Intelligent logistics warehousing system management method and system based on big data

Abstract

The invention discloses an intelligent logistics warehousing system management method and system based on big data, and relates to the technical field of warehousing system management methods, and the method comprises the steps: determining a data storage system between warehousing data storage spaces according to the data load of each warehousing data storage space and a current input data packet; determining a to-be-completed storage event of the intelligent logistics storage system based on big data analysis of the current input data packet; and determining the core storage data according to the storage event to be completed, the storage data of the plurality of original storage nodes and the storage data of the newly added storage nodes, thereby improving the accuracy of the core storage data. Therefore, the multiple intelligent logistics storage plans are determined based on the storage work list, the data management system of each intelligent logistics storage plan is determined according to each intelligent logistics storage plan, the corresponding newly-added storage node and the current working state of the intelligent logistics storage system, and data management of each intelligent logistics storage plan is optimized.

Description

Technical Field

[0001] This invention relates to the technical field of warehouse system management methods, and more particularly to a smart logistics warehouse system management method and system based on big data. Background Technology

[0002] With the development of technology, smart logistics warehouses are gradually being applied to people's lives. Smart logistics warehouses introduce corresponding warehousing systems to control warehousing plans. In existing technologies, pending warehousing events are collected, and corresponding warehousing data is matched with the smart logistics warehousing system based on the pending warehousing events. However, the warehousing data is not core-level categorized and is mixed with various influencing data, resulting in low accuracy of core warehousing data and affecting the data management of various smart logistics warehousing plans. Summary of the Invention

[0003] The purpose of this invention is to overcome the shortcomings of the prior art. This invention provides a smart logistics warehousing system management method and system based on big data.

[0004] This invention provides a big data-based intelligent logistics warehousing system management method, comprising: determining multiple original warehousing nodes and newly added warehousing nodes according to the intelligent logistics warehousing system; determining data association paths based on the warehousing data of the multiple original warehousing nodes and the warehousing data of the newly added warehousing nodes; determining multiple warehousing data storage spaces based on the detection of the data association paths; determining the data storage system between the various warehousing data storage spaces based on the data load of each warehousing data storage space and the current input data packet; and determining the pending warehousing events of the intelligent logistics warehousing system based on big data analysis of the current input data packet within this data storage system. The core warehousing data is determined based on pending warehousing events, warehousing data from multiple existing warehousing nodes, and warehousing data from newly added warehousing nodes. Multiple warehousing data features are identified based on the core warehousing data. The data parsing space corresponding to the newly added warehousing nodes is determined based on these features and the data load of the newly added warehousing nodes, and a warehousing work list is output. Multiple smart logistics warehousing plans are determined based on the warehousing work list. The data management system for each smart logistics warehousing plan is determined based on each plan, the corresponding newly added warehousing nodes, and the current working status of the smart logistics warehousing system, and the data association branches of the core warehousing data are dynamically optimized.

[0005] This invention provides a management system for a big data-based intelligent logistics warehousing system. The management system is applied to the aforementioned big data-based intelligent logistics warehousing system management method. The management system includes: The data association path module is used to determine multiple original and new warehousing nodes based on the smart logistics warehousing system, and to determine the data association path based on the warehousing data of the multiple original and new warehousing nodes. The data storage system module is used to determine multiple warehouse data storage spaces based on the detection of data association paths, and to determine the data storage system between each warehouse data storage space according to the data load of each warehouse data storage space and the current input data packet. The core warehousing data module is used in this data storage system to determine the pending warehousing events of the smart logistics warehousing system based on big data analysis of the current input data packets; and to determine the core warehousing data based on the pending warehousing events, the warehousing data of multiple original warehousing nodes, and the warehousing data of newly added warehousing nodes. The data parsing space module is used to determine multiple warehouse data features based on the identification of core warehouse data, determine the data parsing space corresponding to the new warehouse node based on the multiple warehouse data features and the data load of the new warehouse node, and output the warehouse work list. The data management system module is used to determine multiple smart logistics warehousing plans based on the warehousing work list, determine the data management system for each smart logistics warehousing plan according to each smart logistics warehousing plan, the corresponding newly added warehousing nodes, and the current working status of the smart logistics warehousing system, and dynamically optimize the data association branches of core warehousing data.

[0006] Compared with the prior art, the beneficial effects of the present invention are: In this embodiment of the invention, the data storage system between various warehouse data storage spaces is determined based on the data load of each warehouse data storage space and the current input data packet using the method described in this embodiment. Within this data storage system, pending warehouse events of the smart logistics warehousing system are determined based on big data analysis of the current input data packet. Core warehouse data is determined based on the pending warehouse events, warehouse data from multiple original warehouse nodes, and warehouse data from newly added warehouse nodes. This data storage system incorporates a holistic consideration of pending warehouse events, warehouse data from multiple original warehouse nodes, and warehouse data from newly added warehouse nodes, thereby improving the accuracy of the core warehouse data.

[0007] Therefore, multiple smart logistics warehousing plans are determined based on the warehousing work list. A data management system for each smart logistics warehousing plan is established according to the plan, the corresponding new warehousing nodes, and the current operational status of the smart logistics warehousing system. The data association branches of core warehousing data are dynamically optimized. The introduction of the warehousing work list enables a holistic consideration of each smart logistics warehousing plan, the corresponding new warehousing nodes, and the current operational status of the smart logistics warehousing system. This optimizes the data management of each smart logistics warehousing plan, allowing for precise data control through the data association branches of core warehousing data. Attached Figure Description

[0008] Figure 1 This is a flowchart illustrating the intelligent logistics warehousing system management method based on big data in an embodiment of the present invention. Figure 2 This is a flowchart illustrating step S11 of the intelligent logistics warehousing system management method based on big data in an embodiment of the present invention. Figure 3 This is a flowchart illustrating step S12 in the big data-based smart logistics warehousing system management method of this invention. Figure 4 This is a flowchart illustrating step S13 in the big data-based smart logistics warehousing system management method of this invention. Figure 5 This is a flowchart illustrating step S14 of the intelligent logistics warehousing system management method based on big data in an embodiment of the present invention. Figure 6 This is a flowchart illustrating step S15 of the intelligent logistics warehousing system management method based on big data in an embodiment of the present invention. Figure 7 This is a schematic diagram of the structural composition of the management system of the big data-based smart logistics warehousing system in an embodiment of the present invention. Detailed Implementation

[0009] The technical solutions of the present invention will be clearly and completely described below with reference to the accompanying drawings in the embodiments of the present invention.

[0010] Please see Figures 1 to 7 A big data-based intelligent logistics warehousing system management method is applied to warehousing system management scenarios. The big data-based intelligent logistics warehousing system management method includes: Step S11: Based on the intelligent logistics warehousing system, identify multiple original warehousing nodes and new warehousing nodes, and determine the data association path based on the warehousing data of the multiple original warehousing nodes and the warehousing data of the new warehousing nodes; Step S12: Based on the detection of data association paths, determine multiple warehouse data storage spaces, and determine the data storage system between each warehouse data storage space according to the data load of each warehouse data storage space and the current input data packet; Step S13: In this data storage system, the pending warehousing events of the smart logistics warehousing system are determined based on big data analysis of the current input data packets; the core warehousing data is determined based on the pending warehousing events, the warehousing data of multiple original warehousing nodes, and the warehousing data of newly added warehousing nodes. Step S14: Based on the identification of core warehouse data, determine multiple warehouse data features, determine the data parsing space corresponding to the new warehouse node based on the multiple warehouse data features and the data load of the new warehouse node, and output the warehouse work list; Step S15: Based on the warehouse work list, determine multiple smart logistics warehousing plans. Based on each smart logistics warehousing plan, the corresponding new warehousing nodes, and the current working status of the smart logistics warehousing system, determine the data management system for each smart logistics warehousing plan, and dynamically optimize the data association branches of the core warehousing data.

[0011] refer to Figure 2 In step S11, the specific steps are as follows: S111: Real-time monitoring of the intelligent logistics warehousing system; determination of the data space of warehousing data based on the identification of the intelligent logistics warehousing system; determination of the warehousing data distribution map based on the traversal of the data space of the warehousing data; determination of multiple warehousing data areas based on the identification of the warehousing data distribution map. S112: In each storage data area, multiple storage nodes are determined according to the data storage style, corresponding data distribution weight and data update event of the storage data area. Multiple original storage nodes and new storage nodes are determined based on the identification of multiple storage nodes, and the storage data of each original storage node and the storage data of the new storage node are marked. S113: Based on the storage data of multiple original storage nodes and the storage data of new storage nodes, determine multiple data processing routes. Based on the multiple data processing routes, the distribution locations of multiple original storage nodes, and the distribution locations of new storage nodes, determine the data association path. At this time, the data association path is constructed with the new storage node as the center, and the data association between the new storage node and multiple original storage nodes is controlled.

