Dynamic Inventory Management System and Management Method

Abstract

The present application provides a dynamic inventory management system and management method, which relates to the field of inventory management technology, including: an inventory impact relationship analysis module, which is used to obtain inventory management factors; a scale interaction module, which is used to obtain the interaction characteristics of each time scale and each space scale; a demand setting module, which is used to set the maximum inventory level, the minimum inventory level, and dynamic management requirements; an inventory change analysis module, which is used to analyze inventory changes according to the interaction characteristics of each time scale and each space scale, and obtain time scale control parameters and space scale control parameters. This application can solve the technical problems in the prior art that inventory management is inefficient and lacks flexibility due to neglect of real-time adjustment of dynamic management requirements, achieve the goal of dynamic management requirements, and achieve the technical effect of improving the flexibility and response speed of inventory management.

Description

Technical Field

[0001] The present application relates to the technical field of inventory management, and in particular to a dynamic inventory management system and management method. Background Art

[0002] In modern supply chain management, the efficiency of inventory management directly affects the company's operating costs, customer satisfaction and market competitiveness. Traditional inventory management methods are often based on quantitative analysis of historical sales data and experience, but with the rapid changes in market demand and increasing uncertainty, a single static management method can no longer meet the needs of companies in a complex environment.

[0003] Currently, many companies still rely on manual experience and simple forecasting models to make inventory decisions, lacking systematic and scientific management tools. This situation has led to a series of problems such as low inventory management efficiency and insufficient inventory management flexibility, which has seriously affected the market competitiveness of companies. In addition, existing technologies often ignore the real-time adjustment of dynamic management needs and fail to make full use of big data and intelligent analysis technologies to optimize inventory levels. Summary of the invention

[0004] The purpose of this application is to provide a dynamic inventory management system and management method to solve the technical problems in the prior art that inventory management efficiency is low and flexibility is insufficient due to neglect of real-time adjustment of dynamic management needs.

[0005] In view of the above problems, the present application provides a dynamic inventory management system and management method.

[0006] In a first aspect, the present application provides a dynamic inventory management system, including: an inventory impact relationship analysis module, which is used to analyze inventory impact relationships and obtain inventory management factors, where the inventory management factors are factors that affect dynamic inventory adjustment; a scale interaction module, which is used to collect real-time data and historical data based on the inventory management factors, where the real-time data and historical data both include data at different times and in different regions, and use the real-time data and historical data to perform multi-time scale and spatial scale interactions to obtain interaction features of each time scale and interaction features of each spatial scale, where the time scale interaction features include factor features of different time period scales, and the spatial scale interaction features are factor features of different spatial range scales; a demand setting module, which is used to set a maximum inventory level, a minimum inventory level, and dynamic management requirements; an inventory change analysis module, which is used to use the maximum inventory level and the minimum inventory level as constraints and the dynamic management requirements as a target, perform inventory change analysis according to the interaction features of each time scale and the interaction features of each spatial scale, and obtain time scale control parameters and spatial scale control parameters.

[0007] In a second aspect, the present application further provides a dynamic inventory management method, which is implemented by the dynamic inventory management system described in the first aspect, wherein the dynamic inventory management method includes: analyzing the inventory impact relationship to obtain an inventory management factor, wherein the inventory management factor is a factor that affects the dynamic adjustment of inventory; based on the inventory management factor, collecting real-time data and historical data, wherein the real-time data and historical data both include data from different times and different regions, and using the real-time data and historical data to perform multi-time scale and spatial scale interactions to obtain interaction characteristics of each time scale and interaction characteristics of each spatial scale, wherein the time scale interaction characteristics include factor characteristics of different time period scales, and the spatial scale interaction characteristics are factor characteristics of different spatial range scales; setting a maximum inventory level, a minimum inventory level, and dynamic management requirements; taking the maximum inventory level and the minimum inventory level as constraints and the dynamic management requirements as a goal, performing inventory change analysis according to the interaction characteristics of each time scale and the interaction characteristics of each spatial scale to obtain time scale control parameters and spatial scale control parameters.

[0008] One or more technical solutions provided in this application have at least the following technical effects or advantages:

[0009] The inventory impact relationship analysis module is used to analyze the inventory impact relationship and obtain the inventory management factor, which is the factor that affects the dynamic adjustment of inventory; the scale interaction module is used to collect real-time data and historical data based on the inventory management factor, and the real-time data and historical data both include data at different times and in different regions, and use the real-time data and historical data to perform multi-time scale and spatial scale interaction to obtain the interaction characteristics of each time scale and each spatial scale, and the time scale interaction characteristics include the factor characteristics of different time period scales, and the spatial scale interaction characteristics are the factor characteristics of different spatial range scales; the demand setting module is used to set the maximum inventory level, the minimum inventory level, and the dynamic management demand; the inventory change analysis module is used to take the maximum inventory level and the minimum inventory level as the constraint conditions, the dynamic management demand as the goal, and perform inventory change analysis according to the interaction characteristics of each time scale and each spatial scale interaction characteristics to obtain the time scale control parameter and the spatial scale control parameter. Through the dynamic inventory management system, the comprehensive analysis of various real-time data and historical data, combined with the time scale and spatial scale characteristics, makes a rapid response to changes in market demand, realizes the goal of dynamic management demand, and achieves the technical effect of improving the flexibility and response speed of inventory management.

[0010] The above description is only an overview of the technical solution of the present application. In order to more clearly understand the technical means of the present application, it can be implemented according to the contents of the specification, and in order to make the above and other purposes, features and advantages of the present application more obvious and easy to understand, the specific implementation methods of the present application are specifically cited below. It should be understood that the content described in this section is not intended to identify the key or important features of the embodiments of the present application, nor is it intended to limit the scope of the present application. Other features of the present application will become easy to understand through the following description. BRIEF DESCRIPTION OF THE DRAWINGS

[0011] In order to more clearly illustrate the technical solutions in the present application or the prior art, the drawings required for use in the embodiments or the description of the prior art will be briefly introduced below. Obviously, the drawings in the following description are only exemplary, and for ordinary technicians in this field, other drawings can be obtained based on the provided drawings without paying any creative work.

