A warehouse intelligent management method and system based on order analysis

Abstract

The application discloses a kind of order analysis-based warehouse intelligent management method and system, belong to warehouse management technical field, construct order dataset and fine prediction analysis model, according to fine prediction analysis model, order dataset is handled, and fine prediction data is obtained;According to dynamic adjustment strategy, fine prediction data and the space data of target warehouse are handled, and the adjusted position of each commodity is obtained;According to path optimization algorithm, the space data of target warehouse and the adjusted position of each commodity are handled, and the optimized path of each commodity is obtained Export and the optimized path of warehousing;According to the optimized path of target warehouse export response and warehousing, each commodity is replenished and handled, and the intelligent management of target warehouse is completed;Through fine prediction analysis model, time feature and commodity sequence feature in order data can be more accurately extracted, to improve the accuracy of correlation and sales forecast.

Description

Technical Field

[0001] This application belongs to the field of warehouse management technology, specifically relating to a warehouse intelligent management method and system based on order analysis. Background Technology

[0002] With the rapid development of the e-commerce industry, intelligent warehouse management is becoming an industry trend.

[0003] Currently, in existing warehousing systems, the storage location of goods usually relies on static rules, such as partitioning by category or random storage, lacking in-depth mining of historical order data. This leads to technical problems in existing warehouse management, such as rigid storage location decisions, neglect of correlation, and low operational efficiency. Summary of the Invention

[0004] The purpose of this application is to provide a warehouse intelligent management method and system based on order analysis, which can solve the technical problems of rigid storage location decision-making, neglect of correlation and low operational efficiency in existing warehouse management.

[0005] To solve the above-mentioned technical problems, this application is implemented as follows:

[0006] In a first aspect, embodiments of this application provide a warehouse intelligent management method based on order analysis, the method comprising:

[0007] Obtain multiple historical orders from the target warehouse, and construct an order dataset based on the multiple historical orders;

[0008] A refined predictive analysis model is constructed, and the order dataset is processed based on the refined predictive analysis model to obtain refined predictive data;

[0009] The fine-grained forecast data and the spatial data of the target warehouse are processed according to the dynamic adjustment strategy to obtain the adjusted position of each product.

[0010] The spatial data of the target warehouse and the adjusted position of each commodity are processed according to the path optimization algorithm to obtain the outbound optimization path and the inbound optimization path of each commodity.

[0011] Based on the outbound response of the target warehouse and the inbound optimization path, replenishment is performed on each of the products to complete the intelligent management of the target warehouse;

[0012] The detailed prediction data includes the popularity of each product in the target warehouse and the correlation between the products.

[0013] As an optional implementation of the first aspect of this application, the refined prediction analysis model includes a refined extraction module, a correlation analysis module, and a prediction module; the refined prediction analysis model processes the order dataset to obtain refined prediction data, specifically as follows:

[0014] The fine extraction module extracts data from the order dataset using semantic segmentation and feature extraction algorithms to obtain an information dataset;

[0015] The association analysis module performs data association analysis on the information dataset to obtain the association analysis dataset.

[0016] The prediction module processes the information dataset and the correlation analysis dataset to obtain refined prediction data.

[0017] As an optional implementation of the first aspect of this application, the fine extraction module extracts data from the order dataset according to a semantic segmentation algorithm and a feature extraction algorithm to obtain an information dataset; specifically:

[0018] Scan each of the historical orders in the order dataset to obtain each scan data corresponding to each historical order;

[0019] The semantic segmentation algorithm is used to perform semantic segmentation on each of the scanned data to obtain each time information sequence and each product information sequence corresponding to each of the scanned data.

[0020] The feature extraction algorithm is used to extract features from each time information and each product information sequence to obtain the corresponding time features and product sequence features.

[0021] The order dataset is divided into four different quarterly order datasets based on each of the aforementioned time features;

[0022] The product sequence features corresponding to each quarterly order dataset are divided into four groups to obtain four different quarterly product sequence feature sets. The information dataset is constructed based on the four quarterly product sequence feature sets.

[0023] As an optional implementation of the first aspect of this application, the information dataset is subjected to data association analysis by the association analysis module to obtain an association analysis dataset; specifically:

[0024] The association analysis module performs data association analysis on each quarterly commodity sequence feature set in the information dataset to obtain each association analysis data subset corresponding to each quarterly commodity sequence feature set.

[0025] The association analysis dataset is constructed based on each subset of the association analysis data.

[0026] As an optional implementation of the first aspect of this application, the association analysis module performs data association analysis on the quarterly commodity sequence feature set, specifically as follows:

[0027] Based on the fact that the number of product types in the historical orders corresponding to each quarterly product sequence feature in the quarterly product sequence feature set is greater than one, the quarterly product sequence feature set is filtered to obtain multiple product feature sequence sets.

[0028] Set a support threshold and calculate the support of each product sequence feature in the set of multiple product feature sequences;

[0029] The product sequence features with a support greater than the support threshold in the set of multiple product feature sequences are set as multiple frequent features, and a set of multiple frequent features is constructed based on each of the multiple frequent features.

[0030] Calculate the confidence level of each of the multiple frequent features in the set of multiple frequent features and the support level of the multiple frequent features.

[0031] Calculate the lift of each of the multiple frequent features in the set of multiple frequent features based on the support and confidence of each of the multiple frequent features;

[0032] The association analysis data subset is constructed based on the set of multiple frequent features, the support, confidence, and lift of each of the multiple frequent features.

[0033] As an optional implementation of the first aspect of this application, the prediction module processes the information dataset and the correlation analysis dataset to obtain refined prediction data; specifically:

[0034] The information dataset is subjected to feature processing to obtain the popularity features of each product in the target warehouse;

[0035] Based on the support and confidence of each of the multiple frequent features, each of the multiple frequent features is weighted to obtain each weighted frequent feature corresponding to each of the multiple frequent features.

[0036] Based on the lifting degree of each of the multiple frequent features, each weighted frequent feature corresponding to each of the multiple frequent features is fine-tuned to obtain the fine-tuned frequent feature corresponding to each of the weighted frequent features.