[0012] In the embodiments of this application, the intelligent logistics warehousing system is monitored in real time, and the collected raw data is classified based on data type, source, timestamp, and business attributes; a multi-dimensional data model is constructed, including spatial dimensions (warehouse area, shelf, storage location), time dimensions (inbound, storage, outbound), and business dimensions (product category, value level); data dimensionality reduction techniques (such as PCA) are applied to process high-dimensional data and extract key features; and a data spatial index is established to improve the efficiency of subsequent traversal.

[0013] The data space is traversed using either Breadth-First Search (BFS) or Depth-First Search (DFS) algorithms; the density and distribution characteristics of each data point are calculated to identify data clustering areas; clustering algorithms (such as K-means and DBSCAN) are applied to group similar data; a visual data distribution map is generated, and heatmaps and other techniques are used to display data density; simultaneously, regions are divided based on data density, business relevance, and physical space constraints; graph segmentation algorithms are applied to optimize region boundaries, ensuring strong data correlation within regions and weak correlation between regions; a unique identifier and attribute label are assigned to each data region; and relationships between regions are established to form a hierarchical region structure.

[0014] Furthermore, within each warehouse data region, the data organization method, storage structure, and access patterns of each warehouse data region are evaluated in terms of data storage style. Storage styles include relational storage (structured data), document storage (semi-structured data), graph storage (data with complex relationships), and time-series storage (time-series data). The most suitable storage technology, such as MySQL, MongoDB, Neo4j, or InfluxDB, is selected based on data characteristics. The capacity and performance parameters of each storage node are determined based on data volume and growth forecasts.

[0015] Meanwhile, weights are calculated based on factors such as data access frequency, business importance, and data value; data is divided into high weight (core business data), medium weight (routine business data), and low weight (auxiliary data); high-weight data is allocated more computing and storage resources to ensure high-performance access; data sharding and replication strategies are designed according to the weight distribution to optimize the overall system performance.

[0016] Based on the identification of multiple storage nodes, multiple existing and new storage nodes are determined. The update frequency, update mode, and update scale of various types of data are identified. Update events include high-frequency updates (such as inventory status), medium-frequency updates (such as product information), and low-frequency updates (such as historical data). Corresponding processing mechanisms are designed for different update frequencies, such as real-time updates, batch updates, or scheduled updates. Triggering, processing, and feedback mechanisms for update events are established to ensure data consistency. At the same time, the technical characteristics, business characteristics, and operational characteristics of each storage node are analyzed. Machine learning algorithms (such as K-means clustering) are applied to automatically classify storage nodes. Existing existing storage nodes are identified based on system history records and configuration information. By comparing the current system architecture with the planned architecture, new storage nodes are identified.

[0017] Therefore, a deep analysis is conducted on the data of the original and newly added storage nodes to extract features such as data type, size, update frequency, and correlation. Data processing priorities, real-time requirements, and processing complexity are determined based on business needs. Different data processing workflows are designed, including ETL (Extract-Transform-Load), ELT (Extract-Load-Transform), and stream processing routes. Graph theory algorithms and optimization theory are applied to design the optimal data processing route, considering processing efficiency, resource consumption, and data integrity. Simultaneously, all storage nodes (original and newly added nodes) are precisely located in the physical and logical networks, and the network latency, physical distance, and data transmission cost between nodes are calculated. Data association paths are constructed based on shortest path algorithms, minimum spanning tree algorithms, or network flow algorithms, and the bandwidth, latency, reliability, and security indicators of each path are evaluated.

[0018] The data association path is constructed centered around the newly added warehouse node, and the data association between the newly added warehouse node and multiple original warehouse nodes is managed. The newly added warehouse node serves as the data association center, constructing a star or mesh topology. A data aggregation mechanism from original nodes to the newly added node is designed, including incremental synchronization, full synchronization, and real-time streaming synchronization. The performance of the newly added warehouse node is optimized, including memory expansion, CPU enhancement, and storage acceleration. A balanced distribution of data is achieved on the central node to avoid single-point overload. Simultaneously, establish association rules and constraints between data, including reference integrity, business rules, and data consistency requirements; monitor the data association status in real time to detect association anomalies and conflicts; dynamically adjust the data association strategy based on monitoring results to optimize association efficiency; and regularly verify the correctness and integrity of data associations to ensure data quality. refer to Figure 3 In step S12, the specific steps are as follows: S121: Collect data association paths, determine multiple sub-data storage areas based on the identification of data association paths, and determine multiple warehouse data storage spaces based on the regional location, corresponding data storage mode, and corresponding data storage function of each sub-data storage area; the multiple warehouse data storage spaces are prioritized based on different warehouse data types and corresponding warehouse data levels. S122: In each warehouse data storage space, the current data storage volume is determined based on the identification of the warehouse data storage space, and the data load of the warehouse data storage space is determined based on the current data storage volume, the space specification of the warehouse data storage space and the corresponding data receiving routes, so as to collect the data load of each warehouse data storage space. S123: Collect the current input data packet, determine multiple sub-data sets based on the identification of the current input data packet, and determine the data storage system between the various warehouse data storage spaces according to the multiple sub-data sets, the priority of each warehouse data storage space, and the data load of each warehouse data storage space.

[0019] In the embodiments of this application, data association paths are collected, and based on the topology and data flow of the data association paths, a clustering algorithm is used to identify logically independent sub-data storage areas; the boundaries of each sub-data storage area are determined according to data dependency and business coupling; a feature profile is established for each sub-data storage area, including data type, processing requirements, access mode, etc.; and the rationality of the sub-data storage area division is verified through business scenario testing.

[0020] Based on the physical location, data storage mode, and storage function of the sub-data storage areas, plan the warehouse data storage space; select appropriate storage technologies, such as relational databases, NoSQL databases, and data warehouses, according to data characteristics and business needs; determine the configuration parameters such as capacity, performance, and reliability of each storage space; and design logical or physical isolation mechanisms to ensure data security and performance isolation between different storage spaces.

[0021] Meanwhile, data is categorized into different types based on business functions, such as transaction data, inventory data, logistics data, user data, and analysis data; data is also categorized into different levels based on its importance, timeliness, and sensitivity, such as core level, important level, and general level; the priority of each storage space is evaluated by comprehensively considering the data type and level; and a dynamic priority adjustment mechanism is established to adjust the storage space priority according to business changes and system status.

[0022] Furthermore, warehouse data storage spaces are identified using unique identifiers (such as ID, name, or IP address) to ensure that each space can be explicitly referenced by the system; storage monitoring tools (such as Prometheus, Zabbix) or database built-in monitoring functions (such as MySQL's SHOWTABLESTATUS) are used to obtain the current data volume of each storage space; different types of data (such as order data, inventory data, and logistics data) are distinguished and statistically analyzed separately to ensure fine-grained monitoring; and scheduled tasks (such as every 5 minutes) or event-triggered mechanisms (such as updating immediately after data is written) are established to ensure that the data storage volume is accurate in real time.

[0023] Based on the current data storage volume, the space specifications of the warehouse data storage space, and the corresponding data receiving routes, determine the data load of the warehouse data storage space. The space specifications specify the capacity limit, performance indicators (such as IOPS, throughput), and expansion capabilities of each storage space. Calculate the number of data receiving routes for each storage space, the data traffic and frequency of each route. Based on the current data storage volume, space specifications, and the load of the data receiving routes, calculate the overall load (usually expressed as a percentage). Establish a load assessment model that considers the storage space utilization rate, performance pressure, and growth trend.

[0024] Aggregate the load data of each storage space to a central monitoring system (such as Grafana or ELKStack); display the load of each storage space in real time through dashboards, including storage utilization, performance load, and overall load; set load thresholds (e.g., 80% for warning, 90% for critical), and trigger an alert when the load exceeds the threshold; save historical load data for trend analysis and capacity planning.

[0025] Therefore, data packets from external systems are received through message queues (such as Kafka and RabbitMQ), API interfaces, and file streams (such as FTP and S3); the data packets are parsed to extract key information such as their source, type, timestamp, size, and structure; the data packets are initially classified according to dimensions such as business type, data structure, and source system; and millisecond-level data packet processing is achieved using streaming processing frameworks (such as Apache Flink and Spark Streaming).

