Research On Optimization Of Omni-channel Retail Order Fulfillment Problem Based On Order Split And Inventory Transshipment Strategies

With the rise of new retail,omni-channel retail,which combines online and offline channels for order fulfillment,has become a development trend in the retail industry.To improve delivery efficiency,many companies have begun implementing omni-channel strategies that involve setting up front warehouses and shipping from stores,breaking the traditional layout of online and offline channels and making it an urgent problem to coordinate the online and offline channels for order allocation and omni-channel distribution network location.This article studies the decision problem of retailers in the decision-making window from order arrival to allocation,where a large number of received orders are uniformly distributed to different fulfillment channels based on the omni-channel distribution network.The omni-channel distribution network includes three types of supply points: regional warehouses,fulfillment centers,and stores.Regional warehouses are responsible for replenishing fulfillment centers and stores,and can also ship online orders;fulfillment centers and stores serve as delivery stations closer to customers and ship packages to customer locations.Considering the limited inventory levels of stores and fulfillment centers,in the decision-making process,when inventory cannot meet all the demands of online orders,it is allowed to split orders for fulfillment by other supply points;at the same time,it is also allowed for stores and fulfillment centers to urgently replenish inventory from each other through inventory transshipment to meet demand.To compare the differences between different order split methods,a Product Single Allocation Model(PSA),a Product Multiple Allocation Model(PMA),and an Order Single Allocation Model(OSA)were established.In addition,a greedy algorithm was designed to simulate the process of retailers fulfilling online orders in real time as another benchmark for comparison.Through case analysis,it is known that: 1)There is almost no difference in cost and the number of order splits between Product Multiple Allocation and Product Single Allocation;2)Whether using order split strategy alone,inventory transshipment strategy alone,or using both strategies together,can effectively reduce order fulfillment costs;and 3)Using both the order split and inventory transshipment strategies together can increase the number of inventory transshipments.Building on the omni-channel distribution network mentioned above,this article further studies the location selection problem of the distribution network,where the decision is whether to establish new fulfillment centers or renovate existing stores.In the location selection problem,the uncertainty of online order demand and offline store demand is considered,and a two-stage stochastic programming model is constructed to describe the location selection problem with uncertain demand.When solving the two-stage stochastic programming model,considering that the expected value in the second stage is difficult to solve,the SAA(Sample Average Approximation)method is used to approximate it.In the case analysis,the number of scenarios required for the experiment is first determined through the SAA process,and then sensitivity analyses are conducted on "order split or not," "fulfillment center construction cost," and "regional warehouse replenishment and inventory costs." The results of the case analysis show that: 1)Using order split strategy in the location model has the benefit of reducing order fulfillment costs;2)As the construction cost of the fulfillment center increases,the location plan tends to renovate stores,and as the number of renovated stores increases,the order split rate and the number of orders allocated to stores for fulfillment also increase;and 3)As the regional warehouse replenishment and inventory costs increase,the location plan also tends to renovate stores,and the order split rate and the number of orders allocated to stores for fulfillment also increase,but the number of orders fulfilled by the regional warehouse channel decreases,reflecting the seesaw effect between the regional warehouse channel and the store channel in the location problem.In summary,this article expands the research on order allocation and location selection problems under the background of omni-channel retail,and provides decision support for retail companies to layout omni-channel strategies.