Battery charging control method for battery swap cabinet based on internet of things

Abstract

This invention relates to an IoT-based battery charging control method for battery swapping cabinets. The method includes: acquiring IoT data at the cabinet end to obtain a cabinet-end operation dataset; using a single-bay constraint profiling and locking algorithm to bind the real-time battery status, historical aging characteristics, bay thermal coupling relationships, cabinet internal environmental status, and site replenishment needs within the same cycle, resulting in a single-bay chargeable constraint set; using a thermal coupling conflict cluster staggering algorithm to divide battery bays with thermal impact associations into conflict clusters and stagger fast charging peaks, resulting in a battery charging queue; using a demand gap-based reverse-engineering bay current orchestration algorithm to generate bay charging control commands based on the target battery swapping rhythm; and using an edge control and cloud strategy collaborative algorithm to complete execution status feedback, and generating the battery swapping cabinet charging control result through feedback status judgment and control output encapsulation. This invention can, to a certain extent, ensure the consistency of single-bay constraints within the same cycle, the concurrent safety of adjacent bays, and the stable response of the site replenishment rhythm.

Description

Technical Field

[0001] This invention belongs to the field of battery swapping cabinet charging control technology, specifically relating to a battery charging control method for battery swapping cabinets based on the Internet of Things. Background Technology

[0002] With the development of shared two-wheeled electric vehicles and on-demand delivery services, battery swapping cabinets are widely deployed in commercial areas, communities, office buildings, and rider rest stations. During operation, these cabinets need to charge batteries in multiple compartments concurrently and maintain a certain number of swappable batteries based on the station's demand. Because the batteries within the cabinets differ in state of charge, health, service life, and temperature response, and because different battery compartments also differ in terms of adjacency, airflow connections, and heat dissipation conditions, safety control and energy replenishment scheduling during concurrent charging across multiple compartments become critical issues in the operation of battery swapping cabinets.

[0003] In existing technologies, battery swapping cabinets typically collect battery status data, cabinet interior environmental data, and battery swapping demand data via the Internet of Things (IoT). Based on battery state of charge, temperature, voltage, current, and demand forecasts, they adjust charging plans. During charging, existing technologies also compare environmental data such as cabinet temperature and humidity with safety thresholds, adjusting charging power or switching charging modes accordingly. This type of solution enables basic monitoring, demand response, and environmental safety protection.

[0004] However, existing technologies primarily rely on overall cabinet charging plans or single-compartment environmental thresholds for control, making it difficult to uniformly convert real-time battery status data, historical aging characteristics, compartment thermal coupling relationships, cabinet environmental conditions, and site recharge requirements into single-compartment charging constraints. Furthermore, existing technologies typically limit current or stop charging only after the temperature reaches a threshold, making it difficult to proactively resolve thermal coupling conflicts between adjacent compartments, compartments with the same airflow duct, or compartments experiencing simultaneous temperature increases before the single-compartment temperature reaches the stop-charging threshold. Additionally, battery swapping demand predictions are mostly used for overall charging plan adjustments and are difficult to directly convert into dynamic charging modes, current limits, charging cutoff states of charge, delayed start configurations, and estimated swappable times for each battery compartment. Therefore, this invention proposes an IoT-based battery charging control method for battery swapping cabinets. Summary of the Invention

[0005] In view of the aforementioned existing problems, the present invention is proposed.

[0006] This invention provides a battery charging control method for battery swapping cabinets based on the Internet of Things. The aim is to at least partially solve the problems of existing technologies that mostly use the charging plan of the whole cabinet or the environmental threshold of a single compartment as the basis for control. These technologies make it difficult to uniformly convert real-time battery status data, historical aging characteristics, thermal coupling relationship of compartments, environmental status inside the cabinet and site energy replenishment needs into single-compartment charging constraints for battery compartments. Furthermore, when multiple compartments are charging concurrently, they usually rely on the ex-post current limiting or charging stop of temperature thresholds, which makes it difficult to resolve thermal coupling conflicts between adjacent compartments, compartments with the same air duct and compartments that are heated at the same time in advance. It is also difficult to directly convert the site energy replenishment needs into one or more of the following problems: dynamic charging mode, current limit, charging cutoff state of charge, delayed start time and expected battery swapping time for each battery compartment.

[0007] This invention provides the following technical solution:

[0008] In a first aspect, the present invention provides a battery charging control method for a battery swapping cabinet based on the Internet of Things, comprising:

[0009] Acquire IoT data from the cabinet and process its operational status to obtain the cabinet operational dataset;

[0010] A single-warehouse refillability constraint is constructed on the aforementioned counter-end operation dataset to obtain a single-warehouse refillability constraint set;

[0011] The charging priority is sorted according to the charging constraints of the single compartment to obtain the battery charging queue;

[0012] Configure compartment control parameters for the battery charging queue and generate compartment charging control commands;

[0013] The execution status of the compartment charging control command is collected to obtain charging execution feedback data;

[0014] The charging execution feedback data is used to determine the feedback status and obtain the charging feedback determination result.

[0015] The charging feedback determination result is encapsulated into a control output to generate the charging control result of the battery swapping cabinet.

[0016] As a preferred implementation, the construction of single-warehouse refillability constraints on the counter-end operation dataset to obtain a single-warehouse refillability constraint set includes:

[0017] The battery compartment identifiers of the cabinet-side operation dataset are merged to obtain the battery compartment operation records corresponding to each battery compartment.

[0018] The battery compartment operation records are split into state sources to obtain battery-side state records, compartment-side state records, and demand-side state records;

[0019] The rechargeability constraints are extracted from the battery-side state records to obtain the battery-side rechargeability constraints.

[0020] Thermal coupling constraints are extracted from the warehouse-side state records to obtain warehouse-side thermal coupling constraints;

[0021] The energy replenishment constraint is extracted from the demand-side state record to obtain the demand-side energy replenishment constraint;

[0022] Write the battery-side rechargeability constraint, the storage location-side thermal coupling constraint, and the demand-side energy replenishment constraint into the corresponding battery storage operation record to obtain a battery storage constraint profile.

[0023] The state of the battery compartment constraint profile is written and the version is locked to obtain a single compartment chargeable constraint set.

[0024] As a preferred embodiment, the state writing and version locking of the battery compartment constraint profile includes:

[0025] The single-compartment constraint profile locking algorithm binds the real-time battery status data, historical aging characteristics, compartment thermal coupling relationship, cabinet environment status and site recharge requirements in the battery compartment constraint profile to the same period to obtain the same period profile record.

[0026] The battery compartment identifier, battery identifier, and control cycle identifier in the same period image record are concatenated to obtain the image identifier field;

[0027] Version identifiers are generated for the portrait identifier field and the portrait records in the same period to obtain a portrait lock record;

[0028] The image lock record is written into the corresponding battery compartment constraint image to complete the construction of the single compartment rechargeable constraint set.

[0029] As a preferred embodiment, the step of sorting the charging priority of the single-compartment charging constraint set to obtain the battery charging queue includes:

[0030] The rechargeable state is filtered by the single-compartment rechargeable constraint set to obtain a sortable battery compartment set;

[0031] The sortable battery compartment set is marked with a power replenishment eligibility label to obtain a power replenishment candidate compartment set;

[0032] The set of candidate charging stations is marked with fast charging conflict flags and current limit flags to obtain a set of restricted charging stations;

[0033] The limited charging bay set is sorted according to charging eligibility, fast charging conflict and current limit to obtain a grouped charging sequence;

[0034] The grouped charging sequences are encapsulated into a queue to obtain a battery charging queue.

[0035] As a preferred embodiment, the step of sequentially arranging the limited charging bay set according to charging eligibility, fast charging conflict, and current limitation includes:

[0036] The thermally coupled conflict clusters are obtained by dividing the battery compartments with thermal impact correlation in the set of restricted energy replenishment compartments into conflict clusters using the thermally coupled conflict cluster staggering algorithm.

[0037] The fast charging candidate status of the battery compartment within the thermally coupled conflict cluster is read to obtain the conflict cluster candidate compartment record;

[0038] Fast charging eligibility is allocated to the candidate warehouse records of the conflict clusters to obtain the conflict cluster staggered peak allocation results;

[0039] The collision cluster peak shifting results are written into the corresponding group charging sequence to complete the generation of the group charging sequence.

[0040] As a preferred embodiment, configuring the compartment control parameters of the battery charging queue and generating compartment charging control commands includes:

[0041] The battery charging queue is read to obtain the compartment configuration object;

[0042] Configure the charging mode, current limit, charging cut-off charge state and delayed start for the compartment configuration object to obtain the compartment control parameter record;

[0043] The compartment control parameter records are encapsulated into instructions to generate compartment charging control instructions.

[0044] As a preferred embodiment, configuring the charging mode, current limit, charging cutoff state of charge, and delayed start for the compartment configuration object includes:

[0045] The energy replenishment execution positions are determined by the demand gap reverse-engineering sub-compartment current orchestration algorithm for the sub-compartment configuration object, resulting in a set of energy replenishment execution positions.

[0046] Dynamic charging mode configuration is performed on the set of energy replenishment execution positions to obtain dynamic charging mode records;

[0047] The dynamic charging mode record is configured with current limiting, charging cutoff state of charge and expected swappable time to obtain the energy replenishment control parameter record.

[0048] For sub-compartment configuration objects that are not written into the aforementioned energy replenishment execution unit set but have charging permission, perform maintenance charging or postpone charging configuration to obtain non-urgent control parameter records;

[0049] The generation of the compartment control parameter record is completed based on the energy replenishment control parameter record and the non-urgent control parameter record.

[0050] As a preferred embodiment, the step of sampling the execution status of the compartment charging control command to obtain charging execution feedback data includes:

[0051] The strategy version binding is performed on the compartment charging control command to obtain the strategy version binding record;

[0052] The position distribution record is generated by binding the strategy version to the record.

[0053] A baseline record is generated from the aforementioned warehouse allocation record to obtain the instruction execution baseline record;

[0054] The execution status record is obtained by collecting the status of the instruction execution baseline record;

[0055] The execution status records are associated with warehouse locations to obtain status association records;

[0056] The state association records are collected and processed to obtain charging execution feedback data.

[0057] As a preferred embodiment, the strategy version binding of the compartment charging control command includes:

[0058] The consistency of the cloud policy version identifier, cloud policy summary, profile locking record and control cycle identifier carried by the compartment charging control command is compared by the edge control and cloud policy collaborative algorithm to obtain the multi-dimensional consistency distance.

[0059] Based on the multidimensional consistency distance, the compartment charging control command is divided into full execution state, degraded execution state, or rejection execution state to obtain the strategy version matching result.

