An automatic numbering method for power battery full life cycle management

Abstract

The present application relates to product life cycle management technical field, specifically to a kind of automatic numbering method for power battery full life cycle management, comprising the following steps: the name of cell material supplier, production batch number, production line environmental parameter value and transportation route mileage node are arranged in space dimension mapping, generate basic vector matrix, the parameter characteristic extraction and space dimension mapping arrangement are carried out to baking temperature curve value and pressure node value.The present application carries out cross adjustment to workshop location and equipment resources, and re-assigns equipment resource allocation quota value to abnormal state process node, so that abnormal identification, influence tracing, asset collection, capacity allocation and workshop scheduling form continuous closed loop, and then improve power battery full life cycle information correlation precision, abnormal positioning accuracy, resource utilization matching degree and production recovery efficiency.

Description

Technical Field

[0001] This invention relates to the field of product lifecycle management technology, and in particular to an automatic numbering method for the full lifecycle management of power batteries. Background Technology

[0002] Product lifecycle management technology involves the organization of information, status tracking, and collaborative management of resources throughout the entire process of a product, from demand planning, design and development, raw material supply, manufacturing, quality inspection, warehousing and transportation, sales and delivery, operation and maintenance, to decommissioning and recycling. It establishes the relationship between the product and raw materials, production batches, process parameters, production equipment, suppliers, work orders, inventory, transportation, sales, and after-sales records.

[0003] Existing product lifecycle management typically establishes relationships between product objects and raw materials, production batches, process parameters, production equipment, suppliers, work orders, inventory, transportation, sales, and after-sales records. In practice, it tends to archive, query, and track records at each stage. Different information often forms static correspondences based on fixed numbers, batch fields, time fields, or business primary keys. When abnormalities occur in the power battery supply chain, such as abnormalities in separator materials, contract manufacturing environments, or production stoppages, the existing operating model can only trace back and forth according to supplier, production batch, production date, or inventory records. It is difficult to further measure the degree of similarity in characteristics and time drift between abnormal nodes and different battery objects. This can easily lead to unaffected objects within the same batch being included in the disposal scope, or objects with similar abnormal characteristics across batches being missed. Therefore, improvements are needed. Summary of the Invention

[0004] The purpose of this invention is to address the shortcomings of existing technologies by proposing an automatic numbering method for the full life cycle management of power batteries.

[0005] To achieve the above objectives, the present invention adopts the following technical solution: an automatic numbering method for the full life cycle management of power batteries, comprising the following steps:

[0006] Spatial dimension mapping and arrangement are performed on the cell material supplier name, production batch number, production line environmental parameter value and transportation route mileage node to generate a basic vector matrix. Parameter feature extraction and spatial dimension mapping and arrangement are performed on the baking temperature curve value and pressure node value. The parameter feature extraction and spatial dimension mapping and arrangement results are matrix concatenated with the basic vector matrix to obtain the multidimensional gene battery code.

[0007] The target feature space mapping transformation is performed on the membrane material specifications, the processing environment parameters of the contract manufacturer, and the date of production stoppage to obtain the abnormal target sequence. The time axis node drift and stretching distance is calculated between the corresponding numerical sequence in the multidimensional gene battery encoding and the abnormal target sequence to obtain the list of affected batteries.

[0008] Based on the production shift count and work order status parameters in the affected battery list, density clustering is performed to generate a set of damaged asset prototypes. The current safety stock balance and alternative supplier capacity values ​​corresponding to the set of damaged asset prototypes are used as decision variables and mixed integer linear calculation is performed to obtain the macro capacity allocation plan.

[0009] Based on the corresponding workshop work order scheduling nodes and factory-level inventory transfer direction nodes within the macro-capacity allocation plan, cross-mutation calculations are performed on the remaining production line scheduling days and equipment load status values ​​in the affected battery list to generate a micro-workshop-level scheduling schedule. Abnormal state process nodes and corresponding equipment resource allocation quota values ​​are extracted from the micro-workshop-level scheduling schedule. The abnormal state process nodes and corresponding equipment resource allocation quota values ​​are then reassigned and replaced to obtain the power battery resource redistribution schedule.

[0010] Preferably, the steps for obtaining the fundamental vector matrix are as follows:

[0011] Collect the names of cell material suppliers, production batch numbers, production line environmental parameter values, and transportation route mileage nodes at each corresponding node throughout the entire life cycle of the power battery. Establish a node position index according to the chronological order of the nodes throughout the power battery's life cycle. Write the cell material supplier name into the supplier dimension position according to the node it belongs to, write the production batch number into the batch dimension position according to the node it belongs to, write the production line environmental parameter value into the production line environmental parameter dimension position according to the corresponding collection node, and write the transportation route mileage node into the transportation route dimension position according to the mileage sequence in the transportation route. Align the content of each dimension at the same node position row by row, retain the corresponding dimension empty space for missing node positions, and arrange the duplicate node positions continuously according to the appearance order of the production batch number to generate a basic vector matrix.

[0012] Preferably, the step of obtaining the multidimensional gene battery code is as follows:

[0013] The baking temperature curve values ​​and pressure node values ​​at corresponding nodes in each shift of the coating machine's operating cycle are captured. The baking temperature curve values ​​are segmented according to the start and end nodes of the shift. The starting temperature value, ending temperature value, peak temperature value, valley temperature value, temperature change amplitude, direction of change of continuous sampling points, and number of temperature fluctuations of the baking temperature curve values ​​corresponding to each shift are extracted. The pressure node values, pressure peak value, pressure valley value, pressure change amplitude, direction of change of continuous pressure nodes, and number of pressure fluctuations are extracted according to the corresponding nodes of the same shift. The baking temperature curve value extraction results are written into the temperature feature dimension position according to the shift node order, and the pressure node value extraction results are written into the pressure feature dimension position according to the corresponding pressure node order. The temperature feature dimension position and pressure feature dimension position under the same shift and the same node position are aligned and arranged to generate the parameter feature space mapping arrangement result.

[0014] The shift node position, temperature feature dimension position, and pressure feature dimension position in the parameter feature space mapping arrangement result are read. At the same time, the power battery life cycle node position, supplier dimension position, batch dimension position, production line environmental parameter dimension position, and transportation route dimension position in the basic vector matrix are read. The splicing position is determined by the correspondence between the power battery life cycle node position and the shift node position. For node positions that cannot be matched, the original dimension position is retained. For node positions that can be matched, column splicing is performed according to the fixed arrangement order of supplier dimension position, batch dimension position, production line environmental parameter dimension position, transportation route dimension position, temperature feature dimension position, and pressure feature dimension position. The spliced ​​node positions are numbered continuously according to the chronological order of the power battery life cycle nodes to obtain the multidimensional gene battery code.

[0015] Preferably, the step of obtaining the abnormal target sequence is as follows:

[0016] The multidimensional gene battery code is invoked, and the membrane material specifications, contract manufacturer processing environment parameters, and production stoppage date values ​​are read according to the corresponding positions of the abnormal nodes in the supply chain. The membrane material specifications are mapped to the target feature positions according to the material specification identifier, thickness parameters, and porosity parameters. The contract manufacturer processing environment parameters are mapped to the target feature positions according to the temperature parameters, humidity parameters, and cleanliness parameters. The production stoppage date values ​​are converted into continuous time positions according to the year, month, and day positions. The target feature positions belonging to the same abnormal node in the supply chain are arranged in a fixed order, and the abnormal nodes in different supply chains are arranged continuously according to the order of the production stoppage date values, thus generating an abnormal target sequence.

[0017] Preferably, the step of obtaining the list of affected batteries is as follows:

[0018] Extract the corresponding numerical sequence of consecutive nodes before and after the time position of the abnormal node in the supply chain from the multidimensional gene battery code. Use the time position corresponding to the production stop date value in the abnormal target sequence as the alignment benchmark. Change the matching position of the corresponding numerical sequence node by node according to the adjacent time position. Record the node offset amount generated by each time position offset. Calculate the numerical difference for the position of the membrane material specification parameter, the position of the contract manufacturer processing environment parameter, and the position of the production stop date value after the node offset. Determine the local stretching position according to the direction of the continuous node difference change. Expand the matching span for the position where the difference in the same direction occurs continuously. Shorten the matching span for the position where the difference direction changes. Accumulate the node offset amount, matching span, and numerical difference formed by each matching position in the order of time position to obtain the distance measurement value.