[0026] Use parsers for formats such as JSON, XML, and Protobuf to extract specific fields from the data packet; split the data packet into multiple sub-data sets according to data type, business logic, and storage requirements; clean (remove duplicates and handle missing values) and standardize (unify field naming and format conversion) the sub-data sets; add metadata (such as source, timestamp, priority, and storage requirements) to each sub-data set.

[0027] The data storage system among various warehouse data storage spaces is determined based on multiple sub-data sets, the priorities of each warehouse data storage space, and the data load of each warehouse data storage space. The most suitable warehouse data storage space is matched according to the type, priority, and storage requirements of the sub-data sets. Combined with the load of each storage space collected in S122, data is avoided from being allocated to high-load spaces. Data flow paths between storage spaces are designed to ensure efficient storage and access. Data storage strategies are dynamically adjusted based on real-time load and priority changes. Simultaneously, the load and data inflow of each storage space are continuously monitored. When the load exceeds a threshold, storage resource expansion is automatically triggered (e.g., adding disks, database sharding). Data allocation is optimized through load balancing algorithms (e.g., round-robin, least connections). The adjustment process of the data storage system is recorded for subsequent analysis and optimization.

[0028] Specifically, a schematic diagram of the data collection storage space is shown in Table 1: Table 1. Schematic diagram of storage space

[0029] The system allocates sub-data sets: Order Master Data Set: Priority: High; Matching Storage Space: S1 (Order Data, 65% Load); Allocation Result: Stored in S1; Order Detail Data Set: Priority: High; Matching Storage Space: S1 (Order Data, 65% Load); Allocation Result: Stored in S1; User Related Data Set: Priority: High; Matching Storage Space: S1 (Order Data, 65% Load); However, S1 has a high load, so the system checks S2 (Inventory Data, 85% Load) and S3 (Logistics Data, 55% Load). S2 is under too much load and is unsuitable; S3 has a lower load, but the storage type is incompatible (logistics data); final decision: still store in S1, but trigger a load warning and notify the operations team to prepare for expansion; the system constructs the following data storage path: order data packet → sub-data set → S1 (order master data, order detail data, user associated data); © At the same time, the system sets the following rules: if the load of S1 exceeds 80%, automatically divert some data to S3 (temporary storage); migrate historical data in S1 to S4 (historical data storage space) every morning.

[0030] refer to Figure 4 In step S13, the specific steps are as follows: S131: Monitor the data storage system in real time, collect the current data storage events of the data storage system, construct the corresponding big data combination based on the current data storage events and the current input data packets, and determine the pending storage events of the smart logistics warehousing system based on the data analysis of the big data combination; S132: Based on the detection of pending warehousing events, identify multiple pending warehousing data items, and determine the first core data feature based on the multiple pending warehousing data items and the warehousing data of multiple original warehousing nodes; S133: Determine the second core data feature based on the warehouse data of multiple warehouse data projects to be completed and the warehouse data of newly added warehouse nodes. Determine the core data control mode based on the first core data feature, the second core data feature and the core warehouse data mapping relationship. Determine the corresponding core warehouse data based on the core data control mode and the current input data packet.

[0031] In the embodiments of this application, the data storage system is monitored in real time, and current data storage events are collected. Log events of the storage system are collected in real time through Logstash or Flume. Kafka is used as a message queue to cache and transmit event data. Event types include: data write events: recording the time, size, and target storage space of data writes; data read events: recording the frequency and response time of data reads; storage space expansion events: recording the time and new capacity of storage space expansion; and error events: recording detailed information on data write / read failures.

[0032] Construct corresponding big data combinations based on current data storage events and current input data packets; use ETL tools (such as Apache NiFi) to integrate multi-source data into a unified format; use a data lake (such as AWS S3) to store raw data and support flexible analysis; associate data storage events with current input data packets; extract key fields (such as timestamp, data type, storage space, and event type); and construct structured big data combinations for easy subsequent analysis.

[0033] Specifically, the data storage system of an e-commerce logistics center detected the following anomalies: S1 storage space utilization: 85% (exceeding the warning threshold of 80%); data write latency: 120ms (normal range <100ms); error log: 5 data write failures in the past 10 minutes; the logistics center collected the following data storage events: data write event: time: 2025-09-06 16:30:00; data size: 500MB; target storage space: S1; result: successful; error event: time: 2025-09-06 16:31:00; error type: write failure; error reason: insufficient storage space; affected data: order data packet #12345.

[0034] Machine learning algorithms (such as decision trees and random forests) are used to classify and predict big data combinations; rule engines (such as Drools) are used to define business rules and trigger corresponding events; event identification logic: if the storage space utilization rate is >80%, a storage expansion event is triggered; if the data write failure rate is >1%, a data repair event is triggered; if the input data packet contains promotional orders, an inventory warning event is triggered.

[0035] Specifically, the logistics center combines the following data into a large dataset, as illustrated in Table 2: Table 2. Schematic diagram of large datasets

[0036] The logistics center identified the following pending warehousing events through big data analysis: Storage Expansion Event: Triggering condition: S1 utilization rate 85% > 80%; Action: Expand S1 capacity from 10TB to 15TB; Data Repair Event: Triggering condition: Data write failure rate 2% > 1%; Action: Rewrite order data packet #12345; Inventory Warning Event: Triggering condition: Promotional order #12346 is expected to see a surge in sales; Action: Check the inventory of product P1001 and trigger the replenishment process.

[0037] Furthermore, based on the detection of pending warehousing events, multiple pending warehousing data items are identified. The first core data feature is determined based on the warehousing data of multiple pending warehousing data items and the warehousing data of multiple original warehousing nodes. This approach takes into account the overall consideration of multiple pending warehousing data items and the warehousing data of multiple original warehousing nodes, ensuring the accuracy of the first core data feature.

[0038] At this point, rule engines (such as Drools) or machine learning models (such as random forests and LSTMs) are used to classify and prioritize warehouse events. Event types include: insufficient inventory, order backlog, insufficient storage space, data anomalies, etc. Each warehouse event to be completed is broken down into multiple executable warehouse data items. Data items include project name, objective, required data, execution priority, etc. Priorities are determined based on factors such as business impact (such as customer satisfaction, cost impact) and urgency (such as SLA requirements).

[0039] Based on multiple pending warehousing data projects and warehousing data from multiple original warehousing nodes, the first core data feature was determined. Data from the original warehousing nodes was collected, and relevant data was extracted from the historical database, including: historical sales data (e.g., sales volume over the past 30 days); inventory turnover rate (e.g., turnover times of product P1001); order processing efficiency (e.g., average processing time for region R1). Statistical analysis (e.g., mean, variance, trend analysis) was used to extract data features; machine learning algorithms (e.g., PCA, feature importance analysis) were used to screen key features. The first core data feature was determined by extracting the most representative features from the original warehousing node data, based on the requirements of the pending data projects.

[0040] Specifically, an e-commerce logistics center detected the following pending warehousing events: Insufficient Inventory Event: Event Description: Product P1001's inventory is lower than the safety stock (current inventory 50 units, safety stock 100 units); Breakdown into data items: Item A1: Query historical sales data for product P1001; Item A2: Analyze the replenishment cycle for product P1001; Item A3: Calculate the recommended replenishment quantity for product P1001; Order Backlog Event: Event Description: Order processing delays in region R1 exceed 2 hours; Breakdown into data items: Item B1: Count the number of pending orders in region R1; Item B2: Analyze the order processing bottleneck in region R1; Item B3: Optimize the order allocation strategy in region R1.

[0041] Based on the original warehousing node data (such as historical data from the past 6 months), the logistics center determined the following first core data features: Data items for inventory shortage events: Feature F1: Average weekly sales volume of product P1001 (extracted from historical sales data); Calculation method: Average sales volume over the past 4 weeks = 200 units / week; Feature F2: Replenishment cycle of product P1001 (extracted from replenishment records); Calculation method: Average replenishment cycle = 7 days; Feature F3: Inventory turnover rate of product P1001 (extracted from inventory records); Calculation method: Turnover rate = monthly sales volume / average inventory = 4.2 times / month.

[0042] Data items related to order backlog events: Feature F4: Peak order processing time in region R1 (extracted from order processing records); Calculation method: Peak time is 14:00-16:00, accounting for 40% of daily order volume; Feature F5: Order processing bottleneck in region R1 (extracted from processing logs); Analysis results: Picking process accounts for 60% of processing time; Feature F6: Order allocation efficiency in region R1 (extracted from allocation records); Calculation method: Average allocation delay = 15 minutes.