[0037] The frequent fine-tuning feature and the product popularity feature are processed to obtain the popularity of each product in the target warehouse and the correlation between products, thereby obtaining the refined prediction data.

[0038] As an optional implementation of the first aspect of this application, the step of processing the fine-grained prediction data and the spatial data of the target warehouse according to a dynamic adjustment strategy to obtain the adjusted position of each commodity is as follows:

[0039] Based on the popularity of each product in the detailed prediction data, products with popularity greater than the first threshold are dynamically adjusted to a position near the outbound gate in the target warehouse.

[0040] Based on the correlation between products and products with popularity less than or equal to a first threshold, products with popularity greater than the first threshold are matched sequentially, according to the following matching rules:

[0041] If the correlation between a product whose current popularity is greater than the first threshold and a product whose popularity is less than or equal to the first threshold is greater than the second threshold, then the product whose popularity is less than or equal to the first threshold will be moved to a position near the product whose current popularity is greater than the first threshold.

[0042] If the correlation between a product with a current popularity greater than the first threshold and a product with a popularity less than or equal to the first threshold is less than or equal to the second threshold, then the product with a popularity less than or equal to the first threshold will be moved to the outbound position that is farthest from the target warehouse.

[0043] Secondly, embodiments of this application provide a warehouse intelligent management system based on order analysis, the system comprising:

[0044] Acquisition module: Acquires multiple historical orders from the target warehouse to construct an order dataset based on the multiple historical orders;

[0045] Prediction module: Constructs a refined predictive analysis model, processes the order dataset based on the refined predictive analysis model, and obtains refined predictive data;

[0046] Execution module: Processes the fine prediction data and the spatial data of the target warehouse according to the dynamic adjustment strategy to obtain the adjusted position of each commodity;

[0047] Optimization module: Processes the spatial data of the target warehouse and the adjusted position of each product according to the path optimization algorithm to obtain the outbound optimization path and inbound optimization path for each product;

[0048] Auxiliary module: Replenishes each product based on the outbound response of the target warehouse and the inbound optimization path, thereby completing the intelligent management of the target warehouse;

[0049] The detailed prediction data includes the popularity of each product in the target warehouse and the correlation between the products.

[0050] Thirdly, embodiments of this application provide an electronic device including a processor, a memory, and a program or instructions stored in the memory and executable on the processor, wherein the program or instructions, when executed by the processor, implement the steps of the method described in the first aspect.

[0051] Fourthly, embodiments of this application provide a readable storage medium on which a program or instructions are stored, which, when executed by a processor, implement the steps of the method described in the first aspect.

[0052] Compared with the prior art, this application has the following technical effects:

[0053] (1) Improve prediction accuracy:

[0054] By using a sophisticated predictive analysis model, combined with semantic segmentation and feature extraction algorithms, we can more accurately extract time features and product sequence features from order data. Dividing order data by quarter and constructing a quarterly product sequence feature set helps to capture seasonal sales trends, thereby improving the accuracy of sales forecasting.

[0055] (2) Enhance product association analysis:

[0056] The association analysis module is used to perform data association analysis on the quarterly product sequence feature set, and calculate the support, confidence and lift of multiple frequent features. These indicators help to identify the correlation between products, thereby optimizing the location arrangement of products and improving the level of intelligence in warehouse management.

[0057] (3) Optimize inventory management:

[0058] Based on the popularity and relevance of products, the location of products in the warehouse is dynamically adjusted. High-popularity products are placed near the outbound exit to improve outbound efficiency; products with high relevance are placed in close proximity to reduce picking paths and improve warehouse operation efficiency.

[0059] (4) Improve warehouse operation efficiency:

[0060] By using path optimization algorithms, the optimal paths for outbound and inbound operations are calculated, reducing unnecessary movement and waiting time and improving the overall operational efficiency of the warehouse. The combination of dynamic adjustment strategies and path optimization makes warehouse management more flexible and efficient.

[0061] (5) Supports intelligent replenishment decisions:

[0062] Intelligent replenishment is performed based on outbound response and inbound optimization paths to ensure that warehouse inventory is always kept at a reasonable level and to avoid stockouts or overstocking. By combining historical order data and predictive analysis, replenishment strategies can be better formulated to improve inventory turnover. Attached Figure Description

[0063] Figure 1 This is a flowchart illustrating an intelligent warehouse management method based on order analysis, provided by some embodiments of this application;

[0064] Figure 2 This is a structural diagram of a warehouse intelligent management method based on order analysis provided by some embodiments of this application. Detailed Implementation

[0065] The technical solutions of the embodiments of this application will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of this application, not all embodiments. Based on the embodiments of this application, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of this application.

[0066] The terms "first," "second," etc., used in the specification and claims of this application are used to distinguish similar objects and not to describe a specific order or sequence. It should be understood that such use of data can be interchanged where appropriate so that embodiments of this application can be implemented in orders other than those illustrated or described herein. Furthermore, in the specification and claims, "and / or" indicates at least one of the connected objects, and the character " / " generally indicates that the preceding and following objects are in an "or" relationship.

[0067] The following description, in conjunction with the accompanying drawings, details a warehouse intelligent management method and system based on order analysis provided in this application through specific embodiments and application scenarios.

[0068] Example

[0069] A warehouse intelligent management method based on order analysis includes the following steps:

[0070] S100: Obtain multiple historical orders from the target warehouse to construct an order dataset based on these historical orders;

[0071] It should be noted that S100 collects multiple historical order information from the target warehouse and integrates these order data to construct an order dataset; this historical order data is the basis for subsequent analysis and prediction, and contains the order situation of goods in the warehouse over a period of time.

[0072] S200: Construct a refined predictive analysis model, process the order dataset based on the refined predictive analysis model, and obtain refined predictive data;

[0073] It should be noted that a sophisticated predictive analytics model was used to process the order dataset. This model analyzes historical order data to uncover the popularity (i.e., how frequently a product is ordered) of each item in the target warehouse and the correlation between products (e.g., whether certain products are frequently ordered together). This analysis yields sophisticated predictive data, which provides crucial information for subsequent warehouse management decisions.