[0019] The multidimensional gene battery code positions corresponding to the distance measurement values ​​are read one by one. Each distance measurement value is compared with a preset anomaly limit. Multidimensional gene battery code positions where the distance measurement value is less than the preset anomaly limit are retained, while multidimensional gene battery code positions where the distance measurement value is greater than or equal to the preset anomaly limit are removed. Based on the retained multidimensional gene battery code positions, the corresponding production batch number is traced. The work-in-process node number and the sold battery node number under the production batch number are extracted. The duplicate work-in-process node number and the duplicate sold battery node number are uniquely retained. The affected battery list is generated by continuously arranging the production batch number, work-in-process node number, and sold battery node number according to their correspondence.

[0020] Preferably, the steps for obtaining the prototype set of damaged assets are as follows:

[0021] Extract the production shift count value corresponding to each work-in-process node number and sold battery node number in the affected battery list. Read the work order status parameters recorded at the same node position. Convert the production shift count value into consecutive shift positions according to the shift order. Discretize the work order status parameters according to the corresponding status positions of pending production, in production, paused, and completed. Establish a multi-dimensional feature space coordinate with the consecutive shift positions and status positions corresponding to each node. Count the number of nodes in a preset neighborhood around each multi-dimensional feature space coordinate. Use the coordinates where the number of nodes reaches a preset density limit as the aggregation starting position. Continuously absorb adjacent coordinates in the neighborhood that also reach the preset density limit. Use the position where adjacent coordinates cannot be absorbed as the aggregation boundary. Mark isolated coordinates that do not reach the preset density limit separately. Aggregate the corresponding work-in-process node numbers and sold battery node numbers according to different aggregation boundaries to generate a prototype set of damaged assets.

[0022] Preferably, the steps for obtaining the macro-capacity allocation plan are as follows:

[0023] Read the current safety stock balance value corresponding to the node to which each of the damaged asset prototype sets belongs, read the capacity value of the alternative suppliers according to the same node, align the current safety stock balance value according to the correspondence between the node, production shift value and work order status parameter, align the capacity value of the alternative suppliers according to the corresponding supply node of the alternative suppliers, check whether there are any missing records for the current safety stock balance value corresponding to each node, retain the empty position for the node with missing records, check whether there are any duplicate records for the capacity value of the alternative suppliers corresponding to each supply node, arrange the duplicate records according to the order of the corresponding supply cycle, so that each of the damaged asset prototype sets corresponds to the determined inventory balance position and supply capacity position, and obtain the current safety stock balance value and the capacity value of the alternative suppliers.

[0024] The current safety stock balance available for each attributing node is set as the inventory allocation decision value, and the available alternative supplier capacity value available for release from each alternative supplier is set as the supply capacity allocation decision value. The inventory allocation decision value is limited to not exceeding the current safety stock balance of the corresponding attributing node, and the supply capacity allocation decision value is limited to not exceeding the capacity value of the corresponding alternative supplier. The capacity gap that each attributing node needs to fill is determined according to the number of attributing nodes in the damaged asset prototype set. The inventory allocation decision value and the supply capacity allocation decision value that satisfy the capacity gap constraint are combined item by item. Combinations that exceed the inventory balance limit, exceed the supply capacity limit, or cannot cover the corresponding capacity gap are eliminated. The remaining combinations are sorted according to the degree of capacity gap satisfaction of each attributing node. The allocation combination that can cover the capacity gap of all attributing nodes and has the smallest combination result of inventory occupancy and supply capacity call value is selected. The attributing nodes, inventory allocation decision values, and supply capacity allocation decision values ​​corresponding to the allocation combination are arranged in sequence to generate a macro capacity allocation plan.

[0025] Preferably, the step of obtaining the power battery resource reallocation schedule is as follows:

[0026] Extract the workshop work order scheduling node corresponding to each affiliated node in the macro capacity allocation plan, identify the factory-level inventory transfer direction node according to the inventory transfer-out and inventory transfer-in positions in the macro capacity allocation plan, match the production batch number in the affected battery list according to the workshop work order scheduling node, locate the workshop node corresponding to the production batch number, read the remaining production line scheduling days corresponding to the work orders that have not yet been executed for each workshop node, read the equipment load status value according to the same workshop node, arrange the remaining production line scheduling days according to the execution order of the workshop work order scheduling nodes, arrange the equipment load status value according to the corresponding equipment resource position, and maintain the transfer-in and transfer-out correspondence between the factory-level inventory transfer direction node and the workshop node, to obtain the workshop work order scheduling node, the factory-level inventory transfer direction node, the remaining production line scheduling days, and the equipment load status value;

[0027] The work order execution positions corresponding to each workshop work order scheduling node are arranged as candidate positions from front to back according to the remaining production line scheduling days. The inventory transfer-in workshops corresponding to the factory-level inventory transfer direction nodes are used as positions that can accept work orders. The equipment load status values ​​of different workshop nodes are compared item by item, and the workshop nodes with equipment load status values ​​that allow for the release of equipment resources are selected as the exchange positions. The corresponding work order execution positions are exchanged between different workshop work order scheduling nodes, while maintaining the correspondence between production batch numbers. The remaining production line scheduling days are reread for the exchanged work order execution positions. If the remaining production line scheduling days after the exchange exceed the corresponding remaining production line scheduling days before the exchange, the work order execution positions before the exchange are restored. If the remaining production line scheduling days do not exceed the corresponding remaining production line scheduling days before the exchange, then the work order execution position after the exchange is retained. Then, the equipment resource position where the equipment load status value exceeds the median of the load status values ​​of all equipment in the same workshop is selected, and the corresponding work order execution position is transferred to the acceptable work order position pointed to by the factory-level inventory transfer direction node and where the equipment load status value is lower than the median of the load status values ​​of all equipment in the same workshop. The work order execution position exchange and equipment resource position adjustment are performed cyclically until the remaining production line scheduling days no longer decrease. The work order execution order is rearranged according to the retained workshop work order scheduling node order, the factory-level inventory transfer direction node, and the corresponding equipment resource position to generate a micro-workshop-level scheduling schedule.

[0028] The process node status corresponding to the work order execution sequence is extracted item by item. Abnormal process nodes whose status is inconsistent with the predetermined execution status in the micro-workshop level scheduling are filtered out. The equipment resource allocation quota value corresponding to each abnormal process node is read. The remaining equipment resource allocation quota values ​​within the same execution period are queried according to the workshop node to which the abnormal process node belongs. Equipment resource positions that have not reached the upper limit of the corresponding equipment resource allocation quota are selected as replacement positions. The original equipment resource allocation quota value of the abnormal process node is deducted from the original equipment resource position, and the equipment resource allocation quota value is reassigned to the replacement position according to the work order execution requirements corresponding to the abnormal process node. If the remaining allocation quota of a single replacement position is insufficient to meet the work order execution requirements corresponding to the abnormal process node, replacement positions are selected from largest to smallest according to the remaining allocation quota until all work order execution requirements corresponding to the abnormal process node receive equipment resource allocation quota values. The abnormal process nodes, equipment resource positions, equipment resource allocation values, and work order execution sequence after reassignment and replacement are written back to the micro-workshop level scheduling to generate a power battery resource reallocation schedule.

