Intelligent inventory management method and system
By obtaining the substitution relationship between goods and sales benefits, dividing high and low-yield goods, and adjusting inventory according to inventory and sales demands, the existing inventory management methods are solved, and lower management costs and higher overall benefits are achieved.
Patent Information
- Application Number
- CN202411605619.5
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2024-11-11
- Publication Date
- 2025-04-01
AI Technical Summary
The existing inventory management method has high management costs and limited overall returns, and has failed to fully consider the substitution relationship between commodities and sales returns.
By acquiring the substitution relationship between multiple target products and their initial inventory, predicted sales volume and sales revenue in the preset cycle, determining the substitute product replacement revenue for each target product, dividing the goods into high-yield goods and low-yield goods, and determining the inventory increase of high-yield goods based on the initial inventory, predicted sales volume and conversion needs.
By concentrating resources to supplement inventory of high-yield commodities, reduce inventory resources occupied by low-yield commodities, reduce inventory management costs, and improve the overall profit level.
Smart Images
Figure CN119740953A_ABST
Abstract
Description
Technical Field
[0001] The present application relates to the field of data processing technology, and in particular to an intelligent inventory management method and system. Background Art
[0002] Inventory management is the core link in enterprise supply chain management, which aims to ensure the effective matching of commodity supply and market demand. With the development of e-commerce and big data technology, algorithm-based inventory management methods have been widely used. These methods use intelligent algorithms and real-time data analysis to fully improve the efficiency and accuracy of inventory management.
[0003] Existing inventory management methods include the following processes: collecting historical sales data and market demand information of goods; using machine learning algorithms to predict the market demand for goods in the future; and determining the replenishment quantity for the goods based on the current inventory of the goods and the predicted market demand.
[0004] Although existing methods have improved the efficiency and accuracy of inventory management to a certain extent, there are still problems such as high management costs and limited overall benefits. Summary of the invention
[0005] In view of the above-mentioned problems, the present application is proposed to provide an intelligent inventory management method and system that overcomes the problems or at least partially solves the problems, including:
[0006] An intelligent inventory management method, comprising:
[0007] Obtaining a substitution relationship between a plurality of target commodities, and an initial inventory, predicted sales volume, and sales revenue of each of the target commodities within a preset period; wherein the substitution relationship includes substitute commodities of each of the target commodities among the remaining target commodities, and a substitution rate of the substitute commodities to the target commodity;
[0008] Determine the replacement revenue of the replacement product corresponding to each target product for the target product according to the sales revenue of the replacement product corresponding to the target product and the replacement rate of the replacement product for the target product;
[0009] The target commodity whose sales revenue is greater than or equal to the replacement revenue of the corresponding replacement commodity to the target commodity is set as a high-revenue commodity, and the remaining target commodities are set as low-revenue commodities;
[0010] Determining the conversion demand of each of the high-yield commodities according to the initial inventory and the predicted sales volume of the low-yield commodities, the substitute commodities corresponding to the low-yield commodities, and the substitution rate of the substitute commodities for the low-yield commodities;
[0011] The inventory increment of each of the high-yield commodities is determined according to the initial inventory of the high-yield commodities, the predicted sales volume, and the conversion demand.
[0012] Preferably, the step of obtaining the substitution relationship between the multiple target commodities includes:
[0013] Obtaining platform display information of each target product; wherein the platform display information includes name information, category information, appearance information, price information and evaluation information;
[0014] According to the name information, the category information, the appearance information, the price information and the evaluation information, the substitute products of each of the target products among the remaining target products and the substitution rate of the substitute products for the target product are determined.
[0015] Preferably, the step of determining the substitute products of each target product among the remaining target products and the substitution rate of the substitute products for the target product based on the name information, the category information, the appearance information, the price information and the evaluation information comprises:
[0016] Generate a comprehensive feature vector for each of the target products based on the name information, the category information, the appearance information, the price information, and the evaluation information;
[0017] Using a pre-built substitution rate evaluation model, the comprehensive feature vector of the target product and the comprehensive feature vectors of each of the remaining target products are processed to obtain a substitution rate of each of the remaining target products for the target product;
[0018] The remaining target commodities whose replacement rates for the target commodity are greater than a preset value are set as substitute commodities for the target commodity.
[0019] Preferably, the step of generating a comprehensive feature vector of each target product based on the name information, the category information, the appearance information, the price information and the evaluation information comprises:
[0020] Generate a name feature vector, a category feature vector, an appearance feature vector, a price feature vector and an evaluation feature vector for each target product according to the name information, the category information, the appearance information, the price information and the evaluation information;
[0021] A comprehensive feature vector of each of the target commodities is generated based on the name feature vector, the category feature vector, the appearance feature vector, the price feature vector and the evaluation feature vector.