[0043] Therefore, a second core data feature is determined based on multiple pending warehousing data projects and warehousing data from newly added warehousing nodes. A core data management mode is determined based on the first core data feature, the second core data feature, and the mapping relationship between core warehousing data. The corresponding core warehousing data is then determined based on this core data management mode and the current input data packet. This approach considers both the core data management mode and the current input data packet as a whole, ensuring the accuracy of the corresponding core warehousing data. At the same time, a data storage system is introduced, which considers the overall situation of pending warehousing events, warehousing data from multiple original warehousing nodes, and warehousing data from newly added warehousing nodes, thereby improving the accuracy of the core warehousing data.

[0044] At this point, based on the warehousing data of multiple pending warehousing data projects and newly added warehousing nodes, the second core data feature is determined. Data collection for newly added warehousing nodes involves: obtaining attribute information (such as storage capacity, processing power, geographical location, etc.) of the newly added warehousing nodes through the node registration interface or configuration file; real-time monitoring of the running status of the newly added nodes (such as CPU utilization, memory usage, network latency, etc.); establishing a mapping relationship between pending warehousing data projects and newly added warehousing nodes; analyzing the matching degree between project requirements and node capabilities; using statistical analysis (such as mean, variance, distribution) to extract node performance characteristics; and applying machine learning algorithms (such as cluster analysis) to identify node behavior patterns.

[0045] Specifically, an e-commerce logistics center has added two new warehousing nodes (N1 and N2), and there are pending warehousing data projects: Node information: Node N1: Storage capacity: 10TB; Processing capacity: 1000TPS; Geographic location: East China; Current load: 30%; Node N2: Storage capacity: 5TB; Processing capacity: 500TPS; Geographic location: South China; Current load: 20%; Pending data projects: Project B1: Replenishment plan for product P1001 (requires 2TB storage space, processing capacity 200TPS); Project B2: Order optimization for region R1 (requires 1TB storage space, processing capacity 100TPS).

[0046] Second core data feature extraction: Feature F7: Storage utilization rate of new nodes (Node N1: 30%, Node N2: 20%); Feature F8: Processing capacity margin of new nodes (Node N1: 800TPS, Node N2: 400TPS); Feature F9: Geographic matching degree between project and node (Project B1 and Node N1 are in the same region, matching degree 100%); Feature F10: Resource matching degree between project and node (Project B1's requirements and Node N1's capabilities match 95%).

[0047] Based on the mapping relationship between the first core data feature, the second core data feature, and the core warehousing data, the core data control mode is determined. The first core data feature (extracted from S132) and the second core data feature are standardized. Feature fusion is performed using feature weighting methods (such as TF-IDF) or deep learning models (such as Autoencoder). Mapping relationship construction: A mapping relationship table between features and control strategies is established. The optimal control mode is determined using a rule engine or decision tree algorithm. Control mode classification: Load balancing mode: Data is dynamically allocated according to node load; Priority mode: Resources are allocated based on business priorities; Geographic optimization mode: Data distribution is optimized considering geographical location.

[0048] First core data feature (S132): F1: Weekly sales of product P1001: 200 units; F2: Replenishment cycle: 7 days; F3: Inventory turnover rate: 4.2 times / month; Second core data feature: F7: Storage utilization rate of node N1: 30%; F8: Processing capacity margin of node N1: 800 TPS; F9: Geographic matching degree between project B1 and node N1: 100%.

[0049] Mapping Relationships: Rule 1: If geographical matching degree > 90% and storage utilization < 50%, then adopt the geographical optimization mode; Rule 2: If processing capacity margin > 200% of project requirements, then adopt the load balancing mode; Rule 3: If business priority is high and resources are sufficient, then adopt the priority mode; Determine the control mode: Project B1 (product replenishment): Matching rule 1 (geographical matching degree 100%, storage utilization 30%); Determine the control mode: geographical optimization mode; Project B2 (order optimization): Matching rule 2 (sufficient processing capacity margin); Determine the control mode: load balancing mode.

[0050] Based on the core data management and control model and the current input data packet, the corresponding core warehousing data is determined. The current input data packet is then structured and parsed to extract key fields (such as data type, size, priority, etc.). Management and control model application: Based on the determined management and control model, the corresponding data processing strategy is selected. Data is allocated using scheduling algorithms (such as round-robin, weighted round-robin, least connections). Core data determination: Based on the results of the management and control model, the data most important to the current business is selected. The core data is then marked and prioritized.

[0051] Specifically, the e-commerce logistics center receives new input data packets: Current input data packets: Data packet D1: Type: Product replenishment data; Size: 1.5GB; Priority: High; Source: Supplier system; Data packet D2: Type: Order processing data; Size: 0.8GB; Priority: Medium; Source: Order system; Control mode application: Data packet D1 (product replenishment): Applying geographic optimization mode; Assigned to node N1 (same region as supplier); Core data: Replenishment plan and inventory forecast for product P1001; Data packet D2 (order processing): Applying load balancing mode; Assigned to nodes N1 and N2 (at a 6:4 ratio); Core data: Order allocation scheme and processing priority for region R1; Core warehousing data determination: For node N1: Core data 1: Replenishment plan for product P1001 (Priority 1); Core data 2: 60% of order data for region R1 (Priority 2); For node N2: Core data 1: 40% of order data for region R1 (Priority 1).

[0052] refer to Figure 5 In step S14, the specific steps are as follows: S141: Collect core warehouse data, determine multiple data identification dimensions based on the detection of core warehouse data, and trigger multiple identification of core warehouse data according to multiple data identification dimensions. At this time, in the multiple identification of core warehouse data, perform gradient data classification on core warehouse data in sequence to output multiple warehouse data combinations of different gradient levels. Determine the corresponding warehouse data features based on the identification of each warehouse data combination to output multiple warehouse data features. S142: Collect the current working content of multiple original storage nodes and the current working content of new storage nodes. Based on the matching of the current working content of multiple original storage nodes and the current working content of new storage nodes, determine the autonomous control system of new storage nodes and trigger the data allocation work of new storage nodes. S143: In the data allocation work of the newly added warehouse node, the data load of the newly added warehouse node is collected. Based on the data load of the newly added warehouse node, the corresponding current work content and multiple warehouse data characteristics, a data parsing space corresponding to the newly added warehouse node is constructed to manage the mechanical data of the newly added warehouse node and output a warehouse work list. The warehouse work list records the warehouse events that are highly correlated with the newly added warehouse node.

[0053] In the embodiments of this application, core warehousing data is obtained through API interfaces, database queries, or message queues (such as Kafka); data sources include: order management systems, inventory management systems, logistics tracking systems, etc.; principal component analysis (PCA) or factor analysis is used to identify key dimensions; dimensions are determined by combining business rules and expert experience; common dimension types include: time dimension: divided by time granularity such as hour, day, week, month, etc.; spatial dimension: divided by spatial location such as warehouse, region, shelf, etc.; business dimension: divided by order type, product category, customer level, etc.; dimension weight setting: the weight of each dimension is determined by the analytic hierarchy process (AHP) or entropy weight method.

[0054] The data is sliced, drilled down, and rotated using multidimensional data cube (OLAP) technology; machine learning algorithms (such as K-means and decision trees) are used for data pattern recognition; the recognition process involves grouping the data by dimension; performing statistical analysis on each dimension group; and identifying correlations and abnormal patterns among the data.

[0055] Specifically, the core warehousing data of a certain e-commerce logistics center includes: Order data: Order number: O1001-O5000 (5000 records); Product information: SKU, name, category, price; Order status: Pending, Shipped, Completed; Timestamp: Order time, Shipping time, Completion time; Inventory data: Warehouse number: W001-W010 (10 warehouses); Product inventory: SKU, Current inventory, Safety stock; Inventory location: Warehouse area, Shelf, Storage location; Node data: Node number: N001-N005 (5 processing nodes); Node status: CPU utilization, memory usage, Task queue length; Processing capacity: Number of orders processed per second, Data storage capacity.

[0056] Based on the above core data, the following identification dimensions were determined: Time dimension (weight 30%): D1: Hour dimension (0-23 hours); D2: Day dimension (Monday to Sunday); D3: Month dimension (January to December); Spatial dimension (weight 25%): D4: Warehouse dimension (W001-W010); D5: Regional dimension (North China, East China, South China, West China); Business dimension (weight 45%): D6: Order type dimension (regular orders, rush orders, bulk orders); D7: Product category dimension (electronic products, clothing, food, home furnishings); D8: Customer level dimension (VIP, regular, new customers).