[0029] Compared with the prior art, the advantages and positive effects of the present invention are as follows:

[0030] In this invention, by spatially mapping and arranging the cell material supplier name, production batch number, production line environmental parameter values, and transportation route mileage nodes, and further introducing baking temperature curve values ​​and pressure node values ​​to complete parameter feature extraction and matrix splicing, the power battery number is expanded from a traditional identification to a multi-dimensional genetic battery code that can carry the supply source, manufacturing batch, environmental state, transportation location, and key process characteristics. This allows the same power battery to form a continuous and traceable feature association at different life cycle nodes. When a supply chain anomaly occurs, the separator material specifications, contract manufacturer processing environment parameters, and production stop date values ​​are converted into an abnormal target sequence. Then, by measuring the time axis node drift and stretching distance, the time position offset, feature similarity, and node continuity relationship between the abnormal target sequence and the multi-dimensional genetic battery code are jointly judged, thereby narrowing the scope of the anomaly's impact and avoiding excessive recalls caused by tracing the entire batch based solely on the production batch or a single time node. This ensures that the work-in-process node number and sold battery cells are traced back to their original values. Point numbering can generate a list of affected batteries based on the actual degree of anomaly correlation. After the affected scope is clarified, damaged assets are collected based on the distribution density of production shift values ​​and work order status parameters in a multi-dimensional feature space. This allows affected objects with similar production status and shift attributes to form a prototype set of damaged assets with common resource disposal characteristics. Combined with the current safety stock balance and alternative supplier capacity values, capacity gaps are allocated in a coordinated manner, creating a quantitative constraint relationship between inventory call-up, supply replenishment, and capacity acceptance. Furthermore, the macro-capacity allocation results are implemented at workshop work order scheduling nodes, factory-level inventory transfer direction nodes, remaining production line scheduling days, and equipment load status values. Work order locations and equipment resources are cross-adjusted, and equipment resource allocation quotas are reassigned to abnormal state process nodes. This creates a continuous closed loop of anomaly identification, impact tracing, asset collection, capacity allocation, and workshop scheduling, thereby improving the accuracy of information correlation, anomaly location accuracy, resource utilization matching degree, and production recovery efficiency throughout the entire life cycle of power batteries. Attached Figure Description

[0031] Figure 1 A graph showing the results of numerical extraction of the baking temperature curve;

[0032] Figure 2 The image shows the results of numerical extraction of pressure nodes. Detailed Implementation

[0033] To make the objectives, technical solutions, and advantages of this invention clearer, the invention will be further described in detail below with reference to the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are merely illustrative and not intended to limit the invention.

[0034] Please see Figure 1-2This invention provides a technical solution: an automatic numbering method for the full life cycle management of power batteries, comprising the following steps:

[0035] Spatial dimension mapping and arrangement are performed on the cell material supplier name, production batch number, production line environmental parameter value and transportation route mileage node to generate a basic vector matrix. Parameter feature extraction and spatial dimension mapping and arrangement are performed on the baking temperature curve value and pressure node value. The parameter feature extraction and spatial dimension mapping and arrangement results are matrix concatenated with the basic vector matrix to obtain the multidimensional gene battery code.

[0036] The target feature space mapping transformation is performed on the membrane material specifications, the processing environment parameters of the contract manufacturer, and the date of production stoppage to obtain the abnormal target sequence. The time axis node drift and stretching distance is calculated between the corresponding numerical sequence in the multidimensional gene battery encoding and the abnormal target sequence to obtain the list of affected batteries.

[0037] Based on the production shift count and work order status parameters in the affected battery list, density clustering is performed to generate a set of damaged asset prototypes. The current safety stock balance and alternative supplier capacity values ​​corresponding to the set of damaged asset prototypes are used as decision variables and mixed integer linear calculation is performed to obtain the macro capacity allocation plan.

[0038] Based on the corresponding workshop work order scheduling nodes and factory-level inventory transfer direction nodes within the macro-capacity allocation plan, cross-mutation calculations are performed on the remaining production line scheduling days and equipment load status values ​​in the affected battery list to generate a micro-workshop-level scheduling schedule. Abnormal state process nodes and corresponding equipment resource allocation quota values ​​are extracted from the micro-workshop-level scheduling schedule. The abnormal state process nodes and corresponding equipment resource allocation quota values ​​are then reassigned and replaced to obtain the power battery resource redistribution schedule.

[0039] The steps to obtain the fundamental vector matrix are as follows:

[0040] Collect the names of cell material suppliers, production batch numbers, production line environmental parameter values, and transportation route mileage nodes at each corresponding node throughout the entire life cycle of the power battery. Establish a node position index according to the chronological order of the nodes throughout the power battery's life cycle. Write the cell material supplier name into the supplier dimension position according to the node it belongs to, write the production batch number into the batch dimension position according to the node it belongs to, write the production line environmental parameter value into the production line environmental parameter dimension position according to the corresponding collection node, and write the transportation route mileage node into the transportation route dimension position according to the mileage sequence in the transportation route. Align the content of each dimension at the same node position row by row, retain the corresponding dimension empty space for missing node positions, and arrange the duplicate node positions continuously according to the appearance order of the production batch number to generate a basic vector matrix.

[0041] Specifically, based on the cell material supplier names, production batch numbers, production line environmental parameter values, and transportation route mileage nodes collected at each corresponding node in the entire life cycle of the power battery, a primary node location index is first established according to the actual occurrence order of the life cycle nodes. Then, a secondary expansion sequence number is established for the production batch numbers that appear repeatedly at the same node location. The primary node location index maintains the original life cycle sequence, while the secondary expansion sequence number only increments from 1 within the corresponding primary node. The supplier name is written into the supplier dimension position according to its primary node position. The production batch number is simultaneously bound to the primary node position and the secondary expansion sequence number. The production line environmental parameter values ​​occupy independent dimension positions according to the parameter source node. Values ​​with different dimensions such as temperature, humidity, and cleanliness are not combined in the calculation. The transportation route mileage nodes are based on the actual distance from 0km to subsequent mileage nodes. The location of the transportation route dimension is written sequentially. For example, when the same production batch number corresponds to supplier S1, temperature 23.6℃ and transportation location 126km at node 12, all three items are bound to node 12. If a new production batch number appears at node 12, node 12 is kept unchanged and the second-level expansion sequence number is incremented from 1 to 2. The corresponding relationship of each dimension is copied synchronously with the expansion sequence number. When a certain dimension has no collected value at node 12, an empty space is reserved and node 12 and its second-level expansion sequence number are retained. After all nodes are written, they are arranged in ascending order according to the first-level node position index. Within the same first-level node, they are arranged in ascending order according to the second-level expansion sequence number. The first-level node position index, the second-level expansion sequence number and the supplier dimension position, batch dimension position, production line environment parameter dimension position and transportation route dimension position are bound row by row to form a basic vector matrix with fixed row positioning relationship.

[0042] The steps to obtain the multidimensional gene battery code are as follows:

[0043] The baking temperature curve values ​​and pressure node values ​​at corresponding nodes in each shift of the coating machine's operating cycle are captured. The baking temperature curve values ​​are segmented according to the start and end nodes of the shift. The starting temperature value, ending temperature value, peak temperature value, valley temperature value, temperature change amplitude, direction of change of continuous sampling points, and number of temperature fluctuations of the baking temperature curve values ​​corresponding to each shift are extracted. The pressure node values, pressure peak value, pressure valley value, pressure change amplitude, direction of change of continuous pressure nodes, and number of pressure fluctuations are extracted according to the corresponding nodes of the same shift. The baking temperature curve value extraction results are written into the temperature feature dimension position according to the shift node order, and the pressure node value extraction results are written into the pressure feature dimension position according to the corresponding pressure node order. The temperature feature dimension position and pressure feature dimension position under the same shift and the same node position are aligned and arranged to generate the parameter feature space mapping arrangement result.

[0044] The system reads the shift node position, temperature feature dimension position, and pressure feature dimension position from the parameter feature space mapping arrangement result. Simultaneously, it reads the power battery life cycle node position, supplier dimension position, batch dimension position, production line environmental parameter dimension position, and transportation route dimension position from the basic vector matrix. The splicing position is determined by the correspondence between the power battery life cycle node position and the shift node position. For node positions that cannot be matched, the original dimension position is retained. For node positions that can be matched, column splicing is performed according to the fixed arrangement order of supplier dimension position, batch dimension position, production line environmental parameter dimension position, transportation route dimension position, temperature feature dimension position, and pressure feature dimension position. The spliced ​​node positions are then numbered consecutively according to the chronological order of the power battery life cycle nodes to obtain the multidimensional gene battery code.