[0074] Furthermore, the refined predictive analysis model in S200 includes a refined extraction module, a correlation analysis module, and a prediction module; the refined predictive analysis model processes the order dataset to obtain refined predictive data, specifically as follows:

[0075] S210: The fine extraction module extracts data from the order dataset using semantic segmentation and feature extraction algorithms to obtain an information dataset;

[0076] S220: Perform data association analysis on the information dataset using the association analysis module to obtain the association analysis dataset;

[0077] S230: The prediction module processes the information dataset and the correlation analysis dataset to obtain refined prediction data.

[0078] It should be noted that the fine extraction module in S210 first processes the order dataset using semantic segmentation and feature extraction algorithms. Semantic segmentation identifies and extracts key information from orders, such as product name, quantity, and order time. Feature extraction extracts important features reflecting product characteristics and order behavior, such as sales frequency and seasonality. Through this fine extraction module, the order dataset is transformed into an information dataset. This dataset contains detailed product features and key order information, providing a foundation for subsequent analysis and prediction. The association analysis module in S220 further analyzes the information dataset to identify relationships between products. By using association rule mining techniques, the module can identify frequently ordered product combinations. This analysis helps identify potential relationships between products, such as complementary or substitute products. The output of the association analysis module is an association analysis dataset containing information on the degree of association between products. This information provides crucial information for subsequent predictions, particularly in product location adjustments and inventory management. In S230, the forecasting module integrates the information dataset and the correlation analysis dataset, and uses forecasting algorithms (such as time series analysis and machine learning models) to predict the popularity and demand of goods. By analyzing historical data, the forecasting module can estimate the demand trend of goods over a future period. The output of the forecasting module is a detailed forecast, which includes the popularity of each product (i.e., the predicted demand) and the correlation between products. This data provides important basis for warehouse management decisions, such as the relocation of goods and inventory management strategies.

[0079] It's important to understand that by combining semantic segmentation, feature extraction, association analysis, and prediction algorithms, sophisticated predictive analytics models can more accurately predict product demand and correlations. This accuracy provides reliable data support for warehouse management. Based on predicted product popularity and correlations, warehouse managers can optimize inventory layout, placing best-selling and related products in more easily accessible locations, thus improving warehouse operational efficiency. Accurate predictions help warehouses avoid inventory backlogs and stockouts, reducing inventory costs and improving capital turnover efficiency. Sophisticated predictive data provides comprehensive decision support for warehouse management, helping managers make better decisions regarding inventory management, replenishment strategies, and layout adjustments. By introducing various data analysis and prediction technologies, sophisticated predictive analytics models enhance the system's intelligence, making warehouse management more automated and efficient.

[0080] Furthermore, the fine extraction module in S210 extracts data from the order dataset using semantic segmentation and feature extraction algorithms to obtain an information dataset; specifically:

[0081] S211: Scan each historical order in the order dataset to obtain each scan data corresponding to each historical order;

[0082] S212: Perform semantic segmentation on each scanned data according to the semantic segmentation algorithm to obtain the time information sequence and the product information sequence corresponding to each scanned data.

[0083] S213: Extract features from each time information and each product information sequence according to the feature extraction algorithm to obtain the corresponding time features and product sequence features;

[0084] S214: Divide the order dataset into four different quarterly order datasets based on the time features of each period;

[0085] S215: Divide the product sequence features corresponding to each quarter's order dataset into a group to obtain four different quarterly product sequence feature sets, and construct an information dataset based on the four quarterly product sequence feature sets.

[0086] It should be noted that, firstly, S211 scans each historical order in the order dataset. This includes reading detailed information for each order, such as order time, product name, and product quantity; the result of the scan is each scanned data corresponding to each historical order. This scanned data contains basic order information, providing raw data for subsequent semantic segmentation and feature extraction. S212 processes each scanned data according to a semantic segmentation algorithm (such as the Segment Anything Model (SAM)) used to identify and separate different information categories in the order data. For example, it can identify time information and product information in an order; through semantic segmentation, the module divides each scanned data into a time information sequence and a product information sequence. The time information sequence contains the order's time information, and the product information sequence contains the products involved in the order and their quantity information. S213 uses a feature extraction algorithm to process each time information and product information sequence. The purpose of feature extraction is to extract key features that reflect the characteristics of the data from the raw data; the result of feature extraction is the time feature corresponding to each time information and the product sequence feature corresponding to each product information sequence. Temporal features can include the temporal distribution and seasonality of orders, while product sequence features can include the sales frequency and quantity of products. Then, based on each temporal feature, the order dataset is divided into four distinct quarterly order datasets. This division is based on temporal features (such as season and month) to facilitate subsequent seasonal analysis and forecasting; each quarterly order dataset contains order information for a specific time period, providing a foundation for analyzing seasonal sales trends.

[0087] Finally, S215 divides the product sequence features corresponding to each quarter's order dataset into four different quarterly product sequence feature sets. Based on these four quarterly product sequence feature sets, the module constructs an information dataset. This dataset contains product feature information divided by quarter, providing basic data for subsequent correlation analysis and prediction.

[0088] It's important to understand that through semantic segmentation and feature extraction, the module can accurately identify and extract key information from order data, improving the accuracy of data processing. By dividing the order dataset by quarter, the module supports seasonal analysis of product sales trends, which is crucial for predicting future sales and adjusting inventory strategies. Through the analysis of product sequence characteristics, the module helps warehouse managers better understand product sales patterns, thereby optimizing inventory management and replenishment strategies. By constructing information datasets, the module provides structured data support for subsequent correlation analysis and forecasting, enhancing the system's predictive capabilities. By automating data processing and extraction, the module improves the efficiency of the entire system and reduces the need for manual intervention.

[0089] Furthermore, in S220, the information dataset is subjected to data association analysis by the association analysis module to obtain the association analysis dataset; specifically:

[0090] S221: Based on the association analysis module, perform data association analysis on the product sequence feature set of each quarter in the information dataset to obtain each association analysis data subset corresponding to the product sequence feature set of each quarter;

[0091] S222: Construct an association analysis dataset based on each subset of association analysis data.