[0057] Gradient partitioning is performed using hierarchical clustering or fuzzy C-means clustering; gradient thresholds are set in conjunction with business rules; gradient levels are set as follows: high gradient: key business data, high-value data; medium gradient: routine business data, medium-value data; low gradient: historical data, low-value data; classification criteria: data importance, timeliness, value density, and other indicators; feature extraction methods include: statistical features: mean, variance, quantiles, etc.; time-series features: trend, periodicity, abrupt change points, etc.; business features: conversion rate, turnover rate, satisfaction rate, etc.; feature evaluation: information gain or chi-square test is used to evaluate feature importance; redundant and irrelevant features are removed.

[0058] Specifically, the core data is categorized into the following gradients: High-gradient data (20%): Real-time order data (within the current 2 hours); VIP customer orders; product data with inventory below the safety threshold; status data with node load exceeding 80%; Medium-gradient data (50%): Order data within the last 24 hours; regular customer orders; normal inventory product data; status data with node load between 50% and 80%; Low-gradient data (30%): Historical orders older than 24 hours; detailed information on completed orders; slow-moving product data; node normal operation status data.

[0059] Features were extracted from the data combinations at each gradient: High gradient data features: Feature F1: Average order processing response time (15 minutes); Feature F2: VIP customer order fulfillment rate (98.5%); Feature F3: Low inventory out-of-stock risk (high); Feature F4: High load node processing capacity saturation (85%); Medium gradient data features: Feature F5: Average processing time for ordinary orders (45 minutes); Feature F6: Inventory turnover rate (5.2 times / month); Feature F7: Order accuracy rate (99.2%); Feature F8: Average node utilization rate (65%); Low gradient data features: Feature F9: Historical order completion rate (97.8%); Feature F10: Percentage of slow-moving goods (8%); Feature F11: Data storage growth rate (2% / month); Feature F12: Percentage of system idle time (30%).

[0060] Furthermore, the work content of each node can be obtained through real-time monitoring interfaces, task scheduling systems, or workflow engines; collection frequency: real-time or near real-time (e.g., once every 5 seconds); work content type: tasks in progress: order processing, inventory updates, data synchronization, etc.; system status: CPU utilization, memory usage, network I / O, etc.; queue status: queue length of tasks to be processed, priority distribution, etc.

[0061] The task content is compared using similarity calculations (such as cosine similarity and Jaccard similarity); factors to consider include task type, system load, queue status, etc.; autonomous control system construction includes: control strategy formulation: determining the control mode of nodes based on matching results; control parameters include task allocation strategy, load threshold, priority rules, etc.; dynamic adjustment mechanism: real-time monitoring of matching degree changes; automatic adjustment of control strategy to adapt to changes in task content.

[0062] Specifically, the work content of a certain e-commerce logistics center was collected as follows: Original warehouse node work content: Node N001: Task in progress: Order picking (50 orders); System status: CPU 75%, memory 60%; Queue status: 120 orders to be processed, priority distribution (high: 30, medium: 60, low: 30); Node N002: Task in progress: Inventory counting (3 warehouse areas); System status: CPU 45%, memory 40%; Queue status: 15 tasks to be processed, priority distribution (high: 5, medium: 7, low: 3); Node N003: Task in progress: Data synchronization (with financial system); System status: CPU 85%, memory 70%; Queue status: 80 tasks to be processed, priority distribution (high: 20, medium: 40, low: 20).

[0063] New warehouse node work content: Node N004 (new): Tasks in progress: System initialization; System status: CPU 15%, Memory 25%; Queue status: 0 tasks pending; Based on the above work content, the system performs matching analysis: Similarity calculation: N004 and N001 similarity: 0.35 (low); N004 and N002 similarity: 0.65 (medium); N004 and N003 similarity: 0.25 (low); Control system determination: Optimal matching node: N002 (highest similarity); Control strategy: Task allocation mode: Adopt the "light load priority" strategy of N002; Load threshold: CPU ≤ 70%, Memory ≤ 65%; Priority rule: High priority tasks account for ≤ 30%; Task type: Mainly handles inventory management and low priority orders; Autonomous control parameters: Task receiving limit: 10 tasks per minute; Load balancing threshold: Stop receiving new tasks when CPU > 80%; Priority processing ratio: High:Medium:Low = 3:5:2.

[0064] Data allocation triggering mechanism: based on the data allocation strategy determined by the control system; triggering condition: node status meets control parameter requirements; data allocation algorithm: using greedy algorithm or genetic algorithm to optimize data allocation; factors to consider: data volume, processing complexity, node capabilities, etc.; allocation execution process: splitting the large dataset into small units suitable for processing; dynamically scheduling tasks according to node status; real-time monitoring of allocation execution.

[0065] Specifically, based on the established autonomous control system, the system triggers the data allocation process for N004: Allocation Trigger: Trigger Conditions: N004's CPU (15%) < 70%, Memory (25%) < 65%; Trigger Time: 2024-01-15 14:30:00; Data Allocation Content: Inventory Management Tasks: Inventory count in warehouse A004 (5000 SKUs); Inventory update in warehouse B003 (3000 records); Low-Priority Orders: Processing of ordinary goods orders (100 orders); Processing of return and exchange orders (20 orders).

[0066] Execution process: 14:30:05 - Task splitting completed, generating 8 subtasks; 14:30:10 - Subtasks assigned to N004; 14:30:15 - N004 begins processing the first subtask; 14:35:20 - All subtasks completed; 14:35:25 - Node status updated: CPU 45%, Memory 40%; Allocation results: Successfully processed inventory count task: 5000 SKUs; Successfully processed inventory update task: 3000 records; Successfully processed order task: 120 orders; Processing accuracy: 99.8%; Average processing time: 38 seconds per subtask.

[0067] Therefore, in the data allocation process for new storage nodes, the data load of the new storage nodes is collected. Based on the data load of the new storage nodes, the corresponding current work content, and multiple storage data characteristics, a data parsing space corresponding to the new storage nodes is constructed to manage the mechanical data of the new storage nodes and output a storage work list. The storage work list records storage events that are highly correlated with the new storage nodes.

[0068] At this point, during the data allocation process for newly added warehouse nodes, the data load of these nodes is collected. Node performance metrics are collected in real-time using system monitoring tools (such as Prometheus and Grafana). Collection frequency: once per second to ensure data real-time performance. Load metrics include: CPU utilization: the computational resource usage of the current processing task; memory utilization: the memory usage for data caching and processing; disk I / O: the load of data read / write operations; and network I / O: the bandwidth usage for data transmission. Load calculation: Overall load index = 0.4 × CPU utilization + 0.3 × memory utilization + 0.2 × disk I / O + 0.1 × network I / O.

[0069] A multi-dimensional parsing space is constructed based on load, work content, and data characteristics; the parsing space is represented using a vector space model (VSM); the dimensions of the parsing space are: load dimension: reflecting the processing capacity of the node; work dimension: reflecting the complexity of the current task; and feature dimension: reflecting the business attributes of the data; the parsing space is calculated as: parsing space vector = f(load vector, work vector, feature vector).

[0070] Specifically, the load data for the newly added warehouse node N004 in a certain e-commerce logistics center is as follows: Real-time load indicators: CPU utilization: 65%; Memory utilization: 70%; Disk I / O: 45%; Network I / O: 30%.

[0071] Overall load index = 0.4×65%+0.3×70%+0.2×45%+0.1×30%; ​​Overall load index = 26%+21%+9%+3%=59%; Load level determination: Low load: 0-30%; Medium load: 31-60%; High load: 61-100%; N004 current status: Medium load (59%).

[0072] Construct the parsing space based on N004: Load vector: CPU weight: 0.4; Memory weight: 0.3; Disk weight: 0.2; Network weight: 0.1.

[0073] Load vector = [0.65, 0.70, 0.45, 0.30]; Work vector: Inventory counting weight: 0.5; Order processing weight: 0.3; Data synchronization weight: 0.2; Work vector = [0.5, 0.3, 0.2]; Feature vector: Time feature weight: 0.4; Spatial feature weight: 0.3; Business feature weight: 0.3; Feature vector = [0.4, 0.3, 0.3]; Parsing space construction: Parsing space dimension: 8 dimensions (4 load + 3 work + 1 feature); Parsing space vector = [0.65, 0.70, 0.45, 0.30, 0.5, 0.3, 0.2, 0.33]; Parsing space capacity: 1000 data points; Parsing accuracy: 95%.

[0074] Dynamic allocation and management of data are achieved based on the parsing space; priority queues are used to manage pending tasks; work list generation: calculate the relevance score between tasks and nodes; filter high-relevance tasks (relevance > 80%); relevance calculation: relevance = 0.4 × load relevance + 0.3 × work relevance + 0.3 × feature relevance.

