Warehouse division optimization method and training method and device of warehouse division optimization model

Abstract

The invention discloses a binning optimization method and a training method and device of a binning optimization model, and relates to the technical field of computers. A specific embodiment of the warehouse division optimization method comprises the steps of performing clustering processing on merchants according to industry information and commodity piece type information of the merchants to obtain one or more merchant classes; based on the basic information of the warehouses and the sales volume information of the merchants, determining the warehouses corresponding to the merchant classes; in response to the fact that the number of warehouses corresponding to the merchant class exceeds a set threshold value, adjusting warehouses corresponding to at least part of merchants in the merchant class based on commitment timeliness information of the merchants; the accuracy of warehouse division optimization is improved, and the user experience is improved. A specific embodiment of the training method of the warehouse division optimization model comprises the following steps: taking acquired related information of a target merchant and basic information of a target warehouse as model input, taking historical warehouse division information corresponding to the target merchant as model output, and training to obtain the warehouse division optimization model; and the accuracy of model prediction is improved.

Description

Technical Field

[0001] The present invention relates to the field of computer technology, and in particular to a bin optimization method and a training method and device for a bin optimization model. Background Art

[0002] In the field of logistics and warehousing services, there are a large number of merchants and warehouses. The existing warehouse division methods often only consider the information of a single merchant and rely on insufficient constraints, such as only basic constraints such as the use of warehouse capacity not exceeding the upper limit. In addition, the results of warehouse division often have large errors compared with the actual situation.

[0003] In the process of implementing the present invention, the inventors found that there are at least the following problems in the related art:

[0004] The existing warehouse optimization methods use relatively simple data, resulting in inaccurate optimization results and poor user experience. Summary of the invention

[0005] In view of this, the embodiments of the present invention provide a warehouse division optimization method and a training method and device for a warehouse division optimization model, which can comprehensively consider the relevant information of merchants and the basic information of warehouses, aggregate merchants from the dimensions of industry and item type, and then perform multiple warehouse division adjustments on the aggregated merchant categories, thereby improving the accuracy of warehouse division optimization and enhancing the user experience.

[0006] To achieve the above objective, according to one aspect of an embodiment of the present invention, a warehouse optimization method is provided, comprising:

[0007] Obtain relevant information about merchants and basic information about warehouses; the relevant information about merchants includes industry information, product type information, product sales information, and promised timeliness information;

[0008] Based on the merchant's industry information and product type information, the merchants are clustered to obtain one or more merchant classes;

[0009] Based on the basic information of the warehouse and the sales information of the merchant, determine the warehouse corresponding to each merchant category;

[0010] In response to the number of warehouses corresponding to a merchant class exceeding a set threshold, the warehouses corresponding to at least some merchants in the merchant class are adjusted based on the promised timeliness information of the merchants.

[0011] Optionally, based on the basic information of the warehouse and the sales information of the merchant, the warehouse corresponding to each merchant class is determined, including:

[0012] Based on the basic information of the warehouse and the sales information of the merchant, the warehouse corresponding to each merchant category is determined using the pre-built first-stage optimization model;

[0013] The pre-built first-stage optimization model is obtained through the following steps:

[0014] Based on the acquired merchant historical clustering information, warehouse historical basic information and merchant historical sales information, define the first decision variable and the first constraint condition; wherein the decision variable includes the distribution and proportion of merchant types in the warehouse; the constraint condition includes merchant proportion warehouse splitting constraint, warehouse usage area constraint, piece type industry aggregation constraint, special industry aggregation constraint and fulfillment cost constraint;

[0015] Determine a first objective function according to the first decision variable and the first constraint condition; wherein the first objective function includes minimizing the total number of warehouses of the merchant class, minimizing the industry aggregation of the piece type, or minimizing the total cost;

[0016] A first-stage optimization model is constructed based on the first decision variable, the first constraint condition and the first objective function.

[0017] Optionally, adjusting the warehouses corresponding to at least some merchants in the merchant class based on the promised timeliness information of the merchants includes:

[0018] Based on the promised timeliness information of the merchants, the warehouses corresponding to at least some of the merchants in the merchant class are adjusted by using a pre-built second-stage optimization model;

[0019] The pre-built second-stage optimization model is obtained by the following steps:

[0020] Filter out the historical merchant classes whose corresponding warehouse quantity exceeds the set threshold from the acquired merchant historical clustering information, and obtain the historical commitment timeliness information and historical warehouse basic information corresponding to the historical merchant classes;

[0021] Based on the historical commitment time limit information corresponding to the historical merchant class and the historical basic information of the warehouse, define the second decision variable and the second constraint condition; wherein the second decision variable includes the distribution of merchants in the warehouse, and the second constraint condition includes the merchant non-warehouse constraint and the merchant commitment time limit constraint;

[0022] Determine a second objective function according to the second decision variable and the second constraint condition; wherein the second objective function includes minimizing the difference in merchant area occupation and minimizing the merchant's warehouse relocation cost;

[0023] A second-stage optimization model is constructed based on the second decision variable, the second constraint condition and the second objective function.

[0024] Optionally, the method further comprises:

[0025] In response to the number of warehouses corresponding to the merchant class not exceeding the set threshold, all merchants in the merchant class are allocated to the warehouses corresponding to the merchant class.

[0026] According to a second aspect of an embodiment of the present invention, a training method for a warehouse optimization model is provided, comprising:

[0027] Obtain relevant information of the target merchant, basic information of the target warehouse, and historical warehouse information corresponding to the target merchant; wherein the historical warehouse information corresponding to the target merchant is determined by the first-stage optimization model or the second-stage optimization model;

[0028] The relevant information of the target merchant and the basic information of the target warehouse are used as the model input, and the historical warehouse information corresponding to the target merchant is used as the model output, and the warehouse optimization model is obtained through training.

[0029] According to a third aspect of an embodiment of the present invention, a warehouse optimization device is provided, including:

[0030] The information acquisition module is used to obtain the relevant information of the merchant and the basic information of the warehouse; the relevant information of the merchant includes industry information, product type information, product sales information and promised time information;

[0031] The merchant clustering module is used to cluster merchants according to their industry information and product type information to obtain one or more merchant classes;

[0032] The warehouse determination module is used to determine the warehouse corresponding to each merchant class based on the basic information of the warehouse and the sales information of the merchant;

[0033] The re-adjustment module is used to adjust the warehouses corresponding to at least some merchants in the merchant class based on the merchant's promised timeliness information in response to the number of warehouses corresponding to the merchant class exceeding a set threshold.

[0034] According to a fourth aspect of an embodiment of the present invention, a training device for a warehouse optimization model is provided, comprising:

[0035] The target acquisition module is used to obtain the relevant information of the target merchant, the basic information of the target warehouse, and the historical warehouse information corresponding to the target merchant; wherein the historical warehouse information corresponding to the target merchant is determined by the first-stage optimization model or the second-stage optimization model;

[0036] The model training module is used to take the relevant information of the target merchant and the basic information of the target warehouse as the model input, and the historical warehouse information corresponding to the target merchant as the model output, and train to obtain the warehouse optimization model.

[0037] According to a fifth aspect of an embodiment of the present invention, there is provided an electronic device, including:

[0038] one or more processors;

[0039] a storage device for storing one or more programs,

[0040] When one or more programs are executed by one or more processors, the one or more processors implement the method of any of the above embodiments.

[0041] According to a sixth aspect of an embodiment of the present invention, there is provided a computer-readable medium on which a computer program is stored, and when the program is executed by a processor, the method of any of the above embodiments is implemented.

[0042] According to a seventh aspect of an embodiment of the present invention, there is provided a computer program product, including a computer program, which implements the method of any of the above embodiments when the computer program is executed by a processor.