[0092] It should be noted that the association analysis module in S221 first performs data association analysis on the product sequence feature set for each quarter in the information dataset. Since product sales may be affected by seasonal factors in different quarters, quarterly analysis can more accurately capture the relationships between products. Within each quarter's product sequence feature set, the module uses association rule mining algorithms to identify frequently occurring product combinations. These algorithms can identify products that frequently appear together in orders, thus revealing potential relationships between products. The result of the association analysis is each subset of association analysis data corresponding to the product sequence feature set for each quarter. These subsets contain the association rules between products within that quarter, such as "product A and product B are frequently ordered together." After obtaining the association analysis data subsets for each quarter, the module in S222 integrates these subsets to construct a complete association analysis dataset. This dataset contains the association rules between products across all quarters, providing comprehensive data support for subsequent predictions and decision-making.

[0093] It's important to understand that through association analysis, the module can reveal potential relationships between products, which is crucial for understanding consumer purchasing behavior, optimizing product layout, and managing inventory. Based on these relationships, the system can provide personalized product recommendations to consumers, improving customer satisfaction and loyalty. By understanding these relationships, warehouse managers can more rationally allocate inventory, avoiding overstocking or stockouts. For example, frequently ordered items can be placed in close proximity for faster retrieval. The association analysis dataset provides additional information for subsequent prediction modules, helping to improve prediction accuracy. By considering the relationships between products, the predictive model can more accurately predict product demand and sales trends. The introduction of the association analysis module enhances the system's intelligence, enabling it to automatically discover potential relationships between products and providing more intelligent decision support for warehouse management.

[0094] Furthermore, the correlation analysis module in S221 performs data correlation analysis on the quarterly commodity sequence feature set, specifically as follows:

[0095] S2211: Based on the fact that the number of product types in the historical orders corresponding to each quarterly product sequence feature in the quarterly product sequence feature set is greater than one, the quarterly product sequence feature set is filtered to obtain a set of multiple product feature sequences.

[0096] S2212: Set a support threshold and calculate the support of each product sequence feature in the set of multiple product feature sequences;

[0097] S2213: Set the product sequence features with support greater than the support threshold in the multi-product feature sequence set as multi-product frequent features, and construct a multi-product frequent feature set based on each multi-product frequent feature;

[0098] S2214: Calculate the confidence level of each frequent multinomial feature based on each frequent multinomial feature in the set of frequent multinomial features and the support level of each frequent multinomial feature.

[0099] S2215: Calculate the lift of each frequent multinomial feature in the set of frequent multinomial features based on the support and confidence of each frequent multinomial feature.

[0100] S2216: Construct a subset of association analysis data based on a set of multiple frequent features, the support, confidence, and lift of each multiple frequent feature.

[0101] Specifically, the subset of association analysis in S2216 is represented by the following formula:

[0102] , , ,

[0103] in, This represents a subset of the correlation analysis data for the quarter. Indicates the first Multiple frequent features, Indicates support level, Indicates the confidence level. Indicates the degree of elevation. Represents a set of multiple frequent features. express The subset that maximizes the confidence level. express non-empty subset Indicates the maximum average;

[0104] It should be noted that, firstly, S2211 filters orders based on whether the number of product types in the historical orders corresponding to each quarter's product sequence feature set is greater than one. Only orders containing multiple products are retained because only these orders can provide information on the relationships between products. After filtering, a multi-product feature sequence set is obtained. This set contains all order information that meets the criteria, and these orders contain multiple products, forming the basis for association analysis. Then, module S2212 sets a support threshold to measure the prevalence of product sequence features in the order dataset; support refers to the proportion of orders containing a certain product sequence feature among all orders. For each product sequence feature in the multi-product feature sequence set, the module calculates its support. Secondly, S2213 defines product sequence features in the multi-product feature sequence set whose support is greater than the support threshold as multi-frequent features; these frequent features represent product combinations that frequently appear in orders. A multi-frequent feature set is constructed based on each multi-frequent feature. This set contains all product combinations that meet the support threshold. Next, in step S2214, for each frequent feature in the set of frequent features, the module calculates its confidence score; the confidence score represents the probability that an order containing product A also contains product B. Then, in step S2215, based on the support and confidence score of each frequent feature, the module calculates its lift. The lift score measures the strength of the association between product A and product B, indicating whether the probability of product B appearing in an order containing product A is significantly higher than its probability in all orders. Finally, in step S2216, based on the set of frequent features, the support, confidence score, and lift score of each frequent feature, the module constructs a subset of association analysis data; this subset contains all frequent features and their related indicators, providing comprehensive data support for subsequent analysis and prediction.

[0105] It's important to understand that by calculating support, confidence, and lift, the module can identify frequently occurring product combinations in orders, helping to understand consumer purchasing behavior. By identifying frequent product combinations, warehouse managers can optimize product layout, placing frequently purchased items in close proximity to improve order processing efficiency. Understanding the relationships between products helps managers better predict demand, avoiding inventory buildup or shortages, thereby improving inventory management efficiency. By analyzing these relationships, the system can provide personalized product recommendations to consumers, increasing customer satisfaction. The subset of data from the correlation analysis provides rich information for warehouse management, helping managers make better decisions regarding inventory management, replenishment strategies, and layout adjustments.

[0106] Furthermore, the prediction module in S230 processes the information dataset and the correlation analysis dataset to obtain refined prediction data; specifically:

[0107] S231: Perform feature processing on the information dataset to obtain the popularity features of each product in the target warehouse;

[0108] S232: Based on the support and confidence of each multiple frequent feature, perform weighted processing on each multiple frequent feature to obtain each weighted frequent feature corresponding to each multiple frequent feature.

[0109] S233: Based on the lifting degree of each multi-term frequent feature, fine-tune each weighted frequent feature corresponding to each multi-term frequent feature to obtain the fine-tuned frequent feature corresponding to each weighted frequent feature.

