Order scheduling method and device for physical commodity order management systems

Abstract

The present invention relates to the technical field of order management. Disclosed are an order scheduling method and device for physical commodity order management systems. The method comprises: establishing a commodity library network; constructing a physical commodity storage template, and on the basis of the physical commodity storage template, storing physical commodities in a plurality of physical commodity libraries into a plurality of physical commodity order management systems, respectively to form a plurality of physical commodity lists; configuring a bidirectional order interaction channel, and embedding the bidirectional order interaction channel into the commodity library network to generate an order scheduling connectivity network; receiving physical commodity order demands of the plurality of physical commodity libraries, and inputting the physical commodity order demands and the plurality of physical commodity lists into a scheduling decision-making model to obtain an order scheduling scheme; and configuring the order scheduling connectivity network on the basis of the order scheduling scheme, and executing order scheduling. The present invention solves the technical problems in the prior art of low efficiency and insufficient accuracy in order scheduling, thereby achieving the technical effect of improving the efficiency and accuracy of order scheduling.

Description

Order scheduling method and apparatus for physical order management system Technical Field

[0001] This invention relates to the field of order management technology, and more specifically to an order scheduling method and apparatus for a physical order management system. Background Technology

[0002] With the rapid development of e-commerce and logistics, physical order management systems play a crucial role in supply chain management and logistics distribution. However, with the surge in order volume and the diversification of delivery needs, traditional order scheduling methods are no longer sufficient to meet the requirements of efficiency, accuracy, and real-time performance in modern order management.

[0003] In traditional physical order management systems, order scheduling often relies on manual operation and experience-based judgment. This approach is not only inefficient but also prone to errors and omissions. Furthermore, the lack of real-time monitoring and data analysis of the order scheduling process makes it difficult for businesses to accurately evaluate and optimize the effectiveness of order scheduling. Technical issues

[0004] This application provides an order scheduling method and apparatus for a physical order management system, which addresses the technical problems of low efficiency and insufficient accuracy in order scheduling in the prior art. Technical solutions

[0005] In view of the above problems, this application provides an order scheduling method and apparatus for a physical order management system.

[0006] A first aspect of this application provides an order scheduling method for a physical order management system, the method comprising:

[0007] A product database network is established, consisting of multiple physical product databases, each with a corresponding physical order management system. A physical product storage template is constructed, and physical products from the multiple product databases are stored in multiple physical order management systems based on this template, forming multiple physical product lists. A two-way order interaction channel is set up and embedded into the product database network to generate an order scheduling network. Physical order requests from multiple product databases are received, and the physical order requests and the multiple physical product lists are input into a scheduling decision model to obtain an order scheduling scheme. The physical order requests include multiple inbound requests and multiple outbound requests. The order scheduling network is configured according to the order scheduling scheme, and order scheduling is executed.

[0008] A second aspect of this application provides an order scheduling device for a physical order management system, the device comprising:

[0009] The system comprises the following modules: a product database network establishment module, which establishes a product database network composed of multiple physical product databases, each with a corresponding physical product order management system; a physical product storage template construction module, which constructs physical product storage templates and stores physical products from multiple physical product databases in multiple physical product order management systems based on these templates, forming multiple physical product lists; an order scheduling network generation module, which sets up a two-way order interaction channel and embeds it into the product database network to generate an order scheduling network; an order scheduling scheme acquisition module, which receives physical order requirements from multiple physical product databases, inputs the physical order requirements and the multiple physical product lists into a scheduling decision model to obtain an order scheduling scheme, wherein the physical order requirements include multiple inbound requirements and multiple outbound requirements; and an order scheduling execution module, which configures the order scheduling network according to the order scheduling scheme and executes the order scheduling. Beneficial effects

[0010] One or more technical solutions provided in this application have at least the following technical effects or advantages:

[0011] This application establishes a product database network, which is constructed from multiple physical product databases, each with a corresponding physical order management system. It constructs physical product storage templates and stores physical products from multiple product databases in multiple physical order management systems based on these templates, forming multiple physical product lists. It sets up a two-way order interaction channel, embedding it into the product database network to generate an order scheduling network. It receives physical order requests from multiple product databases, inputs these requests along with multiple physical product lists into a scheduling decision model to obtain an order scheduling plan. The physical order requests include multiple inbound and outbound requests. Based on the order scheduling plan, it configures the order scheduling network and executes the order scheduling. This invention addresses the technical problems of low efficiency and insufficient accuracy in order scheduling in existing technologies. By real-time monitoring of key parameters and states during the order scheduling process, combined with advanced algorithms and models, it achieves the technical effect of improving the efficiency and accuracy of order scheduling. Attached Figure Description

[0012] To more clearly illustrate the technical solutions in the embodiments of the present invention, the accompanying drawings used in the description of the embodiments will be briefly introduced below. Obviously, the accompanying drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0013] Figure 1 is a schematic flowchart of an order scheduling method for a physical order management system provided in an embodiment of this application;

[0014] Figure 2 is a schematic diagram of the structure of an order scheduling device for a physical order management system provided in an embodiment of this application.

