Inventory-based warehouse network optimization method, device, computer equipment and storage medium

Abstract

The present application relates to a warehouse network optimization method, device, computer equipment and storage medium based on inventory. The method includes: obtaining the initial warehouse network structure constructed for the transport order as the warehouse network structure to be disturbed; disturbing and updating the hierarchical status of the candidate warehouses and the remaining candidate warehouses in the warehouse network structure to be disturbed, and determining the updated hierarchical status of each candidate warehouse; reconstructing the warehouse network structure according to the inventory capacity limit and updated hierarchical status of each candidate warehouse to obtain the updated warehouse network structure; respectively predicting the inventory cost of each candidate warehouse in the warehouse network structure to be disturbed and each candidate warehouse in the updated warehouse network structure to obtain the inventory cost of the warehouse network structure to be disturbed and the updated warehouse network structure; according to the inventory cost of the warehouse network structure to be disturbed and the updated warehouse network structure, performing quality assessment on the warehouse network structure to be disturbed and the updated warehouse network structure to determine the optimal warehouse network structure. The use of this method can improve the accuracy of optimization.

Description

Technical Field

[0001] The present application relates to the field of logistics technology, and in particular to an inventory-based warehouse network optimization method, device, computer equipment and storage medium. Background Art

[0002] With the rapid development of e-commerce and logistics technology, warehouse operations are gradually becoming digitalized. For example, warehouse management systems are gradually being developed to automatically manage orders and use corresponding digital auxiliary equipment to automatically complete picking operations. Before picking orders for an order, it is usually necessary to determine the supply chain network of the order. The supply chain network of the order is determined by building a suitable warehouse network structure for the order. Traditional optimization of the warehouse network structure is mainly based on the transportation cost between nodes in the warehouse structure.

[0003] However, traditional optimization based on transportation costs only needs to consider transportation, while the transportation methods between warehouses in different industries and scenarios usually do not change much, which reduces the accuracy of optimization. Summary of the invention

[0004] Based on this, it is necessary to provide an inventory-based warehouse network optimization method, device, computer equipment and storage medium that can improve accuracy in response to the above technical problems.

[0005] A warehouse network optimization method based on inventory, the method comprising:

[0006] Obtaining the initial warehouse network structure constructed for the transport order as the warehouse network structure to be disturbed;

[0007] Disturb and update the hierarchical status of the candidate warehouses in the warehouse network structure to be disturbed and the remaining candidate warehouses, and determine the updated hierarchical status of each of the candidate warehouses;

[0008] Reconstructing the warehouse network structure according to the inventory capacity limit of each candidate warehouse and the updated hierarchical status to obtain an updated warehouse network structure;

[0009] Respectively predicting the inventory costs of each candidate warehouse in the warehouse network structure to be disturbed and each candidate warehouse in the updated warehouse network structure to obtain the inventory costs of the warehouse network structure to be disturbed and the updated warehouse network structure;

[0010] According to the inventory costs of the warehouse-network structure to be disturbed and the updated warehouse-network structure, quality assessment is performed on the warehouse-network structure to be disturbed and the updated warehouse-network structure to determine the optimal warehouse-network structure.

[0011] A warehouse network optimization device based on inventory, the device comprising:

[0012] An acquisition module, used for acquiring an initial warehouse network structure constructed for a transport order as a warehouse network structure to be disturbed;

[0013] A disturbance module, used for disturbing and updating the hierarchical status of the candidate warehouses and the remaining candidate warehouses in the warehouse network structure to be disturbed, and determining the updated hierarchical status of each candidate warehouse;

[0014] A construction module, configured to reconstruct a warehouse network structure according to the inventory capacity limit of each candidate warehouse and the update hierarchical status to obtain an updated warehouse network structure;

[0015] A prediction module, used to respectively predict the inventory cost of each candidate warehouse in the warehouse network structure to be disturbed and each candidate warehouse in the updated warehouse network structure, to obtain the inventory cost of the warehouse network structure to be disturbed and the updated warehouse network structure;

[0016] An evaluation module is used to perform quality evaluation on the warehouse network structure to be disturbed and the updated warehouse network structure according to the inventory costs of the warehouse network structure to be disturbed and the updated warehouse network structure, so as to determine the optimal warehouse network structure.

[0017] A computer device includes a memory and a processor, wherein the memory stores a computer program, and the processor implements the steps of the above-mentioned inventory-based warehouse network optimization method when executing the computer program.

[0018] A computer-readable storage medium having a computer program stored thereon, wherein the computer program, when executed by a processor, implements the steps of any one of the above-mentioned inventory-based warehouse network optimization methods.

[0019] The above-mentioned warehouse network optimization method, device, computer equipment and storage medium based on inventory obtain the initial warehouse network structure constructed for the transport order as the warehouse network structure to be disturbed; the hierarchical state of the candidate warehouses and the remaining candidate warehouses in the warehouse network structure to be disturbed is disturbed and updated, and after the updated hierarchical state of each candidate warehouse is determined, the warehouse network structure is reconstructed according to the inventory capacity limit and the updated hierarchical state of each candidate warehouse to obtain the updated warehouse network structure; and then the inventory cost of each candidate warehouse in the warehouse network structure to be disturbed and each candidate warehouse in the updated warehouse network structure is predicted respectively, and the inventory cost of the warehouse network structure to be disturbed and the updated warehouse network structure is obtained; according to the inventory cost of the warehouse network structure to be disturbed and the updated warehouse network structure, the quality of the warehouse network structure to be disturbed and the updated warehouse network structure is evaluated to determine the optimal warehouse network structure. Since warehouses in different industries and different scenarios have certain industry characteristics, this method optimizes the warehouse network based on the inventory cost of the candidate warehouse, improves the scope of applicable scenarios, and thus improves the accuracy of optimization. At the same time, the method constructs the warehouse network structure based on the inventory capacity limit, fully considers the constraints of the upper limit of the warehouse inventory capacity, and further improves the executability of the warehouse network structure. BRIEF DESCRIPTION OF THE DRAWINGS

[0020] Figure 1This is an application environment diagram of an inventory-based warehouse network optimization method in one embodiment;

[0021] Figure 2 It is a flowchart of a warehouse network optimization method based on inventory in one embodiment;

[0022] Figure 3a is a structural schematic diagram of an initial warehouse network structure in an embodiment;

[0023] Figure 3b is a structural schematic diagram of an initial warehouse network structure in an embodiment;

[0024] Figure 4 It is a flowchart of the steps of reconstructing the warehouse network structure and obtaining the steps of updating the warehouse network structure according to the inventory capacity limit and the update level status of each candidate warehouse in one embodiment;

[0025] Figure 5 A schematic diagram of a flow chart of a warehouse network optimization method based on inventory in another embodiment;

[0026] Figure 6 It is a structural block diagram of a warehouse network optimization device based on inventory in one embodiment;

[0027] Figure 7 FIG. 4 is a diagram showing the internal structure of a computer device in one embodiment. DETAILED DESCRIPTION

[0028] In order to make the purpose, technical solution and advantages of the present application more clearly understood, the present application is further described in detail below in conjunction with the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are only used to explain the present application and are not used to limit the present application.

[0029] The inventory-based warehouse network optimization method provided in this application can be applied to Figure 1 In the application environment shown in the figure, the terminal 102 communicates with the server 104 through the network. When the terminal 102 receives the warehouse network optimization instruction issued by the user, the terminal 102 responds to the warehouse network optimization instruction and implements the warehouse network optimization method based on inventory alone. Alternatively, the terminal 102 sends the warehouse network optimization instruction to the server 104, and the server 104 implements the warehouse network optimization method based on inventory.

[0030] This embodiment takes the server 104 as an example. Specifically, the server 104 obtains the initial warehouse network structure constructed for the transport order as the warehouse network structure to be disturbed; the server 104 disturbs and updates the hierarchical status of the candidate warehouses and the remaining candidate warehouses in the warehouse network structure to be disturbed, and determines the updated hierarchical status of each candidate warehouse; the server 104 reconstructs the warehouse network structure according to the inventory capacity limit and the updated hierarchical status of each candidate warehouse to obtain the updated warehouse network structure; the server 104 predicts the inventory cost of each candidate warehouse in the warehouse network structure to be disturbed and each candidate warehouse in the updated warehouse network structure respectively, and obtains the inventory cost of the warehouse network structure to be disturbed and the updated warehouse network structure; the server 104 performs quality assessment on the warehouse network structure to be disturbed and the updated warehouse network structure according to the inventory cost of the warehouse network structure to be disturbed and the updated warehouse network structure, and determines the optimal warehouse network structure. Among them, the terminal 102 can be but not limited to various personal computers, laptops, smart phones, tablet computers and portable wearable devices, and the server 104 can be implemented by an independent server or a server cluster composed of multiple servers.

[0031] In one embodiment, Figure 2 As shown in the figure, a warehouse network optimization method based on inventory is provided. Figure 1 The server in the example is used to illustrate the following steps:

[0032] Step S202: acquiring an initial warehouse network structure constructed for the transport order as the warehouse network structure to be disturbed.

[0033] Among them, the transport order is the order voucher sent by the user to the supplier. The transport order includes the products that the user wants to order for transportation, and for the supplier, it is the products to be transported. The initial warehouse network structure is the warehouse network structure constructed by the server according to the transport order, which can be understood as the first warehouse network structure constructed for the transport order. The warehouse network structure to be disturbed can be understood as the warehouse network structure that needs to be updated and optimized.

[0034] Specifically, the server obtains candidate warehouses, warehouse constraints and transportation requirements corresponding to the transport order. Among them, the warehouse is a storage facility. The warehouse in this embodiment is used to store products to be transported. The candidate warehouse is to select some warehouses that match the transport order from all warehouses. The candidate warehouse is selected and determined by the staff of the supplier according to the transport order. Warehouse constraints refer to the constraints on the warehouses of the warehouse network structure to be constructed, including the constraints on the number of warehouses of the warehouse network structure to be constructed, the constraints on the number of warehouses at different warehouse levels, etc. The transportation requirements include requirements such as transportation timeliness and transportation cost. The warehouse constraints and transportation requirements are also input by the staff according to the actual situation. Then, the server constructs the initial warehouse network structure according to the candidate warehouses, warehouse constraints and transportation requirements corresponding to the transport order. When the server receives a warehouse network optimization request instruction for the transport order, the server obtains the constructed initial warehouse network structure as the warehouse network structure to be disturbed. Alternatively, when the server receives a warehouse network optimization request instruction for the transport order, the server first obtains the candidate warehouses, warehouse constraints and transportation requirements corresponding to the transport order. The server then constructs an initial warehouse network structure based on the obtained candidate warehouses, warehouse constraints and transportation requirements, thereby obtaining the warehouse network structure to be disturbed.

[0035] In one embodiment, constructing an initial warehouse network structure according to candidate warehouses, warehouse constraints and transportation requirements corresponding to a transportation order includes the following steps:

[0036] S1, determine the initial hierarchical state of each candidate warehouse according to the warehouse constraints.