[0022] Preferably, the substitution rate evaluation model includes a feature fusion layer and a multi-layer fully connected network; wherein the multi-layer fully connected network includes multiple hidden layers and a discard layer connected after each of the hidden layers; the step of using the pre-built substitution rate evaluation model to process the comprehensive feature vector of the target product and the comprehensive feature vector of each of the remaining target products to obtain the substitution rate of each of the remaining target products for the target product includes:
[0023] The comprehensive feature vector of the target product and the comprehensive feature vector of each of the remaining target products are input into the feature fusion layer; wherein the feature fusion layer is used to fuse the comprehensive feature vector of the target product with the comprehensive feature vector of each of the remaining target products to obtain a fused feature vector, and input the fused feature vector into the multi-layer fully connected network; the multi-layer fully connected network is used to perform multi-level mapping on the fused feature vector through linear transformation and nonlinear activation function to obtain the substitution rate of each of the remaining target products for the target product.
[0024] Preferably, the step of determining the conversion demand of each of the high-yield commodities based on the initial inventory and the predicted sales volume of the low-yield commodities, the substitute commodities corresponding to the low-yield commodities, and the substitution rate of the substitute commodities for the low-yield commodities includes:
[0025] When the difference between the sum of the predicted sales volume and the current conversion demand of the low-yield commodity and the current reference inventory is greater than an update threshold, the current conversion demand of the substitute commodity corresponding to the low-yield commodity is updated according to the predicted sales volume, the current conversion demand and the current reference inventory of the low-yield commodity, and the substitution rate of the substitute commodity corresponding to the low-yield commodity for the low-yield commodity, and the current reference inventory of the low-yield commodity is updated according to the predicted sales volume and the current conversion demand of the low-yield commodity; wherein the starting value of the current reference inventory of the low-yield commodity is the initial inventory.
[0026] Preferably, the step of determining the inventory increment of each high-yield commodity according to the initial inventory of the high-yield commodity, the predicted sales volume and the conversion demand includes:
[0027] When the difference between the sum of the predicted sales volume and the current conversion demand of the high-yield commodity and the current expected inventory is greater than an update threshold, the current expected inventory of the high-yield commodity is updated according to the predicted sales volume and the current conversion demand of the high-yield commodity; wherein the starting value of the current expected inventory of the high-yield commodity is the initial inventory, and the difference between the ending value and the initial inventory is the inventory increment.
[0028] An intelligent inventory management system, comprising:
[0029] A commodity information acquisition module, used to acquire the substitution relationship between multiple target commodities, and the initial inventory, predicted sales volume and sales revenue of each target commodity within a preset period; wherein the substitution relationship includes the substitute commodities of each target commodity among the remaining target commodities, and the substitution rate of the substitute commodities to the target commodity;
[0030] A substitution revenue determination module, configured to determine the substitution revenue of the substitute commodity corresponding to each target commodity for the target commodity according to the sales revenue of the substitute commodity corresponding to the target commodity and the substitution rate of the substitute commodity for the target commodity;
[0031] a target commodity classification module, configured to set the target commodity whose sales revenue is greater than or equal to the substitute revenue of the corresponding substitute commodity to the target commodity as a high-revenue commodity, and set the remaining target commodities as low-revenue commodities;
[0032] A conversion demand determination module, configured to determine the conversion demand of each of the high-yield commodities according to the initial inventory and the predicted sales volume of the low-yield commodities, the substitute commodities corresponding to the low-yield commodities, and the substitution rate of the substitute commodities for the low-yield commodities;
[0033] The inventory increment determination module is used to determine the inventory increment of each high-profit commodity according to the initial inventory, the predicted sales volume and the conversion demand of the high-profit commodity.
[0034] A computer device comprises a processor, a memory and a computer program stored in the memory and capable of running on the processor, wherein the computer program can implement any of the above methods when executed by the processor.
[0035] A computer-readable storage medium stores a computer program, and when the computer program is executed by a processor, the method described above can be implemented.
[0036] This application has the following advantages:
[0037] Compared with the problems of high management cost and limited overall benefits of the existing methods, the present application provides a solution for classifying and managing commodities by considering the conversion benefits of substitute products and the self-benefits, specifically: an intelligent inventory management method, comprising: obtaining the substitution relationship between multiple target commodities, and the initial inventory, predicted sales volume and sales revenue of each target commodity within a preset period; wherein the substitution relationship includes the substitute commodities of each target commodity among the remaining target commodities, and the substitution rate of the substitute commodities for the target commodity; determining the sales revenue of the substitute commodities corresponding to the target commodity and the substitution rate of the substitute commodities for the target commodity according to the sales revenue of the substitute commodities corresponding to the target commodity and the substitution rate of the substitute commodities for the target commodity The replacement revenue of the target commodity corresponding to the substitute commodity; setting the target commodity whose sales revenue is greater than or equal to the replacement revenue of the corresponding substitute commodity to the target commodity as a high-profit commodity, and setting the remaining target commodities as low-profit commodities; determining the conversion demand of each high-profit commodity based on the initial inventory and the predicted sales volume of the low-profit commodity, the substitute commodity corresponding to the low-profit commodity, and the replacement rate of the substitute commodity for the low-profit commodity; determining the inventory increment of each high-profit commodity based on the initial inventory, the predicted sales volume and the conversion demand of the high-profit commodity.