[0015] Explanation of reference numerals in the attached diagram: 11. Product database network establishment module; 12. Physical product storage template construction module; 13. Order scheduling network generation module; 14. Order scheduling scheme acquisition module; 15. Order scheduling execution module. Best Mode for Carrying Out the Invention

[0016] Example 1 Modes for Carrying Out the Invention

[0017] This application provides an order scheduling method and apparatus for a physical order management system, addressing the technical problems of low efficiency and insufficient accuracy in order scheduling in the prior art. By real-time monitoring of key parameters and status during the order scheduling process, combined with advanced algorithms and models, the method achieves the technical effect of improving the efficiency and accuracy of order scheduling.

[0018] 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 a part of the embodiments of this application, and not all of them. All other embodiments obtained by those skilled in the art based on the embodiments of this application without creative effort are within the scope of protection of this application.

[0019] It should be noted that the terms "first," "second," etc., in the specification, claims, and accompanying drawings of this application are used to distinguish similar objects and are not necessarily used to describe a specific order or sequence. It should be understood that such data can be interchanged where appropriate so that the embodiments of this application described herein can be implemented in orders other than those illustrated or described herein. Furthermore, the terms "comprising" and "having," and any variations thereof, are intended to cover non-exclusive inclusion; for example, a process, method, system, product, or server that comprises a series of steps or units is not necessarily limited to those steps or units explicitly listed, but may include other steps or modules not explicitly listed or inherent to such processes, methods, products, or devices.

[0020] Example 1

[0021] As shown in Figure 1, this application provides an order scheduling method for a physical order management system, the method comprising:

[0022] Step S100: Establish a product warehouse network, which is constructed from multiple physical product warehouses, wherein each physical product warehouse has a corresponding physical order management system;

[0023] In this embodiment of the application, the commodity warehouse network aims to achieve centralized management, sharing, and rapid circulation of inventory information by integrating the resources of multiple physical commodity warehouses.

[0024] The existing physical goods inventory was reviewed and categorized. Based on factors such as product type, storage conditions, and sales performance, the inventory was divided into different types, such as finished goods inventory, raw material inventory, and semi-finished goods inventory. For each inventory, basic information such as inventory capacity and storage facilities was collected to facilitate subsequent integration and management.

[0025] Next, a corresponding physical order management system will be configured for each physical goods inventory. This system will have functions such as order processing, inventory management, and data analysis. It will be able to record information such as goods receipt, dispatch, and inventory changes in real time, and generate corresponding reports and data. Through the physical order management system, real-time monitoring and management of the inventory will improve the efficiency and accuracy of inventory management.

[0026] Then, by using information technology, such as by establishing a unified database or data platform, multiple physical order management systems can be connected to form a product warehouse network, enabling real-time sharing of inventory information, order information and other data between the various systems.

[0027] Step S200: Construct a physical goods storage template, and based on the physical goods storage template, store physical goods from multiple physical goods libraries in multiple physical order management systems to form multiple physical goods lists;

[0028] In this embodiment, the design principles of the physical goods storage template are first determined. The template includes basic attribute information of the goods, such as product number, name, specifications, unit, and price, to ensure the completeness and accuracy of the product information. Simultaneously, the template is designed according to the storage characteristics of the goods, such as shelf life and storage conditions, to meet the storage needs of different product warehouses.

[0029] Secondly, based on the physical goods storage template, information on the goods in each physical goods warehouse is entered and organized. Each goods warehouse has a separate inbound and outbound port, enabling independent management and control of the goods.

[0030] Next, the organized product information is imported into the corresponding physical order management system. By importing the product information, multiple physical product lists are automatically generated, each list corresponding to a product database.

[0031] To ensure real-time updates to the product list, an effective data synchronization mechanism is established. When changes occur in the product inventory, such as inbound or outbound shipments, the physical order management system captures these changes in real time and automatically updates the corresponding product list by setting up data interfaces and performing scheduled synchronization.

