A PFEP uniform fitting delivery point dynamic allocation method and system for assembly manufacturing scenarios

Abstract

The application provides a PFEP uniform set distribution point dynamic allocation method and system for an assembly manufacturing scene, which comprises the following steps: obtaining a warehouse layout, using density clustering to divide the warehouse into uniform set clusters and determining the adjacent relationship; obtaining material attributes, calculating the volume of a single set of uniform sets and dynamically calibrating the height limit coefficient; based on the production line configuration, the volume of the uniform set and the height coefficient, calculating the work station demand capacity and the minimum safe inventory set number. Iterating the materials, screening the candidate warehouse: the hard constraint is that the available capacity is greater than or equal to the minimum safe inventory, the uniform set constraint is that it is in the same cluster or adjacent to the allocated materials, and the soft constraint is that the capacity deviation is less than or equal to the preset threshold; then, based on the multi-objective scoring function, the recommended warehouse is selected; after obtaining the recommended warehouse of all materials in the set group, the set group is locked and occupied in batches, and if all the operations are successful, the operation is effective. The application greatly shortens the walking distance of uniform set distribution and improves the uniform set distribution efficiency.

Description

Technical Field

[0001] This invention relates to the field of assembly manufacturing material management technology, and in particular to a method and system for dynamic allocation of PFEP kit delivery points for assembly manufacturing scenarios. Background Technology

[0002] In complex assembly processes such as engineering machinery production, line-side storage location allocation typically employs a single-point mapping method from materials to storage locations. This means each material independently selects a storage location, a typical practice of the traditional PFEP (Plan For Every Part) method. This approach fails to impose spatial constraints on storage location allocation at the "set" level (i.e., the collection of all materials required for a complete product). In practice, multiple materials within the same set are often dispersed across different areas of the buffer zone. When complete set delivery is required, picking personnel must traverse long distances between different storage locations, resulting in lengthy picking paths, increased waiting times for complete sets, and in severe cases, even production line downtime. Existing technologies lack mandatory constraints on the spatial aggregation of materials within the same set, making it difficult to guarantee the timeliness and path concentration of set delivery, thus becoming one of the key bottlenecks restricting assembly efficiency. Summary of the Invention

[0003] This invention provides a method and system for dynamic allocation of PFEP (Package Delivery Equipment) points in assembly manufacturing scenarios to solve the problems existing in related technologies. The technical solution is as follows: In a first aspect, embodiments of the present invention provide a method for dynamic allocation of PFEP (Package Partition Delivery) points for assembly manufacturing scenarios, including: Obtain basic configuration data of the production line and layout data of the warehouse area. Based on the layout data of the warehouse area, use a density clustering algorithm based on distance threshold to divide each warehouse location into multiple clusters and determine the adjacency relationship between each cluster. Obtain the material list and attribute data of the current set of units to be allocated; calculate the complete volume of a single set based on the attribute data; and dynamically calibrate the height restriction coefficient; calculate the required capacity of the workstation and the minimum safety stock number of sets based on the basic configuration data of the production line, the complete volume of a single set, and the height restriction coefficient. Based on the material list, each material is traversed. For the current material, the available capacity of each candidate storage location is obtained. A set of candidate storage locations that simultaneously meet the hard constraint, kitting constraint, and soft constraint is selected. From the candidate storage location set, the recommended storage location for the current material is selected based on a multi-objective scoring function. The hard constraint is that the available capacity is not less than the minimum safety stock number of sets. The soft constraint is that the deviation between the available capacity and the workstation demand capacity does not exceed a preset threshold. Once the recommended storage locations for all materials in the current set are obtained, the current set is locked, and all recommended storage locations selected for all materials in the current set are occupied in batches. If all storage locations are successfully occupied, the allocation is confirmed and takes effect.

[0004] In one implementation, based on the reservoir layout data, a density clustering algorithm based on a distance threshold is used to divide each reservoir location into multiple homogeneous clusters, and the adjacency relationships between the homogeneous clusters are determined, including: Using the coordinates of the center point of each storage location as input, a density-based clustering algorithm is used, with parameters of a distance threshold of five meters and a minimum of three storage location center points, to divide the storage locations into clusters. Two clusters that are connected at the same level and whose minimum distance between the center points of the storage locations does not exceed ten meters are defined as adjacent clusters.

[0005] In one implementation, the attribute data includes packaging dimensions, appliance dimensions, turnover frequency, and packaging material; the calculation of the complete set volume based on the attribute data, and the dynamic calibration of the height restriction coefficient, include: Based on the packaging dimensions and appliance dimensions of each material in the current set, the maximum number of pieces to be filled in a single appliance is determined by rotation optimization. The appliance volume is then accumulated by rounding up the required number of pieces to obtain the complete volume of a single set. For each material, a turnover frequency correction coefficient is determined based on the material's turnover frequency, and a packaging material correction coefficient is determined based on the material's packaging material. The preset basic height limit coefficient is multiplied by the packaging material correction coefficient and then by the turnover frequency correction coefficient to obtain the material's height limit coefficient.

[0006] In one implementation, the basic configuration data of the production line includes the physical area of ​​the workstation buffer zone and the aisle width; the calculation method for the required capacity of the workstation is as follows: The safety margin coefficient is determined based on the channel width, and the effective area is obtained by multiplying the physical area of ​​the workstation buffer area by the safety margin coefficient. Configure the area ratio of the complete set buffer zone according to the workstation type, and multiply the effective area by the area ratio of the complete set buffer zone to obtain the effective area of ​​the complete set buffer zone. Divide the effective area of ​​the complete set buffer area by the product of the volume and height restriction coefficient of a single complete set to obtain the required capacity of the workstation.

[0007] In one implementation, the basic production line configuration data also includes production cycle time, changeover stabilization time, emergency replenishment response time, and safety stock coefficient; the method for calculating the minimum safety stock unit quantity is as follows: Add the model changeover stabilization time to the emergency replenishment response time to obtain the time; Divide the time by the production cycle time, round the quotient up, and then multiply the rounded result by a safety stock factor to obtain the minimum safety stock number of units.

[0008] In one implementation, selecting recommended storage locations for the current material based on a multi-objective scoring function includes: For each candidate storage location in the candidate storage location set, a distance score is calculated based on the distance from the candidate storage location to the workstation. A capacity matching score is calculated using a four-segment linear function based on the matching deviation between the available capacity of the candidate storage location and the required capacity of the workstation. A frequency score is calculated based on the current material turnover frequency. A matching score is calculated based on the matching cluster information of the storage locations already allocated to the current set or historical similar process routes. The distance score, capacity matching score, frequency score, and kitting score are multiplied by their respective dynamic weights and then summed to obtain the comprehensive score of the candidate storage location; among which, each dynamic weight is adjusted based on exponential weighted moving average and type change impact detection. Based on the homogeneity priority allocation rule, Pareto optimal solution screening is performed in the candidate storage location set to select the candidate storage location with the highest comprehensive score as the recommended storage location.

[0009] In one implementation, it further includes: If the storage location for any material fails to be occupied, all successful storage location occupancy within that set will be revoked, and the status of that set will be rolled back to the unassigned state.

