为解决随机需求下的城乡配送中心选址问题，实现物流服务的效率与公平，结合排队论构建了需求转移规则下的排队选址模型。为体现城乡客户需求特征差异，构建时间满意度函数作为服务水平的衡量指标。将客户与配送中心看作M/M/p排队系统，采用需求可转移的排队规则，运用超立方模型计算配送中心的可用概率。目标函数包括嫉妒最小、基尼系数最小及字典序最大等3类公平目标，以及覆盖需求量最大、满意度最大两类效率目标。设计禁忌搜索算法求解，通过算例对模型和算法进行验证。结果表明，P中位模型下效率最优，最小嫉妒目标下公平最优，而最大字典序目标可以兼顾公平与效率。%To solve the location problem of urban-rural distribution centers under the random demand condition, queuing location models are built combined with queuing theory under a demand transfer rule. And the models focus on both equity goals and efficiency goals. The time satisfaction functions as the measure of service level reflects the differences between urban and rural demands. The system of customers and distribution centers is considered to be an M/M/p queuing system. The queuing discipline allows demands transfer when the distribution center is busy. The hypercube model is provided to calculate the probability of centers being available. The objective functions include equity objectives such as minimizing envy, minimizing Gini coefficient and maximizing lexicographical order, and efficiency objectives such as maximizing demands covered and satisfaction. The models are solved by a tabu search algorithm and tested on a real world data set from the distribution system. It is proved that the P-median model is efficiency optimization; the minimize envy model is equity optimization; the maximize lexicographical order model can achieve the equity-efficiency tradeoff.
展开▼