[0043] An embodiment of the above invention has the following advantages or beneficial effects: by obtaining relevant information of merchants and basic information of warehouses; wherein the relevant information of merchants includes industry information, commodity type information, commodity sales information and promised time limit information; according to the industry information and commodity type information of merchants, the merchants are clustered to obtain one or more merchant classes; based on the basic information of warehouses and the sales information of merchants, the warehouses corresponding to each merchant class are determined; in response to the number of warehouses corresponding to a merchant class exceeding a set threshold, the warehouses corresponding to at least some merchants in the merchant class are adjusted based on the merchant's promised time limit information; thereby, the relevant information of merchants and the basic information of warehouses can be comprehensively considered, and merchants can be aggregated from the industry and item type dimensions, and then the aggregated merchant classes can be adjusted for multiple warehouse divisions, thereby improving the accuracy of warehouse division optimization and enhancing the user experience. Obtain relevant information of the target merchant, basic information of the target warehouse, and historical warehouse information corresponding to the target merchant; wherein the historical warehouse information corresponding to the target merchant is determined by the first-stage optimization model or the second-stage optimization model; use the relevant information of the target merchant and the basic information of the target warehouse as model input, and the historical warehouse information corresponding to the target merchant as model output, and train to obtain a warehouse optimization model; thereby, many business factors and constraints can be considered to improve the accuracy of model predictions.

[0044] The further effects of the above-mentioned non-conventional optional manner will be described below in conjunction with the specific implementation manner. BRIEF DESCRIPTION OF THE DRAWINGS

[0045] The accompanying drawings are used to better understand the present invention and do not constitute an improper limitation of the present invention.

[0046] Figure 1 is a schematic diagram of the main process of the warehouse optimization method according to an embodiment of the present invention;

[0047] Figure 2 is a schematic diagram of the main process of the training method of the warehouse optimization model according to an embodiment of the present invention;

[0048] Figure 3It is a schematic diagram of the main process of applying the warehouse optimization method according to a preferred embodiment of the present invention;

[0049] Figure 4 is a schematic diagram of the main process of a warehouse optimization algorithm according to a preferred embodiment of the present invention;

[0050] Figure 5 It is a schematic diagram of the main process of solving the warehouse optimization model according to a preferred embodiment of the present invention;

[0051] Figure 6 is a schematic diagram of main modules of a warehouse optimization device according to an embodiment of the present invention;

[0052] Figure 7 is a schematic diagram of main modules of a training device for a warehouse optimization model according to an embodiment of the present invention;

[0053] Figure 8 is an exemplary system architecture diagram to which embodiments of the present invention may be applied;

[0054] Fig. 9 It is a schematic diagram of the structure of a computer system of a terminal device or a server suitable for implementing an embodiment of the present invention. DETAILED DESCRIPTION

[0055] The following is a description of exemplary embodiments of the present invention in conjunction with the accompanying drawings, including various details of the embodiments of the present invention to facilitate understanding, which should be considered as merely exemplary. Therefore, it should be recognized by those of ordinary skill in the art that various changes and modifications may be made to the embodiments described herein without departing from the scope and spirit of the present invention. Similarly, for clarity and conciseness, the description of well-known functions and structures is omitted in the following description.

[0056] It should be noted that the acquisition, storage and application of personal information etc. involved in the embodiments of the present invention are in compliance with the provisions of relevant laws and regulations and do not violate public order and good morals.

[0057] In the field of logistics and warehousing services, there are many merchants and warehouses. The existing warehouse division methods often only consider single merchant information and rely on insufficient constraints, such as only basic constraints such as the use of warehouse capacity not exceeding the upper limit. In addition, the warehouse division results often have large errors compared with the actual situation. That is, the data used by the existing warehouse division optimization methods is relatively single, resulting in inaccurate optimization results and poor user experience.

[0058] In view of this, according to a first aspect of an embodiment of the present invention, a warehouse optimization method is provided.

[0059] Figure 1 FIG. 1 is a schematic diagram of the main process of the warehouse optimization method according to an embodiment of the present invention. Figure 1 As shown, the warehouse optimization method according to an embodiment of the present invention mainly includes the following steps S101 to S104.

[0060] Step S101, obtaining the merchant's relevant information and the warehouse's basic information; the merchant's relevant information includes industry information, product type information, product sales information and promised time limit information.

[0061] Merchants include entities that provide goods or services in the supply chain or logistics network, and can be sellers, brand owners, wholesalers or other suppliers on e-commerce platforms. Merchant-related information includes industry information, product type information, product sales information and promised time information, etc. This information is used to describe the industry field in which the merchant is located (such as food, clothing, electronic products, etc.), the packaging type or volume specification of the product (such as large household appliances, small electronics, bulk goods, etc.), the historical sales data of the product in a specific period, and the merchant's specific commitment to the order fulfillment time (such as next-day delivery, three-day delivery, etc.). Warehouses are logistics nodes that store merchants' goods and are responsible for warehousing operations, including regional warehouses, central warehouses or forward warehouses, etc. The basic information of the warehouse includes its geographical location, storage capacity, sorting efficiency, distribution range, and the ability to connect with specific logistics companies. This information affects the warehouse distribution strategy and fulfillment efficiency of merchants' goods.

[0062] There are many ways to obtain this information. For example, for merchant-related information, merchants actively upload it through the background system, including registration of industry information and product attributes, definition of commitment time limit, etc.; in addition, product sales information can also be generated through historical order data and sales systems. The basic information of the warehouse can be maintained by the warehouse management system (WMS), combined with real-time updates by the operations team. In addition, it can also be obtained through the interface call of the business system, such as the API interface (Application Programming Interface) of the logistics platform, which directly links the real-time data of the warehouse and the merchant. Through the above methods, the core information of merchants and warehouses can be fully collected in the initial stage, providing comprehensive data support for the subsequent formulation of warehouse distribution strategies.

[0063] Step S102: clustering merchants according to their industry information and product type information to obtain one or more merchant classes.

[0064] Clustering is to classify merchants with similar attributes into the same group by analyzing merchant feature data. Specifically, industry information indicates the business field to which the merchant belongs (such as electronics, clothing, food, etc.), while product type information describes the appearance characteristics and logistics attributes of the product (such as volume, weight, packaging type, etc.). These data are used to explore the commonalities and differences between merchants. Clustering can be implemented in a variety of ways, including rule-based clustering or algorithm-based clustering. Rule-based clustering uses predefined rules, such as directly dividing merchants into different categories according to industry categories or product types. This method is simple and intuitive, and is suitable for scenarios where merchant information dimensions are relatively small or clear classification standards already exist. Algorithm-based clustering uses data-driven methods to group merchants using unsupervised learning models (such as K-Means, hierarchical clustering, etc.). The algorithm divides merchants into multiple clusters based on the similarity of merchant information features by calculating the distance between feature vectors. This method can capture hidden patterns in data and is suitable for scenarios with high information dimensions and complex distribution.

[0065] Merchant class is the output result of clustering processing, which represents a group of merchants with similar characteristics in terms of industry information and product type information. A merchant class can represent a specific business model or logistics demand, such as "merchants selling large furniture" or "merchants operating fresh goods". The division of these merchant classes can provide a more targeted data foundation for the subsequent formulation and optimization of warehouse distribution strategies. For example, different merchant classes may require different warehouse layout solutions to meet their specific delivery timeliness requirements and logistics cost control goals. Through clustering processing, the efficiency of merchant management and storage resource allocation can be effectively improved, while achieving refined supply chain operation management.

[0066] Step S103, based on the basic information of the warehouse and the sales information of the merchant, determine the warehouse corresponding to each merchant category.

[0067] The basic information of the warehouse may include the geographical location, storage capacity, facility conditions, service capabilities, and coverage area of ​​the warehouse. The sales information of the merchant reflects the order volume, sales trend, and regional distribution of each merchant class. The method of determining the corresponding warehouse for the merchant class can be carried out through various methods such as rule matching and model optimization. The rule matching method is based on pre-set rules or experience. For example, the merchant class in a certain area is preferentially assigned to the nearest warehouse, or a preliminary correspondence is established according to the service scope of the warehouse and the main sales area of ​​the merchant. This method is simple and intuitive and is suitable for situations where the scene is relatively clear or the relationship between the warehouse and the merchant's demand is relatively fixed. The model optimization method can use data analysis and optimization algorithms to globally optimize the matching of merchant classes and warehouses by establishing mathematical models. Models include linear programming, transportation optimization models, or prediction models based on machine learning, etc. These models can comprehensively evaluate the matching degree between each merchant class and each warehouse by integrating multi-dimensional factors, such as the distribution cost from the warehouse to the merchant's main sales area, the storage demand corresponding to the merchant's sales volume, and the existing resource utilization rate of the warehouse, so as to determine the best allocation plan. For example, the distribution efficiency of different warehouses in meeting the merchant's needs can be calculated based on sales forecast data, and the optimal warehouse resources can be selected in combination with warehouse capacity and regional coverage.