[0032] Furthermore, considering the independent operation of each product warehouse, the physical order management system should also support multi-user concurrent operation and data isolation. Administrators of different product warehouses log into the system using their respective accounts to perform independent order processing and inventory management operations without interfering with each other.

[0033] Finally, the establishment of physical goods storage templates and physical order management systems is not a one-time task, but rather a continuous process of optimization and upgrading as business develops and the market changes. Based on actual needs and data analysis results, the templates and systems are adjusted and improved to enhance the efficiency and accuracy of goods information management.

[0034] Step S300: Set up a two-way order interaction channel, embed the two-way order interaction channel into the product database network, and generate an order scheduling network;

[0035] In this embodiment, the function and role of the two-way order interaction channel are first determined. This channel transmits order information in real time, including status changes such as order generation, modification, and cancellation, and also receives feedback information such as inventory updates and shipment confirmations from various product databases. Through two-way communication, the order management system can comprehensively grasp the order execution status and make timely scheduling decisions.

[0036] Secondly, a technical implementation scheme for a two-way order interaction channel is designed. By selecting appropriate communication protocols and data formats, accurate transmission and parsing of information are ensured. Simultaneously, encryption technology and authentication methods are employed to ensure the reliability and security of communication, preventing information leakage and mistransmission.

[0037] Next, the two-way order interaction channel will be embedded into the product database network. The existing product database network will be modified and upgraded, adding necessary communication interfaces and middleware to achieve seamless integration of order information. Through this embedded channel, various product databases and order management systems can communicate in real time via the network, sharing order data and inventory information.

[0038] During the process of embedding the channel, it is also necessary to formulate corresponding communication protocols and specifications, and clarify the responsibilities and permissions of each party in the communication process. For example, the format and content of order information, transmission frequency, processing flow, etc., should be specified to ensure the accuracy and consistency of information.

[0039] After the channel embedding is completed, an order scheduling network is generated. This network connects all product warehouses and the order management system, forming an organic whole. Through this network, the inventory status and order status of each product warehouse can be monitored in real time, enabling global order scheduling and optimization.

[0040] Step S400: Receive physical order requirements from multiple physical goods warehouses, input the physical order requirements and the multiple physical goods lists into the scheduling decision model, and obtain an order scheduling scheme, wherein the physical order requirements include multiple inbound requirements and multiple outbound requirements;

[0041] In this embodiment, physical order requests from multiple physical goods warehouses are received in real time. These physical order requests include multiple inbound and outbound requests, reflecting the flow and allocation of goods between different warehouses. Inbound requests include the receipt of new goods and returned goods, while outbound requests include order fulfillment and internal transfers.

[0042] The system matches and associates order requirements with multiple lists of physical goods. These lists are updated in real time and contain key data such as inventory information, product type, and quantity for each product warehouse. By matching order requirements with product lists, the system determines the current inventory status of each product warehouse and the types and quantities of goods that need to be allocated.

[0043] Next, the matched data is input into the scheduling decision model. The scheduling decision model is an intelligent system based on algorithms and rules. Based on the input order demand and inventory information, it comprehensively considers various factors, such as inventory level, order priority, transportation cost, and time constraints, to generate the optimal order scheduling plan.

[0044] During the calculation process of the scheduling decision model, the inventory status of each commodity warehouse is analyzed, future demand changes are predicted, and the optimal commodity allocation strategy is determined. For example, it prioritizes certain orders based on their urgency and inventory adequacy; or it selects the most suitable commodity warehouse for shipment based on transportation costs and distance considerations.

[0045] Finally, the scheduling decision model outputs an order scheduling plan. This plan details how to allocate and schedule goods based on current order demand and inventory status to meet order needs and optimize overall operational efficiency. The order scheduling plan includes specific details such as which goods to transfer from which warehouse to which target location, the quantity transferred, and the timing.

[0046] Step S500: Configure the order scheduling network according to the order scheduling scheme and execute order scheduling.

[0047] In this embodiment, the received order scheduling plan is parsed and processed to determine information such as which goods to allocate from which product warehouse, the quantity allocated, the target location, and the estimated completion time. Based on this information, the order scheduling path and process that need to be configured are automatically calculated.

[0048] Next, based on the calculated scheduling path and process, the order scheduling network is configured. According to the requirements of the scheduling scheme, the connection relationships and communication protocols between each node are adjusted to ensure accurate and rapid information transmission.

