Tray warehouse automatic replenishment box warehouse replenishment method and system based on historical delivery data

Abstract

The invention discloses a historical delivery data-based replenishment method and system for a tray warehouse automatic replenishment box warehouse, and the method comprises the steps: firstly configuring a historical delivery data calculation period (quarter / month / week), carrying out the flexible configuration based on the actual business condition, calculating the packaging form (tray / box / piece) data of a delivered material in combination with the package gauge of the material according to the historical delivery data, and carrying out the calculation of the packaging form (tray / box / piece) data; and obtaining the minimum form storage water level cardinal number and the material ABC classification of each material in the box warehouse. And then, a replenishment method is designed based on the stock storage water level cardinal number of the part box and the stock storage condition of the part box library, the capacity of the part box library and a first-in first-out rule need to be considered when the replenishment number is calculated, bin explosion caused by excessive replenishment is avoided, and the delivery requirement can be met. And finally, a box library replenishment task is generated, and the replenishment task off-shelf action is executed by balancing the number of tasks in the task pool, and the replenishment task generation can also be compatible with manual unpacking and robot unpacking according to the configuration condition. And collaborative optimization of efficiency and benefits of the whole warehousing system is realized.

Description

Technical Field

[0001] This invention relates to the field of automatic replenishment control technology for intelligent warehousing systems, specifically to a replenishment method and system for pallet warehouses and box warehouses based on historical shipping data. Background Technology

[0002] In the logistics industry, intelligent warehousing systems have become a core technology for improving warehouse management efficiency and reducing operating costs. With the rapid development of e-commerce and the widespread adoption of omnichannel retail models, logistics orders are characterized by massive growth, a wide variety of SKUs (stock keeping units), order fragmentation, and extremely high customer demands for delivery timeliness. Against this backdrop, the "smart brain" of automated logistics systems—automatic replenishment technology—is becoming increasingly important, directly impacting a company's ability to achieve a precise balance between "inventory costs" and "stockout losses."

[0003] In existing technologies, common automatic replenishment methods mostly rely on preset rules based on fixed formulas (such as safety stock models and reorder point models). When dealing with scenarios involving an extremely large number of SKUs and drastic demand fluctuations (such as the 3C industry), these methods reveal the following inherent flaws and technical shortcomings: Rigid and unresponsive predictive models: Traditional replenishment models struggle to accurately predict newly launched products or materials with short lifecycles and rapidly changing demand. They are unable to dynamically learn and adjust based on multi-dimensional historical shipping data, leading to a disconnect between replenishment decisions and actual demand.

[0004] The replenishment logic is disconnected from the physical constraints of the warehouse area: Most methods only focus on the quantity of inventory without fully considering the actual capacity limitations of the target warehouse area (such as a parcel warehouse). This can easily lead to over-replenishment causing "warehouse overload" or insufficient replenishment failing to meet delivery needs, lacking the intelligent decision-making ability to achieve optimal inventory distribution under limited warehouse capacity.

[0005] Traditional replenishment systems often suffer from limited operational modes and low equipment utilization. They struggle to flexibly adapt to various operational modes, such as manual pallet unpacking and automated robotic pallet unpacking. In particular, when introducing robotic operations, key technical constraints such as the robot's physical working dimensions and grasping feasibility are not incorporated into the pre-replenishment task decision-making logic, resulting in low utilization of automated equipment and even requiring frequent manual intervention.

[0006] The data processing is limited in scope, resulting in low inventory turnover efficiency: Existing methods typically only consider the current inventory level when calculating replenishment quantities, failing to effectively integrate data such as the production date and batch number of materials in the pallet warehouse for linked calculations. This may lead to non-compliance with the "first-in, first-out" principle, increasing the risk of inventory backlog and expiration.

[0007] Therefore, there is an urgent need in this field for an intelligent replenishment method that can overcome the above-mentioned defects. It can make accurate and dynamic demand forecasts based on historical data, make optimal replenishment decisions under complex physical and business constraints (such as warehouse capacity, equipment capacity, and inventory turnover), and flexibly schedule different operational resources, thereby achieving synergistic optimization of the efficiency and effectiveness of the entire warehousing system. Summary of the Invention

[0008] The purpose of this invention is to provide a method and system for automatically replenishing pallet warehouses and box warehouses based on historical shipping data, addressing the aforementioned problems.