[0068] The warehouse corresponding to a merchant class refers to one or more warehouses that provide commodity storage and distribution services for a certain merchant class. The corresponding warehouse is determined to optimize the allocation of logistics resources so that the merchant's order needs can be effectively guaranteed in terms of timeliness and cost. For example, merchants with high sales and wide distribution may be assigned to multiple warehouses to disperse inventory pressure and shorten the distribution distance; while small merchants with sales concentrated in a certain area may only need to be assigned to a single warehouse to reduce management complexity. This matching relationship can provide data support for subsequent logistics scheduling and warehouse distribution strategy optimization, while improving the overall supply chain operation efficiency.

[0069] Step S104, in response to the number of warehouses corresponding to a merchant class exceeding a set threshold, the warehouses corresponding to at least some merchants in the merchant class are adjusted based on the merchant's promised timeliness information.

[0070] The set threshold refers to the upper limit of the number of warehouses allocated to a merchant class that is pre-defined in actual business. This threshold can be set based on comprehensive factors such as logistics network planning, management complexity, and cost control. For example, too many warehouse allocations may lead to increased storage and transportation costs, as well as increased management difficulty, so reasonable restrictions need to be set. When the number of warehouses allocated to a certain merchant class exceeds the set threshold, the warehouse allocation can be adjusted based on the merchant's promised timeliness information. The promised timeliness information is the merchant's requirement for order delivery time, which is related to user experience, sales strategy, and logistics timeliness. The goal of the adjustment is to reduce the number of warehouses while meeting the timeliness requirements, thereby achieving efficient use of resources.

[0071] Specifically, the adjustment can be made in a variety of ways. For example, based on the redistribution of timeliness priority, the promised timeliness of merchants in the merchant category is graded, for example, into different levels such as high timeliness (next day delivery), medium timeliness (2-3 days delivery) and low timeliness (ordinary delivery). For merchants with low timeliness requirements, they can be adjusted from multiple warehouses to a central warehouse to simplify management; while for merchants with high timeliness requirements, the warehouse closest to the main distribution area is given priority to ensure delivery speed. It is also possible to optimize the matching relationship between merchants and warehouses by constructing a mathematical model, comprehensively considering factors such as the merchant's promised timeliness, warehouse capacity, transportation cost and service scope. For example, integer programming or allocation models are used to find the minimum warehouse combination that meets the timeliness requirements, and adjust the merchant's corresponding warehouse according to the model results. Through these adjustment methods, a balance can be achieved between the merchant's promised timeliness and the number of warehouse allocations, which not only ensures that the merchant's service level is not affected, but also reduces the logistics cost and complexity of warehousing management. At the same time, the adjustment results provide data support for subsequent logistics scheduling and optimization models, further improving the stability and flexibility of the entire logistics system.

[0072] The embodiment of the present invention obtains relevant information of merchants and basic information of warehouses; wherein the relevant information of merchants includes industry information, commodity type information, commodity sales information and promised time limit information; according to the industry information and commodity type information of merchants, the merchants are clustered to obtain one or more merchant classes; based on the basic information of warehouses and the sales information of merchants, the warehouses corresponding to each merchant class are determined; in response to the number of warehouses corresponding to a merchant class exceeding a set threshold, the warehouses corresponding to at least some merchants in the merchant class are adjusted based on the promised time limit information of the merchants; thereby, the relevant information of merchants and the basic information of warehouses can be comprehensively considered, and merchants are aggregated from the industry and item type dimensions, and then the aggregated merchant classes are subjected to multiple warehouse adjustments, thereby improving the accuracy of warehouse optimization and enhancing the user experience.

[0073] Optionally, based on the basic information of the warehouse and the sales information of the merchant, the warehouse corresponding to each merchant class is determined, including: based on the basic information of the warehouse and the sales information of the merchant, using the pre-constructed first-stage optimization model to determine the warehouse corresponding to each merchant class. The pre-constructed first-stage optimization model is obtained through the following steps: based on the acquired merchant historical clustering information, warehouse historical basic information and merchant historical sales information, define the first decision variable and the first constraint condition; wherein the decision variable includes the distribution and proportion of merchant classes in the warehouse; the constraint conditions include merchant proportion warehouse splitting constraints, warehouse usage area constraints, piece-type industry aggregation constraints, special industry aggregation constraints and fulfillment cost constraints; determine the first objective function according to the first decision variable and the first constraint condition; wherein the first objective function includes minimizing the total number of warehouses of merchant classes, minimizing piece-type industry aggregation or minimizing total cost; construct the first-stage optimization model based on the first decision variable, the first constraint condition and the first objective function.

[0074] Optionally, based on the basic information of the warehouse and the sales information of the merchant, the pre-built first-stage optimization model can be used to determine the warehouse corresponding to each merchant class to achieve a more efficient and reasonable warehouse allocation strategy. The construction of the first-stage optimization model is based on the merchant's historical clustering information, the warehouse's historical basic information and the merchant's historical sales information, and the model is gradually defined and optimized through multiple steps.

[0075] Decision variables are used to describe the distribution and proportion of merchant types in warehouses, and are the core of model solving. For example, decision variables can represent the specific situation of the proportion of orders of a certain merchant type to different warehouses. Constraints are used to limit the feasible range of model solving so that the allocation results meet the actual operational needs. Specifically, constraints include the following key aspects: merchant proportion split warehouse constraints, which are used to control the allocation ratio of merchant orders in multiple warehouses to avoid excessive split warehouses leading to increased costs and management complexity; warehouse usage area constraints, which are used to limit the carrying capacity of a single warehouse to avoid warehouse resource overload; piece-type industry aggregation constraints and special industry aggregation constraints, so that the same or similar commodity piece types and special industry merchants are gathered in the same warehouse as much as possible to improve operational efficiency; fulfillment cost constraints, which are used to control the logistics fulfillment costs caused by allocation decisions to ensure overall economy. The objective function is used to guide the solution direction of the optimization model. For example, the objectives may include minimizing the total number of warehouses of merchant classes, thereby reducing management complexity and resource reuse; minimizing the aggregation of piece-type industries to optimize the operational efficiency of warehouses; or minimizing the total cost to maximize economic benefits by balancing logistics costs, storage costs and timeliness requirements. The choice of objective function should be weighed according to specific business needs to ensure that the optimization results of the model are consistent with the actual operation goals. The first-stage optimization model is constructed based on the first decision variable, the first constraint condition and the first objective function. The model can be constructed using mathematical programming methods, such as linear programming or integer programming. By integrating the above variables, conditions and functions, a complete mathematical expression is formed, and the optimization algorithm is used to solve the model to determine the optimal allocation plan between merchant classes and warehouses. The results of the optimization model can not only achieve the reasonable allocation of warehouse resources, but also provide an important reference for subsequent dynamic adjustment and optimization.

[0076] The embodiment of the present invention can efficiently solve the matching problem between merchant types and warehouses by constructing a first-stage optimization model, make full use of logistics network resources, improve allocation efficiency, and control logistics costs, thereby providing scientific support for logistics management decisions.

