Logistics cargo batch loading data analysis method, device and equipment and storage medium

Abstract

The application discloses a logistics cargo batch loading data analysis method, which can effectively avoid subjective deviation and recognition lag problems caused by artificial judgment, improve the accuracy and response efficiency of batch loading time detection by constructing unified and standardized batch loading judgment logic and automatically triggering and real-time judging at key operation nodes of cargo circulation; by correlating multi-dimensional operation data and automatically identifying batch causes through data mining, the key influencing factors causing batch loading can be quickly located, the cumbersome process of artificial one-by-one checking and tracing is saved, and the efficiency and objectivity of problem analysis are greatly improved; based on the attribution analysis result, the routing planning algorithm and standard operation procedure are continuously optimized, a complete intelligent management closed loop from judgment, analysis to optimization execution is formed, the phenomenon of cargo batch loading is reduced from the source, the integrity and continuity of logistics loading are improved, and the overall cargo circulation efficiency and operation control quality are optimized.

Description

Technical Field

[0001] This invention relates to the field of logistics and express delivery services, and in particular to a method, apparatus, equipment, and storage medium for analyzing data on the batch loading and distribution of logistics goods. Background Technology

[0002] The express logistics industry is rapidly developing towards large-scale and networked operations. Simultaneously, customers' demands for the accuracy, reliability, and end-to-end stability of express services are constantly increasing. Various operational and service-related issues arising from batch loading are becoming increasingly prominent, and existing technologies are struggling to effectively address these problems. In traditional logistics operations, the determination of whether batch loading has occurred relies primarily on the experience of on-site operators or passive feedback from subsequent stages, lacking a set of objective, unified, and quantifiable judgment rules. This reliance on experience and passive feedback not only leads to a significant lag in the identification of batch loading issues but also results in chaotic statistical standards, failing to provide accurate and effective support for management decisions. More importantly, due to the lack of corresponding standardized control mechanisms, batch loading events are often easily concealed or overlooked, making it difficult to trace the specific occurrence and responsible parties, creating significant management blind spots and ambiguous areas of responsibility. This makes it difficult to fundamentally resolve batch loading-related issues, thus hindering the adaptation to the large-scale and networked development trend of the express logistics industry and failing to meet customers' ever-increasing demands for service quality.

[0003] It is evident that existing technologies still need improvement and enhancement. Summary of the Invention

[0004] In view of the shortcomings of the prior art, the purpose of this invention is to provide a data analysis method for batch loading of logistics goods, which aims to solve the technical problem that the lack of standardized control mechanism in the prior art makes it easy for batch loading events to be covered up or ignored, making it difficult to trace the specific occurrence and the relevant responsible persons, forming obvious management blind spots and ambiguous areas of responsibility, making it difficult to fundamentally solve the problems related to batch loading.

[0005] To achieve the above objectives, the present invention adopts the following technical solution: A method for analyzing logistics cargo batch loading data, comprising the following steps: The logic for determining the batch loading of logistics goods is established in advance based on business scenarios, time windows and physical attributes of goods; Real-time acquisition of cargo flow status data; triggering the judgment logic based on the cargo flow status data at key operation nodes to automatically determine and mark batch events. Acquire the tagged batch events and associate them with operation dimension data, cargo dimension data and operating environment dimension data. Identify the key influencing factors that cause batching through data mining methods and generate an attribution analysis report. Based on the attribution analysis report, optimize the routing planning algorithm and / or standard operating procedures to form a closed loop of intelligent management for logistics cargo loading.

[0006] The aforementioned data analysis method for batch loading of logistics goods includes, in part, a judgment logic for batch loading of logistics goods pre-established based on business scenarios, time windows, and physical attributes of the goods, comprising: For distribution to distribution or distribution to network business scenarios, when the main order is confirmed to be dispatched after the loading handover order at the responsible station, if the number of unloading scanned items received by the next station corresponding to the loading handover order is less than the number of items in the inventory of the main order at the responsible station within a preset time threshold, it is determined that a batch loading event has occurred. In the loading business scenario, when the loading handover order of the master order is confirmed for dispatch at the responsible station, if the number of scanned items in the current loading handover order is less than the number of items in the inventory of the master order at the current responsible station, and the inventory weight and / or volume of the master order meets the preset threshold conditions, then it is marked as a pre-judged batching state.

[0007] The aforementioned method for analyzing logistics cargo batch loading data includes the following steps: real-time acquisition of cargo flow status data; triggering the judgment logic based on the cargo flow status data at key operation nodes to automatically determine and mark batching events. Real-time acquisition of cargo flow status data, including waybill data, scan data, inventory data, and handover document data; Real-time monitoring of key operational events at each distribution center, including loading scan events, unloading scan events, and departure confirmation events; When a critical operation event is detected, the associated cargo flow status data is queried based on the master order number carried by the critical operation event. The data is then input into the rule engine to trigger the batching determination logic, automatically calculate and mark the batching event. The status of the batching event includes pre-batching, confirmed batching, or no batching.

[0008] The aforementioned method for analyzing logistics cargo batch loading data includes acquiring marked batching events and associating them with operational dimension data, cargo dimension data, and operational environment dimension data. Data mining techniques are used to identify key influencing factors leading to batching, and an attribution analysis report is generated. Specifically, this includes: Retrieve all batch events marked as confirmed batches from the database; The tagged batch events are integrated and associated with operation-dimensional data, cargo-dimensional data, and operational environment-dimensional data. The operation-dimensional data includes the responsible distribution center, operation team, handover order number, departure time, and unloading time. The cargo-dimensional data includes the master order number, a list of sub-order numbers, total cargo volume, total weight, number of sub-items, and the arrival timestamps and time differences for each sub-item. The operational environment-dimensional data includes vehicle load saturation at departure, the current throughput load of the distribution center, and the daily routing plan information. Statistical analysis, association rule mining, or machine learning algorithms are used to conduct in-depth analysis of the operational dimension data, cargo dimension data, and operational environment dimension data to identify high-frequency batching patterns and potential root causes. Based on the aforementioned high-frequency batching mode and the potential root cause generation visualization attribution analysis report, the batching rate trend, the ranking of responsible parties, the distribution of attribution categories, and the location of specific problem links are displayed.

[0009] The aforementioned method for analyzing logistics cargo batch loading data includes the following steps: Based on the attribution analysis report, optimizing the routing planning algorithm to form a closed loop for intelligent management of logistics cargo loading. The attribution analysis report is fed back to the route planning system; During the waybill generation stage, the physical attributes and destination information of the master waybill to be allocated and all its sub-items are obtained. Dynamic routing and batch optimization algorithms are called, with the allocation of all sub-items of the same shipment to the same handover batch as the core optimization objective. Multi-objective optimization calculations are performed in combination with transportation costs and timeliness requirements. Generate a transportation plan that includes batching suggestions and push the transportation plan to the distribution and scheduling system for execution.

[0010] The aforementioned method for analyzing logistics cargo batch loading data, wherein the dynamic routing and batch optimization algorithm is a graph theory-based path search algorithm, specifically including: The logistics network is abstracted as a graph structure, where nodes are distribution centers, edges are transportation routes, and edge weights include distance, timeliness, and cost. For each shipment, when searching for the path from the starting node to the target node in the graph, a batching constraint factor is introduced. If a certain path causes the sub-items to be separated, then a higher weight value is assigned to that path. Choose the path with the lowest total cost as the final routing scheme, and ensure that all sub-components follow the same path.

