Retail supply chain optimization system and method based on big data analysis

Abstract

The invention provides a retail supply chain optimization system and method based on big data analysis, and relates to the technical field of big data analysis. The system comprises a data acquisition module used for acquiring order data, logistics data and user behavior data and generating structured data; the data processing module is used for preprocessing the structured data and establishing a multi-dimensional association relationship among the data based on a preprocessing result; the inventory prediction module is used for predicting future order demands, calculating regional inventory demands and generating inventory distribution data; the inventory optimization module is used for generating replenishment data and an inventory adjustment strategy based on the inventory distribution data; the behavior analysis module is used for generating commodity recommendation data and promotion priorities based on the user behavior data in combination with the adjusted inventory distribution data; and the logistics allocation module is used for dynamically adjusting the logistics distribution path and the distribution priority. According to the technical scheme, the operation efficiency and flexibility of the retail supply chain can be improved.

Description

Background Art

[0002] With the rapid expansion of the retail industry and the increasing intensification of market competition, supply chain management has gradually become an important link for enterprises to improve operational efficiency and enhance customer satisfaction. Retail enterprises need to coordinate various links from procurement, inventory to logistics to ensure that goods can meet market demands in a timely and accurate manner. However, the types of data involved in supply chain management are diverse, including transaction information, logistics dynamics, and consumer behavior, etc. The diversity and complexity of data sources pose huge challenges to the optimized management of the supply chain.

[0003] In practical applications, the retail supply chain usually needs to process a large amount of heterogeneous data, which vary greatly in format, source, and update frequency. Traditional systems are difficult to efficiently process and integrate this data, resulting in blocked information flow between various links, thus limiting the overall visibility and collaboration of the supply chain. In addition, the demand in the retail market is highly volatile and is significantly affected by factors such as seasons and promotional activities, which requires the supply chain to have higher flexibility and precision to quickly respond to market changes. In inventory management, many existing systems still rely on rules of thumb or static models based on historical data and cannot respond to demand changes in a timely manner, which may lead to overstocking or shortages of goods.

[0004] Therefore, with the rapid development of the retail industry, the existing supply chain management technologies face great limitations in data integration, demand response, and resource allocation, and there is an urgent need for new solutions to achieve higher efficiency and flexibility to meet the diverse needs in the modern retail environment. Summary of the Invention

[0005] The purpose of the embodiments of the present disclosure is to provide an optimized system for a retail supply chain based on big data analysis, an optimized method for a retail supply chain based on big data analysis, an electronic device, and a computer-readable storage medium, so as to at least to a certain extent improve the operation efficiency and flexibility of the retail supply chain, thereby achieving accurate inventory control, intelligent recommendation, and efficient logistics scheduling.

[0006] Other features and advantages of the present disclosure will become apparent through the following detailed description, or be learned in part through the practice of the present disclosure.

[0007] According to the first aspect of the embodiments of the present disclosure, there is provided a retail supply chain optimization system based on big data analysis, the system comprising: a data collection module, configured to collect order data, logistics data, and user behavior data, and format the order data, logistics data, and user behavior data to generate structured data; a data processing module, configured to preprocess the structured data, and establish a multi-dimensional association relationship between the order data, logistics data, and user behavior data based on the preprocessing result; an inventory forecasting module, configured to construct a time series model and a regional feature model based on the multi-dimensional association relationship, forecast future order demands, calculate regional inventory demands, and generate inventory distribution data; an inventory optimization module, configured to dynamically calculate replenishment quantities based on the inventory distribution data, in combination with logistics constraint conditions and the current inventory status, and generate replenishment data and inventory adjustment strategies; a behavior analysis module, configured to construct a user preference model and a commodity demand model based on the user behavior data, and generate commodity recommendation data and promotion priorities in combination with the adjusted inventory distribution data; and a logistics allocation module, configured to dynamically adjust the logistics distribution path and distribution priorities based on the replenishment data and the promotion priorities.

[0008] According to the second aspect of the embodiments of the present disclosure, there is provided a retail supply chain optimization method based on big data analysis, the method comprising: collecting order data, logistics data, and user behavior data, and formatting the order data, logistics data, and user behavior data to generate structured data; preprocessing the structured data, and establishing a multi-dimensional association relationship between the order data, logistics data, and user behavior data based on the preprocessing result; constructing a time series model and a regional feature model based on the multi-dimensional association relationship, forecasting future order demands, calculating regional inventory demands, and generating inventory distribution data; dynamically calculating replenishment quantities based on the inventory distribution data, in combination with logistics constraint conditions and the current inventory status, and generating replenishment data and inventory distribution adjustment strategies; constructing a user preference model and a commodity demand model based on the user behavior data, and generating commodity recommendation data and promotion priorities in combination with the inventory distribution adjustment strategies; and dynamically adjusting the logistics distribution path and distribution priorities based on the replenishment data and the promotion priorities.

[0009] According to the third aspect of the embodiments of the present disclosure, there is provided an electronic device, comprising: a processor; and a memory, having stored thereon computer-readable instructions, which when executed by the processor, implement the above-mentioned retail supply chain optimization method based on big data analysis.

[0010] According to the fourth aspect of the embodiments of the present disclosure, there is provided a computer-readable storage medium, having stored thereon a computer program, which when executed by a processor, implements the retail supply chain optimization method based on big data analysis as described above.

[0011] The technical solutions provided by the embodiments of the present disclosure may include the following beneficial effects:

[0012] In the retail supply chain optimization system based on big data analysis in the embodiments of the present disclosure, through the data collection module, the unified collection and structured processing of multi-source data such as orders, logistics, and user behaviors are realized, solving the integration problem brought by data heterogeneity and providing a reliable data basis for subsequent analysis. The data processing module can deeply explore the correlation between data by preprocessing the structured data and establishing multi-dimensional correlation relationships, making the information chain in supply chain management clearer and improving the overall data utilization efficiency. The inventory forecasting module analyzes the dynamic changes in order demand by constructing time series models and regional feature models, calculates the inventory demand quantity in combination with regional characteristics, and provides accurate forecasting results for inventory decisions in the supply chain, helping to cope with the fluctuations in market demand. The inventory optimization module dynamically calculates the replenishment quantity and generates an inventory adjustment strategy based on inventory distribution data, combined with logistics constraints and the current inventory status, and can realize the dynamic optimization and adjustment of inventory, avoiding inventory backlogs and shortages in the traditional static management mode. The behavior analysis module generates product recommendation data and promotion priorities based on user behavior data by constructing user preference models and product demand models, enabling the supply chain to better match consumer demands and improving the accuracy and efficiency of retail operations. The logistics allocation module dynamically optimizes the delivery route and delivery priority in combination with replenishment data and promotion priorities, and realizes the efficient allocation of resources in complex delivery scenarios, improving the logistics response speed and the flexibility of the overall supply chain. Through the collaborative work of each module, the technical solutions of the present disclosure effectively improve the overall operation efficiency and flexibility of the retail supply chain.