[0077] Optionally, adjusting the warehouses corresponding to at least some of the merchants in the merchant class based on the merchant's commitment time information includes: adjusting the warehouses corresponding to at least some of the merchants in the merchant class using a pre-constructed second-stage optimization model based on the merchant's commitment time information. The pre-constructed second-stage optimization model is obtained by the following steps: screening out historical merchant classes whose corresponding warehouse numbers exceed a set threshold from the acquired merchant historical clustering information, and obtaining the historical commitment time information and warehouse historical basic information corresponding to the historical merchant class; defining the second decision variable and the second constraint condition based on the historical commitment time information and warehouse historical basic information corresponding to the historical merchant class; wherein the second decision variable includes the distribution of merchants in the warehouse, and the second constraint condition includes the merchant non-warehouse constraint and the merchant commitment time constraint; determining the second objective function based on the second decision variable and the second constraint condition; wherein the second objective function includes minimizing the difference in merchant area occupancy and minimizing the merchant warehouse relocation cost; constructing the second-stage optimization model based on the second decision variable, the second constraint condition and the second objective function.

[0078] Through the pre-built second-stage optimization model, warehouses of some merchants in the merchant class are adjusted to make the warehouse allocation more consistent with the merchant's time commitment, while reducing the difference in warehouse resource occupancy and warehouse moving costs. From the merchant historical clustering information, historical merchant classes whose number of warehouses exceeds the set threshold are selected. Due to their wide distribution, these merchant classes may face problems such as over-allocation of warehouse resources or difficulty in achieving timeliness. Therefore, for these merchant classes, their historical commitment timeliness information and historical basic warehouse information are obtained. This information will be used to analyze the distribution of merchant classes in different warehouses.

[0079] Based on the historical commitment time information of the merchant class and the historical basic information of the warehouse, the decision variables and constraints of the second-stage optimization model are defined. The second decision variable mainly describes the distribution of merchants in different warehouses, considering how to allocate resources between different warehouses to ensure that the merchant's commitment to timeliness is met. For example, the distribution of merchants in a certain warehouse may need to be adjusted according to the timeliness requirements so that the warehouse can process and deliver orders within the specified time. The constraints of the second-stage optimization model include the merchant non-warehouse constraint and the merchant commitment time constraint. The merchant non-warehouse constraint requires that the distribution of certain merchants between different warehouses should not be too dispersed, that is, to ensure that a merchant operates in the same warehouse as much as possible, thereby avoiding frequent warehouse switching and resource waste. The merchant commitment time constraint ensures that the warehouse allocation strategy can meet the merchant's timeliness commitment. The processing capacity and location of each warehouse will affect the merchant's fulfillment time. Therefore, it is necessary to ensure that the allocation strategy in the model can achieve the merchant's timeliness requirements. After determining the second decision variables and constraints, it is necessary to construct the second objective function. The second objective function minimizes the differences in warehouse space occupied by merchants by optimizing warehouse allocation, reduces unnecessary space waste, and thus improves warehouse utilization. At the same time, in order to minimize the merchant's warehouse relocation costs, while ensuring timeliness, merchants should try to avoid frequent relocation to other warehouses to reduce costs and operational complexity. Based on these second decision variables, second constraints, and second objective functions, a second-stage optimization model is constructed. This model can further optimize the matching relationship between merchant types and warehouses by considering multiple factors such as timeliness, warehouse resource usage, and warehouse relocation costs, thereby effectively improving merchants' fulfillment efficiency and reducing operating costs.

[0080] Optionally, the method further includes: in response to the number of warehouses corresponding to a merchant class not exceeding a set threshold, allocating all merchants in the merchant class to the warehouses corresponding to the merchant class.

[0081] If the number of warehouses corresponding to a merchant class does not exceed the set threshold, direct allocation is adopted to directly allocate all merchants in the merchant class to the warehouses corresponding to the merchant class. Specifically, when the number of warehouses in a merchant class is small and the warehouse resources are sufficient to meet the needs of merchants under the current circumstances, excessive adjustments may bring unnecessary complexity and costs. Therefore, the system chooses to simplify the process and directly allocate all merchants to existing warehouses. It can reduce friction and uncertainty in the allocation process. At the same time, direct allocation can make merchants' inventory and distribution more centralized, thereby improving warehouse resource utilization and operational efficiency.

[0082] According to a second aspect of an embodiment of the present invention, a training method for a warehouse optimization model is provided.

[0083] Figure 2is a schematic diagram of the main process of the training method of the warehouse optimization model according to an embodiment of the present invention; Figure 2 As shown, the training method of the warehouse optimization model according to an embodiment of the present invention mainly includes the following steps S201 to S202.

[0084] Step S201, obtaining relevant information of the target merchant, basic information of the target warehouse and historical warehouse information corresponding to the target merchant; wherein the historical warehouse information corresponding to the target merchant is determined by the first stage optimization model or the second stage optimization model.

[0085] Step S202: Use the relevant information of the target merchant and the basic information of the target warehouse as model input, and the historical warehouse information corresponding to the target merchant as model output, and train to obtain a warehouse optimization model.

[0086] First, obtain the relevant information of the target merchant, the basic information of the target warehouse, and the historical warehouse information corresponding to the target merchant. The relevant information of the target merchant includes the merchant's industry, product type, sales volume and other data, which are crucial to understanding the merchant's operating model and needs. The basic information of the target warehouse includes the warehouse capacity, geographical location, facility conditions and other aspects. The historical warehouse information is obtained through the previous optimization model (such as the first-stage optimization model or the second-stage optimization model), which records the distribution of the target merchant in different warehouses in the past. The relevant information of the target merchant and the basic information of the target warehouse are used as input data, and the historical warehouse information is used as output data to train the warehouse optimization model. The training process of this model learns the relationship between input and output through the algorithm, aiming to optimize the merchant's warehouse strategy, so that the merchant can achieve the best inventory management and logistics efficiency under the condition of limited warehouse resources. During the training process, the model continuously optimizes the warehouse solution by adjusting the model parameters according to historical data and actual needs, thereby improving the matching degree between merchants and warehouses, and finally providing an efficient and low-cost warehouse allocation strategy for the target merchant. The training of the warehouse optimization model is not only a process of mathematical modeling, but also includes a deep understanding of the actual business scenario. Through continuous optimization, the model can gradually improve the accuracy and adaptability of warehouse distribution decisions, thereby ensuring that merchants can better respond to market demands in terms of warehousing and distribution.

[0087] Figure 3 FIG. 1 is a schematic diagram of the main process of applying the warehouse optimization method according to a preferred embodiment of the present invention. Figure 3As shown in the figure, the flowchart describes the merchant contract and demand processing process, which mainly involves the merchant submitting information, system processing and result output. Starting from the business goal, the core purpose of the system operation is clarified. Merchants submit information to the system by uploading relevant contracts and demand documents, including contract-related requirements and specific demand goals. This information will go through two upload steps, corresponding to the basic terms of the contract and the supplementary terms of the contract, and will be automatically parsed and structured by the system. The system organizes and analyzes the uploaded information to generate a contract demand matching plan. The generation process of this plan makes the merchant's needs consistent with the requirements of existing resources and contract terms, and ensures the accuracy and feasibility of subsequent allocation work. On this basis, the system will give the merchant a contract demand matching plan. When the contract demand plan is completed, the system enters the subsequent steps, including the verification and revision of the plan. Merchants can confirm the generated matching plan. If there is a problem with the plan, it will also be fed back, and the system will adjust and optimize the allocation plan. The process outputs three main results, including contract information, merchant demand information and demand matching solutions. These results provide important reference for subsequent specific business links such as warehousing and distribution, and can efficiently complete the merchant contract demand matching task.

[0088] Figure 4 It is a schematic diagram of the main process of the warehouse optimization algorithm according to a preferred embodiment of the present invention. Figure 5 FIG. 1 is a schematic diagram of the main process of solving the warehouse optimization model according to a preferred embodiment of the present invention. Figure 4 and Figure 5 As shown, the flowchart describes the overall framework of the merchant processing logic and optimization strategy, which is divided into two parts: the main process and the optimization model.