[0009] The technical solution of the present invention is as follows: A method for automatically replenishing pallet and box warehouses based on historical shipping data includes the following steps: Historical data collection and processing steps: According to the configured period, retrieve historical shipping order data from the warehouse management system; Material classification steps: Based on the quantity of shipped materials in historical shipping order data, dynamically calculate and maintain the ABC classification of materials; Inventory level calculation steps: Based on historical shipping order data, ABC classification, and warehouse capacity data, the upper and lower limits of the inventory level for each material in the warehouse are calculated using a predictive model. Replenishment task generation steps: Based on the upper and lower limits of inventory level, preset replenishment rules, and material turnover of pallet warehouse inventory, replenishment tasks are automatically generated through an optimized algorithm and allocated to the task pool. Replenishment task execution steps: Monitor the number of tasks in the task pool. When the number of tasks is lower than the set threshold, automatically trigger the execution of the replenishment task, control the corresponding automated equipment or generate manual instructions to complete the replenishment and unpacking action from the pallet warehouse to the box warehouse.

[0010] By employing the methods described above, dynamic inventory level calculation based on predictive models (such as LSTM) and multidimensional data, and by utilizing time-series predictive models to process historical shipping data, seasonality, and promotional information, replenishment decisions are shifted from passive response to proactive forecasting. This significantly reduces stockout rates or inventory redundancy caused by inaccurate forecasts. It also enhances the system's robustness in the face of market fluctuations: enabling the replenishment system to adapt to complex scenarios such as sales peaks and promotional activities, maintaining stable and reasonable inventory levels.

[0011] Furthermore, in the inventory level calculation step, the lower limit of the inventory level is calculated. The formula is: , in, This represents the average number of pallets split into smaller boxes per day. The number of pallets used for full-case shipments. From 0 to Configurable weighting coefficients within the range, This refers to the number of remaining boxes for full-case shipment. This represents the average number of parts disassembled per box per day. The conversion rate between boxes and pieces; Calculate the upper limit of the inventory level The formula is: , in, This represents the total number of warehouse space occupied by the parcel lockers. This represents the current number of bits in the database. This is the percentage of material shipments calculated based on historical data. The conversion rate is between boxes and pieces.

[0012] By dynamically linking the total number of storage locations, current occupancy rate, and material shipment ratio using the above methods, the mathematical model ensures that replenishment volume remains within the physical carrying capacity of the storage area, fundamentally avoiding the risk of "warehouse overload" and improving storage location utilization. Furthermore, by linking water level calculations with ABC classification, the system ensures that high-value, high-turnover A-class materials always have optimal inventory levels, thus optimizing overall capital utilization.

[0013] Furthermore, the basis for material turnover specifically includes: Calculate the minimum replenishment date for materials in the pallet warehouse based on the configured turnover unit; where the turnover unit is month, the minimum replenishment date is the last day of the month containing the minimum production date; when the turnover unit is week, the minimum replenishment date is the last day of the week containing the minimum production date; when the turnover unit is day, the minimum replenishment date is the minimum production date itself. Pallet warehouse inventory with a production date earlier than or equal to the minimum replenishment date is included in the available inventory base of the box warehouse materials.

[0014] Furthermore, the process of determining the dismantling method when generating a replenishment task is as follows: Prioritize generating robot disassembly tasks; Before generating the robot disassembly task, the 3D vision module is called to obtain the image of the pallet to be disassembled, and the robot arm grasping path is simulated in the virtual environment through a collision detection algorithm. If the simulation is successful, a robot disassembly task will be generated; if the simulation fails, the task type will be automatically changed to a manual disassembly task.

[0015] By employing the methods described above, including virtual simulation and collision detection before task generation, the success rate of robot disassembly tasks is improved, reducing task interruptions and human intervention, thereby enhancing the overall utilization rate of the robot equipment. This mechanism enables the system to intelligently determine which situations are suitable for robot operation and which require manual intervention, forming an efficient and smooth human-robot collaborative workflow.

[0016] Furthermore, in the replenishment task execution step, when replenishing tasks to the task pool, a load-balanced weighted round-robin scheduling algorithm is adopted to dynamically allocate replenishment tasks according to the current task processing speed and queue length of each picking station, so that the number of tasks to be processed in the task pool of each station is maintained near a set threshold.