[0013] It should be understood that the above general description and the following detailed description are only exemplary and explanatory, and cannot limit the present disclosure. BRIEF DESCRIPTION OF THE DRAWINGS

[0014] The drawings herein are incorporated into the specification and form a part of the specification, showing embodiments consistent with the present disclosure and used together with the specification to explain the principles of the present disclosure. Obviously, the drawings in the following description are only some embodiments of the present disclosure, and those of ordinary skill in the art can obtain other drawings without creative efforts based on these drawings.

[0015] Figure 1 Schematically shows a composition diagram of a retail supply chain optimization system based on big data analysis according to some embodiments of the present disclosure.

[0016] Figure 2Schematically shows a schematic diagram of the composition of another retail supply chain optimization system based on big data analysis according to some embodiments of the present disclosure.

[0017] Figure 3 Schematically shows a flowchart of the implementation process of a behavior analysis module according to some embodiments of the present disclosure.

[0018] Figure 4 Schematically shows a flowchart of a retail supply chain optimization method based on big data analysis according to some embodiments of the present disclosure.

[0019] Figure 5 Schematically shows a schematic diagram of the structure of a computer system of an electronic device according to some embodiments of the present disclosure.

[0020] Figure 6 Schematically shows a schematic diagram of a computer-readable storage medium according to some embodiments of the present disclosure.

[0021] In the drawings, the same or corresponding reference numerals denote the same or corresponding parts. Detailed Description of Specific Embodiments

[0022] Here, exemplary embodiments will be described in detail, and examples thereof are shown in the drawings. When the following description refers to the drawings, unless otherwise indicated, the same numbers in different drawings represent the same or similar elements. The embodiments described in the following exemplary embodiments do not represent all embodiments consistent with this specification. On the contrary, they are merely examples of devices and methods consistent with some aspects of this specification as detailed in the appended claims.

[0023] The terms used in this specification are for the purpose of describing specific embodiments only and are not intended to limit this specification. The singular forms "a", "the", and "said" used in this specification and the appended claims are also intended to include the plural forms unless the context clearly indicates otherwise. It should also be understood that the term "and / or" as used herein refers to and includes any or all possible combinations of one or more of the associated listed items.

[0024] It should be understood that although the terms first, second, third, etc. may be used in this specification to describe various information, such information should not be limited to these terms. These terms are only used to distinguish the same type of information from each other. For example, without departing from the scope of this specification, the first information may also be referred to as the second information, and similarly, the second information may also be referred to as the first information. Depending on the context, the word "if" as used herein may be interpreted as "when" or "while" or "in response to determining".

[0025] Example embodiments will now be described more fully with reference to the accompanying drawings. However, the example embodiments can be implemented in various forms and should not be construed as limited to the examples set forth herein; rather, these embodiments are provided so that this disclosure will be thorough and complete, and will fully convey the concept of the example embodiments to those skilled in the art.

[0026] In addition, the described features, structures, or characteristics may be combined in any suitable manner in one or more embodiments. In the following description, numerous specific details are provided to give a thorough understanding of the embodiments of the present disclosure. However, those skilled in the art will realize that the technical solutions of the present disclosure may be practiced without one or more of the specific details, or may be implemented using other methods, components, devices, steps, etc. In other cases, well-known methods, devices, implementations, or operations are not shown or described in detail to avoid obscuring aspects of the present disclosure.

[0027] Furthermore, the accompanying drawings are merely schematic illustrations and are not necessarily drawn to scale. The block diagrams shown in the drawings are only functional entities and do not necessarily correspond to physically independent entities. That is, these functional entities may be implemented in software form, or implemented in one or more hardware modules or integrated circuits, or implemented in different networks and / or processor devices and / or microcontroller devices.

[0028] In an example embodiment of the present disclosure, first, a retail supply chain optimization system based on big data analysis is provided, and the retail supply chain optimization system based on big data analysis can be applied to terminal devices. Figure 1 The composition schematic diagram of the retail supply chain optimization system based on big data analysis according to some embodiments of the present disclosure is schematically shown. Refer to Figure 1 As shown, the retail supply chain optimization system based on big data analysis may include the following modules:

[0029] A data collection module 1, which can be used to collect order data, logistics data, and user behavior data, and format the order data, logistics data, and user behavior data to generate structured data;

[0030] A data processing module 2, which can be used to preprocess the structured data and establish a multi-dimensional correlation relationship between the order data, logistics data, and user behavior data based on the preprocessing results;

[0031] An inventory prediction module 3, which can be used to construct a time series model and a regional feature model based on the multi-dimensional correlation relationship, predict future order demands, calculate regional inventory demands, and generate inventory distribution data;

[0032] The inventory optimization module 4 can be used to dynamically calculate the replenishment quantity based on inventory distribution data, combined with logistics constraints and the current inventory status, and generate replenishment data and inventory adjustment strategies;

[0033] The behavior analysis module 5 can be used to build user preference models and commodity demand models based on user behavior data, and generate commodity recommendation data and promotion priorities in combination with the adjusted inventory distribution data;

[0034] The logistics allocation module 6 can be used to dynamically adjust the logistics distribution path and distribution priority based on the replenishment data and promotion priorities.

[0035] In the actual operation process, the retail supply chain optimization system based on big data analysis executes in modular steps. First, the data collection module 1 collects order data, logistics data, and user behavior data, generates unified structured data through formatting processing, and provides standardized input for subsequent data analysis. The data collection module 1 realizes real-time collection of multi-dimensional and multi-type data by docking different data sources, and at the same time performs format conversion and field mapping on the original data, so as to ensure the accuracy and consistency of multi-source data. Subsequently, the data processing module 2 receives the structured data generated by the data collection module 1 and performs data preprocessing on it, including operations such as data cleaning, deduplication, and missing value filling, to improve data quality; at the same time, the data processing module 2 establishes multi-dimensional association relationships between order data, logistics data, and user behavior data based on the preprocessing results. By constructing association rules and data mapping models, this module effectively mines the internal connections between various types of data, laying a foundation for accurate analysis and optimization of all links in the supply chain. Next, the inventory forecasting module 3 uses the multi-dimensional association relationships generated by the data processing module 2 to build time series models and regional feature models to predict future order demands and calculate regional inventory demands. The time series model can analyze the dynamic change trend of order data, while the regional feature model combines logistics and user behavior data, comprehensively considers regional factors, and finally generates accurate inventory distribution data to provide forward-looking support for inventory management.

[0036] Based on the inventory distribution data, the inventory optimization module 4 dynamically calculates the replenishment quantity in combination with the logistics constraints and the current inventory status, generating replenishment data and inventory adjustment strategies. By introducing a real-time optimization algorithm, this module can quickly respond to changes in market demand, dynamically adjust the inventory, thereby avoiding inventory backlogs or shortages and ensuring the smooth operation of the supply chain. At the same time, the behavior analysis module 5 constructs user preference models and commodity demand models based on user behavior data. By deeply analyzing users' browsing, clicking, and purchasing behaviors, this module generates product recommendation data and promotion priorities in combination with the adjusted inventory distribution data. Thus, the behavior analysis module 5 can effectively connect user needs with the results of inventory optimization, thereby realizing personalized product recommendations and precise promotion strategies. Finally, the logistics allocation module 6 dynamically adjusts the logistics distribution path and distribution priority using the replenishment data generated by the inventory optimization module 4 and the promotion priorities generated by the behavior analysis module 5. In addition, through path optimization algorithms and task priority allocation strategies, this module can flexibly allocate resources in complex distribution scenarios, ensuring the timeliness and efficiency of logistics distribution. Through the collaborative work of the above modules, the retail supply chain optimization system realizes the full-process dynamic optimization from data collection to inventory optimization and logistics allocation during operation, significantly improving the overall operation efficiency and flexibility of the retail supply chain.