[0060] For the sub-compartment charging control instructions that belong to the full execution state or the degraded execution state in the strategy version matching results, perform edge end version registration to obtain edge end strategy registration records;

[0061] The edge-end policy registration record is bound to the corresponding battery compartment to obtain the compartment distribution record;

[0062] The command execution baseline record is generated based on the position issuance record.

[0063] As a preferred embodiment, the step of determining the feedback status of the charging execution feedback data to obtain a charging feedback determination result includes:

[0064] The charging execution feedback data is read to obtain the feedback status record;

[0065] The execution deviation of the feedback status record and the corresponding compartment charging control command is extracted to obtain the feedback deviation record;

[0066] The feedback deviation records are categorized by state to obtain feedback state labels;

[0067] The feedback status markers are summarized to obtain the charging feedback determination result.

[0068] Secondly, this application provides an electronic device, comprising:

[0069] A processor; and a memory for storing processor-executable instructions; wherein the processor is configured to invoke instructions stored in the memory, which, when executed by the processor, cause the electronic device to perform the method disclosed in the first aspect above.

[0070] Thirdly, this application provides a computer-readable storage medium, which...

[0071] The medium stores computer program instructions that, when executed by a processor, cause the method disclosed in the first aspect above to be implemented.

[0072] The battery charging control method for battery swapping cabinets based on the Internet of Things provided by this invention has the following beneficial effects:

[0073] 1. Real-time battery status data, historical aging characteristics, thermal coupling relationship of battery compartments, environmental status inside the cabinet, and site charging requirements are uniformly converted into single-compartment chargeable constraints for battery compartments. This enables the cabinet to make a unified judgment on a single compartment based on data from the same period when multiple compartments are charging concurrently. It can identify multiple mild abnormal accumulation situations and maintain the consistency of single-compartment constraints within the same period even when batteries in the cabinet have different service life and health status.

[0074] 2. Before issuing the fast charging command, conflict clusters are divided and fast charging qualifications are staggered for battery compartments with thermal impact. The heat residue and heat load inheritance across control cycles are taken into account in the staggered scoring. This ensures that battery compartments that are fast charging at the same time are isolated from each other in terms of space and temperature rise trajectory, and the fast charging load rotates naturally across the cycle scale. To a certain extent, this can reduce the repeated derating and charging stop caused by the superposition of temperature rise in adjacent compartments and the local accelerated aging caused by the concentration of fast charging tasks in a few battery compartments.

[0075] 3. The data on site replenishment, the number of swappable batteries in the cabinet, and the estimated swappable time of each battery compartment are back-calculated on the control cycle scale to determine the dynamic charging mode, current limit, charging cutoff state of charge, delayed start time, and estimated swappable time of each battery compartment. The multi-dimensional consistency distance is used as the hierarchical collaboration criterion between the edge and the cloud, so that the cabinet can configure fast charging intensity according to the target swappable rhythm during peak site replenishment periods, and charging in the cabinet can still continue to operate when the cloud strategy version is switched. Attached Figure Description

[0076] Figure 1 An exemplary flowchart of a battery charging control method for a battery swapping cabinet based on the Internet of Things provided in an embodiment of the present invention.

[0077] Figure 2 A comparison diagram showing the effect of the IoT-based battery charging control method for battery swapping cabinets provided in this embodiment of the invention with existing technologies. Detailed Implementation

[0078] To make the technical means, creative features, and achieved objectives and effects of this invention easier to understand, the invention is further described below with reference to specific embodiments. However, the following embodiments are merely preferred embodiments of this invention and not all of them. Other embodiments obtained by those skilled in the art based on the embodiments described herein without creative effort are all within the protection scope of this invention. Unless otherwise specified, the experimental methods in the following embodiments are conventional methods, and the materials and reagents used in the following embodiments are commercially available unless otherwise specified.

[0079] Example 1 combined Figure 1 The following is an exemplary flowchart of a battery charging control method for a battery swapping cabinet based on the Internet of Things (IoT). The specific implementation steps are as follows:

[0080] Acquire IoT data from the cabinet and process its operational status to obtain the cabinet operational dataset;

[0081] The aforementioned cabinet-side IoT data refers to the data set collected by the cabinet-side controller within the current control cycle from the battery compartment node, charging control node, cabinet environment node, local historical database, and site service interface through the cabinet bus, storage interface, and gateway communication module. The data set includes at least five categories according to the data source identifier: real-time battery status data, charging execution data, cabinet environment status data, historical aging characteristic data, and site recharging data. The arrival time of each piece of data at the cabinet-side controller is marked by the data reception time identifier.

[0082] The aforementioned operational status aggregation processing refers to the process by which the cabinet-side controller uses the control cycle identifier, cabinet identifier, and battery compartment identifier as merging keys to write the cabinet-side IoT data into the battery-side field, compartment-side field, and demand-side field according to the data source identifier and field purpose.

[0083] The cabinet-side operation dataset refers to a data set obtained after the operation status aggregation processing, which is encapsulated according to the control cycle identifier, with the battery compartment identifier as the primary key, and includes battery-side fields, compartment-side fields, demand-side fields, common status fields, data missing markers, and duplicate record indexes.

[0084] First, a low-voltage control compartment is set up inside the battery swapping cabinet. The cabinet-side controller is fixedly installed on the guide rail mounting base inside the low-voltage control compartment, and the power port of the cabinet-side controller is connected to the low-voltage power supply module of the battery swapping cabinet. Then, the first communication port of the cabinet-side controller is connected to each battery compartment node, the second communication port of the cabinet-side controller is connected to each charging control node, the third communication port of the cabinet-side controller is connected to the cabinet environment node, the storage interface of the cabinet-side controller is connected to the local historical database, the network communication port of the cabinet-side controller is connected to the gateway communication module, and the gateway communication module establishes a network communication connection with the site service interface.

[0085] After the connection is established, the cabinet controller registers each battery compartment node, charging control node, cabinet environment node, local history database, and site service interface, generating a node address table. The node address table records the cabinet identifier, node type, node address, corresponding battery compartment identifier, corresponding battery identifier, and data source identifier. Specifically, battery compartment nodes and charging control nodes are bound to their corresponding battery compartment identifiers, cabinet environment nodes are bound to their cabinet identifiers, the local history database is bound to its corresponding battery identifier, and the site service interface is bound to its cabinet identifier.

[0086] After generating the node address table, the cabinet controller reads the cabinet configuration table and the compartment thermal coupling basic table. The cabinet configuration table is written by the cabinet controller before the battery swapping cabinet is put into operation and is updated synchronously when the cloud strategy version is updated. The cabinet configuration table records the cabinet type identifier, battery model identifier, control cycle length, fast charging temperature upper limit, charging temperature upper limit, equalization threshold, insulation resistance safety lower limit, aging threshold, health lower limit, fast charging temperature rise warning threshold, thermal diffusion isolation threshold, non-urgent charging state threshold, derating step size, current lower limit, and safe charging time window. The thresholds related to battery safety are registered according to the battery manufacturer's factory specifications and the battery swapping cabinet safety strategy table; the thresholds related to cabinet thermal coupling are registered according to the cabinet type thermal test results; and the time window related to site charging is registered according to the statistical results of the battery swapping interval of historical battery swapping orders during peak hours.

[0087] The compartment thermal coupling baseline table records cabinet identification, battery compartment identification, compartment coordinates, adjacent compartment identification, air duct number, basic thermal influence coefficient, and structural distance. The compartment coordinates are registered in the cabinet design documents; adjacent compartment identifications are generated when the compartment coordinates are spaced one compartment unit horizontally or vertically; the air duct number is registered in the cabinet air duct structure; and the basic thermal influence coefficient is registered in the cabinet thermal test results. When generating compartment thermal coupling relationships, the cabinet controller reads the compartment thermal coupling baseline table and updates the thermal influence association markers based on the current compartment temperature status, cooling fan speed, and cabinet door opening / closing status.

[0088] The cabinet controller reads the node address, corresponding battery compartment identifier, corresponding battery identifier, and data source identifier from the node address table according to the control cycle, and generates a data reading task. The data reading task records the control cycle identifier, cabinet identifier, node address, corresponding battery compartment identifier, corresponding battery identifier, and data source identifier. The cabinet controller sends a read command to the corresponding node based on the data reading task; the battery compartment node returns real-time battery status data, the charging control node returns charging execution data for the corresponding battery compartment, the cabinet environment node returns cabinet internal environment status data, the local history database returns historical aging characteristic data associated with the corresponding battery identifier, and the site service interface returns site recharging data.

[0089] The historical aging characteristic data is stored in the local historical database according to the battery identifier, and includes at least the internal resistance increment, health status, cumulative number of charging cycles, and the time of the last complete charge end. Among them, the internal resistance increment is generated based on the difference between the ratio of voltage change to current change during the constant current charging stage of the same battery and the factory internal resistance reference value. The health status is obtained from the capacity retention rate field returned by the battery management system or updated by the local historical database based on the ratio of cumulative charging capacity to nominal capacity.

[0090] When the battery identifier changes, the cabinet controller rereads the historical aging feature data according to the new battery identifier; when the historical aging feature data is not read, the cabinet controller writes a historical feature missing mark and writes the corresponding battery compartment into the restricted state in the subsequent single compartment charging constraint construction.

[0091] After receiving data from each node, the cabinet controller matches the returned data with the corresponding data reading task and writes the successfully matched data into the data receiving buffer. During writing, the cabinet controller adds a control cycle identifier, cabinet identifier, battery compartment identifier, battery identifier, data source identifier, and data receiving time identifier to each piece of returned data, forming cabinet-side IoT data. Return data that is not matched with a data reading task is written to the unmatched buffer and is not included in the operational status aggregation processing for the current control cycle.

[0092] When a data reading task does not receive data returned by the corresponding node within the return deadline of this control cycle, the cabinet controller generates a missing placeholder record based on the data reading task, and writes the control cycle identifier, cabinet identifier, battery compartment identifier, battery identifier, data source identifier, and data missing marker into the missing placeholder record; the missing placeholder record enters the running status aggregation processing of this control cycle, and is used to determine the data integrity status of the corresponding battery compartment when constructing the rechargeable constraints of the subsequent single compartment.

[0093] After obtaining the IoT data from the cabinet, the cabinet controller performs operational status aggregation processing on the IoT data. Specifically, it first performs periodic merging of the IoT data according to the control cycle identifier to obtain a periodic data area; then, it performs cabinet merging of the periodic data area according to the cabinet identifier to obtain a cabinet record; subsequently, it performs compartment merging of the cabinet records according to the battery compartment identifier, writing data with the same battery compartment identifier into the same battery compartment record, and writing data with only a cabinet identifier but no battery compartment identifier into the common status field of the cabinet record.