[0012] Figure 1 This is a schematic diagram of the structure of the dynamic inventory management system for this application;

[0013] Figure 2 This is a flow chart of the dynamic inventory management method of this application.

[0014] Description of reference numerals:

[0015] Inventory impact relationship analysis module 11, scale interaction module 12, demand setting module 13, inventory change analysis module 14. DETAILED DESCRIPTION

[0016] This application solves the technical problems in the prior art of low inventory management efficiency and insufficient flexibility due to neglect of real-time adjustment of dynamic management needs by providing a dynamic inventory management system and management method, thereby achieving the goal of dynamic management needs and achieving the technical effect of improving the flexibility and response speed of inventory management.

[0017] Below, the technical solutions in the present application will be clearly and completely described with reference to the accompanying drawings. Obviously, the described embodiments are only part of the embodiments of the present application, rather than all of the embodiments of the present application. It should be understood that the present application is not limited to the example embodiments described herein. Based on the embodiments of the present application, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present application. It should also be noted that, for the convenience of description, only the parts related to the present application are shown in the accompanying drawings, rather than all of them.

[0018] For example, please refer to the attached Figure 1 , this application provides a dynamic inventory management system, including:

[0019] The inventory impact relationship analysis module 11 is used to analyze the inventory impact relationship and obtain inventory management factors, where the inventory management factors are factors that affect the dynamic adjustment of inventory.

[0020] Specifically, in a dynamic inventory management system, the first step is to analyze the inventory impact relationship to identify and obtain inventory management factors. Inventory management factors refer to factors that have a significant impact on the dynamic adjustment of inventory, including but not limited to demand fluctuations, supply chain delivery time, storage space limitations, capital occupation, etc. Specifically, data mining and statistical analysis techniques can be used to extract relevant information from historical sales data, market trends, and supplier delivery records. For example, by analyzing the sales data of the past six months, it is found that the demand for a certain product increases significantly in a specific season (such as holidays or promotions), which becomes a key inventory management factor. In addition, external factors, such as changes in consumer preferences, need to be considered. These factors may have an impact on different time scales and spatial scales. By constructing a factor feature matrix, these impacts can be quantified and the most important inventory management factors can be identified.

[0021] For example, in an experiment, by analyzing the sales data of 500 SKUs (stock keeping units) in different seasons, four main factors were identified: seasonal demand, promotional activities, inventory turnover rate and supply chain stability. The results showed that seasonal demand and promotional activities had an impact of 60% on inventory levels, which provided an important basis for subsequent dynamic inventory adjustments. The identification of inventory management factors not only provides data support for inventory decisions, but also helps managers optimize inventory strategies in actual operations to cope with market changes and improve operational efficiency.

[0022] The scale interaction module 12 is used to collect real-time data and historical data based on the inventory management factors, and the real-time data and historical data both include data from different times and different regions. The real-time data and historical data are used to perform multi-time scale and space scale interactions to obtain interaction features of each time scale and each space scale. The time scale interaction features include factor features of different time period scales, and the space scale interaction features are factor features of different spatial range scales.

[0023] Specifically, based on inventory management factors, real-time data and historical data are collected. Real-time data and historical data not only contain basic information such as sales volume and inventory levels, but also involve data from different times and regions to fully capture the factors affecting dynamic changes in inventory. Among them, real-time data refers to the inventory status, sales situation and market trends at the current moment, and historical data refers to the inventory status, sales situation and market trends in the past, which are usually collected and obtained through sensors or point-of-sale systems. For example, a retailer’s sales data on weekends can quickly reflect customer purchase behavior, while historical data covers sales records in the past few months or years. Historical data can provide long-term trends and patterns, providing background information for inventory management.

[0024] After data collection is completed, data analysis is performed using the interactive method of multiple time scales and spatial scales. Multi-time scale analysis is to analyze data according to different time periods (such as days, weeks, and months) to identify how inventory management factors affect inventory changes in these time periods to obtain interactive features of each time scale. Each time scale interactive feature includes inventory status, sales and market trends at different time points. Multi-time scale analysis can reveal sales peaks in specific periods. For example, the sales of a product during the Spring Festival may be 50% higher than usual. At the same time, the analysis of spatial scale interactive features focuses on the performance of different regions or markets, and obtains inventory status, sales and market trends in different regions as spatial scale interactive features. For example, by comparing sales data from different cities or regions, regional demand differences can be identified. In southern cities, the sales volume of a certain summer beverage when the temperature rises is significantly higher than that of northern cities. By analyzing these spatial scale data, it is possible to identify which regions are most sensitive to changes in inventory demand, so that corresponding inventory adjustments can be made.

[0025] Through multi-dimensional interactive analysis of real-time data and historical data, the interactive characteristics of each time scale and spatial scale can be extracted. These characteristics will provide strong data support for subsequent inventory decisions, enabling managers to flexibly respond to market changes and optimize inventory management strategies.

[0026] The demand setting module 13 is used to set the maximum inventory level, the minimum inventory level, and dynamically manage demand.