[0010] Secondly, embodiments of the present invention provide a PFEP kit delivery point dynamic allocation system for assembly manufacturing scenarios, which executes the above-described PFEP kit delivery point dynamic allocation method for assembly manufacturing scenarios, including: The data coupling engine is used to acquire basic configuration data of the production line and layout data of the warehouse area. Based on the layout data of the warehouse area, a density clustering algorithm based on distance threshold is used to divide each warehouse location into multiple clusters and determine the adjacency relationship between each cluster. The capacity calculation engine is used to obtain the material list of the current set to be allocated and the attribute data of each material, calculate the complete volume of a single set based on the attribute data, and dynamically calibrate the height restriction coefficient; based on the production line basic configuration data, the complete volume of a single set and the height restriction coefficient, it calculates the required capacity of the workstation and the minimum safety stock number of sets. The optimized matching engine iterates through each material in the material list, obtains the available capacity of each candidate storage location for the current material, filters out a set of candidate storage locations that simultaneously meet hard constraints, kitting constraints, and soft constraints, and then selects the recommended storage location for the current material from the candidate storage location set based on a multi-objective scoring function. The hard constraint is that the available capacity is not less than the minimum safety stock number of sets; the soft constraint is that the deviation between the available capacity and the required capacity of the workstation does not exceed a preset threshold. The transaction control engine is used to lock the current set of materials and batch occupy the recommended storage locations selected for all materials in the current set of materials. If all are successfully occupied, the allocation is confirmed and takes effect.

[0011] Thirdly, embodiments of the present invention provide an electronic device comprising a memory and a processor. The memory and the processor communicate with each other via an internal connection path. The memory stores instructions, and the processor executes the instructions stored in the memory. When the processor executes the instructions stored in the memory, it causes the processor to perform the method described in any of the above embodiments.

[0012] Fourthly, embodiments of the present invention provide a computer-readable storage medium that stores a computer program, wherein when the computer program is run on a computer, the methods in any of the embodiments described above are executed.

[0013] The advantages or beneficial effects of the above technical solutions include at least the following: This invention divides storage locations into kitting clusters and defines adjacency relationships. It combines kitting constraints with a two-way verification of hard and soft constraints based on dynamic workstation demand capacity and available capacity of candidate storage locations to filter the candidate storage location set. This forces materials within the same kitting group to be highly concentrated in physical space, achieving precise capacity matching and avoiding material shortages or waste. Finally, a recommended storage location for the current material is selected based on a multi-objective scoring function. This solution effectively solves the problem of scattered picking paths caused by the traditional single-point mapping of materials to storage locations, significantly shortens the walking distance for kitting delivery, reduces production line waiting time, and thus breaks through the bottleneck restricting assembly efficiency, improving kitting delivery efficiency.

[0014] The above overview is for illustrative purposes only and is not intended to be limiting in any way. In addition to the illustrative aspects, embodiments, and features described above, further aspects, embodiments, and features of the invention will become readily apparent from the accompanying drawings and the following detailed description. Attached Figure Description

[0015] In the accompanying drawings, unless otherwise specified, the same reference numerals throughout the various drawings denote the same or similar parts or elements. These drawings are not necessarily drawn to scale. It should be understood that these drawings depict only some embodiments disclosed in the invention and should not be construed as limiting the scope of the invention.

[0016] Figure 1 This is a flowchart illustrating the PFEP kit delivery point dynamic allocation method for assembly manufacturing scenarios according to the present invention. Figure 2 This is a structural block diagram of an electronic device according to an embodiment of the present invention. Detailed Implementation

[0017] In the following description, only certain exemplary embodiments are briefly described. As those skilled in the art will recognize, the described embodiments can be modified in various ways without departing from the spirit or scope of the invention. Therefore, the drawings and description are considered to be exemplary in nature and not restrictive.

[0018] Example 1 This embodiment provides a method for dynamic allocation of PFEP (Package Delivery Equipment) points in assembly manufacturing scenarios, referencing... Figure 1 As shown, the method specifically includes the following steps: Step S1: Obtain basic configuration data of the production line and layout data of the warehouse area. Based on the layout data of the warehouse area, use a density clustering algorithm based on distance threshold to divide each warehouse location into multiple clusters and determine the adjacency relationship between each cluster.

[0019] This embodiment uses an API gateway to adapt to existing enterprise information systems such as Enterprise Resource Planning (ERP), Manufacturing Execution System (MES), Warehouse Management System (WMS), and Product Lifecycle Management (PLM), enabling the input of basic production line configuration data and warehouse layout data. Specifically, the standardized API gateway adapts to the ERP system to read material master data, supplier information, and inventory ledgers; it adapts to the MES system to subscribe to real-time production line work order status, production cycle time, and process route data via the MQTT protocol; it adapts to the WMS system to query real-time warehouse location status, available capacity, and locking information via RESTful API; and it adapts to the PLM system to obtain data such as unit group BOM relationships, material attributes, and packaging specifications via SOAP Web services.

[0020] In this embodiment, a warehouse layout plan is obtained from the WMS, and warehouse layout data is obtained based on the warehouse layout plan. Specifically, the warehouse layout data refers to a set of physical spatial information describing the storage locations of each material within the assembly line buffer area (or warehouse), including at least the center point coordinates (X, Y) of each storage location in the workshop's two-dimensional plane coordinate system. When the original storage location data is a polygonal area, it is first converted to point coordinates through geometric centroid calculation. The formula for geometric centroid calculation is: X=Σ(x i ) / n, Y=Σ(y i ) / n; Where, x i y i Let x and y be the x and y coordinates of the i-th point; n is the total number of points; X and Y are the calculated centroid coordinates, i.e., the average position of all points.

[0021] Subsequently, a density-based spatial clustering algorithm, with a distance threshold of 5 meters and a minimum number of points (three reservoir center points) required, was used to cluster all reservoir locations. The spatial clustering algorithm automatically identifies density-connected regions and groups reservoir locations that meet the criteria into the same homogeneous cluster. In the spatial clustering algorithm, the conditions for grouping reservoir locations into the same homogeneous cluster are: a reservoir location must contain at least three other reservoir center points (including itself) within a 5-meter radius, and these reservoir locations must be interconnected through density reachability relationships.

[0022] After clustering, the adjacency relationships between each set of clusters are further determined. For any two sets of clusters, the minimum distance between the center points of all storage locations within the two clusters is calculated. If the minimum distance does not exceed 10 meters and the two clusters are located on the same physical layer (i.e., connected on the same layer, without obstacles or across layers), then the two clusters are marked as adjacent clusters. The adjacent cluster information is stored in the form of an adjacency list for subsequent use in set constraint conditions.

[0023] By defining the kitting clusters and determining adjacency relationships, the originally discrete and unorganized storage locations are automatically aggregated into a finite number of kitting clusters based on spatial density. This ensures that the storage locations within each cluster are physically adjacent, providing clear spatial boundaries for the subsequent forced aggregation of materials from the same kit. Through the definition of adjacent clusters (minimum center point distance ≤ 10 meters and connected on the same floor), long-distance cross-regional handling is avoided while maintaining a certain degree of flexibility. Pickers only need to move short distances between adjacent clusters to complete kitting delivery, reducing kitting waiting time and improving production line delivery efficiency.

[0024] In addition, this embodiment obtains basic production line configuration data through a combination of MES system interface and manual configuration. This basic configuration data includes production cycle time, physical area of ​​workstation buffer zones, aisle width, changeover stabilization time, emergency replenishment response time, and safety stock coefficient. Specifically, the production cycle time is read in real-time from the MES system, representing the current assembly line's design cycle time, for example, 12 minutes / unit; the physical area of ​​the workstation buffer zone is entered through the workshop layout management system or by manual measurement; the aisle width is extracted from the WMS warehouse layout data, showing the width of the main aisles surrounding the buffer zone; the changeover stabilization time is calculated based on historical changeover records, with a default value of 4 hours, but can also be manually configured; the emergency replenishment response time is calculated based on historical material delivery data (e.g., the time from triggering a material shortage to the material arriving at the workstation); and the safety stock coefficient is set based on production fluctuation experience, with a default value of 0.15, but can be dynamically adjusted.

[0025] Basic production line configuration data is collected from MES, WMS, and the basic data management module via an API gateway and stored in the production line configuration table for subsequent calculation of workstation capacity requirements and minimum safety stock units. When changes occur in the production line, such as increased cycle time or expanded buffer area, the system supports manual updates or automatic synchronization via MES messages to ensure the real-time accuracy of the calculated parameters.