[0011] The aforementioned method for analyzing logistics cargo batch loading data includes the following steps: Based on the attribution analysis report, standard operating procedures are optimized to form a closed loop for intelligent management of logistics cargo loading. Based on the systemic operational problems identified in the attribution analysis report, we extracted the high-incidence patterns in batches that are related to specific cargo types and operational procedures. Based on preset rule templates, suggestions for revising standard operating procedures are automatically generated, including adding operation checkpoints, modifying operation processes, and adding review steps. The revision suggestions will be pushed to the operations management end. After review, the electronic work instructions will be automatically updated and the relevant operators will be notified.

[0012] An intelligent analysis and management device for batch loading of logistics goods, comprising: The judgment logic construction module is used to pre-establish the judgment logic for batch loading of logistics goods based on the business scenario type of distribution and transshipment, loading and dispatch, the time window of cargo flow and cargo attributes. The batch event determination module is used to acquire cargo flow status data in real time. At key operation nodes such as loading, unloading, and dispatch confirmation, the determination logic is triggered based on the cargo flow status data to automatically determine and mark batch events. The attribution analysis module is used to acquire the tagged batch events and associate them with operation dimension data, cargo dimension data and operating environment dimension data. It uses data mining methods to identify the key influencing factors that lead to batching and generates an attribution analysis report. The closed-loop optimization module is used to optimize the routing planning algorithm or revise the standard operating procedure based on the attribution analysis report, forming a closed-loop intelligent management system for logistics cargo loading that consists of "judgment-analysis-optimization-execution".

[0013] An intelligent analysis and management device for batch loading of logistics goods, comprising: At least one processor; and a memory communicatively connected to the at least one processor; wherein the memory stores instructions executable by the at least one processor to cause the electronic device to perform the logistics cargo batch loading data analysis method described above.

[0014] A computer-readable storage medium stores a computer program that, when executed by a processor, implements the aforementioned method for analyzing batch loading data of logistics goods.

[0015] Beneficial effects: This invention provides a data analysis method for batch loading of logistics goods. By combining business scenarios, time windows, and the physical attributes of goods to construct a unified and standardized batch loading judgment logic, and realizing automatic triggering and real-time judgment at key operational nodes in the flow of goods, it can effectively avoid the subjective bias and recognition lag caused by manual judgment, and improve the accuracy and response efficiency of batch loading time detection. By associating multi-dimensional operational data and using data mining methods to automatically identify the causes of batch loading, it can quickly locate the key influencing factors that lead to batch loading, saving the tedious process of manual investigation and tracing, and greatly improving the efficiency and objectivity of problem analysis. Based on the attribution analysis results, it continuously optimizes the routing planning algorithm and standard operating procedures, forming a complete intelligent management closed loop from judgment, analysis to optimization execution, reducing the phenomenon of batch loading of goods from the source, improving the integrity and continuity of logistics loading, and optimizing the overall efficiency of goods flow and the quality of operational control. Attached Figure Description

[0016] Figure 1 The first flowchart of the logistics cargo batch loading data analysis method provided by the present invention; Figure 2 The second flowchart of the logistics cargo batch loading data analysis method provided by the present invention; Figure 3 The third flowchart of the logistics cargo batch loading data analysis method provided by the present invention; Figure 4 The fourth flowchart of the logistics cargo batch loading data analysis method provided by the present invention; Figure 5 The fifth flowchart of the logistics cargo batch loading data analysis method provided by the present invention; Figure 6 The sixth flowchart of the logistics cargo batch loading data analysis method provided by the present invention; Figure 7 The seventh flowchart of the logistics cargo batch loading data analysis method provided by the present invention; Figure 8 The eighth flowchart of the logistics cargo batch loading data analysis method provided by the present invention; Figure 9 A schematic diagram of a structure for an intelligent analysis and management device for batch loading of logistics goods provided by the present invention; Figure 10 A schematic diagram of the structure of the intelligent analysis and management equipment for batch loading of logistics goods provided by the present invention. Detailed Implementation

[0017] This invention provides a method for analyzing data on the batch loading and distribution of logistics goods. To make the objectives, technical solutions, and effects of this invention clearer and more explicit, the invention will be further described in detail below with reference to the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are only for explaining the invention and are not intended to limit the invention.

[0018] In the description of this invention, it should be understood that the terms "upper," "lower," "left," and "right," etc., indicating orientation or positional relationships based on the orientation or positional relationships shown in the accompanying drawings, are only for the convenience of describing the invention and simplifying the description, and do not indicate or imply that the device or element referred to must have a specific orientation, or a specific orientational structure and operation. Therefore, they should not be construed as limitations on the invention. Furthermore, "first" and "second" are only for descriptive purposes and should not be construed as indicating or implying relative importance or implicitly specifying the number of indicated technical features. Therefore, a feature defined with "first" or "second" may explicitly or implicitly include one or more of that feature. In the description of this invention, unless otherwise stated, "multiple" means two or more.

[0019] Please see Figures 1 to 8 As shown, the present invention provides a method for analyzing logistics cargo batch loading data, which includes the following steps: Step 1: Establish a pre-defined logic for batch loading of logistics goods based on business scenarios, time windows, and the physical attributes of the goods. Relying on existing business standards of logistics companies, and considering differences in business scenarios such as trunk transportation and regional transshipment, as well as the time windows for goods circulation and physical attributes such as weight, volume, and number of pieces, build a standardized and quantifiable batch loading judgment logic. This forms a proprietary judgment rule system, avoiding ambiguity and providing a unified basis for subsequent self-judgments. Please refer to [link / reference needed]. Figure 2 As shown, it includes the following steps: Step 101: For distribution-to-distribution or distribution-to-network business scenarios, after the main order is confirmed to be dispatched at the responsible station's loading handover form, if the number of unloading scans corresponding to the loading handover form received by the next station within a preset time threshold is less than the number of inventory items of the main order at the responsible station, then a batch loading event is determined to have occurred. For regional transit, the confirmation of dispatched loading handover forms at the responsible station is clearly defined as the trigger node for determination. A preset time threshold adapted to the timeliness of cargo flow is set. The system monitors the unloading scan data of the next flow node in real time. When, within this time threshold, the number of unloading scans corresponding to the loading handover form of this responsible station received by the next station is less than the number of inventory items of the main order at the responsible station, after excluding non-batch situations such as normal distribution and separate loading of special goods, a batch loading event is directly determined to have occurred for this shipment.

[0020] For example, the master order of the responsible node includes sub-orders 1, 2, 3, 4, and 5. If the responsible node scans sub-orders 1, 2, 3, 4, and 5 using handover document A for loading, it indicates that there is no batch loading event for this responsible node. If the responsible node scans sub-orders 1, 2, 3, and 4 using handover document A for loading, it indicates that the responsible node's pre-judgment of whether to use batch loading is "yes". If the next transfer node scans sub-orders 1, 2, 3, 4, and 5 using handover document A for unloading, it indicates that there is ultimately no batch loading for this responsible node. If the next transfer node scans sub-orders 1, 2, 3, and 4 using handover document A for unloading, it indicates that there is batch loading for this responsible node. Each node in the cargo flow process can be predicted and confirmed through this judgment logic, which can accurately pinpoint the specific node where batch loading behavior occurs, providing data support for intervention and liability determination.