[0094] After the battery compartments are consolidated, the cabinet controller aggregates the battery compartment records and common status fields according to the data source identifier. It writes the real-time battery status data into the battery status field, the charging execution data into the charging status field, the cabinet environment status data into the compartment status field, the historical aging characteristic data into the historical characteristic field, and the site replenishment data into the demand status field. In the case of duplicate returned data under the same control cycle, the same cabinet identifier, the same battery compartment identifier, the same battery identifier, and the same data source identifier, the cabinet controller retains the returned data with the later reception time according to the data reception time identifier and writes the replaced returned data into the duplicate record index.

[0095] After completing the above processing, the counter controller encapsulates the data in the periodic data area that has completed the merging of counters, storage locations, and status, as well as the corresponding data missing markers, missing placeholder records, and duplicate record indexes, into a counter operation dataset according to the control cycle identifier; the counter operation dataset serves as the input for the subsequent construction of single-storage refillability constraints.

[0096] In this embodiment, the algorithm parameters registered in the counter configuration table adopt the following preferred values: envelope sharpness coefficient ranges from 4 to 12; temperature rise aging coupling coefficient ranges from 0.05 to 0.20; demand-driven relaxation coefficient ranges from 0.02 to 0.10; safety threshold lower bound is registered as 0.6 to 0.8 of the value range of each normalized term; thermal coupling spatial attenuation scale ranges from 1.5 to 3.0 of the number of warehouse spacings; synchronization evaluation window ranges from 5 to 10 control cycles; thermal coupling conflict baseline is set at 0.4 to 0 of the maximum value of the basic thermal influence coefficient. The following parameters are specified: 0.6 times the registration value; risk sensitivity threshold tightening coefficient ranging from 0.10 to 0.30; residual heat penalty coefficient ranging from 0.2 to 0.5; heat load inheritance coefficient ranging from 0.1 to 0.3; safe energy replenishment time window ranging from 5 to 15 minutes based on the peak battery swapping rhythm of the site; risk attenuation coefficient ranging from 1 to 3; gap amplification coefficient ranging from 0.5 to 1.5; thermal safety attenuation index ranging from 1 to 2; graded execution threshold ranging from 0.15 to 0.30; arbitration time window registered as 2 to 5 times the control cycle length. The specific values ​​of each of the above parameters are written into the cabinet configuration table by the cabinet controller before commissioning and are updated synchronously when the cloud strategy version is updated.

[0097] A single-warehouse refillability constraint is constructed on the aforementioned counter-end operation dataset to obtain a single-warehouse refillability constraint set;

[0098] The single-cell chargeability constraint construction refers to the set of constraints on whether a single battery cell can be charged, with what current, and in what charging mode within the current control cycle. It consists of three categories: battery-side chargeability constraints, cell-side thermal coupling constraints, and demand-side energy replenishment constraints.

[0099] The single-cell chargeable constraint set refers to the data set obtained by encapsulating the single-cell chargeable constraints according to the battery cell identifier. It includes the battery cell constraint profile corresponding to each battery cell, and each battery cell constraint profile consists of battery-side chargeable constraints, cell-side thermal coupling constraints, demand-side energy replenishment constraints, risk envelope values, and profile locking records.

[0100] After obtaining the cabinet-side operation dataset, the cabinet-side controller establishes a constraint construction task for the current control cycle using the control cycle identifier and cabinet identifier, and writes the cabinet-side operation dataset into the constraint construction cache. The cabinet-side controller uses the battery compartment identifier as the merge key and scans the data in the constraint construction cache in ascending order of battery compartment identifier: when data with the same battery compartment identifier is encountered, it is written to the battery-side field, compartment-side field, or demand-side field of the corresponding battery compartment operation record according to the data source identifier; when more than two data entries appear under the same source identifier, they are written from first to last according to the data reception time identifier, with the later-arriving data overwriting the earlier-arriving field, and the overwritten earlier-arriving field along with its data reception time identifier is written to the duplicate record index. For data that only has a cabinet identifier and represents the overall status of the cabinet, it is attached to the common status field of each battery compartment operation record under the same cabinet according to the control cycle identifier and cabinet identifier; for data with missing data markers, the missing data markers are written to the corresponding battery compartment operation record along with the corresponding field.

[0101] The data missing marker is used to indicate that the corresponding source field has not obtained a valid return value within the current control cycle, but it does not change the control cycle identifier and the data reception time identifier of the acquired field in the battery compartment operation record; when the local data missing marker exists but the acquired five types of heterogeneous data meet the same cycle binding condition, the cabinet controller allows the battery compartment to continue to enter the profile locking process and writes the constraint corresponding to the missing field into the restricted state; when the data reception time identifier does not meet the same cycle binding condition, the cabinet controller does not generate a profile locking record and writes the corresponding battery compartment into the freshness abnormal record.

[0102] After obtaining the battery compartment operation record, the real-time battery status data, charging execution data, and historical aging characteristic data in the battery compartment operation record are written into the battery-side status record according to the data source identifier and field usage; the compartment temperature status, adjacent compartment status, air duct association status, and cabinet environment status obtained by attaching the common status field are written into the compartment-side status record; and the site replenishment data, the current number of available batteries, and the replenishment demand status within the target control cycle are written into the demand-side status record.

[0103] Then, chargeability constraints are extracted from the battery-side state records. Specifically, based on the state of charge, battery temperature, individual cell voltage difference, and insulation resistance readings in the real-time battery state data, combined with the internal resistance increment and health status in the historical aging characteristics returned from the local historical database, the charging permission state, current limit state, and charging mode limit state of the battery compartment are determined respectively: when the battery temperature is higher than the fast charging temperature limit registered in the cabinet configuration table but not higher than the charging temperature limit, or the individual cell voltage difference is higher than the equalization threshold registered in the cabinet but has not triggered the prohibition of charging conditions, the charging permission state is written to the restricted permission state, the current limit state is written to the derating current state, and the charging mode limit state is downgraded from the fast charging permission state to the normal charging state; when the battery temperature is higher than the charging temperature limit, or the insulation resistance is lower than the safety lower limit registered in the cabinet, the charging permission state is written to the prohibition of charging state, and the current limit state is written to the zero current state.

[0104] When the internal resistance increment in the historical aging characteristics is higher than the aging threshold or the health status is lower than the health lower limit, the charging permission status is written to the restricted permission status, the current limit status is written to the derating current status, and the charging mode limit status is reduced from the normal charging status to the maintenance charging status.

[0105] When the battery temperature is not higher than the fast charging temperature limit, the single-cell voltage difference is not higher than the equalization threshold, the insulation resistance is not lower than the safety lower limit, the internal resistance increment is not higher than the aging threshold, and the health status is not lower than the health lower limit, the charging permission status is written to the allowed charging status, the current limit status is written to the rated current status, and the charging mode limit status is written to the fast charging permission status. After completing the above writing, the battery-side chargeable constraints are obtained.

[0106] When the current battery compartment simultaneously meets two or more of the above triggering conditions, the conditions for prohibiting charging, maintenance charging, and derating charging are determined and written into the file in the order of prohibiting charging, maintenance charging, and derating charging, and the remaining triggering conditions are written into the restricted reason field as auxiliary reasons.

[0107] After obtaining the battery-side charging constraints, thermal coupling constraints are extracted from the compartment-side state records. Specifically, the thermal impact association between the current battery compartment and other battery compartments is determined according to the states of adjacent compartments. Specifically, the cabinet controller reads the compartment topology table from the cabinet configuration table. The compartment topology table records the row number, column number, and air duct number of each battery compartment in the cabinet coordinate system. The condition for determining structural adjacency is that the difference between the row number and column number of two compartments in the compartment topology table is not greater than one, and the condition for determining the same air duct influence range is that the two compartments belong to the same air duct. When the current battery compartment and at least one other battery compartment satisfy the structural adjacency relationship or the same air duct influence range in the compartment topology table, a thermal impact association mark is registered in the compartment-side state record corresponding to the current battery compartment, and the battery compartment identifiers of the other battery compartments that satisfy any of the above conditions with the current battery compartment are written into the thermal impact association compartment list.

[0108] The fast charging conflict status is determined based on the temperature status of the battery compartment and the environmental status inside the cabinet: when the current battery compartment has a thermal impact associated marker and the temperature status of the battery compartment is higher than the fast charging temperature rise warning threshold registered in the cabinet configuration table, it is further determined whether the adjacent battery compartments that have a thermal impact associated with the current battery compartment have fast charging configuration in the previous control cycle: if they do, the fast charging conflict status is written to the conflict pending delay status; if they do not, the conflict pending rate reduction status is written.

[0109] When there are no thermal impact associated markers in the current battery compartment or the compartment temperature is not higher than the fast charging temperature rise warning threshold, a conflict-free state is written. Based on this, the thermal diffusion isolation state is determined according to the compartment temperature and the internal environment.

[0110] When the temperature of the storage compartment is higher than the thermal diffusion isolation threshold registered in the cabinet configuration table, write the thermal diffusion isolation status; otherwise, write the no-isolation status. After completing the above writing, the thermal coupling constraint on the storage compartment side is obtained.

[0111] After obtaining the thermal coupling constraints on the storage location side, the energy replenishment constraints are extracted from the demand-side status records. Specifically, the cabinet-level energy replenishment gap is determined according to the difference between the target available battery count and the current swappable battery count in the site energy replenishment data; when the cabinet-level energy replenishment gap is greater than zero and the current battery-side charging constraint of the battery storage compartment is written to the allowed charging state or restricted permission state, the energy replenishment participation state of the battery storage compartment is written to the participation in energy replenishment state.

[0112] When the cabinet-level power replenishment gap is not greater than zero or the current battery compartment has a charging permit but its state of charge is higher than the non-urgent threshold registered at the cabinet, write the non-urgent status.

[0113] When there is missing station replenishment data, or when the target available battery count returned by the station's business interface within the current control cycle changes beyond the target jitter threshold registered in the cabinet configuration table compared to the previous control cycle, the station's replenishment rhythm is determined to be unstable, and the replenishment participation status of the current battery compartment is written to the replenishment pending state. After completing the above writing, the demand-side replenishment constraint is obtained.

[0114] The target jitter threshold is registered in the cabinet configuration table as a percentage of the target available battery count, with a value range of 10% to 30%. When the change in the target available battery count relative to the previous control cycle exceeds this percentage, it is determined that the station's power replenishment rhythm is unstable.

[0115] After obtaining the battery-side charging constraints, the storage space-side thermal coupling constraints, and the demand-side energy replenishment constraints, the above three types of constraints are written into the corresponding battery storage operation record according to the battery storage identifier, using the same control cycle identifier, cabinet identifier, battery storage identifier, and battery identifier, thus obtaining the battery storage constraint profile.