[0037] The hierarchical status refers to the warehouse level of each candidate warehouse, which includes the supply factory, central warehouse, regional warehouse, etc. The initial hierarchical status is the initial warehouse level assigned to the candidate warehouse by the warehouse constraint. For example, the initial hierarchical status of any candidate warehouse can be a central warehouse, a regional warehouse, or an unselected warehouse. The central warehouse is the warehouse at the upper level of the regional warehouse.

[0038] Specifically, after the server obtains the candidate warehouses and warehouse constraints, it assigns an initial hierarchical state to each candidate warehouse according to the warehouse constraints, and each candidate warehouse obtains the initial hierarchical state. For example, suppose there are 10 candidate warehouses in total, and the warehouse constraints require 2 warehouses as central warehouses and 3 warehouses as regional warehouses. The server randomly selects 5 warehouses from the 10 candidate warehouses, determines 2 of them as central warehouses, and the other 3 as regional warehouses. The 5 candidate warehouses that are not selected are temporarily designated as unused warehouses.

[0039] In one embodiment, when determining the initial hierarchical state for each candidate warehouse, the initial hierarchical state can be determined for each candidate warehouse based on the FirstFit (first adaptation) algorithm and warehouse constraints. The basic idea of ​​the FirstFit algorithm is to assign a suitable state to each candidate point one by one. When each candidate point has a state, it represents an initial solution, that is, the framework of an initial warehouse network structure is formed. Specifically in this embodiment, the server assigns initial hierarchical states to the candidate warehouses one by one according to the order in which the staff inputs the candidate warehouses and according to the warehouse constraints. For example, the first candidate warehouse and the second candidate warehouse are determined as central warehouses, the third candidate warehouse, the fourth candidate warehouse and the fifth candidate warehouse are determined as regional warehouses, and the last five candidate warehouses input are temporarily determined as warehouses that are not selected. In this embodiment, the initial hierarchical state of each candidate warehouse is determined by the FirstFit algorithm to save the server's allocation time and improve the efficiency of the construction.

[0040] S2, construct the initial warehouse network structure according to the initial hierarchical status and transportation demand of each candidate warehouse, and use the initial warehouse network structure as the warehouse network structure to be disturbed.

[0041] Specifically, when the server determines the initial hierarchical status of each candidate warehouse, the structural framework of the initial warehouse network structure can be directly determined. The warehouse network structure framework can be understood as a structure in which candidate warehouses have been selected but the supply and replenishment relationship between the candidate warehouses has not yet been determined. Therefore, the server determines the supply and replenishment relationship between each candidate warehouse based on the transportation demand. According to the supply and replenishment relationship, the supply chain of each candidate warehouse is constructed to obtain a complete warehouse network structure, that is, the initial warehouse network structure of this embodiment. For example, assuming that the transportation demand includes transportation time, the transportation time requirement is that the warehouse with the shortest transportation time is used as the supply and replenishment warehouse. When the server constructs the supply chain of each candidate warehouse, it is based on the warehouse with the shortest transportation time. Figure 3a As shown, a schematic diagram of the structure of an initial warehouse network structure is provided, referring to Figure 3a , the initial warehouse network structure includes four levels, including 2 supply factories, 2 central warehouses, 4 regional warehouses and multiple customers. The supply chains of the four levels are determined according to the transportation demand. Assuming that the central warehouse with the shortest transportation time for replenishing the first regional warehouse is the first central warehouse, the server selects the first regional warehouse and the first central warehouse to build a supply and replenishment chain. When the supply chains between all levels are constructed, the constructed warehouse network structure is obtained. The transportation time server can determine it according to the distance between the warehouses and the mode of transportation. In addition, the conditions for constructing the warehouse network structure can also include rules including adaptive selection of the best network level and relaxation of the level restrictions of optional warehouses. That is, the customer's selected warehouse level is not restricted, and the final output network level may be less than the candidate network level range. For example Figure 3bAs shown, when building a supply chain for customers based on transportation needs, it is determined that customers can directly purchase goods from the central warehouse. Then, the regional warehouse network level can be eliminated from the warehouse network structure, so that the warehouse network structure only includes three levels, thereby flattening the entire warehouse network and reducing costs and improving efficiency.

[0042] Step S204: disturb and update the hierarchical status of the candidate warehouses in the warehouse network structure to be disturbed and the remaining candidate warehouses, and determine the updated hierarchical status of each candidate warehouse.

[0043] Among them, the disturbance update corresponds to the concept of neighborhood space search in the neighborhood search, that is, the hierarchical state of some candidate warehouses of the current solution is modified to obtain the hierarchical state of the candidate warehouses of the new solution.

[0044] Specifically, the server modifies the hierarchical state of each candidate warehouse in the currently determined warehouse network structure to be disturbed, and the modified hierarchical state is the updated hierarchical state of each candidate warehouse in the warehouse network structure to be disturbed. Since the warehouse network structure always needs to meet the warehouse constraint conditions, after the hierarchical state of the candidate warehouse in the warehouse network structure to be disturbed is disturbed and updated, it is also necessary to disturb and update the remaining hierarchical states that have not been selected to form the warehouse network structure according to the updated hierarchical states of the candidate warehouses in the warehouse network structure to be disturbed. For example, 7 of the 10 candidate warehouses are constructed as the warehouse network structure to be disturbed, and there are 3 remaining candidate warehouses that have not been selected. After the 7 candidate warehouses in the warehouse network structure to be disturbed are disturbed and updated, the remaining 3 candidate warehouses are further disturbed and updated according to the result of the disturbance update, so that the hierarchical states of the 10 candidate warehouses always meet the warehouse constraint conditions.

[0045] In one embodiment, the specific implementation of the perturbation update of the hierarchical state of each candidate warehouse can define the corresponding perturbation strategy according to the actual scenario requirements, and the degree of perturbation of different perturbation strategies defined for different scenarios is different. However, it should be understood that no matter how the perturbation strategy is defined, it is necessary to follow a common goal, that is, the perturbation update should have a greater possibility of improving the quality of the solution. In this embodiment, the perturbation strategy includes two strategies: hierarchical state change and hierarchical state exchange. The strategy of hierarchical state change is to directly modify the initial hierarchical state of the candidate warehouse, and the strategy of hierarchical state exchange is to exchange the initial hierarchical states of multiple candidate warehouses.

[0046] Step S206, reconstructing the warehouse network structure according to the inventory capacity limit and update level status of each candidate warehouse to obtain an updated warehouse network structure.

[0047] Among them, the inventory capacity limit is the maximum inventory capacity of the candidate warehouse. Specifically, after the disturbance update obtains the updated hierarchical status of the candidate warehouse, since the hierarchical status of each candidate warehouse has changed, for example, the original central warehouse may be disturbed and updated to become a regional warehouse or a warehouse that is not selected. Therefore, the server rebuilds the warehouse network structure according to the updated hierarchical status and transportation requirements.

[0048] The principle of reconstructing the warehouse network structure is roughly the same as that of constructing the initial warehouse network structure. The difference is that in this embodiment, the constraint of the inventory capacity limit is added when reconstructing the warehouse network structure. That is, when the server reconstructs the supply and replenishment chain between candidate warehouses, it is necessary to consider the inventory capacity limit of the candidate warehouse. When the server is looking for an upstream warehouse for supplying a warehouse W, it is necessary to consider whether the newly added inventory area brought by warehouse W to the upstream warehouse for supply exceeds the risk of the inventory capacity upper limit area of ​​the upstream warehouse. Once the newly added inventory area brought by warehouse W exceeds the inventory capacity upper limit area of ​​the candidate upstream warehouse, the upstream candidate warehouse is rejected as the upstream supply warehouse of warehouse W.

[0049] Step S208, predicting the inventory cost of each candidate warehouse in the warehouse network structure to be disturbed and each candidate warehouse in the updated warehouse network structure respectively, and obtaining the inventory cost of the warehouse network structure to be disturbed and the updated warehouse network structure.

[0050] Among them, inventory cost refers to inventory-related costs. In actual scenarios, in addition to transportation costs, inventory-related costs also account for a large proportion of costs. Inventory-related costs usually include the following parts: inventory rental costs, inventory capital occupation costs, inventory loss and discount costs, and operation-related costs.

[0051] Specifically, the server first predicts the inventory cost of each candidate warehouse in the warehouse network structure to be disturbed and the updated warehouse network structure. Then, the sum of the inventory costs of all candidate warehouses in the warehouse network result to be disturbed is the inventory cost of the warehouse network structure to be disturbed. Similarly, the sum of the inventory costs of each candidate warehouse in the updated warehouse network structure is the inventory cost of the updated warehouse network structure.

[0052] Step S210, based on the inventory costs of the warehouse-network structure to be disturbed and the updated warehouse-network structure, quality assessment is performed on the warehouse-network structure to be disturbed and the updated warehouse-network structure to determine the optimal warehouse-network structure.

[0053] Specifically, the server compares the inventory costs of the warehouse network structure to be disturbed and the updated warehouse network structure, and selects the warehouse network structure with the lower inventory cost as the best warehouse network structure according to the comparison result. For example, if the inventory cost of the warehouse network structure to be disturbed is lower than the inventory cost of the updated warehouse network structure, the quality of the warehouse network structure to be disturbed is better than the updated warehouse network structure, and the warehouse network structure to be disturbed is selected as the best warehouse network structure. If the inventory cost of the warehouse network structure to be disturbed is higher than the inventory cost of the updated warehouse network structure, the quality of the updated warehouse network structure is better than the warehouse network structure to be disturbed, and the updated warehouse network structure is selected as the best warehouse network structure.

[0054] In the above-mentioned warehouse network optimization method based on inventory, since warehouses in different industries and scenarios have certain industry characteristics, this method optimizes the warehouse network based on the inventory cost of the candidate warehouse, improves the scope of applicable scenarios, and thus improves the accuracy of optimization. At the same time, this method constructs the warehouse network structure based on the inventory capacity limit, fully considering the constraints of the warehouse inventory capacity upper limit, and further improves the executability of the warehouse network structure.

[0055] In one embodiment, Figure 4 As shown, step S206 includes:

[0056] Step S402, according to the update level status of each candidate warehouse, select a candidate warehouse from each candidate warehouse as a warehouse network node, and the warehouse network nodes constitute a warehouse network structure framework.

[0057] Step S404, determining the upstream supply warehouse of each warehouse network node from the warehouse network structure framework.

[0058] Step S406, screening candidate upstream supply warehouses whose inventory capacity limits meet the requirements of the warehouse network nodes from the candidate upstream supply warehouses corresponding to each warehouse network node.

[0059] Step S408, according to the transportation demand of the transportation order, determine the upstream supply warehouse of the warehouse network node from the candidate upstream supply warehouses whose inventory capacity limit meets the demand of the warehouse network node.

[0060] Step S410, establishing a supply and replenishment chain of the warehouse network structure framework according to the upstream supply warehouses of each warehouse network node in the warehouse network structure framework, and obtaining an updated warehouse network structure.

[0061] Among them, the warehouse network node is a candidate warehouse reselected according to the update level status of the candidate warehouse to build a warehouse network structure. It can be understood as a candidate warehouse in the updated warehouse network structure to be built.