[0021] Step 102: For loading operations, when the loading handover document for the master order is confirmed for dispatch at the responsible station, if the number of scanned items on the current loading handover document is less than the number of items in stock for the master order at the current responsible station, and the inventory weight and / or volume of the master order meets the preset threshold conditions, then it is marked as a pre-judged batching state. For trunk line transportation, the dispatch confirmation time of the loading handover document at the responsible station is used as the judgment node. The loading scan data and inventory data are compared in real time. If the number of scanned items on the current loading handover document is less than the number of items in stock for the master order at the current responsible station, and the inventory weight and / or volume of the master order meets the preset threshold conditions, that is, if the remaining weight and volume of the goods do not exceed the vehicle's loading capacity and can be transported in the same vehicle, and the entire vehicle assembly has not been completed, then the goods are marked as a pre-judged batching state. Subsequently, manual review is required based on the scan data and handover documents of the next flow node to confirm whether it is ultimately determined to be confirmed as a batching.

[0022] Special scenario judgment conditions can also be added. For example, if a shipment of goods is split into sub-items and transferred to different areas by different vehicles, but all can be delivered to the same terminal station at the same time, manual review is required to exclude the split-batch judgment.

[0023] A standardized judgment rule system is established, categorizing the aforementioned judgment conditions, unified judgment logic, reasonable error standards, special scenario descriptions, and data source specifications into general rules, trunk line transportation-specific rules, regional transit-specific rules, and supplementary rules for special scenarios. Each rule is assigned a unique entry number, and the applicable scenarios, judgment standards, and triggering conditions for each rule are clearly defined. This standardized judgment rule library is integrated into the system's rule management module and stored in an encrypted format. A clear rule update process is also defined, allowing enterprise operations personnel to manually submit rule modification requests based on business adjustments. Updates are then jointly reviewed by the technical and operations departments. Automatic update triggering conditions can also be set; when the system detects that the accuracy rate of a certain scenario segmentation is below 95%, it automatically reminds operations personnel to review and optimize the rules, ensuring that the rule system adapts to business development needs. The rule library supports rapid retrieval by scenario and rule number, providing a clear and unified basis for subsequent system triggering of judgment logic.

[0024] Step 2: Acquire real-time cargo flow status data. At key operation nodes, trigger the judgment logic based on the cargo flow status data to automatically determine and mark batching events. Real-time collection of end-to-end logistics data is achieved through multi-interface linkage, locking key cargo flow operation nodes, automatically triggering preset judgment logic, completing the automated judgment and status marking of batching events, and realizing real-time monitoring and rapid identification of batching behavior. Please refer to [link / reference]. Figure 3 As shown, it includes the following steps: Step 201: Real-time acquisition of cargo flow status data, including waybill data, scan data, inventory data, and handover document data; real-time access to the enterprise's entire network cargo flow status data via Ethernet interface, wireless communication interface, and API interface, with data transmission synchronization frequency reaching the second level, ensuring data synchronization with actual cargo flow. Specifically, the cargo flow status data covers waybill data, scan data, inventory data, and handover document data: Waybill data includes core fields such as master order number, sub-order number, consignor and consignee information, cargo category, number of cargo pieces, weight and volume, transportation time, and route path; Scan data covers scanning information for all nodes such as cargo warehousing, sorting, loading, transshipment, and unloading, including scanning time, location, employee ID, cargo sub-order number, and scanning status; Inventory data is synchronized in real-time with dynamic data such as the number of inventory pieces waiting to be loaded, the number of loaded pieces, and the number of pieces waiting to be sorted in the warehouse at each responsible station; Handover document data includes electronic handover documents, scanned copies of paper documents, and verifiable information such as handover document number, signature information, and number of handedover pieces.

[0025] Furthermore, during the data access process, the system sets up a data verification mechanism to mark data with missing key fields, incorrect formats, or data anomalies and push this information to the operation and maintenance terminal, reminding staff to verify and correct it in a timely manner. Simultaneously, a data backup node is established to prevent data loss and ensure the stability and accuracy of data access. At the same time, the corresponding judgment logic in the judgment system is automatically triggered at key operational nodes in the goods flow. Specifically, the core key operational nodes are first identified, and based on the enterprise's operational processes, five core nodes are determined: the inbound scanning completion node, the sorting completion node, the loading confirmation node, the transit handover scanning node, and the unloading scanning node. Each node has preset trigger conditions. When the goods complete the corresponding node operation, the scanning terminal or system module immediately sends a trigger signal to the judgment rule management module. The signal includes the operation node type, the corresponding main / sub-order number of the goods, and the responsible node information. After receiving the trigger signal, the judgment rule management module automatically matches the judgment logic operation corresponding to the node, completing the preliminary judgment of the batch loading status. The entire triggering and initiation process takes several seconds, achieving real-time judgment.

[0026] Step 202: Monitor key operation events at each distribution center in real time. These key operation events include loading scan events, unloading scan events, and departure confirmation events. The system backend continuously monitors the core operation behaviors of all distribution centers and operating stations across the network in real time, focusing on key nodes in the flow of goods that are prone to batching problems. It identifies three types of key operations: loading scan events, unloading scan events, and departure confirmation events. When each type of event is triggered, the terminal device will immediately send a signal to the system judgment module to ensure that no operation behavior is missed and no judgment delay occurs.

[0027] Step 203: When a key operation event is detected, query the associated cargo flow status data based on the master order number carried by the key operation event, and input them into the rule engine to trigger the batching determination logic. Automatically calculate and mark the batching event. The status of the batching event includes pre-batching, confirmed batching, or no batching. When any key operation event such as loading scan, unloading scan, or departure confirmation is detected, the system automatically extracts the master order number information carried by the event. Based on the master order number, it accurately queries and matches the associated waybill data, scanning data, inventory data, and handover document data. The integrated full data is then synchronously input into the system's rule engine. The rule engine automatically matches the batching determination logic corresponding to the business scenario in step 1, completes quantitative calculation and determination, and automatically marks the batching event status. The batching event status is divided into three categories: pre-determined batching, confirmed batching, or no batching. The no-batching status corresponds to complete loading of goods without separation; the pre-determined batching status corresponds to suspected incomplete loading during the loading process, requiring verification at subsequent nodes; and the confirmed batching status corresponds to the situation where, after full-link data verification, it is determined that sub-items of goods have been separated and not transferred in the same batch. After the determination is completed, the system automatically records the determination rule basis, time node, operation geographical location, and corresponding handover documents to form a preliminary determination record.

[0028] Step 3: Obtain the tagged batch events and correlate them with operational, cargo, and operational environment data. Identify key influencing factors causing the batching using data mining methods and generate an attribution analysis report. Focusing on the confirmed batch events, integrate multi-dimensional data from operations, cargo, and the operational environment. Conduct in-depth analysis using various data mining algorithms to pinpoint the root causes of the batching and generate a visualized attribution analysis report, providing data support for subsequent optimization and rectification. Please refer to [link / reference needed]. Figure 4 As shown, it includes the following steps: Step 301: Extract all batching events marked as confirmed batching from the database; The system filters and extracts all batching events marked as confirmed batching after full-link verification from the backend database, removes invalid data that was predicted to be batched or not batched, and only performs attribution analysis on actual batching events that have occurred, to ensure that the analysis objects are accurate and the analysis results are consistent with the actual problems. The extracted event data is synchronously associated with the corresponding judgment records and flow data to facilitate subsequent data integration.