[0049] During the configuration of the connectivity network, confirm the inventory status and the quantity of goods available for allocation; or coordinate with the transportation department to arrange suitable transportation tools and routes.

[0050] After configuring the order scheduling network, order scheduling begins. Based on the scheduling plan, transfer instructions are automatically sent to the relevant product warehouses, specifying the type, quantity, and destination of the goods to be transferred. Upon receiving the instructions, the product warehouses perform picking, packing, and shipping operations as required.

[0051] Furthermore, step S100 in the method provided in the application embodiment further includes:

[0052] Multiple physical goods databases are identified, multiple database nodes are generated, and a unique node identifier is set for each database node.

[0053] Obtain attribute information from multiple physical product databases, construct multiple product database attribute vectors based on the attribute information, and map and bind them to multiple database nodes;

[0054] Multiple product database nodes are clustered and grouped based on multiple product database attribute vectors to form multiple product database clusters, and cluster networks are built to obtain multiple internal networks of the clusters.

[0055] Construct a backbone network for multiple cluster internal networks, and obtain a product database network based on the multiple cluster internal networks and the backbone network.

[0056] In this embodiment, the physical goods warehouses involved in the goods warehouse network are first determined. These warehouses are distributed in different geographical locations and have different inventory levels and product types. For each physical goods warehouse, a corresponding warehouse node is generated in the goods warehouse network. The warehouse node is the basic unit in the goods warehouse network, representing the location and role of the physical goods warehouse in the network.

[0057] To ensure the uniqueness and identifiability of each library node in the network, a unique node identifier is assigned to each library node. This identifier can be numbers, letters, or a combination thereof, accurately distinguishing different library nodes within the network.

[0058] Next, the product inventory management system collects attribute information for each physical product inventory. This attribute information includes the product inventory's geographical location, inventory level, product type, sales data, etc.

[0059] After obtaining the attribute information, this information is converted into a mathematical representation, i.e., a product library attribute vector is constructed. An attribute vector is a multi-dimensional numerical representation, with each dimension corresponding to one attribute. For example, a product library attribute vector might include inventory level, sales revenue, geographic location rating, and operational efficiency rating.

[0060] During the construction of attribute vectors, data processing and transformation are performed, such as data cleaning, standardization, and normalization, to ensure the comparability between different attributes and the validity of the vectors. Finally, the constructed product library attribute vectors are mapped and bound to the corresponding library nodes through a database, data structure, or a dedicated mapping file, associating the identifier of each library node with its attribute vector.

[0061] Using appropriate clustering algorithms, such as K-means, hierarchical clustering, or density clustering, the database nodes are clustered based on multiple product database attribute vectors. Nodes with similar attributes are grouped together to form product database clusters. During the clustering process, clusters are formed based on the similarity of various dimensions in the attribute vectors, such as inventory level, sales revenue, geographical location, and operational efficiency.

[0062] After forming the product database cluster, a cluster network is built to establish connections between the database nodes within the cluster, ensuring efficient information transmission and sharing within the cluster. The cluster network is built using technologies such as local area networks (LANs) and dedicated networks to ensure communication security and stability within the cluster. After the cluster network is built, multiple internal networks are obtained. These internal networks are important components of the product database network, connecting the various database nodes within the cluster and enabling efficient information transmission and sharing.

[0063] To connect the various cluster networks into a complete product database network, a backbone network is constructed. Before building the backbone network, the design principles and objectives are first clarified. Design principles include reliability, scalability, and high performance, ensuring the backbone network can support fast and stable data transmission and communication between networks within the clusters. Next, the backbone network topology is selected, as it determines the connection method between nodes. Common topologies include star, ring, and mesh. Then, the transmission technologies and protocols for the backbone network are determined. Appropriate transmission technologies and protocols are selected based on actual needs. For example, high-speed fiber optic transmission technology is chosen to achieve long-distance, high-bandwidth data transmission; simultaneously, standard network protocols are used to ensure interoperability between different cluster networks. Next, the logical parameters of the backbone network are configured, including IP address allocation, routing settings, and network bandwidth management. By properly configuring these parameters, the efficient operation of the backbone network and accurate data transmission are ensured. After the backbone network is built, performance testing and optimization are performed. By simulating actual data transmission scenarios, network throughput, latency, and packet loss rate are tested, and corresponding optimizations are made based on the test results. Once the backbone network is built and optimized, the internal networks of each cluster will be connected to the backbone network to form a product database network.