[0045] Specifically, based on the baking temperature curve values ​​and pressure node values ​​captured at corresponding nodes during each shift of the coating machine's operating cycle, a corresponding sampling interval is extracted according to the start and end times of each shift. Within the shift interval, the first valid temperature sample value, the last valid temperature sample value, the temperature peak value, and the temperature trough value are read, and the temperature change amplitude between the temperature peak value and the temperature trough value is recorded. The direction of change of continuous sampling points is not counted according to arbitrary tiny fluctuations, but rather the direction judgment threshold is determined first based on the sensor calibration allowable error and the adjacent sampling difference of the stable operation segment of the equipment. The threshold is the larger of the two values. For example, if the temperature sensor calibration allowable error is 0.3℃ and the maximum absolute value of the adjacent temperature difference in the historical stable operation segment is 0.18℃, then the temperature direction judgment threshold is 0.3℃. Adjacent sampling differences greater than 0.3℃ are recorded as an increase, and those less than -0℃ are recorded as an increase. 3℃ is recorded as a decrease, and the rest as a hold. The pressure direction determination threshold adopts the same setting process. For example, the allowable error for pressure sensor calibration is 0.005MPa, and the maximum absolute value of the pressure difference between adjacent segments in a stable operation segment is 0.003MPa, so 0.005MPa is taken. Continuous segments in the same direction form one direction segment. When the transition occurs between the rising segment and the falling segment, one fluctuation count is accumulated. The hold segment only connects the segments in the same direction before and after it and does not add fluctuation count separately. Then, the pressure peak value, pressure valley value, pressure change amplitude, direction of change of continuous pressure nodes, and pressure fluctuation count are extracted according to the same shift node. The temperature features are arranged according to the shift node order, and the pressure features are arranged according to the order of appearance of pressure nodes. When the shift number and node position are consistent, row alignment is performed. When one side is missing, the corresponding dimension space is retained to obtain the parameter feature space mapping arrangement result.

[0046] Based on the shift node positions, temperature feature dimension positions, and pressure feature dimension positions in the parameter feature space mapping arrangement results, and simultaneously reading the power battery full life cycle node positions, supplier dimension positions, batch dimension positions, production line environmental parameter dimension positions, and transportation route dimension positions from the basic vector matrix, a first-level matching constraint is established using the production batch number, searching for splicing positions only between records with the same production batch number. A second-level matching constraint is then established using the life cycle node occurrence time and the shift start and end times. When the life cycle node occurrence time falls within the corresponding shift time interval, it is determined as a candidate splicing position. If it occurs at the boundary between two consecutive shifts, it is processed according to the shift to which the time point belongs. If multiple candidate nodes exist for the same production batch number in the same shift, the life cycle node occurrence time is compared with the nodes of each shift. The absolute time difference of the sampling center time is used to select the position with the smallest time difference. If the time differences are still the same, the previous position is selected according to the order of the node positions in the basic vector matrix. This avoids artificially widening the corresponding range of nodes through additional empirical limits. After determining the splicing position, column splicing is performed in a fixed order of supplier dimension position, batch dimension position, production line environment parameter dimension position, transportation route dimension position, temperature feature dimension position, and pressure feature dimension position. Life cycle nodes that cannot correspond to shifts continue to retain their original rows, and empty positions are reserved in the temperature feature dimension position and pressure feature position. Duplicate nodes that can correspond continue to use the second-level expansion sequence number of the basic vector matrix. After all splicing is completed, the life cycle node positions and second-level expansion sequence numbers are continuously numbered. The continuous numbering is kept one-to-one associated with the original production batch number to obtain the multidimensional gene battery code.

[0047] The steps for obtaining the abnormal target sequence are as follows:

[0048] The multidimensional gene battery coding is invoked to read the membrane material specifications, contract manufacturer processing environment parameters, and shutdown date values ​​according to the corresponding positions of the abnormal nodes in the supply chain. The membrane material specifications are mapped to the target feature positions according to the material specification identifier, thickness parameters, and porosity parameters. The contract manufacturer processing environment parameters are mapped to the target feature positions according to the temperature parameters, humidity parameters, and cleanliness parameters. The shutdown date values ​​are converted into continuous time positions according to the year, month, and day positions. The target feature positions belonging to the same abnormal node in the supply chain are arranged in a fixed order, and the abnormal nodes in different supply chains are arranged continuously according to the order of the shutdown date values, thus generating an abnormal target sequence.

[0049] Specifically, based on the identified supply chain anomaly nodes in the multidimensional gene battery coding, the membrane material specifications, contract manufacturer processing environment parameters, and shutdown date values ​​associated with each node are read sequentially. Each supply chain anomaly node is then divided into a fixed-length anomaly node block. Within the block, the first three positions are sequentially filled with the material specification identifier, thickness parameter, and porosity parameter. The next three positions are sequentially filled with the contract manufacturer's temperature parameter, humidity parameter, and cleanliness parameter. Finally, the continuous time position corresponding to the shutdown date is written. The material specification identifier remains discrete, while the thickness, porosity, temperature, humidity, and cleanliness parameters retain their original units of measurement and occupy independent positions, without cross-dimensional merging in this step. The continuous time position is the earliest among all the aforementioned supply chain anomaly nodes. The production stoppage date is used as the time origin, with the earliest date recorded as 0. The remaining dates are incremented according to the number of natural days between them and the time origin. For example, when the abnormal node dates are March 1, 2026, March 4, 2026, and March 9, 2026, the corresponding consecutive time positions are recorded as 0, 3, and 8, respectively. When multiple abnormal nodes appear in the supply chain on the same date, they are first arranged according to the original node positions in the multidimensional gene battery code. If the original node positions are the same, they are arranged according to the order of appearance of the production batch number. When a certain parameter is missing, only the corresponding empty position in the abnormal node block is retained. Other feature positions within the same block are not changed, and the position length of subsequent node blocks is not compressed. After all abnormal node blocks are completed, they are connected in ascending order of consecutive time positions, and the fixed position order within each node block is maintained to obtain the abnormal target sequence.

[0050] The steps to obtain the list of affected batteries are as follows:

[0051] The numerical sequence of consecutive nodes before and after the time position of the abnormal node in the supply chain is extracted from the multidimensional gene battery code. The time position corresponding to the production stoppage node date value in the abnormal target sequence is used as the alignment benchmark. The matching position of the corresponding numerical sequence is changed node by node according to the adjacent time position. The node offset caused by each time position shift is recorded. The numerical difference is calculated for the position of the membrane material specification parameter, the position of the contract manufacturer processing environment parameter, and the position of the production stoppage node date value after the node shift. The local stretching position is determined according to the direction of the change of the difference of consecutive nodes. The matching span is expanded for the position where the difference in the same direction occurs continuously, and the matching span is shortened for the position where the difference direction changes. The node offset, matching span, and numerical difference formed by each matching position are accumulated in the order of time position to obtain the distance measurement value.

[0052] The multidimensional gene battery code positions corresponding to the distance measurement values ​​are read one by one. Each distance measurement value is compared with a preset anomaly limit. Multidimensional gene battery code positions with distance measurement values ​​less than the preset anomaly limit are retained, while multidimensional gene battery code positions with distance measurement values ​​greater than or equal to the preset anomaly limit are removed. Based on the retained multidimensional gene battery code positions, the corresponding production batch number is traced. The work-in-process node number and the sold battery node number under the production batch number are extracted. The duplicate work-in-process node number and the duplicate sold battery node number are uniquely retained. The affected battery list is generated by continuously arranging the production batch number, work-in-process node number, and sold battery node number according to their correspondence.

