Abstract: In this paper, the uncertainty of logistics demand network is considered. The interval analysis idea is applied to measure uncertain variables and parameters with interval numbers. The mixed integer programming model for logistics network design under interval demand mode is built. The risk coefficient and the maximum constraint deviation are defined to transform the objective function and constraints into certainty. A interval hierarchical optimization genetic algorithm is designed to solve the problem, and to calculate interval optimal solution and node decision-making scheme for objective function under different scenario states. It is shown by a tested example that the operability of the algorithm is more stronger and the solution result has superiority of interval optimal solution and scenario decision.

Key words: logistics network design, uncertainty, mixed integer programming, interval variable, genetic algorithm

摘要： 考虑物流网络需求的不确定性，运用区间分析理念以区间数度量不确定性变量与参数，建立区间需求模式下的物流网络设计的混合整数规划模型，定义风险系数与最大约束偏差，对模型进行目标函数与约束条件的确定性转化，设计问题求解的区间递阶优化遗传算法，对不同情景状态下目标函数的区间最优解与节点决策方案进行运算。算例测试表明该算法可操作性更强，求解结果具有区间最优解与情景决策的优越性。

关键词: 物流网络设计, 不确定性, 混合整数规划, 区间变量, 遗传算法