[0089] like Figure 4 As shown in the figure, in the main process part, the processing starts with the input of contract information, merchant needs and business goals. After this information enters the system, after preliminary data sorting, the merchant demand splitting module is started. The demand splitting will conduct a preliminary analysis based on the merchant's attributes and demand characteristics to generate a preliminary demand plan. After the preliminary demand plan is generated, it enters the merchant demand matching stage and further refines the plan in combination with existing resources and warehouse capabilities. Finally, the refined demand matching plan is verified and confirmed. If the plan is approved, the final solution is output to complete the processing of merchant needs. As shown in the figure, Figure 5As shown in the figure, in the optimization model part, the framework is divided into two stages. The goal of the first stage optimization model is to minimize logistics costs, contract terms default rate, and consider factors such as industry characteristic distribution to generate a preliminary warehouse distribution plan. The core of this stage is to formulate a preliminary allocation strategy based on the overall needs of merchants. After the preliminary warehouse distribution plan is generated, the plan will enter the second stage optimization model for further refinement. The second stage optimization model mainly focuses on the decomposition and matching of merchants' specific needs. The core is to subdivide the merchant class needs in the preliminary plan into specific merchants and ensure the realization of the optimization goals under constraints. The model at this stage outputs the final merchant matching plan, including the specific correspondence between merchants and warehouses. Through the two-stage optimization model, the merchant demand allocation is gradually refined and improved to ensure the achievement of business goals.

[0090] This embodiment proposes a two-stage merchant inventory warehouse optimization method. The algorithm processing part mainly includes two main parts: the first is data preprocessing, which inputs the basic information of various warehouses and merchants. For the basic information of merchants, it is necessary to cluster and label the merchants according to the first-level industry, second-level industry, and piece type of the merchants. For the aggregated merchant class, the area occupancy and other information are calculated and summarized. The scale of the merchant dimension is reduced by clustering to facilitate subsequent calculations. The second part of the process is the establishment and solution of the operations model. For the merchants and warehouses in the first stage, the area usage during off-peak / big promotion periods, the production capacity elasticity of the park / single warehouse, and the aggregation of piece types and industries are considered. The optimization model is established with the goal of minimizing the warehouse moving cost, annual warehouse rent, and annual ferry cost, and the first stage of solution is carried out. For the solution results of the first stage, for the merchant class that has been separately warehoused, all merchants under this class can be directly assigned to the designated warehouse. However, for the merchant class that is split according to a certain ratio in the results, each split part needs to be mapped to the specific merchant under this class. Therefore, a second stage of solution is required. For the merchant class that has undergone warehouse splitting, the goal is to approximate the occupied area, minimize the warehouse moving cost, and consider the constraints of the merchant dimension to obtain a detailed warehouse moving plan.

[0091] Specifically, the reason for clustering merchants is to solve the problem that the merchant dimension is too large, the subsequent algorithm is difficult to converge, and the detailed warehouse results of the merchants cannot be obtained. Experiments have shown that when the merchant scale is thousands, the traditional algorithm model that considers the cost of moving warehouses can no longer complete convergence within an hour. Since this embodiment ultimately hopes to put merchants of the same level of industry, the same secondary industry, and the same type of goods together as much as possible, merchants are aggregated from the perspective of primary industry-secondary industry-type, and merchants of the same level of industry, the same secondary industry, and the same type of goods are clustered into the same category. The merchants in a merchant class range from dozens to hundreds. The scale of the problem is reduced in the form of merchant classes, and the unpacking of merchants of the same category is minimized in algorithm optimization. The purpose of aggregating industries is mainly for management considerations. Industry-specific warehouses are more convenient for the management of warehouses and merchants. Different industries are also convenient for specifying different management indicators. Second, the correlation between commodities in the same industry is relatively high. Aggregating industries themselves can reduce the situation of unpacking packages between multiple warehouses. The reason for gathering merchants of the same piece type is mainly to improve the warehouse's floor space efficiency. From the perspective of the warehouse's process, placing the same type of goods together often facilitates the warehouse to improve the process and increase the warehouse volume floor space efficiency. At the same time, the aggregation of ultra-small goods brings the opportunity to introduce automated equipment, which can further improve the warehouse's labor efficiency. The improvement in floor space efficiency will eventually reduce the required area and reduce warehouse rental costs by withdrawing warehouses. The merchant's piece type is marked by the merchant's average piece volume, and the classification of a merchant is determined according to the industry and piece type. For example, home appliances-digital 3C-medium piece type is a separate merchant class, and the subsequent algorithm optimization link will ensure that merchants of this merchant class are not divided into warehouses as much as possible.

[0092] The first stage optimization model can be a merchant industry aggregation based on merchant categories. Based on merchant clustering data and warehouse basic data, first, as shown in Tables 1, 2 and 3, the model related sets, parameters and decision variables are given.

[0093] Table 1 Correlation set of the first stage model

[0094]

[0095] Table 2. Relevant parameters of the first stage model

[0096]

[0097] Table 3 The first decision variables of the first stage model

[0098]

[0099] Specifically, the first-stage optimization model is as follows:

[0100] min cof1p1+cof2p2+cost rent +cost carry +costferry #(1)

[0101]

[0102]

[0103] Among them, constraint (2) is the warehouse usage flag constraint. The warehouse will be used only when there are merchants in the warehouse. Constraints (3) and (4) are merchant ratio splitting constraints, which describe the merchant class splitting situation. Merchants are allowed to split warehouses, but the sum of the merchant class splitting ratios in each warehouse should be 1. Describes whether there is a certain type of merchant in the warehouse. Constraints (5) and (6) are warehouse area constraints, which limit the warehouse merchant area to no more than 80% during normal hours and no more than 100% during promotions. Constraints (7)-(9) are piece type and industry aggregation constraints, which mark the piece types and primary and secondary industries in the warehouse respectively. Combined with the objective function, piece type and industry aggregation are achieved. Constraints (10) and (11) are business-required special industry aggregation constraints. Combined with the input from the business side, it ensures that certain special industries must be aggregated, such as establishing a special warehouse for alcohol. (12)-(14) are fulfillment cost constraints, which describe the fulfillment cost. Since the vehicle type in the ferry link is fixed, starting from the loading capacity of the vehicle type, considering the daily order volume of the merchants in the warehouse, the number of trips required for each warehouse is calculated. Since the vehicle departure cost and mileage cost of the trip are fixed, the annualized ferry cost corresponding to each warehouse can be quantitatively calculated in the objective function. Constraints (12) and (13) model the number of trains required by each warehouse. For example, if each train can transport 1,000 orders, considering a maximum of 2 segments, and the actual number of orders in the warehouse is 1,500, then equations (12) and (13) are equivalent to: Combined with constraint (14), to make the above equation valid, It means that at least 2 vehicles are needed to complete the transportation of the ferry link. The objective function consists of five parts. The first part, p1, is the total number of warehouses of the merchant class. This part has the highest priority and prioritizes to ensure that the merchant class does not split the warehouse as much as possible. The second part, p2, is the merchant piece type industry aggregation. It has the second priority and is used to minimize the number of warehouses of the same piece type and the same industry of the same piece type, so as to achieve the purpose of aggregating piece types and industries as much as possible. The remaining three parts correspond to the annualized warehouse rental cost, warehouse moving cost, and annualized ferry cost, respectively, corresponding to minimizing the sum of the three to achieve the optimal total cost under the premise of piece type and industry aggregation. Among them, the ferry cost cost ferry The quantitative calculation is based on the number of trips required for each warehouse and the warehouse-to-pickup distance. For example, in the above example, assuming that the warehouse-to-pickup distance is 100 km, the cost of a single vehicle is 100 yuan, and the transportation cost per kilometer is 1 yuan, then because Therefore, the corresponding cost is 365×400=146,000 yuan.

[0104] The final decision variables of the above model include which merchant types are in the warehouse and what is the corresponding proportion. The above operations research model can be easily solved using common commercial solvers. However, if there is a situation where a certain merchant type actually corresponds to a large number of merchants and a single warehouse cannot accommodate them, the merchant type will be split. For example, one warehouse has 40% of the medium-sized items - home appliances - digital 3C, and the other has 60% of the medium-sized items - home appliances - digital 3C. At this time, it is necessary to further decide the specific merchants actually allocated to each warehouse. Therefore, a second-stage optimization model is added: the merchant splitting model.