[0053] Specifically, based on the numerical sequence of consecutive nodes before and after the time position of the abnormal node in the supply chain in the multidimensional gene battery coding, and the time position corresponding to the production stoppage date in the abnormal target sequence, the thickness parameter, porosity parameter, contract manufacturer temperature parameter, humidity parameter, cleanliness parameter, and consecutive time position are normalized to 0 to 1 according to the minimum and maximum values ​​of the same dimension. The same material specification identifier is recorded as 0, and different identifiers as 1. Empty dimensions are not included in the local distance averaging. Then, constrained dynamic time warping (DTW) is used to calculate node drift and local stretching relationships. The time warping bandwidth is taken as the maximum absolute value of the actual node offset of historically confirmed supply chain abnormal events. For example, if the offset of 10 historical events is 1, 1, 2, 2, 3, 3, 4, 4, 5, and 7 nodes, and the maximum offset is 7, then the bandwidth is taken as 7 nodes. Positions exceeding the bandwidth are not included in the path search. The cumulative distance is calculated as follows:

[0054] ; recursion, where, Represents the preceding encoded sequence Each node and the abnormal target sequence before The cumulative distance of each node. Representing the The encoding node and the first Local distances of anomalous nodes Represents the sequence number of the encoded node. Represents the sequence number of the abnormal node. and These represent the sequence number of the preceding node, respectively. The minimum value is taken from the three valid preceding cumulative distances. The distance is backtracked along the minimum cumulative distance to the endpoint. Continuous horizontal or vertical movement is recorded as the local stretching position, and diagonal movement is recorded as the synchronous matching position. The cumulative distance is then divided by the number of valid matching nodes in the backtracking path to obtain the distance measurement value.

[0055] Based on the obtained distance measurement values ​​and their corresponding multidimensional gene battery coding locations, the distance measurement values ​​corresponding to affected and unaffected batches are first extracted from historical supply chain events where anomaly tracing has been completed. The midpoint of two adjacent values ​​for all different distance values ​​is used as a preset anomaly limit candidate value. For each candidate value, the proportion of historical affected batches with distances less than the candidate value and the proportion of historical unaffected batches with distances greater than or equal to the candidate value are calculated. Then, the candidate values ​​are compared using the sum of sensitivity and specificity corresponding to the Youden index minus 1. When the indices are the same, the candidate value with the higher retention rate of affected batches is selected first. For example, if the distances of historical affected batches are 0.14, 0.18, 0.21, 0.24, and 0.27, and the distances of unaffected batches are 0.25, 0.31, 0.36, 0.42, and 0.48, the candidate value 0.245 can be retained. Four affected batches were identified, and five unaffected batches were excluded. A candidate value of 0.28 was used to retain five affected batches and exclude four unaffected batches. Both corresponded to a Youden index of 0.8. Based on the priority rule of retaining affected batches, 0.28 was selected as the preset anomaly limit. After a new anomaly event was traced, it was recalibrated using the same process. Then, the current distance measurement values ​​were compared one by one. The multidimensional gene battery code positions that were less than the preset anomaly limit were retained, while those that were greater than or equal to the preset anomaly limit were removed. The production batch number was then traced from the retained positions to extract the work-in-process node number and the sold battery node number under the corresponding batch. When the same number appeared repeatedly, the first associated position was retained and its repeated references were merged. Finally, the affected batteries were obtained by arranging them continuously according to the correspondence between the production batch number, the work-in-process node number, and the sold battery node number.

[0056] The steps to obtain the prototype set of damaged assets are as follows:

[0057] Extract the production shift count value corresponding to each work-in-process node number and sold battery node number in the affected battery list. Read the work order status parameters recorded at the same node position. Convert the production shift count value into consecutive shift positions according to the shift sequence. Discretize the work order status parameters according to the corresponding status positions of pending production, in production, paused, and completed. Establish a multi-dimensional feature space coordinate with the consecutive shift positions and status positions corresponding to each node. Count the number of nodes in the preset neighborhood around each multi-dimensional feature space coordinate. Use the coordinates where the number of nodes reaches the preset density limit as the starting position of aggregation. Continuously absorb adjacent coordinates in the neighborhood that also reach the preset density limit. Use the position where it is no longer possible to absorb adjacent coordinates as the aggregation boundary. Mark isolated coordinates that do not reach the preset density limit separately. Aggregate the corresponding work-in-process node numbers and sold battery node numbers according to different aggregation boundaries to generate a prototype set of damaged assets.

[0058] Specifically, based on the production shift count and work order status parameters corresponding to each work-in-process node number and sold battery node number in the affected battery list, the production shift count is converted into consecutive shift positions according to shift order. The four types of work order status parameters—pending production, in production, paused, and completed—are retained as discrete status positions. Gower distance is used to calculate the mixed feature distance, and the shift position uses the normalized absolute difference. Work order statuses are recorded as 0 for the same status and 1 for different statuses. The average of the two distances is then used for density clustering via DBSCAN. Identify the core nodes, among which, Representative node In the neighborhood radius The set of nodes within, Represents the current node. Represents the neighborhood radius. Represents the number of neighboring nodes. This represents the minimum number of neighboring nodes. The minimum number of neighboring nodes is increased by 1 to get a lower limit of 3, and increased by 1 to get an upper limit of 5. Then, the results of 3, 4, and 5 in the historical processing batches are compared. For example, when there are 48 historical nodes, taking 3 results in 6 cross-batch erroneous mergings, taking 4 results in no cross-batch mergings but 2 isolated boundary points, and taking 5 results in 1 consecutive batch split. Therefore, 4 is taken. Then, the fourth nearest neighbor distance of each node is sorted in ascending order, and the adjacent distance increment is calculated. The distance with the largest increment is taken as the neighborhood radius. For example, the largest increment of 0.11, 0.13, 0.15, 0.18, and 0.31 is 0.13, so 0.18 is taken. The density of reachable nodes is continuously absorbed from the core node, and the remaining nodes are marked separately to obtain the prototype set of damaged assets.

[0059] The steps to obtain a macro-level capacity allocation plan are as follows:

[0060] Read the current safety stock balance value corresponding to the node to which each damaged asset prototype set belongs, and read the capacity value of the alternative suppliers according to the same node. Align the current safety stock balance value according to the correspondence between the node, production shift value, and work order status parameter. Align the capacity value of the alternative suppliers according to the corresponding supply node of the alternative suppliers. Check whether there are any missing records for the current safety stock balance value corresponding to each node. For the node with missing records, leave the empty position. Check whether there are any duplicate records for the capacity value of the alternative suppliers corresponding to each supply node. Arrange the duplicate records according to the order of the corresponding supply cycle. So that each damaged asset prototype set corresponds to the determined inventory balance position and supply capacity position, and obtain the current safety stock balance value and the capacity value of the alternative suppliers.

[0061] The current safety stock balance available for each attributable node is set as the inventory allocation decision value, and the available alternative supplier capacity is set as the supply capacity allocation decision value. The inventory allocation decision value is limited to not exceeding the current safety stock balance of the corresponding attributable node, and the supply capacity allocation decision value is limited to not exceeding the capacity of the corresponding alternative supplier. The capacity gap that each attributable node needs to fill is determined according to the number of attributable nodes in the damaged asset prototype set. The inventory allocation decision value and supply capacity allocation decision value that satisfy the capacity gap constraint are combined item by item. Combinations that exceed the inventory balance limit, exceed the supply capacity limit, or cannot cover the corresponding capacity gap are eliminated. The remaining combinations are sorted according to the degree of capacity gap satisfaction of each attributable node. The allocation combination that can cover the capacity gap of all attributable nodes and has the smallest combination result of inventory occupancy and supply capacity call value is selected. The attributable nodes, inventory allocation decision values, and supply capacity allocation decision values ​​corresponding to the allocation combination are arranged in order to generate a macro capacity allocation plan.