[0037] Next, with reference to Figure 2 the schematic diagram of the composition of the retail supply chain optimization system shown, the above retail supply chain optimization system based on big data analysis will be described in detail in other embodiments of the present disclosure.

[0038] The data collection module 1 can be used to collect order data, logistics data, and user behavior data, and format the order data, logistics data, and user behavior data to generate structured data. Among them, the order data can represent a set of information related to retail transactions, including but not limited to the name, category, quantity, unit price, total order amount, order placement time, order number, payment method, and relevant customer information of the goods. The logistics data can represent a set of information related to the flow process of goods in the supply chain, including but not limited to the distribution path, transportation method, shipping time, estimated arrival time, actual arrival time, logistics node status, package weight and volume, receipt record, and relevant abnormal events. The user behavior data can represent the interaction behavior records of users on the retail platform, including but not limited to the browsing path, click operations, search keywords, products added to the shopping cart, favorite products, order placement and payment behaviors, evaluation content, and other information preferred by users. The structured data can represent a set of data that conforms to predefined specifications after formatting the original data. Its data structure includes rows, columns, and fields, and can be directly read and processed by a database or other analysis tools. The structured data is generated by subjecting the order data, logistics data, and user behavior data to field normalization, format unification, and redundancy removal processes, and includes fields such as standardized timestamps, product numbers, quantities, geographical locations, etc., providing standardized inputs for the construction and calculation of subsequent analysis models.

[0039] In some embodiments, referring to Figure 2 as shown, the data collection module 1 may include an order data collection unit 11, a logistics data collection unit 12, a behavior data collection unit 13, and a formatting processing unit 14. Among them, the order data collection unit can be used to collect order data including but not limited to the order number, product details, and transaction amount through the order management interface; the logistics data collection unit can be used to collect logistics data including but not limited to the distribution path, transportation status, and receipt record through the logistics information interface; the behavior data collection unit can be used to collect user behavior data including but not limited to browsing trajectories, product click records, and preference information through the user behavior monitoring interface; the formatting processing unit can be used to format the order data, logistics data, and user behavior data to generate structured data.

[0040] Specifically, the formatting process may include the following steps:

[0041] First, for the order data, the formatting processing unit maps the fields in the original order information to construct a field conversion matrix M order =[m i，j , where m i，jIndicates that the i-th original order field (such as order number, product details, transaction amount) is mapped to the j-th target field. Through this mapping, the order data is unified into a standardized format, making its field names, data types, and lengths conform to predefined rules.

[0042] Subsequently, for logistics data, the formatting processing unit performs timestamp normalization to standardize the original timestamps of logistics events (such as shipping time, receipt time). The standardization formula is:

[0043] t log =t o +△ t

[0044] where, t log represents the standardized logistics time, t o is the original timestamp, and △t is the time zone offset. This step ensures that the time fields of logistics data are unified into Coordinated Universal Time (UTC), providing support for cross-regional logistics analysis. At the same time, for the logistics path and transportation status fields, the status information (such as "in transit") is converted into a predefined digital coding format using a coding conversion table to further improve the operability of the data.

[0045] Next, for user behavior data, the formatting processing unit extracts the browsing, clicking, and purchasing behaviors of users and quantifies them through a behavior scoring formula:

[0046]

[0047] where, b u，p represents the behavior score of user u for product p, and respectively represent the number of browsing, clicking, and purchasing times, and w 1 , w 2 , w 3 are the corresponding weight factors. Through this quantification step, user behavior data is standardized into a numerical form, facilitating subsequent modeling and analysis.

[0048] Finally, the formatting processing unit integrates the order data, logistics data, and user behavior data into a unified structured data set, defined as D = {d 1 , d 2 ,..., d m}, where each record d i = [f i，1 , f i,2 ,…, f i，kStandardized fields including orders, logistics, and user behavior. Through the above formatting steps, the generated structured data has high consistency and high availability, providing standardized input for subsequent analysis and optimization. Exemplarily, the form of the structured data can be as shown in Table 1 below:

[0049] Table 1

[0050]

[0051] The data processing module 2 can be used to preprocess the structured data and establish multi-dimensional correlation relationships between order data, logistics data, and user behavior data based on the preprocessing results. Among them, the multi-dimensional correlation relationship can represent the potential correlation between different data dimensions by deeply analyzing order data, logistics data, and user behavior data, so as to reveal the interaction and influence between different data elements in the supply chain operation process.

[0052] In some embodiments, as shown in Figure 2 , the data processing module 2 may include a data cleaning unit 21, a data standardization unit 22, and a data association unit 23. Among them, the data cleaning unit can be used to preprocess the structured data by removing duplicate data, filling in missing data, and eliminating abnormal data; the data standardization unit can be used to standardize the data processed by the data cleaning unit by unifying the field format, time format, and data unit; the data association unit can be used to receive the data processed by the data standardization unit and establish multi-dimensional correlation relationships between order data, logistics data, and user behavior data based on association rules and feature extraction algorithms.

[0053] Specifically, the establishment of the multi-dimensional correlation relationship may include the following process:

[0054] First, perform feature extraction on the standardized order data set D order ={o 1 , o 2 , …, o m}, logistics data set D logistics ={l 1 , l 2 , …, l m}, and user behavior data set D user ={u 1 , u 2 , …, u m} to generate an order feature matrix logistics feature matrix and user behavior feature matrix where and respectively represent the j-th order, logistics, or user behavior characteristic value in the i-th record.

[0055] Secondly, based on the association between order data and logistics data, rules between commodity categories and logistics timeliness are mined to generate an association rule set where each rule r l is defined as For example, "commodity category is fresh" (X) derives "needs to be delivered preferentially" (Y). The support degree s(X, Y) and confidence degree of the association rule The calculation processes are respectively:

[0056]

[0057] Among them, Count(X∩Y) represents the number of records that satisfy both X and Y, and m is the total number of records.

[0058] Subsequently, for the association between user behavior data and order data, the relationship between user behavior characteristics (such as the number of views, the number of clicks) and commodity sales volume is mined to generate a user behavior association rule set For example, "high click-through rate" (X) derives "high sales volume" (Y), revealing the impact of user behavior on the order conversion rate.

[0059] Finally, combining the above rule sets and Based on the order feature matrix F order the logistics feature matrix F logistics and the user behavior feature matrix F user , a feature interaction matrix A = [a i，j is constructed, where:

[0060]

[0061] a i，j represents the interaction intensity between the i-th feature of the order data and the j-th feature of the logistics or user behavior data. Through the interaction matrix, a multi-dimensional association graph is constructed Among them, the node set includes the features of order, logistics, and user behavior data, the edge set ε represents the association between features, and the edge weight is jointly determined by the interaction intensity and the confidence degree of the association rule.