[0062] Specifically, when the server obtains the updated hierarchical status of each candidate warehouse through disturbance update, the candidate warehouses that can be used as candidate warehouses in the warehouse network structure are re-determined according to the updated hierarchical status of each candidate warehouse, and the warehouse network nodes are obtained. The warehouse network structure without a determined supply and replenishment relationship is formed by the warehouse network nodes, that is, the warehouse network structure framework. After obtaining the warehouse network structure framework, the server determines the candidate upstream supply warehouses of each warehouse network structure according to the updated hierarchical status of each warehouse network node in the warehouse network structure framework. For example, in the warehouse network structure framework, the warehouse network node with the updated hierarchical status of the regional warehouse, all the warehouse network nodes that are central warehouses in the warehouse network structure framework are candidate upstream supply warehouses for the regional warehouse.

[0063] After the server determines the candidate upstream supply warehouses for each warehouse network node in the warehouse network structure, it calculates the additional inventory area brought by the warehouse network node for each candidate upstream supply warehouse. The candidate upstream supply warehouses whose additional inventory area brought by the warehouse network node will not exceed the inventory capacity limit of the candidate upstream supply warehouse are screened. For example, when a warehouse network node has three candidate upstream supply warehouses A, B and C, the three candidate upstream supply warehouses are screened according to the additional inventory area brought by the warehouse network node, and the candidate upstream supply warehouses whose inventory capacity limit can meet the additional inventory area brought by the warehouse network node are screened. The candidate upstream supply warehouses obtained by screening can be any one or more of A, B and C.

[0064] Then, the server determines the upstream supply warehouse of the warehouse network node from the screened candidate upstream supply warehouses according to the transportation demand. After obtaining the upstream supply warehouses of each warehouse network node, it means that the supply and replenishment relationship between each warehouse network node in the warehouse network structure framework has been determined. The server can directly establish the supply and replenishment chain of the warehouse network structure framework according to the determined upstream supply warehouse to obtain the updated warehouse network structure of this embodiment. For example, the original candidate upstream supply warehouses of the warehouse network node include 4 candidate warehouses. After screening by inventory capacity limit, there are 2 candidate upstream supply warehouses remaining in the candidate upstream supply warehouses. Then, according to the transportation demand, the server selects the warehouse with the shortest transportation time from these 2 candidate upstream supply warehouses as the upstream supply warehouse of the warehouse network node.

[0065] In one embodiment, step S406 includes: obtaining area data, demand quantity, and the number of customers served by the warehouse network node for products that the warehouse network node needs the candidate upstream supply warehouse to supply; determining the inventory area occupied by the products required by the warehouse network node based on the area data and demand quantity of the products, and obtaining the first newly added inventory area of ​​the candidate upstream supply warehouse; determining the newly added safety inventory of the products required by the warehouse network node based on the number of customers served by the warehouse network node, and determining the second newly added inventory area of ​​the candidate upstream supply warehouse based on the newly added safety inventory and the area data; summing the first newly added inventory area and the second newly added inventory area with the used inventory area of ​​the candidate upstream supply warehouse, and when the sum is less than or equal to the inventory area upper limit of the candidate upstream supply warehouse, determining that the inventory capacity limit of the candidate upstream supply warehouse meets the needs of the warehouse network node.

[0066] Among them, area data refers to data used to calculate the inventory area occupied by the product, the demand quantity is the number of products required by the currently determined warehouse network node, the number of service customers is the number of customers that the warehouse network node needs to serve, and the number of service customers of a warehouse network node should include the number of directly served customers and the number of indirectly served customers. The customers of the direct server can be understood as the served customers, that is, the customers directly supplied by the warehouse network node, and the indirectly served customers can be understood as the customers served by the downstream warehouse covered by the warehouse network node. Due to the different specifications of the products, the area data of the products in this embodiment can have three dimensions of inventory area conversion coefficients, including piece-to-surface ratio (the number of pieces of goods stored per square meter of inventory area), weight-to-area ratio (the number of kilograms of goods stored per square meter of inventory area), and volume-to-area ratio (the number of cubic meters of goods stored per square meter of inventory area). The area data of each product is given from these three dimensions according to the actual specifications of the product.

[0067] Specifically, when the server calculates the additional inventory area brought by the current warehouse network node to its corresponding upstream candidate supply warehouse, it calculates the additional inventory area brought by its own demand based on the area data of the demand product of the warehouse network node and the demand quantity, that is, the first additional inventory area. Assume that, taking the piece quantity dimension as an example, the piece-to-surface ratio of the product is 0.5 pieces / square meter. When the target inventory of the product in a candidate warehouse is 1,000 pieces, the corresponding inventory area that the product needs to occupy in the candidate warehouse is 1,000 pieces / (0.5 pieces / square meter) = 2,000 square meters.

[0068] In addition, in addition to the inventory area brought by the downstream supply demand of the regular storage of the warehouse itself, the safety stock also needs to be considered. That is, specifically for this application, due to the consideration of the safety stock of the warehouse, when the same warehouse network node is added as the downstream for different candidate upstream supply warehouses, the inventory brought by the downstream warehouse network node and its corresponding inventory area increase are different, and the increase process is not linear. The newly added inventory area of ​​the increased safety stock is related to the number of customers served by the downstream warehouse network node. Safety stock = lead time standard deviation under a certain service level of the warehouse * warehouse demand standard deviation * warehouse replenishment lead time. The calculation method of each component is as follows:

[0069] The standard deviation of the lead time under a certain service level of the warehouse is calculated using the cumulative distribution function:

[0070]

[0071] The cumulative distribution function in the above formula can be understood as the calculation of the assumed customer demand x subject to a mathematical expectation of μ and a standard deviation of σ. 2 The normal distribution is recorded as: X~N(μ,σ 2 ).

[0072] The corresponding function of the cumulative distribution function in Excel is NORMSINV(). For example, if the service level is 95%, we can get NORMSINV(0.95)=1.645, and the corresponding lead time standard deviation is 1.645. In actual products, the lead time standard deviations corresponding to different service levels can be checked in advance and stored in the comparison table. In subsequent use, you only need to check the comparison table to get the values ​​of different service levels. The service level of the candidate warehouse at different levels is given by the customer, and the value can only be between 50% and 99.9%. In actual products, the default value obtained from the experience value of the inventory module can also be used.

[0073] When customer demand conforms to an independent normal distribution and has the same variance, the standard deviation of warehouse demand = standard deviation of customer demand * number of customer outlets served. Among them, the number of "customer outlets served" is the number of customers served by the corresponding warehouse under the current iteration of the warehouse-network structure. That is, the number of customers served by the current warehouse-network node. It should be understood that for each additional downstream warehouse W-2 added to the warehouse-network node W-1, the number of downstream customer outlets added by the warehouse-network node W-1 is the number of all customer outlets (including directly and indirectly served customer outlets) in the downstream warehouse W-2 that use the warehouse-network node W-1 for replenishment. The warehouse replenishment lead time is provided by customer input, which is divided into warehouse-level inputs. For example: the replenishment lead time for the same warehouse as a central warehouse and a regional warehouse should be different. In this implementation, without considering the fluctuations in the replenishment lead time due to transportation differences, the lead time can be considered a fixed value. The customer demand standard is obtained based on the statistics of the demand order data entered by the customer, and the calculation formula is as follows:

[0074]

[0075]

[0076]

[0077] Therefore, when the warehouse network node searches for an upstream supply warehouse, the newly added safety stock is calculated based on the number of service customers (the number of downstream customer outlets) brought by the current warehouse network node. The server calculates the newly added inventory area brought by the newly added safety stock, that is, the second newly added inventory area, based on the newly added safety stock and the area quantity. Then, the calculated first newly added inventory area and the second newly added inventory area are summed with the inventory area currently used by the candidate upstream supply warehouse to obtain the total required inventory area that needs to be provided on the candidate upstream supply warehouse after the newly added warehouse network node is used as a downstream warehouse. When the total required inventory area does not exceed (is less than or equal to) the inventory area upper limit of the candidate upstream supply warehouse, it is determined that the use of the candidate upstream supply warehouse will not cause the capacity to exceed the limit, and the candidate upstream supply warehouse can be used as the upstream supply warehouse of the warehouse network node. On the contrary, when the total required inventory area exceeds the inventory area upper limit of the candidate upstream supply warehouse, it is determined that the newly added area will cause the capacity of the candidate upstream supply warehouse to exceed the limit, and the server determines that the candidate upstream supply warehouse cannot be used as the upstream supply warehouse of the warehouse network node.

[0078] In this embodiment, when optimizing and reconstructing the warehouse network structure, the warehouse inventory capacity limit is used as a construction constraint to avoid excessive concentration of inventory in some warehouses. In addition, when calculating whether the inventory area exceeds the limit, a universal calculation method is provided for the input area data of different dimensions, which also improves the executability of the constructed warehouse network.

[0079] In one embodiment, step S208 includes the following steps:

[0080] Step 1: Determine the inventory cost calculation dimensions based on the customer requirements corresponding to the transport order.

[0081] Step 2: Based on the inventory cost calculation dimension, obtain the target inventory occupancy area, unit cargo cost, turnover inventory, safety inventory, inventory capital occupancy coefficient, inventory loss discount coefficient, outlet operation price and outlet operation labor price of the products in each candidate warehouse in the warehouse network structure to be disturbed and the updated warehouse network structure.

[0082] Among them, the calculation dimensions are divided into three different calculation dimensions: piece quantity, weight and volume according to the customer's inventory needs. When the customer's demand is the piece quantity dimension, the target inventory area, turnover inventory, safety inventory, outlet operation price and outlet operation labor price obtained are the data of the piece quantity dimension. When the customer's demand is the weight dimension, the target inventory area, turnover inventory, safety inventory, outlet operation price and outlet operation labor price obtained are the data of the weight dimension. When the customer's demand is the volume dimension, the target inventory area, turnover inventory, safety inventory, outlet operation price and outlet operation labor price obtained are the data of the volume dimension.

[0083] Step 3: Predict the inventory rental cost of each candidate warehouse in the warehouse network structure to be disturbed and the updated warehouse network structure according to the target inventory occupancy area of ​​the product, and obtain the inventory rental cost of the warehouse network structure to be disturbed and the updated warehouse network structure respectively.

[0084] Specifically, the inventory rental cost is usually calculated according to the target inventory or average inventory, and the target inventory = turnover inventory + safety inventory. Average inventory = turnover inventory / 2 + safety stock. This embodiment defaults to calculation according to the target inventory, that is, the inventory rental cost is calculated based on the target inventory area and the rent per square meter of the warehouse. The target inventory area can be calculated based on any one of the three dimensions of piece quantity, weight and volume, including target inventory piece quantity / number of goods stored per unit area, or target inventory weight / weight stored per unit area, or target inventory volume / volume stored per unit area. Based on this, the inventory rental cost is calculated according to three different calculation dimensions, and the corresponding calculation formula is as follows:

[0085] Daily inventory rental cost = warehouse rent per square meter per day * target inventory quantity / number of goods stored per unit area; or, daily inventory rental cost = warehouse rent per square meter per day * target inventory weight / storage weight per unit area; or, daily inventory rental cost = warehouse rent per square meter per day * target inventory volume / storage volume per unit area.