[0064] The product database network covers all physical product databases and achieves connectivity and collaborative work between product databases through multiple layers of construction, including nodes, attribute vectors, clustering, internal cluster networks, and backbone networks.

[0065] Furthermore, the method also includes:

[0066] Obtain historical order scheduling data from multiple physical goods warehouses, and construct a product feature vector set, an order feature vector set, and a scheduling feature vector set based on the historical order scheduling data;

[0067] The product feature vector set, order feature vector set, and scheduling feature vector set are divided according to the internal networks of the multiple clusters, and the scheduling training within the cluster is carried out based on the division results to obtain multiple cluster decision layers.

[0068] The global decision layer is obtained by coordinating training based on the product feature vector set, order feature vector set, scheduling feature vector set, and multiple cluster decision layers.

[0069] By aggregating multiple cluster decision layers with the global decision layer, a scheduling decision model is obtained.

[0070] In this embodiment, historical order scheduling data is collected from multiple physical goods warehouses. This historical order scheduling data includes past order information, product inventory status, scheduling strategies, and transportation records.

[0071] Based on historical order scheduling data, product-related features are extracted, such as product type, sales volume, inventory level, and shelf life. These features are then converted into numerical forms to construct a product feature vector set. Historical order data is analyzed to extract order-related features, such as order volume, order type, order time, and customer location. These features are also converted into numerical vectors to form an order feature vector set. Various factors exist during the scheduling process, such as transportation costs, transportation time, and transportation methods. Features related to scheduling are extracted, and a scheduling feature vector set is constructed.

[0072] Subsequently, based on the layout and characteristics of the internal networks of multiple clusters, the product feature vector set, order feature vector set, and scheduling feature vector set are divided. This division should take into account factors such as the geographical location, resource capabilities, and business characteristics of the clusters to ensure that the feature vector sets can be reasonably allocated to each cluster for processing.

[0073] The partitioning process can be implemented through a data distribution algorithm to ensure that each cluster's internal network can obtain a subset of feature vectors that matches its characteristics and needs.

[0074] After the feature vector set is partitioned, scheduling training is performed within each cluster. This allows each cluster to learn and generate a corresponding scheduling decision model based on its own feature vector subset. Within each cluster, necessary preprocessing operations, such as data cleaning, standardization, and normalization, are performed on the assigned feature vector subset to eliminate noise and outliers in the data and improve training performance.

[0075] Based on the characteristics and needs of the cluster, select appropriate machine learning algorithms or deep learning models, such as decision trees, random forests, and neural networks, for training. Use a subset of processed feature vectors as training data to train the selected model. During training, optimize the model's performance by adjusting its parameters and structure, enabling it to accurately make scheduling decisions based on the input feature vectors. After training, evaluate the model using a validation set to observe its performance on unknown data. Based on the evaluation results, fine-tune the model, such as adjusting hyperparameters and improving the model structure, to enhance its accuracy and generalization ability.

[0076] Through in-cluster scheduling training, each cluster will obtain a scheduling decision model for its specific subset of feature vectors. These models constitute the cluster decision layer, which can make efficient scheduling decisions within the cluster based on the input feature vectors.

[0077] After obtaining multiple cluster decision layers, coordination training is performed to obtain a global decision layer. This ensures that the various cluster decision layers can work collaboratively to form a unified scheduling decision system. Coordination training uses optimization algorithms, such as genetic algorithms and particle swarm optimization, to adjust the parameters of each cluster decision layer in order to achieve the optimal scheduling effect globally. Through this process, the global decision layer is obtained.

[0078] Multiple cluster decision-making layers are integrated with the global decision-making layer through weight allocation. The weight allocation is determined based on cluster performance, historical data performance, or other relevant indicators.

[0079] Next, the outputs of multiple cluster decision layers and the global decision layer are integrated and formatted. After data integration and formatting, the decision layers are fused. Based on the selected fusion strategy, the outputs of multiple cluster decision layers are combined with the output of the global decision layer. After fusing multiple decision layers, the scheduling decision model is obtained.

[0080] Furthermore, the method further includes receiving physical order requests from multiple physical goods lists, inputting the physical order requests and the multiple physical goods lists into a scheduling decision model to obtain an order scheduling scheme.

[0081] Analyze physical order requirements to obtain standardized order feature vectors, and transform multiple lists of physical goods into standardized product feature vectors;

[0082] The standardized order feature vector and the standardized product feature vector are input into multiple cluster decision layers of the scheduling decision model to obtain cluster scheduling sub-schemes for each cluster's internal network.