[0110] S234: Process the frequent fine-tuning features and product popularity features to obtain the popularity of each product in the target warehouse and the correlation between products, thereby obtaining refined prediction data.

[0111] Specifically, the popularity of each product in S234 is represented by the following formula:

[0112] ,

[0113] in, Indicates the first The popularity of a product Indicates the first The basic popularity of a product Indicates containing the first The sum of the fine-tuning values ​​of all frequent features of a product. Represents the basic heat calculation function, the first... The basic popularity of a product is equal to the frequency with which that product appears in the data set. Indicates the first One product, Represents an information dataset. Indicates the first A frequently adjusted feature, This indicates fine-tuning of the intensity parameters. Indicates the first A weighted frequent feature, and All represent weighted parameters;

[0114] The degree of correlation between the upper and lower levels in S234 is expressed by the following formula:

[0115] ,

[0116] in, Indicates the first The first item and the first The correlation between individual products Indicates containing the first The first item and the first The set consisting of all the frequent features of a product.

[0117] It should be noted that, firstly, the prediction module in S231 performs feature processing on the information dataset; the information dataset contains product sequence features divided by quarter, such as sales frequency and quantity. The goal of feature processing is to extract indicators that reflect product popularity from these features; by calculating indicators such as sales volume and growth rate of each product in different time periods, the module obtains the product popularity characteristics of each product in the target warehouse, which reflect the product's popularity and sales trend in the market. Secondly, in S232, based on the support and confidence of each multiple frequent feature in the association analysis dataset, the module performs weighted processing on these frequent features; support and confidence reflect the frequency and reliability of product combinations appearing in orders; by assigning a weight to each frequent feature (based on its support and confidence), the module obtains each weighted frequent feature corresponding to each multiple frequent feature; these weighted features emphasize more frequent and reliable product combinations. Next, in S233, the weighted frequent features are fine-tuned based on the lift of each multiple frequent feature. Lift measures the strength of the association between product combinations, indicating whether the appearance of one product significantly increases the probability of the appearance of another. By adjusting the weights of the weighted frequent features (based on their lift), the module obtains the fine-tuned frequent features corresponding to each weighted frequent feature. These fine-tuned features more accurately reflect the actual association between products. Finally, in S234, the prediction module integrates the fine-tuned frequent features and product popularity features. By combining product popularity and the association between products, the module can gain a more comprehensive understanding of the sales potential and mutual influence of products. Ultimately, the module obtains the popularity of each product in the target warehouse and the association between products, thus constructing refined prediction data. This dataset contains the predicted popularity of products (i.e., sales potential in the future) and the association information between products, providing important basis for warehouse management decisions.

[0118] It's important to understand that by combining product popularity characteristics and correlation analysis results, the forecasting module can more accurately predict product sales trends and demand. Product popularity characteristics reflect the market appeal of a product, while correlation analysis reveals potential relationships between products; combining the two improves forecast accuracy. Detailed forecast data provides warehouse managers with detailed information on product popularity and correlation. Managers can use this information to optimize inventory layout, ensuring sufficient stock of best-selling and related products while avoiding overstocking. Detailed forecast data provides comprehensive data support for warehouse management decisions. Managers can develop replenishment strategies, adjust product layout, and optimize promotional activities based on forecast results to improve warehouse operational efficiency and customer satisfaction. The introduction of the forecasting module enhances the system's intelligence. By automatically processing and analyzing data, the module can generate detailed forecast results, providing intelligent decision support for warehouse management. By accurately forecasting product demand and optimizing inventory management, warehouses can better meet customer needs, reduce stockouts, and thus improve customer experience and loyalty.

[0119] S300: Based on the dynamic adjustment strategy, the fine forecast data and the spatial data of the target warehouse are processed to obtain the adjusted position of each product;

[0120] It should be noted that, based on a dynamic adjustment strategy, combined with refined forecast data (product popularity and relevance) and spatial data of the target warehouse (such as warehouse layout and shelf distribution), the location of each product is adjusted. The purpose of the dynamic adjustment strategy is to place products in more suitable locations based on their characteristics and historical order data, thereby improving warehouse operational efficiency.

[0121] Furthermore, in S300, the refined forecast data and the spatial data of the target warehouse are processed according to a dynamic adjustment strategy to obtain the adjusted location of each product; specifically:

[0122] S310: Based on the popularity of each product in the detailed prediction data, dynamically adjust products with popularity greater than the first threshold to a position near the outbound gate in the target warehouse;

[0123] S320: Based on the correlation between products and products with popularity less than or equal to the first threshold, products with popularity greater than the first threshold are matched sequentially, according to the following matching rules:

[0124] S321: If the correlation between a product whose current popularity is greater than the first threshold and a product whose popularity is less than or equal to the first threshold is greater than the second threshold, then the product whose popularity is less than or equal to the first threshold will be moved to a position near the product whose current popularity is greater than the first threshold.

[0125] S322: If the correlation between a product with a current popularity greater than the first threshold and a product with a popularity less than or equal to the first threshold is less than or equal to the second threshold, then the product with a popularity less than or equal to the first threshold will be moved to the outbound position furthest from the target warehouse.