[0086] Step 4: According to the product's turnover inventory, safety inventory, unit goods cost and inventory capital occupancy coefficient, the inventory capital occupancy cost of each candidate warehouse in the warehouse network structure to be disturbed and the updated warehouse network structure is predicted respectively, and the inventory capital occupancy cost of the warehouse network structure to be disturbed and the updated warehouse network structure is obtained respectively.

[0087] Specifically, daily inventory capital utilization cost = average inventory * unit cost of goods * inventory capital utilization coefficient / 365. Among them, average inventory = turnover inventory / 2 + safety inventory. The inventory capital utilization coefficient is entered by the user and can be calculated by default using the average annualized investment return or bank regular interest. In terms of goods cost, the unit cost of goods of different products is different. When the unit cost of goods of a product is unknown, it can also be changed to the unit sales amount of the product.

[0088] Step 5: According to the product's turnover inventory, safety inventory, unit goods cost and inventory loss discount coefficient, the inventory loss discount cost of each candidate warehouse in the warehouse network structure to be disturbed and the updated warehouse network structure is predicted respectively, and the inventory loss discount cost of the warehouse network structure to be disturbed and the updated warehouse network structure is obtained respectively.

[0089] Specifically, daily inventory loss discount cost = average inventory * unit cost of goods * daily inventory loss discount coefficient. Among them, average inventory = turnover inventory / 2 + safety inventory. Since natural loss or price reduction brings greater losses in fresh food, 3C and other industries, inventory loss discount needs to be considered in fresh food, 3C and other industries, and other industries can decide according to actual conditions. Similarly, when the unit cost of goods of a product is unknown, it can also be calculated by the unit sales amount of the product.

[0090] Step 6: According to the product's turnover inventory, safety inventory, outlet operation price and outlet operation labor price, the cargo operation cost of each candidate warehouse in the warehouse network structure to be disturbed and the updated warehouse network structure is predicted respectively, and the cargo operation cost of the warehouse network structure to be disturbed and the updated warehouse network structure is obtained respectively.

[0091] Among them, cargo operating costs include cargo handling costs and warehouse fixed operating costs.

[0092] Specifically, the cargo operation cost can also be divided into two categories: considering the safety stock and not considering the safety stock. When the safety stock is not considered, the cargo operation cost is calculated using the turnover inventory. Cargo operation cost = turnover inventory piece quantity * operation price of a single piece of cargo in the turnover inventory of the outlet + turnover inventory weight * operation price of cargo per kilogram of the turnover inventory of the outlet + turnover inventory volume * operation price of cargo per cubic meter of the turnover inventory of the outlet + target inventory piece quantity * operation labor price of a single piece of cargo in the target inventory of the outlet + target inventory weight * operation labor price of cargo per kilogram of the target inventory of the outlet + target inventory volume * operation labor price of cargo per cubic meter of the target turnover inventory. When the safety stock is considered, the target inventory = turnover inventory + safety stock is used for calculation. That is, the turnover inventory is replaced by the target inventory for calculation. It should be understood that since the present embodiment allows the input of operation prices in three different dimensions of piece quantity, weight and volume, but in order to avoid repeated calculation of costs, the prices of the three dimensions have and only one dimension greater than 0, and the prices of the other two dimensions are 0. That is, when the piece quantity dimension is used, the operation price per kilogram of goods in the outlet turnover inventory and the operation price per cubic meter of goods in the above formula for calculating the goods operation cost are 0. Similarly, when the weight dimension is used, the operation price per piece of goods in the outlet turnover inventory and the operation price per cubic meter of goods in the outlet turnover inventory are 0. In addition, the prices of different products under different candidate warehouses can be different, which is determined according to the actual situation.

[0093] The fixed operating costs of warehouses include warehouse management costs, personnel costs, water and electricity costs, etc. All costs that are not related to the amount of goods can be recorded as fixed operating costs of warehouses. The fixed operating costs of warehouses for each candidate warehouse can be different according to the actual situation. The monthly fixed operating costs of warehouses for each candidate warehouse are entered by the user at one time. Among them, since the fixed operating costs of warehouses are not related to the inventory of goods, there is no need to distinguish products for calculation, and only the candidate warehouse itself needs to be considered. When calculating the fixed operating costs of warehouses, the user enters detailed cost items and then adds them up as the cost items of each warehouse. Each detailed cost can be entered on a monthly basis. For example, water and electricity costs are fixed expenses for candidate warehouses every month and are included in the fixed operating costs of warehouses. Monthly fixed rent for the site. The fixed rent for the candidate warehouse site does not change regardless of the amount of goods and should be included in the fixed operating costs. The cost of building a new warehouse is evenly distributed to each month and included in the fixed operating costs; labor wages, if the number of warehouse personnel is fixed, the labor wages are also included in the fixed operating costs. If the number of labor is related to the inventory demand area, the labor wages are related to the amount of goods and should be included in the labor operating costs of goods. It can be understood that the total fixed operating cost of the warehouse is the total fixed cost expenditure during the solution calculation period. For example: if the customer enters 1 year of data for calculation, the total fixed operating cost = 12*monthly fixed operating cost.

[0094] Step 7, respectively sum up the inventory rental cost, inventory capital occupation cost, inventory loss discount cost and goods operation cost of the warehouse network structure to be disturbed and the updated warehouse network structure, and obtain the inventory cost of the warehouse network structure to be disturbed and the updated warehouse network structure respectively.

[0095] Specifically, after obtaining the inventory rental cost, inventory capital occupation cost, inventory loss discount cost and goods operation cost of the warehouse network structure to be disturbed and the updated warehouse network structure, the inventory rental cost, inventory capital occupation cost, inventory loss discount cost and goods operation cost are summed to obtain the inventory cost. Inventory cost = inventory rental cost + inventory capital occupation cost + inventory loss discount cost + goods operation cost.

[0096] In this embodiment, the method optimizes the warehouse network based on the inventory cost of the candidate warehouse, improves the scope of applicable scenarios, and thus improves the accuracy of optimization. In addition, according to the inventory cost calculation method of different calculation dimensions provided by this embodiment, a unified calculation interface is provided for the calculation of complex inventory costs in real scenarios, further improving the scope of application of each scenario and improving the accuracy of optimization.

[0097] In one embodiment, Figure 5 As shown, another inventory-based warehouse network optimization method is provided, and after step S210, the following steps are further included:

[0098] Step S212: when the quality of the updated warehouse-network structure is better than that of the warehouse-network structure to be disturbed, the updated warehouse-network structure is used as the new warehouse-network structure to be disturbed.

[0099] Step S214: when the quality of the updated warehouse-network structure is inferior to that of the warehouse-network structure to be disturbed, a new warehouse-network structure to be disturbed is determined from the updated warehouse-network structure and the warehouse-network structure to be disturbed according to the acceptance probability of the updated warehouse-network structure.

[0100] Specifically, when the server compares the inventory costs of the warehouse-network structure to be disturbed and the updated warehouse-network structure and determines the best warehouse-network structure, the server again determines a new warehouse-network structure to be disturbed from the warehouse-network structure to be disturbed and the updated warehouse-network structure according to the comparison result. When the inventory cost of the updated warehouse-network structure is lower than the inventory cost of the warehouse-network structure to be disturbed, that is, the quality of the updated warehouse-network structure is better than the inventory cost of the warehouse-network structure to be disturbed, the updated warehouse-network structure is used as the new warehouse-network structure to be disturbed. When the inventory cost of the updated warehouse-network structure is lower than the inventory cost of the warehouse-network structure to be disturbed, that is, the quality of the updated warehouse-network structure is worse than the inventory cost of the warehouse-network structure to be disturbed, the server can use the warehouse-network structure to be disturbed as the new structure to be disturbed, or adopt a certain strategy to select any one of the warehouse-network structures from the updated warehouse-network structure and the warehouse-network structure to be disturbed as the new warehouse-network structure to be disturbed.

[0101] In one embodiment, the server evaluates and determines whether to accept the updated warehouse network structure as the new warehouse network structure to be disturbed based on the idea of ​​simulated annealing algorithm. The simulated annealing algorithm is derived from the solid annealing principle in the field of metallurgy and is a probability-based algorithm. The calculation formula for the probability of being accepted is as follows:

[0102]

[0103] T t =u*T t-1

[0104] Among them, Pro is the probability of acceptance, T is a variable, t represents the number of iterations, and u is the rate of decrease of T value. As the iteration proceeds, the value of T gradually decreases, and this process simulates the annealing process in the metallurgical industry. That is, as the iteration proceeds, the probability of acceptance of the warehouse network structure with poor quality gradually decreases. The rate of decrease of T value u can be adjusted according to the actual problem characteristics. For example, the rate of decrease u every 100 iterations can be assigned to 0.95. That is, T t =0.95*T t-1 . A and B are evaluation values ​​set according to the pursuit of actual problem scenarios. A represents the evaluation value of the warehouse network structure to be disturbed, and B represents the evaluation value of the updated warehouse network structure. In this embodiment, since the optimal warehouse network structure is selected based on inventory cost, A is the cost evaluation value of the warehouse network structure to be disturbed, and B is the evaluation value of the updated warehouse network structure. If the selected optimal warehouse network structure is more concerned with timeliness, A can be the actual transportation timeliness (or transportation timeliness achievement rate) of the warehouse network structure to be disturbed, and B can be the actual transportation timeliness (or transportation timeliness achievement rate) of the updated warehouse network structure. When the acceptance probability of the updated warehouse network structure is calculated, a probability random number is generated to compare with the acceptance probability to determine whether to accept the updated warehouse network structure as the new disturbed warehouse network structure. When the generated probability random number falls within the accepted probability range, the updated warehouse network structure is determined to be the new warehouse network structure to be disturbed. When the probability random number does not fall within the accepted probability range, the warehouse network structure to be disturbed is determined to be the new warehouse network structure to be disturbed. In addition, the acceptable warehouse network structure can be restricted based on the taboo search algorithm.

[0105] In this embodiment, by adopting certain strategies, such as simulated annealing algorithm and taboo search algorithm to determine whether to accept the updated warehouse network structure with poor quality as the solution for the next disturbance update, it is possible to avoid iterative disturbance update falling into local optimum.

[0106] Step S216, when the quality of the currently determined optimal warehouse network structure is better than the existing optimal warehouse network structure, the currently determined optimal warehouse network structure is updated to the existing optimal warehouse network structure, and the number of disturbance updates is reset to zero, and the process returns to step S204 until the inventory cost of the currently determined optimal warehouse network structure is higher than that of the existing optimal warehouse network structure and the number of disturbance updates meets the preset requirements, and the existing optimal warehouse network structure is output.