[0026] Step S2: Obtain the material list and attribute data of the current set to be allocated, calculate the complete volume of a single set based on the attribute data, and dynamically calibrate the height restriction coefficient; calculate the required capacity of the workstation and the minimum safety stock number of sets based on the production line basic configuration data, the complete volume of a single set, and the height restriction coefficient.

[0027] This embodiment obtains the work order information of the currently assigned work sets from the MES system through the API gateway, including the work set identifier and the bill of materials it contains; and obtains attribute data such as packaging dimensions (e.g., length, width, height), equipment dimensions, turnover frequency (e.g., number of times per week), and packaging material (carton, turnover box, or rack) of each material from the PLM system. For example, a certain drive axle work set contains 12 types of materials such as the main reducer gear set and differential gasket.

[0028] After obtaining the packaging dimensions and appliance dimensions for each material, the packaging module filling efficiency is calculated for each material. Specifically, for each material, its packaging dimensions are matched with the internal dimensions of a standard appliance, and the number of material pieces that a single appliance can hold is calculated both in the original orientation and after a 90° horizontal rotation. The larger value is selected as the actual number of pieces that can be filled per appliance for that material. Simultaneously, the filling efficiency (i.e., the ratio of the actual volume occupied by the material to the appliance volume) is calculated to quantify the benefits of rotation optimization.

[0029] It should be noted that this embodiment does not consider vertical placement (i.e., placement with changes in height direction), and the stacking restrictions in the height direction are uniformly controlled by the subsequent height restriction coefficient η.

[0030] Then, using standard fixtures as the smallest space allocation unit, the required number of pieces of the material in a single set is divided by the number of pieces loaded in a single fixture and rounded up to obtain the number of fixtures required for the material. This number is then multiplied by the volume corresponding to a single fixture (or by multiplying the volume of a single material by the actual number of pieces required) to obtain the volume contribution of the material in the set. Finally, the volume contributions of all materials in the current set are summed to obtain the complete set volume V. kit This calculation method improves the space utilization of the equipment through rotation optimization, eliminates waste inside the equipment, and makes the calculation of the kit volume more accurate.

[0031] The following uses two typical materials as examples to illustrate: Taking the main reducer gear set as an example, the packaging dimensions of the main reducer gear set are 420mm in length, 300mm in width, and 250mm in height. The internal dimensions of the standard fixture are 600mm in length, 400mm in width, and 280mm in height.

[0032] Align the packaging with the appliance at 420mm (length) along the 600mm direction and 300mm (width) along the 400mm direction. One piece can be placed along the length of the appliance (600 / 420), one piece along the width (400 / 300), and one piece along the height (280 / 250). A total of 1 × 1 × 1 = 1 piece.

[0033] After rotating the packaging horizontally by 90°, align the 300mm length with the 600mm direction of the appliance, and align the 420mm width with the 400mm direction of the appliance. The length direction floor (600 / 300) = 2 pieces, and the width direction floor (400 / 420) = 0 pieces (cannot accommodate).

[0034] The original orientation allows for one item to be loaded, but after rotating 90°, loading becomes impossible. Therefore, one item per appliance is used. The filling efficiency is calculated as follows: the material packaging volume is 0.42 × 0.3 × 0.25 = 0.0315 m³. 3 The volume of the appliance is 0.6 × 0.4 × 0.28 = 0.0672 m³. 3 The filling efficiency is calculated as (0.0315 × 1) / 0.0672 = 0.46875. This example illustrates that rotation is not always beneficial; by comparing and automatically selecting a better direction, blind rotation can lead to a decrease in filling efficiency.

[0035] To give another example, taking differential gaskets, according to the above calculation and comparison method, in the original orientation, each fixture can hold 18 pieces; after rotating the packaging horizontally by 90°, each fixture can hold 24 pieces. The comparison shows that taking the larger value of 24 pieces as the single-fixture filling quantity for this material increases the filling efficiency from 0.643 to 0.857. Rotation optimization significantly improves the space utilization of the fixtures.

[0036] It should be noted that this embodiment does not consider vertical placement (i.e., placement with changes in height direction), and the stacking restrictions in the height direction are uniformly controlled by the subsequent height restriction coefficient η.

[0037] Then, using standard fixtures as the smallest space allocation unit, the required number of pieces of the material in a single set is divided by the number of pieces loaded in a single fixture and rounded up to obtain the number of fixtures required for the material. This number is then multiplied by the volume corresponding to a single fixture (or by multiplying the volume of a single material by the actual number of pieces required) to obtain the volume contribution of the material in the set. Finally, the volume contributions of all materials in the current set are summed to obtain the complete set volume V. kitThis calculation method improves the space utilization of the equipment through rotation optimization, eliminates waste inside the equipment, and makes the calculation of the kit volume more accurate.

[0038] At the same time, for each material, a height limit coefficient is dynamically calibrated. Specifically, a turnover frequency correction coefficient is determined based on the turnover frequency of the material, and a packaging material correction coefficient is determined based on the packaging material of the material. The preset basic height limit coefficient is multiplied by the packaging material correction coefficient and then by the turnover frequency correction coefficient to obtain the height limit coefficient of the material.

[0039] In this embodiment, the basic height limitation coefficient is set to 0.80. The packaging material correction coefficient is determined based on the packaging material. For example, cardboard boxes are made of softer material with lower pressure resistance, requiring a reduced stacking height; therefore, the packaging material correction coefficient for cardboard boxes is 0.95. Turnover boxes are made of harder material with moderate pressure resistance, allowing for a more appropriate increase in stacking height; therefore, the packaging material correction coefficient for turnover boxes is 1.05. Material racks are metal structures with strong pressure resistance; therefore, their corresponding packaging material correction coefficient is 1.15.

[0040] The frequency correction factor is determined based on the turnover frequency. For example, high-frequency materials with a turnover of more than 50 times per week are given a factor of 0.90, medium-frequency materials with a turnover of 20 to 50 times per week are given a factor of 0.95, and low-frequency materials with a turnover of less than 20 times per week are given a factor of 1.00.

[0041] By multiplying the preset basic height limit coefficient, packaging material correction coefficient, and turnover frequency correction coefficient, the height limit coefficient η of the material can be obtained.

[0042] Subsequently, the required capacity of workstations is calculated based on the basic configuration data of the production line, specifically: Determine the safety margin factor K based on the channel width. safety According to formula S effective =S physical ×K safety The effective area is obtained by multiplying the physical area of ​​the workstation buffer area by the safety margin factor. In this embodiment, when the aisle width is greater than or equal to 1.5 meters, the safety margin factor is 0.85; when the aisle width is between 1.0 meter and 1.5 meters, the safety margin factor is 0.80; and when the aisle width is less than 1.0 meter, the safety margin factor is 0.75. Configure the kittingBufferRatio based on the workstation type, according to the formula S. kitting =S effective×kittingBufferRatio multiplies the effective area by the kitting buffer area ratio to obtain the effective area of ​​the kitting buffer. The kitting buffer area ratio refers to the proportion of the workstation buffer area specifically used for storing complete kits of materials (i.e., kitted materials) to the total effective area. The value of this ratio varies depending on the workstation type: assembly workstations typically require more kitting buffer space, so the ratio is between 0.5 and 0.7; processing workstations have relatively higher single-material buffer requirements, so the kitting buffer ratio is between 0.3 and 0.5. Specific values ​​need to be configured after workstation operation analysis and measurement. The effective area S of the set buffer area kitting Divide by the complete set volume V kit The product of the height restriction factor η, i.e., C required =S kitting / (V kit ×η) to obtain the required capacity of the workstation.