[0038] By dividing all commodities into low-profit commodities and high-profit commodities based on the comparison between the conversion revenue of substitute products and their own revenue, and determining the inventory increment of high-profit commodities based on the initial inventory, predicted sales volume and the conversion demand caused by insufficient inventory of low-profit commodities, resources can be concentrated on replenishing the inventory of high-profit commodities, reducing the inventory resources occupied by low-profit commodities, thereby reducing the cost of inventory management, and at the same time avoiding low-profit commodities occupying the market share of high-profit commodities, thereby improving the overall profit level. BRIEF DESCRIPTION OF THE DRAWINGS
[0039] In order to more clearly illustrate the technical solution of the present application, the drawings required for use in the description of the present application will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present application. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying any creative work.
[0040] Figure 1 is a flowchart of the steps of the intelligent inventory management method provided by an embodiment of the present application;
[0041] Figure 2 It is a structural block diagram of an intelligent inventory management system provided by an embodiment of the present application. DETAILED DESCRIPTION
[0042] In order to make the objects, features and advantages of the present application more obvious and understandable, the present application is further described in detail below in conjunction with the accompanying drawings and specific implementation methods. Obviously, the described embodiments are part of the embodiments of the present application, rather than all of the embodiments. Based on the embodiments in the present application, all other embodiments obtained by ordinary technicians in the field without creative work are within the scope of protection of the present application.
[0043] By analyzing the prior art, the inventors discovered that there is a substitution relationship between different commodities, and their sales revenues are different. However, the existing inventory management method fails to fully consider the substitution relationship and sales revenue between commodities, but instead performs independent inventory management for each commodity, which can easily cause low-profit commodities to occupy too many inventory resources, thereby increasing the cost of inventory management. In addition, it can also easily cause low-profit commodities to occupy the market share of high-profit commodities, thereby affecting the overall profit level.
[0044] Reference Figure 1 In an embodiment of the present application, compared with the problems of high management cost and limited overall benefits of the existing method, the present application provides a solution for classifying and managing commodities by considering the conversion benefits of substitutes and their own benefits, specifically:
[0045] An intelligent inventory management method, comprising:
[0046] S110, obtaining a substitution relationship between a plurality of target commodities, and an initial inventory, predicted sales volume, and sales revenue of each target commodity within a preset period; wherein the substitution relationship includes substitute commodities of each target commodity among the remaining target commodities, and a substitution rate of the substitute commodity to the target commodity;
[0047] S120, determining the replacement revenue of the replacement product corresponding to each target product for the target product according to the sales revenue of the replacement product corresponding to the target product and the replacement rate of the replacement product for the target product;
[0048] S130, setting the target commodity whose sales revenue is greater than or equal to the replacement revenue of the corresponding replacement commodity to the target commodity as a high-revenue commodity, and setting the remaining target commodities as low-revenue commodities;
[0049] S140, determining the conversion demand of each of the high-yield commodities according to the initial inventory and the predicted sales volume of the low-yield commodities, the substitute commodities corresponding to the low-yield commodities, and the substitution rate of the substitute commodities for the low-yield commodities;
[0050] S150: Determine the inventory increment of each of the high-yield commodities according to the initial inventory of the high-yield commodities, the predicted sales volume, and the conversion demand.
[0051] By dividing all commodities into low-profit commodities and high-profit commodities based on the comparison between the conversion revenue of substitute products and their own revenue, and determining the inventory increment of high-profit commodities based on the initial inventory, predicted sales volume and the conversion demand caused by insufficient inventory of low-profit commodities, resources can be concentrated on replenishing the inventory of high-profit commodities, reducing the inventory resources occupied by low-profit commodities, thereby reducing the cost of inventory management, and at the same time avoiding low-profit commodities occupying the market share of high-profit commodities, thereby improving the overall profit level.
[0052] Next, the intelligent inventory management method in this exemplary embodiment will be further described.
[0053] As described in step S110, the substitution relationship between multiple target products, as well as the initial inventory, predicted sales volume and sales revenue of each target product within a preset period are obtained; wherein the substitution relationship includes the substitute products of each target product among the remaining target products, and the substitution rate of the substitute products for the target product.
[0054] The substitution relationship between multiple target commodities is obtained through historical sales data or in combination with intelligent algorithms and platform display information; wherein, the substitution relationship includes the substitute commodities of each target commodity among the remaining target commodities (i.e., when the inventory of the target commodity is zero, the commodity that converts the lost sales of the target commodity into its own sales with a higher probability), and the substitution rate of the substitute commodity to the target commodity (i.e., when the inventory of the target commodity is zero, the probability of converting the lost sales of the target commodity into its own sales). It can be understood that each target commodity can have zero substitute commodities or more than one substitute commodity among the remaining target commodities; for the target commodity with more than one substitute commodity, the sum of the substitution rates of all the substitute commodities to the target commodity is a value greater than 0 and less than 100%.