[0062] Through the above steps, the data association unit can systematically establish the multi-dimensional association relationship between order, logistics, and user behavior data, providing accurate data support for supply chain optimization.

[0063] The inventory forecasting module 3 can be used to construct a time series model and a regional feature model based on multi-dimensional correlation relationships, predict future order demands, calculate regional inventory demands, and generate inventory distribution data. Among them, the domain feature model can represent a mathematical model constructed by analyzing the characteristics related to geographical regions in order data, logistics data, and user behavior data, which can reflect the relationship between different regions and order demands. The inventory distribution data can represent a data set generated based on the regional feature model and the time series model, which is used to represent the inventory demands and their distribution status in each region.

[0064] In some embodiments, referring to Figure 2 as shown, the inventory forecasting module 3 may include a demand forecasting unit 31, a feature analysis unit 32, and an inventory distribution calculation unit 33. Among them, the demand forecasting unit can be used to construct a time series model based on multi-dimensional correlation relationships to perform trend analysis on order data and predict future order demands; the feature analysis unit can be used to calculate regional inventory demands through the regional feature model based on the prediction results of the demand forecasting unit and in combination with the regional distribution information in the logistics data; the inventory distribution calculation unit can be used to generate inventory distribution data based on the inventory demand data and in combination with the current inventory status.

[0065] In some embodiments, based on multi-dimensional correlation relationships, constructing a time series model to perform trend analysis on order data and predict future order demands specifically includes the following technical steps:

[0066] Based on multi-dimensional correlation relationships, construct a multi-dimensional time series matrix X t =[x i，j，t , where x i，j，t represents the order quantity of the i-th type of commodity in the j-th region at time t, i ∈ {1, 2,..., n}, j ∈ {1, 2,..., m}, n represents the total number of commodity types, and m represents the total number of regions.

[0067] Based on the time series matrix X t , construct a standardized sequence matrix Z t through detrending and standardization operations, where μ i，j represents the historical average order quantity of the i-th type of commodity in the j-th region, and σ i，j represents its historical standard deviation.

[0068] Extract features from the standardized sequence matrix Z t through a dynamic factor model, and construct a time series prediction model Z t+1 =F·Z t +G·U t , where F is the time correlation factor matrix, G is the perturbation coefficient matrix, and U tThe disturbance vector representing time t.

[0069] Based on the time series prediction model, combined with the future time step T, calculate the order demand prediction matrix where represents the order demand matrix at future time t+T.

[0070] Further, after obtaining the order demand prediction matrix, the feature analysis unit receives the order demand prediction matrix where x i，j，t+T represents the predicted order quantity of the i-th type of commodity in the j-th region at future time t+T. Combined with the regional distribution timeliness matrix D=[d j，k , where d j，k represents the average distribution duration from region j to distribution center k, calculate the regional inventory demand quantity matrix Q=[q i，j , and its calculation formula is:

[0071]

[0072] where, T s represents the inventory safety cycle, and by introducing the distribution timeliness adjustment factor, the accuracy of regional inventory demand is optimized.

[0073] The inventory distribution calculation unit 33 is based on the regional inventory demand quantity matrix Q, combined with the current inventory status S=[s i，j , where s i，j represents the existing inventory quantity of the i-th type of commodity in the j-th region, and calculate the inventory distribution adjustment matrix A=[a i，j . The calculation formula for the inventory adjustment quantity is:

[0074] a i，j =q i，j -s i，j

[0075] If a i，j >0, it means that inventory needs to be replenished. If a i，j <0, it means that inventory needs to be reduced or transferred. Through this adjustment quantity, generate the final inventory distribution data I={i 1 , i 2 ,..., i m}, where each record i j =[r j , q j , a j , r j represents the region number, q j represents the regional inventory demand quantity, and a j represents the inventory adjustment quantity.

[0076] The inventory optimization module 4 can be used to dynamically calculate the replenishment quantity based on inventory distribution data, combined with logistics constraint conditions and the current inventory status, and generate replenishment data and inventory adjustment strategies. Among them, the logistics constraint conditions can represent a set of constraint parameter sets formed by logistics resource allocation, transportation capacity, and distribution network limitations during the process of supply chain inventory management. The logistics constraint conditions include, but are not limited to, transportation capacity constraints, distribution timeliness constraints, warehousing capacity constraints, distribution route constraints, and cost constraints.

[0077] In some embodiments, referring to Figure 2 As shown, the inventory optimization module 4 may include an inventory status analysis unit 41, a replenishment quantity calculation unit 42, and a strategy generation unit 43. Among them, the inventory status analysis unit can be used to analyze the current inventory status based on inventory distribution data and generate inventory surplus quantity and inventory demand difference data; the replenishment quantity calculation unit can be used to calculate the required replenishment quantity through a dynamic replenishment algorithm based on the inventory demand difference data and logistics constraint conditions; the strategy generation unit can be used to generate replenishment data and inventory adjustment strategies based on the required replenishment quantity and combined with the current inventory status.

[0078] Specifically, the inventory status analysis unit 41 based on the inventory distribution data I = {i 1 , i 2 ,..., i m}, analyzes the current inventory status and generates an inventory surplus quantity matrix S = [s i，j and an inventory demand difference matrix △Q = [△q i，j . Among them, s i，j represents the existing inventory quantity of the i-th type of commodity in the j-th area, and △q i，j represents the difference between the regional demand and the existing inventory quantity. The calculation formula is:

[0079] △q i，j = max(0, q i，j - s i，j )

[0080] Among them, q i，j represents the demand for the i-th type of commodity in the j-th area. By analyzing the inventory status, this unit clarifies the inventory gaps in each area and lays a foundation for replenishment decisions.

[0081] Next, the replenishment quantity calculation unit 42 calculates the replenishment quantity matrix R = [r i，j based on the inventory demand difference matrix △Q and logistics constraint conditions using a dynamic optimization model. The replenishment quantity calculation formula is:

[0082]

[0083] Among them, r i，jrepresents the replenishment quantity of the i-th category of goods in the j-th region, λ j，k represents the logistics complexity coefficient between region j and distribution center k, d j，k is the transportation timeliness, T s is the inventory safety cycle. This formula comprehensively considers logistics constraints, timeliness, and regional logistics complexity to ensure more accurate calculation of the replenishment quantity.

[0084] Finally, the strategy generation unit 43 generates replenishment data B = [b i，j and a dynamic inventory adjustment strategy based on the replenishment quantity matrix R and in combination with the current inventory status S The generation formula for the replenishment data is as follows:

[0085] b i，j = r i，j + α·s i，j

[0086] where α is the inventory holding coefficient used to adjust the safety inventory level after replenishment. The inventory adjustment strategy includes replenishment priority ranking and allocation plans. The strategy generation unit combines the regional urgency and logistics capabilities to optimize the replenishment plan through a dynamic programming algorithm, improving the overall replenishment efficiency.