[0083] The multiple cluster scheduling sub-schemes are input into the global decision layer for global coordination and optimization to generate an order scheduling scheme.

[0084] In this embodiment, the physical order requirements include the order type, order quantity, types of goods involved in the order, specifications and quantity of goods, delivery deadline, delivery address, and special requirements. Order types include sales orders, purchase orders, and return orders.

[0085] Based on the order details, features that influence scheduling decisions are extracted. These features include the total order amount, the total weight or volume of the goods in the order, the urgency of delivery, the geographical location of the delivery address, and restrictions on the mode of transport.

[0086] Since different orders may have different units and dimensions, these features are standardized by scaling or normalization to ensure that all features are on the same scale.

[0087] Features that influence scheduling decisions are extracted from the physical goods list. These features include inventory levels, transportation costs, weight, and volume. Similarly, since different goods have different units and dimensions of measurement for their features, standardization is performed. For example, inventory levels are normalized so that inventory levels for different goods can be compared on the same scale. The standardized goods features are then combined into a feature vector.

[0088] Standardized order feature vectors and product feature vectors are input into multiple cluster decision layers of the scheduling decision model. Each cluster decision layer processes and analyzes the input feature vectors based on its internal network structure and the trained model. Each cluster decision layer generates a cluster scheduling sub-scheme for its internal network based on the input feature vectors. These sub-schemes include product allocation, transportation route selection, and delivery time arrangement.

[0089] Multiple cluster scheduling sub-schemes are input into the global decision layer. The global decision layer performs detailed analysis and evaluation of each sub-scheme, determining its advantages, disadvantages, and applicable scope. Based on the analysis results, global coordination is performed to resolve conflicts and contradictions between different cluster sub-schemes, ensuring the consistency and coordination of the overall scheduling strategy. Optimization algorithms and models are used to comprehensively optimize multiple cluster scheduling sub-schemes, including the optimization calculation of objective functions such as cost minimization, efficiency maximization, and resource utilization improvement.

[0090] After global coordination and optimization calculations, an order scheduling plan is generated.

[0091] Furthermore, step S500 in the method provided in the application embodiment further includes:

[0092] Build a data mart, connect the multiple physical order management systems to the data mart, and receive multiple lists of physical goods;

[0093] In the data mart, multiple lists of physical goods are updated according to the order scheduling scheme to obtain multiple updated lists of physical goods. The multiple updated lists of physical goods are multiple lists of physical goods after the order scheduling is executed according to the order scheduling scheme.

[0094] Based on the multiple updated physical goods lists, an order scheduling network is set up to perform order scheduling.

[0095] In this embodiment, a data mart is a collection of data for a specific department or user group, used to support specific decision analysis processes. In an order scheduling scenario, the purpose of building a data mart is to centrally store, integrate, and manage data related to order scheduling. The steps for building a data mart include identifying requirements, data integration, data cleaning and transformation, data modeling, and data storage and management.

[0096] Specifically, the first step is to clarify which business scenarios and decision-making processes the data mart needs to support.

[0097] Next, data from different source systems, such as order management systems and inventory management systems, is collected and integrated. The data is then cleaned to remove duplicates, errors, or invalid data, and necessary transformations are performed to meet analytical needs. Next, a data mart model is designed, including fact tables and dimension tables. Finally, appropriate data storage technologies, such as relational databases and columnar storage, are selected, and a data management mechanism is established.

[0098] The physical order management system is the core system for processing physical orders, recording key data such as order status, product information, and customer information. Connecting these systems to a data mart enables real-time data synchronization and sharing. The connection process begins with developing a data interface to achieve bidirectional communication between the order management system and the data mart. Next, scheduled tasks or real-time triggers are set up to ensure that data from the order management system is synchronized to the data mart in real-time or periodically. Finally, the synchronized data is verified to ensure its accuracy and completeness.

[0099] Once the physical order management system successfully connects to the data mart, it begins receiving multiple lists of physical goods. These lists contain key data such as product inventory and location information.

[0100] Upon receiving the product list, the product data in the data mart is updated according to the generated order scheduling plan. Updates include, but are not limited to, adjustments to product quantity, updates to product location, and changes to product status.

[0101] After the update is complete, the product list in the data mart reflects the latest status after order scheduling.

[0102] The order scheduling network is a network structure used to perform order scheduling. It determines the optimal path for goods from the origin to the destination based on the updated product list and order scheduling scheme.