[0027] Specifically, the maximum inventory level refers to the upper limit of inventory that can be accommodated within a specific time. The maximum inventory level usually considers multiple factors, such as storage space, capital cost, and item turnover speed. For example, if a retailer has a storage space of 1,000 square meters and can accommodate 100 items per square meter, the maximum inventory level can be set to 100,000 items. The minimum inventory level is the lower limit to prevent supply interruptions caused by insufficient inventory. It is usually set based on factors such as order demand fluctuations, safety inventory experience values, and supply chain delivery time. For example, after analyzing historical sales data, the average daily sales volume of a product is 200 pieces, and the supplier's delivery cycle is 5 days. In order to ensure that sales are not affected by insufficient inventory, the minimum inventory level can be set to 1,000 pieces. Dynamic management demand refers to the demand for dynamic adjustment of inventory according to market changes and actual sales during the operation process. This demand needs to combine real-time data and historical data, and consider seasonal demand, promotional activities and other influencing factors. For example, if the demand for a product is expected to increase by 30% during the Spring Festival, the dynamic management demand needs to be adjusted in a timely manner.

[0028] The inventory change analysis module 14 is used to perform inventory change analysis based on the maximum inventory level and the minimum inventory level as constraints and the dynamic management demand as a goal, and to obtain time scale control parameters and space scale control parameters according to the interaction characteristics of each time scale and each space scale.

[0029] Specifically, the maximum inventory level and the minimum inventory level provide boundary conditions for inventory management. The maximum inventory level is the upper limit of inventory that the enterprise can bear, aiming to avoid inventory backlogs and financial risks; while the minimum inventory level is the lower limit to prevent supply disruptions due to insufficient inventory. For example, suppose that the maximum inventory level of a company is 80,000 pieces and the minimum inventory level is 10,000 pieces, which means that the inventory level must always be kept between these two values. Dynamic management of demand refers to the flexible adjustment of inventory targets based on changes in market demand and sales forecasts. For example, during the upcoming holidays, the demand for a certain product is expected to increase by 30%. Therefore, the company sets the dynamic management demand target for this product to 96,000 pieces (80,000 pieces × 1.3) to ensure that the upcoming high demand is met. Next, the inventory change analysis is performed based on the interactive features of each time scale and the interactive features of the spatial scale. The interactive features of the time scale may reflect the sales trend in a specific time period (such as weekends and holidays), while the interactive features of the spatial scale show the sales performance in different regions. For example, by analyzing historical data, it is found that sales during holidays are usually 150% of normal sales, while sales in the southern market are often 20% higher than those in the northern market. This information will be used to evaluate inventory change requirements under different time and space conditions. After completing the inventory change analysis, the time scale control parameters and spatial scale control parameters are obtained. The time scale control parameters are used to guide the timing of inventory replenishment, while the spatial scale control parameters indicate how to allocate inventory between different regions. For example, the analysis results show that in the two weeks before the holidays, the inventory in the southern market should be replenished 30% in advance, while the northern market can remain relatively stable and replenish 10%. Therefore: Time scale control parameters: The replenishment time of the southern market is set to two weeks in advance. Spatial scale control parameters: The replenishment volume of the southern market is 30% of the target inventory, and the replenishment volume of the northern market is 10%. Through the above analysis, enterprises can effectively formulate reasonable inventory replenishment plans to ensure that appropriate inventory levels are maintained during high demand periods, optimize inventory management, and improve the company's ability to respond to market fluctuations, ensuring that customer needs can be met in a timely manner.

[0030] Furthermore, the scale interaction module 12 in the dynamic inventory management system is also used for:

[0031] Data are extracted from the real-time data and historical data respectively according to the inventory management factors to construct various inventory management factor data sets; the historical projection relationship between preset time scales and spatial scales is identified according to the historical data; the inventory management factor data sets are relationally aligned and connected according to the temporal relationship and the regional distribution relationship, and an interactive projection relationship is established according to the temporal alignment relationship and the regional alignment relationship between the real-time data and the historical data; the interactive projection relationship and real-time data are used as input features, and the feature interaction space constructed based on the historical projection relationship is projected at multiple time scales and multiple spatial scales to obtain the interactive features of each time scale and the interactive features of each spatial scale.

[0032] Specifically, based on inventory management factors, firstly, information related to inventory management factors is extracted from real-time data and historical data to construct data sets of various inventory management factors. Inventory management factors may include sales volume, inventory turnover rate, seasonal demand, etc. For example, in a retail environment, real-time data may show that a certain product has sold 100 pieces in the past 24 hours, while historical data may show that the average sales volume of the product in the same time period is 80 pieces. Through this extraction, a set of data containing various factors, namely, data sets of various inventory management factors, can be formed for subsequent analysis. Next, the historical projection relationship between the preset time scale and the spatial scale is identified based on historical data. The preset time scale may include different cycles such as day, week, month, etc., while the spatial scale involves different regions or markets. For example, by analyzing the data of the past year, it is found that the sales volume in a certain region in summer is 70% higher than that in winter, which forms a projection relationship between time and space. On this basis, the inventory management factor data sets are aligned and connected according to the temporal relationship and regional distribution relationship. In simple terms, the temporal alignment relationship between real-time data and historical data is compared, such as by analyzing the sales volume changes of specific products at different times, so as to ensure the timeliness and accuracy of the data. At the same time, regional alignment involves the matching of sales data in different markets, ensuring that data for the same product in different regions can be effectively compared.

[0033] Through these alignments, an interactive projection relationship is established, that is, data from different time and space dimensions are integrated to form a comprehensive view. The interactive projection relationship and real-time data are used as input features, and multi-time scale and multi-space scale projection can be performed in the feature interaction space, which can reveal how the performance of inventory management factors changes under different time and space conditions, thereby obtaining interactive features of each time scale and spatial scale. For example, suppose that a specific product shows in past sales data that sales increased by 150% before and after holidays, while the average sales on non-holidays remained stable. Through multi-scale projection analysis, these sales change patterns can be identified, helping companies make more accurate predictions and adjustments in future inventory management. Such analysis not only improves the flexibility of inventory management, but also provides solid data support for rapid response to market demand.