[0055] The initial inventory, predicted sales volume and sales revenue of each target commodity within the preset period are obtained through inventory management data and market evaluation data. It can be understood that the initial inventory is the initial inventory quantity of the target commodity within the preset period. The predicted sales volume is the predicted sales volume of the target commodity within the preset period, and the predicted sales volume can be predicted based on the historical sales data and market demand information of the target commodity. The sales revenue is the profit of a single unit sold by the target commodity within the preset period, and the sales revenue can be calculated based on the sales price, procurement cost and management cost of the target commodity.
[0056] In an embodiment of the present application, the specific process of "obtaining the substitution relationship between multiple target products" in step S110 can be further explained in combination with the following description.
[0057] As described in the following steps, the platform display information of each target product is obtained; wherein the platform display information includes name information, category information, appearance information, price information and evaluation information.
[0058] It can be understood that the name information is the name or title of the target product. The category information is the category or type to which the target product belongs. The appearance information is the image of the target product. The price information is the selling price of the target product. The evaluation information is the comments of consumers on the target product.
[0059] As described in the following steps, based on the name information, the category information, the appearance information, the price information and the evaluation information, the substitute products of each target product among the remaining target products and the substitution rate of the substitute products for the target product are determined.
[0060] As a first example, a comprehensive feature vector of each target product is generated based on the name information, the category information, the appearance information, the price information and the evaluation information; the comprehensive feature vector of the target product and the comprehensive feature vector of each of the remaining target products are processed using a pre-built substitution rate evaluation model to obtain the substitution rate of each of the remaining target products for the target product; the remaining target products whose substitution rate for the target product is greater than a preset value are set as substitute products for the target product. It can be understood that the substitution rate evaluation model is a deep learning model trained based on historical sample data; wherein the historical sample data includes platform display information of multiple sample products and the substitution rate between each of the sample products.
[0061] By using a deep learning model to evaluate the substitution rate between commodities, the complex nonlinear relationship between features can be fully captured, achieving higher evaluation accuracy.
[0062] As a second example, based on the name information, the category information, the appearance information, the price information and the evaluation information, the cosine similarity between each of the remaining target products and the target product is determined; the remaining target products whose cosine similarity with the target product is greater than a preset value are set as substitute products for the target product, and based on the cosine similarity, the replacement rate of the substitute products for the target product is determined.
[0063] By using cosine similarity to calculate the substitution rate between products, it has low computational complexity and is easy to quickly deploy and implement.
[0064] As a third example, the Euclidean distance between each of the remaining target products and the target product is determined based on the name information, the category information, the appearance information, the price information and the evaluation information; the remaining target products whose Euclidean distance to the target product is less than a preset value are set as substitute products for the target product, and the substitution rate of the substitute products for the target product is determined based on the Euclidean distance.
[0065] By using the Euclidean distance to calculate the substitution rate between commodities, the method has low computational complexity and is easy to deploy and implement quickly.
[0066] In one embodiment of the present application, the specific process of "generating a comprehensive feature vector for each target product based on the name information, the category information, the appearance information, the price information and the evaluation information" in the first example can be further explained in combination with the following description.
[0067] As described in the following steps, based on the name information, the category information, the appearance information, the price information and the evaluation information, a name feature vector, a category feature vector, an appearance feature vector, a price feature vector and an evaluation feature vector are generated for each target product.
[0068] Use text embedding technology (such as TF-IDF, Word2Vec and BERT, etc.) to convert the name information into a name feature vector, convert the category information into a category feature vector, and convert the evaluation information into an evaluation feature vector; use a color histogram to convert the appearance information into an appearance feature vector; and convert the evaluation information into an evaluation feature vector through standardization.
[0069] As described in the following steps, a comprehensive feature vector of each target product is generated based on the name feature vector, the category feature vector, the appearance feature vector, the price feature vector and the evaluation feature vector.
[0070] The name feature vector, the category feature vector, the appearance feature vector, the price feature vector and the evaluation feature vector are concatenated to obtain the comprehensive feature vector.
[0071] In one embodiment of the present application, the substitution rate evaluation model includes a feature fusion layer and a multi-layer fully connected network; wherein the multi-layer fully connected network includes multiple hidden layers and a discard layer connected after each of the hidden layers; the specific process of "using a pre-built substitution rate evaluation model to process the comprehensive feature vector of the target product and the comprehensive feature vector of each of the remaining target products to obtain the substitution rate of each of the remaining target products for the target product" in the first example can be further explained in combination with the following description.