[0126] It should be noted that S310, based on the popularity of each product in the detailed prediction data, filters out products with popularity exceeding a first threshold and dynamically adjusts them to positions near the outbound exit in the target warehouse. Product popularity reflects a product's market appeal and sales potential; high-popularity products typically mean higher outbound frequency. Placing high-popularity products near the outbound exit reduces the walking distance for picking personnel, improves picking efficiency, speeds up order processing, reduces picking time, and improves overall warehouse operational efficiency. By optimizing picking routes, it reduces wasted manpower and time. In S320, based on the correlation between products and products with popularity less than or equal to the first threshold, products with popularity exceeding the first threshold are matched sequentially, and their positions are adjusted according to the matching rules. The matching rules are as follows: Rule 1 (S321): If the correlation between a product with current popularity greater than the first threshold and a product with popularity less than or equal to the first threshold is greater than the second threshold, then the product with popularity less than or equal to the first threshold will be moved to a location near the product with current popularity greater than the first threshold. High correlation usually means that products are frequently ordered together; placing these products in close proximity can further reduce the walking distance of picking personnel and improve picking efficiency. Close placement of related products also improves the collaborative efficiency of order processing, reduces order processing time, speeds up product outbound delivery, and enhances customer experience. Rule 2 (S322): If the correlation between a product with current popularity greater than the first threshold and a product with popularity less than or equal to the first threshold is less than or equal to the second threshold, then the product with popularity less than or equal to the first threshold will be moved to an outbound location far from the target warehouse. Low correlation usually means that products are rarely ordered together. Placing these goods away from the outbound exit reduces interference with the picking paths of high-demand items and makes better use of warehouse space. A well-planned layout improves warehouse space utilization, prevents low-related items from affecting the picking paths of high-demand items, and increases picking efficiency. By prioritizing based on popularity and matching based on relevance, the warehouse can process orders more quickly, reducing picking time and labor costs. Dynamically adjusting the location of goods based on their popularity and relevance makes the warehouse layout more rational and improves space utilization. This dynamic adjustment strategy allows for real-time adjustments to product locations based on market changes and sales performance, maintaining high warehouse operational efficiency.

[0127] It should be understood that, based on actual operation, after implementing this embodiment to dynamically adjust the goods in the target warehouse, compared with before the dynamic adjustment, picking efficiency increased by 30%, order processing time was shortened by 20%, and the space utilization rate of the target warehouse increased by 15%, reducing inventory backlog and waste.

[0128] S400: Based on the path optimization algorithm, the spatial data of the target warehouse and the adjusted position of each product are processed to obtain the outbound optimization path and inbound optimization path for each product.

[0129] It should be noted that path optimization algorithms (such as Dijkstra's algorithm, A* algorithm, or ant colony optimization) are used to comprehensively consider the spatial data of the target warehouse and the adjusted position of each product to calculate the optimized outbound and inbound paths for each product. The goal of path optimization algorithms is to find the shortest path, reduce the handling distance and time of products within the warehouse, and thus improve warehouse operational efficiency.

[0130] Furthermore, in S400, the spatial data of the target warehouse and the adjusted position of each product are processed according to the path optimization algorithm to obtain the optimized outbound path and the optimized inbound path for each product; specifically:

[0131] S410: Establish a three-dimensional coordinate system for the target warehouse based on its spatial data;

[0132] S420: Based on the target warehouse coordinate system and the adjusted position of each product, establish the product coordinates corresponding to the adjusted position of each product;

[0133] S430: Connect key nodes of the target warehouse according to the target warehouse coordinate system to obtain multiple target connection channels;

[0134] S440: The path optimization algorithm optimizes the path for each product coordinate in real time based on real-time orders, product coordinates, and target connection channels, resulting in the optimized outbound path and optimized inbound path for each product.

[0135] It should be noted that S410 constructs a three-dimensional coordinate system based on the spatial data of the target warehouse (such as shelf layout, aisle location, entrance / exit location, etc.), mapping each physical location in the warehouse to the coordinate system. This three-dimensional coordinate system accurately describes the location of each point within the warehouse, including key nodes such as shelves, aisles, and entrances / exits. A unique coordinate value is assigned to each key node in the warehouse, facilitating subsequent path planning and optimization. S420 determines the specific coordinates of each product in the coordinate system based on the target warehouse coordinate system and the adjusted position of each product. The adjusted positions of the products obtained in step S300 (such as shelf number, layer number, column number, etc.) are mapped to the three-dimensional coordinate system to obtain the specific coordinates of the products. The coordinates of each product are stored in the system for quick retrieval during subsequent path planning. S430 connects key nodes within the warehouse using the target warehouse coordinate system, forming multiple target connection channels. All channels within the warehouse are identified, including main channels, secondary channels, and vertical channels. Based on the physical connection relationships of the channels, the connection relationships between channels are established in the coordinate system, forming a target connection channel network. In S440, the path optimization algorithm optimizes the path for each product coordinate in real time based on real-time orders, product coordinates, and target connection channels, resulting in optimized outbound and inbound paths. The system receives real-time order information and determines the list of products that need to be outbound or inbound. Based on product coordinates and target connection channels, it uses path optimization algorithms (such as A* algorithm, Dijkstra algorithm, etc.) to plan the optimal path for each product. Taking into account real-time conditions in the warehouse (such as channel congestion, equipment failure, etc.), the system dynamically adjusts and optimizes the paths to guide warehouse operators to perform efficient operations.

[0136] It's important to understand that establishing a three-dimensional coordinate system and product coordinates allows for a precise description of the location and aisle connections of each product within the warehouse, thereby improving the accuracy of path planning. Precise spatial positioning and aisle connections provide a reliable data foundation for path planning, reducing errors caused by inaccurate location information. Path optimization algorithms can plan optimal outbound and inbound routes for each product, reducing the walking distance and time for warehouse workers and improving operational efficiency. Optimal path planning considers real-time conditions such as aisle congestion and equipment malfunctions, ensuring efficient and smooth operations. Reducing the walking distance and time for workers lowers labor costs; simultaneously, optimized path planning reduces equipment wear and energy consumption, lowering equipment costs. Efficient path planning reduces unnecessary movement and waiting time, thus lowering operating costs. Path optimization algorithms can dynamically adjust and optimize based on real-time orders and warehouse conditions, enhancing the system's flexibility and adaptability. Real-time data processing and dynamic path planning enable the system to quickly respond to changes within the warehouse, ensuring operational efficiency and accuracy. Improving warehouse operational efficiency and accuracy accelerates order processing and increases customer satisfaction.

[0137] S500: Replenishes each item based on the outbound response and inbound optimization path of the target warehouse, thus completing the intelligent management of the target warehouse.

[0138] It should be noted that replenishment is performed on each item based on the outbound response (i.e., real-time orders or actual outbound status) and the optimized inbound path of the target warehouse. In this way, the system can replenish inventory in a timely manner according to actual demand, ensuring sufficient supply of goods in the warehouse, while optimizing inventory layout and operation processes, and completing intelligent management of the target warehouse.