[0029] Step 302: Integrate the marked batch events, and associate them with operation dimension data, cargo dimension data, and operating environment dimension data. The operation dimension data includes the responsible distribution center, operation team, handover order number, departure time, and unloading time. The cargo dimension data includes the master order number, a list of sub-order numbers, total cargo volume, total weight, number of sub-items, and the arrival timestamps and time differences for each sub-item. The operating environment dimension data includes the vehicle load saturation at departure, the current throughput load of the distribution center, and the daily routing plan information. The system uniformly integrates the extracted confirmed batch events, categorizing them by single shipment and single event. The system automatically associates and matches three core data dimensions to complete, deduplicate, and validate the data, ensuring its integrity and validity. Operational data includes the responsible distribution center, operational team and personnel, designed transport batches, departure time, unloading time, and handover document number. Cargo data includes the master document number, a list of sub-document numbers, total cargo volume, total weight, number of sub-items, timestamps of each sub-item's arrival at the station, time differences in progress between sub-items, and cargo flow timeliness requirements. Operating environment data includes vehicle load saturation at departure, current throughput load of the distribution center, daily routing plan information, and real-time routing adjustment records. The integrated data forms a complete analytical dataset for each batch event, laying the foundation for in-depth data mining and analysis.

[0030] Step 303: Utilize statistical analysis, association rule mining, or machine learning algorithms to conduct in-depth analysis of the operational dimension data, cargo dimension data, and operational environment dimension data to identify high-frequency batching patterns and potential root causes. Employ a multi-algorithm fusion mining mode combining statistical analysis, association rule mining, and machine learning to perform hierarchical in-depth analysis of the integrated multi-dimensional data, ensuring accurate and reliable analysis results: First, through descriptive and comparative statistical analysis, calculate the batching rate and frequency for each responsible entity and scenario, and screen core indicators highly correlated with batching behavior; then, use the Apriori association rule mining algorithm to set minimum support and confidence levels suitable for enterprise operations, mine hidden relationships between data, and identify high-frequency batching behavior patterns; finally, use a decision tree machine learning model, with historical batching data as training samples, to train and optimize the model, and through feature importance ranking, clarify the influence weight of each factor on batching behavior, accurately locate the potential root causes of batching, covering various root causes such as human operational oversight, vehicle load saturation, lack of process specifications, and unreasonable route planning.

[0031] Specifically, the statistical analysis algorithm employs a combination of descriptive and comparative statistics as the foundational layer for multi-algorithm fusion. This core approach is used for initial screening of key data and calculation of core indicators, providing direction for subsequent in-depth analysis. First, the integrated multi-dimensional data is categorized and statistically analyzed. For the operational dimension, the batch loading rate, batch event frequency, and batch item percentage are calculated for responsible distribution centers, operational teams, and operators, clarifying who performed the operation, who made the error, and the frequency of errors. For the cargo dimension, the differences in batch loading rates across different cargo categories, order structures, and timeliness requirements are statistically analyzed. For the operational environment dimension, the batch loading rates under different vehicle load saturation levels and distribution center throughput loads are statistically analyzed. Secondly, comparative statistical analysis was conducted, setting normal loading events as the control group and batch loading events as the experimental group. The differences between the two groups in various dimensions were compared to initially identify the operation team, operators, vehicle loading saturation, and cargo category as potential key influencing factors. At the same time, the correlation coefficient between each factor and the batch loading rate was calculated, and core related indicators with a correlation coefficient ≥ 0.6 were screened out, while indicators with no obvious correlation were eliminated to reduce the redundancy of subsequent algorithm calculations.

[0032] Specifically, the association rule mining algorithm is as follows: The Apriori algorithm is selected as the intermediate layer analysis in the multi-algorithm fusion, and its core function is to mine hidden relationships between data of different dimensions and identify high-frequency behavioral patterns. First, the core data after statistical analysis and screening is discretized, dividing continuous data into discrete intervals and encoding categorical data to ensure that the data is suitable for the algorithm's computational requirements. Second, the core parameters of the algorithm are set. Based on the operational scale of the express delivery company, the minimum support is set to 3% (i.e., the proportion of a certain association combination occurring in the total number of events ≥ 3%) and the minimum confidence is set to 50% (i.e., the probability of batch loading occurring ≥ 50% when a certain precondition is met), to avoid mining association rules without practical business significance. Then, the algorithm is started to scan the multi-dimensional data of all batch loading events to mine frequent itemsets and association rules. Finally, the mined association rules are screened, redundant rules are removed, and high-frequency association rules with practical business significance are retained to form a list of high-frequency behavioral patterns, clarifying the association strength between different behavioral combinations and batch loading, and providing direct evidence for root cause analysis.

[0033] Specifically, the machine learning model is as follows: a decision tree model is selected as the deep analysis of multi-algorithm fusion. The core is used to accurately identify the weight of key influencing factors, locate potential root causes, and achieve preliminary prediction of batch risk. First, a training dataset and a test dataset are constructed. The historical data of the enterprise in the past 6 months are selected, of which 70% is used as the training set (including batch loading events and normal loading events) and 30% is used as the test set. The training set data covers the core indicators screened by statistical analysis and the high-frequency association combination mined by association rules. The labels are set as "batch (1)" and "normal (0)". Second, a decision tree model is constructed. The core indicators of operation dimension, cargo dimension and operating environment dimension are used as input features, and "whether to batch" is used as the output label. The model training is started. The model accuracy is optimized by iteratively adjusting the depth of the decision tree and the number of leaf nodes to ensure that the accuracy of the model test set is ≥90% to meet the actual analysis needs. After training, the weight of each influencing factor is clarified by ranking the importance of the model features, and the key influencing factors are accurately locked. At the same time, the potential root causes are located by the node splitting logic of the decision tree. For high-frequency association combinations, the model further locates the specific root causes.

[0034] Furthermore, the multi-algorithm fusion and collaboration logic specifically involves three algorithms working in a layered and collaborative manner, with cross-validation to ensure the accuracy and reliability of the analysis results. First, statistical analysis algorithms are used to filter and initially locate data, reducing computational load and providing core analysis data for the subsequent two algorithms. Second, association rule mining algorithms are used to mine hidden associations based on statistical data, identifying high-frequency behavioral patterns and compensating for the inability of statistical analysis to discover indirect associations. Finally, machine learning models are used to deeply validate and supplement the results of the first two algorithms, clarifying the weight of influencing factors and locating potential root causes. Simultaneously, the discovered root causes are cross-validated to eliminate contradictory or unsupported analysis results, ultimately integrating them into three major analysis results: "high-frequency behavioral patterns, key influencing factors, and potential root causes." For example, the high-frequency behavioral pattern is "when loading shifts handle ordinary / loose parts in high-saturation loading scenarios, operators are prone to batch operations," the key influencing factors are "operator proficiency, vehicle loading saturation, operation team management, and cargo category," and the potential root causes are "inadequate operator training, lack of specific operating procedures during high-saturation loading, and missing team review processes."