[0062] Specifically, based on the identified attribution nodes in the damaged asset prototype set, the current safety stock balance and alternative supplier capacity values ​​corresponding to each attribution node are read one by one. The inventory balance position is aligned using a three-level combination key: attribution node, production shift value, and work order status parameter. Inventory balances for different production shifts or work order statuses under the same attribution node are kept separate. The alternative supplier capacity position is aligned using a three-level combination key: alternative supplier, supply node, and supply cycle. The capacity of the same alternative supplier in different supply cycles is kept separate. For example, if alternative supplier S1 has 1200 units and 900 units of releaseable capacity in week 1 and week 2 respectively, it occupies two supply cycle positions respectively, and 2100 units is not considered a callable value for the same cycle. The system distinguishes between missing and zero values ​​in the current safety stock balance. (Missing data is listed in the table.) The key combination indicates that no inventory record has been retrieved and the space is left empty. A value of 0 indicates that a record has been retrieved but there is currently no available inventory. For duplicate records of alternative supplier capacity values, the start and end positions of the supply cycle are compared first. If the cycles do not overlap, they are retained in chronological order. If the cycles are exactly the same, they are arranged in the original order of appearance and bound with the same supply cycle identifier. If the cycles partially overlap, the overlapping start and end points are used as segment boundaries to split the original supply cycle into non-overlapping continuous positions. The original capacity records corresponding to each segment are attached separately. Overlapping records are not added together. After all positions are aligned, each belonging node is checked for the existence of a corresponding inventory balance position and supply capacity position according to the damaged asset prototype set number. Missing positions are left empty, forming the current safety stock balance value and alternative supplier capacity value corresponding to each item in the damaged asset prototype set.

[0063] Based on the current safety stock level, alternative supplier capacity, and capacity gap, both types of resources are converted into equivalent battery quantities. A mixed-integer linear programming approach is then used to assign ownership to nodes. The inventory allocation decision quantity is denoted as Alternative suppliers To the home node The supply capacity allocation decision quantity is denoted as ,set up , and satisfy ,in, Represents the node to which the node belongs Inventory allocation decision volume Represents the node to which the node belongs The current safety stock level. Representative of alternative suppliers To the home node The decision-making volume of supply capacity allocation. This represents the production capacity of the alternative suppliers. Represents the node to which the node belongs The capacity gap Represents the sequence number of the node to which it belongs. Represents the candidate supplier number. Represents the total number of alternative suppliers. Represents serial numbers 1 to The supply capacity allocation decision quantities are accumulated, and the inventory allocation decision quantities and supply capacity allocation decision quantities are non-negative integers. The branch and bound method is used to solve the problem. First, branches that exceed the current safety stock balance, exceed the capacity of alternative suppliers, and cannot cover the capacity gap are eliminated. Then, the sum of the inventory allocation decision quantities and the supply capacity allocation decision quantities of the remaining combinations are compared. The combination with the smallest sum is selected. If the sums are the same, the combination with the smaller inventory allocation decision quantity is selected first, and then the combination with the fewer alternative suppliers is selected to obtain the macro capacity allocation plan.

[0064] The steps for obtaining the power battery resource reallocation schedule are as follows:

[0065] Extract the workshop work order scheduling node corresponding to each affiliated node in the macro capacity allocation plan. Identify the factory-level inventory transfer direction node according to the inventory transfer-out and transfer-in locations in the macro capacity allocation plan. Match the production batch number in the affected battery list with the workshop work order scheduling node to locate the workshop node corresponding to the production batch number. Read the remaining production line scheduling days corresponding to the unexecuted work orders of each workshop node. Read the equipment load status value according to the same workshop node. Arrange the remaining production line scheduling days according to the execution order of the workshop work order scheduling nodes. Arrange the equipment load status value according to the corresponding equipment resource location. Maintain the transfer-in and transfer-out correspondence between the factory-level inventory transfer direction node and the workshop node. Obtain the workshop work order scheduling node, factory-level inventory transfer direction node, remaining production line scheduling days, and equipment load status value.

[0066] The work order execution positions corresponding to each workshop work order scheduling node are arranged in a candidate order from front to back according to the remaining production line scheduling days. The inventory transfer-in workshops corresponding to the factory-level inventory transfer direction nodes are used as the accepting work order positions. The equipment load status values ​​of different workshop nodes are compared item by item, and workshop nodes with available equipment resources in their equipment load status values ​​are selected as the exchange positions. The corresponding work order execution positions are exchanged between different workshop work order scheduling nodes, while maintaining the production batch number correspondence. The remaining production line scheduling days are reread for the exchanged work order execution positions. If the remaining production line scheduling days after the exchange exceed the corresponding remaining production line scheduling days before the exchange, the work order execution position before the exchange is restored. If the remaining production line scheduling days do not exceed the corresponding remaining production line scheduling days before the exchange, then the work order execution position after the exchange is retained. Then, select the equipment resource position where the equipment load status value exceeds the median of the load status values ​​of all equipment in the same workshop, and transfer the corresponding work order execution position to the position of the work order that can be accepted by the factory-level inventory transfer direction node and the equipment load status value is lower than the median of the load status values ​​of all equipment in the same workshop. Repeat the work order execution position exchange and equipment resource position adjustment until the remaining production line scheduling days no longer decrease. Then, rearrange the work order execution order according to the retained workshop work order scheduling node order, factory-level inventory transfer direction node, and corresponding equipment resource position to generate a micro-workshop-level scheduling schedule.

[0067] The process node status corresponding to the work order execution sequence is extracted item by item. Abnormal process nodes whose status is inconsistent with the predetermined execution status in the micro-workshop level scheduling are filtered out. The equipment resource allocation quota value corresponding to each abnormal process node is read. The remaining equipment resource allocation quota values ​​within the same execution period are queried according to the workshop node to which the abnormal process node belongs. Equipment resource locations that have not reached the upper limit of the corresponding equipment resource allocation quota are selected as replacement locations. The original equipment resource allocation quota value of the abnormal process node is deducted from the original equipment resource location. The equipment resource allocation quota value is then reassigned to the replacement location according to the work order execution requirements corresponding to the abnormal process node. If the remaining allocation quota of a single replacement location is insufficient to meet the work order execution requirements corresponding to the abnormal process node, replacement locations are selected from largest to smallest according to the remaining allocation quota until all work order execution requirements corresponding to the abnormal process node receive equipment resource allocation quota values. The abnormal process nodes, equipment resource locations, equipment resource allocation quota values, and work order execution sequence after reassignment and replacement are written back to the micro-workshop level scheduling to generate a power battery resource reallocation schedule.

[0068] Specifically, based on the workshop work order scheduling nodes corresponding to each attribution node within the macro-capacity allocation plan, as well as the inventory transfer-out and inventory transfer-in locations, the factory-level inventory transfer direction nodes are first determined according to the relationship between the inventory transfer-out location and the inventory transfer-in location. Then, based on the workshop work order scheduling nodes and the production batch numbers in the affected battery list, the corresponding workshop nodes are located, and the remaining production line scheduling days and equipment load status values ​​of the unexecuted work orders are read. The workshop work order scheduling nodes are arranged in an initial candidate order according to the remaining production line scheduling days from smallest to largest. The inventory transfer-in workshops pointed to by the factory-level inventory transfer direction nodes are limited to the locations that can accept work orders. The median of the load status values ​​of all equipment in the same workshop is calculated as the dynamic boundary for this round. For example, when the load status values ​​of 5 equipment are 52%, 61%, 67%, 74%, and 86%, the median is 67%. Equipment with a load status value higher than 67% is considered a dynamic boundary for this round. The source position enters the set to be exchanged; positions below 67% enter the set that can be accepted; positions equal to 67% maintain their original schedule. Then, work orders are selected from the set to be exchanged in descending order of remaining production line scheduling days, and positions are selected from the set that can be accepted in ascending order of equipment load status. Paired work order position exchanges are performed while maintaining the production batch number association. After each exchange, the total number of remaining production line scheduling days for the work orders involved is reread. If the total increases, the original position is restored; if the total does not increase, the exchanged position is retained. For work orders that are still above the median position, equipment position transfer is performed along the factory-level inventory transfer direction node as a mutation. After completing one round of traversal of all candidate positions, the total number of remaining production line scheduling days is recalculated. The loop stops when the total number of remaining production line scheduling days no longer decreases compared to the previous round; otherwise, the next round of crossover and mutation continues. Finally, the work order execution order is rearranged to obtain the micro-workshop level scheduling.