[0107] Specifically, after the server determines the new warehouse network structure to be disturbed, the quality of the currently determined optimal warehouse network structure and the existing optimal warehouse network structure is evaluated according to the inventory cost. When the inventory cost of the currently determined optimal warehouse network structure is lower than the existing optimal warehouse network structure, it means that the quality of the currently determined optimal warehouse network structure is better than the existing optimal warehouse network structure. The server updates the currently determined optimal warehouse network structure as the existing optimal warehouse network structure, and then starts a new round of disturbance iteration, that is, returns to step S204. The quality of the optimal warehouse network structure determined for the first time in the next round is compared with the best warehouse network structure determined for the last time in this round. At the same time, since a new round of iteration is started, the server needs to reset the number of disturbance updates to zero at the same time and then return to the disturbance update step S204, and start a new round of iteration to find a warehouse network structure with better quality than the currently determined optimal warehouse network structure. It should be understood that if the current optimal warehouse network structure is the best warehouse network structure determined for the first time, there is no optimal warehouse network structure before the current optimal warehouse network structure, that is, the existing optimal warehouse network structure is an empty solution. Then the server directly uses the current optimal warehouse-network structure as the existing optimal warehouse-network structure and returns to the disturbance update step S204, which is equivalent to starting the first iteration of this round.

[0108] Step S218, when the quality of the currently determined optimal warehouse network structure is inferior to the existing optimal warehouse network structure and the number of disturbance updates does not meet the preset requirements, the number of disturbance updates increases by 1, and returns to step S204 until the inventory cost of the currently determined optimal warehouse network structure is higher than the existing optimal warehouse network structure and the number of disturbance updates meets the preset requirements, and the existing optimal warehouse network structure is output.

[0109] Specifically, when the inventory cost of the currently determined optimal warehouse network structure is higher than the existing optimal warehouse network structure and the number of disturbance updates does not meet the preset requirements, it means that the quality of the currently determined optimal warehouse network structure is inferior to the existing optimal warehouse network structure, and the server further determines whether the current number of disturbance updates has reached the preset number of iterations set for one round. When the number of disturbance updates reaches the preset number of iterations, it means that the iteration termination condition configured for one round of iterations is met, and the existing optimal warehouse network structure can be directly output. Otherwise, when the number of disturbance updates does not reach the preset number of times, it means that although a warehouse network structure with better quality than the existing optimal warehouse network structure has not been found, the iteration of this round has not met the configured iteration termination condition, then it is still directly returned to step S204, and the next disturbance update is continued in this round, without starting a new round of iteration. It should be understood that when returning to step S204 for disturbance update or iterative disturbance update of this round, the number of disturbance updates should be incremented by 1.

[0110] It can be seen that the iteration termination condition set for the iteration of this embodiment is that the iteration is stopped only when the preset number of iterations set for a round has been met, but a solution better than the existing optimal solution has not been found. If a better solution has not been found but the current round of iteration has not yet ended, a new iteration of this round will be started. As long as a solution better than the existing optimal solution is found, the number of iterations is reset to zero and a new round of iterations begins. Compared with the traditional method of directly setting a fixed number of iteration conditions, the iteration termination conditions in this embodiment are set in rounds, which can not only find an approximate optimal solution at an appropriate time, but also further improve the quality of the solution, and ensure that a high-quality warehouse network structure is output as much as possible.

[0111] In another embodiment, the quality of the warehouse network structure to be disturbed and the updated warehouse network structure can be evaluated in stages to determine the optimal warehouse network structure. Step S210 includes: performing multi-level quality evaluations on the warehouse network structure to be disturbed and the updated warehouse network structure according to the warehouse constraints, transportation requirements, and inventory costs of the transportation order to determine the optimal warehouse network structure, including the following steps:

[0112] Step S2101, performing quality assessment on the updated warehouse network structure and the warehouse network structure to be disturbed respectively according to warehouse constraints, and obtaining constraint assessment values ​​of the updated warehouse network structure and the warehouse network structure to be disturbed respectively.

[0113] Among them, the quality evaluation of the warehouse network structure according to the warehouse constraint condition refers to evaluating whether the warehouse network structure satisfies the warehouse constraint condition. Since the warehouse constraint condition may contain constraints on the number of warehouses in multiple different hierarchical states, there may also be multiple constraint evaluation values. For example, the warehouse constraint condition includes constraints on the number of central warehouses and constraints on the number of regional warehouses. Therefore, the constraint evaluation value includes the central warehouse constraint evaluation value and the regional warehouse constraint evaluation value.

[0114] Specifically, in this embodiment, the constraint evaluation value is represented by a 01 variable. Assume that X CDC represents the constraint evaluation value of the central warehouse, with X RDC Represents the regional warehouse constraint evaluation value. When the number of central warehouses (regional warehouses) in the warehouse network structure meets the constraints on the number of central warehouses (regional warehouses) in the constraints, for example, when the number of central warehouses (regional warehouses) is 2 and the warehouse network structure includes exactly 2 central warehouses (regional warehouses), X CDC (X RDC ) assigns the variable 1, otherwise assigns the variable 0. CDC and X RDC When both are 1, it means that the warehouse network structure fully satisfies the warehouse constraints. CDC or X RDC When it is 0, it means that the warehouse network structure does not fully meet the warehouse constraints. CDC or X RDC The value formula of is as follows:

[0115]

[0116]

[0117] Among them, NUM C Indicates the specified central warehouse quantity constraint, NUM R Indicates the specified regional warehouse quantity constraint, NUM CDC Indicates the number of central warehouses actually used in the warehouse network structure, NUM RDC Indicates the number of regional warehouses actually used in the warehouse network structure. When the number of central warehouses (regional warehouses) actually used is equal to the constraint of the specified number of central warehouses (regional warehouses), the constraint evaluation value (X CDC and X RDC ) assigns the variable 1, otherwise assigns the variable 0. When comparing the updated warehouse network structure with the initial warehouse network structure using the constraint evaluation value, the X of the two warehouse network structures are compared. CDC and X RDC The assignment variables are used to determine the warehouse network structure with high constraint quality.

[0118] Step S2102: The transportation demand includes transportation time efficiency and transportation cost. The transportation time efficiency of the updated warehouse network structure and the warehouse network structure to be disturbed are determined respectively, and the time efficiency evaluation values ​​of the updated warehouse network structure and the warehouse network structure to be disturbed are obtained respectively; the operating costs of the updated warehouse network structure and the warehouse network structure to be disturbed are determined respectively, and the transportation cost evaluation values ​​of the updated warehouse network structure and the warehouse network structure to be disturbed are obtained respectively.

[0119] Among them, the transportation time efficiency of the warehouse network structure refers to the time required for transportation and distribution of the transportation order based on the supply chain in the warehouse network structure. The transportation time efficiency includes the transportation time efficiency achievement rate, which is the ratio of the number of orders completed within the upper limit of the transportation time efficiency requirement to the total number of orders. The transportation cost of the warehouse network structure refers to the transportation cost incurred when the transportation order is delivered based on the supply chain in the warehouse network structure.

[0120] Specifically, when the server determines the transportation time of updating the warehouse network structure and the warehouse network structure to be disturbed, it calculates the transportation time actually required to complete the transportation order with the updated warehouse network structure or the warehouse network structure to be disturbed. Figure 3a Taking the warehouse network structure shown in the figure as an example, when the transport order belongs to the first customer's order, the transport warehouse corresponding to the first customer's order can be determined as the first central warehouse and the second regional warehouse according to the warehouse network structure. That is, the server calculates the actual transport time required for transportation based on the distance between the first central warehouse and the second regional warehouse, the distance between the second regional warehouse and the customer, and the speed of the transportation vehicle used for transportation. Then, the number of orders completed within the specified upper limit of the transport time requirement is compared with the total number of orders to obtain the transport time achievement rate, thereby obtaining the time evaluation value.

[0121] In this embodiment, the timeliness evaluation value can be directly represented by the transportation timeliness achievement rate. For example, if the transportation timeliness achievement rate obtained by calculating the ratio is 90%, the timeliness evaluation value is 90%. By directly comparing the timeliness evaluation values ​​of the updated warehouse network structure and the warehouse network structure to be disturbed, the warehouse network structure with better timeliness quality is determined. Alternatively, the timeliness evaluation value can be represented by assigning a variable 0 or 1. The timeliness evaluation value assignment formula is as follows:

[0122]

[0123] Among them, X SLA is the timeliness evaluation value, SLA SATISFIED It indicates the delivery time efficiency rate (i.e. the number of orders completed within the delivery time efficiency limit divided by the total number of orders), and SLA indicates the specified delivery time efficiency limit. For example, if the specified delivery time efficiency limit SLA is 24 hours, if more than 90% of the orders are delivered within 24 hours, then SLA SATISFIED >90%, then X SLA Assign a value of 1. Otherwise, when the actual delivery time of more than 10% of the orders is greater than 24 hours, it means that the delivery time achievement rate is less than 90%, so X SLA The value is assigned to 0. The warehouse network structure with better timeliness quality is determined by directly comparing the variable values ​​of the timeliness evaluation values ​​of the updated warehouse network structure and the warehouse network structure to be disturbed.

[0124] When the server determines the transportation cost of updating the warehouse network structure and the warehouse network structure to be disturbed, the transportation cost generated when completing the transportation order with the updated warehouse network structure or the warehouse network structure to be disturbed is calculated. The transportation cost is the transportation cost evaluation value of the updated warehouse network structure or the warehouse network structure to be disturbed. When comparing two warehouse network structures by the transportation cost evaluation value, the larger the transportation cost evaluation value, the higher the cost generated by the warehouse network structure and the worse the cost quality. Among them, the transportation cost server can calculate it through the transportation pricing mode. Different transportation pricing modes use different parameters to calculate the transportation cost, and the calculated transportation cost is different. The transportation pricing mode server selected for each transportation order is determined according to the transportation mode specified by the user. The transportation pricing mode includes but is not limited to weight distance pricing mode, special vehicle transportation mode, and logistics product distribution transportation mode. When the transportation mode specified by the transportation order is physical express delivery, the corresponding transportation pricing mode is the pricing formula of the logistics product distribution transportation mode.

[0125] In one embodiment, determining the transportation cost includes: obtaining the transportation mode of each supply stage of the transportation order; determining the transportation pricing mode of each supply stage according to the transportation mode; calculating the transportation cost of each supply stage according to each transportation pricing mode; and summing the transportation cost of each supply stage to obtain the transportation cost of the warehouse network structure.

[0126] Specifically, during the transportation process, different transportation modes may be used in different supply stages, and thus the transportation pricing modes used in each supply stage are also different. For example, if the current stage is from the regional warehouse to the customer, and the user specifies the "logistics express" transportation mode, then the transportation cost of the supply stage should be calculated according to the transportation pricing mode corresponding to the "logistics express" transportation mode. Alternatively, if the current stage is from the central warehouse to the regional warehouse, and the user specifies the "whole truck transportation" mode for transportation, then the transportation cost of the supply stage should be calculated using the transportation pricing mode corresponding to the "whole truck transportation" mode. Therefore, when calculating the transportation cost, the transportation mode specified by the user for each supply stage of the transportation order can be obtained. The transportation pricing mode that should be used for calculating the transportation cost of each supply stage is determined according to the transportation mode. Then, the transportation cost of the supply stage is calculated according to the pricing parameters corresponding to each transportation pricing mode, and the sum of the transportation costs of each supply stage is the total transportation cost. The transportation cost calculated according to each supply stage is added to the operating cost to obtain the cost evaluation value.