[0116] After obtaining the battery compartment constraint profile, the battery compartment constraint profile is state-written and version-locked. Specifically, a single-compartment constraint profile locking algorithm is used to bind the real-time battery status data, historical aging characteristics, compartment thermal coupling relationship, cabinet environmental status, and site recharging requirements in the battery compartment constraint profile to the same cycle. Before performing the same-cycle binding, according to the arrival time registered in the data reception time identifier of each of the five types of heterogeneous data, the difference between the latest arrival time and the start time of the current control cycle is taken as the same-cycle data lag, and this lag is compared with the freshness tolerance upper limit registered in the cabinet configuration table. For comparison, the calculation formula for the same-period binding validity indication is as follows:

[0117] ,

[0118] in, This indicates the validity of the same-cycle binding of the i-th battery compartment within control cycle t; t represents the start time of the current control cycle, and i represents the battery compartment index. Indicates real-time battery status data, Indicating historical aging characteristics, Indicates the thermal coupling relationship of positions. Indicates the state of the environment inside the cabinet. This indicates the arrival time of each site's power replenishment demand, as registered in the data reception time identifier. { } indicates taking the maximum value of the lag of the above five types of fields relative to the start time of the current control cycle; This represents a conditional indicator function, which takes a value of 1 if the condition is true and zero if the condition is false. This indicates the upper limit of freshness tolerance, which is registered in the counter configuration table according to one-quarter of the control cycle length.

[0119] Among them, when When the value is one, the five types of heterogeneous data constitute an effective same-period profile, which is then used to generate the subsequent risk envelope.

[0120] when When the value is zero, the corresponding battery compartment will not be included in the current control cycle's profile locking, and the data with lag exceeding the limit among the five types of heterogeneous data will be locked. Write the field with abnormal freshness to the record.

[0121] for For a battery compartment with a value of one, a risk envelope value is further generated to characterize its overall adequacy. The formula for calculating the risk envelope value is as follows:

[0122] ,

[0123] in, λ represents the risk envelope value of the i-th battery compartment within the control period t; λ represents the envelope sharpness coefficient, which is registered in the cabinet configuration table. In the equation k=1,…,5, the normalization terms generated by the battery temperature rise rate, internal resistance increment, battery health status, remaining safe service life, and battery alarm status are respectively. The value range of the normalization terms is from zero to one, where zero indicates that the corresponding state has not triggered the constraint, and one indicates that the corresponding state has reached the constraint boundary. The source item weights related to the state are jointly determined by the missing data flag and the confidence level corresponding to the normalization item. When the confidence level is full, it is taken as one, and when there are missing data, it is reduced according to the proportion of missing fields. This represents the coupling correction term between temperature rise and aging, used to reflect the physical coupling between high temperature accelerating aging and aging intensifying heat generation. This represents the temperature rise-aging coupling coefficient, which is registered in the cabinet configuration table. This indicates a demand-driven relaxation term, which is only valid if all normalized terms are not higher than the lower bound of the safety threshold registered at the counter. Enabled at any time; This represents the demand-driven relaxation coefficient; This represents the normalized urgency of refueling calculated from the site refueling data.

[0124] After obtaining the risk envelope value, a version identifier for the same period profile record is generated using the following formula:

[0125] ,

[0126] in, This indicates the version identifier of the same period image of the i-th battery compartment within the control period t; This represents the function for calculating the summary. The image identifier field is formed by concatenating the battery compartment identifier, battery identifier, and control cycle identifier in sequence. This represents a profile record for the same period, which is obtained by writing real-time battery status data, historical aging characteristics, thermal coupling relationship of the storage location, environmental status inside the cabinet, and site energy replenishment needs in the order of the fields. This indicates that q is a quantization operator for a specific point. q is registered by the data type of the field in the counter configuration table. The status field takes a lower bit width and the risk envelope value takes a higher bit width, so that the perturbation of the decimal places of the floating-point data will not change the version identifier. Indicates by The lag of data in the same period obtained from the calculation; This indicates the version number of the current cabinet configuration table, so that any upgrade to the cabinet configuration table will naturally cause a change in the version identifier; || indicates a field concatenation operation.

[0127] In this embodiment, the digest calculation function uses the SHA-256 algorithm, with an output bit width of 256 bits, thus the version identifier in the image locking record... The field length is 256 bits; the length of the profile lock record registered in the counter configuration table is registered according to the above 256 bits, which serves as the basis for determining the field integrity verification in the subsequent warehouse control parameter configuration process.

[0128] After the version identifier is generated, the version identifier, along with the same-period profile record, the same-period binding validity indicator, and the risk envelope value, are written into the battery compartment constraint profile to obtain the profile lock record; for this control period Battery compartment constraint profiles that cannot generate profile locking records due to zero values ​​or missing control cycle identifiers, cabinet identifiers, or battery compartment identifiers are written into constraint construction anomaly records and do not enter the single-compartment chargeable constraint set for this control cycle. For battery compartment constraint profiles that can generate profile locking records but only have partial data missing markers, the corresponding missing state is written into the restricted state, and the battery compartment constraint profile is continued to be written into the single-compartment chargeable constraint set.

[0129] After completing the status writing and version locking of each battery compartment constraint profile within this control cycle, the constraint profiles of each battery compartment are encapsulated according to the battery compartment identifier to obtain a single-compartment chargeable constraint set. The single-compartment chargeable constraint set serves as the input for the charging priority sorting of the single-compartment chargeable constraint set in the next step.

[0130] In one embodiment, the single-warehouse constraint profile locking algorithm further includes:

[0131] In the actual operation of the battery swapping cabinet, real-time battery status data is returned from the battery compartment node via the cabinet bus, historical aging characteristics are returned from the local historical database via the storage interface, compartment thermal coupling relationship and cabinet environment status are jointly returned by the cabinet environment node and adjacent compartment nodes, and site power replenishment demand is returned from the site service interface via the gateway communication module. The links through which the above five types of heterogeneous data pass differ in physical medium, transmission protocol, sampling cycle and interface response delay, so that although the five types of data are requested within the same control cycle, they often arrive at the cabinet controller at different times.

[0132] When riders' concentrated battery swapping causes delays in the station's business interface response, or when the local historical database is performing aging feature writing and causing storage interface latency, the aforementioned link differences will be further amplified, resulting in five types of data that are "nominally in the same period but actually span multiple periods." In this scenario, if the profile is still directly assembled according to the order in which the fields are complete, mismatches will occur, such as matching the fast charging conflict status of the current control period with the historical aging features of the previous control period, or matching the real-time battery status data of the current control period with the station's charging needs of the previous control period. This will lead to distortion of the input for subsequent charging priority ranking.

[0133] The single-warehouse constraint profile locking algorithm compares the data reception time identifiers of the five types of heterogeneous data before profile writing. It only allows the data to be bound in the same period if the lag of the latest arriving field relative to the start time of the current control period does not exceed the freshness tolerance limit. This eliminates cross-period mismatched battery warehouse profiles from the source, ensuring that the battery warehouse constraint profiles read in subsequent steps always correspond to the real state within the same control period.

[0134] Addressing the issue of varying service life and heat dissipation conditions among batteries of the same model within battery swapping cabinets, different battery compartments within the cabinet often exhibit multiple minor anomalies across five dimensions—temperature rise rate, internal resistance increment, battery health status, remaining safe service life, and battery alarm status—within the same control cycle. These anomalies may not trigger a single-threshold charging stop threshold. For example, a battery might have a slightly higher temperature rise rate, a slightly larger internal resistance increment, and a slightly lower health status, but none of these individual indicators exceed the single-threshold charging stop criteria commonly used in existing technologies. In this scenario, battery compartments judged using the single-threshold method will still be considered normal and continue to perform fast charging tasks. This leads to the accumulation and amplification of multiple minor anomalies, accelerating battery aging and even inducing potential thermal runaway.

[0135] The single-compartment constraint profiling and locking algorithm normalizes the five dimensions mentioned above during the profiling stage and weights them using envelope sharpness coefficients. This ensures that when any dimension approaches the constraint boundary, its contribution to the risk envelope value becomes dominant, while when all dimensions are low, the overall value is suppressed. Simultaneously, the algorithm explicitly registers a coupling correction term between the temperature rise rate and the battery health state, reflecting the physical coupling relationship between high temperature accelerating aging and the greater heat generation during charging of aged batteries. This allows these mutually amplified states to be identified jointly in the profiling, rather than being submerged in two separate mild anomaly criteria. Therefore, battery compartments that have been in a state of multiple mild anomaly accumulation for a long time can be explicitly marked during the profiling stage, preventing them from being included in the fast charging candidate list.

[0136] This solution addresses the common operational scenario of battery swapping stations during peak charging periods where "safety margin is sufficient but charging slots are scarce." Specifically, if all five normalized risk parameters for multiple battery compartments within the station are within the lower safety threshold, while the charging gap reported by the station's business interface remains greater than zero and rider waiting queues continue to accumulate, the single-compartment constraint profile locking algorithm introduces a demand-driven relaxation item registered in the station's configuration table when generating the risk envelope value. This item is only activated when all normalized parameters are within the lower safety threshold, and a limited downward adjustment is applied to the risk envelope value based on the normalized charging urgency calculated from the station's charging data. Therefore, battery compartments within the overall safety margin experience a corresponding decrease in risk envelope value as charging urgency increases, making them more likely to be included in fast charging candidates in subsequent charging priority ranking. Conversely, once any normalized parameter reaches or exceeds the lower safety threshold, the demand-driven relaxation item automatically disables, and the risk envelope value is no longer adjusted by the station's charging urgency, thus ensuring that the station's charging pressure does not translate into a concession on the single-compartment safety boundary. This mechanism enables the algorithm to release limited capacity during peak energy replenishment periods without compromising the safety boundary of a single warehouse, thus avoiding the deteriorating path of "the greater the energy replenishment pressure, the more aggressive the charging of a single warehouse" in existing technologies.

[0137] For common version switching scenarios when online policy distribution is carried out between the cabinet controller and the cloud policy center, such as cloud policy table upgrades, cabinet configuration table upgrades, or changes in the correspondence between battery compartment identifiers and battery identifiers due to battery replacement, if the profile version identifier is directly generated from the original floating-point data, the slight disturbance of the last bit of the floating-point data during the data collection process will cause the version identifier to change frequently, resulting in a "pseudo-inconsistency" between the edge policy registration and the cloud policy version. This causes the warehouse-level control commands that are normally distributed in subsequent steps to be incorrectly rejected for execution.