[0017] The algorithm described above ensures the smooth flow of replenishment tasks at each stage by monitoring and dynamically allocating the task load of each station in real time, reducing the waiting time of equipment and personnel, and improving the overall efficiency of the delivery operation.

[0018] This application also includes a replenishment system for an automated pallet warehouse replenishment system based on historical shipping data, used to implement a replenishment method for an automated pallet warehouse replenishment system based on historical shipping data, including: The data acquisition and processing module is configured to acquire and process historical shipping order data from the warehouse management system according to the configured period. The material classification module is configured to dynamically calculate and maintain the ABC classification of materials based on historical shipping order data. The intelligent water level calculation module is configured to calculate the upper and lower limits of the inventory water level for each material in the box warehouse based on historical shipping order data, ABC classification, and the warehouse capacity data through a predictive model. The replenishment task decision and generation module is configured to automatically generate replenishment tasks based on the upper and lower limits of inventory levels, preset replenishment rules, and material turnover of pallet warehouse inventory through optimized algorithms. The task scheduling and execution module is configured to manage the task pool, monitor the number of tasks in the task pool, and automatically trigger and schedule replenishment tasks to the corresponding automated equipment or generate manual instructions when the number of tasks is lower than the set threshold.

[0019] Furthermore, the prediction model integrated in the intelligent water level calculation module is a Long Short-Term Memory Network (LSTM) model or a Transformer time series prediction model. The input of the model includes at least the shipping data of the past N periods, seasonal factors, and promotional activity information. The output is the predicted value of the parcel-level shipping volume for the next T days. This predicted value is used to calculate the lower limit of the inventory water level.

[0020] Furthermore, the optimization algorithm used in the replenishment task decision and generation module is a multi-objective optimization algorithm. The multi-objective optimization algorithm takes the minimum total replenishment cost and the minimum stockout risk as the objective functions, and solves the problem with the upper limit of the box warehouse capacity, the robot working efficiency, and the task pool capacity as constraints.

[0021] By embedding the "minimum replenishment date" logic into the calculation using the above method, the system prioritizes replenishing older batches, automatically adhering to the first-in, first-out (FIFO) principle and shortening the average inventory age. By optimizing the algorithm to simultaneously balance cost, efficiency, and risk, the system solves the problem of traditional rules struggling to balance multiple conflicting objectives (such as minimum replenishment frequency vs. minimum stockout risk), thus reducing overall replenishment costs.

[0022] Furthermore, the replenishment task decision and generation module also includes a robot task preprocessing unit, which includes: A 3D vision subunit, configured to acquire point cloud data of the pallet to be removed; The collision detection and path simulation subunit is configured to simulate the gripping path of a robotic arm in a virtual environment. The task decision subunit is configured to determine whether to generate a robot disassembly task or a manual disassembly task based on the simulation results.

[0023] This application also includes a non-transitory computer-readable storage medium storing a computer program that, when executed by a processor, implements the steps of a method for automatically replenishing a pallet warehouse and a box warehouse based on historical shipping data.

[0024] Compared with existing technologies, the advantages of this invention are: By designing replenishment rules and calculation methods, materials requiring partial replenishment are pre-stocked in the parcel warehouse based on turnover criteria, significantly shortening the delivery cycle. Furthermore, the modular replenishment rules improve equipment utilization and save on labor costs. Flexible parameter configuration also better adapts to the diverse replenishment needs of different industries. Attached Figure Description

[0025] Figure 1 This is a flowchart of the method described in this application.

[0026] Figure 2 This is a flowchart for calculating the water level in the container warehouse in this application.

[0027] Figure 3 A flowchart is generated for the replenishment task in this application. Detailed Implementation

[0028] It should be noted that relational terms such as "first" and "second" are used merely to distinguish one entity or operation from another, and do not necessarily require or imply any such actual relationship or order between these entities or operations. Furthermore, the terms "comprising," "including," or any other variations thereof are intended to cover non-exclusive inclusion, such that a process, method, article, or apparatus that comprises a list of elements includes not only those elements but also other elements not expressly listed, or elements inherent to such a process, method, article, or apparatus. Without further limitations, an element defined by the phrase "comprising one..." does not exclude the presence of other identical elements in the process, method, article, or apparatus that includes said element.