[0127] In this embodiment, the transportation pricing modes corresponding to various transportation modes include but are not limited to "weight distance pricing mode", "special vehicle transportation mode" and "logistics product distribution transportation mode".

[0128] The transportation cost of the "weight-distance pricing model" is cost = weight (transportation weight) * distance (transportation distance) * price (transportation price per unit weight per unit distance). The transportation weight refers to the weight of the product being transported. The transportation distance is the distance between the demand points involved in the supply stage, such as the distance from the central warehouse to the regional warehouse, and the distance from the regional warehouse to the customer. The demand points involved in the supply stage are directly determined based on the supply chain of the customer corresponding to the transportation order in the warehouse network structure.

[0129] The transportation cost of "special car transportation method" cost = NV (number of vehicles used) * distance (transportation distance) * price (unit distance cost of each vehicle).

[0130] The transportation cost of "logistics product distribution transportation mode" is cost = basePrice + (weight-baseWeight) * weightPrice. Among them, basePrice represents the first weight price, weight represents the transportation weight (the weight of the transported product), baseWeight represents the first weight, and weightPrice represents the additional weight price.

[0131] In this embodiment, the transportation cost is calculated by considering multiple transportation modes in each supply stage, and the consideration of transportation cost is refined, thereby optimizing the transportation cost of the warehouse network structure and reducing the operating cost of the warehouse network structure.

[0132] Step S2103, based on the constraint evaluation value, the time evaluation value, the transportation cost evaluation value and the inventory cost, the quality of the disturbed warehouse network structure and the updated warehouse network structure is evaluated to determine the optimal warehouse network structure.

[0133] Specifically, since the quality evaluation of this embodiment is a hierarchical evaluation, when the transportation demand includes multiple requirements of different dimensions, the number of levels of the hierarchical evaluation can be determined according to the number of demand dimensions. That is, when the transportation demand includes transportation time requirements, transportation cost requirements, and warehouse constraint requirements, plus inventory costs, a total of four levels of evaluation can be divided, and the optimal warehouse network structure is finally determined based on these three evaluation values ​​and inventory cost comparison. It should be understood that since the hierarchical evaluation is specifically compared step by step according to the level level, that is, after the comparison of the evaluation value of one level is completed, it is compared based on the evaluation value of the next level. Therefore, the multiple evaluation values ​​obtained are of different levels, and the specific level relationship of each evaluation value is determined according to the actual demand configuration. For example, assuming that the actual pursuit of inventory cost can set the level of inventory cost to the highest, the server will give priority to inventory cost for quality comparison. And the actual pursuit of constraint conditions can set the level of constraint evaluation value to the highest, and the server will give priority to constraint evaluation value for quality comparison. Or, if the actual pursuit of transportation time efficiency can set the time efficiency evaluation value to the highest, the server will give priority to time efficiency evaluation value for quality comparison. Therefore, if the level of inventory cost>the level of constraint evaluation value>the level of timeliness evaluation value>the level of transportation cost evaluation value, then when the quality evaluation of the updated warehouse network structure and the warehouse network structure to be disturbed is performed based on the inventory cost, constraint evaluation value, timeliness evaluation value and transportation cost evaluation value to determine the best warehouse network structure, the updated warehouse network structure and the warehouse network structure to be disturbed are first compared according to the inventory cost. When the inventory costs of the updated warehouse network structure and the warehouse network structure to be disturbed are equal, the updated warehouse network structure and the warehouse network structure to be disturbed are compared according to the constraint evaluation value. When the updated warehouse network structure and the warehouse network structure to be disturbed are determined to be equal in quality according to the constraint evaluation value, the updated warehouse network structure and the warehouse network structure to be disturbed are compared according to the timeliness evaluation value. Similarly, when the updated warehouse network structure and the warehouse network structure to be disturbed are determined to be equal in quality according to the timeliness evaluation value, the updated warehouse network structure and the warehouse network structure to be disturbed are compared according to the transportation cost evaluation value.

[0134] When the quality of the updated warehouse network structure is determined to be better than the warehouse network structure to be disturbed based on the comparison of inventory cost, constraint evaluation value, time evaluation value and transportation cost evaluation value, the updated warehouse network structure is determined to be the optimal warehouse network structure. Alternatively, when the quality of the updated warehouse network structure is determined to be worse than the warehouse network structure to be disturbed based on the comparison of inventory cost, constraint evaluation value, time evaluation value and transportation cost evaluation value, the warehouse network structure to be disturbed is determined to be the optimal warehouse network structure.

[0135] The meanings of the various symbols involved in the quality assessment in this embodiment are shown in Table 1 below.

[0136] Table 1, Symbol meaning table

[0137]

[0138]

[0139] In this embodiment, a warehouse network structure with better quality is determined through hierarchical quality evaluation, thereby ensuring that a warehouse network structure with better quality is finally obtained.

[0140] In one embodiment, step S204 includes: obtaining a preset perturbation strategy and the use probability of each perturbation strategy; generating a corresponding probability random number for each perturbation strategy, and selecting a use strategy from each perturbation strategy according to each probability random number; and performing a perturbation update on the hierarchical status of each candidate warehouse in the warehouse network structure to be disturbed and the remaining candidate warehouses according to the use strategy, and determining the updated hierarchical status of each candidate warehouse.

[0141] Specifically, due to the different requirements of different problem scenarios, different perturbation strategies are applicable to different problem scenarios. Therefore, after the perturbation strategy is set, the probability of using the perturbation strategy is configured according to the corresponding scenario, that is, the usage probability of different perturbation strategies in different scenarios is configured. When the server starts to perturb and update the hierarchical state, the preset perturbation strategy and the usage probability of each perturbation strategy are obtained. According to the current problem scenario, the perturbation strategy used for perturbation is selected from each perturbation strategy as the usage strategy. For example, for problem scenarios with a relatively small search space (i.e., the optimization space is not large), a perturbation strategy with a slightly smaller degree of perturbation can be assigned a greater probability of use, thereby increasing the depth of the search. When the search space is relatively large (i.e., the optimization space is relatively large), a perturbation strategy with a larger degree of perturbation can be assigned a greater probability of use, so that the breadth of the search is improved.

[0142] After obtaining the usage probability of the disturbance strategy, a corresponding number of probability random numbers are generated according to the number of disturbance strategies. For example, three disturbance strategies randomly generate three probability random numbers, and each disturbance strategy corresponds to a probability random number. The probability random number means that the range of the random number is consistent with the probability range. Since the probability is usually represented by 0% to 100%, the range of the random number generated in this embodiment is between 0 and 1, that is, the probability random number. After the probability random number is generated, the probability random number is compared with the corresponding usage probability. When the probability random number falls within the usage probability range, it means that the disturbance strategy corresponding to the usage probability needs to be selected as the usage strategy this time. For example, the usage probability of the disturbance strategy is 50%. When the generated probability random number is 0 to 0.5, it means that it falls into the range, otherwise it does not fall into the range. Therefore, since each disturbance strategy has a corresponding probability random number, there may be multiple disturbance strategies selected for a disturbance update. It should be understood that when any disturbance strategy is not selected according to the probability random number, the probability random number can be generated multiple times for selection. Alternatively, directly select the perturbation strategy with the highest probability of use as the use strategy, which can be configured according to the actual situation. In addition, since the perturbation update method based on the basic idea of ​​neighborhood search is a heuristic method, this kind of parameter tuning for the problem always takes time and is inevitable. Therefore, it is impossible to give a specific fixed use probability value. In order to be more efficient, it should be configured accordingly for the actual problem being processed.

[0143] In one embodiment, the disturbance strategy includes a strategy for hierarchical state change. The hierarchical state of each candidate warehouse in the warehouse network structure to be disturbed and the remaining candidate warehouses is disturbed and updated according to the use strategy to determine the updated hierarchical state of each candidate warehouse, including: randomly selecting a preset number of candidate warehouses from the warehouse network structure to be disturbed as state-changed warehouses; performing hierarchical state change on the hierarchical state of each state-changed warehouse to obtain the updated hierarchical state of each state-changed warehouse; and performing hierarchical state change on the hierarchical state of the remaining candidate warehouses according to the warehouse constraint conditions to obtain the updated hierarchical state of the unselected candidate warehouses.

[0144] Specifically, the strategy for changing the hierarchical state is a strategy for modifying the hierarchical state of the candidate warehouses in the current warehouse network structure to be disturbed, and the number of candidate warehouses whose states are modified is determined by a preset number. In this embodiment, the strategy for changing the hierarchical state includes a strategy for modifying the hierarchical state of a candidate warehouse, which is called a Change strategy. That is, the preset number is 1, indicating that the server randomly selects a candidate warehouse as the candidate warehouse to be modified for hierarchical state modification. For example, a "central warehouse" is changed to a "regional warehouse", a "not selected" candidate warehouse is changed to a "central warehouse", or a "regional warehouse" is modified to a "not selected" candidate warehouse, etc. The strategy for changing the hierarchical state can also include a strategy for modifying the hierarchical state of multiple candidate warehouses, which is called a Pillar Change strategy. It can be understood as a strategy for modifying the hierarchical state of a group of candidate warehouses, and the number of candidate warehouses included in the group can be defined by itself, that is, the preset number is greater than 1. For example, the preset number is 3, indicating that a group includes 3 candidate warehouses. Then, the server randomly selects three candidate warehouses from the warehouse network structure to be disturbed and modifies the hierarchical status of these three candidate warehouses. Since the number of candidate warehouses modified by the PillarChange strategy is greater than that of the Change strategy, the degree of disturbance of the PillarChange strategy is greater than that of the Change strategy.

[0145] In addition, after the hierarchical status of a preset number of candidate warehouses in the warehouse network structure to be disturbed is modified, the initial hierarchical status of the remaining candidate warehouses is randomly modified according to the warehouse constraint conditions to ensure that the updated hierarchical status after the disturbance update satisfies the warehouse constraint conditions. For example, after a "central warehouse" is changed to a "regional warehouse", the number of central warehouses and regional warehouses in the warehouse constraint conditions is not satisfied. At this time, there is one more regional warehouse and one less central warehouse. Therefore, a random selection is made from the remaining candidate warehouses whose status is not a central warehouse and its status is changed to a central warehouse. When the candidate warehouse selected to be modified as a central warehouse is a regional warehouse, the warehouse constraint conditions have been met after the modification, and the disturbance is terminated. When the candidate warehouse selected to be modified as a central warehouse is a candidate that is not selected, the status after the modification is one more regional warehouse and one less candidate warehouse that is not selected, then one candidate warehouse is selected from the remaining candidate warehouses whose status is a regional warehouse and its status is changed to a candidate warehouse that is not selected.

[0146] In one embodiment, the disturbance strategy includes a hierarchical state swap strategy. The hierarchical state of each candidate warehouse in the warehouse network structure to be disturbed and the remaining candidate warehouses is disturbed and updated according to the use strategy to determine the updated hierarchical state of each candidate warehouse, including: randomly selecting a preset number of candidate warehouses from the warehouse network structure to be disturbed to form a state hierarchical swap group; swapping the hierarchical states of the candidate warehouses in the two groups of state hierarchical swap groups to obtain the updated hierarchical state of each candidate warehouse.