[0087] The behavior analysis module 5 can be used to build a user preference model and a commodity demand model based on user behavior data, and generate commodity recommendation data and promotion priorities in combination with the adjusted inventory distribution data. Among them, the user preference model can represent a mathematical model constructed based on user behavior data by analyzing the browsing, clicking, and purchasing behavior characteristics of users on the retail platform, and is used to describe the interest preferences of users for different commodity categories, brands, price ranges, and functional attributes. The commodity demand model can represent a mathematical model constructed by analyzing order data and the user preference model and using an algorithm for mining commodity association rules to extract the demand association relationships between commodities, and is used to describe the demand interaction characteristics between commodities.

[0088] In some embodiments, as shown in Figure 2 the behavior analysis module 5 may include a user preference modeling unit 51, a commodity demand modeling unit 52, and a recommendation strategy generation unit 53. Among them, the user preference modeling unit can be used to identify the characteristics of users' browsing, clicking, and purchasing behaviors based on user behavior data through a behavior feature extraction algorithm, and build a user preference model; the commodity demand modeling unit can be used to extract the demand association relationships between commodities based on the user preference model and order data using an algorithm for mining commodity association rules, and build a commodity demand model; the recommendation strategy generation unit can be used to generate commodity recommendation data and promotion priorities based on the user preference model and the commodity demand model in combination with the adjusted inventory distribution data. The specific implementation process of the behavior analysis module is as Figure 3As shown, specifically:

[0089] In some embodiments, based on user behavior data, the characteristics of user browsing, clicking, and purchasing behaviors are identified through a behavior feature extraction algorithm, and a user preference model is constructed. Specifically, the following technical steps are included:

[0090] The first step is to construct a user behavior matrix B based on user behavior data i，j，t =[b i，j，t , where b i，j，t represents the behavior score of user i for product j at time t. The behavior score formula is defined as Among them, respectively represent the number of browsing, clicking, and purchasing times, and w 1 , w 2 , w 3 represent behavior weights.

[0091] The second step is to apply the singular value decomposition method to the user behavior matrix to decompose the behavior score into a user latent preference vector P i =[p i，k and a product feature vector Q j =[q j，k , where B = P·Q T , P i represents the preference weight of user i in the latent preference dimension, and Q j represents the feature weight of product j.

[0092] The third step is to optimize the decomposition model through regularized least squares. The objective function is defined as Among them, λ represents the regularization coefficient.

[0093] The fourth step is to construct a time-dynamic user preference model U by combining the user's historical behavior and the time decay factor i,t =P i +β·T t , where T t =exp(-γt), β represents the time influence coefficient, and γ represents the time decay weight.

[0094] In this embodiment, the commodity demand modeling unit 52 extracts the demand association relationship between commodities by analyzing the user preference model and order data, specifically including the following technical steps:

[0095] The first step is based on the commodity scoring matrix R = [r i，j of the user preference model, where r i，j represents the preference score of user i for product j. Combining the commodity co-occurrence matrix in the order data Calculate the support and confidence of commodity association rules. The formula for support is:

[0096]

[0097] where Count(j 1 ∩j 2 ) represents the number of orders that contain both commodity j 1 and j 2 , and N represents the total number of orders. The formula for confidence is:

[0098]

[0099] where Count(j 1 ) represents the number of orders that contain commodity j 1 . Screen high - associated commodity pairs based on support and confidence, and construct a commodity demand association network where the node set is the commodity set, the edge set ε represents the association relationship between commodities, and the edge weight is determined by the association strength.

[0100] Second, construct a commodity demand feature matrix D = [d j，k through matrix factorization, where d j，k represents the feature weight of commodity j in the k - th demand dimension. The commodity demand feature matrix is obtained through the following optimization objective function:

[0101]

[0102] where λ represents the regularization coefficient, and ‖·‖ represents the L 2 norm of the vector. Through the above steps, extract the demand association characteristics between commodities and generate a commodity demand model.

[0103] The recommendation strategy generation unit 53 generates commodity recommendation data and promotion priorities based on the user preference model and the commodity demand model, combined with the adjusted inventory distribution data. The specific process is as follows:

[0104] First, based on the user preference model matrix U i,t and the commodity demand feature matrix D, calculate the recommendation score S = [s i，j of user i for commodity j. The recommendation score formula is:

[0105]

[0106] where D j represents the weight vector of commodity j in the demand feature dimension.

[0107] Second, combine the inventory distribution adjustment data A = [a j, generate a product recommendation list where the recommendation order is determined by the descending order of (s i，j +ω·a j ), and ω is the inventory adjustment priority weight coefficient.

[0108] In the third step, based on the recommendation scores and inventory data, calculate the promotion priority matrix P = [p j , where,

[0109]

[0110] δ is the promotion adjustment factor. Through this formula, the promotion priority of products with high inventory but low recommendation scores is increased, thereby achieving precise marketing.

[0111] The logistics allocation module 6 can be used to dynamically adjust the logistics distribution path and distribution priority based on replenishment data and promotion priorities. Thereby optimizing the logistics distribution path, reducing transportation costs and distribution time, while enhancing the response speed of high-priority replenishment tasks, and achieving the efficient utilization of logistics resources and the flexibility and reliability of the overall operation of the supply chain.

[0112] In some embodiments, as shown in Figure 2 , the logistics allocation module 6 includes a path planning unit 61, a priority adjustment unit 62, and a task assignment unit 63. Among them, the path planning unit can be used to construct a distribution path model based on the replenishment data and the distribution area information in the logistics data, and optimize the logistics distribution path through the shortest path algorithm; the priority adjustment unit can be used to analyze the commodity demand priority in the distribution task based on the promotion priority and commodity recommendation data, and dynamically adjust the task order of the logistics distribution; the task assignment unit can be used to match logistics resources and generate specific distribution instructions based on the distribution path generated by the path planning unit and the task order generated by the priority adjustment unit. The specific process is as follows:

[0113] The path planning unit 61 constructs a distribution path model based on the replenishment data B = [b i,j and the distribution area information in the logistics data where represents the set of distribution nodes, including the distribution center and the target area; ε represents the path between distribution nodes, and the weight of the path is determined by the transportation time matrix T = [t u，v , where t u，v represents the transportation time from node u to node v.

[0114] To optimize the logistics distribution path, this unit uses the shortest path algorithm (such as Dijkstra or Floyd algorithm) for path search, and the optimization objective is:

[0115]

[0116] Among them, x u，v ∈ {0, 1} indicates whether the path between nodes u and v is selected. The optimization result generates the optimal delivery path and outputs the access order of the delivery nodes, providing a basis for subsequent priority adjustment.

[0117] The priority adjustment unit 62 is based on the promotion priority matrix P = [p j and the product recommendation data R = [r i，j , and analyzes the demand priorities of the products in the delivery task. First, according to the promotion priority p j of each product j and the user recommendation score r i，j , calculate the comprehensive priority U = [u j of the product. The calculation formula for the comprehensive priority is as follows:

[0118]

[0119] Among them, α and β are weight coefficients, respectively measuring the influence degrees of the promotion priority and the user recommendation score, and n is the number of users.