[0034] Furthermore, the scale interaction module 12 in the dynamic inventory management system is also used for:

[0035] According to the preset time scale and space scale, scale adjacency relationship analysis is performed to establish an adjacency scale; according to the adjacency scale, a first scale feature and a second scale feature are obtained from the historical data, and the first scale feature and the second scale feature are adjacency scales; the first scale feature and the second scale feature are feature enhanced through the attention mechanism to obtain first scale enhanced features and second scale enhanced features; inner product calculation and a softmax calculation are performed on the first scale enhanced features and the second scale enhanced features to establish an adjacency matrix of the first scale and the second scale, and the adjacency matrix reflects the projection relationship between the first scale feature and the second scale feature, and the historical projection relationship is obtained.

[0036] Specifically, the process of identifying the historical projection relationship between the preset time scale and the spatial scale according to the historical data includes: first, according to the preset time scale (for example, day, week, month) and spatial scale (such as region, market), analysis is performed to determine which scales are adjacent, so as to construct the adjacent scale. The adjacent scale is the adjacent scale, including the scales adjacent in time and the scales adjacent in space. The purpose of this step is to lay the foundation for subsequent feature extraction and analysis. Among them, the preset time scale and spatial scale are set by the professional and technical personnel in this field in combination with the actual situation. For example, the historical time scale and spatial scale can be extracted from the historical data. Once the adjacent scale is established, the relevant feature data is extracted from the historical data. For example, the first scale feature may be the sales data of a certain area in the past month, and the second scale feature may be the average sales data of the area in the past year. By extracting these features, the relationship between the two can be identified, that is, the mutual influence between these adjacent scales.

[0037] Next, the first-scale features and the second-scale features are enhanced through the attention mechanism. The attention mechanism is a technology that imitates the human ability to concentrate and can effectively enhance the importance of features. In this process, the correlation between the first-scale features and the change in inventory demand is calculated to obtain the attention weights. Based on these weights, the features are enhanced to obtain the first-scale enhanced features and the second-scale enhanced features. For example, if the first-scale features show a significant increase in sales during a specific promotion period, the enhanced features will highlight this information to reflect its importance to inventory management.

[0038] Then, the inner product calculation is performed on the first scale enhanced features and the second scale enhanced features, and a softmax calculation is performed. The purpose of this process is to quantify the relationship between the first scale and the second scale and construct an adjacency matrix. Inner product calculation and a softmax calculation are commonly used operations in machine learning and deep learning, especially in feature processing and attention mechanisms. Inner product calculation is the process of multiplying two vectors and summing them. A softmax calculation is a function that converts any real vector into a probability distribution. Its output range is between 0 and 1, and the sum of all outputs is 1. In this application, the softmax calculation is used to convert the inner product result into an attention weight so that the weight of each feature can be used as a probability distribution. For example, assuming that a vector of inner product results is obtained, it can be converted into an attention weight vector through softmax. Through the above steps, the inner product calculation and softmax calculation can not only quantify the relationship between the feature and the demand change, but also convert it into a probability distribution that is easy to understand and apply. As an adjacency matrix, the adjacency matrix can reflect the projection relationship between features of different scales, indicating the mutual influence of these features under specific conditions. For example, if the sales volume of the first scale is significantly higher than that of the second scale during a promotion period, the adjacency matrix will show a strong correlation between them, and the historical projection relationship is obtained. The historical projection relationship can help better understand the performance of inventory management factors under different time and space conditions, thereby providing data support for subsequent dynamic inventory adjustments.

[0039] Furthermore, the scale interaction module 12 in the dynamic inventory management system is also used for:

[0040] The inventory management factors in the scale features and the corresponding inventory demand changes are represented as vectors; for each scale, the correlation between the feature vector and the inventory demand change is calculated to obtain the attention weight; based on the attention weight, the local features are enhanced to obtain the scale enhanced features.

[0041] Specifically, expressing the inventory management factors and the corresponding inventory demand changes in the scale features as vectors is a basic step to achieve feature enhancement. Each feature vector contains not only the current inventory management factors (such as sales volume, inventory turnover rate, etc.), but also the related inventory demand changes (such as the specific amount of increase or decrease). For example, it can be expressed as: feature vector = (sales volume, inventory turnover rate, demand change). Next, for each feature vector at each scale, calculate the correlation between it and the inventory demand change. Specifically, correlation analysis or similarity calculation can be used. Usually, existing methods such as dot product and cosine similarity can be used to quantify this correlation. Through calculation, the attention weight of each feature can be obtained, indicating the importance of each feature in predicting inventory demand changes. For example, if a feature (such as sales volume) shows a high correlation in affecting inventory demand changes, a higher attention weight will be assigned. Based on the attention weight, the local features are enhanced to obtain scale-enhanced features, that is, each feature vector is multiplied by its corresponding attention weight, so as to amplify those features that are highly correlated with demand changes, while suppressing features that are less correlated with demand changes. For example, assuming that the attention weight of sales volume is 0.7 and the weight of inventory turnover rate is 0.3, the feature vector is enhanced to obtain scale-enhanced features. Through this process, scale-enhanced features are obtained, which can more accurately reflect the impact of inventory management factors on inventory demand changes at different time and space scales. This not only improves the accuracy of inventory management, but also provides a more reliable data basis for subsequent decision-making. For example, in practical applications, if sales volume significantly affects demand changes in a certain period, the enhanced features will be able to better predict the future inventory demand of the product, thereby effectively avoiding out-of-stock or overstock situations.

[0042] Furthermore, the scale interaction module 12 in the dynamic inventory management system is also used for:

[0043] A scale axis is constructed according to the preset time scale and space scale; based on the scale axis, a scale window adjacency analysis is performed according to the time adjacency standard relationship and the space adjacency standard relationship to determine the scale adjacency density; when the scale adjacency density does not satisfy the adjacency standard scale relationship, a relationship inheritance scale is established based on the time adjacency standard relationship and the space adjacency standard relationship; the relationship inheritance scale is inserted into the scale axis as an auxiliary scale, the scale adjacency relationship is analyzed, and the adjacency scale is determined.