[0147] Specifically, the strategy for hierarchical state swapping is a strategy for swapping the hierarchical states of candidate warehouses in the current warehouse network structure, and the number of candidate warehouses whose states are swapped is determined by a preset number. In this embodiment, the strategy for hierarchical state swapping includes a strategy for swapping the hierarchical states of any two candidate warehouses, which is called a Swap strategy. That is, the preset number is 2, indicating that the server randomly selects 2 candidate warehouses from the warehouse network structure to be disturbed as candidate warehouses to be swapped for hierarchical state swapping. For example, swap the hierarchical states of a "central warehouse" and a "regional warehouse", or swap the hierarchical states of a "not selected" and a "regional warehouse", and so on. The Swap strategy has a greater degree of disturbance than the Change strategy. The strategy for hierarchical state swapping can also include a strategy for swapping the hierarchical states of two groups of candidate warehouses, which is called a Pillar Swap strategy. It can be understood as a strategy for swapping the hierarchical status of two groups of candidate warehouses. The number of candidate warehouses included in each group can be defined by itself, but the number of candidate warehouses between the two groups is equal, that is, the preset number is greater than 1. For example, the preset number is 3, which means that one group includes 3 candidate warehouses. Then, the server randomly selects 6 candidate warehouses and divides the 6 candidate warehouses into 2 groups, with 3 candidate warehouses in each group. Then the hierarchical status of the candidate warehouses in the two groups is exchanged. Since the disturbance degree of the Pillar Swap strategy is greater than that of the Pillar Change strategy, the Change strategy and the Swap strategy. And because the disturbance strategy is a swap strategy, the warehouse constraints can always be met. In this instance, there is no need to modify the hierarchical status of the unselected candidate warehouses and maintain the original hierarchical status.

[0148] When the search space is small (i.e. the optimization space is not large), a larger probability of use can be assigned to the Change strategy and the Swap strategy to increase the depth of the search. When the search space is large (i.e. the optimization space is large), a larger probability of use can be assigned to the Pillar Change strategy and the Pillar Swap strategy to increase the breadth of the search.

[0149] In this embodiment, by configuring the probability of using different disturbance strategies in different scenarios, the warehouse network structure construction in different scenarios can have better performance, so as to find the optimal warehouse network structure as quickly as possible and improve efficiency.

[0150] It should be understood that although Figure 2 , 4 The steps in the flowchart of -5 are shown in sequence as indicated by the arrows, but these steps are not necessarily executed in the order indicated by the arrows. Unless otherwise specified in this document, there is no strict order restriction for the execution of these steps, and these steps can be executed in other orders. Moreover, Figure 2 , 4 -At least part of the steps in 5 may include multiple steps or multiple stages. These steps or stages are not necessarily executed at the same time, but can be executed at different times. The execution order of these steps or stages is not necessarily sequential, but can be executed in turn or alternately with other steps or at least part of the steps or stages in other steps.

[0151] In one embodiment, Figure 6 As shown, a warehouse network optimization device based on inventory is provided, including: an acquisition module 602, a disturbance module 604, a construction module 606, a prediction module 608 and an evaluation module 610, wherein:

[0152] The acquisition module 602 is used to acquire the initial warehouse network structure constructed for the transport order as the warehouse network structure to be disturbed.

[0153] The disturbance module 604 is used to disturb and update the hierarchical status of the candidate warehouses and the remaining candidate warehouses in the warehouse network structure to be disturbed, and determine the updated hierarchical status of each candidate warehouse.

[0154] The construction module 606 is used to reconstruct the warehouse network structure according to the inventory capacity limit and update level status of each candidate warehouse to obtain an updated warehouse network structure.

[0155] The prediction module 608 is used to predict the inventory cost of each candidate warehouse in the warehouse network structure to be disturbed and each candidate warehouse in the updated warehouse network structure, respectively, to obtain the inventory cost of the warehouse network structure to be disturbed and the updated warehouse network structure.

[0156] The evaluation module 610 is used to perform quality evaluation on the warehouse network structure to be disturbed and the updated warehouse network structure according to the inventory costs of the warehouse network structure to be disturbed and the updated warehouse network structure, and determine the optimal warehouse network structure.

[0157] In one embodiment, the construction module 608 is also used to select candidate warehouses from each candidate warehouse as warehouse network nodes according to the updated hierarchical status of each candidate warehouse, and the warehouse network nodes constitute the warehouse network structure framework; determine the candidate downstream warehouses of each warehouse network node from the warehouse network structure framework; screen the candidate downstream warehouses that meet the inventory capacity limit of the warehouse network node from the candidate downstream warehouses corresponding to each warehouse network node; determine the downstream warehouses of the warehouse network node from the candidate downstream warehouses that meet the inventory capacity limit of the warehouse network node according to the transportation requirements of the transportation order; establish the supply and replenishment chain of the warehouse network structure framework according to the downstream warehouses of each warehouse network node in the warehouse network structure framework, and obtain an updated warehouse network structure.

[0158] In one embodiment, the construction module 608 is also used to: obtain the area data, demand quantity and the number of customers served by the warehouse network node of the products that the warehouse network node needs the candidate upstream supply warehouse to supply; determine the inventory area occupied by the products required by the warehouse network node based on the area data and demand quantity of the products, and obtain the first newly added inventory area of ​​the candidate upstream supply warehouse; determine the newly added safety inventory of the products required by the warehouse network node based on the number of customers served by the warehouse network node, and determine the second newly added inventory area of ​​the candidate upstream supply warehouse based on the newly added safety inventory and the area data; sum the first newly added inventory area and the second newly added inventory area with the used inventory area of ​​the candidate upstream supply warehouse, and when the sum is less than or equal to the inventory area upper limit of the candidate upstream supply warehouse, determine that the inventory capacity limit of the candidate upstream supply warehouse meets the needs of the warehouse network node.

[0159] In one embodiment, the prediction module 608 is also used to determine the inventory cost calculation dimension according to the customer requirements corresponding to the transportation order; based on the inventory cost calculation dimension, the target inventory occupancy area, unit goods cost, turnover inventory, safety inventory, inventory capital occupancy coefficient, inventory loss discount coefficient, outlet operation price and outlet operation labor price of the product in each candidate warehouse in the warehouse network structure to be disturbed and the updated warehouse network structure are obtained respectively; the inventory rental cost of each candidate warehouse in the warehouse network structure to be disturbed and the updated warehouse network structure is predicted according to the target inventory occupancy area of ​​the product, and the inventory rental cost of the warehouse network structure to be disturbed and the updated warehouse network structure is obtained respectively; the inventory capital occupancy cost of each candidate warehouse in the warehouse network structure to be disturbed and the updated warehouse network structure is predicted according to the turnover inventory, safety inventory, unit goods cost and inventory capital occupancy coefficient of the product, and the inventory capital occupancy cost of each candidate warehouse in the warehouse network structure to be disturbed and the updated warehouse network structure is obtained respectively. The inventory capital occupation cost of the warehouse network structure to be disturbed and the updated warehouse network structure is predicted; according to the product's turnover inventory, safety inventory, unit goods cost and inventory loss discount coefficient, the inventory loss discount cost of each candidate warehouse in the warehouse network structure to be disturbed and the updated warehouse network structure is predicted respectively, and the inventory loss discount cost of the warehouse network structure to be disturbed and the updated warehouse network structure is obtained respectively; according to the product's turnover inventory, safety inventory, outlet operation price and outlet operation labor price, the goods operation cost of each candidate warehouse in the warehouse network structure to be disturbed and the updated warehouse network structure is predicted respectively, and the goods operation cost of the warehouse network structure to be disturbed and the updated warehouse network structure is obtained respectively; the inventory rental cost, inventory capital occupation cost, inventory loss discount cost and goods operation cost of the warehouse network structure to be disturbed and the updated warehouse network structure are summed up respectively, and the inventory cost of the warehouse network structure to be disturbed and the updated warehouse network structure is obtained respectively.

[0160] In one embodiment, it also includes an iteration module, which is used to use the updated warehouse-network structure as the new warehouse-network structure to be disturbed and the optimal warehouse-network structure when the inventory cost of the updated warehouse-network structure is lower than the warehouse-network structure to be disturbed; when the inventory cost of the updated warehouse-network structure is higher than the warehouse-network structure to be disturbed, use the updated warehouse-network structure as the new warehouse-network structure to be disturbed, and use the warehouse-network structure to be disturbed as the optimal warehouse-network structure; when the inventory cost of the currently determined optimal warehouse-network structure is lower than the inventory cost of the existing optimal warehouse-network structure, update the currently determined optimal warehouse-network structure to the existing optimal warehouse-network structure, reset the number of disturbance updates to zero, and return to the disturbance update step until the inventory cost of the currently determined optimal warehouse-network structure is higher than the existing optimal warehouse-network structure and the number of disturbance updates meets the preset requirements, and output the existing optimal warehouse-network structure.

[0161] In one embodiment, the iteration module is also used to return to the disturbance update step when the inventory cost of the currently determined optimal warehouse network structure is higher than the existing optimal warehouse network structure and the number of disturbance updates does not meet the preset requirements, and output the existing optimal warehouse network structure until the inventory cost of the currently determined optimal warehouse network structure is higher than the existing optimal warehouse network structure and the number of disturbance updates meets the preset requirements.

[0162] In one embodiment, the iteration module is also used to determine the acceptance probability of the updated warehouse-network structure and generate a probability random number according to the simulated annealing algorithm; when the probability random number falls within the accepted probability range, the updated warehouse-network structure is determined to be a new warehouse-network structure to be disturbed; when the probability random number does not fall within the accepted probability range, the warehouse-network structure to be disturbed is determined to be a new warehouse-network structure to be disturbed.

[0163] For the specific definition of the inventory-based warehouse network optimization device, please refer to the definition of the inventory-based warehouse network optimization method above, which will not be repeated here. Each module in the above-mentioned inventory-based warehouse network optimization device can be implemented in whole or in part through software, hardware and a combination thereof. The above-mentioned modules can be embedded in or independent of the processor in the computer device in the form of hardware, or can be stored in the memory of the computer device in the form of software, so that the processor can call and execute the operations corresponding to the above modules.

[0164] In one embodiment, a computer device is provided. The computer device may be a server, and its internal structure diagram may be as follows: Figure 7 As shown. The computer device includes a processor, a memory and a network interface connected through a system bus. Among them, the processor of the computer device is used to provide computing and control capabilities. The memory of the computer device includes a non-volatile storage medium and an internal memory. The non-volatile storage medium stores an operating system, a computer program and a database. The internal memory provides an environment for the operation of the operating system and the computer program in the non-volatile storage medium. The database of the computer device is used to store warehouse constraints, transportation requirements and other related data. The network interface of the computer device is used to communicate with an external terminal through a network connection. When the computer program is executed by the processor, a warehouse network optimization method based on inventory is implemented.

