To solve vehicle routing problem with different fleets, two methodologies are developed. The first methodology adopts twophase strategy. In the first phase, the improved savings method is used to assign customers to a...To solve vehicle routing problem with different fleets, two methodologies are developed. The first methodology adopts twophase strategy. In the first phase, the improved savings method is used to assign customers to appropriate vehicles. In the second phase, the iterated dynasearch algorithm is adopted to route each selected vehicle with the assigned customers. The iterated dynasearch algorithm combines dynasearch algorithm with iterated local search algorithm based on random kicks. The second methodplogy adopts the idea of cyclic transfer which is performed by using dynamic programming algorithm, and the iterated dynasearch algorithm is also embedded in it. The test results show that both methodologies generate better solutions than the traditional method, and the second methodology is superior to the first one.展开更多
Stochastic vehicle routing problems ( VRPs) play important roles in logistics, though they have not been studied systematically yet. The paper summaries the definition, properties and classification of stochastic VRPs...Stochastic vehicle routing problems ( VRPs) play important roles in logistics, though they have not been studied systematically yet. The paper summaries the definition, properties and classification of stochastic VRPs, makes further discussion about two strategies in stochastic VRPs, and at last overviews dynamic and stochastic VRPs.展开更多
针对同时取送货车辆路径问题,考虑客户商品需求差异及车辆异型的因素,以最小化车辆碳排放成本及总配送距离之和为目标,建立多商品分批次取送货的异构绿色车辆路径问题(multi-commodity heterogeneous green vehicle routing problem wit...针对同时取送货车辆路径问题,考虑客户商品需求差异及车辆异型的因素,以最小化车辆碳排放成本及总配送距离之和为目标,建立多商品分批次取送货的异构绿色车辆路径问题(multi-commodity heterogeneous green vehicle routing problem with split pickup and delivery,MCHGVRPSPD)的数学模型,且提出1种增强型变邻域搜索(ehanced variable neighborhood search,EVNS)算法对数学模型进行求解。在EVNS的初始阶段,采用距离–容量平衡法(distance–capacity balancing,DCB)生成初始解;在全局搜索扰动阶段,结合1种自适应扰动操作,防止算法过早收敛陷入局部最优;在局部搜索阶段,采用4种带容量约束的邻域搜索操作,以探测更优质的邻域解空间。最后,采用GA,VNS和ALNS算法进行测试案例仿真实验,验证EVNS算法求解MCHGVRPSPD的有效性。结果表明:与3种对比算法相比,EVNS算法在求解质量方面提升了15%~25%的性能,同时在收敛性和稳定性方面更优,是1种求解MCHGVRPSPD的有效算法。展开更多
针对带时间窗的时间依赖型同时取送货车辆路径问题(Time Dependent Vehicle Routing Problem with Simultaneous Pickup-Delivery and Time Windows,TDVRPSPDTW),本文建立以车辆固定成本、驾驶员成本、燃油消耗及碳排放成本之和为优化...针对带时间窗的时间依赖型同时取送货车辆路径问题(Time Dependent Vehicle Routing Problem with Simultaneous Pickup-Delivery and Time Windows,TDVRPSPDTW),本文建立以车辆固定成本、驾驶员成本、燃油消耗及碳排放成本之和为优化目标的数学模型;并在传统蚁群算法的基础上,利用节约启发式构造初始解初始化信息素,改进状态转移规则,引入局部搜索策略,提出一种带自适应大邻域搜索的混合蚁群算法(Ant Colony Optimization with Adaptive Large Neighborhood Search,ACO-ALNS)进行求解;最后,分别选取基准问题算例和改编生成TDVRPSPDTW算例进行实验。实验结果表明:本文提出的ACO-ALNS算法可有效解决TDVRPSPDTW的基准问题;相较于模拟退火算法和带局部搜索的蚁群算法,本文算法求解得到的总配送成本最优值平均分别改善7.56%和2.90%;另外,相比于仅考虑碳排放或配送时间的模型,本文所构建的模型综合多种因素,总配送成本平均分别降低4.38%和3.18%,可有效提高物流企业的经济效益。展开更多
基金The National Natural Science Founda-tion of China ( No.70471039)the National Social Science Foundation of China (No.07BJY038)the Program for New Century Excellent Talents in University (No.NCET-04-0886)
文摘To solve vehicle routing problem with different fleets, two methodologies are developed. The first methodology adopts twophase strategy. In the first phase, the improved savings method is used to assign customers to appropriate vehicles. In the second phase, the iterated dynasearch algorithm is adopted to route each selected vehicle with the assigned customers. The iterated dynasearch algorithm combines dynasearch algorithm with iterated local search algorithm based on random kicks. The second methodplogy adopts the idea of cyclic transfer which is performed by using dynamic programming algorithm, and the iterated dynasearch algorithm is also embedded in it. The test results show that both methodologies generate better solutions than the traditional method, and the second methodology is superior to the first one.
基金Supported by the National Natural Science Fundation of China( No. 70071028 and 79700019).
文摘Stochastic vehicle routing problems ( VRPs) play important roles in logistics, though they have not been studied systematically yet. The paper summaries the definition, properties and classification of stochastic VRPs, makes further discussion about two strategies in stochastic VRPs, and at last overviews dynamic and stochastic VRPs.
文摘针对同时取送货车辆路径问题,考虑客户商品需求差异及车辆异型的因素,以最小化车辆碳排放成本及总配送距离之和为目标,建立多商品分批次取送货的异构绿色车辆路径问题(multi-commodity heterogeneous green vehicle routing problem with split pickup and delivery,MCHGVRPSPD)的数学模型,且提出1种增强型变邻域搜索(ehanced variable neighborhood search,EVNS)算法对数学模型进行求解。在EVNS的初始阶段,采用距离–容量平衡法(distance–capacity balancing,DCB)生成初始解;在全局搜索扰动阶段,结合1种自适应扰动操作,防止算法过早收敛陷入局部最优;在局部搜索阶段,采用4种带容量约束的邻域搜索操作,以探测更优质的邻域解空间。最后,采用GA,VNS和ALNS算法进行测试案例仿真实验,验证EVNS算法求解MCHGVRPSPD的有效性。结果表明:与3种对比算法相比,EVNS算法在求解质量方面提升了15%~25%的性能,同时在收敛性和稳定性方面更优,是1种求解MCHGVRPSPD的有效算法。
文摘针对带时间窗的时间依赖型同时取送货车辆路径问题(Time Dependent Vehicle Routing Problem with Simultaneous Pickup-Delivery and Time Windows,TDVRPSPDTW),本文建立以车辆固定成本、驾驶员成本、燃油消耗及碳排放成本之和为优化目标的数学模型;并在传统蚁群算法的基础上,利用节约启发式构造初始解初始化信息素,改进状态转移规则,引入局部搜索策略,提出一种带自适应大邻域搜索的混合蚁群算法(Ant Colony Optimization with Adaptive Large Neighborhood Search,ACO-ALNS)进行求解;最后,分别选取基准问题算例和改编生成TDVRPSPDTW算例进行实验。实验结果表明:本文提出的ACO-ALNS算法可有效解决TDVRPSPDTW的基准问题;相较于模拟退火算法和带局部搜索的蚁群算法,本文算法求解得到的总配送成本最优值平均分别改善7.56%和2.90%;另外,相比于仅考虑碳排放或配送时间的模型,本文所构建的模型综合多种因素,总配送成本平均分别降低4.38%和3.18%,可有效提高物流企业的经济效益。