[0075] Specifically, N004's data management and work list generation includes: Task pool: Task T001: Urgent order processing (relevant requirements: high load, order processing, time-sensitive); Task T002: Routine inventory update (relevant requirements: medium load, data synchronization, routine business); Task T003: System backup (relevant requirements: low load, system maintenance, time-insensitive); Task T004: Data migration (relevant requirements: high load, data synchronization, spatially relevant); Task T005: Report generation (relevant requirements: medium load, business processing, feature analysis).

[0076] Correlation Calculation: T001 Correlation: Load Correlation: 0.85 (High Load Matching); Work Correlation: 0.90 (Order Processing Matching); Feature Correlation: 0.80 (Time-Sensitive Matching); Total Correlation = 0.4 × 0.85 + 0.3 × 0.90 + 0.3 × 0.80 = 0.85; T002 Correlation: Load Correlation: 0.60 (Medium Load Matching); Work Correlation: 0.70 (Data Synchronization Matching); Feature Correlation: 0.65 (Routine Business Matching); Total Correlation = 0.4 × 0.60 + 0.3 × 0.70 + 0.3 × 0.65 = 0.65.

[0077] T003 Correlation: Load Correlation: 0.30 (Low Load Mismatch); Operational Correlation: 0.40 (System Maintenance Partial Match); Feature Correlation: 0.50 (Time-Insensitive Match); Total Correlation = 0.4 × 0.30 + 0.3 × 0.40 + 0.3 × 0.50 = 0.39; T004 Correlation: Load Correlation: 0.80 (High Load Match); Operational Correlation: 0.85 (Data Synchronization Match); Feature Correlation: 0.75 (Spatial Correlation Match); Total Correlation = 0.4 × 0.80 + 0.3 × 0.85 + 0.3 × 0.75 = 0.80; T005 Correlation: Load Correlation: 0.65 (Medium Load Match); Operational Correlation: 0.60 (Business Processing Match); Feature relevance: 0.70 (feature analysis matching); Total correlation = 0.4 × 0.65 + 0.3 × 0.60 + 0.3 × 0.70 = 0.65.

[0078] Job List Generation: High-Relevance Tasks (>80%): T001: Urgent Order Processing (Relevance: 85%); T004: Data Migration (Relevance: 80%); Job List Details: Urgent Order Processing; Estimated Processing Time: 30 minutes; Required Resources: CPU 80%, Memory 75%; Priority: High; Deadline: 16:30; Data Migration: Estimated Processing Time: 45 minutes; Required Resources: CPU 70%, Memory 65%; Priority: High; Deadline: 17:00; Execution Monitoring: 14:40:00 - Start processing T001; 15:10:00 - T001 completed; 15:15:00 - Start processing T004; 16:00:00 - T004 completed; 16:05:00 - Update node status: CPU 50%, Memory 45%.

[0079] refer to Figure 6 In step S15, the specific steps are as follows: S151: Collect the warehouse work list, determine multiple warehouse plan items based on the parsing of the warehouse work list, determine the smart logistics warehouse plan based on the parsing of each warehouse plan item, collect multiple smart logistics warehouse plans, determine the corresponding warehouse storage routes based on the identification of multiple smart logistics warehouse plans, determine the storage conflict area based on the matching of multiple warehouse storage routes, and optimize each smart logistics warehouse plan based on the storage conflict area. S152: Collect each smart logistics warehousing plan, determine the first data management parameter based on each smart logistics warehousing plan and the corresponding newly added warehousing node, and determine the second data management parameter based on each smart logistics warehousing plan and the current working status of the smart logistics warehousing system; S153: Determine the data management system for each smart logistics warehousing plan based on the mapping relationship between the first data management parameter, the second data management parameter, and the data management system. At this time, the corresponding data management method is presented in the data management system, and the data association branch of the core warehousing data is dynamically optimized based on the data management method, the core warehousing data, and the new content of the newly added warehousing nodes.

[0080] In the embodiments of this application, a daily warehouse work list is automatically collected from a task scheduling system, WMS (Warehouse Management System), or MES (Manufacturing Execution System). The list includes: task number, task type (inbound / outbound / transfer), target storage location, time window, execution priority, required equipment, etc. Task change events are retrieved periodically via API interface or received in real time using a message queue (such as Kafka). Data cleaning tools (such as Python Pandas or ETL tools) are used to deduplicate, format, and structure the list.

[0081] Using a rule engine (such as Drools) or an NLP-based task semantic parsing model, each work record is transformed into an executable "warehouse planning project". Each project includes: task objectives, resource requirements, time constraints, execution path suggestions, etc. Project breakdown example: Original task: "Move goods A from the receiving area to area B at 10:00 on 2025-09-06"; Decomposed into: Picking project: Pick goods A from the receiving area; Transportation project: Use AGV carts to transport goods A to area B; Shelving project: Shelve goods A to the designated storage location in area B.

[0082] Multiple planned items are combined into a "warehouse plan" using time-series scheduling algorithms (such as genetic algorithms and dynamic programming). Each plan includes: execution order; resource allocation (such as AGVs, forklifts, operators); estimated start / end time; and dependencies (such as task B can only be executed after task A is completed). Example output: Plan number: P001; Items included: J001 (picking), J002 (transportation), J003 (shelving); Execution time: 10:00–10:30; Resource allocation: AGV_02.

[0083] Based on the starting point, ending point, and transit areas in each plan, a "storage route" is generated; the optimal path is calculated using a graph algorithm (such as Dijkstra's algorithm), taking into account factors such as: shortest distance; avoiding congested areas; and compliance with safety passage rules; Route representation: Route number: R001; Path: Inbound area → Main aisle → Area B aisle → Area B shelf; Estimated time: 8 minutes.

[0084] Multiple storage routes are overlaid and compared in the spatiotemporal dimension, and conflicts are identified using the "spatiotemporal grid method". Conflict types: temporal conflict: multiple tasks occupy the same area at the same time period; spatial conflict: there is a risk of collision between vehicles or personnel at the route intersection; conflict area marking: for example: the B area channel is occupied by P001 and P002 at the same time from 10:10 to 10:15 → marked as a conflict area.

[0085] Optimize each smart logistics warehousing plan based on storage conflict areas. Optimization strategies include: delaying or advancing the execution time of one of the plans; replanning routes for conflicting tasks to avoid conflict areas; changing execution equipment or personnel to reduce resource competition. Optimization goals include: minimizing task delays; maximizing resource utilization; and ensuring operational safety.

[0086] Furthermore, data is collected from various smart logistics warehousing plans. Based on each smart logistics warehousing plan and the corresponding newly added warehousing nodes, the first data management parameter is determined. Based on the current working status of each smart logistics warehousing plan and the smart logistics warehousing system, the second data management parameter is determined. This approach takes into account the overall working status of each smart logistics warehousing plan and the smart logistics warehousing system, ensuring the accuracy of the second data management parameter.

[0087] At this point, information is collected from multiple smart logistics warehousing plans that have been generated and optimized in S151; each plan includes: plan number, task type, execution time, resource allocation, target node, route planning, etc.; collection method: obtain plan data through system internal API or database query; use message queue (such as RabbitMQ or Kafka) to synchronize warehousing plan changes in real time.

[0088] The first data management parameters are determined based on the warehousing plan and newly added warehousing nodes. New nodes can be new warehouses, new shelves, new AGV docking points, etc. Node information includes: node ID, type, location, capacity, current load, etc. The first data management parameters describe the compatibility between the warehousing plan and the new nodes. Common parameters include: node compatibility (whether the node supports the plan); path adjustment cost (the impact of the new node on the original path); resource utilization (whether the new node improves resource utilization efficiency). Calculation method: A scoring model (such as weighted scoring) is used to match and score each plan with the new node. Formula example: ; in , , These are the weighting coefficients.

[0089] Specifically, an e-commerce logistics center is adding a smart shelf in Zone E as a new warehousing node, and needs to re-evaluate the following three warehousing plans. A schematic table of the new warehousing node is provided, as shown in Table 3:

[0090] Table 3. Schematic diagram of newly added warehousing nodes

[0091] Determine the first data management parameters: Add a new node: E-zone smart shelf (located near the main aisle, current load 20%); Adaptability calculation: P001: Original path does not pass through E-zone, low adaptability (score 60); P002: Original path passes near E-zone, medium adaptability (score 80); P003: Irrelevant to E-zone, low adaptability (score 50); First data management parameters: P001: {Adaptability: 60, Path adjustment cost: High}; P002: {Adaptability: 80, Path adjustment cost: Medium}; P003: {Adaptability: 50, Path adjustment cost: Low}.