[0043] The required capacity of a workstation refers to the number of complete sets of products that the workstation buffer area needs to store to ensure the continuous operation of the assembly line without interrupting production. The required capacity is calculated based on factors such as the effective area of ​​the workstation buffer area and the volume occupied by a single set of materials, reflecting how many complete sets of materials the workstation buffer area can theoretically hold. This value is used to compare with the available capacity of candidate storage locations to determine whether the storage capacity meets the production needs of the workstation.

[0044] In addition, the minimum safety stock quantity is calculated using a formula, the expression of which is:

[0045] Specifically: The transformation stabilization time T switch Emergency replenishment response time T emergency Add them together to get the time; where the transformation settling time T switch The default time is 4 hours, and the emergency replenishment response time is T. emergency The default setting is 30 minutes. Divide the time by the production cycle T akt The resulting quotient is rounded up, and then the rounded result is multiplied by a safety stock factor λ to obtain the minimum number of safety stock units.

[0046] The minimum safety stock refers to the minimum material reserve that the workstation buffer area must maintain to cope with fluctuations such as production changeovers and emergency replenishment. This value serves as a hard constraint, requiring that the available capacity of candidate storage locations must not be less than this value to ensure that the production line will not stop due to material shortages, thereby guaranteeing the continuity and stability of production.

[0047] Step S3: Iterate through each material in the material list, obtain the available capacity of each candidate storage location for the current material, and filter out the set of candidate storage locations that simultaneously meet the hard constraint, kitting constraint, and soft constraint conditions.

[0048] It's important to explain that the available capacity of a candidate storage location refers to the number of sets of materials that location can currently hold, expressed in units of sets. The available capacity is obtained as follows: It can be directly read from the WMS system's real-time interface or cached data to determine the maximum designed capacity of each storage location. Alternatively, it can be calculated based on the storage location volume, the volume occupied by a single set of materials, and height restrictions to determine the theoretical maximum number of sets that can be stored. The maximum designed capacity is then subtracted from the currently occupied sets (i.e., the amount of allocated but not yet released materials). The storage location's locking status, such as whether it is temporarily locked by other transactions, is then considered. The final result is the number of sets currently available for new allocation; this number is the available capacity. The available capacity is dynamically changing; the system retrieves the latest value from the WMS before each allocation.

[0049] In this embodiment, the material list of the current set to be allocated is obtained, and each material is traversed in turn. Taking the first material as an example, all unlocked and idle candidate storage locations and their available capacity are first obtained from the WMS. Then, hard constraints, kitting constraints and soft constraints are performed on each candidate storage location in turn to obtain the set of candidate storage locations that meet the hard constraints, kitting constraints and soft constraints.

[0050] The hard constraint is that the available capacity of the candidate storage location must not be less than the minimum safety stock number of units. Specifically, it is determined whether the available capacity of the candidate storage location is not less than the pre-calculated minimum safety stock number of units. If the available capacity of the candidate storage location is less than this value, it will be directly eliminated.

[0051] The kitting constraint is that if there are already allocated materials in the current kitting group, then the kitting cluster to which the candidate storage location belongs must be the same as or adjacent to the kitting cluster to which any allocated material belongs. Specifically, it is determined whether the kitting cluster to which the candidate storage location belongs is the same as or adjacent to the kitting cluster to which any allocated material belongs in the current kitting group. If there are no allocated materials in the current kitting group, this condition is not restricted; if there are already allocated materials, then the kitting cluster to which the candidate storage location belongs must be the same as or adjacent to the kitting cluster to which any allocated material belongs (i.e., the minimum distance between the center points of the storage locations in the two clusters does not exceed ten meters and they are connected on the same floor). Candidate storage locations that do not meet this condition are eliminated.

[0052] The soft constraint is that the deviation between the available capacity and the required capacity of the workstation does not exceed a preset threshold. Specifically, the deviation between the available capacity and the required capacity of the workstation is calculated by dividing the absolute value of the difference between the available capacity and the required capacity of the workstation by the required capacity of the workstation and then multiplying by 100%. The expression for this deviation is: Deviation δ = |C available -C required | / C required ×100%, where C available C represents the available capacity of the candidate storage location. required The required capacity for each workstation. If the matching deviation does not exceed a preset threshold (e.g., 20%), the soft constraint condition is met; otherwise, it is eliminated.

[0053] Candidate storage locations that simultaneously meet the aforementioned hard constraints, kitting hard constraints, and soft constraints are included in the candidate storage location set for further optimization by the multi-objective scoring function. This embodiment ensures that storage locations meet the minimum safety stock level at each workstation through hard constraints, preventing line stoppages due to material shortages; the kitting hard constraint forces new materials to spatially cluster with already allocated materials (in the same or adjacent clusters), shortening the walking distance for subsequent kitting picking; and the soft constraint avoids extreme cases of excessively large or small capacity, improving buffer space utilization while meeting production needs. This three-layer screening mechanism ensures the feasibility of candidate storage locations from three dimensions: safety stock, spatial clustering, and capacity matching, providing a high-quality foundation for the generation of the final recommended storage locations.

[0054] Step S4: Select the recommended storage location for the current material from the candidate storage location set based on a multi-objective scoring function.

[0055] After obtaining the candidate storage location set, for each candidate storage location in the candidate storage location set, a score is calculated in four dimensions: distance score, capacity matching score, frequency score, and kit matching score.

[0056] The distance score is calculated based on the Manhattan or Euclidean distance from the candidate storage location to the workstation (or the designated picking point at the workstation). A smaller distance results in a higher score. Closer distances mean less time for picking personnel or AGVs to transport materials, helping to reduce travel time and improve kitting efficiency. This score encourages the allocation of materials to storage locations close to the workstation.

[0057] The capacity matching score is calculated based on the deviation δ between the available capacity of the candidate storage location and the required capacity of the workstation. Specifically, after calculating the deviation δ, a four-segment linear function is used to calculate the score according to the interval in which the deviation δ falls. When δ ≤ 10%, the capacity matching score C is... score The score is 100; when 10% < δ ≤ 30%, the capacity matching score is C. score=100 -2×(δ-10), and 60 points are obtained when δ=30%; When 30% < δ ≤ 50%, the capacity matching score Cscore = 60 - 1.5 × (δ - 30), with 30 points awarded when δ = 50%; when δ > 50%, Cscore = 60 - 1.5 × (δ - 30). score =max(30-0.5×(δ-50), 10), with a minimum base score of 10. This piecewise function prevents a sharp drop in score due to capacity matching deviation within a certain range, thus avoiding the complete exclusion of candidate storage locations with large capacity but superior distance or matching scores, preserving the trade-off space for multi-objective optimization. The capacity matching score reflects the degree of fit between the storage location's capacity and the actual demand of the workstation. A small deviation between the available capacity and the workstation's required capacity may indicate low storage location utilization or an overestimation of demand; a large deviation may lead to wasted space or frequent material shortages. The four-segment linear function allows for a certain range of deviations, avoiding excessive penalties for small deviations, while still allowing storage locations with exceptionally large capacity but superior performance in other dimensions to be selected.

[0058] Frequency scores are determined based on material turnover frequency; turnover frequency refers to the number of times a material is retrieved from its storage location and put into production per unit of time. Higher turnover frequency results in a higher score. Turnover frequency can be normalized to a percentage system, for example, high-frequency materials correspond to high scores, and low-frequency materials correspond to low scores. Common mapping methods include linear normalization or quantiles based on historical distributions. Because high-frequency materials require frequent inbound and outbound movements, placing them in distant or difficult-to-access locations increases handling workload. Therefore, high-frequency materials should receive higher frequency scores, making them more likely to be assigned to storage locations close to workstations or easily accessible locations, thereby optimizing overall logistics efficiency.