[0139] It should be noted that the executor of the order analysis-based intelligent warehouse management method provided in this application embodiment can be an order analysis-based intelligent warehouse management system, or a control module within that system for executing the order analysis-based intelligent warehouse management method. This application embodiment uses the execution of the order analysis-based intelligent warehouse management method by an order analysis-based intelligent warehouse management system as an example to illustrate the order analysis-based intelligent warehouse management method provided in this application embodiment.

[0140] A warehouse intelligent management system based on order analysis includes:

[0141] Acquisition module: Acquires multiple historical orders from the target warehouse to construct an order dataset based on these historical orders;

[0142] Prediction module: Constructs a refined predictive analysis model, processes the order dataset based on the refined predictive analysis model, and obtains refined prediction data;

[0143] Execution module: Processes the fine forecast data and spatial data of the target warehouse according to the dynamic adjustment strategy to obtain the adjusted position of each product;

[0144] Optimization module: Based on the path optimization algorithm, the spatial data of the target warehouse and the adjusted position of each product are processed to obtain the optimized outbound path and the optimized inbound path for each product;

[0145] Auxiliary module: Replenishes each product based on the outbound response and inbound optimization path of the target warehouse, thus completing the intelligent management of the target warehouse;

[0146] The detailed forecast data includes the popularity of each product in the target warehouse and the correlation between products.

[0147] The warehouse intelligent management system based on order analysis in this application embodiment can be a device, or a component, integrated circuit, or chip in a terminal. The device can be a mobile electronic device or a non-mobile electronic device. For example, mobile electronic devices can be mobile phones, tablets, laptops, PDAs, ultra-mobile personal computers (UMPCs), netbooks, or personal digital assistants (PDAs), etc., while non-mobile electronic devices can be servers, network-attached storage (NAS), personal computers (PCs), etc. This application embodiment does not impose specific limitations.

[0148] The warehouse intelligent management system based on order analysis in this application embodiment can be a device with an operating system. This operating system can be Android, iOS, or other possible operating systems; this application embodiment does not specifically limit it.

[0149] The warehouse intelligent management system based on order analysis provided in this application embodiment can achieve... Figures 1 to 2 The various processes implemented in the method embodiment of the order analysis-based intelligent warehouse management method are not described in detail here to avoid repetition.

[0150] According to an order analysis-based intelligent warehouse management system in this embodiment, a sophisticated predictive analysis model (such as deep learning or ensemble learning algorithms) combined with historical order data accurately predicts future demand and fluctuation trends for goods. The predicted data includes product popularity (sales trends) and product correlation (such as the "beer and diapers" effect), providing data support for inventory layout and replenishment strategies. Based on product popularity and correlation, the system dynamically adjusts storage locations, placing high-demand products near the outbound exit and storing related products nearby. This optimized layout shortens order picking paths by 20%-40%, reducing the distance traveled by personnel or equipment. Path optimization algorithms (such as A* algorithm or ant colony optimization) generate the shortest or least congested inbound and outbound paths. Based on outbound responses (real-time orders) and optimized inbound paths, replenishment tasks are automatically triggered to ensure sufficient inventory. Through a closed-loop feedback mechanism, actual execution data (such as replenishment effectiveness and path efficiency) is fed back to the predictive model and dynamic adjustment strategies, enabling system self-optimization.

[0151] Optionally, this application embodiment also provides an electronic device, including a processor, a memory, and a program or instructions stored in the memory and executable on the processor. When the program or instructions are executed by the processor, they implement the various processes of the above-described embodiment of a warehouse intelligent management method based on order analysis and achieve the same technical effect. To avoid repetition, they will not be described again here.

[0152] This application also provides a readable storage medium storing a program or instructions. When the program or instructions are executed by a processor, they implement the various processes of the above-described embodiment of an intelligent warehouse management method based on order analysis and achieve the same technical effect. To avoid repetition, they will not be described again here.

[0153] The processor is the processor in the electronic device described in the above embodiments. The readable storage medium includes computer-readable storage media, such as computer read-only memory (ROM), random access memory (RAM), magnetic disk, or optical disk.

[0154] It should be noted that, in this document, the terms "comprising," "including," or any other variations thereof are intended to cover non-exclusive inclusion, such that a process, method, article, or apparatus that comprises a list of elements includes not only those elements but also other elements not expressly listed, or elements inherent to such a process, method, article, or apparatus. Without further limitations, an element defined by the phrase "comprising one..." does not exclude the presence of other identical elements in the process, method, article, or apparatus that includes that element. Furthermore, it should be noted that the scope of the methods and apparatuses in the embodiments of this application is not limited to performing functions in the order shown or discussed, but may also include performing functions substantially simultaneously or in the reverse order, depending on the functions involved. For example, the described methods may be performed in a different order than described, and various steps may be added, omitted, or combined. Additionally, features described with reference to certain examples may be combined in other examples.

[0155] Through the above description of the embodiments, those skilled in the art can clearly understand that the methods of the above embodiments can be implemented by means of software plus necessary general-purpose hardware platforms. Of course, they can also be implemented by hardware, but in many cases the former is a better implementation method. Based on this understanding, the technical solution of this application, in essence, or the part that contributes to the prior art, can be embodied in the form of a software product. This computer software product is stored in a storage medium (such as ROM / RAM, magnetic disk, optical disk) and includes several instructions to cause a terminal (which may be a mobile phone, computer, server, air conditioner, or network device, etc.) to execute the methods described in the various embodiments of this application.

[0156] The embodiments of this application have been described above with reference to the accompanying drawings. However, this application is not limited to the specific embodiments described above. The specific embodiments described above are merely illustrative and not restrictive. Those skilled in the art can make many other forms under the guidance of this application without departing from the spirit and scope of the claims, and all of these forms are within the protection scope of this application.