[0035] Step 304: Based on the high-frequency batching pattern and potential root causes, generate a visualized attribution analysis report, displaying the batching rate trend, the ranking of responsible entities, the distribution of attribution categories, and pinpointing specific problem stages. Based on the mined high-frequency batching pattern and potential root causes, the system automatically generates a visualized attribution analysis report. The report data is taken from actual analysis results and judgment files, ensuring traceability. It corely displays four main components: a line graph showing the overall trend of batching rate changes over time and scenario, intuitively reflecting the effectiveness of batching control; a hierarchical bar chart showing the ranking of batching indicators for the three levels of responsible distribution centers, operation teams, and operators, clearly identifying high-risk responsible entities; pie charts and donut charts showing the distribution and percentage of the four core attribution categories—human operational factors, operating environment factors, cargo characteristic factors, and process standardization factors—and their sub-items; and a cargo flow flowchart to accurately pinpoint the specific operational stages and problem nodes where batching occurs. The report also includes a summary of core problems and targeted optimization directions, supporting online viewing, export printing, and tiered access control, facilitating quick understanding of the core issues by operational personnel at all levels.

[0036] Step 4: Based on the attribution analysis report, optimize the routing planning algorithm and / or standard operating procedures to form a closed loop of intelligent management for logistics cargo loading. Based on the conclusions of the attribution analysis report, conduct targeted optimization of the routing planning algorithm or revision of the standard operating procedures to avoid batch loading problems from the source, achieving a closed loop of intelligent management throughout the entire process of "judgment-analysis-optimization-execution," and continuously improving logistics loading efficiency and control quality.

[0037] Please refer to Figure 5 As shown, based on the attribution analysis report, the route planning algorithm is optimized to form a closed loop for intelligent management of logistics cargo loading, including the following steps: Step 411: Feed back the attribution analysis report to the routing planning system; feed back the full amount of the visualized attribution analysis report to the enterprise routing planning system to locate the root causes of routing issues such as unreasonable routing planning, missing batching logic, and path conflicts that lead to batching, clarify the core direction of routing optimization, and provide accurate basis for subsequent algorithm adjustments and scheme optimization.

[0038] Step 412: In the waybill generation stage, obtain the physical attributes and destination information of the master order to be allocated and all its sub-items, call the dynamic routing and batching optimization algorithm, and take the allocation of all sub-items of the same shipment to the same handover batch as the core optimization objective, and perform multi-objective optimization calculations in combination with transportation cost and timeliness requirements; In the initial stage of waybill generation and distribution plan formulation, the system automatically obtains the physical attributes such as weight and volume of the master order to be allocated and all its sub-items, as well as the receiving and shipping destination information, call the dynamic routing and batching optimization algorithm, and take the allocation of all sub-items of the same shipment to the same handover batch as the core optimization objective, while taking into account transportation cost control and timeliness requirements, and carry out multi-objective constraint optimization calculations.

[0039] Please refer to Figure 6 As shown, the dynamic routing and batch optimization algorithm is a graph theory-based path search algorithm, and its specific implementation logic is as follows: Step 4121: Abstract the logistics network into a graph structure, where nodes are distribution centers and edges are transportation routes. Edge weights include distance, timeliness, and cost. Based on the company's actual logistics transportation system, abstract all distribution centers, regional operation stations, and terminal outlets into independent nodes of a directed graph, assigning each node a unique code to accurately correspond to its actual physical location. Define the trunk transportation routes and regional transit routes between nodes as directed edges of the graph. Based on actual logistics transportation management needs, set multi-dimensional composite weights for each edge. The weight parameters specifically include three core indicators: transportation distance, delivery timeliness, and unit logistics cost. Simultaneously, route stability and transit loading / unloading difficulty are included as auxiliary weight parameters in the calculation. All weight parameters are pre-calibrated based on the company's historical operational data, constructing a logistics network topology graph that fits the actual business, providing a basic model support for subsequent path search and batch processing.

[0040] Step 4122: For each shipment, when searching for the path from the starting node to the target node in the graph, a batching constraint factor is introduced. If a path causes the separation of sub-items, a higher weight value is assigned to that path. For each main shipment containing multiple sub-items, when starting the path search, all sub-items under the same main shipment are marked as a mandatory batching group, and the highest batching priority is set. The batching constraint factor is introduced simultaneously as the core judgment condition for path selection. During the process of traversing the logistics network topology and searching for alternative paths from the starting node to the target node, it is verified in real time whether each alternative path can achieve the same path and batch transfer of all sub-items within the batching group. If a certain alternative path has sub-item splitting, cross-line transfer, or different batch handover, it is determined that the path violates the batching constraint. At this time, a preset multiple weight is applied to the composite edge weight of the alternative path, which significantly increases the overall transportation cost of the path, and the path is marked as a non-compliant path. If the alternative path can satisfy the same path flow of all sub-items and has no risk of splitting or separation, a higher weight value is not assigned, and the original composite weight value is retained to ensure the selection priority of compliant paths.

[0041] Step 4123: Select the path with the lowest total cost as the final routing scheme, ensuring consistency across all sub-items. After calculating the weights and verifying the batching of all candidate paths, the shortest path search logic is used to sort the weighted total costs of each path, eliminating non-compliant paths with penalty weights, and selecting the compliant path with the lowest total cost as the optimal routing scheme for the shipment. This optimal path strictly ensures that all sub-items under the same master order share a unique transportation path, synchronously transfer and hand over, and are included in the same loading and handover batch, completely avoiding the batch loading problem caused by sub-item separation from the source of routing planning. At the same time, the system performs a secondary verification of the optimal path to check whether the path timeliness and transportation cost meet the preset control requirements. If there is room for path optimization in special scenarios, the weight parameters can be fine-tuned based on real-time operating environment data, and the final routing scheme can be locked after recalculation, ensuring that the scheme combines batching compliance, transportation economy, and timeliness compliance, adapting to the needs of various logistics operation scenarios.

[0042] Step 413: Generate a transportation plan including batch consolidation suggestions and push the plan to the distribution and scheduling system for execution. After the algorithm completes its calculations, a complete transportation plan is generated, including batch consolidation allocation suggestions, optimal transportation routes, transshipment batch arrangements, and departure time planning. The system synchronously pushes this plan to the distribution and scheduling system, which then executes loading, transshipment, and scheduling operations according to the plan, thus implementing the optimization results. Subsequent monitoring of batch rate changes under this plan verifies the optimization effect.

[0043] The aforementioned data analysis method for batch loading and distribution of logistics goods includes optimizing standard operating procedures based on the attribution analysis report to form a closed loop for intelligent management of logistics goods loading and distribution. Please refer to [link / reference needed]. Figure 7 As shown, the steps include: Step 421: Based on the systemic operational problems identified in the attribution analysis report, extract the high-frequency patterns in batches related to specific cargo types and operational steps; for systemic operational problems such as operational omissions and missing processes identified in the attribution analysis report, extract the high-frequency problem scenarios and problem steps related to specific cargo types and key operational steps, identify the target for process optimization.

[0044] Step 422: Based on the preset rule template, automatically generate standard operating procedure (SOP) revision suggestions, including adding operation checkpoints, modifying operation processes, and adding review steps; based on the system's preset operation process revision rule template, combined with the batch high-incidence pattern and root cause of the problem, automatically generate SOP revision suggestions, specifically including adding operation checkpoints for easily batched steps, adjusting unreasonable operation process sequences, adding double-person review and verification steps for key steps, and supplementing special operation specifications for high-load loading scenarios, etc., which are in line with actual operational needs and are executable.