[0138] Conversely, if the version identifier is simply generated by concatenating the battery compartment identifier, battery identifier, and control cycle identifier, the version identifier will not change in scenarios such as upgrades to the cabinet configuration table or changes in the risk envelope calculation method. This will cause subsequent steps to fail to recognize that the strategy method has changed and misuse instructions issued under the old method.

[0139] In the version identifier generation stage, the single-warehouse constraint profile locking algorithm adopts a fixed-point quantization method that registers the status field and risk envelope value in the profile record according to the data type of the field. The status field takes a lower bit width and the risk envelope value takes a higher bit width, thereby shielding the floating-point last bit disturbance and retaining meaningful state changes.

[0140] Simultaneously, the version number of the counter configuration table and the lag amount of the latest arrival field are explicitly written into the version identifier. This ensures that the version identifier changes automatically when any of the following occurs: counter configuration table upgrade, quantitative caliber change, or change in the freshness of binding within the same period. Therefore, the profile version identifier will not experience spurious changes due to floating-point last-digit disturbances, nor will it be missed when the policy caliber has changed. This supports the collaborative determination between the edge and cloud in subsequent steps regarding policy version, profile locking records, and control period identifiers.

[0141] By determining the validity of the same-cycle binding, generating multi-source coupling risk envelopes, and generating robust version identifiers in the above scenarios, the single-compartment constraint profile locking algorithm integrates the original separate criteria of "temperature-based stop, current-based drop, and mode-based determination" that were performed serially by the cabinet based on single-field thresholds into a same-cycle profile locking record for the battery compartment. This allows subsequent charging priority sorting, battery charging queue generation, compartment control parameter configuration, and execution status feedback to directly use the profile locking record as input. Furthermore, in the edge control and cloud strategy collaboration algorithm, the profile locking record is used as one of the criteria for determining multi-dimensional consistency distance. Thus, in the actual operation of concurrent charging in multiple compartments within the cabinet, the algorithm achieves same-cycle consistency of single-compartment constraints, comprehensive identifiability of multi-source risks, and robust comparability of strategy versions.

[0142] The charging priority is sorted according to the charging constraints of the single compartment to obtain the battery charging queue;

[0143] The charging priority sorting refers to the process of screening, marking, and arranging each battery compartment according to the image locking record, risk envelope value, energy replenishment participation status, fast charging conflict status, thermal diffusion isolation status, and current limit status in the constraint profile of each battery compartment in the single compartment chargeable constraint set, so that each battery compartment belongs to one of the priority energy replenishment group, derating charging group, delayed charging group, or non-charging group in the current control cycle.

[0144] The battery charging queue refers to a queue file encapsulated in the order of priority charging, priority charging group, derated charging group, delayed charging group, and non-charging group after being sorted by the charging priority. Each record corresponds to a battery compartment and carries a control cycle identifier, cabinet identifier, battery compartment identifier, battery identifier, profile lock record, risk envelope value, charging qualification mark, fast charging conflict mark, thermal diffusion isolation mark, current limit mark, intra-cluster peak shifting score, conflict cluster peak shifting result, queue grouping mark, and queue position identifier.

[0145] After obtaining the single-compartment chargeable constraint set, the cabinet controller reads the single-compartment chargeable constraint set corresponding to the current control cycle from the charging priority sorting input buffer, and sequentially reads the image locking record, risk envelope value, battery-side chargeable constraint, compartment-side thermal coupling constraint and demand-side energy replenishment constraint in each battery compartment constraint image according to the battery compartment identifier.

[0146] For battery compartment constraint images that are not written to the image locking record, or whose binding validity indication in the same period in the image locking record is zero, write them to the sorting anomaly record and do not enter the charging priority sorting of this control period.

[0147] For battery compartment constraint profiles with valid profile locking records, continue with the rechargeable status screening. Subsequent determinations involving the "previous control cycle" in this step will use the charging control results of the previous control cycle stored in the cabinet's result cache as the reading source; this reading source will not be repeated later.

[0148] Then, the rechargeable constraint set of the single compartment is filtered for rechargeable status. Specifically, the charging permission status and current limit status in the constraint profile of each battery compartment are read one by one according to the battery compartment identifier: when the charging permission status is a permitted charging status or a restricted permission status and the current limit status is not a zero current status, the battery compartment identifier, battery identifier, profile lock record, risk envelope value, charging permission status, and current limit status of the battery compartment are written into the sortable battery compartment set in ascending order of battery compartment identifier, and a filtering position identifier is generated for this record. The filtering position identifier is used to read one by one in the same order in the subsequent energy replenishment qualification marking and fast charging conflict marking stages; when the charging permission status is a prohibited charging status or the current limit status is a zero current status, the battery compartment identifier, battery identifier, charging permission status, current limit status, and judgment source field of the battery compartment are written into the non-charging record. The judgment source field record is triggered by the prohibited charging status or by the zero current status, and the battery compartment written into the non-charging record is not written into the sortable battery compartment set;

[0149] When the charging permission status is "charge prohibited" or the current limit status is "zero current," the battery compartment is written to the "non-charging" record and not to the sortable battery compartment set. After completing the above filtering, the sortable battery compartment set is obtained.

[0150] After obtaining the sortable battery compartment set, the sortable battery compartment set is marked as eligible for recharging. Specifically, the energy replenishment participation status in the constraint profile of each battery compartment is read one by one according to the battery compartment identifier: When the energy replenishment participation status is "participating in energy replenishment", the battery compartment identifier, battery identifier, profile lock record and risk envelope value of the battery compartment are written into the energy replenishment candidate compartment set, and an energy replenishment qualification flag is written into the corresponding battery compartment record with the energy replenishment qualification field. The energy replenishment qualification field is used as the basis for reading whether the battery compartment participates in the energy replenishment task in the subsequent battery charging queue encapsulation, compartment configuration object generation and energy replenishment selection score calculation; When the energy replenishment participation status is "not urgently needed", the battery compartment identifier, battery identifier and current charge status of the battery compartment are written into the "not urgently needed" candidate record, and a "not urgently needed" flag is written into the energy replenishment qualification field; When the energy replenishment participation status is "energy replenishment pending", the battery compartment identifier, battery identifier and energy replenishment pending source field of the battery compartment are written into the energy replenishment pending record, and an energy replenishment pending flag is written into the energy replenishment qualification field. The energy replenishment pending source field record is triggered by the lack of site energy replenishment data or by the unstable site energy replenishment rhythm. After completing the above marking, a set of candidate positions for energy replenishment is obtained.

[0151] The charging qualification flag indicates that the battery compartment participates in the charging task of the current control cycle; the fast charging candidate status indicates that the battery compartment is allowed to participate in the charging task in fast charging mode. Battery compartments with a charging qualification flag but not written into the fast charging candidate status do not enter the fast charging reserved bit competition, and enter the derating charging group or delayed charging group according to the current limiting status, thermal diffusion isolation status and queue grouping flag.

[0152] After obtaining the set of candidate charging bays, fast charging conflict marking and current limitation marking are applied to the set of candidate charging bays. Specifically, the fast charging conflict state, thermal diffusion isolation state, charging mode limitation state, and current limitation state in the constraint profile of each battery bay are read one by one according to the battery bay identifier: when the fast charging conflict state is a conflict-free state and the charging mode limitation state is a fast charging allowed state, a fast charging candidate state is written in the fast charging conflict field and a conflict-free mark is written in the conflict mark field in the battery bay record corresponding to that battery bay;

[0153] When the fast charging conflict state is in the conflict-pending derating state, write the derating candidate state in the fast charging conflict field and write the derating conflict flag in the conflict flag field; when the fast charging conflict state is in the conflict-pending delay state, or the thermal diffusion isolation state is in the thermal diffusion isolation state, write the delay candidate state in the fast charging conflict field and write the delay conflict flag in the conflict flag field; when the current limiting state is in the derating current state, write the current limited state in the current limiting field and write the current limiting flag in the current flag field.

[0154] The fast charging conflict field, conflict flag field, current limit field, and current flag field are used as the basis for reading the corresponding battery compartments in subsequent thermal coupling conflict cluster peak shifting processing, group charging sequence generation, and compartment control parameter configuration.

[0155] After completing the above writing, the set of candidate charging bays with fast charging conflict flags, thermal diffusion isolation flags, and current limit flags will be encapsulated into a set of restricted charging bays.

[0156] After obtaining the set of restricted charging bays, the restricted charging bays are sorted according to charging eligibility, fast charging conflict, and current limitation. Specifically, the charging eligibility flags corresponding to each battery bay in the restricted charging bay set are read one by one according to the battery bay identifier. Battery bays that have not written charging eligibility flags are excluded from the charging sorting path and retained in the non-urgent candidate record or the charging pending record.

[0157] Then, read the conflict mark field corresponding to each battery compartment, and write the battery compartment identifier, battery identifier, profile lock record, risk envelope value and current fast charging conflict field value of the battery compartment with derating conflict mark or delayed conflict mark into the conflict cluster peak-shifting pending processing cache area according to the battery compartment identifier. The conflict cluster peak-shifting pending processing cache area serves as the reading source for the subsequent thermally coupled conflict cluster peak-shifting algorithm.

[0158] Then, the current flag field corresponding to each battery compartment is read. For battery compartments with current limit flags, the fast charging candidate state in the fast charging conflict field is overwritten as the derating candidate state. At the same time, the non-conflict flag in the conflict flag field is cleared and rewritten as the derating conflict flag. A current transfer reason field is added to the battery compartment record corresponding to the battery compartment to record the source of the state overwrite triggered by the current limit state.

[0159] When performing thermally coupled conflict cluster staggering, the thermally coupled conflict cluster staggering algorithm is used to divide the battery compartments with thermal influence associations in the restricted energy replenishment compartment set into conflict clusters. Specifically, the thermal coupling association strength is calculated for each pair of battery compartments in the restricted energy replenishment compartment set according to the battery compartment identifier. The formula for calculating the thermal coupling association strength is as follows:

[0160] ,

[0161] in, This indicates the thermal coupling strength between the i-th battery compartment and the j-th battery compartment during the control period t; This refers to the basic thermal influence coefficient registered by factory thermal testing or on-site trial operation, which is registered by the cabinet configuration table according to cabinet type and storage location. The basic thermal influence coefficient is determined as follows: During the factory thermal test or on-site trial operation, the rated charging current is applied to compartment i alone and maintained until the temperature rise reaches a steady state. The temperature rise slopes of compartment i and compartment j are collected simultaneously. The ratio of the temperature rise slope of compartment j to that of compartment i is used as the basic thermal influence coefficient. and the results The storage location thermal coupling basic table is written into the cabinet configuration table according to the cabinet type and storage location layout. This represents the Manhattan distance between the two compartments in the coordinate system within the cabinet, expressed in the number of compartment spacing units. The thermal coupling space attenuation scale is indicated by the cabinet configuration table, which is registered according to the air-cooled or natural heat dissipation structure. This represents the attenuation term generated by the structural distance between warehouses; the greater the structural distance, the weaker the coupling. This indicates the synchronization degree of the temperature rise trajectory, used to characterize the shape similarity of the temperature rise sequences of the two compartments within the evaluation window. This indicates the air duct correlation value. When two compartments are within the same air duct influence range, it is registered in the cabinet configuration table. When they cross air ducts, the value is zero. This represents the environmental amplification function generated by the combined effects of the cabinet's internal temperature, cooling fan speed, and cabinet door opening / closing status.