[0059] The calculation of kit matching score is divided into two cases. One is for materials that are not the first, which is calculated directly based on the kit cluster information of the assigned storage location in the current kit group. Specifically, if the kit cluster of the candidate storage location is the same as or adjacent to the cluster of the assigned material, a high score is assigned; otherwise, a low score is assigned. For example, the same cluster gets 100 points, adjacent clusters get 70 points, and others get 0 points.

[0060] Another approach is to target the first material. For the first material in the current set to be allocated, since there is no information on the already allocated storage locations for this set, a predictive occupancy correction based on historically similar process routes is used to calculate the kitting score. The specific steps are as follows: The first step is to calculate the similarity of the process routes between the current set of equipment and the historical set of equipment. Specifically, this involves obtaining the process route of the current set of equipment, which is the set M of all material types included in the current set of equipment to be assigned. a And retrieve all allocated unit sets from the historical database to obtain the set M of material types contained in a historically allocated unit set.b The similarity between the set of material types in the current set to be allocated and the set of material types in a previously allocated set is calculated using the following formula: Sim(M a M b )=|M a ∩M b | / |M a ∪M b |; The process route similarity is calculated by dividing the number of material types in the intersection of two sets by the number of material types in the union of the two sets. This ratio ranges from 0 to 1; a higher ratio indicates greater similarity between the two processes. If Sim is greater than 0.6, the two processes are considered similar.

[0061] The second step is to screen similar historical storage sets and calculate their storage location cluster centers. Specifically, historical storage sets with a similarity greater than 0.6 are selected as valid references for transferable experience. If no historical storage sets meet the criteria, the cold start baseline score is directly used as the initial matching score.

[0062] For each similar historical storage set, obtain the coordinates (x, y) of its allocated storage locations. i y i and the corresponding material turnover frequency f i According to the formula Center(X, Y) = Σ(f i ×(x i y i )) / Σf i The weighted average centroid, also known as the cluster center, is calculated by multiplying the x-coordinates of all storage locations by their corresponding turnover frequencies, summing the results, and then dividing by the total turnover frequencies to obtain the x-coordinate of the cluster center. The y-coordinate is calculated similarly. This centroid represents the spatial center of gravity of the materials in that historical storage unit.

[0063] The third step is to calculate the workstation correlation degree of the candidate storage location. Specifically, for the current candidate storage location, calculate its Manhattan distance to the workstation center, and divide the Manhattan distance by the maximum span of the storage area to obtain the normalized distance. Assume the Manhattan distance from the candidate storage location to the workstation is d, and the maximum span of the storage area is D. max According to formula R station =1-(d / D max The workstation correlation degree is calculated, and its value is between 0 and 1. The closer the value is to 1, the closer the candidate warehouse location is to the workstation.

[0064] The fourth step is to calculate the kitting score. Specifically, a cold start baseline score of 62 is used as the base score. This score is obtained through simulation calibration. Specifically, in 100 random scenarios, different baseline scores are tested, and the median distance between the initial material allocation position and the actual position after system stabilization is calculated. The baseline score with the smallest median distance is selected, resulting in 62. This score ensures that the initial material allocation during the cold start phase is closest to the steady-state result.

[0065] Combining the similarity of the process route and the correlation of workstations, according to formula K score =62+(Sim-0.6)×20+R station ×10 Calculate the matching score K score If multiple similar historical work sets exist, the scores of each set can be weighted and averaged according to the sum of turnover frequency or similarity value to obtain the final score. Alternatively, the scores can be calculated separately and then weighted and averaged according to the sum of turnover frequency or similarity value of each set to obtain the final matching score. For example, suppose there is a historical work set that has already been assigned, whose process route includes three materials: A, B, and C. The coordinates of the three assigned storage locations are (10, 20), (12, 22), and (11, 21), corresponding to turnover frequencies of 30, 45, and 25 times for the corresponding materials. The process route of the new work set to be assigned currently includes four materials: A, B, C, and E. The intersection of the material sets of the new work set and the historical work set is {A, B, C}, and the union is {A, B, C, E}. Therefore, the process route similarity is three-quarters, or 0.75, which is greater than the set similarity threshold of 0.6, satisfying the knowledge transfer condition. Based on the coordinates of the historically assigned storage locations and their turnover frequency, the weighted average cluster center is calculated: the horizontal axis is (30×10+45×12+25×11) divided by the total frequency of 100, which equals 11.35; the vertical axis is (30×20+45×22+25×21) divided by 100, which equals 21.05. For the current candidate storage location, it is assumed that the Manhattan distance to the workstation, after normalization, yields a workstation correlation degree of 0.7. Substituting the above parameters into the matching score calculation formula K... score =62+(0.75-0.6)×20+0.7×10=72, the final score is 72 points. This score is the matching score of the first material in the new set at this candidate storage location.

[0066] After determining the distance score, capacity matching score, frequency score, and kitting score, the comprehensive score of the candidate storage location is calculated according to the multi-objective scoring function:

[0067] In the formula, The distance score for candidate location j; The capacity matching score for candidate storage location j; The frequency score of candidate location j; The score for the homogeneous matching of candidate library position j; w1(t), w2(t), w3(t), and w4(t) correspond to the dynamic weights of distance score, capacity matching score, frequency score, and matching score, respectively. The dynamic weights can be adjusted based on exponentially weighted moving average and type change impact detection.

[0068] It's important to explain that the exponentially weighted moving average is a smoothing algorithm used to gradually update weights based on historical weight values ​​and current target values ​​(such as the performance of each score), ensuring that weight changes are neither too drastic nor too inaccurate, while still reflecting recent trends. Changeover shock detection, on the other hand, occurs when the system detects a model change on the production line (e.g., switching from product A to product B), and this state persists for more than the steady-state detection window T. stable At that time, it was determined to be a type-change impact event.

[0069] During normal production, each dynamic weight is updated smoothly using an exponentially weighted moving average, as shown in the formula: w i (t)=β×score i (t)+(1-β)×w i (t-1) Where β = 0.1; w i (t) represents the smoothed value of the i-th weight at the current time t; w i (t-1) represents the smoothed value of the i-th weight at the previous time t-1; score i (t) represents the observation value of the i-th score (such as distance score, capacity matching score, etc.) at the current time.

[0070] When a trigger event occurs, the system switches to accelerated convergence mode, and the update formula is changed to: w i (t)=w i (t-1)+α×(target wi -w i (t-1)); The changeover response acceleration coefficient α is set to 0.5, so that the weights quickly approach the target weights corresponding to the new production line. wi The target weight is based on the production cycle time T of the new product. akt Preset: If T aktFor matches lasting less than 15 minutes, a distance-first strategy is used, with a weight vector of [0.40, 0.20, 0.25, 0.15] (corresponding to distance, capacity, frequency, and matching score, respectively). If 15≤T akt For intervals of less than 30 minutes, a balanced strategy is adopted, with a weight vector of [0.30, 0.25, 0.25, 0.20]. If T akt For a time of ≥30 minutes, a capacity-first strategy is adopted, with a weight vector of [0.20, 0.30, 0.25, 0.25].

[0071] Through the above mechanism, the system can smoothly track weight changes during long-term operation and quickly adapt during product switching, so that the multi-objective scoring function always fits the current production status. For example, after changing models, the weight of the kit matching score can be appropriately increased to strengthen material aggregation, or the weight of the capacity matching score can be adjusted according to the cycle time to avoid material shortage.

[0072] After calculating the comprehensive score of each storage location in the candidate storage location set, the Pareto optimal solution selection is further performed. This embodiment introduces a homogeneous matching priority rule as a supplement to the traditional Pareto rule. The homogeneous matching priority rule states that if the homogeneous matching score K of storage location a is... score If the price of storage location b is more than 15 points higher than storage location b, and the cluster where a is located already contains other materials in the current set, then a will directly have priority to control b.