[0120] Subsequently, arrange the products in the delivery task in descending order according to the comprehensive priority to generate the dynamically adjusted task order where each t k represents the product number and its priority in task k. Through this sorting, it is ensured that high-demand or promoted products are given priority in delivery.

[0121] The task allocation unit 63 matches the logistics resources and generates specific delivery instructions based on the optimal delivery path generated by the path planning unit 61 and the task order k generated by the priority adjustment unit 62. First, define the logistics resource matrix R = [r k , where r

[0122] represents the transportation capacity of the k-th logistics vehicle.

[0123]

[0124] Among them, d j represents the demand quantity of the product in task j, and c k represents the remaining transportation capacity of vehicle k, and it is necessary to satisfy the constraint condition d j ≤ c k . Allocate logistics vehicles through the linear programming method to ensure that the loading capacity of each vehicle is optimal and at the same time satisfy the transportation capacity constraint.

[0125] After the allocation is completed, the task allocation unit generates a set of delivery instructions Each instruction i k may contain the following information: the delivery route the task sequence and the assigned vehicle number k. The delivery instructions are transmitted to the logistics operation end in real time through the wireless network to achieve efficient execution.

[0126] In the above-mentioned retail supply chain optimization system based on big data analysis, through the efficient collection and formatting of order data, logistics data, and user behavior data by the data collection module, the standardization of multi-source heterogeneous data is realized, providing a unified data basis for subsequent analysis, and enhancing the reliability and consistency of data processing. The data processing module further improves the data quality by data cleaning, standardization, and the establishment of multi-dimensional correlation relationships, eliminates redundancy and abnormal interference, and at the same time enhances the correlation between order, logistics, and user behavior data, providing highly relevant basic support for prediction and optimization. The inventory prediction module uses time series models and regional feature models to accurately predict future order demands, calculate regional inventory demands, and generate inventory distribution data, thereby improving the timeliness and regional accuracy of inventory prediction and laying a data foundation for inventory optimization. The inventory optimization module combines inventory distribution data, logistics constraints, and the current inventory status, and uses dynamic replenishment algorithms to generate replenishment data and inventory adjustment strategies, realizing precise and efficient inventory management, reducing the risks of inventory backlog and out-of-stock, and optimizing replenishment efficiency. The behavior analysis module deeply explores the potential connections between user behavior data and order data by constructing user preference models and commodity demand models, and generates commodity recommendation data and promotion priorities in combination with inventory distribution data, thereby improving user satisfaction and commodity sales volume, and enhancing the market response ability of the supply chain. The logistics allocation module dynamically adjusts the logistics delivery route and task sequence through the collaborative optimization of path planning, priority adjustment, and task allocation, effectively reducing transportation costs and improving the response timeliness of high-priority tasks, ensuring the efficient use of logistics resources and the flexible operation of the supply chain. Through the real-time interaction and dynamic adjustment of data among modules, a full-chain intelligent linkage from data collection, processing, prediction to optimization, analysis, and allocation is constructed, significantly improving the overall operation efficiency and flexibility of the supply chain, and meeting the requirements of complex and changing market demands and the dynamic changes of the logistics environment.

[0127] It should be noted that although several modules or units of the retail supply chain optimization system based on big data analysis are mentioned in the above detailed description, this division is not mandatory. In fact, according to the embodiments of the present disclosure, the features and functions of the two or more modules or units described above can be embodied in one module or unit. Conversely, the features and functions of one module or unit described above can be further divided and embodied by multiple modules or units.

[0128] Secondly, in an exemplary embodiment of the present disclosure, a retail supply chain optimization method based on big data analysis is also provided. Refer to Figure 4 As shown, the retail supply chain optimization method based on big data analysis may include the following steps:

[0129] Step S410, collect order data, logistics data, and user behavior data, and perform formatting processing on the order data, logistics data, and user behavior data to generate structured data;

[0130] Step S420, preprocess the structured data, and establish a multi-dimensional association relationship between the order data, logistics data, and user behavior data based on the preprocessing results;

[0131] Step S430, construct a time series model and a regional feature model based on the multi-dimensional association relationship, predict future order demands, calculate regional inventory demands, and generate inventory distribution data;

[0132] Step S440, based on the inventory distribution data, combine the logistics constraint conditions and the current inventory status, dynamically calculate the replenishment quantity, and generate replenishment data and inventory distribution adjustment strategies;

[0133] Step S450, construct a user preference model and a commodity demand model based on the user behavior data, and generate commodity recommendation data and promotion priorities in combination with the inventory distribution adjustment strategy;

[0134] Step S460, dynamically adjust the logistics distribution path and distribution priority based on the replenishment data and promotion priorities.

[0135] Next, the above retail supply chain optimization method based on big data analysis will be further described in an exemplary embodiment.

[0136] First, by collecting order data, logistics data and user behavior data, and formatting them, structured data is generated to provide a unified input for subsequent data analysis. Furthermore, the structured data is preprocessed, and through data cleaning, deduplication and standardization operations, a multidimensional correlation relationship between order data, logistics data and user behavior data is established based on the preprocessing results to explore the potential correlation between data and provide support for supply chain optimization decisions. Then, a time series model and a regional feature model are constructed through multidimensional correlation relationships to achieve accurate prediction of future order demand, calculate regional inventory demand, and then generate inventory distribution data to lay the foundation for dynamic inventory management. Based on inventory distribution data, combined with logistics constraints and current inventory status, the replenishment quantity is dynamically calculated to generate replenishment data and inventory distribution adjustment strategies to optimize inventory configuration and improve the supply chain's adaptability to market fluctuations. In addition, based on user behavior data, a user preference model and a commodity demand model are constructed, and commodity recommendation data and promotion priorities are generated by combining inventory distribution adjustment strategies, so that the supply chain can match user needs more accurately. Finally, based on replenishment data and promotion priorities, the logistics distribution routes and distribution priorities are dynamically adjusted to optimize the utilization of logistics resources, improve distribution efficiency and service quality, and achieve dynamic optimization of the entire process of the retail supply chain.

[0137] It should be noted that, although the steps of the method in the present disclosure are described in a specific order in the drawings, this does not require or imply that the steps must be performed in this specific order, or that all the steps shown must be performed to achieve the desired results. Additionally or alternatively, some steps may be omitted, multiple steps may be combined into one step, and / or one step may be decomposed into multiple steps, etc.

[0138] In addition, in an exemplary embodiment of the present disclosure, an electronic device capable of implementing the above-mentioned retail supply chain optimization method based on big data analysis is also provided.

[0139] Those skilled in the art will appreciate that various aspects of the present disclosure may be implemented as systems, methods or program products. Therefore, various aspects of the present disclosure may be specifically implemented in the following forms, namely: complete hardware embodiments, complete software embodiments (including firmware, microcode, etc.), or embodiments combining hardware and software aspects, which may be collectively referred to herein as "circuits", "modules" or "systems".