[0162] Furthermore, the formula for calculating the synchronization degree of the temperature rise trajectory is:

[0163] ,

[0164] in, This indicates the synchronization degree of the temperature rise trajectory between the i-th battery compartment and the j-th battery compartment within the evaluation window length w; w represents the synchronization degree evaluation window length, which is registered by the cabinet configuration table at several times the control cycle. This indicates that the i-th battery compartment is in the control cycle. The change in internal temperature; The L1 norm represents the temperature rise change sequence within the evaluation window, used to normalize the temperature rise sequence of each compartment before comparison. The sum of the cumulative terms in the formula reflects the total deviation of the normalized temperature rise of the two compartments within the evaluation window. Then, the window mean of this deviation is subtracted from one to obtain the synchronization value with a range of zero to one. The closer the value is to one, the more synchronized the temperature rise trajectories of the two compartments are.

[0165] After obtaining the thermal coupling correlation strength, the thermal coupling conflict baseline is tightened according to the risk envelope value before determining whether to register a thermal coupling conflict edge between the two compartments. The formula for determining the thermal coupling conflict edge is as follows:

[0166] ,

[0167] in, This indicates whether a thermal coupling conflict edge is registered between the i-th battery compartment and the j-th battery compartment within the control period t. A value of one indicates that a conflict edge is registered, and a value of zero indicates that no conflict edge is registered. This represents a conditional indicator function, which takes a value of 1 if the condition is true and zero if the condition is false. The thermal coupling conflict baseline is indicated and registered in the cabinet configuration table; The risk-sensitive threshold tightening coefficient is registered in the counter configuration table; , These represent the risk envelope values ​​of the i-th battery compartment and the j-th battery compartment, respectively, generated by the risk envelope value calculation formula described in the previous step within the control period t. This indicates the larger of the two risk envelope values; This represents the effective conflict boundary after tightening the risk envelope value. The larger the risk of the two positions, the lower the effective conflict boundary, and the more likely the two positions are to be registered as conflict edges. , These represent the fast charging candidate states of the i-th and j-th battery compartments within the control period t, respectively. Conflict edge determination is only enabled when both compartments are in the fast charging candidate state, thus avoiding duplicate determination of non-candidate compartments.

[0168] Furthermore, it will satisfy The connection between any two battery compartments with a value of 1 is registered as a thermally coupled conflict edge. The connected subgraph in the graph formed by the thermally coupled conflict edges is determined as a thermally coupled conflict cluster, thus obtaining the thermally coupled conflict cluster. For battery compartments that have not registered thermally coupled conflict edges with any other battery compartments, they are written as non-conflict fast charging candidate compartments into the priority charging group of the group charging sequence.

[0169] After obtaining the thermally coupled conflict clusters, the fast-charging candidate status of the battery compartments within the thermally coupled conflict clusters is read to obtain the conflict cluster candidate compartment records. Specifically, according to the thermally coupled conflict cluster identifier and the battery compartment identifier, the corresponding charging qualification flag, fast-charging candidate status, derating candidate status, delay candidate status, current-limited status, risk envelope value, and profile lock record for each battery compartment within each thermally coupled conflict cluster are read one by one, and the above fields are written into the conflict cluster candidate compartment records with the battery compartment identifier as the primary key.

[0170] After obtaining the candidate battery compartment records for conflict clusters, fast charging eligibility is allocated to these candidate battery compartment records to obtain the conflict cluster peak shifting results. Specifically, battery compartments corresponding to delayed candidate states and current-limited states are first screened out within each thermally coupled conflict cluster. The remaining battery compartments with charging eligibility markers generate a peak shifting score within the cluster using the following formula:

[0171] ,

[0172] in, This represents the intra-cluster peak misalignment score of the i-th battery compartment within the k-th thermally coupled conflict cluster C_k(t) and within the control period t; This represents the energy replenishment selection score of the i-th battery compartment within the control period t. It is obtained by backfilling the energy replenishment selection score generated by the demand gap reverse-engineering compartment current arrangement algorithm in subsequent steps. If it is not backfilled, it is substituted with the initial score generated by the energy replenishment participation state, derating candidate state, delay candidate state and current-limited state in the set of restricted energy replenishment compartments. This indicates the residual heat penalty coefficient, which is registered in the cabinet configuration table. Indicates the strength of thermal coupling correlation; This indicates whether the j-th battery compartment was configured to fast charging in the previous control cycle. It is set to 1 if it is configured to fast charging and to 0 if it is not configured. This indicates the amount of residual heat accumulated on the current battery compartment due to the fast charging configuration of other battery compartments in the same cluster during the previous control cycle. The larger the amount of residual heat, the lower the peak shaving score within the cluster. This represents the heat load inheritance coefficient, which is registered in the cabinet configuration table. This represents the cumulative fast-charging heat load of the i-th battery compartment in the previous control cycle, calculated from the actual compartment charging current and actual charging time retained in the battery swapping cabinet charging control results of the previous control cycle.

[0173] When the current control cycle is the first control cycle after the cabinet controller starts up. and All values ​​are set to zero; when the charging control result of the battery swapping cabinet in the previous control cycle exists but the actual sub-compartment charging current or actual charging time of the corresponding battery compartment is not saved, Take zero, The value is taken from the fast charging configuration status retained in the charging control results of the battery swapping cabinet in the previous control cycle; if it is not retained, it is taken as zero.

[0174] Furthermore, the number of fast-charging configurations in the battery compartments within a cluster is limited according to the intra-cluster peak shifting constraint. The determination formula for the intra-cluster peak shifting constraint is as follows:

[0175] ,

[0176] in, (t) indicates whether the i-th battery compartment is configured to fast charging state within the control period t. It takes one when configured to fast charging state and zero when not configured. This represents the k-th thermally coupled conflict cluster within the control period t; the determination formula indicates that within the same thermally coupled conflict cluster and within the same control period, the number of battery compartments configured for fast charging does not exceed one.

[0177] Under the constraint of intra-cluster peak shifting, the remaining battery compartments with energy replenishment qualification markers in each thermally coupled conflict cluster are sorted from high to low according to the intra-cluster peak shifting score. When the intra-cluster peak shifting scores of two or more battery compartments are equal, they are sorted in the following order: risk envelope value from low to high, cumulative fast charging heat load from the previous control cycle from low to high, and battery compartment identifier from small to large, as the secondary, tertiary, and quaternary sorting keys. The sorting results are written into the intra-cluster peak shifting sorting record with the cluster identifier as the primary key and the intra-cluster position as the secondary key. The fast charging conflict field of the battery compartment with the first intra-cluster position is overwritten as fast charging reserved status, and the fast charging reserved status is written into the peak shifting result field of the corresponding battery compartment record, which is the only battery compartment in the thermally coupled conflict cluster that is configured as fast charging state in the current control cycle.

[0178] Within the thermally coupled conflict cluster, battery compartments with charging eligibility markers (excluding those in fast-charging reserved state) are written to either a derating transfer state or a delayed transfer state according to the following rules: When the fast-charging conflict marker of a battery compartment is a derating conflict marker and the current limiting state is not zero current, it is written to a derating transfer state; when the fast-charging conflict marker of a battery compartment is a delayed conflict marker, or the thermal diffusion isolation state is thermal diffusion isolation, it is written to a delayed transfer state; except for the above cases, the writing state is determined according to the fast-charging conflict marker category of the preceding battery compartment in the cluster's staggered sorting record: when the fast-charging conflict marker of the preceding battery compartment is a derating conflict marker, it is written to a derating transfer state; when the fast-charging conflict marker of the preceding battery compartment is a delayed conflict marker, it is written to a delayed transfer state; when the battery compartment is second in the sorting record and the preceding battery compartment is in fast-charging reserved state, the writing state is determined according to the current limiting state of the battery compartment itself: when the current limiting state is not zero current, it is written to a derating transfer state; when it is zero current, it is written to a delayed transfer state. After completing the above writing, the collision cluster peak offset results are obtained.

[0179] After obtaining the collision cluster staggering results, a grouped charging sequence for queue encapsulation is generated according to the processing results of the charging qualification flag, fast charging collision flag, current limit flag, and collision cluster staggering results. Specifically, each battery compartment is grouped and written in the following mutually exclusive order: First, it is determined whether it is in a non-charging record. Battery compartments in a non-charging record are written into a non-charging group and will not enter the subsequent grouping determination; Second, it is determined whether it is in a fast charging reserved state or a non-collision fast charging candidate compartment that has not registered a thermal coupling collision edge with any other battery compartment in the collision cluster staggering results. Battery compartments that meet either condition are written into a priority charging group; Third, it is determined whether it is in a delayed transfer state or has a thermal diffusion isolation flag or a charging pending flag in the collision cluster staggering results. Battery compartments that meet either condition are written into a delayed charging group; Finally, it is determined whether it is in a derating transfer state or has a current limit flag in the collision cluster staggering results. Battery compartments that meet either condition are written into a derating charging group. The mutually exclusive assignment order ensures that each battery compartment belongs to only one group in the grouped charging sequence, and the final group to which the battery compartment is assigned is written to the queue group tag field in the battery compartment record. After writing is completed, the grouped charging sequence is obtained.

[0180] Finally, the grouped charging sequences are encapsulated into a queue to obtain a battery charging queue. Specifically, in the order of priority charging group, derating charging group, delayed charging group, and non-charging group, the battery compartment records in each group are written to the same queue file in sequence; during writing, the control cycle identifier, cabinet identifier, battery compartment identifier, battery identifier, profile lock record, risk envelope value, charging qualification flag, fast charging conflict flag, thermal diffusion isolation flag, current limit flag, intra-cluster peak shifting score, conflict cluster peak shifting result, and queue grouping flag are retained.