[0072] As described in the following steps, the comprehensive feature vector of the target product and the comprehensive feature vector of each of the remaining target products are input into the feature fusion layer; wherein the feature fusion layer is used to fuse the comprehensive feature vector of the target product with the comprehensive feature vector of each of the remaining target products to obtain a fused feature vector, and input the fused feature vector into the multi-layer fully connected network; the multi-layer fully connected network is used to perform multi-level mapping on the fused feature vector through linear transformation and nonlinear activation function to obtain the substitution rate of each of the remaining target products for the target product.
[0073] The substitution rate evaluation model includes a feature fusion layer, a multi-layer fully connected network and an output layer; the feature fusion layer is used to fuse the two comprehensive feature vectors by addition, multiplication, concatenation or other attention mechanisms to obtain a fused feature vector; the multi-layer fully connected network includes multiple hidden layers and a dropout layer connected after each hidden layer, the hidden layer is used to realize linear transformation through a weight matrix and realize nonlinear transformation through ReLU or other activation functions, and the dropout layer is used to prevent overfitting by randomly discarding some neurons; the output layer is used to output a scalar representing the substitution rate of the target product for the target product, and its activation function is Sigmoid or Softmax, which is used to map the output to between 0 and 1.
[0074] In the first specific implementation, the substitution relationship between 6 target products sold on an e-commerce platform at the same time, as well as the initial inventory, predicted sales volume and sales revenue of each target product in the next month are obtained. The results are shown in Table 1:
[0075] Table 1 Product information table
[0076]
[0077]
[0078] As described in step S120, the replacement revenue of the replacement product corresponding to each target product for the target product is determined according to the sales revenue of the replacement product corresponding to the target product and the replacement rate of the replacement product for the target product.
[0079] The substitution benefit of the substitute product corresponding to the target product to the target product is calculated according to the following formula:
[0080]
[0081] Among them, OG is the substitution benefit of all the alternative commodities to the target commodity, n is the number of the alternative commodities, G i is the sales revenue of the ith substitute product, P i is the substitution rate of the ith substitute product for the target product.
[0082] In the second specific implementation, continuing the first specific implementation, according to the sales revenue of the substitute product corresponding to the target product and the substitution rate of the substitute product for the target product, the substitution revenue of the substitute product corresponding to each target product for the target product is determined, and the result is shown in Table 2:
[0083] Table 2 Commodity Substitution Income Statement
[0084] Replacement income (yuan) Product A 0 Product B 9.6 Product C 7.8 Product D 1.55 Product E 0 Product F 3.6
[0085] As described in step S130, the target commodity whose sales revenue is greater than or equal to the replacement revenue of the corresponding replacement commodity to the target commodity is set as a high-revenue commodity, and the remaining target commodities are set as low-revenue commodities.
[0086] The target commodity whose sales revenue (i.e., G) is greater than or equal to the corresponding substitution revenue (i.e., OG) of the substitution commodity for the target commodity is set as a high-revenue commodity, and the remaining target commodities are set as low-revenue commodities. It can be understood that when the inventory of the high-revenue commodity is zero, the revenue generated by the lost sales of the high-revenue commodity is greater than or equal to the revenue generated by the sales of the substitution commodity converted from the lost sales of the high-revenue commodity, and when the inventory of the low-revenue commodity is zero, the revenue generated by the lost sales of the low-revenue commodity is less than the revenue generated by the sales of the substitution commodity converted from the lost sales of the low-revenue commodity.
[0087] In the third specific implementation, continuing the second specific implementation, the target product whose sales revenue is greater than or equal to the replacement revenue of the corresponding replacement product for the target product is set as a high-revenue product, and the remaining target products are set as low-revenue products. The results are shown in Table 3:
[0088] Table 3 Commodity classification results
[0089] Classification results Product A High-yield commodities Product B Low-yield products Product C Low-yield products Product D High-yield commodities Product E High-yield commodities Product F Low-yield products
[0090] As described in step S140, the conversion demand of each of the high-yield commodities is determined based on the initial inventory and the predicted sales volume of the low-yield commodities, the substitute commodities corresponding to the low-yield commodities, and the substitution rate of the substitute commodities for the low-yield commodities.
[0091] For the low-yield goods, the actual inventory is not increased. When there is overflow demand, the overflow demand is transferred to the replacement goods according to the replacement rate (for the low-yield goods whose demand is transferred, when there is overflow demand, the overflow demand is also transferred to the replacement goods according to the replacement rate) until the overflow demand is less than the update threshold.
[0092] In one embodiment of the present application, the specific process of "determining the conversion demand for each of the high-yield commodities based on the initial inventory and the predicted sales volume of the low-yield commodities, the substitute commodities corresponding to the low-yield commodities, and the substitution rate of the substitute commodities for the low-yield commodities" in step S140 can be further explained in combination with the following description.