[0140] Refer to the following Figure 5 hereinafter describes an electronic device 500 according to such an embodiment of the present disclosure. Figure 5 The electronic device 500 shown is merely an example and should not bring any limitation to the functions and scope of use of the embodiments of the present disclosure.

[0141] As shown Figure 5 in FIG. 500, the electronic device 500 is presented in the form of a general-purpose computing device. The components of the electronic device 500 may include, but are not limited to: at least one of the above-mentioned processing units 510, at least one of the above-mentioned storage units 520, a bus 530 connecting different system components (including the storage unit 520 and the processing unit 510), and a display unit 540.

[0142] Among them, the storage unit stores program code, and the program code can be executed by the processing unit 510, so that the processing unit 510 executes the steps according to various exemplary embodiments of the present disclosure described in the "Exemplary Method" section of this specification. The storage unit 520 may include a readable medium in the form of a volatile storage unit, such as a random access storage unit (RAM) 521 and / or a cache storage unit 522, and may further include a read-only storage unit (ROM) 523.

[0143] The storage unit 520 may also include a program / utilities 524 having a set (at least one) of program modules 525. Such program modules 525 include, but are not limited to: an operating system, one or more application programs, other program modules, and program data. Each or some combination of these examples may include the implementation of a network environment.

[0144] The bus 530 may represent one or more of several types of bus structures, including a storage unit bus or a storage unit controller, a peripheral bus, a graphics acceleration port, a processing unit, or a local bus using any of the various bus structures.

[0145] The electronic device 500 may also communicate with one or more external devices 570 (such as a keyboard, a pointing device, a Bluetooth device, etc.), and may also communicate with one or more devices that enable a user to interact with the electronic device 500, and / or communicate with any device that enables the electronic device 500 to communicate with one or more other computing devices (such as a router, a modem, etc.). Such communication may be performed through an input / output (I / O) interface 550. And, the electronic device 500 may also communicate with one or more networks (such as a local area network (LAN), a wide area network (WAN), and / or a public network, such as the Internet) through a network adapter 560. As shown in the figure, the network adapter 560 communicates with other modules of the electronic device 500 through the bus 530. It should be understood that although not shown in the figure, other hardware and / or software modules may be used in combination with the electronic device 500, including but not limited to: microcode, device drivers, redundant processing units, external disk drive arrays, RAID systems, tape drives, and data backup storage systems, etc.

[0146] Through the description of the above embodiments, those skilled in the art can easily understand that the example embodiments described herein can be implemented by software or by a combination of software and necessary hardware.

[0147] In an exemplary embodiment of the present disclosure, there is also provided a computer-readable storage medium having a program product stored thereon that can implement the methods described in this specification above. In some possible embodiments, various aspects of the present disclosure can also be implemented in the form of a program product, which includes program code. When the program product runs on a terminal device, the program code is used to cause the terminal device to execute the steps according to various exemplary embodiments of the present disclosure described in the "Exemplary Methods" section above of this specification.

[0148] Referring to Figure 6 As shown, a program product 600 for implementing the above-mentioned retail supply chain optimization method based on big data analysis according to an embodiment of the present disclosure is described. It can be a portable compact disc read-only memory (CD-ROM) and includes program code, and can run on a terminal device, such as a personal computer. However, the program product of the present disclosure is not limited to this. In this document, a readable storage medium can be any tangible medium that contains or stores a program, and this program can be used by or in combination with an instruction execution system, apparatus, or device.

[0149] The program product can adopt any combination of one or more readable media. The readable medium can be a readable signal medium or a readable storage medium. The readable storage medium can be, for example, but not limited to, an electrical, magnetic, optical, electromagnetic, infrared, or semiconductor system, apparatus, or device, or any combination of the above. More specific examples (non-exhaustive list) of the readable storage medium include: an electrical connection having one or more wires, a portable disk, a hard disk, a random access memory (RAM), a read-only memory (ROM), an erasable programmable read-only memory (EPROM or flash memory), an optical fiber, a portable compact disc read-only memory (CD-ROM), an optical storage device, a magnetic storage device, or any suitable combination of the above.

[0150] The computer-readable signal medium can include a data signal propagated in a baseband or as part of a carrier wave, which carries the readable program code. Such a propagated data signal can take various forms, including but not limited to electromagnetic signals, optical signals, or any suitable combination of the above. The readable signal medium can also be any readable medium other than the readable storage medium, and this readable medium can send, propagate, or transmit a program for use by or in combination with an instruction execution system, apparatus, or device.

[0151] The program code contained on a readable medium can be transmitted using any suitable medium, including but not limited to wireless, wired, optical fiber cable, electromagnetic waves, etc., or any suitable combination of the foregoing.

[0152] The program code for performing the operations of the present disclosure can be written in any combination of one or more programming languages, including object-oriented programming languages such as Java, C++, etc., and also including conventional procedural programming languages such as the "C" language or similar programming languages. The program code can be executed entirely on the user's computing device, partially on the user's device, executed as a stand-alone software package, partially on the user's computing device and partially on a remote computing device, or entirely on a remote computing device or server. In the case of a remote computing device, the remote computing device can be connected to the user's computing device through any type of network, including a local area network (LAN) or a wide area network (WAN), or, alternatively, can be connected to an external computing device (e.g., by connecting through the Internet using an Internet service provider).

[0153] In addition, the above-mentioned drawings are only schematic illustrations of the processes included in the method according to the exemplary embodiments of the present disclosure, and are not for limiting purposes. It is easy to understand that the processes shown in the above-mentioned drawings do not indicate or limit the chronological order of these processes. Additionally, it is also easy to understand that these processes can be executed, for example, synchronously or asynchronously in multiple modules.

[0154] From the description of the above embodiments, those skilled in the art can easily understand that the example embodiments described herein can be implemented by software, or can be implemented by a combination of software and necessary hardware. Therefore, the technical solution according to the embodiments of the present disclosure can be embodied in the form of a software product, which can be stored in a non-volatile storage medium (which can be a CD-ROM, USB flash drive, mobile hard disk, etc.) or on a network, including several instructions to enable a computing device (which can be a personal computer, server, touch terminal, or network device, etc.) to execute the method according to the embodiments of the present disclosure.

[0155] After considering the specification and practicing the invention disclosed herein, those skilled in the art will readily conceive of other embodiments of the present disclosure. This application is intended to cover any variations, uses, or adaptations of the present disclosure that follow the general principles of the present disclosure and include known common knowledge or conventional technical means in the technical field not disclosed in the present disclosure. The specification and embodiments are only to be regarded as exemplary, and the true scope and spirit of the present disclosure are pointed out by the claims.

[0156] It should be understood that the present disclosure is not limited to the exact structures described above and shown in the drawings, and various modifications and changes can be made without departing from its scope. The scope of the present disclosure is limited only by the appended claims.