[0181] The queue position identifier is generated by concatenating the queue group identifier and the position within the group according to the field. The queue group identifier is registered as one, two, three, and four in the order of priority charging group, derating charging group, delayed charging group, and non-charging group. The position within the group is determined by the peak-shifting score within the cluster from high to low within the priority charging group. When the peak-shifting scores within the cluster are equal, they are determined by the risk envelope value from low to high, the cumulative fast-charging heat load of the previous control cycle from low to high, and the battery compartment identifier from small to large as the second, third, and fourth level sorting keys, respectively. The position within the group is determined by the peak-shifting score within the cluster from high to low within the derating charging group and delayed charging group. When the peak-shifting scores within the cluster are equal, they are determined by the risk envelope value from low to high. The position within the group is determined by the battery compartment identifier from small to large within the non-charging group. A queue position identifier is generated for each battery compartment record according to the above rules.

[0182] Once all battery compartment records in each group are written to the same queue file and a queue position identifier is generated for each battery compartment record, queue encapsulation is completed, resulting in a battery charging queue. The battery charging queue is then written to the compartment control parameter configuration input buffer as input for the next step of configuring the compartment control parameters for the battery charging queue.

[0183] In one embodiment, the thermally coupled conflict cluster staggering algorithm further includes:

[0184] In actual operation of the battery swapping cabinet, multiple battery compartments within the cabinet are coupled through the same cabinet sheet metal, the same air duct, and the same internal air environment. This results in a significant thermal impact correlation between adjacent battery compartments when the fast charging current is high. When two adjacent battery compartments are simultaneously configured for fast charging within the same control cycle, their respective temperature rises are superimposed through conduction through the compartment walls and crosstalk in the air duct. This causes the two compartments, which were originally below the fast charging temperature limit, to reach the fast charging temperature limit one after the other during fast charging, thus triggering the cabinet's current limiting or charging stop protection. During peak charging periods at the site, this process manifests as repeated switching between "charging starts with derating and charging stops shortly after derating," which not only fails to complete charging within the time gap but also accelerates battery aging at high temperatures. Before the fast charging command is issued, the thermally coupled conflict cluster staggered peak algorithm identifies battery compartment pairs with thermally related conditions within the current control cycle based on structural distance between compartments, temperature rise trajectory synchronization, airflow correlation, and cabinet environmental conditions. The connected subgraph formed by these thermally related conditions is then divided into thermally coupled conflict clusters, and it is constrained that at most one battery compartment within the same conflict cluster is configured for fast charging within the same control cycle. Thus, battery compartments simultaneously entering fast charging within the cabinet are always spatially isolated by structural distance, different airflow paths, or asynchronous temperature rise trajectories, fundamentally preventing repeated derating and charging stoppages induced by the superposition of temperature rises in adjacent compartments.

[0185] For operational scenarios where battery compartments within a cabinet have varying service lives and health conditions, traditional fast-charging candidate selection methods, which prioritize candidates based on state of charge difference or urgency, can lead to situations where older, less healthy battery compartments are repeatedly selected as fast-charging candidates due to their larger remaining rechargeable capacity. This further exacerbates the aging and thermal runaway risks associated with these compartments. The proposed thermally coupled conflict cluster staggering algorithm introduces a tightening mechanism at the conflict edge determination stage, based on the risk envelope value described in the previous step, to tighten the thermally coupled conflict baseline. This increases the risk of the higher-risk compartment, lowers the effective conflict boundary between them, and makes it easier to register a thermally coupled conflict edge between them. Consequently, higher-risk battery compartments are more likely to be included in conflict clusters and subject to staggering constraints during the cluster partitioning stage. This mechanism moves the engineering intent of "high-risk battery compartments should be kept away from fast charging" from the single-compartment shutdown level to the cluster partitioning level, resulting in more stringent isolation of concurrent fast-charging combinations between higher-risk battery compartments and their adjacent compartments during the determination process.

[0186] In scenarios where battery swapping cabinets operate in cycles based on control periods and there is residual heat between adjacent control periods, if the battery compartment that was configured for fast charging in the previous control period has not yet cooled down to ambient temperature at the start of the current control period, and if the residual heat across cycles is not taken into account and the cluster peak shaving score is calculated only based on the temperature rise trajectory within the current control period, the same battery compartment will be repeatedly selected as fast charging reserved state in multiple adjacent control periods, causing the compartment to bear too much fast charging load and enter the aging acceleration zone.

[0187] The thermally coupled conflict cluster staggering algorithm introduces a residual heat penalty term accumulated on the current battery compartment due to the fast charging configuration of other battery compartments in the same cluster in the previous control cycle, as well as a heat load inheritance term calculated from the accumulated fast charging heat load of the battery compartment itself in the previous control cycle, into the cluster staggering score. This results in a corresponding reduction in the cluster staggering score of battery compartments that undertook fast charging tasks in the previous control cycle or those near fast charging positions in the previous control cycle and affected by residual heat. As a result, the fast charging load in the cabinet naturally rotates across control cycles, avoiding the concentration of fast charging tasks in a few battery compartments and preventing localized accelerated aging.

[0188] In the case of heat diffusion isolation, when the temperature of a battery compartment exceeds the heat diffusion isolation threshold registered in the cabinet configuration table, the compartment itself must not only be removed from the fast charging candidate, but the fast charging activities of its adjacent compartments will also further increase its temperature through the air duct and the air environment inside the cabinet, which poses a risk of aggravated heat diffusion.

[0189] In the fast-charging qualification staggering allocation stage, the thermally coupled conflict cluster staggering algorithm directly writes the delayed transfer state to the battery compartment in the thermal diffusion isolation state and sets its fast-charging candidate state to zero in the current control cycle, so that it does not participate in the fast-charging reserved bit competition of its conflict cluster. At the same time, since the fast-charging candidate state of the battery compartment in the thermal diffusion isolation state is set to zero in the conflict edge determination, the effective fast-charging candidate range of its conflict cluster is reduced accordingly, avoiding the battery compartment in the thermal diffusion isolation state from being incorrectly allocated to the derating transfer state because it is still included in the candidate. Thus, the two requirements of "self-cooling" and "avoiding secondary temperature increase of neighboring compartments" are uniformly implemented at the algorithm level.

[0190] For scenarios where the number of eligible battery compartments exceeds the capacity of the conflict cluster during peak charging periods at a site—that is, multiple battery compartments with charging eligibility tags exist within the same conflict cluster, but only one battery compartment is allowed to be configured for fast charging within the cluster—the thermally coupled conflict cluster staggering algorithm does not perform coarse-grained hard discarding within the cluster. Instead, it systematically divides the remaining battery compartments with charging eligibility tags according to the staggering score within the cluster. The battery compartment ranked first is written into the fast charging reserved state, and the remaining battery compartments are written into the derating transfer state or delayed transfer state according to their respective fast charging conflict tags and current limit states. These are then handed over to the subsequent compartment control parameter configuration stage to continue undertaking the charging task in the derating charging mode or delayed charging mode. Thus, battery compartments within the conflict cluster that have not obtained a fast charging reserved position are not simply excluded, but continue to participate in the charging of the current control cycle in a secondary mode that matches their state. This maximizes the utilization of the cabinet's rechargeable capacity to respond to the site's charging gap while ensuring the safety of concurrency within the cluster.

[0191] By employing connected subgraph cluster partitioning, risk-sensitive conflict edge tightening, cross-cycle thermal residue and thermal load inheritance, candidate zeroing for thermal diffusion isolation, and ordered staggered peak partitioning within clusters in the aforementioned scenarios, the thermally coupled conflict cluster staggered peak algorithm transforms the post-event current limiting or charging stoppage originally performed at the cabinet end after the temperature reaches the threshold into pre-event cluster partitioning and staggered peak allocation completed before the fast charging command is issued. The allocation results are written into the group charging sequence in the form of fast charging retention state, derating transfer state, and delayed transfer state, serving as inputs for the compartment control parameter configuration and demand gap reverse-engineering compartment current orchestration algorithm in subsequent steps. This ensures concurrent safety between adjacent compartments, fast charging load balancing across cycles, and maximized response to site energy replenishment gaps in the actual operation of multi-compartment concurrent charging within the cabinet.

[0192] Configure compartment control parameters for the battery charging queue and generate compartment charging control commands;

[0193] The compartment charging control command refers to the set of commands obtained by batch packaging according to cabinet identification and control cycle identification after the compartment control parameters are configured, and by recording each compartment control parameter according to the battery compartment identification to generate compartment-level control commands. Each compartment-level control command carries dynamic charging mode, actual compartment charging current, charging cut-off state of charge, delayed start time, estimated battery swapping time, image lock record and control reason mark.

[0194] After obtaining the battery charging queue, the cabinet-side controller reads the battery charging queue corresponding to the current control cycle from the compartment control parameter configuration input buffer. Then, according to the queue position identifier, it sequentially reads the control cycle identifier, cabinet identifier, battery compartment identifier, battery identifier, profile lock record, risk envelope value, energy replenishment qualification mark, fast charging conflict mark, thermal diffusion isolation mark, current limit mark, intra-cluster peak shifting score, conflict cluster peak shifting result, queue grouping mark, current state of charge, current charging voltage, and charging permission status of each battery compartment in the queue file.

[0195] Each battery compartment record undergoes a field integrity check sequentially: if the profile lock record field is empty or its length is not equal to the profile lock record length registered in the cabinet configuration table, it is determined that the profile lock record is missing; if the queue grouping mark field is empty or its value is not within the four registered values ​​of priority charging group, derating charging group, delayed charging group, and non-charging group, it is determined that the queue grouping mark is missing; if the queue position identifier field is empty or its value has been occupied by another battery compartment record within this control cycle, it is determined that the queue position identifier is missing. For battery compartment records exhibiting any of the above conditions, their battery compartment identifier, battery identifier, missing field name, and missing reason are written into the sub-compartment configuration anomaly record. The sub-compartment configuration anomaly record is used by the cabinet controller as the source of the anomaly output field in the control output encapsulation stage; it does not enter the sub-compartment control parameter configuration for this control cycle. Battery compartment records that pass the field integrity check continue to execute queue position reading.

[0196] Next, the queue position of the battery charging queue is read to obtain the compartment configuration object. Specifically, the battery compartment records in the priority charging group, derating charging group, delayed charging group and non-charging group are read sequentially according to the queue position identifier, and the aforementioned fields in each battery compartment record are bound to obtain the compartment configuration object;

[0197] If the same battery compartment identifier appears in two or more queue records within the current control period, the following rules are used to determine which record to retain: First, compare the image lock record in each queue record with the image lock record generated by the single-compartment constraint image lock algorithm for the corresponding battery compartment in the single-compartment chargeable constraint set within the current control period, and only retain the queue record that is equal to it; if there are still two or more candidate records after the above comparison, retain the one with the first position in the queue position identifier from smallest to largest; write the battery compartment identifier, battery identifier, image lock record, and queue position identifier of the remaining queue records into the duplicate queue record index.