[0093] As described in the following steps, when the difference between the sum of the predicted sales volume and the current conversion demand of the low-yield commodity and the current reference inventory is greater than an update threshold, the current conversion demand of the substitute commodity corresponding to the low-yield commodity is updated based on the predicted sales volume, the current conversion demand and the current reference inventory of the low-yield commodity, and the substitution rate of the substitute commodity corresponding to the low-yield commodity for the low-yield commodity, and the current reference inventory of the low-yield commodity is updated based on the predicted sales volume and the current conversion demand of the low-yield commodity; wherein the starting value of the current reference inventory of the low-yield commodity is the initial inventory.
[0094] The inspection procedure is performed on all the target commodities in a loop in sequence: for the low-revenue commodity, the current overflow demand (EL) of the low-revenue commodity is calculated according to the following formula: r ): the current overflow demand (EL) of the low-yield commodity r ) = the predicted sales volume (SL) of the low-profit commodity + the current conversion demand (TL) of the low-profit commodity r )-Current reference inventory of the low-yield item (BL r ); when the current overflow demand (EL r ) is greater than the update threshold (E mim), the current conversion demand (TO) of the substitute product corresponding to the low-yield product is updated according to the following formula r ): The updated conversion demand of the alternative product (TO r+1 ) = the current conversion demand of the alternative commodity (TO r ) + the current overflow demand for the low-yield commodity (EL r )×the substitution rate of the substitute product for the low-yield product (P i ), and update the current reference inventory (BL) of the low-yield commodity according to the following formula: r ): Updated reference inventory of the low-yield commodity (BL r+1 ) = the current reference inventory of the low-yield commodity (BL r ) + the current overflow demand for the low-yield commodity (EL r ) to meet the current overflow demand for the low-yield commodity (EL r ) is cleared; wherein, the starting value (TL1) of the current conversion demand of the low-yield commodity is 0; the starting value (BL1) of the current reference inventory of the low-yield commodity is the initial inventory (BL0) of the low-yield commodity; the starting value (TH1) of the current conversion demand of the high-yield commodity is 0; until no update operation occurs in a round of inspection, the inspection procedure is terminated.
[0095] In the fourth specific implementation, continuing the third specific implementation, the conversion demand of each high-yield commodity is determined according to the initial inventory and the predicted sales volume of the low-yield commodity, the substitute commodity corresponding to the low-yield commodity, and the substitution rate of the substitute commodity to the low-yield commodity. The results are shown in Table 4:
[0096] Table 4 High-yield commodity conversion demand table
[0097] Conversion demand (end value) (pieces) Product A 400 Product D 400 Product E 0
[0098] As described in step S150, the inventory increment of each high-profit commodity is determined based on the initial inventory of the high-profit commodity, the predicted sales volume and the conversion demand.
[0099] For the high-yield commodities, actual inventory is increased as needed, and when there is overflow demand, the inventory level is updated until the overflow demand is less than the update threshold.
[0100] In one embodiment of the present application, the specific process of "determining the inventory increment of each high-yield commodity based on the initial inventory of the high-yield commodity, the predicted sales volume and the conversion demand" in step S150 can be further explained in combination with the following description.
[0101] As described in the following steps, when the difference between the sum of the predicted sales volume and the current conversion demand of the high-yield commodity and the current expected inventory is greater than an update threshold, the current expected inventory of the high-yield commodity is updated according to the predicted sales volume and the current conversion demand of the high-yield commodity; wherein the starting value of the current expected inventory of the high-yield commodity is the initial inventory, and the difference between the ending value and the initial inventory is the inventory increment.
[0102] The inspection procedure is performed on all the target commodities in turn in a loop: For the high-yield commodity, the current overflow demand (EH r ): the current overflow demand for the high-yield commodity (EH r ) = the predicted sales volume of the high-yield commodity (SH) + the current conversion demand of the high-yield commodity (TH r )-the current expected inventory of the high-yield commodity (BH r ); when the current overflow demand (EH r ) is greater than the update threshold (E min ), the current expected inventory (BH r ): The updated expected inventory of the high-yield commodity (BH r+1 ) = the current expected inventory of the high-yield commodity (BH r ) + the current overflow demand for the high-yield commodity (EH r ) to meet the current overflow demand for the high-yield commodity (EH r ) is emptied; wherein, the starting value (TH1) of the current conversion demand of the high-yield commodity is 0; the starting value (BH1) of the current expected inventory of the high-yield commodity is the initial inventory (BH0); the ending value (BH e ) and the initial inventory (BH0) is the inventory increment; until no update operation occurs in a round of inspection, the inspection procedure is terminated.