[0103] When setting up a connectivity network, a logistics network model is first established based on the location and relationships of logistics nodes such as warehouses and distribution centers. Based on factors such as order demand, product location, and transportation costs, the optimal path from the origin to the destination is planned. Then, resources such as transportation vehicles and personnel are allocated according to the path planning results. Finally, the network model, path planning, and resource allocation results are integrated to form the order scheduling connectivity network.

[0104] Once the order scheduling network is completed, order scheduling will begin.

[0105] Furthermore, the method further includes updating multiple lists of physical goods according to the order scheduling scheme to obtain multiple updated lists of physical goods, and the method also includes:

[0106] The order scheduling scheme is parsed, the product scheduling information in the order scheduling scheme is extracted, and a product transfer instruction set is generated. Each product transfer instruction includes the source product library, the destination library to which the product is transferred, the product number, the scheduling quantity, and the scheduling timestamp.

[0107] According to the aforementioned product transfer instruction set, multiple lists of physical products are processed, including:

[0108] Find the corresponding product number in the physical product list of the source product library, reduce the product inventory quantity, and add a transfer outbound mark to the product number. The transfer outbound mark includes the target product library, the scheduling quantity, and the scheduling timestamp.

[0109] Search for the corresponding product number in the physical product list of the target product library. If it does not exist, create a new product number, increase the product inventory quantity, and add a transfer-in mark. The added transfer-in mark includes the source product library, the scheduling quantity, and the scheduling timestamp.

[0110] After processing the multiple lists of physical goods, multiple updated lists of physical goods are obtained.

[0111] Furthermore, the device is also used to perform the following functions:

[0112] Extract the order scheduling execution result and add it to the end of the execution result chain;

[0113] Receive scheduling feedback from multiple physical goods warehouses, score the order scheduling execution results based on the scheduling feedback, obtain the scheduling execution score, add it to the execution result chain, and store it together with the order scheduling execution results;

[0114] Set up a result monitoring window, and monitor the execution results of the last three orders in the execution result chain based on the result monitoring window to obtain the window monitoring results;

[0115] When the window monitoring result shows that three consecutive scheduling execution scores are all lower than the execution score threshold, the scheduling decision model is optimized.

[0116] It should be noted that the order of the embodiments described above is merely for descriptive purposes and does not represent the superiority or inferiority of the embodiments. Furthermore, specific embodiments have been described above. Other embodiments are within the scope of the appended claims. In some cases, the actions or steps described in the claims can be performed in a different order than that shown in the embodiments and still achieve the desired result. Additionally, the processes depicted in the drawings do not necessarily require a specific or sequential order to achieve the desired result. In some embodiments, multitasking and parallel processing are also possible or may be advantageous.

[0117] The above description is only a preferred embodiment of this application and is not intended to limit this application. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of this application should be included within the protection scope of this application.

[0118] This specification and accompanying drawings are merely illustrative examples of this application and are intended to cover any and all modifications, variations, combinations, or equivalents within the scope of this application. Clearly, those skilled in the art can make various alterations and modifications to this application without departing from its scope. Therefore, if such modifications and modifications fall within the scope of this application and its equivalents, this application intends to include such modifications and modifications.

Claims

1. An order scheduling method for a physical order management system, characterized in that, The method includes: A product warehouse network is established, which is constructed from multiple physical product warehouses, each of which has a corresponding physical order management system; Construct a physical goods storage template, and based on the physical goods storage template, store physical goods from multiple physical goods databases in multiple physical goods order management systems to form multiple physical goods lists; Set up a two-way order interaction channel, embed the two-way order interaction channel into the product database network, and generate an order scheduling network; The system receives physical order requests from multiple physical goods warehouses, inputs the physical order requests and the multiple physical goods lists into a scheduling decision model, and obtains an order scheduling scheme. The physical order requests include multiple inbound requests and multiple outbound requests. Configure the order scheduling network according to the order scheduling scheme, and execute order scheduling.

2. The method according to claim 1, characterized in that, Establish a product inventory network, including: Multiple physical goods databases are identified, multiple database nodes are generated, and a unique node identifier is set for each database node. Obtain attribute information from multiple physical product databases, construct multiple product database attribute vectors based on the attribute information, and map and bind them to multiple database nodes; Multiple product database nodes are clustered and grouped based on multiple product database attribute vectors to form multiple product database clusters, and cluster networks are built to obtain multiple internal networks of the clusters. Construct a backbone network for multiple cluster internal networks, and obtain a product database network based on the multiple cluster internal networks and the backbone network.