[0029] The features and performance of the present invention will be further described in detail below with reference to embodiments.

[0030] Please see Figure 1-3 A method for automatically replenishing pallet and box warehouses based on historical shipping data, the specific implementation process of which is as follows: Figure 1 As shown, the historical shipment data calculation period (quarterly / monthly / weekly) is first configured, which can be flexibly configured based on actual business conditions. Based on historical shipment data, and combined with the material packaging specifications, the packaging form (pallet / box / piece) of the shipped materials is calculated to obtain the minimum form storage level base and material ABC classification for each material in the box / piece warehouse. Then, a replenishment method is designed based on the box / piece inventory storage level base and the box / piece warehouse inventory status. When calculating the replenishment quantity, the capacity of the box / piece warehouse and the first-in, first-out (FIFO) rule need to be considered to avoid over-replenishment and warehouse overflow while meeting shipment needs. Finally, a box / piece warehouse replenishment task is generated, and the number of tasks in the task pool is balanced to execute the replenishment task removal action. The replenishment task generation can also be configured to be compatible with both manual unpacking and robotic unpacking.

[0031] This application example demonstrates a method for optimizing replenishment tasks based on historical shipping data. The specific implementation process is shown below: In this application example, such as Figure 2 As shown in the process, historical shipment data is dynamically maintained through specific periods. Materials are categorized into ABC classes. By analyzing ABC classification and shipping data, accurate warehouse inventory requirements can be obtained. Precise replenishment decisions can respond to rapidly changing market demands.

[0032] The ABC classification calculation rule differs from the industry standard. Without material unit prices, it calculates ABC classification based on the shipment volume within the order cycle, using a 5:3:2 ratio. Based on historical order information and the warehouse shipment logistics route, it summarizes the warehouse shipment data and the shipment percentage of materials within the cycle according to warehouse area and material category. Warehouse shipment data is divided into four categories: full pallet shipments, palletized shipments, full carton shipments, and cartonized shipments. A secondary analysis is then performed based on these four data categories, following these steps: Excluding data on shipments of entire boards, shipments of entire boards have the highest efficiency. Calculate the average number of pallets and boxes to be broken down daily based on packaging data (pallet and box conversion rate). ; Calculate the number of pallets required for full-case shipments based on packaging data (pallet and box conversion rate). ) and remaining boxes ; The formula for calculating the lower limit of the inventory level in the parts box warehouse is as follows:

[0033] in This is the minimum quantity for the box warehouse. Weighting coefficients ( This can be configured in the system. Conversion rates for boxes and pieces; Get the total number of storage locations in the storage area Calculate the currently occupied storage space , This represents the percentage of materials shipped. To account for the conversion rate between boxes and pieces, 10% of the storage space needs to be reserved to meet temporary buffering needs within the storage area. The formula for calculating the upper limit of the material inventory level in the box and piece storage area is as follows: , The calculated upper and lower limits of material water level are visualized. Based on the material type, you can set it to calculate only the upper and lower limits of the inventory level for specified types of A, B, or C materials; by default, you calculate the upper and lower limits of the inventory level for all materials.

[0034] In this application example, such as Figure 3As shown in the process, based on system configuration parameters, replenishment tasks are generated and automatically executed by balancing the various business tasks. Flexible parameter configuration can handle changing business needs and is compatible with different practical scenarios. All replenishment rules are configured as system parameters, and the replenishment execution cycle is the replenishment task generation cycle. The upper limit for the number of replenishment tasks is the maximum number of tasks that can be executed on a daily basis, preventing an overabundance of tasks that cannot be completed. This can be flexibly set based on actual staffing levels and work efficiency. The dismantling method is compatible with both manual and mechanical operation modes. For robot dismantling, the limitations of the robot's working size and the corresponding box structure data must be considered.