[0044] Specifically, a scale axis is created based on a preset time scale (such as day, week, month) and a spatial scale (such as region, market). This scale axis provides a framework for subsequent analysis, which is convenient for comparing and analyzing data from different time periods and regions. The time adjacency standard relationship and the spatial adjacency standard relationship are determined by professional and technical personnel in this field based on the experience value of the inventory management scale. For example, the consistency relationship of the time standard: day, week, month, quarter, half a year, and full year. The closer the adjacency, the more reliable the result of the mapping interaction. Therefore, according to the time adjacency standard relationship and the spatial adjacency standard relationship, the scale window adjacency analysis is performed on the preset time scale and spatial scale to determine the scale adjacency tightness. The scale adjacency tightness is the tightness value corresponding to each pair of adjacent time and space scales in the scale axis. The tightness value reflects the consistency degree of the change continuity of each pair of adjacent time and space scales with the time adjacency standard relationship and the spatial adjacency standard relationship. Specifically, the difference between each pair of adjacent time and space scales in the scale axis and the time adjacency standard relationship and the spatial adjacency standard relationship can be compared. The greater the difference, the lower the scale adjacency tightness. For example, if the preset time scale is directly from day to month, the corresponding scale adjacency is low and does not meet the adjacency standard scale relationship. A scale of one week can be inserted as an intermediate relationship connection scale.

[0045] Inserting the established relationship-based scale as an auxiliary scale into the scale axis can better analyze the scale adjacency relationship, which can not only supplement the original scale information, but also make the mutual influence between scales clearer. In this way, adjacent scales can be extracted in the updated scale axis to obtain adjacent scales, which is more accurate and efficient when processing data of different time and space. It not only enhances the rationality of the data, but also provides a more solid foundation for subsequent inventory decisions.

[0046] Furthermore, the demand setting module 13 in the dynamic inventory management system is also used for:

[0047] According to the storage space constraints, capital cost ratio and item turnover speed experience values, the maximum inventory level is obtained. The maximum inventory level is the upper limit of the capital risk constraint to avoid inventory backlogs; according to order demand fluctuations, supply chain delivery time, and safety stock experience values, the minimum inventory level is obtained. The minimum inventory level is the lower limit to prevent supply interruptions caused by insufficient inventory; according to the sales coverage area of ​​inventory management, the spatial scale management requirements are determined, and according to the capital turnover rate targets and sales cycles, the time scale management requirements are determined, wherein the spatial scale management requirements and time scale management requirements can be changed by setting dynamic change cycles.

[0048] Specifically, first, the maximum inventory level is calculated based on the experience of storage space constraints, capital cost and item turnover speed. The maximum inventory level is an important indicator to prevent inventory backlogs and is subject to capital risk constraints. For example, assuming that the effective storage space of a warehouse is 1,000 square meters and 100 items can be stored per square meter, then the theoretical maximum inventory is 100,000 items. However, considering the liquidity risk of funds, the company hopes to set the maximum inventory level to 80,000 items to avoid capital occupation due to excessive inventory. In addition, if the annual turnover rate of the commodity is 4 times, the item turnover speed per cycle can reach 20,000 pieces (80,000 pieces / 4), which shows that the company can effectively manage inventory and ensure capital liquidity. Next, the minimum inventory level is calculated based on the experience of order demand fluctuations, supply chain delivery time and safety stock. The minimum inventory level is the lower limit to prevent supply interruptions due to insufficient inventory. Assuming that historical data shows that the average daily sales volume of a product is 200 units and the supplier's delivery cycle is 5 days, the minimum inventory level can be set at 1,000 units (200 units / day × 5 days) to ensure that basic customer needs can still be met when there are delays in the supply chain.

[0049] After determining the maximum and minimum inventory levels, it is necessary to determine the spatial scale management requirements based on the sales coverage area of ​​inventory management. For example, if the product is mainly sold in the south and north, the company can evaluate the sales and market demand differences in each region and adjust the inventory configuration. If the sales volume in the south is higher, the company can consider increasing the inventory ratio in the south to 70%, while the northern region is set at 30%. In addition, according to the capital turnover rate target and sales cycle, the company also needs to determine the time scale management requirements. For example, the capital turnover rate target may be set to 4 times a year, and the sales cycle depends on the sales characteristics of the product. If the sales volume of a product is expected to increase by 30% during the promotion season, the company needs to adjust the inventory level before that period to ensure that it can meet the increased demand. The time scale management requirements can be set as a dynamic change cycle, such as evaluating and adjusting the inventory level every quarter to respond to market changes. Through the above steps, the company can effectively set the maximum and minimum inventory levels, and dynamically adjust the spatial scale and time scale management requirements according to market changes, which not only improves the efficiency of inventory management, but also enhances the company's ability to respond to market changes.

[0050] Furthermore, the inventory change analysis module 14 in the dynamic inventory management system is also used for:

[0051] Determine the safety stock according to the maximum inventory level and the minimum inventory level; perform dynamic time series decomposition and regional scope decomposition according to the dynamic management needs to obtain a dynamic target scale and establish an adjacent interaction path; within the safety stock range, perform scale feature mapping and matching on the interaction features of each time scale and the interaction features of each space scale according to the adjacent interaction path to obtain the inventory change features corresponding to the dynamic target scale; and obtain the time scale control parameters and space scale control parameters according to the inventory change features corresponding to the dynamic target scale by mapping according to the time series relationship and the regional scope relationship.