[0103] In the fifth specific implementation, continuing with the fourth specific implementation, the inventory increment of each high-yield commodity is determined according to the initial inventory, the predicted sales volume and the conversion demand of the high-yield commodity. The result is shown in Table 5:
[0104] Table 5 High-yield commodity inventory increment table
[0105] Inventory increment (pieces) Product A 0 Product D 1900 Product E 2000
[0106] Reference Figure 2In one embodiment of the present application, an intelligent inventory management system is also provided, including:
[0107] The commodity information acquisition module 210 is used to acquire the substitution relationship between multiple target commodities, and the initial inventory, predicted sales volume and sales revenue of each target commodity within a preset period; wherein the substitution relationship includes the substitute commodities of each target commodity among the other target commodities, and the substitution rate of the substitute commodities to the target commodity;
[0108] The substitution revenue determination module 220 is used to determine the substitution revenue of the substitute product corresponding to each target product for the target product according to the sales revenue of the substitute product corresponding to the target product and the substitution rate of the substitute product for the target product;
[0109] The target commodity classification module 230 is used to set the target commodity whose sales revenue is greater than or equal to the substitute revenue of the corresponding substitute commodity to the target commodity as a high-profit commodity, and set the rest of the target commodities as low-profit commodities;
[0110] The conversion demand determination module 240 is used to determine the conversion demand of each of the high-yield commodities according to the initial inventory and the predicted sales volume of the low-yield commodities, the substitute commodities corresponding to the low-yield commodities, and the substitution rate of the substitute commodities to the low-yield commodities;
[0111] The inventory increment determination module 250 is used to determine the inventory increment of each high-profit commodity according to the initial inventory, the predicted sales volume and the conversion demand of the high-profit commodity.
[0112] As for the system embodiment, since it is basically similar to the method embodiment, the description is relatively simple, and the relevant parts can be referred to the description of the method embodiment.
[0113] In one embodiment of the present application, a computer device is also provided, including a processor, a memory, and a computer program stored in the memory and capable of running on the processor. When the computer program is executed by the processor, the intelligent inventory management method as described in any of the above embodiments can be implemented.
[0114] In one embodiment of the present application, a computer-readable storage medium is further provided, on which a computer program is stored. When the computer program is executed by a processor, the intelligent inventory management method as described in any of the above embodiments can be implemented.
[0115] Each embodiment in this specification is described in a progressive manner, and each embodiment focuses on the differences from other embodiments. The same or similar parts between the embodiments can be referenced to each other.
[0116] Although the preferred embodiments of the present application have been described, those skilled in the art may make additional changes and modifications to these embodiments once they have learned the basic creative concept. Therefore, the appended claims are intended to be interpreted as including the preferred embodiments and all changes and modifications that fall within the scope of the embodiments of the present application.
[0117] Finally, it should be noted that, in this article, relational terms such as first and second, etc. are only used to distinguish one entity or operation from another entity or operation, and do not necessarily require or imply any such actual relationship or order between these entities or operations. Moreover, the terms "include", "comprise" or any other variants thereof are intended to cover non-exclusive inclusion, so that a process, method, article or terminal device including a series of elements includes not only those elements, but also other elements not explicitly listed, or also includes elements inherent to such process, method, article or terminal device. In the absence of further restrictions, the elements defined by the sentence "comprise a ..." do not exclude the existence of other identical elements in the process, method, article or terminal device including the elements.
[0118] The intelligent inventory management method and system provided by the present application are introduced in detail above. Specific examples are used in this article to illustrate the principles and implementation methods of the present application. The description of the above embodiments is only used to help understand the method of the present application and its core idea; at the same time, for general technical personnel in this field, according to the idea of the present application, there will be changes in the specific implementation method and application scope. In summary, the content of this specification should not be understood as a limitation on the present application.
Claims
1. An intelligent inventory management method, characterized in that: include: Obtaining a substitution relationship between a plurality of target commodities, and an initial inventory, predicted sales volume, and sales revenue of each of the target commodities within a preset period; wherein the substitution relationship includes substitute commodities of each of the target commodities among the remaining target commodities, and a substitution rate of the substitute commodities to the target commodity; Determine the replacement revenue of the replacement product corresponding to each target product for the target product according to the sales revenue of the replacement product corresponding to the target product and the replacement rate of the replacement product for the target product; The target commodity whose sales revenue is greater than or equal to the replacement revenue of the corresponding replacement commodity to the target commodity is set as a high-revenue commodity, and the remaining target commodities are set as low-revenue commodities; Determining the conversion demand of each of the high-yield commodities according to the initial inventory and the predicted sales volume of the low-yield commodities, the substitute commodities corresponding to the low-yield commodities, and the substitution rate of the substitute commodities for the low-yield commodities; The inventory increment of each of the high-yield commodities is determined according to the initial inventory of the high-yield commodities, the predicted sales volume, and the conversion demand.
2. The method according to claim 1, characterized in that The step of obtaining the substitution relationship between the multiple target commodities includes: Obtaining platform display information of each target product; wherein the platform display information includes name information, category information, appearance information, price information and evaluation information; According to the name information, the category information, the appearance information, the price information and the evaluation information, the substitute products of each of the target products among the remaining target products and the substitution rate of the substitute products for the target product are determined.