[0073] In this embodiment, based on the homogeneity priority allocation rule, the Pareto optimal solution screening is performed in the candidate storage location set to select the candidate storage location with the highest comprehensive score as the recommended storage location. The specific method is as follows: Each pair of storage locations in the candidate storage location set is compared sequentially. For any two candidate storage locations, it is first determined whether the matching priority domination condition is met; if it is, the dominated storage location is directly removed from the candidate set. If the matching priority domination condition is not met, the comparison is performed according to the traditional Pareto rule, and the dominated storage location is eliminated. After multiple rounds of comparison, the remaining storage locations not dominated by any other storage location constitute the Pareto front (i.e., the non-dominated solution set).

[0074] Finally, the storage location with the highest overall score from the Pareto front is selected as the recommended storage location for the current material. This rule ensures that storage locations with significant advantages in completeness and partial aggregation are given priority, thereby guiding subsequent materials for the same set to concentrate in areas where materials are already aggregated, further strengthening the spatial aggregation of completeness delivery.

[0075] It should be explained that the traditional Pareto selection rule is that if storage location A is not inferior to storage location B in all objectives (distance score, capacity matching score, frequency score, and kitting score), and is superior to storage location B in at least one objective, then A dominates B, and B is eliminated.

[0076] Step S5: After obtaining the recommended storage locations for all materials in the current set, lock the current set and batch occupy the recommended storage locations selected for all materials in the current set. If all are successfully occupied, confirm the allocation and make it effective.

[0077] After selecting recommended storage locations for all materials within the current set, the commit operation is executed using a set-level atomic transaction. It's important to understand that a set-level atomic transaction treats the allocation of storage locations for all materials within a set (i.e., all materials required for a complete product) as a single, indivisible unit of execution. Specifically, first, the set's status is switched from "calculation complete" to "preview," and all materials in that set are locked. At this point, no other concurrent transaction can modify or occupy the selected recommended storage locations for these materials.

[0078] Then, the Saga coordinator partitions the materials within a kit group according to the kit cluster boundary: it iterates through the materials in the kit group, classifies them according to the kit cluster to which each material's recommended storage location belongs, and generates a mapping table (map). <clusterId, list <material>Materials within the same cluster are divided into shards. The purpose of sharding is to enable parallel submission, local fault tolerance, and reduce lock granularity. The system executes storage location occupancy commands in parallel for each shard.

[0079] If all material storage locations within a segment are successfully occupied, the status of the set will be advanced to "KITTING_LOCKED", and finally confirmed as "COMMITTED", the allocation will take effect, and the storage capacity will be reduced accordingly.

[0080] If, during the occupancy process, any material (such as an adjustment shim in a shard) fails to be occupied due to insufficient capacity, concurrency conflicts, or system anomalies, the set-level compensation mechanism is immediately triggered: the Saga coordinator executes compensation transactions for all successfully committed shards (such as shard 1 and shard 2), rolling back the material status in these shards to "UNCONFIGURED" and releasing the occupied storage space capacity. Ultimately, all materials within the set group are restored to the UNCONFIGURED state, achieving "all or nothing" at the business semantic level and preventing the generation of "half-set" dirty configurations.

[0081] After the transaction is completed, the allocation result is synchronized to the warehouse management system database through a dual-write consistency mechanism: on the one hand, idempotency is guaranteed by using the transaction ID and asynchronous backfilling is achieved through the Kafka message queue; on the other hand, the client is pushed through Redis Pub / Sub broadcast and WebSocket, with the end-to-end latency controlled within 500 milliseconds, ensuring that operators or upper-level systems can know the allocation result in real time.

[0082] In addition, if the transaction commit fails due to insufficient overall storage capacity, the system automatically triggers a two-level exception degradation strategy: Level 1 "Expansion Recommendation": The system filters adjacent clusters whose available capacity is not less than the minimum safety stock number of units N. min Candidate storage locations. For each cross-cluster candidate storage location, the distance score is adjusted according to the following formula: D score =D score ×(1+0.3×adjacentPenalty) The cross-cluster penalty coefficient, adjacentPenalty, is 0.3. The expected composite score (combining capacity matching, frequency, and homogeneous matching scores) is recalculated, and the top three scores are ranked from highest to lowest for recommendation. The predicted D is also output. score With K score Changes (e.g., the difference between the original scheme and the cross-cluster scheme).

[0083] The second level, "Set Splitting Suggestion," states that when expansion recommendations still cannot meet the requirements, the system will split the current set group into 2 to 3 sub-sets. There are two splitting methods: 1) Splitting based on physical proximity, concentrating materials in the same set cluster or adjacent clusters within each sub-set; 2) Splitting based on assembly sequence, allocating materials assembled earlier to nearby storage locations and later materials to slightly more distant locations using a time buffer. Each sub-set must still meet set constraints. The splitting plan is pushed to management for confirmation before execution.

[0084] In another embodiment, the system has a built-in three-level cold start detection mechanism. Specifically, when the data accumulation is less than 7 days or the number of allocated sets of samples is less than 50, it is determined to be in Level 1 full cold start state; when the data accumulation is between 7 and 30 days or the number of samples is between 50 and 200 sets, it is in Level 2 partial cold start state; when the data accumulation exceeds 30 days and the number of samples exceeds 200 sets, it enters Level 3 steady-state operation state.

[0085] During the Level 1 complete cold start phase, the system uses default parameters, namely the changeover settling time T. switch With a timeframe of 4 hours, a safety stock factor λ of 0.15, and an emergency replenishment response time T... emergency The default time is 30 minutes, and the cold start baseline score is 62 points.

[0086] The system triggers an online parameter calibration every time it completes the processing of 20 new sets of equipment, which is the changeover stabilization time T. switch The system is recalibrated based on the stabilization time recorded in historical replacement records; the safety stock factor λ is dynamically adjusted based on the coefficient of variation of material consumption frequency. The coefficient of variation refers to the variation of material consumption frequency (or turnover frequency), calculated as: Coefficient of Variation = Standard Deviation / Mean. The system statistically analyzes the turnover frequency of each material in all completed unit sets and calculates the ratio of its standard deviation to its mean. If the coefficient of variation is large (indicating large frequency fluctuations), the safety stock factor λ is appropriately increased; if the coefficient of variation is small, the safety stock factor λ can be decreased. Specific adjustment rules can be fitted using historical data.

[0087] For extreme cold start scenarios (i.e., situations where there is no historical data and default parameters cannot be obtained), simulation training data is automatically generated based on production line cycle time, BOM (Bill of Materials), and warehouse layout. This ensures that the system can obtain reasonable initial parameter configurations such as changeover stabilization time, safety stock coefficient, and cold start baseline score before formal production. This simulation data generation strategy eliminates the need for manual experience-based parameter tuning when launching new production lines or new products, achieving a closed loop from "manual assignment" to "data-driven self-calibration."

[0088] Example 2 This embodiment provides a dynamic allocation system for PFEP (Pre-Failure Toll Collection) kit delivery points in assembly manufacturing scenarios, executing the dynamic allocation method for PFEP kit delivery points in assembly manufacturing scenarios as described in Embodiment 1. The system specifically includes: The data coupling engine is used to acquire basic configuration data of the production line and layout data of the warehouse area. Based on the layout data of the warehouse area, a density clustering algorithm based on distance threshold is used to divide each warehouse location into multiple clusters and determine the adjacency relationship between each cluster. The capacity calculation engine is used to obtain the material list of the current set to be allocated and the attribute data of each material, calculate the complete volume of a single set based on the attribute data, and dynamically calibrate the height restriction coefficient; based on the production line basic configuration data, the complete volume of a single set and the height restriction coefficient, it calculates the required capacity of the workstation and the minimum safety stock number of sets. The optimized matching engine iterates through each material in the material list, obtains the available capacity of each candidate storage location for the current material, filters out a set of candidate storage locations that simultaneously meet hard constraints, kitting constraints, and soft constraints, and then selects the recommended storage location for the current material from the candidate storage location set based on a multi-objective scoring function. The hard constraint is that the available capacity is not less than the minimum safety stock number of sets; the soft constraint is that the deviation between the available capacity and the required capacity of the workstation does not exceed a preset threshold. The transaction control engine is used to lock the current set of materials and batch occupy the recommended storage locations selected for all materials in the current set of materials. If all are successfully occupied, the allocation is confirmed and takes effect.