[0052] Specifically, safety stock refers to the amount of additional inventory retained to cope with demand fluctuations and supply chain uncertainties. For example, assuming that the maximum inventory level is 80,000 pieces and the minimum inventory level is 10,000 pieces, the safety stock is set to 5,000 pieces based on historical data and demand fluctuations. This setting ensures that when demand suddenly increases or supply is delayed, the company can still meet customer demand and reduce the risk of out-of-stock. Next, according to dynamic management needs, dynamic time series decomposition and regional scope decomposition are carried out to decompose the overall demand into demand characteristics of different time periods and regions. For example, if the analysis finds that the demand for a certain product is large during holidays, the company can decompose the overall sales demand in time before and after the holidays. At the same time, regional scope decomposition can determine the inventory demand of each region by analyzing the sales performance of different markets. Through the above decomposition, a dynamic target scale can be obtained. The dynamic target scale reflects the inventory demand under specific conditions in different time periods and regions. For example, if the demand during holidays is expected to increase by 30%, the dynamic target scale may be set to 96,000 pieces. After determining the dynamic target scale, an adjacent interaction path is established. The adjacent interaction path describes the mutual influence relationship between different time scales and spatial scales within the safety stock range. For example, by analyzing historical data, it is found that the sales trends of the southern and northern markets are similar. Therefore, a corresponding interaction path is established between these markets so that the inventory strategy of the north can be adjusted in time when the sales in the south increase. According to the adjacent interaction path, the scale feature mapping and matching of the interaction characteristics of each time scale and the interaction characteristics of the spatial scale are carried out, and the inventory characteristics of different time and space are compared with each other to identify the key factors affecting inventory changes. For example, if it is found that the sales in the south increase by 20% during a specific promotion period, while the sales in the north are stable during the same period, the enterprise needs to give priority to the inventory adjustment in the southern market. Finally, based on the inventory change characteristics corresponding to the dynamic target scale, the time scale control parameters and the spatial scale control parameters are obtained through the mapping of the time series relationship and the regional scope relationship. For example, if the inventory change characteristics in the south indicate that the stock should be replenished in advance before the holidays, the time scale control parameter can be set to replenish one week in advance, while the spatial scale control parameter may point to increasing the inventory in the southern market by 30%, which enhances the flexibility and responsiveness of inventory management and ensures that the enterprise can maintain its competitiveness in the rapidly changing market.

[0053] In summary, the dynamic inventory management system provided by this application has the following technical effects:

[0054] The inventory impact relationship analysis module is used to analyze the inventory impact relationship and obtain the inventory management factor, which is the factor that affects the dynamic adjustment of inventory; the scale interaction module is used to collect real-time data and historical data based on the inventory management factor, and the real-time data and historical data both include data at different times and in different regions, and use the real-time data and historical data to perform multi-time scale and spatial scale interaction to obtain the interaction characteristics of each time scale and each spatial scale, and the time scale interaction characteristics include the factor characteristics of different time period scales, and the spatial scale interaction characteristics are the factor characteristics of different spatial range scales; the demand setting module is used to set the maximum inventory level, the minimum inventory level, and the dynamic management demand; the inventory change analysis module is used to take the maximum inventory level and the minimum inventory level as the constraint conditions, the dynamic management demand as the goal, and perform inventory change analysis according to the interaction characteristics of each time scale and each spatial scale interaction characteristics to obtain the time scale control parameter and the spatial scale control parameter. Through the dynamic inventory management system, the comprehensive analysis of various real-time data and historical data, combined with the time scale and spatial scale characteristics, makes a rapid response to changes in market demand, realizes the goal of dynamic management demand, and achieves the technical effect of improving the flexibility and response speed of inventory management.

[0055] Embodiment 2: Based on the same inventive concept as the dynamic inventory management system in the aforementioned embodiment 1, this application also provides a dynamic inventory management method, please refer to the attached Figure 2 , the dynamic inventory management method comprises:

[0056] The inventory impact relationship is analyzed to obtain inventory management factors, which are factors that affect the dynamic adjustment of inventory.

[0057] Based on the inventory management factors, real-time data and historical data are collected, and the real-time data and historical data both include data from different times and different regions. The real-time data and historical data are used to perform multi-time scale and spatial scale interactions to obtain interaction characteristics of each time scale and interaction characteristics of each spatial scale. The time scale interaction characteristics include factor characteristics of different time period scales, and the spatial scale interaction characteristics are factor characteristics of different spatial range scales.

[0058] Set maximum and minimum stock levels, and dynamically manage demand.

[0059] With the maximum inventory level and the minimum inventory level as constraints and the dynamic management demand as a goal, inventory change analysis is performed according to the interaction characteristics of each time scale and each space scale, and time scale control parameters and space scale control parameters are obtained.

[0060] Furthermore, the real-time data and historical data are used to perform multi-time scale and space scale interactions to obtain interaction features of each time scale and each space scale, including:

[0061] Data are extracted from the real-time data and historical data respectively according to the inventory management factors to construct various inventory management factor data sets; the historical projection relationship between preset time scales and spatial scales is identified according to the historical data; the inventory management factor data sets are relationally aligned and connected according to the temporal relationship and the regional distribution relationship, and an interactive projection relationship is established according to the temporal alignment relationship and the regional alignment relationship between the real-time data and the historical data; the interactive projection relationship and real-time data are used as input features, and the feature interaction space constructed based on the historical projection relationship is projected at multiple time scales and multiple spatial scales to obtain the interactive features of each time scale and the interactive features of each spatial scale.

[0062] Further, identifying a historical projection relationship between a preset time scale and a spatial scale according to the historical data includes:

[0063] According to the preset time scale and space scale, scale adjacency relationship analysis is performed to establish an adjacency scale; according to the adjacency scale, a first scale feature and a second scale feature are obtained from the historical data, and the first scale feature and the second scale feature are adjacency scales; the first scale feature and the second scale feature are feature enhanced through the attention mechanism to obtain first scale enhanced features and second scale enhanced features; inner product calculation and a softmax calculation are performed on the first scale enhanced features and the second scale enhanced features to establish an adjacency matrix of the first scale and the second scale, and the adjacency matrix reflects the projection relationship between the first scale feature and the second scale feature, and the historical projection relationship is obtained.

