协作配送中多车型和顾客时间窗对企业成本节约的影响研究

The influence of multiple vehicle types and customer time windows on the cost savings of the alliance members under the collaborative vehicle routing problem

协作配送是实现企业降本增效的重要途径。传统研究多假设联盟车辆同质且未考虑顾客时间窗,但实际中联盟企业的车辆异质以及顾客对配送时效的要求普遍存在。本文考虑每个企业的多车型以及企业之间车型的差异、顾客期望时间窗,建立了以一次性生成2n-1(n为企业数量)个子联盟对应协作配送问题的数学模型,利用自适应大邻域搜索算法求解最优路径,运用Shapley值法分摊大联盟总配送成本。通过设计并求解多组算例,本文验证并探讨了各企业参与协作所实现的经济效益,重点分析了车辆多样性和顾客时间窗对大联盟及每个成员配送成本节约的影响。结果表明:在时间窗约束下,车辆装载能力不完全相同的企业协作配送能够保证货物准时送达,提高车辆平均装载率并减少49.55%~72.77%的配送成本;与大联盟中所有成员均拥有相同的车型相比,某个成员新增一类不同车型有益于减少大联盟的协作总配送成本,降低该成员在大联盟中分摊的成本及分摊比重;某个成员所服务顾客的期望时间窗数量越少或该成员的顾客期望等待时间越长时,不同顾客时间窗之间的协调程度越高,大联盟的协作总配送成本、该成员的成本分摊及分摊比重越小。

As China has become a global logistics power,the single vehicle type cannot meet the large-scale demand,enterprises tend to apply diversified vehicles,and customers have higher requirements for delivery timeliness.Under the situation where enterprises individually distribute goods in competition with others,enterprises face a major challenge to economically and ecologically realize punctual delivery.In the studies of collaborative distribution,it is found that an effective way can be used by enterprises to both reduce costs and increase efficiency,attracting extensive attention from scholars at home and abroad for its economic,environmental and social benefits.The differences in vehicle type and customer time window are in line with real needs,though not much research has been done by now,which introduces the diversification of vehicle types of each enterprise,the differences of vehicle types among different enterprises and the customer′s expected delivery time window into the multi-enterprise cooperative delivery.On the other hand,the relevant research indicated that the members in the alliance always hope that the shared distribution cost after reaching the collaboration is as low as possible,and the distribution cost of the collaboration is highly variable due to the characteristics of partners.In the research of collaborative distribution,it can be seen that most efforts have been on the influence of the number of partners and the overlap degree of customer distribution on the total distribution cost of the collaboration alliance,however the research needs to further focus on the relationship between vehicle types and customer time windows as well as the total distribution cost of the alliance.In addition,it is essential to study the distribution cost allocation changes for each member of the collaboration alliance from an individual perspective,effectively analyze the influence of member characteristics on the collaboration distribution cost.As a result,it is significant to explore how the diversification of vehicle types and customer time windows served by members influence the total distribution cost savings of the collaboration alliance and the cost-sharing savings for each member.In this work,to explore the multi-owner collaborative vehicle routing problem with heterogeneous fleets and hard time windows(MOHFCVRP-HTW),two factors are introduced into the problem of traditional multi-owner collaborative vehicle routing problem(MOCVRP),such as the differences in vehicle types for each enterprise and between different enterprise,the customer′s expected delivery time window.First of all,based on the goal of minimizing the total distribution cost of vehicles delivered to customers,a mathematical model which 2n-1 sub-alliance corresponding to the problem of cooperative delivery of goods was constructed.Secondly,the route optimization model of 2n-1 sub-alliance corresponding to goods transshipment and the quantification model of cooperative total distribution cost are established.Then,the optimal routing was solved by an adaptive large neighborhood search algorithm,and the total distribution cost was shared by the Shapley value method of the cooperative game.Finally,multiple examples were designed and solved to explore the economic benefits obtained by each enterprise in collaboration,the influence of the diversity of vehicle types owned by members,the number of customers with a time window,and the width of the customer time window on the total distribution cost savings for the major collaboration alliance,and the cost-sharing savings for members under the constraints of customer time windows were analyzed.The main results are as follows.Firstly,under the constraints of customers′expected time window,enterprises with different vehicle loading capacities form alliances not only to exchange customer orders but also share resources of vehicles.That can realize the complementary advantages of multiple vehicle types,as well as the complementary geographical location of customer orders in space and a certain degree of coordination in the expected time window,improve the average vehicle loading rate and reduce the delivery cost of each member by 49.55%~72.77%.Secondly,compared with all members owning the same type of vehicle,if a member from the collaboration alliance adds a different type of vehicle,then the total delivery cost of the major collaboration alliance can be significantly saved,and the cost allocation and proportion of the member can witness a significant reduction trend.Thirdly,with the gradual decrease of the number of customers′expected time windows of a member,or the gradual increase of the time range in which customers expect vehicles to service,the coordination degree between different customers′expected time windows will be higher,and the total collaborative delivery cost,delivery cost allocation and allocation proportion of the member will also show a significant downward trend.The findings of this study contribute to the theory of cooperative distribution in a variety of aspects.Firstly,starting from the practical situation of collaborative delivery,two realistic factors are taken into account,which are the diversification of vehicle types for each enterprise and the differences in vehicle types between different enterprises,and the expected service time window of customers.Then,a mathematical model is built to generate 2n-1 sub-alliance corresponding to the cooperative distribution problem.By designing and solving an example,the effectiveness and rationality of the model and the algorithm were verified,which both makes up for the deficiency that the current research pays less attention to the differences in vehicle types and expected time window of customers,and provides ideas for solving the quantification and allocation of the total distribution cost in this kind of collaborative distribution.Secondly,from the perspective of the number of time windows,and the width of the customers′expected time window,the influence of the characteristics of time window on the total distribution cost savings of the major collaboration alliance and cost-sharing savings for each member are analyzed in detail,thereby further deepening the understanding of the fact that the higher the coordination degree between the expected time windows of customers,the lower the total distribution cost of the collaboration alliance.