[0089] It should be noted that the functions of each module in this embodiment have been disclosed in the method of Embodiment 1, and will not be described again here.

[0090] This embodiment of the system was tested at workstation J23 of the drive axle assembly line of a certain construction machinery company from October to December 2025. The test sample size was 1850 sets, involving 320 types of materials in 8 sets. The trial operation achieved the following results: 1. Improved picking efficiency By dividing storage locations into kitting clusters and defining adjacent clusters, kitting constraints force allocated materials within the same kit group to be located in the same or adjacent clusters as candidate storage locations, resulting in a high degree of physical concentration of all materials in that kit group. Compared to the traditional single-point mapping of materials to storage locations, picking personnel do not need to walk long distances across areas, thus reducing the average picking path by more than 60%.

[0091] 2. Improved delivery on-time rate The minimum safety stock unit count is calculated based on factors such as changeover stabilization time, production cycle time, and safety stock coefficient, and used as a hard constraint to ensure that the available capacity of candidate storage locations is not lower than this threshold, thereby avoiding material shortages due to insufficient storage capacity. Simultaneously, unit-level atomic transactions ensure that all materials are configured successfully on the first attempt, eliminating delays caused by missing materials, thus significantly improving the on-time delivery rate of complete sets.

[0092] 3. Improved space utilization DBSCAN clustering allows storage locations to naturally aggregate according to spatial density, avoiding dead zones caused by material dispersion; the calculation of workstation capacity requirements introduces aisle safety margin coefficients and height restriction coefficients (material-frequency dual factors), making the effective area estimation of the buffer zone more accurate; soft constraints avoid extreme allocations of excessively large or small capacities. These methods collectively increase the number of storage units per unit area.

[0093] 4. Shortened changeover response time The product model changeover impact detection mechanism can automatically identify product model switching and call the changeover option re-update formula (α=0.5 to accelerate convergence). At the same time, it resets the target weight vector (distance priority / balance / capacity priority) according to the new production line cycle time. During this process, no manual replanning is required, and the location recalculation of 320 materials can be completed in 1.8 hours, while the traditional manual mode takes 72 hours.

[0094] 5. Concurrency safety In the pre-configuration state, all materials in the entire set are locked, and idempotency is guaranteed through transaction IDs when Saga segments are submitted in parallel, avoiding dirty data caused by multiple users modifying the same storage location or the same material configuration at the same time. With 50 planners operating concurrently, the average daily conflict events decreased from 23 to 0.3.

[0095] 6. Automatic exception resolution A two-tiered anomaly degradation strategy was adopted: the first tier, "Expansion Recommendation," automatically searches for available storage locations in adjacent clusters and calculates the distance score after penalty; the second tier, "Set Splitting Suggestion," splits the set group into sub-sets based on physical proximity or assembly sequence and allocates them independently. During the trial operation, all 17 set constraint conflict anomalies were automatically resolved within 25 minutes, avoiding production line downtime.

[0096] Example 3 This embodiment provides an electronic device. Figure 2 A structural block diagram of an electronic device according to an embodiment of the present invention is shown. Figure 2 As shown, the electronic device includes a memory 100 and a processor 200. The memory 100 stores a computer program that can run on the processor 200. When the processor 200 executes the computer program, it implements the PFEP kitting delivery point dynamic allocation method for assembly manufacturing scenarios described in the above embodiments. The number of memories 100 and processors 200 can be one or more.

[0097] The electronic device also includes: The communication interface 300 is used to communicate with external devices and perform data exchange and transmission.

[0098] If the memory 100, processor 200, and communication interface 300 are implemented independently, they can be interconnected via a bus to communicate with each other. This bus can be an Industry Standard Architecture (ISA) bus, a Peripheral Component Interconnect (PCI) bus, or an Extended Industry Standard Architecture (EISA) bus, etc. This bus can be divided into address bus, data bus, control bus, etc.

[0099] Optionally, in a specific implementation, if the memory 100, processor 200, and communication interface 300 are integrated on a single chip, then the memory 100, processor 200, and communication interface 300 can communicate with each other through an internal interface.

[0100] This invention provides a computer-readable storage medium storing a computer program that, when executed by a processor, implements the method provided in this invention.

[0101] This invention also provides a chip, which includes a processor for calling and executing instructions stored in a memory, causing a communication device on which the chip is installed to perform the method provided in this invention.

[0102] This invention also provides a chip, including: an input interface, an output interface, a processor, and a memory. The input interface, output interface, processor, and memory are connected through an internal connection path. The processor is used to execute code in the memory. When the code is executed, the processor is used to execute the method provided in this invention.

[0103] It should be understood that the aforementioned processor can be a Central Processing Unit (CPU), or other general-purpose processors, digital signal processors (DSPs), application-specific integrated circuits (ASICs), field-programmable gate arrays (FPGAs), or other programmable logic devices, discrete gate or transistor logic devices, discrete hardware components, etc. General-purpose processors can be microprocessors or any conventional processor. It is worth noting that the processor can be a processor supporting the Advanced Reduced Instruction Set Computing (RISC) machine (ARM) architecture.

[0104] Further, optionally, the aforementioned memory may include read-only memory and random access memory, and may also include non-volatile random access memory. The memory may be volatile or non-volatile, or may include both. Non-volatile memory may include read-only memory (ROM), programmable read-only memory (PROM), erasable programmable read-only memory (EPROM), electrically erasable programmable read-only memory (EEPROM), or flash memory. Volatile memory may include random access memory (RAM), which serves as an external cache. Many forms of RAM are available by way of example, but not limitation. Examples include static random access memory (SRAM), dynamic random access memory (DRAM), synchronous dynamic random access memory (SDRAM), double data rate synchronous dynamic random access memory (DDR SDRAM), enhanced synchronous dynamic random access memory (ESDRAM), synchronous linked dynamic random access memory (SLDRAM), and direct rambus RAM (DR RAM).

[0105] In the above embodiments, implementation can be achieved, in whole or in part, by software, hardware, firmware, or any combination thereof. When implemented in software, it can be implemented, in whole or in part, as a computer program product. A computer program product includes one or more computer instructions. When the computer program instructions are loaded and executed on a computer, all or part of the flow or function according to the present invention is generated. The computer can be a general-purpose computer, a special-purpose computer, a computer network, or other programmable device. The computer instructions can be stored in a computer-readable storage medium or transferred from one computer-readable storage medium to another.

[0106] In the description of this specification, references to terms such as "one embodiment," "some embodiments," "example," "specific example," or "some examples," etc., indicate that a specific feature, structure, material, or characteristic described in connection with that embodiment or example is included in at least one embodiment or example of the invention. Furthermore, the specific features, structures, materials, or characteristics described may be combined in any suitable manner in one or more embodiments or examples. Moreover, without contradiction, those skilled in the art can combine and integrate the different embodiments or examples described in this specification, as well as the features of those different embodiments or examples.

[0107] Furthermore, the terms "first" and "second" are used for descriptive purposes only and should not be construed as indicating or implying relative importance or implicitly specifying the number of technical features indicated. Thus, a feature defined as "first" or "second" may explicitly or implicitly include at least one of that feature. In the description of this invention, "a plurality of" means two or more, unless otherwise explicitly specified.