[0165] Those skilled in the art will understand that Figure 7 The structure shown in the figure is only a block diagram of a part of the structure related to the solution of the present application, and does not constitute a limitation on the computer device to which the solution of the present application is applied. The specific computer device may include more or fewer components than those shown in the figure, or combine certain components, or have a different arrangement of components.

[0166] In one embodiment, a computer device is further provided, including a memory and a processor, wherein a computer program is stored in the memory, and the processor implements the steps in the above method embodiments when executing the computer program.

[0167] In one embodiment, a computer-readable storage medium is provided, on which a computer program is stored. When the computer program is executed by a processor, the steps in the above-mentioned method embodiments are implemented.

[0168] Those of ordinary skill in the art can understand that all or part of the processes in the above-mentioned embodiment methods can be completed by instructing the relevant hardware through a computer program, and the computer program can be stored in a non-volatile computer-readable storage medium. When the computer program is executed, it can include the processes of the embodiments of the above-mentioned methods. Among them, any reference to memory, storage, database or other media used in the embodiments provided in this application can include at least one of non-volatile and volatile memory. Non-volatile memory can include read-only memory (ROM), magnetic tape, floppy disk, flash memory or optical memory, etc. Volatile memory can include random access memory (RAM) or external cache memory. As an illustration and not limitation, RAM can be in various forms, such as static random access memory (SRAM) or dynamic random access memory (DRAM).

[0169] The technical features of the above embodiments may be combined arbitrarily. To make the description concise, not all possible combinations of the technical features in the above embodiments are described. However, as long as there is no contradiction in the combination of these technical features, they should be considered to be within the scope of this specification.

[0170] The above-mentioned embodiments only express several implementation methods of the present application, and the descriptions thereof are relatively specific and detailed, but they cannot be understood as limiting the scope of the invention patent. It should be pointed out that, for a person of ordinary skill in the art, several variations and improvements can be made without departing from the concept of the present application, and these all belong to the protection scope of the present application. Therefore, the protection scope of the patent of the present application shall be subject to the attached claims.

Claims

1. A warehouse network optimization method based on inventory, characterized in that: The method comprises: Obtaining the initial warehouse network structure constructed for the transport order as the warehouse network structure to be disturbed; Disturb and update the hierarchical status of the candidate warehouses in the warehouse network structure to be disturbed and the remaining candidate warehouses, and determine the updated hierarchical status of each of the candidate warehouses; Reconstructing the warehouse network structure according to the inventory capacity limit of each candidate warehouse and the updated hierarchical status to obtain an updated warehouse network structure; Respectively predicting the inventory costs of each candidate warehouse in the warehouse network structure to be disturbed and each candidate warehouse in the updated warehouse network structure to obtain the inventory costs of the warehouse network structure to be disturbed and the updated warehouse network structure; According to the warehouse constraints, transportation demands and inventory costs of the transportation order, a multi-level quality evaluation is performed on the warehouse network structure to be disturbed and the updated warehouse network structure to determine the optimal warehouse network structure, wherein the warehouse constraints, transportation demands and inventory costs are different levels of quality evaluation, and the multi-level quality evaluation is a step-by-step comparison of evaluation values ​​according to the hierarchical levels.

2. The method according to claim 1, characterized in that The reconstructing the warehouse network structure according to the inventory capacity limit of each candidate warehouse and the updated hierarchical state to obtain the updated warehouse network structure includes: According to the update level status of each candidate warehouse, a candidate warehouse is selected from each candidate warehouse as a warehouse network node, and the warehouse network nodes constitute a warehouse network structure framework; Determine the candidate upstream supply warehouses of each of the warehouse network nodes from the warehouse network structure framework; Selecting candidate upstream supply warehouses whose inventory capacity limits meet the requirements of the warehouse network nodes from the candidate upstream supply warehouses corresponding to each warehouse network node; According to the transportation demand of the transportation order, determine the upstream supply warehouse of the warehouse network node from the candidate upstream supply warehouses whose inventory capacity limit satisfies the warehouse network node; According to the upstream supply warehouse of each of the warehouse network nodes in the warehouse network structure framework, a supply and replenishment chain of the warehouse network structure framework is established to obtain an updated warehouse network structure.

3. The method according to claim 2, characterized in that The step of selecting, from the candidate upstream supply warehouses corresponding to each of the warehouse network nodes, candidate upstream supply warehouses whose inventory capacity limits meet the requirements of the warehouse network nodes, includes: Obtaining the area data, demand quantity, and number of service customers of the product that the warehouse network node needs the candidate upstream supply warehouse to supply; Determine the inventory area occupied by the product required by the warehouse network node according to the area data and the demand quantity of the product, and obtain the first newly added inventory area of ​​the candidate upstream supply warehouse; Determine the newly added safety inventory of the product required by the warehouse network node according to the number of service customers of the warehouse network node, and determine the second newly added inventory area of ​​the candidate upstream supply warehouse according to the newly added safety inventory and the area data; The first newly added inventory area and the second newly added inventory area are summed with the used inventory area of ​​the candidate upstream supply warehouse. When the sum is less than or equal to the upper limit of the inventory area of ​​the candidate upstream supply warehouse, it is determined that the inventory capacity limit of the candidate upstream supply warehouse meets the needs of the warehouse network node.

4. The method according to claim 1, characterized in that: The respectively predicting the inventory cost of each candidate warehouse in the warehouse network structure to be disturbed and each candidate warehouse in the updated warehouse network structure to obtain the inventory cost of the warehouse network structure to be disturbed and the updated warehouse network structure includes: Determine inventory cost calculation dimensions based on customer requirements corresponding to the transport order; Based on the inventory cost calculation dimension, the target inventory occupied area, unit goods cost, turnover inventory, safety inventory, inventory capital occupation coefficient, inventory loss discount coefficient, outlet operation price and outlet operation labor price of the products in each candidate warehouse in the warehouse network structure to be disturbed and the updated warehouse network structure are respectively obtained; Predicting the inventory rental costs of each candidate warehouse in the warehouse network structure to be disturbed and the updated warehouse network structure according to the target inventory occupied area of ​​the product, and obtaining the inventory rental costs of the warehouse network structure to be disturbed and the updated warehouse network structure respectively; According to the turnover inventory, the safety inventory, the unit cost of goods and the inventory capital occupation coefficient of the product, the inventory capital occupation cost of each candidate warehouse in the warehouse network structure to be disturbed and the updated warehouse network structure is predicted respectively, and the inventory capital occupation cost of the warehouse network structure to be disturbed and the updated warehouse network structure is obtained respectively; According to the turnover inventory, the safety inventory, the unit cost of goods and the inventory loss discount coefficient of the product, the inventory loss discount cost of each candidate warehouse in the warehouse network structure to be disturbed and the updated warehouse network structure is predicted respectively, and the inventory loss discount cost of the warehouse network structure to be disturbed and the updated warehouse network structure is obtained respectively; According to the turnover inventory, the safety inventory, the outlet operation price and the outlet operation labor price of the product, the cargo operation cost of each candidate warehouse in the warehouse network structure to be disturbed and the updated warehouse network structure is predicted respectively, and the cargo operation cost of the warehouse network structure to be disturbed and the updated warehouse network structure is obtained respectively; The inventory rental cost, inventory capital occupation cost, inventory loss discount cost and goods operation cost of the warehouse network structure to be disturbed and the updated warehouse network structure are summed up respectively to obtain the inventory costs of the warehouse network structure to be disturbed and the updated warehouse network structure respectively.

5. The method according to claim 1, characterized in that After performing multi-level quality evaluation on the warehouse network structure to be disturbed and the updated warehouse network structure according to the warehouse constraints, transportation requirements and inventory costs of the transportation order and determining the optimal warehouse network structure, the method further includes: When the quality of the updated warehouse network structure is better than that of the warehouse network structure to be disturbed, the updated warehouse network structure is used as a new warehouse network structure to be disturbed; When the quality of the updated warehouse network structure is inferior to that of the warehouse network structure to be disturbed, a new warehouse network structure to be disturbed is determined from the updated warehouse network structure and the warehouse network structure to be disturbed according to the acceptance probability of the updated warehouse network structure; When the quality of the currently determined optimal warehouse-network structure is better than the existing optimal warehouse-network structure, the currently determined optimal warehouse-network structure is updated to the existing optimal warehouse-network structure, and the number of disturbance updates is reset to zero, and the disturbance update step is returned until the inventory cost of the currently determined optimal warehouse-network structure is higher than the existing optimal warehouse-network structure and the number of disturbance updates meets the preset requirements, and the existing optimal warehouse-network structure is output.

6. The method according to claim 5, characterized in that The method further comprises: When the quality of the currently determined optimal warehouse network structure is inferior to that of the existing optimal warehouse network structure and the number of disturbance updates does not meet the preset requirements, return to the disturbance update step until the inventory cost of the currently determined optimal warehouse network structure is higher than that of the existing optimal warehouse network structure and the number of disturbance updates meets the preset requirements, and then output the existing optimal warehouse network structure.

7. The method according to claim 5, characterized in that The determining of a new warehouse network structure to be disturbed according to the acceptance probability of the updated warehouse network structure includes: Determine the acceptance probability of the updated warehouse network structure according to the simulated annealing algorithm and generate a probabilistic random number; When the probability random number falls within the accepted probability range, determining the updated warehouse network structure as a new warehouse network structure to be disturbed; When the probability random number does not fall within the accepted probability range, the warehouse-network structure to be disturbed is determined to be a new warehouse-network structure to be disturbed.

8. A warehouse network optimization device based on inventory, characterized in that: The device comprises: An acquisition module, used for acquiring an initial warehouse network structure constructed for a transport order as a warehouse network structure to be disturbed; A disturbance module, used for disturbing and updating the hierarchical status of the candidate warehouses and the remaining candidate warehouses in the warehouse network structure to be disturbed, and determining the updated hierarchical status of each candidate warehouse; A construction module, configured to reconstruct a warehouse network structure according to the inventory capacity limit of each candidate warehouse and the update hierarchical status to obtain an updated warehouse network structure; A prediction module, used to respectively predict the inventory cost of each candidate warehouse in the warehouse network structure to be disturbed and each candidate warehouse in the updated warehouse network structure, to obtain the inventory cost of the warehouse network structure to be disturbed and the updated warehouse network structure; An evaluation module is used to perform multi-level quality evaluation on the warehouse network structure to be disturbed and the updated warehouse network structure according to the warehouse constraints, transportation requirements and inventory costs of the transportation order, so as to determine the optimal warehouse network structure, wherein the warehouse constraints, transportation requirements and inventory costs are different levels of quality evaluation, and the multi-level quality evaluation is a step-by-step comparison of evaluation values ​​according to the hierarchical levels.

9. A computer device comprising a memory and a processor, wherein the memory stores a computer program, wherein: When the processor executes the computer program, the steps of the method according to any one of claims 1 to 7 are implemented.

10. A computer-readable storage medium having a computer program stored thereon, characterized in that: When the computer program is executed by a processor, the steps of the method according to any one of claims 1 to 7 are implemented.

Images