Research On Multi-return Logistics Distribution Based On Socialized Inventory

With the rapid 'development of e-commerce,the range of online shopping products has been greatly expanded,and online purchases of daily necessities such as takeaway catering,fresh fruits and vegetables have also become a reality.Relying on e-commerce platforms and intra-city distribution logistics,consumers can purchase the products they need without leaving their homes.The intra-city distribution companies deliver goods in a short time,which greatly improves the consumer experience of residents' lives.The transaction volume of the same city distribution and the scale of users continue to rise,and the transaction volume has approached the trillion mark.The number of users will reach 482 million in 2020,with broad market prospects.Correspondingly,the distribution cost of distribution in the same city is high and the distribution efficiency is low,which seriously restricts the development of distribution in the same city.Under this background,this paper studies the same-city distribution of socialized inventory and discusses how the same-city distribution optimizes the distribution path,thereby improving the distribution quality and reducing the distribution cost,which has important theoretical and practical significance.Socialized inventory is a new feature presented by the rapid development of distribution in the same city.The same city distribution optimization of socialized inventory is a special type of multi-return mixed pickup logistics distribution problem.Unlike traditional logistics,it does not have a central warehouse and is supplied by multiple stores in the same city;the goods distributed are not uniformly loaded in the warehouse.Instead,it runs back and forth between the store and the customer point.The vehicle has the characteristics of integration of pickup and delivery;and consumers are more sensitive to the distribution cost and delivery time,which needs to be completed in a shorter time at a lower cost.The interweaving of the above characteristics makes it difficult to optimize the distribution of socialized inventory in the same city,and the current optimal scheduling method cannot be applied.There are two mainstream modeling ideas for same-city distribution,which are virtual copy of store points and customer points according to time and order.By comparing the two modeling ideas,the modeling ideas suitable for the problem in this article are selected to construct integers Linear programming model.In terms of algorithms,since the vehicles in the problem in this paper are homogeneous,the Lagrangian relaxation algorithm only needs to solve once to obtain the lower bound of the problem during the sub-problem solving process,and the efficiency is higher.However,due to the high complexity of the problem in this paper,the traditional Lagrangian relaxation algorithm can only provide the lower bound of the problem,and cannot meet the practical needs.A solution method of a feasible solution to the problem needs to be introduced.The idea of sub-optimization is to accept the path that is inferior to the optimal path within a certain range,so as to cover more paths,so that a feasible solution to the problem can be effectively constructed.Therefore,this paper introduces the sub-optimization idea into the sub-problem solving process,and provides an effective method for constructing the upper bound of the problem for the traditional algorithm,,so as to find a better feasible solution through the continuous approximation of the upper and lower bounds.In order to verify the effectiveness of the algorithm,the standard Solomon example is improved,18 examples with different orders are constructed,and the algorithm solution results are compared with the CPLEX solver,which is better than the solution time and solution results.CPLEX solver.When solving small-scale problems,the distribution logistics companies in the same city can directly use the algorithm of this paper to optimize the scheduling of capacity.When the scale of the problem is large,a suitable stopping condition can be set for the Lagrangian relaxation algorithm,so that it can output the feasible solution of the problem found in a limited time,so as to obtain a better feasible solution.