3. The method according to claim 2, characterized in that, The method further includes: Obtain historical order scheduling data from multiple physical goods warehouses, and construct a product feature vector set, an order feature vector set, and a scheduling feature vector set based on the historical order scheduling data; The product feature vector set, order feature vector set, and scheduling feature vector set are divided according to the internal networks of the multiple clusters, and the scheduling training within the cluster is carried out based on the division results to obtain multiple cluster decision layers. The global decision layer is obtained by coordinating training based on the product feature vector set, order feature vector set, scheduling feature vector set, and multiple cluster decision layers. By aggregating multiple cluster decision layers with the global decision layer, a scheduling decision model is obtained.

4. The method according to claim 3, characterized in that, The system receives physical order requests from multiple physical goods databases, inputs these physical order requests along with the lists of physical goods into a scheduling decision model, and obtains an order scheduling plan, including: Analyze physical order requirements to obtain standardized order feature vectors, and transform multiple lists of physical goods into standardized product feature vectors; The standardized order feature vector and the standardized product feature vector are input into multiple cluster decision layers of the scheduling decision model to obtain cluster scheduling sub-schemes for each cluster's internal network. The multiple cluster scheduling sub-schemes are input into the global decision layer for global coordination and optimization to generate an order scheduling scheme.

5. The method according to claim 1, characterized in that, Executing order scheduling includes: Build a data mart, connect the multiple physical order management systems to the data mart, and receive multiple lists of physical goods; In the data mart, multiple lists of physical goods are updated according to the order scheduling scheme to obtain multiple updated lists of physical goods. The multiple updated lists of physical goods are multiple lists of physical goods after the order scheduling is executed according to the order scheduling scheme. Based on the multiple updated physical goods lists, an order scheduling network is set up to perform order scheduling.

6. The method according to claim 5, characterized in that, According to the order scheduling scheme, multiple lists of physical goods are updated to obtain multiple updated lists of physical goods, including: The order scheduling scheme is parsed, the product scheduling information in the order scheduling scheme is extracted, and a product transfer instruction set is generated. Each product transfer instruction includes the source product library, the destination library to which the product is transferred, the product number, the scheduling quantity, and the scheduling timestamp. According to the aforementioned product transfer instruction set, multiple lists of physical products are processed, including: Find the corresponding product number in the physical product list of the source product library, reduce the product inventory quantity, and add a transfer outbound mark to the product number. The transfer outbound mark includes the target product library, the scheduling quantity, and the scheduling timestamp. Search for the corresponding product number in the physical product list of the target product library. If it does not exist, create a new product number, increase the product inventory quantity, and add a transfer-in mark. The added transfer-in mark includes the source product library, the scheduling quantity, and the scheduling timestamp. After processing the multiple lists of physical goods, multiple updated lists of physical goods are obtained.

7. The method according to claim 1, characterized in that, The method further includes: Extract the order scheduling execution result and add it to the end of the execution result chain; Receive scheduling feedback from multiple physical goods warehouses, score the order scheduling execution results based on the scheduling feedback, obtain the scheduling execution score, add it to the execution result chain, and store it together with the order scheduling execution results; Set up a result monitoring window, and monitor the execution results of the last three orders in the execution result chain based on the result monitoring window to obtain the window monitoring results; When the window monitoring result shows that three consecutive scheduling execution scores are all lower than the execution score threshold, the scheduling decision model is optimized.

8. An order scheduling device for a physical order management system, characterized in that, The device includes: A product database network establishment module is used to establish a product database network, which is constructed from multiple physical product databases, each of which has a corresponding physical order management system. A physical goods storage template construction module is used to construct physical goods storage templates. Based on the physical goods storage templates, physical goods from multiple physical goods libraries are stored in multiple physical goods order management systems to form multiple physical goods lists. An order scheduling network generation module is provided, which sets up a two-way order interaction channel and embeds the two-way order interaction channel into the product database network to generate an order scheduling network. The order scheduling scheme acquisition module receives physical order requirements from multiple physical goods warehouses, inputs the physical order requirements and the multiple physical goods lists into the scheduling decision model, and obtains the order scheduling scheme. The physical order requirements include multiple inbound requirements and multiple outbound requirements. The order scheduling execution module configures the order scheduling network according to the order scheduling scheme and executes order scheduling.

Images