应急物流的分批配送规划及蚁群优化求解  

Split delivery routing of emergency logistics and ant colony optimization

作　　者：徐志宇[1] 彭嘉臻[2] 许维胜[1] 

机构地区：[1]同济大学电子与信息工程学院,上海201804 [2]特拉华大学交通工程系,美国19716

出　　处：《计算机工程与应用》2011年第24期1-3,8,共4页Computer Engineering and Applications

基　　金：国家自然科学基金(No.70871091);上海市重点学科建设项目(No.B004)~~

摘　　要：针对应急物流的特点,建立分批配送车辆路径规划(SDVRP)模型,实现三大目标:(1)未满足需求最少化;(2)总配送时间最短化;(3)各灾点失衡度最低化。以加权求和的方式转化为单一优化目标,分批配送也节约了救援路径数/车辆数。借鉴并改进最大-最小蚁群算法,设定信息素增量的上下限,避免了结果陷入局部最优。给出数值算例,验证了模型和算法的有效性。The Split Delivery Vehicle Routing Problem（SDVRP） is applied for emergency logistics and a model is estab-lished,which aims at the minimization of（1）the unsatisfied demand,（2）the total delivery time and（3）the supply unbalance among disaster areas.The weighted aggression is performed to compromise the multiple objectives to a single one.Since the dynamic demands of disaster areas are split then satisfied,the number of relief tours and vehicles is reduced.Inspired by the Max-Min Ant System（MMAS）,it sets the upper and lower limits for the increment of pheromone so as to avoid the local optima.A numeric example is presented to validate the effectiveness of the proposed model and algorithm.

关 键 词：分批配送车辆路径规划(SDVRP) 多目标优化(MOP) 应急物流 最大-最小蚂蚁系统(MMAS) 

分 类 号：TP18[自动化与计算机技术—控制理论与控制工程] O221.6[自动化与计算机技术—控制科学与工程]