[0105] In order to reduce the scale of the merchant dimension, the merchants are clustered, but in the actual results, the merchant class may split warehouses in proportion. Therefore, the merchant split model is introduced to determine which merchants go to which warehouses. At the same time, the model at this stage will deal with the constraints of the merchant dimension: the commitment time limit of the contracts signed by different merchants is different, and the performance time limit that can be met by the waves of different warehouses is also different. Therefore, there is still a matching problem between merchants and warehouses. We divide the time limit levels of warehouses and merchants into three types: 1, 2, and 3. The time limit requirement of 1 is the highest, and the requirement of 3 is the lowest. Warehouses that can meet the time limit requirement of 1 can also meet the requirements of time limits 2 and 3. The merchant split model is only for merchant classes that have split warehouses. Each merchant class that splits warehouses will solve the model separately once, with the purpose of specifically matching the merchants of this merchant class to the warehouses assigned to this class in the first-stage results. Therefore, the merchant scale of the model at this stage is only the specific number of merchants of this merchant class, and the warehouse scale is only the warehouses with this merchant class in the first-stage results. The overall scale is much smaller than all merchants*all warehouses.

[0106] The relevant sets, parameters and decision variables of the second stage optimization model are shown in Tables 4, 5 and 6 below.

[0107] Table 4. Correlation set of the second stage optimization model

[0108] I The specific merchant corresponding to a certain merchant class <![CDATA[I t1 ]]> A collection of merchants with a commitment validity period of 1 <![CDATA[I t2 ]]> A collection of merchants with a commitment validity period of 2 J Warehouse Collection <![CDATA[J t2 ]]> A collection of warehouses that can meet timeliness requirements 2 and 3 <![CDATA[J t3 ]]> A collection of warehouses that can meet the timeliness requirement 3

[0109] Table 5. Relevant parameters of the second stage optimization model

[0110] <![CDATA[S j ]]> The total area of ​​the merchant class considered in warehouse j (first-stage results) <![CDATA[s i ]]> Total area of ​​business <![CDATA[c ij ]]> The cost of moving all merchants i to warehouse j <![CDATA[cof1]]> Objective function coefficient <![CDATA[cof2]]> Objective function coefficient

[0111] Table 6 The second decision variables of the second stage optimization model

[0112] <![CDATA[x ij ]]> Whether merchant i is in warehouse j, 0-1 variable

[0113]

[0114] The first part of the second phase optimization model objective is to minimize the difference in merchant area occupancy, and the second part is to minimize the merchant relocation cost. On the one hand, the area of ​​the merchants after the division is guaranteed to be as consistent as possible with that of the first phase, ensuring the upper limit of the warehouse area. For example, a certain merchant category originally occupied 1,000 square meters in total. In the first phase, 40% of this category was allocated to warehouse A, and the remaining 60% was allocated to warehouse B. Then, the goal is to ensure that after the merchants are specific, the sum of the areas of all merchants in warehouse A is 400 square meters, and in warehouse B is 600 square meters. On the other hand, minimize the merchant relocation cost and reduce the number of relocations. The model's constraint (16) is the merchant non-warehouse constraint, ensuring that each merchant only enters one warehouse. Formulas (17) and (18) are the merchant commitment time constraints, ensuring that merchants with high commitment time will not be adjusted to warehouses that cannot meet the time requirements. Finally, through the solution of the above two stages, we can finally get which warehouses each merchant goes to.

[0115] According to the comparison results of a certain regional warehouse group before and after optimization, the merchant inventory warehouse optimization method proposed in this embodiment effectively concentrates industries and types of items. In terms of total cost, the total warehouse relocation cost required by merchants in this embodiment is about 2.2 million; the annualized ferry cost from the warehouse to the first sorting center before the warehouse relocation is about 11.56 million, and it is reduced to about 9.37 million after the warehouse relocation, saving about 2.19 million in ferrying. The warehouse rental cost that can be saved by retreating the area is 7.2 million per year, and the total annualized income of the costs involved in the model is about 7.2 million.

[0116] According to a third aspect of an embodiment of the present invention, a warehouse optimization device is provided.

[0117] Figure 6 is a schematic diagram of the main modules of the warehouse optimization device according to an embodiment of the present invention. Figure 6 As shown, a warehouse optimization device 600 includes:

[0118] The information acquisition module 601 is used to acquire the relevant information of the merchant and the basic information of the warehouse; the relevant information of the merchant includes industry information, commodity type information, commodity sales information and promised time limit information;

[0119] The merchant clustering module 602 is used to cluster merchants according to their industry information and product type information to obtain one or more merchant classes;

[0120] The warehouse determination module 603 is used to determine the warehouse corresponding to each merchant class based on the basic information of the warehouse and the sales information of the merchant;

[0121] The re-adjustment module 604 is used to adjust the warehouses corresponding to at least some merchants in the merchant class based on the merchant's promised timeliness information in response to the number of warehouses corresponding to the merchant class exceeding a set threshold.

[0122] Optionally, the warehouse determination module 603 is further used for:

[0123] Based on the basic information of the warehouse and the sales information of the merchant, the warehouse corresponding to each merchant category is determined using the pre-built first-stage optimization model;

[0124] The pre-built first-stage optimization model is obtained through the following steps:

[0125] Based on the acquired merchant historical clustering information, warehouse historical basic information and merchant historical sales information, define the first decision variable and the first constraint condition; wherein the decision variable includes the distribution and proportion of merchant types in the warehouse; the constraint condition includes merchant proportion warehouse splitting constraint, warehouse usage area constraint, piece type industry aggregation constraint, special industry aggregation constraint and fulfillment cost constraint;

[0126] Determine a first objective function according to the first decision variable and the first constraint condition; wherein the first objective function includes minimizing the total number of warehouses of the merchant class, minimizing the industry aggregation of the piece type, or minimizing the total cost;

[0127] A first-stage optimization model is constructed based on the first decision variable, the first constraint condition and the first objective function.

[0128] Optionally, the re-determining module 603 is further used for:

[0129] Based on the promised timeliness information of the merchants, the warehouses corresponding to at least some of the merchants in the merchant class are adjusted by using a pre-built second-stage optimization model;

[0130] The pre-built second-stage optimization model is obtained by the following steps:

[0131] Filter out the historical merchant classes whose corresponding warehouse quantity exceeds the set threshold from the acquired merchant historical clustering information, and obtain the historical commitment timeliness information and historical warehouse basic information corresponding to the historical merchant classes;

[0132] Based on the historical commitment time limit information corresponding to the historical merchant class and the historical basic information of the warehouse, define the second decision variable and the second constraint condition; wherein the second decision variable includes the distribution of merchants in the warehouse, and the second constraint condition includes the merchant non-warehouse constraint and the merchant commitment time limit constraint;

[0133] Determine a second objective function according to the second decision variable and the second constraint condition; wherein the second objective function includes minimizing the difference in merchant area occupation and minimizing the merchant's warehouse relocation cost;

[0134] A second-stage optimization model is constructed based on the second decision variable, the second constraint condition and the second objective function.

[0135] Optionally, the device 600 further includes a full allocation module, and the full allocation module is used to:

[0136] In response to the number of warehouses corresponding to the merchant class not exceeding the set threshold, all merchants in the merchant class are allocated to the warehouses corresponding to the merchant class.

[0137] It should be noted that the specific implementation content of the warehouse optimization device in the embodiment of the present invention has been described in detail in the above warehouse optimization method, so the repeated content will not be described here.

[0138] According to a fourth aspect of an embodiment of the present invention, a training device for a warehouse optimization model is provided.

[0139] Figure 7 is a schematic diagram of the main modules of the training device of the warehouse optimization model according to an embodiment of the present invention, such as Figure 7 As shown, a training device 700 for a bin optimization model includes:

[0140] The target acquisition module 701 is used to acquire relevant information of the target merchant, basic information of the target warehouse, and historical warehouse information corresponding to the target merchant; wherein the historical warehouse information corresponding to the target merchant is determined by the first-stage optimization model or the second-stage optimization model;

[0141] The model training module 702 is used to take the relevant information of the target merchant and the basic information of the target warehouse as the model input, and the historical warehouse information corresponding to the target merchant as the model output, and train to obtain the warehouse optimization model.