[0035] Replenishment tasks are automatically generated based on the upper and lower limits of the parcel locker water level and replenishment rules, including the following steps: Set up a timed task, specify the execution cycle, and loop through all valid material water level upper and lower limit data. The current inventory status of the parcel warehouse is calculated according to the water level calculation rules. If the full quantity calculation is configured, the sum of the available inventory and the quantity of replenishment in transit in the parcel warehouse is used as the current inventory base of the parcel warehouse. If the turnover basis is configured, the sum of the available inventory and the quantity of replenishment in transit in the parcel warehouse is calculated based on the pallet warehouse inventory and the material turnover basis, and is used as the inventory base of the current parcel warehouse. The material is set to a lower limit for water level and the total available inventory of materials in the box warehouse are compared. If the total available inventory is less than the lower limit, a replenishment task is generated. When the number of replenishment tasks reaches the upper limit, the generation of the current material replenishment task is terminated. Based on the configuration settings of the dismantling method, a replenishment task type is generated. The logistics route is selected by the task type, and different dismantling modes are executed. Robot dismantling is given priority, and if the robot dismantling rules are not met, it will automatically switch to manual dismantling. Calculate the total number of replenishment tasks for the day, categorize them by task type, and terminate the generation of replenishment tasks for the day when the total number of replenishment tasks reaches the set task limit. After a replenishment task is successfully generated, the total quantity of the current replenishment task is added to the quantity in transit and stored in the upper and lower limit configuration of the material. Deleting or canceling a replenishment task will deduct the quantity of materials in transit for replenishment tasks. In the first calculation cycle of the day, delete the replenishment tasks that were not executed the previous day and reset the replenishment quantities in transit; The steps for calculating the inventory of boxes and containers based on turnover are as follows: The minimum replenishment date for palletized warehouses is calculated based on shipment turnover, with the turnover period being monthly. The replenishment date is the last day of the month in which the minimum production date falls; the turnover is based on weeks, with the minimum replenishment date being the last day of the week in which the minimum production date falls; the turnover is based on days, with the minimum replenishment date being the minimum production date. The current inventory base of the parts and cases warehouse materials is calculated based on the minimum replenishment date. The total inventory with a production date less than or equal to the minimum replenishment date is considered as the inventory base of the parts and cases warehouse materials. The system automatically executes replenishment tasks and removes items from shelves based on the maximum number of tasks in the task pool set for each region / location. When the number of tasks in the task pool is less than the maximum, the system automatically adds the next replenishment task to the task pool.

[0036] The embodiments described above merely illustrate specific implementation methods of this application, and while the descriptions are detailed and specific, they should not be construed as limiting the scope of protection of this application. It should be noted that those skilled in the art can make various modifications and improvements without departing from the concept of the technical solution of this application, and these modifications and improvements all fall within the scope of protection of this application.

Claims

1. A method for automatically replenishing pallet warehouses and box warehouses based on historical shipping data, characterized in that, Includes the following steps: Historical data collection and processing steps: According to the configured period, retrieve historical shipping order data from the warehouse management system; Material classification steps: Based on the quantity of shipped materials in historical shipping order data, dynamically calculate and maintain the ABC classification of materials; Inventory level calculation steps: Based on historical shipping order data, ABC classification, and warehouse capacity data, the upper and lower limits of the inventory level for each material in the warehouse are calculated using a predictive model. Replenishment task generation steps: Based on the upper and lower limits of inventory level, preset replenishment rules, and material turnover of pallet warehouse inventory, replenishment tasks are automatically generated through an optimized algorithm and allocated to the task pool. Replenishment task execution steps: Monitor the number of tasks in the task pool. When the number of tasks is lower than the set threshold, automatically trigger the execution of the replenishment task, control the corresponding automated equipment or generate manual instructions to complete the replenishment and unpacking action from the pallet warehouse to the box warehouse.

2. The method for automatically replenishing pallet warehouses and box warehouses based on historical shipping data according to claim 1, characterized in that, In the inventory level calculation step, the formula for calculating the lower limit (Min) of the inventory level is: ， in, This represents the average number of pallets split into smaller boxes per day. The number of pallets used for full-case shipments. From 0 to Configurable weighting coefficients within the range, This refers to the number of remaining boxes for full-case shipment. This represents the average number of parts disassembled per box per day. The conversion rate between boxes and pieces; The formula for calculating the maximum upper limit of the inventory level is: in, This represents the total number of warehouse space occupied by the parcel lockers. This represents the current number of bits in the database. This is the percentage of material shipments calculated based on historical data. The conversion rate is between boxes and pieces.