3. The method according to claim 2, characterized in that The step of determining the substitute products of each target product among the remaining target products and the substitution rate of the substitute products for the target product based on the name information, the category information, the appearance information, the price information and the evaluation information comprises: Generate a comprehensive feature vector for each of the target products based on the name information, the category information, the appearance information, the price information, and the evaluation information; Using a pre-built substitution rate evaluation model, the comprehensive feature vector of the target product and the comprehensive feature vectors of each of the remaining target products are processed to obtain a substitution rate of each of the remaining target products for the target product; The remaining target commodities whose replacement rates for the target commodity are greater than a preset value are set as substitute commodities for the target commodity.
4. The method according to claim 3, characterized in that The step of generating a comprehensive feature vector for each target product based on the name information, the category information, the appearance information, the price information and the evaluation information comprises: Generate a name feature vector, a category feature vector, an appearance feature vector, a price feature vector and an evaluation feature vector for each target product according to the name information, the category information, the appearance information, the price information and the evaluation information; A comprehensive feature vector of each of the target commodities is generated based on the name feature vector, the category feature vector, the appearance feature vector, the price feature vector and the evaluation feature vector.
5. The method according to claim 3, characterized in that: The substitution rate evaluation model includes a feature fusion layer and a multi-layer fully connected network; wherein the multi-layer fully connected network includes multiple hidden layers and a discard layer connected after each of the hidden layers; the step of using the pre-built substitution rate evaluation model to process the comprehensive feature vector of the target product and the comprehensive feature vector of each of the remaining target products to obtain the substitution rate of each of the remaining target products for the target product includes: The comprehensive feature vector of the target product and the comprehensive feature vector of each of the remaining target products are input into the feature fusion layer; wherein the feature fusion layer is used to fuse the comprehensive feature vector of the target product with the comprehensive feature vector of each of the remaining target products to obtain a fused feature vector, and input the fused feature vector into the multi-layer fully connected network; the multi-layer fully connected network is used to perform multi-level mapping on the fused feature vector through linear transformation and nonlinear activation function to obtain the substitution rate of each of the remaining target products for the target product.
6. The method according to claim 1, characterized in that The step of determining the conversion demand of each of the high-yield commodities based on the initial inventory and the predicted sales volume of the low-yield commodities, the substitute commodities corresponding to the low-yield commodities, and the substitution rate of the substitute commodities for the low-yield commodities comprises: When the difference between the sum of the predicted sales volume and the current conversion demand of the low-yield commodity and the current reference inventory is greater than an update threshold, the current conversion demand of the substitute commodity corresponding to the low-yield commodity is updated according to the predicted sales volume, the current conversion demand and the current reference inventory of the low-yield commodity, and the substitution rate of the substitute commodity corresponding to the low-yield commodity for the low-yield commodity, and the current reference inventory of the low-yield commodity is updated according to the predicted sales volume and the current conversion demand of the low-yield commodity; wherein the starting value of the current reference inventory of the low-yield commodity is the initial inventory.
7. The method according to claim 1, characterized in that The step of determining the inventory increment of each high-yield commodity according to the initial inventory of the high-yield commodity, the predicted sales volume and the conversion demand comprises: When the difference between the sum of the predicted sales volume and the current conversion demand of the high-yield commodity and the current expected inventory is greater than an update threshold, the current expected inventory of the high-yield commodity is updated according to the predicted sales volume and the current conversion demand of the high-yield commodity; wherein the starting value of the current expected inventory of the high-yield commodity is the initial inventory, and the difference between the ending value and the initial inventory is the inventory increment.
8. An intelligent inventory management system, characterized in that: include: A commodity information acquisition module, used to acquire the substitution relationship between multiple target commodities, and the initial inventory, predicted sales volume and sales revenue of each target commodity within a preset period; wherein the substitution relationship includes the substitute commodities of each target commodity among the remaining target commodities, and the substitution rate of the substitute commodities to the target commodity; A substitution revenue determination module, configured to determine the substitution revenue of the substitute commodity corresponding to each target commodity for the target commodity according to the sales revenue of the substitute commodity corresponding to the target commodity and the substitution rate of the substitute commodity for the target commodity; a target commodity classification module, configured to set the target commodity whose sales revenue is greater than or equal to the substitute revenue of the corresponding substitute commodity to the target commodity as a high-revenue commodity, and set the remaining target commodities as low-revenue commodities; A conversion demand determination module, configured to determine the conversion demand of each of the high-yield commodities according to the initial inventory and the predicted sales volume of the low-yield commodities, the substitute commodities corresponding to the low-yield commodities, and the substitution rate of the substitute commodities for the low-yield commodities; The inventory increment determination module is used to determine the inventory increment of each high-profit commodity according to the initial inventory, the predicted sales volume and the conversion demand of the high-profit commodity.
9. A computer device, characterized in that: The method comprises a processor, a memory, and a computer program stored in the memory and capable of running on the processor, wherein the computer program implements the method according to any one of claims 1 to 7 when executed by the processor.
10. A computer-readable storage medium, characterized in that: The computer-readable storage medium stores a computer program, and when the computer program is executed by a processor, the method according to any one of claims 1 to 7 is implemented.