[0045] Step 423: Push the revision suggestions to the operations management terminal. After review and approval, the electronic work instructions will be automatically updated, and relevant operators will be notified. The generated standard operating procedure revision suggestions will be pushed to the operations management terminal. After review and confirmation by management personnel, the system will automatically update the electronic work instructions and simultaneously issue process update notifications to the corresponding distribution centers, operation teams, and frontline operators, urging operators to operate according to the revised specifications. Subsequently, the process will be continuously optimized based on post-mortem data to form a closed-loop management system.

[0046] To further improve the closed-loop intelligent management of logistics cargo allocation, the data analysis method for batch allocation of logistics cargo also includes: Step 5: Establish a trusted traceability and evidence storage system based on blockchain technology. To further ensure the traceability of batch loading control and the objectivity of liability determination, leveraging the decentralized, tamper-proof, and traceable characteristics of blockchain technology, a dedicated trusted traceability and evidence storage system will be established. This system will connect to the company's existing logistics data management platform to achieve intelligent closed-loop control of the entire batch loading process, from identification, analysis, and intervention to traceability and accountability, thus solidifying the foundation of data trust and resolving liability disputes. Please refer to [link / reference needed]. Figure 8 As shown, it includes the following steps: Step 501: Establish a trusted traceability and evidence storage system based on blockchain technology, including the following steps: At key nodes in the flow of each sub-item of goods, perform hash calculations on the corresponding operation information to generate unique hash values, upload and store them to the blockchain network to form a complete and tamper-proof digital certificate chain, clearly recording the flow trajectory, separation and convergence nodes, corresponding time, geographical location and related sub-item information of each sub-item.

[0047] Specifically, this system is built on blockchain technology and connects to the enterprise's existing logistics data management platform. At key nodes in the flow of sub-items, such as warehousing, sorting, loading, and transshipment, the system automatically collects information such as operation time, operators, operation content, and cargo status. It then uses the SHA-256 hash algorithm to calculate a unique and immutable hash value. This hash value, along with the corresponding operation information, is then synchronously uploaded to the blockchain network. The hash values ​​of each node are sequentially linked to form a complete digital credential chain, accurately recording the entire flow trajectory of each sub-item, its separation and convergence nodes, geographical location, and related sub-item information, providing a reliable basis for subsequent traceability and liability determination.

[0048] Step 502: Based on the aforementioned trusted traceability and evidence storage system, responsibility determination is achieved. The specific steps are as follows: When a dispute arises regarding the batch loading or when it is necessary to define responsibility, the authorized entity can objectively and indisputably trace the specific transfer links, corresponding handover vouchers, and related operating entities involved in the batch loading by querying the evidence storage records on the blockchain, providing authoritative technical basis for dispute resolution and responsibility determination.

[0049] Specifically, when determining liability based on a blockchain-based evidence storage system, it is necessary to first clarify the authorizing entity and its corresponding permissions. Authorizing entities include the express delivery company's operations management department, regulatory agencies, shippers, consignees, and other parties involved in the dispute. The system will assign tiered query permissions according to the type of entity. For example, parties involved in the dispute can query the evidence storage information of their own associated waybills, while regulatory agencies can query relevant evidence storage data across the entire network, thus ensuring the security and controllability of the evidence storage data. After the authorizing entity completes permission confirmation, it can initiate an evidence storage query request through the company's dedicated blockchain evidence storage query port by entering core retrieval information such as the main waybill number and sub-waybill number. The system will automatically match the corresponding evidence storage record in the blockchain network. After the query is completed, the authorizing entity can intuitively obtain the tamper-proof evidence storage details on the blockchain, accurately tracing the specific flow links, corresponding handover documents, and related operating entities involved in the batch loading. Finally, using this evidence storage record as authoritative technical evidence, the responsible parties for the batch loading are clarified, such as the operators and their teams corresponding to operational errors, and the distribution centers corresponding to process loopholes, quickly defining the attribution of responsibility, effectively resolving disputes between shippers and the company, and among various nodes within the company, and efficiently promoting dispute resolution.

[0050] Please see Figure 9 As shown, an intelligent analysis and management device for batch loading of logistics goods includes: The judgment logic construction module 601 is used to pre-establish judgment logic for batch loading of logistics goods based on the business scenario type of distribution and transshipment, loading and dispatch, the time window of cargo flow, and the physical attributes of the goods. In this embodiment, the judgment logic construction module 601 adopts an embedded industrial-grade hardware unit, which is integrated and deployed to the logistics headquarters management server to ensure unified management and distribution of rules, without being distributed to various distribution centers, thus avoiding inconsistent rules and cumbersome operation and maintenance. It is equipped with customized rule configuration and verification software, which is mainly used to pre-build standardized batch loading judgment logic based on the business scenario of distribution and transshipment, loading and dispatch, the time window of cargo flow, and physical attributes such as cargo weight, volume, and number of pieces. During implementation, the module has a built-in rule editing engine that distinguishes between two scenarios: distribution / outlets and loading / shipping. It sets quantitative judgment thresholds and unified judgment logic, automatically eliminating non-batch scenarios such as normal distribution and separate loading of special goods. The generated standardized rule library is stored in an encrypted format, supporting manual updates by headquarters operations personnel and intelligent iteration by the system based on judgment accuracy. The rules are synchronized to each distribution judgment terminal via the intranet, operating stably 24 hours a day, providing a unified and unambiguous basis for batch judgment across the entire network, and completely eliminating judgment bias.

[0051] The batch event determination module 602 is used to acquire real-time cargo flow status data. At key operation nodes such as loading, unloading, and dispatch confirmation, it triggers the determination logic based on the cargo flow status data to automatically determine and mark batch events. The batch event determination module 602 adopts a hardware combination of edge computing terminal and data acquisition card, which is distributed and deployed on local servers in various distribution centers and core operation sites. It connects to the front-end acquisition equipment nearby to reduce data transmission latency. It is paired with real-time determination software and connects to the determination logic construction module 601 and the front-end logistics acquisition equipment through a dedicated intranet interface. Its core function is to acquire real-time cargo flow status data and trigger the determination logic and mark batch events at key operation nodes. During implementation, the module collects end-to-end data on waybills, scans, inventory, and handover documents in seconds via Ethernet, wireless communication, and multiple API interfaces. It monitors three key operational events in real time: loading scan, unloading verification, and dispatch confirmation. After an event is triggered, the module automatically retrieves the judgment logic issued by headquarters to complete quantitative calculations, accurately marking three states: pre-judged batching, confirmed batching, and no batching. It simultaneously records the judgment time, operation site, and associated voucher information, generates a real-time judgment record, and transmits it synchronously to the attribution analysis module. This enables the identification of batching events at the nearest location and full traceability, ensuring the real-time nature of the judgment.