3. The method for automatically replenishing pallet warehouses and box warehouses based on historical shipping data according to claim 1, characterized in that, The basis for material turnover specifically includes: Calculate the minimum replenishment date for materials in the pallet warehouse based on the configured turnover unit; where the turnover unit is month, the minimum replenishment date is the last day of the month containing the minimum production date; when the turnover unit is week, the minimum replenishment date is the last day of the week containing the minimum production date; when the turnover unit is day, the minimum replenishment date is the minimum production date itself. Pallet warehouse inventory with a production date earlier than or equal to the minimum replenishment date is included in the available inventory base of the box warehouse materials.

4. The method for automatically replenishing pallet warehouses and box warehouses based on historical shipping data according to claim 1, characterized in that, The process of generating a replenishment task involves deciding on the pallet splitting method, specifically: Prioritize generating robot disassembly tasks; Before generating the robot disassembly task, the 3D vision module is called to obtain the image of the pallet to be disassembled, and the robot arm grasping path is simulated in the virtual environment through a collision detection algorithm. If the simulation is successful, a robot disassembly task will be generated; if the simulation fails, the task type will be automatically changed to a manual disassembly task.

5. The method for automatically replenishing pallet warehouses and box warehouses based on historical shipping data according to claim 1, characterized in that, In the replenishment task execution step, when replenishing tasks to the task pool, a load-balanced weighted round-robin scheduling algorithm is adopted to dynamically allocate replenishment tasks according to the current task processing speed and queue length of each picking station, so that the number of tasks to be processed in the task pool of each station is maintained near a set threshold.

6. A replenishment system for pallet warehouses and box warehouses based on historical shipping data, characterized in that, A method for automatically replenishing a pallet warehouse and a box warehouse based on historical shipping data, as described in any one of claims 1-5, includes: The data acquisition and processing module is configured to acquire and process historical shipping order data from the warehouse management system according to the configured period. The material classification module is configured to dynamically calculate and maintain the ABC classification of materials based on historical shipping order data. The intelligent water level calculation module is configured to calculate the upper and lower limits of the inventory water level for each material in the box warehouse based on historical shipping order data, ABC classification, and the warehouse capacity data through a predictive model. The replenishment task decision and generation module is configured to automatically generate replenishment tasks based on the upper and lower limits of inventory levels, preset replenishment rules, and material turnover of pallet warehouse inventory through optimized algorithms. The task scheduling and execution module is configured to manage the task pool, monitor the number of tasks in the task pool, and automatically trigger and schedule replenishment tasks to the corresponding automated equipment or generate manual instructions when the number of tasks is lower than the set threshold.

7. A replenishment system for an automatic pallet warehouse and box warehouse based on historical shipping data according to claim 6, characterized in that, The prediction model integrated in the intelligent water level calculation module is either a Long Short-Term Memory Network (LSTM) model or a Transformer time series prediction model. The input of the model includes at least the shipping data of the past N periods, seasonal factors, and promotional activity information. The output is the predicted value of the parcel-level shipping volume for the next T days. This predicted value is used to calculate the lower limit of the inventory water level.

8. A replenishment system for an automatic pallet warehouse and box warehouse based on historical shipping data according to claim 6, characterized in that, The optimization algorithm used in the replenishment task decision and generation module is a multi-objective optimization algorithm. The multi-objective optimization algorithm takes the minimum total replenishment cost and the minimum stockout risk as the objective functions, and solves the problem under the constraints of the upper limit of the box warehouse capacity, the robot working efficiency, and the task pool capacity.

9. A replenishment system for an automatic pallet warehouse and box warehouse based on historical shipping data according to claim 6, characterized in that, The replenishment task decision and generation module also includes a robot task preprocessing unit, which includes: A 3D vision subunit, configured to acquire point cloud data of the pallet to be removed; The collision detection and path simulation subunit is configured to simulate the gripping path of a robotic arm in a virtual environment. The task decision subunit is configured to determine whether to generate a robot disassembly task or a manual disassembly task based on the simulation results.

10. A non-transitory computer-readable storage medium storing a computer program, characterized in that, When the computer program is executed by the processor, it implements the steps of the automatic replenishment method for pallet warehouses and box warehouses based on historical shipping data as described in any one of claims 1-5.

Images