Claims

1. A retail supply chain optimization system based on big data analysis, characterized in that: include: A data collection module, used to collect order data, logistics data and user behavior data, and format the order data, logistics data and user behavior data to generate structured data; A data processing module, used to pre-process the structured data and establish a multi-dimensional correlation relationship between the order data, logistics data and user behavior data based on the pre-processing result; An inventory forecasting module is used to construct a time series model and a regional characteristic model based on the multi-dimensional association relationship, predict future order demand, calculate regional inventory demand and generate inventory distribution data; An inventory optimization module, for dynamically calculating replenishment quantities based on the inventory distribution data, combined with logistics constraints and current inventory status, and generating replenishment data and inventory adjustment strategies; A behavior analysis module, for building a user preference model and a commodity demand model based on the user behavior data, and generating commodity recommendation data and promotion priorities in combination with the adjusted inventory distribution data; The logistics allocation module is used to dynamically adjust the logistics distribution path and distribution priority based on the replenishment data and the promotion priority.

2. The retail supply chain optimization system based on big data analysis according to claim 1 is characterized in that: The data acquisition module comprises: An order data collection unit, used to collect the order data including but not limited to order number, product details and transaction amount through an order management interface; A logistics data collection unit, used to collect the logistics data including but not limited to the delivery route, transportation status and receipt record through the logistics information interface; A behavior data collection unit, used to collect the user behavior data including but not limited to browsing tracks, product click records and preference information through a user behavior monitoring interface; A formatting processing unit is used to format the order data, logistics data and user behavior data to generate the structured data.

3. The retail supply chain optimization system based on big data analysis according to claim 1 is characterized in that: The data processing module comprises: A data cleaning unit, used for preprocessing the structured data by removing duplicate data, filling missing data and eliminating abnormal data; A data standardization unit, used for standardizing the data processed by the data cleaning unit by unifying the field format, time format and data unit; The data association unit is used to receive the data processed by the data standardization unit and establish a multi-dimensional association relationship between the order data, logistics data and user behavior data based on association rules and feature extraction algorithms.

4. The retail supply chain optimization system based on big data analysis according to claim 1 is characterized in that: The inventory forecasting module includes: A demand forecasting unit, configured to construct the time series model based on the multi-dimensional association relationship to perform trend analysis on the order data and forecast future order demand; A feature analysis unit, configured to calculate the regional inventory demand through the regional feature model based on the prediction result of the demand prediction unit and in combination with the regional distribution information in the logistics data; The inventory distribution calculation unit is used to generate the inventory distribution data based on the inventory demand data and in combination with the current inventory status.

5. The retail supply chain optimization system based on big data analysis according to claim 1 is characterized in that: The constructing of the time series model based on the multi-dimensional association relationship to perform trend analysis on the order data and predict future order demand includes: Based on the multidimensional correlation relationship, a multidimensional time series matrix X is constructed. t =[x i，j，t ], where x i，j，t represents the number of orders for the i-th category of goods in the j-th region at time t, i∈{1,2,...,n}, j∈{1,2,...,m}, n represents the total number of product categories, and m represents the total number of regions; Based on the time series matrix X t , construct the standardized sequence matrix Z through detrending and standardization operations t ,in μ i，j represents the historical average order volume of the i-th category of goods in the j-th region, σ i，j represents its historical standard deviation; The standardized sequence matrix Z is transformed into t Perform feature extraction and build time series prediction model Z t+1 =F·Z t +G·U t , where F is the time correlation factor matrix, G is the perturbation coefficient matrix, and U t represents the disturbance vector at time t; Based on the time series forecasting model, combined with the future time step T, the order demand forecast matrix is ​​calculated in Represents the order demand matrix at future time t+T.

6. The retail supply chain optimization system based on big data analysis according to claim 1 is characterized in that: The inventory optimization module includes: An inventory status analysis unit, used to analyze the current inventory status based on the inventory distribution data, and generate inventory remaining quantity and inventory demand difference data; A replenishment quantity calculation unit, used to calculate the required replenishment quantity through a dynamic replenishment algorithm based on the inventory demand difference data and the logistics constraint conditions; A strategy generating unit is used to generate the replenishment data and inventory adjustment strategy based on the required replenishment quantity and in combination with the current inventory status.

7. The retail supply chain optimization system based on big data analysis according to claim 1 is characterized in that: The behavior analysis module includes: A user preference modeling unit, for identifying the characteristics of user browsing, clicking and purchasing behaviors through a behavior feature extraction algorithm based on the user behavior data, and building a user preference model; A commodity demand modeling unit, configured to extract demand association relationships between commodities using a commodity association rule mining algorithm based on the user preference model and the order data, and to construct a commodity demand model; A recommendation strategy generating unit is used to generate the commodity recommendation data and promotion priority based on the user preference model and the commodity demand model in combination with the adjusted inventory distribution data.

8. The retail supply chain optimization system based on big data analysis according to claim 1 is characterized in that: Based on the user behavior data, identifying the characteristics of user browsing, clicking and purchasing behaviors through a behavior feature extraction algorithm, and building a user preference model, includes: Based on the user behavior data, construct a user behavior matrix B i，j，t =[b i，j，t ], where b i，j，t represents the behavior score of user i on product j at time t, and the behavior score formula is defined as in, They represent the number of views, clicks, and purchases, respectively, and w1, w2, and w3 represent the behavior weights; The singular value decomposition method is applied to the user behavior matrix to decompose the behavior score into the user potential preference vector Pi = [pi, k] and the product feature vector Q j =[q j , k], where B = P·Q T , Pi represents the preference weight of user i on the potential preference dimension, Q j represents the feature weight of product j; The decomposition model is optimized by regularized least squares, and the objective function is defined as Among them, λ represents the regularization coefficient; Combining the user's historical behavior and time decay factor to construct the user preference model U based on time dynamics i，t =P i +β·T t , where T t =exp(-γt), β represents the time influence coefficient, and γ represents the time attenuation weight.

9. The retail supply chain optimization system based on big data analysis according to claim 1, characterized in that: The logistics deployment module includes: A path planning unit, configured to construct a distribution path model based on the replenishment data and the distribution area information in the logistics data, and optimize the logistics distribution path by using a shortest path algorithm; A priority adjustment unit, configured to analyze the commodity demand priority in the distribution task based on the promotion priority and the commodity recommendation data, and dynamically adjust the task sequence of the logistics distribution; The task allocation unit is used to match logistics resources and generate specific delivery instructions based on the delivery path generated by the path planning unit and the delivery task sequence generated by the priority adjustment unit.

10. A retail supply chain optimization method based on big data analysis, characterized in that: include: Collecting order data, logistics data and user behavior data, and formatting the order data, logistics data and user behavior data to generate structured data; Preprocessing the structured data, and establishing a multi-dimensional correlation relationship between the order data, logistics data and user behavior data based on the preprocessing result; Based on the multi-dimensional association relationship, a time series model and a regional characteristic model are constructed to predict future order demand, calculate regional inventory demand and generate inventory distribution data; Based on the inventory distribution data, combined with logistics constraints and current inventory status, dynamically calculate the replenishment quantity, generate replenishment data and inventory distribution adjustment strategy; Based on the user behavior data, a user preference model and a commodity demand model are constructed, and commodity recommendation data and promotion priorities are generated in combination with the inventory distribution adjustment strategy; Based on the replenishment data and the promotion priority, the logistics distribution path and distribution priority are dynamically adjusted.

Images