[0092] The second data management parameters are determined based on the warehousing plan and the current working status. The current working status includes: AGV location and status (idle / working); shelf utilization; aisle congestion; and system load (CPU / memory / network). The second data management parameters describe the feasibility of executing the warehousing plan under the current system status. Common parameters include: execution priority (based on task urgency); system load capacity (whether the system can support the plan); and resource contention (whether multiple plans compete for the same resource). Calculation method: Each plan is evaluated using a status assessment model (such as a decision tree or neural network). Example rule: If the system load > 80%, the priority is reduced; if the number of idle AGVs < 2, it is marked as resource contention.

[0093] Specifically, the current system status is as follows: Number of idle AGVs: 1 (AGV03 is idle); System load: 70%; Main channel congestion: Moderate congestion; Feasibility assessment: P001: AGV01 is performing other tasks, resource contention is high (score 70); P002: AGV02 is idle, but the main channel is congested, execution feasibility is medium (score 75); P003: AGV03 is idle, no congestion, execution feasibility is high (score 90); Results: Second data management parameters: P001: {Execution priority: Medium, System load capacity: Medium, Resource contention: High}; P002: {Execution priority: High, System load capacity: Medium, Resource contention: Medium}; P003: {Execution priority: High, System load capacity: High, Resource contention: Low}; Optionally, the first parameter focuses on the compatibility of the plan with the new node; the second parameter focuses on the feasibility of the plan under the current system status.

[0094] Therefore, based on the mapping relationship between the first data management parameter, the second data management parameter, and the data management system, the data management system for each smart logistics warehousing plan is determined. At this point, the corresponding data management method is presented within this system. Based on this data management method, the core warehousing data, and the newly added content of the new warehousing nodes, the data association branches of the core warehousing data are dynamically optimized. This approach considers the overall mapping relationship between the first data management parameter, the second data management parameter, and the data management system, ensuring the accuracy of the data management system for each smart logistics warehousing plan. Simultaneously, a warehousing work list is introduced, realizing a holistic consideration of each smart logistics warehousing plan, the corresponding new warehousing nodes, and the current working status of the smart logistics warehousing system. This optimizes the data management of each smart logistics warehousing plan, enabling precise data control through the data association branches of the core warehousing data.

[0095] At this point, the data management system for each smart logistics warehousing plan is determined based on the mapping relationship between the first data management parameter, the second data management parameter, and the data management system. The mapping relationship of the data management system is as follows: The system predefines multiple data management system templates, each corresponding to a different management strategy (e.g., high-priority real-time management, medium-priority batch management, low-priority delayed management, etc.); Mapping rules: The first data management parameter (node ​​adaptability) determines the basic management type; the second data management parameter (system status feasibility) determines the management intensity and priority; Mapping logic: Matching is performed using a rule engine or machine learning model (e.g., decision tree, neural network); A schematic table illustrating the mapping relationship of the data management system is shown in Table 4. Table 4. Schematic diagram of mapping relationships in the data management system.

[0096] High-priority real-time management: Real-time data collection, analysis, and feedback; suitable for urgent tasks or critical nodes; Medium-priority batch management: Data collection and analysis in batches (e.g., every 10 minutes); suitable for routine tasks or medium-load nodes.

[0097] Low-priority delay management: Data processing with delays (e.g., once per hour); suitable for non-urgent tasks or low-load nodes; presentation: management mode is indicated by different colors or icons on the system interface; example: red: high-priority real-time management; yellow: medium-priority batch management; green: low-priority delay management.

[0098] Key data in the system, such as inventory information, order status, and resource allocation; New content: New data for new warehouse nodes, such as inventory information for new shelves and status data for new AGVs; Dynamic optimization: Adjusting data association branches according to data management methods: Real-time management: Establishing real-time data associations to ensure synchronized data updates; Batch management: Establishing batch data associations to synchronize data by batch; Delayed management: Establishing loose data associations to delay data synchronization; Optimization algorithms: Using graph algorithms (such as shortest path and minimum spanning tree) to optimize data association paths; Using machine learning models (such as reinforcement learning) to dynamically adjust association strategies.

[0099] Specifically, a smart logistics warehousing system is executing three warehousing plans (P001, P002, P003); P001: Add node N001 (high adaptability), system status is high feasibility; P002: Add node N002 (medium adaptability), system status is medium feasibility; P003: Add node N003 (low adaptability), system status is high feasibility; Core warehousing data: Inventory information: 100 units of product A, 200 units of product B; Order status: O001 incomplete, O002 completed; New additions: N001: Add shelf, inventory of 50 units of product C; N002: Add AGV, status is idle; N003: Add sorting area, status is ready for use.

[0100] Define the data management system: P001: High adaptability + High feasibility → High priority real-time management; P002: Medium adaptability + Medium feasibility → Medium priority batch management; P003: Low adaptability + High feasibility → Not executed, requires manual intervention; Present the data management method: P001: Red label (real-time management); P002: Yellow label (batch management); P003: Gray label (not executed).

[0101] Dynamically optimize data association branches: P001: Establish real-time association: The inventory of product C in N001 is synchronized to the core inventory data in real time; Optimization path: N001 → Core inventory system (shortest path); P002: Establish batch association: The AGV status of N002 is synchronized to the core resource data every 10 minutes; Optimization path: N002 → Batch processing module → Core resource system (minimum spanning tree); P003: Do not execute, wait for manual intervention; Result: The inventory of product C in P001 is updated in real time, and the core inventory data changes from 100+200=300 units to 100+200+50=350 units; The AGV status of P002 is updated every 10 minutes, and the AGV status in the core resource data is updated synchronously; P003 waits for manual processing and does not participate in data association optimization.

[0102] Please see Figure 7 , Figure 7 This is a schematic diagram of the structural composition of the management system of the big data-based smart logistics warehousing system in an embodiment of the present invention; the management system of the big data-based smart logistics warehousing system includes: The data association path module 21 is used to determine multiple original storage nodes and new storage nodes according to the smart logistics warehousing system, and to determine the data association path based on the storage data of multiple original storage nodes and the storage data of new storage nodes. The data storage system module 22 is used to determine multiple warehouse data storage spaces based on the detection of data association paths, and to determine the data storage system between each warehouse data storage space according to the data load of each warehouse data storage space and the current input data packet. The core warehousing data module 23 is used in this data storage system to determine the pending warehousing events of the smart logistics warehousing system based on big data analysis of the current input data packets; and to determine the core warehousing data based on the pending warehousing events, the warehousing data of multiple original warehousing nodes, and the warehousing data of newly added warehousing nodes. The data parsing space module 24 is used to determine multiple warehouse data features based on the identification of core warehouse data, determine the data parsing space corresponding to the new warehouse node based on the multiple warehouse data features and the data load of the new warehouse node, and output the warehouse work list. The data management system module 25 is used to determine multiple smart logistics warehousing plans based on the warehousing work list, determine the data management system of each smart logistics warehousing plan according to each smart logistics warehousing plan, the corresponding newly added warehousing nodes and the current working status of the smart logistics warehousing system, and dynamically optimize the data association branches of the core warehousing data.

[0103] The technical features of the above embodiments can be combined arbitrarily. For the sake of brevity, not all combinations of the technical features in the above embodiments are described. However, as long as there is no contradiction in the combination of these technical features, they should be considered to be within the scope of this specification.

Claims

1. A smart logistics warehousing system management method based on big data, characterized in that, include: Based on the intelligent logistics warehousing system, multiple original warehousing nodes and new warehousing nodes are identified, and data association paths are determined based on the warehousing data of the multiple original warehousing nodes and the warehousing data of the new warehousing nodes. Multiple warehouse data storage spaces are determined based on the detection of data association paths, and the data storage system between the warehouse data storage spaces is determined based on the data load of each warehouse data storage space and the current input data packet. In this data storage system, the pending warehousing events of the smart logistics warehousing system are determined based on big data analysis of the current input data packets; The core warehousing data is determined based on the pending warehousing events, warehousing data from multiple original warehousing nodes, and warehousing data from newly added warehousing nodes. Based on the identification of core warehousing data, multiple warehousing data features are determined. Based on these features and the data load of the new warehousing nodes, the data parsing space corresponding to the new warehousing nodes is determined, and a warehousing work list is output. Based on the warehouse work list, multiple smart logistics warehousing plans are determined. According to each smart logistics warehousing plan, the corresponding new warehousing nodes, and the current working status of the smart logistics warehousing system, the data management system of each smart logistics warehousing plan is determined, and the data association branches of the core warehousing data are dynamically optimized.