[0069] Based on the work order execution order in the micro-workshop level scheduling, the status of the corresponding process node is read item by item and compared with the predetermined execution status recorded at the same work order execution location. If the status codes are inconsistent, it is determined to be an abnormal process node. Then, the equipment resource allocation quota value at the original equipment resource location of the abnormal process node is read. At the same time, the allocated quota and rated allocable capacity of other equipment resource locations within the same workshop node and the same execution time period are read. The upper limit of the equipment resource allocable quota is determined by multiplying the rated unit time processing capacity of the corresponding equipment in the execution time period by the available execution time and deducting the locked maintenance period. For example, if the rated processing capacity of the equipment is 120 pieces per hour, and there is a 1-hour locked maintenance period within an 8-hour execution time period, then the upper limit of the equipment resource allocable quota is set according to 840 pieces for 7 hours. If 650 pieces have been allocated, then the remaining allocable quota is 190 pieces, which is only available during the 7-hour execution time period. Replacement positions are selected from those corresponding to the same equipment resource type as the abnormal state process node. These are arranged in descending order of remaining allocable quotas. If the remaining allocable quotas are the same, the position with the earlier execution order in the micro-workshop level scheduling is selected. First, the equipment resource allocation quota value corresponding to the abnormal state process node is deducted from the original equipment resource position. Then, the value is reassigned to the first replacement position. If a single replacement position is insufficient, the assignment is split. For example, if the work order execution requirement is 420 units, and the first and second replacement positions have 250 and 210 units remaining respectively, 250 units are allocated first, then 170 units are allocated to the second position, until all work order execution requirements have obtained equipment resource allocation quota values. Finally, the abnormal state process node, equipment resource position, equipment resource allocation quota value, and work order execution order are written back to the micro-workshop level scheduling to obtain the power battery resource redistribution schedule.

[0070] The above are merely preferred embodiments of the present invention and are not intended to limit the present invention in any other way. Any person skilled in the art may make changes or modifications to the above-disclosed technical content to create equivalent embodiments that can be applied to other fields. However, any simple modifications, equivalent changes, and modifications made to the above embodiments based on the technical essence of the present invention without departing from the scope of the present invention shall still fall within the protection scope of the present invention.

Claims

1. An automatic numbering method for the full life cycle management of power batteries, characterized in that, Includes the following steps: Spatial dimension mapping and arrangement of battery cell material supplier names, production batch numbers, production line environmental parameter values, and transportation route mileage nodes are performed to generate a basic vector matrix. Parameter features are extracted and spatial dimension mapping and arrangement are performed on baking temperature curve values ​​and pressure node values. The result of parameter feature extraction and spatial dimension mapping is combined with the basic vector matrix to obtain the multidimensional gene battery code. The abnormal target sequence is obtained by performing target feature space mapping transformation on the diaphragm material specifications, contract manufacturing environment parameters, and shutdown date values. The time axis node drift and stretching distance of the corresponding numerical sequence in the multidimensional gene battery encoding and the abnormal target sequence are calculated to obtain the list of affected batteries. Based on the production shift count and work order status parameters in the affected battery list, a density clustering operation is performed to generate a prototype set of damaged assets. The current safety stock balance and the capacity of alternative suppliers corresponding to the set of damaged asset prototypes are used as decision variables and mixed integer linear calculation is performed to obtain the macro capacity allocation plan. Based on the corresponding workshop work order scheduling nodes and factory-level inventory transfer direction nodes within the macro capacity allocation plan, cross-mutation calculations are performed on the remaining production line scheduling days and equipment load status values ​​in the affected battery list to generate a micro workshop-level scheduling schedule. Extract the abnormal state process nodes and corresponding equipment resource allocation quota values ​​from the micro-workshop level scheduling, and reassign and replace them according to the abnormal state process nodes and corresponding equipment resource allocation quota values ​​to obtain the power battery resource reallocation schedule.

2. The automatic numbering method for the full life cycle management of power batteries according to claim 1, characterized in that, The steps for obtaining the fundamental vector matrix are as follows: Collect the names of cell material suppliers, production batch numbers, production line environmental parameter values, and transportation route mileage nodes at each corresponding node throughout the entire life cycle of the power battery. Establish a node position index according to the chronological order of the nodes throughout the power battery's life cycle. Write the cell material supplier name into the supplier dimension position according to the node it belongs to, write the production batch number into the batch dimension position according to the node it belongs to, write the production line environmental parameter value into the production line environmental parameter dimension position according to the corresponding collection node, and write the transportation route mileage node into the transportation route dimension position according to the mileage sequence in the transportation route. Align the content of each dimension at the same node position row by row, retain the corresponding dimension empty space for missing node positions, and arrange the duplicate node positions continuously according to the appearance order of the production batch number to generate a basic vector matrix.

3. The automatic numbering method for the full life cycle management of power batteries according to claim 1, characterized in that, The steps for obtaining the multidimensional gene battery code are as follows: The baking temperature curve values ​​and pressure node values ​​at corresponding nodes in each shift of the coating machine's operating cycle are captured. The baking temperature curve values ​​are segmented according to the start and end nodes of the shift. The starting temperature value, ending temperature value, peak temperature value, valley temperature value, temperature change amplitude, direction of change of continuous sampling points, and number of temperature fluctuations of the baking temperature curve values ​​corresponding to each shift are extracted. The pressure node values, pressure peak value, pressure valley value, pressure change amplitude, direction of change of continuous pressure nodes, and number of pressure fluctuations are extracted according to the corresponding nodes of the same shift. The baking temperature curve value extraction results are written into the temperature feature dimension position according to the shift node order, and the pressure node value extraction results are written into the pressure feature dimension position according to the corresponding pressure node order. The temperature feature dimension position and pressure feature dimension position under the same shift and the same node position are aligned and arranged to generate the parameter feature space mapping arrangement result. The shift node position, temperature feature dimension position, and pressure feature dimension position in the parameter feature space mapping arrangement result are read. At the same time, the power battery life cycle node position, supplier dimension position, batch dimension position, production line environmental parameter dimension position, and transportation route dimension position in the basic vector matrix are read. The splicing position is determined by the correspondence between the power battery life cycle node position and the shift node position. For node positions that cannot be matched, the original dimension position is retained. For node positions that can be matched, column splicing is performed according to the fixed arrangement order of supplier dimension position, batch dimension position, production line environmental parameter dimension position, transportation route dimension position, temperature feature dimension position, and pressure feature dimension position. The spliced ​​node positions are numbered continuously according to the chronological order of the power battery life cycle nodes to obtain the multidimensional gene battery code.

4. The automatic numbering method for the full life cycle management of power batteries according to claim 1, characterized in that, The steps for obtaining the abnormal target sequence are as follows: The multidimensional gene battery code is invoked, and the membrane material specifications, contract manufacturer processing environment parameters, and production stoppage date values ​​are read according to the corresponding positions of the abnormal nodes in the supply chain. The membrane material specifications are mapped to the target feature positions according to the material specification identifier, thickness parameters, and porosity parameters. The contract manufacturer processing environment parameters are mapped to the target feature positions according to the temperature parameters, humidity parameters, and cleanliness parameters. The production stoppage date values ​​are converted into continuous time positions according to the year, month, and day positions. The target feature positions belonging to the same abnormal node in the supply chain are arranged in a fixed order, and the abnormal nodes in different supply chains are arranged continuously according to the order of the production stoppage date values, thus generating an abnormal target sequence.