[0052] The attribution analysis module 603 is used to acquire marked batch events and associate them with operational, cargo, and operational environment data. It identifies key influencing factors leading to batching through data mining methods and generates an attribution analysis report. The attribution analysis module 603 is equipped with a high-performance AI data processing chip and integrated into the logistics headquarters' big data analysis platform. This centralized processing of batch data from all network points avoids wasted computing power and inconsistent results in distributed analysis. It includes a built-in software package for statistical analysis, association rule mining, and machine learning multi-algorithm fusion. It directly connects with the batch event determination module 602 and its core function is to acquire marked batch events, associate multi-dimensional data, mine key influencing factors, and generate an attribution analysis report. During implementation, the module automatically collects confirmed batching event data uploaded by each distribution center, integrates information from three dimensions: operation, goods, and operating environment, and completes data cleaning, deduplication, and anomaly verification. Through layered algorithms, it deeply mines high-frequency batching patterns, locates core root causes, and distinguishes influencing factors such as human operation, process standardization, operating environment, and routing planning. Finally, it generates a visual attribution report, which intuitively presents the overall batching rate trend, the ranking of each responsible entity, and the distribution of problematic links. The report is pushed to the closed-loop optimization module in real time, providing accurate data support for global management and optimization.

[0053] The closed-loop optimization module 604 is used to optimize the routing planning algorithm or revise the standard operating procedure based on the attribution analysis report, forming a closed-loop intelligent management system for logistics cargo allocation that consists of "judgment-analysis-optimization-execution". The closed-loop optimization module 604 is a pure software plug-in module that seamlessly integrates into the enterprise headquarters' waybill system, scheduling system, performance evaluation system, and operation management system. It requires no additional hardware and relies on the existing system's computing power. Its core function is to optimize the routing algorithm or revise the operating procedure based on the attribution analysis report, forming a complete intelligent management closed loop. During implementation, the module addresses batching issues caused by unreasonable routing by invoking a dynamic routing batching algorithm to optimize transportation plans with the goal of transferring items in the same batch, and then pushes the optimized plans to the distribution and scheduling system for execution. For operational oversights and missing processes, it automatically generates suggestions for revising operating procedures, updates electronic operating manuals after headquarters review, and simultaneously distributes them to frontline operating terminals. At the same time, it calculates the batching rate of each distribution center and work group weekly / monthly, synchronizes it to the performance evaluation system, and iterates control strategies based on optimization results to achieve a closed-loop control system of "judgment-analysis-optimization-execution-review," steadily reducing the batch loading rate.

[0054] The Trusted Traceability and Evidence Storage Module 605 is used to encrypt and store information on the blockchain for all stages of batch loading, including batch determination, transfer operations, and optimized disposal, based on the full-process operation data of the device. This ensures that the data related to batch loading is tamper-proof and traceable, providing a reliable support for the intelligent closed-loop management of logistics cargo loading. This module combines an encrypted security chip with blockchain node software, integrated into the logistics headquarters' data security server, and connected to the enterprise's dedicated consortium blockchain network. It achieves full-link data communication with the batch event determination module 602, the attribution analysis module 603, and the closed-loop optimization module 604, balancing data security and traceability. Its core function is to achieve trusted evidence storage, responsibility definition, and dispute resolution throughout the entire batch loading process. During implementation, the module automatically collects all information, including operation time, operators, workstations, judgment results, optimization records, and handover vouchers, at key flow nodes such as goods receiving, sorting, loading, transshipment, and unloading, as well as at each stage of batch determination, attribution analysis, and closed-loop optimization. It then encrypts the information using the SHA-256 hash algorithm to generate a unique and immutable hash value, which is simultaneously uploaded to the consortium blockchain network storage. The hash values ​​of each node are sequentially linked to form a complete digital voucher chain. When receiving query requests from authorized entities such as enterprise operations departments, regulators, and consignors / consignees, the module retrieves the on-chain evidence data according to hierarchical permissions. This objectively traces the occurrence nodes, operating entities, and handling processes of batch loading, accurately defining responsibility, preventing shirking of responsibility and data tampering, providing authoritative technical support for dispute resolution and accountability, and perfecting the intelligent closed-loop management of the entire process.

[0055] Please see Figure 10 As shown, an intelligent analysis and management device 700 for batch loading of logistics goods includes: At least one processor 701; and a memory 702 communicatively connected to the at least one processor 701, the two achieving high-speed data interaction via a motherboard bus to ensure real-time instruction execution and data transmission; wherein, the memory 702 stores instructions executable by the at least one processor 701, the instructions being executed by the at least one processor 701 to cause the electronic device to perform the logistics cargo batch loading data analysis method described above. In this embodiment, the memory 702 includes DDR4 32GB high-speed memory and a 1TB SSD solid-state drive. The DDR4 high-speed memory is used for temporary storage of real-time collected logistics flow data, batching determination results, attribution analysis intermediate data, algorithm operation cache, etc., to meet the temporary data storage needs of multi-task parallel processing and ensure that the data read and write speed matches the real-time control requirements of the entire link; the SSD solid-state drive is used for persistent storage of computer-readable instructions, preset batching determination rule base, historical batching data, blockchain evidence data, system operation logs, etc., and has the characteristics of storage security, high reading efficiency, shock resistance and durability, suitable for industrial-grade long-term operation scenarios, and preventing data loss and damage.

[0056] The processor 701 is an Intel Core i7-12700H high-performance processor 701, which has multi-core and multi-thread processing capabilities, supports multi-task parallel operation, and can efficiently call computer-readable instructions in memory 702 to fully execute all the technical steps of the intelligent analysis and management method for batch logistics goods described above. It covers the entire process of judgment logic construction, batch event judgment, attribution analysis, closed-loop optimization, and trusted evidence storage, meeting the computing power requirements for real-time processing, in-depth analysis and intelligent control of massive logistics data.

[0057] The electronic device 700 also includes at least one storage medium 703, at least one power supply 704, at least one communication interface 705, and at least one input / output interface 706. Various extended hardware components work in conjunction with the core hardware to improve the device's functionality.

[0058] The storage medium 703 uses a large-capacity mechanical hard drive for persistent storage of massive application and historical business data. It can be expanded according to enterprise operational needs. The storage medium 703 contains modular program components, each corresponding to a series of instructions within a method. The processor 701 communicates with the storage medium 703 and can call upon these instructions to assist in executing the entire process. The power supply 704 is an industrial-grade regulated power supply, ensuring uninterrupted and stable power supply to the equipment 24 hours a day, preventing data loss and operational interruptions due to power outages. The communication interface 705 includes an Ethernet wired interface and a wireless communication interface. The Ethernet interface is used for wired connection between the device and scanning equipment, inventory management equipment, and control terminals at various distribution centers and operating sites, enabling stable access to end-to-end logistics data and instruction issuance. The wireless communication interface 705 is used for wireless interaction between the device and management personnel's mobile terminals and the enterprise consortium blockchain network, enabling operations such as pushing analysis reports, sending anomaly warnings, uploading and querying blockchain-based evidence data.

[0059] The electronic device is equipped with an operating system adapted to server operation, such as Windows Server, Unix, Linux, FreeBSD, etc., with strong system compatibility and stable operation, and can be adapted to various logistics management software and algorithm programs. The computer-readable instructions stored in the memory 702 are adapted to the operating system. When the instructions are executed by the processor 701, they can accurately realize all the functions mentioned above, such as batch allocation determination, attribution analysis, closed-loop optimization, and reliable evidence storage. This provides logistics companies with one-stop hardware support for batch loading data analysis, intelligent control and responsibility traceability, and realizes a closed-loop intelligent management of the entire process.