[0142] It should be noted that the specific implementation content of the training device for the warehouse optimization model in the embodiment of the present invention has been described in detail in the training method for the warehouse optimization model above, so the repeated content will not be described here.

[0143] According to a fifth aspect of an embodiment of the present invention, there is provided an electronic device, including:

[0144] one or more processors;

[0145] a storage device for storing one or more programs,

[0146] When one or more programs are executed by one or more processors, the one or more processors implement the method provided by the first aspect of the embodiment of the present invention and / or the second aspect of the embodiment of the present invention.

[0147] According to a sixth aspect of an embodiment of the present invention, a computer-readable medium is provided, on which a computer program is stored. When the program is executed by a processor, the method provided by the first aspect of the embodiment of the present invention and / or the second aspect of the embodiment of the present invention is implemented.

[0148] According to a seventh aspect of an embodiment of the present invention, a computer program product is provided, including a computer program, which, when executed by a processor, implements the method provided by the first aspect of the embodiment of the present invention and / or the second aspect of the embodiment of the present invention.

[0149] Figure 8 An exemplary system architecture 800 is shown to which the warehouse division optimization method or warehouse division optimization device according to an embodiment of the present invention can be applied.

[0150] like Figure 8 As shown, system architecture 800 may include terminal devices 801, 802, 803, a network 804, and a server 805. Network 804 is used to provide a medium for communication links between terminal devices 801, 802, 803 and server 805. Network 804 may include various connection types, such as wired, wireless communication links, or optical fiber cables, etc.

[0151] Users can use terminal devices 801, 802, and 803 to interact with server 805 through network 804 to receive or send messages, etc. Various communication client applications can be installed on terminal devices 801, 802, and 803, such as shopping applications, web browser applications, search applications, instant messaging tools, email clients, social platform software, etc. (only examples).

[0152] The terminal devices 801 , 802 , and 803 may be various electronic devices having a display screen and supporting web browsing, including but not limited to smart phones, tablet computers, laptop computers, desktop computers, and the like.

[0153] The server 805 may be a server that provides various services, such as a backend management server (only an example) that provides support for shopping websites browsed by users using the terminal devices 801, 802, and 803. The backend management server may analyze and process the received data such as the warehouse optimization request, and feed back the processing result (warehouse optimization result - only an example) to the terminal device.

[0154] It should be noted that the warehouse optimization method provided in the embodiment of the present invention is generally executed by the server 805, and accordingly, the warehouse optimization device is generally set in the server 805.

[0155] It should be noted that the training method of the warehouse optimization model provided in the embodiment of the present invention is generally run by the server 805, and accordingly, the training device of the warehouse optimization model is generally set in the server 805.

[0156] It should be understood that Figure 8 The number of terminal devices, networks and servers in the embodiment is only for illustration. Any number of terminal devices, networks and servers may be provided according to implementation requirements.

[0157] Reference below Fig. 9 , which shows a schematic diagram of the structure of a computer system 900 of a terminal device suitable for implementing an embodiment of the present invention. Fig. 9 The terminal device shown is only an example and should not bring any limitation to the functions and scope of use of the embodiments of the present invention.

[0158] like Fig. 9 As shown, the computer system 900 includes a central processing unit (CPU) 901, which can perform various appropriate actions and processes according to a program stored in a read-only memory (ROM) 902 or a program loaded from a storage part 908 into a random access memory (RAM) 903. In the RAM 903, various programs and data required for the operation of the system 900 are also stored. The CPU 901, the ROM 902, and the RAM 903 are connected to each other via a bus 904. An input / output (I / O) interface 905 is also connected to the bus 904.

[0159] The following components are connected to the I / O interface 905: an input section 906 including a keyboard, a mouse, etc.; an output section 907 including a cathode ray tube (CRT), a liquid crystal display (LCD), etc., and a speaker, etc.; a storage section 908 including a hard disk, etc.; and a communication section 909 including a network interface card such as a LAN card, a modem, etc. The communication section 909 performs communication processing via a network such as the Internet. A drive 910 is also connected to the I / O interface 905 as needed. A removable medium 911, such as a magnetic disk, an optical disk, a magneto-optical disk, a semiconductor memory, etc., is installed on the drive 910 as needed, so that a computer program read therefrom is installed into the storage section 908 as needed.

[0160] In particular, according to the embodiments disclosed in the present invention, the process described above with reference to the flowchart can be implemented as a computer software program. For example, the embodiments disclosed in the present invention include a computer program product, which includes a computer program carried on a computer-readable medium, and the computer program includes a program code for executing the method shown in the flowchart. In such an embodiment, the computer program can be downloaded and installed from the network through the communication part 909, and / or installed from the removable medium 911. When the computer program is executed by the central processing unit (CPU) 901, the above-mentioned functions defined in the system of the present invention are executed.

[0161] It should be noted that the computer-readable medium shown in the present invention may be a computer-readable signal medium or a computer-readable storage medium or any combination of the above two. The computer-readable storage medium may be, for example, but not limited to, an electrical, magnetic, optical, electromagnetic, infrared, or semiconductor system, device or device, or any combination of the above. More specific examples of computer-readable storage media may include, but are not limited to: an electrical connection with one or more wires, a portable computer disk, a hard disk, a random access memory (RAM), a read-only memory (ROM), an erasable programmable read-only memory (EPROM or flash memory), an optical fiber, a portable compact disk read-only memory (CD-ROM), an optical storage device, a magnetic storage device, or any suitable combination of the above. In the present invention, a computer-readable storage medium may be any tangible medium containing or storing a program that can be used by or in combination with an instruction execution system, device or device. In the present invention, a computer-readable signal medium may include a data signal propagated in a baseband or as part of a carrier wave, which carries a computer-readable program code. This propagated data signal may take a variety of forms, including but not limited to electromagnetic signals, optical signals, or any suitable combination of the above. The computer-readable signal medium may also be any computer-readable medium other than a computer-readable storage medium, which may send, propagate or transmit a program for use by or in conjunction with an instruction execution system, apparatus or device. The program code contained on the computer-readable medium may be transmitted using any appropriate medium, including but not limited to: wireless, wire, optical cable, RF, etc., or any suitable combination of the above.

[0162] The flow chart and block diagram in the accompanying drawings illustrate the possible architecture, functions and operations of the system, method and computer program product according to various embodiments of the present invention. In this regard, each box in the flow chart or block diagram can represent a module, a program segment, or a part of a code, and the above-mentioned module, program segment, or a part of a code contains one or more executable instructions for realizing the specified logical function. It should also be noted that in some alternative implementations, the functions marked in the box can also occur in a different order from the order marked in the accompanying drawings. For example, two boxes represented in succession can actually run basically in parallel, and they can sometimes run in the opposite order, depending on the functions involved. It should also be noted that each box in the block diagram or flow chart, and the combination of the boxes in the block diagram or flow chart can be implemented with a dedicated hardware-based system that runs the specified function or operation, or can be implemented with a combination of dedicated hardware and computer instructions.

[0163] The modules involved in the embodiments of the present invention may be implemented in software or in hardware. The modules described may also be set in a processor, for example, they may be described as: a processor including an information acquisition module, a merchant clustering module, a warehouse determination module, and a readjustment module. The names of these modules do not, in some cases, constitute a limitation on the module itself. For example, the information acquisition module may also be described as a "module for obtaining relevant information of merchants and basic information of warehouses." Alternatively, a processor includes a target acquisition module and a model training module, wherein the names of these modules do not, in some cases, constitute a limitation on the module itself. For example, the target acquisition module may also be described as a "module for obtaining relevant information of target merchants, basic information of target warehouses, and historical warehouse information corresponding to target merchants."