2. The intelligent logistics warehousing system management method based on big data according to claim 1, characterized in that, The process of identifying multiple original and new warehousing nodes based on the intelligent logistics warehousing system, and determining data association paths based on the warehousing data of the original and new warehousing nodes, includes: Real-time monitoring of the intelligent logistics warehousing system; determination of the data space of warehousing data based on the identification of the intelligent logistics warehousing system; determination of the warehousing data distribution map based on the traversal of the data space of warehousing data; determination of multiple warehousing data areas based on the identification of the warehousing data distribution map. In each warehouse data area, multiple storage nodes are determined based on the data storage style, corresponding data distribution weight, and data update events of the warehouse data area. Based on the identification of multiple storage nodes, multiple original warehouse nodes and new warehouse nodes are determined, and the warehouse data of each original warehouse node and the warehouse data of the new warehouse node are marked. Multiple data processing routes are determined based on the storage data of multiple original storage nodes and the storage data of newly added storage nodes. Data association paths are determined based on the multiple data processing routes, the distribution locations of multiple original storage nodes, and the distribution locations of newly added storage nodes. At this time, the data association paths are constructed with the newly added storage nodes as the center, and the data association between the newly added storage nodes and multiple original storage nodes is controlled.

3. The intelligent logistics warehousing system management method based on big data according to claim 1, characterized in that, The process of determining multiple warehouse data storage spaces based on data association path detection, and determining the data storage system between these warehouse data storage spaces based on the data load of each warehouse data storage space and the current input data packet, includes: Collect data association paths, identify multiple sub-data storage areas based on the identification of data association paths, and determine multiple warehouse data storage spaces based on the regional location, corresponding data storage mode, and corresponding data storage function of each sub-data storage area; the multiple warehouse data storage spaces are prioritized based on different warehouse data types and corresponding warehouse data levels; In each warehouse data storage space, the current data storage volume is determined based on the identification of the warehouse data storage space. Based on the current data storage volume, the space specifications of the warehouse data storage space and the corresponding data receiving routes, the data load of the warehouse data storage space is determined to collect the data load of each warehouse data storage space. Collect the current input data packet, determine multiple sub-data sets based on the identification of the current input data packet, and determine the data storage system between the various warehouse data storage spaces according to the multiple sub-data sets, the priority of each warehouse data storage space, and the data load of each warehouse data storage space.

4. The intelligent logistics warehousing system management method based on big data according to claim 1, characterized in that, In this data storage system, the pending warehousing events of the smart logistics warehousing system are determined based on big data analysis of the current input data packets. The core warehousing data is determined based on pending warehousing events, warehousing data from multiple existing warehousing nodes, and warehousing data from newly added warehousing nodes. The system monitors the data storage system in real time and collects the current data storage events. Based on the current data storage events and the current input data packets, it constructs corresponding big data combinations and determines the pending storage events of the smart logistics warehousing system based on the data analysis of these big data combinations.

5. The intelligent logistics warehousing system management method based on big data according to claim 4, characterized in that, In this data storage system, the pending warehousing events of the smart logistics warehousing system are determined based on big data analysis of the current input data packets. The core warehousing data is determined based on pending warehousing events, warehousing data from multiple existing warehousing nodes, and warehousing data from newly added warehousing nodes. This also includes: Based on the detection of pending warehousing events, multiple pending warehousing data items are identified, and the first core data feature is determined based on the multiple pending warehousing data items and the warehousing data of multiple original warehousing nodes. The second core data feature is determined based on the warehousing data of multiple pending warehousing data projects and newly added warehousing nodes. The core data management mode is determined based on the first core data feature, the second core data feature and the mapping relationship of core warehousing data. The corresponding core warehousing data is determined based on the core data management mode and the current input data packet.

6. The intelligent logistics warehousing system management method based on big data according to claim 1, characterized in that, The process involves identifying multiple warehouse data features based on core warehouse data, determining the data parsing space corresponding to the new warehouse node based on these features and the data load of the new warehouse node, and outputting a warehouse work list, including: Collect core warehouse data, determine multiple data identification dimensions based on the detection of core warehouse data, and trigger multiple identification of core warehouse data according to multiple data identification dimensions. At this time, in the multiple identification of core warehouse data, the core warehouse data is sequentially classified into gradient data to output multiple warehouse data combinations of different gradient levels. Based on the identification of each warehouse data combination, the corresponding warehouse data features are determined to output multiple warehouse data features.

7. The intelligent logistics warehousing system management method based on big data according to claim 6, characterized in that, The process of identifying multiple warehouse data features based on the core warehouse data, determining the data parsing space corresponding to the new warehouse node based on the multiple warehouse data features and the data load of the new warehouse node, and outputting a warehouse work list also includes: The system collects the current working content of multiple original storage nodes and the current working content of new storage nodes. Based on the matching of the current working content of multiple original storage nodes and the current working content of new storage nodes, it determines the autonomous control system of new storage nodes and triggers the data allocation work of new storage nodes. In the data allocation process for new warehouse nodes, the data load of the new warehouse nodes is collected. Based on the data load of the new warehouse nodes, the corresponding current work content, and multiple warehouse data characteristics, a data parsing space corresponding to the new warehouse nodes is constructed to manage the mechanical data of the new warehouse nodes and output a warehouse work list. The warehouse work list records warehouse events that are highly correlated with the new warehouse nodes.

8. The intelligent logistics warehousing system management method based on big data according to claim 1, characterized in that, The process involves determining multiple smart logistics warehousing plans based on a warehousing work list, establishing a data management system for each smart logistics warehousing plan based on the plan, corresponding new warehousing nodes, and the current operational status of the smart logistics warehousing system, and dynamically optimizing the data association branches of core warehousing data, including: The system collects a warehouse work list, identifies multiple warehouse planning items based on the parsing of the warehouse work list, and determines smart logistics warehouse plans based on the parsing of each warehouse planning item. This process collects multiple smart logistics warehouse plans. Simultaneously, based on the identification of multiple smart logistics warehouse plans, the corresponding warehouse storage routes are determined. Based on the matching of multiple warehouse storage routes, storage conflict areas are identified, and each smart logistics warehouse plan is optimized based on the storage conflict areas.

9. The intelligent logistics warehousing system management method based on big data according to claim 8, characterized in that, The process of determining multiple smart logistics warehousing plans based on a warehousing work list, establishing a data management system for each smart logistics warehousing plan based on each plan, corresponding new warehousing nodes, and the current working status of the smart logistics warehousing system, and dynamically optimizing the data association branches of core warehousing data, also includes: Collect various smart logistics warehousing plans, determine the first data management parameter based on each smart logistics warehousing plan and the corresponding newly added warehousing node, and determine the second data management parameter based on each smart logistics warehousing plan and the current working status of the smart logistics warehousing system; The data management system for each smart logistics warehousing plan is determined based on the mapping relationship between the first data management parameter, the second data management parameter, and the data management system. At this time, the corresponding data management method is presented in the data management system, and the data association branch of the core warehousing data is dynamically optimized based on the data management method, the core warehousing data, and the new content of the newly added warehousing nodes.

10. A management system for a smart logistics warehousing system based on big data, characterized in that, The management system of the big data-based smart logistics warehousing system is applied to the big data-based smart logistics warehousing system management method as described in any one of claims 1-9, wherein the management system of the big data-based smart logistics warehousing system includes: The data association path module is used to determine multiple original and new warehousing nodes based on the smart logistics warehousing system, and to determine the data association path based on the warehousing data of the multiple original and new warehousing nodes. The data storage system module is used to determine multiple warehouse data storage spaces based on the detection of data association paths, and to determine the data storage system between each warehouse data storage space according to the data load of each warehouse data storage space and the current input data packet. The core warehousing data module is used in this data storage system to determine the pending warehousing events of the smart logistics warehousing system based on big data analysis of the current input data packets; and to determine the core warehousing data based on the pending warehousing events, the warehousing data of multiple original warehousing nodes, and the warehousing data of newly added warehousing nodes. The data parsing space module is used to determine multiple warehouse data features based on the identification of core warehouse data, determine the data parsing space corresponding to the new warehouse node based on the multiple warehouse data features and the data load of the new warehouse node, and output the warehouse work list. The data management system module is used to determine multiple smart logistics warehousing plans based on the warehousing work list, determine the data management system for each smart logistics warehousing plan according to each smart logistics warehousing plan, the corresponding newly added warehousing nodes, and the current working status of the smart logistics warehousing system, and dynamically optimize the data association branches of core warehousing data.

Images