 坚持物流配送中心选址与企业长期发展相结合，综合考虑选址方案带来的建设运营成本和长期配送费用，运用粗糙集方法和群决策特征根法对所有定性因素进行综合评判，并把定性要素综合评判值与定量要素相结合构建决策模型。利用遗传算法进行最优决策方案选择，实现了对企业物流配送中心选址过程中所有定性和定量影响因素的综合考虑，同时增加了配送费用考虑年限参数[λ]，使得模型具有了更强的实用性。通过仿真实验，模型的有效性和实用性得到了充分的验证。

关键词: 多物流配送中心, 优化选址, 粗糙集, 群决策, 遗传算法