[0164] As another aspect, the present invention also provides a computer-readable medium, which may be included in the device described in the above embodiment; or it may exist independently and not be assembled into the device. The above computer-readable medium carries one or more programs, and when the above one or more programs are run by a device, the device includes: obtaining relevant information of merchants and basic information of warehouses; wherein the relevant information of merchants includes industry information, commodity type information, commodity sales information and promised time limit information; clustering merchants according to the industry information and commodity type information of merchants to obtain one or more merchant classes; based on the basic information of warehouses and the sales information of merchants, determining the warehouses corresponding to each merchant class; in response to the number of warehouses corresponding to the merchant class exceeding the set threshold, adjusting the warehouses corresponding to at least some merchants in the merchant class based on the merchant's promised time limit information. Alternatively, the device implements the following method: obtaining relevant information of the target merchant, basic information of the target warehouse and historical warehouse information corresponding to the target merchant; wherein the historical warehouse information corresponding to the target merchant is determined by the first stage optimization model or the second stage optimization model; using the relevant information of the target merchant and the basic information of the target warehouse as the model input, and the historical warehouse information corresponding to the target merchant as the model output, and training to obtain the warehouse optimization model.

[0165] The computer program product provided by an embodiment of the present invention includes a computer program, which, when executed by a processor, implements the warehouse optimization method in the embodiment of the first aspect of the present invention and / or the training method of the warehouse optimization model in the embodiment of the second aspect of the present invention.

[0166] The technical solution according to the embodiment of the present invention has the following advantages or beneficial effects: by obtaining relevant information of merchants and basic information of warehouses; wherein the relevant information of merchants includes industry information, commodity type information, commodity sales information and promised time limit information; according to the industry information and commodity type information of merchants, the merchants are clustered to obtain one or more merchant classes; based on the basic information of warehouses and the sales information of merchants, the warehouses corresponding to each merchant class are determined; in response to the number of warehouses corresponding to a merchant class exceeding a set threshold, the warehouses corresponding to at least some merchants in the merchant class are adjusted based on the promised time limit information of the merchants; thereby, the relevant information of merchants and the basic information of warehouses can be comprehensively considered, and merchants can be aggregated from the industry and item type dimensions, and then the aggregated merchant classes can be adjusted for multiple warehouse divisions, thereby improving the accuracy of warehouse division optimization and enhancing the user experience. Obtain relevant information of the target merchant, basic information of the target warehouse, and historical warehouse information corresponding to the target merchant; wherein the historical warehouse information corresponding to the target merchant is determined by the first-stage optimization model or the second-stage optimization model; use the relevant information of the target merchant and the basic information of the target warehouse as model input, and the historical warehouse information corresponding to the target merchant as model output, and train to obtain a warehouse optimization model; thereby, many business factors and constraints can be considered to improve the accuracy of model predictions.

[0167] The above specific implementations do not constitute a limitation on the protection scope of the present invention. It should be understood by those skilled in the art that various modifications, combinations, sub-combinations and substitutions may occur depending on design requirements and other factors. Any modification, equivalent substitution and improvement made within the spirit and principle of the present invention shall be included in the protection scope of the present invention.

[0168] It should be noted that the acquisition, storage and application of user personal information involved in the technical solution of the present disclosure are in compliance with the provisions of relevant laws and regulations and do not violate public order and good morals.

Claims

1. A warehouse optimization method, characterized in that: include: Obtaining relevant information of the merchant and basic information of the warehouse; wherein the relevant information of the merchant includes industry information, product type information, product sales information and promised time limit information; Clustering the merchants according to the industry information and product type information of the merchants to obtain one or more merchant classes; Based on the basic information of the warehouse and the sales information of the merchant, determine the warehouse corresponding to each of the merchant categories; In response to the number of warehouses corresponding to the merchant class exceeding a set threshold, the warehouses corresponding to at least some of the merchants in the merchant class are adjusted based on the promised timeliness information of the merchants.

2. The method according to claim 1, characterized in that Based on the basic information of the warehouse and the sales information of the merchant, the warehouse corresponding to each merchant class is determined, including: Based on the basic information of the warehouse and the sales information of the merchant, the warehouse corresponding to each merchant class is determined using the pre-built first-stage optimization model; The pre-built first-stage optimization model is obtained by the following steps: Based on the acquired merchant historical clustering information, warehouse historical basic information and merchant historical sales information, define the first decision variable and the first constraint condition; wherein the decision variable includes the distribution and proportion of merchant types in the warehouse; the constraint condition includes merchant proportion warehouse splitting constraint, warehouse usage area constraint, piece type industry aggregation constraint, special industry aggregation constraint and fulfillment cost constraint; Determine a first objective function according to the first decision variable and the first constraint condition; wherein the first objective function includes minimizing the total number of warehouses of merchant types, minimizing the aggregation of piece-type industries, or minimizing the total cost; A first-stage optimization model is constructed based on the first decision variable, the first constraint condition and the first objective function.

3. The method according to claim 1, characterized in that Adjusting the warehouses corresponding to at least some of the merchants in the merchant class based on the promised timeliness information of the merchants includes: Based on the promised timeliness information of the merchant, using the pre-built second-stage optimization model, adjusting the warehouses corresponding to at least some of the merchants in the merchant class; The pre-built second-stage optimization model is obtained by the following steps: Filter out the historical merchant classes whose corresponding warehouse quantity exceeds the set threshold from the acquired merchant historical clustering information, and obtain the historical commitment time limit information and warehouse historical basic information corresponding to the historical merchant classes; Based on the historical commitment time limit information and the historical basic warehouse information corresponding to the historical merchant class, define a second decision variable and a second constraint condition; wherein the second decision variable includes the distribution of merchants in the warehouse, and the second constraint condition includes a merchant non-warehouse constraint and a merchant commitment time limit constraint; Determine a second objective function according to the second decision variable and the second constraint condition; wherein the second objective function includes minimizing the difference in merchant area occupation and minimizing the merchant relocation cost; A second-stage optimization model is constructed based on the second decision variable, the second constraint condition and the second objective function.

4. The method according to claim 1, characterized in that: The method further comprises: In response to the number of warehouses corresponding to the merchant class not exceeding a set threshold, all merchants in the merchant class are allocated to the warehouses corresponding to the merchant class.

5. A training method for a warehouse optimization model, characterized in that: include: Obtain relevant information of the target merchant, basic information of the target warehouse, and historical warehouse information corresponding to the target merchant; wherein the historical warehouse information corresponding to the target merchant is determined by the first-stage optimization model or the second-stage optimization model; The relevant information of the target merchant and the basic information of the target warehouse are used as model inputs, and the historical warehouse information corresponding to the target merchant is used as model output, and the warehouse optimization model is obtained through training.

6. A warehouse optimization device, characterized in that: include: Information acquisition module, used to obtain relevant information of merchants and basic information of warehouses; The merchant's relevant information includes industry information, product type information, product sales information and promised time limit information; A merchant clustering module, used to cluster the merchants according to the industry information and product type information of the merchants to obtain one or more merchant classes; A warehouse determination module, used to determine the warehouse corresponding to each merchant class based on the basic information of the warehouse and the sales information of the merchant; The re-adjustment module is used to adjust the warehouses corresponding to at least some merchants in the merchant class based on the promised time information of the merchants in response to the number of warehouses corresponding to the merchant class exceeding a set threshold.

7. A training device for a warehouse optimization model, characterized in that: include: A target acquisition module is used to acquire relevant information of a target merchant, basic information of a target warehouse, and historical warehouse information corresponding to the target merchant; wherein the historical warehouse information corresponding to the target merchant is determined by the first-stage optimization model or the second-stage optimization model; The model training module is used to take the relevant information of the target merchant and the basic information of the target warehouse as model input, and the historical warehouse information corresponding to the target merchant as model output, and train to obtain a warehouse optimization model.

8. An electronic device, characterized in that: include: one or more processors; a storage device for storing one or more programs, When the one or more programs are executed by the one or more processors, the one or more processors implement the method according to any one of claims 1 to 5.

9. A computer readable medium having a computer program stored thereon, characterized in that: When the program is executed by a processor, the method according to any one of claims 1 to 5 is implemented.

10. A computer program product, comprising a computer program, characterized in that When the computer program is executed by a processor, the method according to any one of claims 1 to 5 is implemented.

Images