[0064] Further, the first scale feature and the second scale feature are respectively enhanced through the attention mechanism to obtain the first scale enhanced feature and the second scale enhanced feature, including:

[0065] The inventory management factors in the scale features and the corresponding inventory demand changes are represented as vectors; for each scale, the correlation between the feature vector and the inventory demand change is calculated to obtain the attention weight; based on the attention weight, the local features are enhanced to obtain the scale enhanced features.

[0066] Furthermore, the scale adjacency relationship analysis is performed according to the preset time scale and space scale to establish the adjacency scale, including:

[0067] A scale axis is constructed according to the preset time scale and space scale; based on the scale axis, a scale window adjacency analysis is performed according to the time adjacency standard relationship and the space adjacency standard relationship to determine the scale adjacency density; when the scale adjacency density does not satisfy the adjacency standard scale relationship, a relationship inheritance scale is established based on the time adjacency standard relationship and the space adjacency standard relationship; the relationship inheritance scale is inserted into the scale axis as an auxiliary scale, the scale adjacency relationship is analyzed, and the adjacency scale is determined.

[0068] Furthermore, the setting of the maximum inventory level, the minimum inventory level, and the dynamic management requirements include:

[0069] According to the storage space constraints, capital cost ratio and item turnover speed experience values, the maximum inventory level is obtained. The maximum inventory level is the upper limit of the capital risk constraint to avoid inventory backlogs; according to order demand fluctuations, supply chain delivery time, and safety stock experience values, the minimum inventory level is obtained. The minimum inventory level is the lower limit to prevent supply interruptions caused by insufficient inventory; according to the sales coverage area of ​​inventory management, the spatial scale management requirements are determined, and according to the capital turnover rate targets and sales cycles, the time scale management requirements are determined, wherein the spatial scale management requirements and time scale management requirements can be changed by setting dynamic change cycles.

[0070] Furthermore, taking the maximum inventory level and the minimum inventory level as constraints and the dynamic management demand as a goal, an inventory change analysis is performed according to the interaction characteristics of each time scale and each space scale, and a time scale control parameter and a space scale control parameter are obtained, including:

[0071] Determine the safety stock according to the maximum inventory level and the minimum inventory level; perform dynamic time series decomposition and regional scope decomposition according to the dynamic management needs to obtain a dynamic target scale and establish an adjacent interaction path; within the safety stock range, perform scale feature mapping and matching on the interaction features of each time scale and the interaction features of each space scale according to the adjacent interaction path to obtain the inventory change features corresponding to the dynamic target scale; and obtain the time scale control parameters and space scale control parameters according to the inventory change features corresponding to the dynamic target scale by mapping according to the time series relationship and the regional scope relationship.

[0072] The various embodiments in this specification are described in a progressive manner, and each embodiment focuses on the differences from other embodiments. Figure 1The dynamic inventory management system and specific examples in the first embodiment are also applicable to the dynamic inventory management method of the present embodiment. Through the above detailed description of the dynamic inventory management system, those skilled in the art can clearly understand the dynamic inventory management method of the present embodiment, so for the sake of brevity of the specification, it will not be described in detail here. As for the method disclosed in the embodiment, since it corresponds to the system disclosed in the embodiment, the description is relatively simple, and the relevant parts can be referred to the description of the method part.

[0073] The above description of the disclosed embodiments enables those skilled in the art to implement or use the present application. Various modifications to these embodiments will be apparent to those skilled in the art, and the general principles defined herein may be implemented in other embodiments without departing from the spirit or scope of the present application. Therefore, the present application will not be limited to the embodiments shown herein, but will conform to the widest scope consistent with the principles and novel features disclosed herein.

[0074] Obviously, those skilled in the art can make various changes and modifications to the present application without departing from the spirit and scope of the present application. Thus, if these modifications and variations of the present application belong to the scope of the present application and its equivalent technology, the present application is also intended to include these modifications and variations.