5. The automatic numbering method for the full life cycle management of power batteries according to claim 1, characterized in that, The steps for obtaining the list of affected batteries are as follows: Extract the corresponding numerical sequence of consecutive nodes before and after the time position of the abnormal node in the supply chain from the multidimensional gene battery code. Use the time position corresponding to the production stop date value in the abnormal target sequence as the alignment benchmark. Change the matching position of the corresponding numerical sequence node by node according to the adjacent time position. Record the node offset amount generated by each time position offset. Calculate the numerical difference for the position of the membrane material specification parameter, the position of the contract manufacturer processing environment parameter, and the position of the production stop date value after the node offset. Determine the local stretching position according to the direction of the continuous node difference change. Expand the matching span for the position where the difference in the same direction occurs continuously. Shorten the matching span for the position where the difference direction changes. Accumulate the node offset amount, matching span, and numerical difference formed by each matching position in the order of time position to obtain the distance measurement value. The multidimensional gene battery code positions corresponding to the distance measurement values ​​are read one by one. Each distance measurement value is compared with a preset anomaly limit. Multidimensional gene battery code positions where the distance measurement value is less than the preset anomaly limit are retained, while multidimensional gene battery code positions where the distance measurement value is greater than or equal to the preset anomaly limit are removed. Based on the retained multidimensional gene battery code positions, the corresponding production batch number is traced. The work-in-process node number and the sold battery node number under the production batch number are extracted. The duplicate work-in-process node number and the duplicate sold battery node number are uniquely retained. The affected battery list is generated by continuously arranging the production batch number, work-in-process node number, and sold battery node number according to their correspondence.

6. The automatic numbering method for the full life cycle management of power batteries according to claim 1, characterized in that, The steps for obtaining the prototype set of damaged assets are as follows: Extract the production shift count value corresponding to each work-in-process node number and sold battery node number in the affected battery list. Read the work order status parameters recorded at the same node position. Convert the production shift count value into consecutive shift positions according to the shift order. Discretize the work order status parameters according to the corresponding status positions of pending production, in production, paused, and completed. Establish a multi-dimensional feature space coordinate with the consecutive shift positions and status positions corresponding to each node. Count the number of nodes in a preset neighborhood around each multi-dimensional feature space coordinate. Use the coordinates where the number of nodes reaches a preset density limit as the aggregation starting position. Continuously absorb adjacent coordinates in the neighborhood that also reach the preset density limit. Use the position where adjacent coordinates cannot be absorbed as the aggregation boundary. Mark isolated coordinates that do not reach the preset density limit separately. Aggregate the corresponding work-in-process node numbers and sold battery node numbers according to different aggregation boundaries to generate a prototype set of damaged assets.

7. The automatic numbering method for the full life cycle management of power batteries according to claim 1, characterized in that, The steps for obtaining the macro-capacity allocation plan are as follows: Read the current safety stock balance value corresponding to the node to which each of the damaged asset prototype sets belongs, read the capacity value of the alternative suppliers according to the same node, align the current safety stock balance value according to the correspondence between the node, production shift value and work order status parameter, align the capacity value of the alternative suppliers according to the corresponding supply node of the alternative suppliers, check whether there are any missing records for the current safety stock balance value corresponding to each node, retain the empty position for the node with missing records, check whether there are any duplicate records for the capacity value of the alternative suppliers corresponding to each supply node, arrange the duplicate records according to the order of the corresponding supply cycle, so that each of the damaged asset prototype sets corresponds to the determined inventory balance position and supply capacity position, and obtain the current safety stock balance value and the capacity value of the alternative suppliers. The current safety stock balance available for each attributing node is set as the inventory allocation decision value, and the available alternative supplier capacity value available for release from each alternative supplier is set as the supply capacity allocation decision value. The inventory allocation decision value is limited to not exceeding the current safety stock balance of the corresponding attributing node, and the supply capacity allocation decision value is limited to not exceeding the capacity value of the corresponding alternative supplier. The capacity gap that each attributing node needs to fill is determined according to the number of attributing nodes in the damaged asset prototype set. The inventory allocation decision value and the supply capacity allocation decision value that satisfy the capacity gap constraint are combined item by item. Combinations that exceed the inventory balance limit, exceed the supply capacity limit, or cannot cover the corresponding capacity gap are eliminated. The remaining combinations are sorted according to the degree of capacity gap satisfaction of each attributing node. The allocation combination that can cover the capacity gap of all attributing nodes and has the smallest combination result of inventory occupancy and supply capacity call value is selected. The attributing nodes, inventory allocation decision values, and supply capacity allocation decision values ​​corresponding to the allocation combination are arranged in sequence to generate a macro capacity allocation plan.

8. The automatic numbering method for the full life cycle management of power batteries according to claim 1, characterized in that, The steps for obtaining the power battery resource reallocation schedule are as follows: Extract the workshop work order scheduling node corresponding to each affiliated node in the macro capacity allocation plan, identify the factory-level inventory transfer direction node according to the inventory transfer-out and inventory transfer-in positions in the macro capacity allocation plan, match the production batch number in the affected battery list according to the workshop work order scheduling node, locate the workshop node corresponding to the production batch number, read the remaining production line scheduling days corresponding to the work orders that have not yet been executed for each workshop node, read the equipment load status value according to the same workshop node, arrange the remaining production line scheduling days according to the execution order of the workshop work order scheduling nodes, arrange the equipment load status value according to the corresponding equipment resource position, and maintain the transfer-in and transfer-out correspondence between the factory-level inventory transfer direction node and the workshop node, to obtain the workshop work order scheduling node, the factory-level inventory transfer direction node, the remaining production line scheduling days, and the equipment load status value; The work order execution positions corresponding to each workshop work order scheduling node are arranged as candidate positions from front to back according to the remaining production line scheduling days. The inventory transfer-in workshops corresponding to the factory-level inventory transfer direction nodes are used as positions that can accept work orders. The equipment load status values ​​of different workshop nodes are compared item by item, and the workshop nodes with equipment load status values ​​that allow for the release of equipment resources are selected as the exchange positions. The corresponding work order execution positions are exchanged between different workshop work order scheduling nodes, while maintaining the correspondence between production batch numbers. The remaining production line scheduling days are reread for the exchanged work order execution positions. If the remaining production line scheduling days after the exchange exceed the corresponding remaining production line scheduling days before the exchange, the work order execution positions before the exchange are restored. If the remaining production line scheduling days do not exceed the corresponding remaining production line scheduling days before the exchange, then the work order execution position after the exchange is retained. Then, the equipment resource position where the equipment load status value exceeds the median of the load status values ​​of all equipment in the same workshop is selected, and the corresponding work order execution position is transferred to the acceptable work order position pointed to by the factory-level inventory transfer direction node and where the equipment load status value is lower than the median of the load status values ​​of all equipment in the same workshop. The work order execution position exchange and equipment resource position adjustment are performed cyclically until the remaining production line scheduling days no longer decrease. The work order execution order is rearranged according to the retained workshop work order scheduling node order, the factory-level inventory transfer direction node, and the corresponding equipment resource position to generate a micro-workshop-level scheduling schedule. The process node status corresponding to the work order execution sequence is extracted item by item. Abnormal process nodes whose status is inconsistent with the predetermined execution status in the micro-workshop level scheduling are filtered out. The equipment resource allocation quota value corresponding to each abnormal process node is read. The remaining equipment resource allocation quota values ​​within the same execution period are queried according to the workshop node to which the abnormal process node belongs. Equipment resource positions that have not reached the upper limit of the corresponding equipment resource allocation quota are selected as replacement positions. The original equipment resource allocation quota value of the abnormal process node is deducted from the original equipment resource position, and the equipment resource allocation quota value is reassigned to the replacement position according to the work order execution requirements corresponding to the abnormal process node. If the remaining allocation quota of a single replacement position is insufficient to meet the work order execution requirements corresponding to the abnormal process node, replacement positions are selected from largest to smallest according to the remaining allocation quota until all work order execution requirements corresponding to the abnormal process node receive equipment resource allocation quota values. The abnormal process nodes, equipment resource positions, equipment resource allocation values, and work order execution sequence after reassignment and replacement are written back to the micro-workshop level scheduling to generate a power battery resource reallocation schedule.

Images