[0108] The above description is merely a specific embodiment of the present invention, but the scope of protection of the present invention is not limited thereto. Any person skilled in the art can easily conceive of various variations or substitutions within the technical scope disclosed in the present invention, and these should all be included within the scope of protection of the present invention. Therefore, the scope of protection of the present invention should be determined by the scope of the claims.< / material>

Claims

1. A method for dynamic allocation of PFEP (Package Delivery Equipment) fulfillment points for assembly manufacturing scenarios, characterized in that, include: Obtain basic configuration data of the production line and layout data of the warehouse area. Based on the warehouse layout data, use a density clustering algorithm based on a distance threshold to divide each warehouse location into multiple clusters and determine the adjacency relationship between each cluster. Obtain the material list and attribute data of the current set of units to be allocated; calculate the complete volume of a single set based on the attribute data; and dynamically calibrate the height restriction coefficient; calculate the required capacity of the workstation and the minimum safety stock number of sets based on the basic configuration data of the production line, the complete volume of a single set, and the height restriction coefficient. Based on the material list, each material is traversed. For the current material, the available capacity of each candidate storage location is obtained. A set of candidate storage locations that simultaneously meet the hard constraint, the kitting constraint, and the soft constraint is selected. From the set of candidate storage locations, a recommended storage location for the current material is selected based on a multi-objective scoring function. The hard constraint is that the available capacity is not less than the minimum safety stock unit set. The soft constraint is that the deviation between the available capacity and the workstation demand capacity does not exceed a preset threshold. Once the recommended storage locations for all materials in the current set are obtained, the current set is locked, and all recommended storage locations selected for all materials in the current set are occupied in batches. If all storage locations are successfully occupied, the allocation is confirmed and takes effect.

2. The method for dynamic allocation of PFEP kit delivery points for assembly manufacturing scenarios according to claim 1, characterized in that, Based on the reservoir layout data, a density clustering algorithm based on a distance threshold is used to divide each reservoir location into multiple homogeneous clusters, and the adjacency relationships between each homogeneous cluster are determined, including: Using the coordinates of the center point of each storage location as input, a density-based clustering algorithm is used, with parameters of a distance threshold of five meters and a minimum of three storage location center points, to divide the storage locations into clusters. Two clusters that are connected at the same level and whose minimum distance between the center points of the storage locations does not exceed ten meters are defined as adjacent clusters.

3. The method for dynamic allocation of PFEP kit delivery points for assembly manufacturing scenarios according to claim 1, characterized in that, The attribute data includes packaging dimensions, appliance dimensions, turnover frequency, and packaging material; the volume of a single complete set is calculated based on the attribute data, and the height restriction coefficient is dynamically calibrated, including: Based on the packaging dimensions and appliance dimensions of each material in the current set, the maximum number of pieces to be filled in a single appliance is determined by rotation optimization. The appliance volume is then accumulated by rounding up the required number of pieces to obtain the complete volume of a single set. For each material, a turnover frequency correction coefficient is determined based on the turnover frequency of the material, and a packaging material correction coefficient is determined based on the packaging material of the material. The preset basic height limit coefficient is multiplied by the packaging material correction coefficient and then by the turnover frequency correction coefficient to obtain the height limit coefficient of the material.

4. The method for dynamic allocation of PFEP kit delivery points for assembly manufacturing scenarios according to claim 1, characterized in that, The basic configuration data of the production line includes the physical area of ​​the workstation buffer zone and the width of the passageway; the calculation method for the required capacity of the workstation is as follows: The safety margin coefficient is determined based on the channel width, and the effective area is obtained by multiplying the physical area of ​​the workstation buffer area by the safety margin coefficient. Configure the area ratio of the complete set of buffer zones according to the workstation type, and multiply the effective area by the area ratio of the complete set of buffer zones to obtain the effective area of ​​the complete set of buffer zones. The required capacity of the workstation is obtained by dividing the effective area of ​​the complete set buffer area by the product of the volume of a single complete set and the height restriction coefficient.

5. The method for dynamic allocation of PFEP kit delivery points for assembly manufacturing scenarios according to claim 4, characterized in that, The basic configuration data for the production line also includes production cycle time, changeover stabilization time, emergency replenishment response time, and safety stock coefficient; the calculation method for the minimum safety stock unit quantity is as follows: Add the model changeover stabilization time to the emergency replenishment response time to obtain the time; Divide the time by the production cycle time, round the quotient up, and then multiply the rounded result by the safety stock coefficient to obtain the minimum safety stock number of units.

6. The method for dynamic allocation of PFEP kit delivery points for assembly manufacturing scenarios according to claim 1, characterized in that, Recommended storage locations for the current material, selected based on a multi-objective scoring function, include: For each candidate storage location in the candidate storage location set, a distance score is calculated based on the distance from the candidate storage location to the workstation. A capacity matching score is calculated using a four-segment linear function based on the matching degree deviation between the available capacity of the candidate storage location and the required capacity of the workstation. A frequency score is calculated based on the current material turnover frequency. A matching score is calculated based on the matching cluster information of the current set of storage locations or historical similar process routes. The distance score, capacity matching score, frequency score, and kitting score are multiplied by their respective dynamic weights and then summed to obtain the comprehensive score of the candidate storage location; wherein, each dynamic weight is adjusted based on exponential weighted moving average and type change impact detection. Based on the homogeneity priority allocation rule, Pareto optimal solution screening is performed in the candidate storage location set to select the candidate storage location with the highest comprehensive score as the recommended storage location.

7. The method for dynamic allocation of PFEP kit delivery points for assembly manufacturing scenarios according to claim 1, characterized in that, Also includes: If the storage location for any material fails to be occupied, all successful storage location occupancy within that set will be revoked, and the status of that set will be rolled back to the unassigned state.

8. A dynamic allocation system for PFEP kit delivery points in assembly manufacturing scenarios, characterized in that, The method for dynamic allocation of PFEP kit delivery points for assembly manufacturing scenarios as described in any one of claims 1 to 7 includes: The data coupling engine is used to acquire basic configuration data of the production line and layout data of the warehouse area. Based on the layout data of the warehouse area, a density clustering algorithm based on a distance threshold is used to divide each warehouse location into multiple clusters and determine the adjacency relationship between each cluster. The capacity calculation engine is used to obtain the material list of the current set to be allocated and the attribute data of each material, calculate the complete volume of a single set based on the attribute data, and dynamically calibrate the height restriction coefficient; based on the production line basic configuration data, the complete volume of a single set and the height restriction coefficient, calculate the workstation required capacity and the minimum safety stock number of sets. An optimized matching engine is used to traverse each material based on the material list, obtain the available capacity of each candidate storage location for the current material, filter out a set of candidate storage locations that simultaneously meet hard constraints, kitting constraints, and soft constraints, and filter out the recommended storage location for the current material from the set of candidate storage locations based on a multi-objective scoring function; wherein, the hard constraint is that the available capacity is not less than the minimum safety stock number of sets; the soft constraint is that the deviation between the available capacity and the required capacity of the workstation does not exceed a preset threshold; The transaction control engine is used to lock the current set of materials and batch occupy the recommended storage locations selected for all materials in the current set of materials. If all are successfully occupied, the allocation is confirmed and takes effect.

9. An electronic device, characterized in that, include: A processor and a memory, wherein the memory stores instructions that are loaded and executed by the processor to implement the PFEP kitting delivery point dynamic allocation method for assembly manufacturing scenarios as described in any one of claims 1 to 7.

10. A computer-readable storage medium, characterized in that, The computer-readable storage medium stores a computer program, which, when executed by a processor, implements the PFEP kit delivery point dynamic allocation method for assembly manufacturing scenarios as described in any one of claims 1 to 7.

Images