Claims

1. Dynamic inventory management system, characterized by: The dynamic inventory management system includes: An inventory impact relationship analysis module is used to analyze inventory impact relationships and obtain inventory management factors, where the inventory management factors are factors that affect dynamic inventory adjustments; A scale interaction module is used to collect real-time data and historical data based on the inventory management factor, wherein the real-time data and historical data both include data at different times and in different regions, and to use the real-time data and historical data to perform multi-time scale and space scale interactions to obtain interaction features of each time scale and each space scale, wherein the time scale interaction features include factor features of different time period scales, and the space scale interaction features are factor features of different spatial range scales; Demand setting module, used to set maximum inventory level, minimum inventory level, and dynamically manage demand; An inventory change analysis module is used to perform inventory change analysis based on the maximum inventory level and the minimum inventory level as constraint conditions and the dynamic management demand as a target, and to obtain a time scale control parameter and a space scale control parameter according to the interaction characteristics of each time scale and each space scale, wherein the time scale control parameter is used to guide the timing of inventory replenishment, and the space scale control parameter indicates how to allocate inventory between different regions; Among them, taking the maximum inventory level and the minimum inventory level as constraint conditions and the dynamic management demand as the goal, the inventory change analysis is performed according to the interaction characteristics of each time scale and each space scale, and the time scale control parameters and space scale control parameters are obtained, including: Determine the safety stock based on the maximum stock level and the minimum stock level; According to the dynamic management demand, dynamic time series decomposition and regional scope decomposition are performed to obtain a dynamic target scale and establish an adjacent interaction path, wherein the dynamic target scale reflects the inventory demand under specific conditions in different time periods and regions, and the adjacent interaction path describes the mutual influence relationship between different time scales and spatial scales within the safety inventory range; Within the safety stock range, according to the adjacent interaction paths, scale feature mapping and matching are performed on the interaction features of each time scale and each space scale, and inventory features of different times and spaces are compared with each other to identify key factors affecting inventory changes and obtain inventory change features corresponding to the dynamic target scale; According to the inventory change characteristics corresponding to the dynamic target scale, the time scale control parameters and the space scale control parameters are obtained according to the time series relationship and the regional scope relationship mapping; The real-time data and historical data are used to perform multi-time scale and spatial scale interactions to obtain the interaction features of each time scale and each spatial scale, including: Extracting data from the real-time data and historical data according to the inventory management factors, respectively, to construct data sets for each inventory management factor; Identifying a historical projection relationship between a preset time scale and a spatial scale based on the historical data; Relational alignment connection is performed on the inventory management factor data sets according to the time series relationship and the regional distribution relationship, and an interactive projection relationship is established according to the time series alignment relationship and the regional alignment relationship between the real-time data and the historical data; The interactive projection relationship and real-time data are used as input features, and multi-time scale and multi-space scale projection is performed on the feature interaction space constructed based on the historical projection relationship to obtain the interactive features of each time scale and the interactive features of each space scale; Wherein, identifying the historical projection relationship between the preset time scale and the spatial scale according to the historical data includes: According to the preset time scale and space scale, conduct scale adjacency relationship analysis and establish adjacency scale; Acquire a first scale feature and a second scale feature from the historical data according to the adjacent scale, wherein the first scale feature and the second scale feature are two scales adjacent in time and space; The first scale feature and the second scale feature are enhanced respectively by using an attention mechanism to obtain a first scale enhanced feature and a second scale enhanced feature; Performing inner product calculation and one softmax calculation on the first scale enhanced features and the second scale enhanced features, establishing an adjacency matrix of the first scale and the second scale, wherein the adjacency matrix reflects the projection relationship between the first scale features and the second scale features, and obtaining the historical projection relationship; The step of performing scale adjacency relationship analysis according to a preset time scale and space scale to establish an adjacency scale includes: Constructing a scale axis according to the preset time scale and space scale; Based on the scale axis, performing scale window adjacency analysis according to a temporal adjacency standard relationship and a spatial adjacency standard relationship to determine scale adjacency closeness; When the scale adjacency closeness does not satisfy the adjacency standard scale relationship, a relationship undertaking scale is established based on the temporal adjacency standard relationship and the spatial adjacency standard relationship; The relationship-taking scale is used as an auxiliary scale and inserted into the scale axis, and the scale adjacent relationship is analyzed to determine the adjacent scale.

2. The dynamic inventory management system according to claim 1, characterized in that: The first scale feature and the second scale feature are respectively enhanced by the attention mechanism to obtain the first scale enhanced feature and the second scale enhanced feature, including: The inventory management factors in the scale characteristics and the corresponding inventory demand changes are expressed as vectors; For each feature vector at each scale, calculate its correlation with the change in inventory demand and obtain the attention weight; The local features are enhanced based on the attention weights to obtain scale-enhanced features.

3. The dynamic inventory management system according to claim 1, characterized in that: The setting of maximum inventory level, minimum inventory level, and dynamic management demand includes: According to the storage space constraints, capital cost ratio and the empirical value of the turnover speed of goods, the maximum inventory level is obtained. The maximum inventory level is the upper limit of the capital risk constraint to avoid inventory backlog; According to order demand fluctuations, supply chain delivery time, and safety stock experience, the minimum inventory level is obtained. The minimum inventory level is the lower limit to prevent supply interruptions caused by insufficient inventory. According to the sales coverage area of ​​inventory management, the spatial scale management requirements are determined, and according to the capital turnover rate target and sales cycle, the time scale management requirements are determined, wherein the spatial scale management requirements and the time scale management requirements can be changed by setting a dynamic change cycle.

4. A dynamic inventory management method, characterized in that: The dynamic inventory management system is implemented by any one of claims 1 to 3, and the dynamic inventory management method comprises: Analyze the inventory impact relationship to obtain inventory management factors, where the inventory management factors are factors that affect dynamic inventory adjustment; Based on the inventory management factor, real-time data and historical data are collected, and the real-time data and historical data both include data from different times and different regions. The real-time data and historical data are used to perform multi-time scale and spatial scale interactions to obtain interaction features of each time scale and each spatial scale. The time scale interaction features include factor features of different time period scales, and the spatial scale interaction features are factor features of different spatial range scales. Set maximum and minimum inventory levels, and dynamically manage demand; Taking the maximum inventory level and the minimum inventory level as constraint conditions and the dynamic management demand as a target, an inventory change analysis is performed according to the interaction characteristics of each time scale and each space scale, and a time scale control parameter and a space scale control parameter are obtained, wherein the time scale control parameter is used to guide the timing of inventory replenishment, and the space scale control parameter indicates how to allocate the dynamic target scale of inventory among different regions; Among them, taking the maximum inventory level and the minimum inventory level as constraint conditions and the dynamic management demand as the goal, the inventory change analysis is performed according to the interaction characteristics of each time scale and each space scale, and the time scale control parameters and space scale control parameters are obtained, including: Determine the safety stock based on the maximum stock level and the minimum stock level; According to the dynamic management demand, dynamic time series decomposition and regional scope decomposition are performed to obtain a dynamic target scale and establish an adjacent interaction path, wherein the dynamic target scale reflects the inventory demand under specific conditions in different time periods and regions, and the adjacent interaction path describes the mutual influence relationship between different time scales and spatial scales within the safety inventory range; Within the safety stock range, according to the adjacent interaction paths, scale feature mapping and matching are performed on the interaction features of each time scale and each space scale, and inventory features of different times and spaces are compared with each other to identify key factors affecting inventory changes and obtain inventory change features corresponding to the dynamic target scale; According to the inventory change characteristics corresponding to the dynamic target scale, the time scale control parameters and the space scale control parameters are obtained according to the time series relationship and the regional scope relationship mapping.

Images