[0060] A computer-readable storage medium stores a computer program that, when executed by a processor 701, implements the aforementioned data analysis method for batch loading of logistics goods. In specific implementations, the storage medium can be any one or more of the following: USB flash drive, portable hard drive, solid-state drive, optical disc, flash memory, etc., to meet the equipment operation requirements of the aforementioned express delivery companies.

[0061] It is understood that those skilled in the art can make equivalent substitutions or modifications to the technical solution and inventive concept of the present invention, and all such substitutions or modifications should fall within the protection scope of the appended claims.

Claims

1. A method for analyzing data on batch loading and distribution of logistics goods, characterized in that, Including the following steps: The logic for determining the batch loading of logistics goods is established in advance based on business scenarios, time windows and physical attributes of goods; Real-time acquisition of cargo flow status data; triggering the judgment logic based on the cargo flow status data at key operation nodes to automatically determine and mark batch events. Acquire the tagged batch events and associate them with operation dimension data, cargo dimension data and operating environment dimension data. Identify the key influencing factors that cause batching through data mining methods and generate an attribution analysis report. Based on the attribution analysis report, optimize the routing planning algorithm and / or standard operating procedures to form a closed loop of intelligent management for logistics cargo loading.

2. The method for analyzing logistics cargo batch loading data according to claim 1, characterized in that, Based on business scenarios, time windows, and the physical attributes of goods, a pre-established logic for determining the batch loading of logistics goods is established, including: For distribution to distribution or distribution to network business scenarios, when the main order is confirmed to be dispatched after the loading handover order at the responsible station, if the number of unloading scanned items received by the next station corresponding to the loading handover order is less than the number of items in the inventory of the main order at the responsible station within a preset time threshold, it is determined that a batch loading event has occurred. In the loading business scenario, when the loading handover order of the master order is confirmed for dispatch at the responsible station, if the number of scanned items in the current loading handover order is less than the number of items in the inventory of the master order at the current responsible station, and the inventory weight and / or volume of the master order meets the preset threshold conditions, then it is marked as a pre-judged batching state.

3. The method for analyzing logistics cargo batch loading data according to claim 1, characterized in that, Real-time acquisition of cargo flow status data; triggering the judgment logic based on the cargo flow status data at key operation nodes; automatically determining and marking batch events, including the following steps: Real-time acquisition of cargo flow status data, including waybill data, scan data, inventory data, and handover document data; Real-time monitoring of key operational events at each distribution center, including loading scan events, unloading scan events, and departure confirmation events; When a critical operation event is detected, the associated cargo flow status data is queried based on the master order number carried by the critical operation event. The data is then input into the rule engine to trigger the batching determination logic, automatically calculate and mark the batching event. The status of the batching event includes pre-batching, confirmed batching, or no batching.

4. The method for analyzing logistics cargo batch loading data according to claim 1, characterized in that, Acquire the tagged batch events and correlate them with operational, cargo, and operational environment data. Use data mining techniques to identify key influencing factors leading to batching and generate an attribution analysis report, specifically including: Retrieve all batch events marked as confirmed batches from the database; The tagged batch events are integrated and associated with operation-dimensional data, cargo-dimensional data, and operational environment-dimensional data. The operation-dimensional data includes the responsible distribution center, operation team, handover order number, departure time, and unloading time. The cargo-dimensional data includes the master order number, a list of sub-order numbers, total cargo volume, total weight, number of sub-items, and the arrival timestamps and time differences for each sub-item. The operational environment-dimensional data includes vehicle load saturation at departure, the current throughput load of the distribution center, and the daily routing plan information. Statistical analysis, association rule mining, or machine learning algorithms are used to conduct in-depth analysis of the operational dimension data, cargo dimension data, and operational environment dimension data to identify high-frequency batching patterns and potential root causes. Based on the aforementioned high-frequency batching mode and the potential root cause generation visualization attribution analysis report, the batching rate trend, the ranking of responsible parties, the distribution of attribution categories, and the location of specific problem links are displayed.

5. The method for analyzing logistics cargo batch loading data according to claim 1, characterized in that, Based on the attribution analysis report, the routing planning algorithm is optimized to form a closed loop for intelligent management of logistics cargo loading, including the following steps: The attribution analysis report is fed back to the route planning system; During the waybill generation stage, the physical attributes and destination information of the master waybill to be allocated and all its sub-items are obtained. Dynamic routing and batch optimization algorithms are called, with the allocation of all sub-items of the same shipment to the same handover batch as the core optimization objective. Multi-objective optimization calculations are performed in combination with transportation costs and timeliness requirements. Generate a transportation plan that includes batching suggestions and push the transportation plan to the distribution and scheduling system for execution.

6. The method for analyzing logistics cargo batch loading data according to claim 5, characterized in that, The dynamic routing and batch optimization algorithm is a graph theory-based path search algorithm, which specifically includes: The logistics network is abstracted as a graph structure, where nodes are distribution centers, edges are transportation routes, and edge weights include distance, timeliness, and cost. For each shipment, when searching for the path from the starting node to the target node in the graph, a batching constraint factor is introduced. If a certain path causes the sub-items to be separated, then a higher weight value is assigned to that path. Choose the path with the lowest total cost as the final routing scheme, and ensure that all sub-components follow the same path.

7. The method for analyzing logistics cargo batch loading data according to claim 1, characterized in that, Based on the attribution analysis report, the standard operating procedures are optimized to form a closed loop for intelligent management of logistics cargo loading, including the following steps: Based on the systemic operational problems identified in the attribution analysis report, we extracted the high-incidence patterns in batches that are related to specific cargo types and operational procedures. Based on preset rule templates, suggestions for revising standard operating procedures are automatically generated, including adding operation checkpoints, modifying operation processes, and adding review steps. The revision suggestions will be pushed to the operations management end. After review, the electronic work instructions will be automatically updated and the relevant operators will be notified.

8. An intelligent analysis and management device for batch loading of logistics goods, characterized in that, include: The judgment logic construction module is used to pre-establish the judgment logic for batch loading of logistics goods based on the business scenario type of distribution and transshipment, loading and dispatch, the time window of cargo flow and cargo attributes. The batch event determination module is used to acquire cargo flow status data in real time. At key operation nodes such as loading, unloading, and dispatch confirmation, the determination logic is triggered based on the cargo flow status data to automatically determine and mark batch events. The attribution analysis module is used to acquire the tagged batch events and associate them with operation dimension data, cargo dimension data and operating environment dimension data. It uses data mining methods to identify the key influencing factors that lead to batching and generates an attribution analysis report. The closed-loop optimization module is used to optimize the routing planning algorithm or revise the standard operating procedure based on the attribution analysis report, forming a closed-loop intelligent management system for logistics cargo loading that consists of "judgment-analysis-optimization-execution".

9. An intelligent analysis and management device for batch loading of logistics goods, characterized in that, include: At least one processor; And a memory communicatively connected to the at least one processor; wherein the memory stores instructions executable by the at least one processor, the instructions being executed by the at least one processor to cause the device to perform the logistics cargo batch loading data analysis method according to any one of claims 1-7.

10. A computer-calibrated storage medium, characterized in that, The computer-readable storage medium stores a computer program that, when executed by a processor, implements the logistics cargo batch loading data analysis method according to any one of claims 1-7.

Images