Claims

1. A warehouse intelligent management method based on order analysis, characterized in that, The method comprises: acquiring a plurality of historical orders of a target warehouse to construct an order dataset according to the plurality of historical orders; constructing a fine prediction analysis model, processing the order dataset according to the fine prediction analysis model, and obtaining fine prediction data; processing the fine prediction data and spatial data of the target warehouse according to a dynamic adjustment strategy to obtain an adjusted position of each commodity; processing the spatial data of the target warehouse and the adjusted position of each commodity according to a path optimization algorithm to obtain an optimized outbound path and an optimized inbound path of each commodity; restocking each commodity according to the outbound response of the target warehouse and the optimized inbound path to complete intelligent management of the target warehouse; wherein the fine prediction data comprises the heat of each commodity in the target warehouse and the correlation between commodities; the fine prediction analysis model comprises a fine extraction module, an association analysis module, and a prediction module; the fine prediction analysis model processes the order dataset to obtain fine prediction data, specifically: the fine extraction module extracts data from the order dataset according to a semantic segmentation algorithm and a feature extraction algorithm to obtain an information dataset; the association analysis module analyzes the information dataset to obtain an association analysis dataset; the prediction module processes the information dataset and the association analysis dataset to obtain fine prediction data; the fine extraction module extracts data from the order dataset according to a semantic segmentation algorithm and a feature extraction algorithm to obtain an information dataset; specifically: scanning each historical order in the order dataset to obtain each scan data corresponding to each historical order; performing semantic segmentation processing on each scan data according to the semantic segmentation algorithm to obtain each time information sequence and each commodity information sequence corresponding to each scan data; performing feature extraction on each time information and each commodity information sequence according to the feature extraction algorithm to obtain each time feature and each commodity sequence feature corresponding thereto; dividing the order dataset into four different quarterly order datasets according to each time feature; dividing the commodity sequence features corresponding to each quarterly order dataset into a group to obtain four different quarterly commodity sequence feature sets, and constructing the information dataset according to the four quarterly commodity sequence feature sets; the association analysis module analyzes the information dataset to obtain an association analysis dataset; specifically: the association analysis module analyzes each quarterly commodity sequence feature set in the information dataset respectively to obtain each association analysis data subset corresponding to each quarterly commodity sequence feature set; constructing the association analysis dataset according to each association analysis data subset; the association analysis module analyzes the quarterly commodity sequence feature set, specifically: screening the seasonal commodity sequence feature set to obtain a multi-item commodity feature sequence set, according to the fact that the number of types of commodities in the historical orders corresponding to each seasonal commodity sequence feature in the seasonal commodity sequence feature set is greater than one; setting a support threshold, and calculating the support of each commodity sequence feature in the multi-item commodity feature sequence set; setting the commodity sequence features with support greater than the support threshold in the multi-item commodity feature sequence set as multi-item frequent features, and constructing a multi-item frequent feature set according to each multi-item frequent feature; calculating the confidence of each multi-item frequent feature according to each multi-item frequent feature in the multi-item frequent feature set and the support of the multi-item frequent feature; calculating the lift of each multi-item frequent feature in the multi-item frequent feature set according to the support and confidence of each multi-item frequent feature; constructing the association analysis data subset according to the multi-item frequent feature set, the support, confidence and lift of each multi-item frequent feature; the prediction module processes the information data set and the association analysis data set to obtain fine prediction data; specifically: performing feature processing on the information data set to obtain commodity heat features of each commodity in the target warehouse; performing weighted processing on each multi-item frequent feature according to the support and confidence of each multi-item frequent feature to obtain each weighted frequent feature corresponding to each multi-item frequent feature; performing fine tuning on each weighted frequent feature corresponding to each multi-item frequent feature according to the lift of each multi-item frequent feature to obtain a fine-tuned frequent feature corresponding to each weighted frequent feature; processing the fine-tuned frequent features and the commodity heat features to obtain the heat of each commodity and the correlation between commodities in the target warehouse, and further obtaining the fine prediction data. 2.The warehouse intelligent management method based on order analysis of claim 1, wherein, the fine prediction data and the spatial data of the target warehouse are processed according to the dynamic adjustment strategy to obtain the adjusted position of each commodity; specifically: according to the heat of each commodity in the fine prediction data, adjusting the commodities with heat greater than a first threshold to positions close to the delivery port in the target warehouse; according to the correlation between commodities and commodities with heat less than or equal to the first threshold, sequentially matching the commodities with heat greater than the first threshold, and the matching rule is as follows: if the correlation between the current commodity with heat greater than the first threshold and the commodity with heat less than or equal to the first threshold is greater than a second threshold, the commodity with heat less than or equal to the first threshold is adjusted to a position close to the current commodity with heat greater than the first threshold; if the correlation between the current commodity with heat greater than the first threshold and the commodity with heat less than or equal to the first threshold is less than or equal to the second threshold, the commodity with heat less than or equal to the first threshold is adjusted to a position away from the delivery port of the target warehouse.

3. A warehouse intelligent management system based on order analysis, capable of realizing the warehouse intelligent management method based on order analysis in any one of claims 1-2, characterized in that, the system comprises: an acquisition module: acquiring a plurality of historical orders of a target warehouse to construct an order data set according to the plurality of historical orders; The prediction module: constructs a fine prediction analysis model, processes the order dataset according to the fine prediction analysis model, and obtains fine prediction data; The execution module: processes the fine prediction data and the space data of the target warehouse according to a dynamic adjustment strategy, and obtains the adjusted position of each commodity; The optimization module: processes the space data of the target warehouse and the adjusted position of each commodity according to a path optimization algorithm, and obtains the outbound optimization path and the inbound optimization path of each commodity; The auxiliary module: replenishes each commodity according to the outbound response of the target warehouse and the inbound optimization path, and completes the intelligent management of the target warehouse; The fine prediction data includes the heat of each commodity in the target warehouse and the correlation between commodities.

4. An electronic device, comprising: The processor, the memory, and the program or instructions stored on the memory and executable on the processor are included. The program or instructions are executed by the processor to implement the steps of the warehouse intelligent management method based on order analysis according to any one of claims 1-2.

5. A readable storage medium characterized by, The program or instructions are stored on the readable storage medium, and the program or instructions are executed by the processor to implement the steps of the warehouse intelligent management method based on order analysis according to any one of claims 1-2.

Images