[0198] After obtaining the sub-compartment configuration object, the charging mode, current limit, charging cutoff state of charge, and delayed start are configured for the sub-compartment configuration object. Specifically, independent buffers are established in the cabinet controller according to four processing paths: a replenishment calculation buffer, a derating calculation buffer, a delayed calculation buffer, and a non-charging control buffer. Each buffer uses the battery compartment identifier as the primary key and maintains the recording order in ascending order according to the queue position identifier. The sub-compartment configuration object is written to the corresponding buffer according to the queue grouping mark: when the queue grouping mark is a priority replenishment group, all binding fields of the sub-compartment configuration object are written to the replenishment calculation buffer; when the queue grouping mark is a derating charging group, it is written to the derating calculation buffer; when the queue grouping mark is a delayed charging group, it is written to the delayed calculation buffer; when the queue grouping mark is a non-charging group or the charging permission state is a prohibited charging state, it is written to the non-charging control buffer. The energy replenishment calculation cache serves as the source of reading for the subsequent demand gap-based current orchestration algorithm; the derating calculation cache and the delayed calculation cache serve as the source of reading for the subsequent dynamic charging mode configuration; and the non-charging control cache serves as the source of reading for the subsequent non-charging mode write and delayed start configuration.

[0199] Then, the demand gap reverse-engineering compartment current orchestration algorithm is invoked to determine the compartment location for energy replenishment, configure the dynamic charging mode, configure the current limit, configure the charging cutoff state of charge, configure the delayed start, and configure the expected battery swapping time for the compartment configuration objects in the energy replenishment calculation path. The processing of the demand gap reverse-engineering compartment current orchestration algorithm includes steps such as generating an effective energy replenishment gap, calculating the energy replenishment selection score, generating the reverse current, generating the actual compartment charging current, and generating the thermal safety current upper limit.

[0200] Specifically, the demand gap reverse-engineering current arrangement algorithm first calculates the effective energy gap perceived by the time window by subtracting the number of battery compartments that will naturally reach a swappable charge state within the safe energy replenishment time window, based on the difference between the target available battery count and the current swappable battery count in the site replenishment data. The calculation formula for the effective energy gap is as follows:

[0201] ,

[0202] in, This indicates the number of effective energy replenishment gaps within the control period t; This represents the target number of available batteries returned by the site service interface within the control period t; This indicates the number of swappable batteries currently in the cabinet at the start of control cycle t; This represents the estimated swappable time of the battery in the i-th battery compartment, which is retained in the charging control result of the battery swapping cabinet in the previous control cycle within control cycle t. The safe energy replenishment time window is indicated by the configuration table at the counter, which is registered according to the peak power swapping schedule of the site. This represents a conditional indicator function, which takes a value of 1 if the condition is true and zero if the condition is false. This indicates the number of battery compartments that will naturally reach a swappable charge state within the safe recharge time window during the current control cycle. This means that the value is zero when the difference is less than zero, to avoid the effective energy gap being negative.

[0203] when When the value is greater than zero, the demand gap reverse-engineering compartmentalized current orchestration algorithm enters the energy replenishment selection scoring calculation; when When the value is not greater than zero, no new set of energy replenishment execution positions will be created, and the sub-position configuration objects in the energy replenishment calculation path will be written into the non-urgent control path.

[0204] Furthermore, the demand gap reverse-engineering compartmentalized current orchestration algorithm generates a replenishment selection score for the compartmentalized configuration objects in the replenishment calculation path according to the following formula:

[0205] ,

[0206] in, This represents the energy replenishment selection score of the i-th battery compartment within the control period t; This indicates the charging participation status of the i-th battery compartment, which is derived from the charging eligibility flag. It is set to 1 when the charging eligibility flag is written and to 0 when it is not written. The indicator represents the fast charging conflict release status of the i-th battery compartment. It is obtained by converting the conflict cluster staggering result generated by the thermal coupling conflict cluster staggering algorithm. When it is in the fast charging retention state, it takes a value of 1; when it is in the derating transfer state, it takes a value less than 1 according to the derating release coefficient registered in the cabinet configuration table; and when it is in the delayed transfer state, it takes a value of zero. This indicates the electrical status of the replaceable charge, which is registered in the configuration table at the cabinet end; This indicates the current state of charge of the battery in the i-th battery compartment during the control period t; This represents the estimated time required for the battery in the i-th battery compartment to charge to a swappable state under the current current-limited condition; This represents a constant used to prevent the denominator from being zero, and is registered in the counter configuration table; This represents the risk decay factor generated from the risk envelope value; This represents the gap amplification factor generated by the effective energy filling gap.

[0207] Furthermore, the formula for calculating the risk attenuation factor is as follows:

[0208] ,

[0209] in, This represents the risk decay factor of the i-th battery compartment within the control period t. The higher the risk envelope value and the smaller the risk decay factor, the lower the energy replenishment selection score of the battery compartment will be. This represents the risk attenuation coefficient, which is registered in the counter configuration table. The risk envelope value of the i-th battery compartment within the control period t is obtained by the calculation formula of the risk envelope value generated by the single compartment constraint profile locking algorithm described in the previous step; This represents the natural exponential function.

[0210] Furthermore, the formula for calculating the notch amplification factor is as follows:

[0211] ,

[0212] in, This represents the gap amplification factor generated by the effective energy replenishment gap. The larger the effective energy replenishment gap, the larger the gap amplification factor, which further amplifies the battery compartment that is close to achieving the charge-swappable electrical state in the energy replenishment selection score. This indicates the gap magnification factor, which is registered in the cabinet configuration table; This represents the gap normalization constant registered by the counter configuration table according to the typical gap size of the site; Indicates the number of effective energy gaps; This represents the hyperbolic tangent function, which makes the notch amplification factor converge to one when the effective energy gap is near zero, and tend to one when the effective energy gap is large. .

[0213] Furthermore, the demand gap-based compartmentalized current orchestration algorithm arranges the compartmentalized configuration objects in the energy replenishment calculation path from high to low according to the energy replenishment selection score, and selects compartmentalized configuration objects sequentially from the arrangement results based on the effective energy replenishment gap: the selected compartmentalized configuration objects whose cumulative number does not exceed the effective energy replenishment gap are written into the energy replenishment execution compartment set; the compartmentalized configuration objects that are not selected but have charging permits are written into the non-urgent control path. After completing the above selection, the energy replenishment execution compartment set is obtained.

[0214] After obtaining the set of charging execution locations, the demand gap reverse-engineering current orchestration algorithm dynamically configures the charging mode of the set of charging execution locations to obtain a dynamic charging mode record. Specifically, according to the battery compartment identifier, the queue grouping flag, fast charging conflict flag, current limit flag, and charging permission status of each compartment configuration object in the set of charging execution locations are read one by one: when the queue grouping flag is a priority charging group and there are no derating conflict flags and current limit flags, the fast charging mode is written.

[0215] When a queue group is marked as a priority charging group but has a current limit flag, write to the normal charging mode;

[0216] When a queue group is marked as a derating charging group, write the derating charging mode; when a queue group is marked as a delayed charging group, write the delayed charging mode.

[0217] When a queue group is marked as a non-charging group or its charging permission status is set to "charging prohibited," a non-charging mode is written. After completing the above write, a dynamic charging mode record is obtained.

[0218] Furthermore, the demand gap-based reverse-engineering current orchestration algorithm configures current limits for the dynamic charging mode record. Specifically, the reverse-engineering current required for the i-th battery compartment to achieve the target battery swap completion time within the control period t is generated using the following formula:

[0219] ,

[0220] in, This represents the reverse current of the i-th battery compartment during the control period t; This indicates the rated capacity of the battery in the i-th battery compartment, which is registered in the local historical database based on the battery identifier; This represents the charging efficiency coefficient of the battery in the i-th battery compartment. This represents the current charging voltage of the battery in the i-th battery compartment during the control period t; This indicates the target battery swap completion time for the sub-compartment configuration object, and its value is the sum of the start time t of the current control cycle and the target battery swap duration;

[0221] The target battery swapping time is obtained by dividing the number of effective energy replenishment gaps by the expected battery swapping rhythm. The expected battery swapping rhythm is recorded by the number of times riders go to the cabinet for battery swapping per unit time returned by the station business interface, with the unit being vehicles per minute. When the expected battery swapping rhythm is zero or the station business interface does not return the expected battery swapping rhythm, the safe energy replenishment time window registered in the cabinet configuration table is used as the target battery swapping time. This indicates the remaining time from the start of the current control cycle to the completion of the target battery swap.

[0222] Furthermore, the demand gap-based reverse-engineering current allocation algorithm is calculated by taking the minimum value of the reverse-engineered current, the charging current limit registered by the battery-side charging constraints, the risk-driven thermal safety current limit, and the intra-cluster charging current limit of the conflict cluster constraints, and then modulating it with the fast charging conflict release status indication to obtain the actual compartment charging current. The calculation formula for the actual compartment charging current is as follows:

[0223] ,

[0224] in, This represents the actual compartment charging current of the i-th battery compartment during the control period t; This indicates the charging current limit determined by the current limit state in the battery-side charging constraints. When the current limit state is the rated current state, the rated charging current registered in the cabinet configuration table is used; when it is the derating current state, the derating current registered in the battery-side charging constraints is used; and when it is the zero current state, zero is used. This indicates the upper limit of the thermally safe current generated by the risk envelope value; This represents the upper limit of the intra-cluster charging current determined by the collision cluster staggering result generated by the thermally coupled collision cluster staggering algorithm. When in fast charging reserve state, the rated charging current is used; when in derating transfer state, the intra-cluster derating current registered in the cabinet configuration table is used; and when in delayed transfer state, zero is used. This indicates the fast charging conflict release status of the i-th battery compartment.

[0225] Furthermore, the formula for calculating the upper limit of the thermal safety current is as follows:

[0226] ,

[0227] in, This represents the upper limit of the thermal safety current of the i-th battery compartment during the control period t; This indicates the rated charging current registered in the cabinet configuration table; α represents the risk envelope value of the i-th battery compartment within the control period t; α represents the thermal safety degradation index, which is registered in the cabinet configuration table. The higher the risk envelope value, the faster the upper limit of the thermal safety current decreases relative to the rated charging current.

[0228] After obtaining the actual compartment charging current, the demand gap reverse-engineering compartment current orchestration algorithm writes it into the current limit field of the dynamic charging mode record; when When the current is zero, the sub-compartment configuration object is switched from the energy replenishment calculation path to the delayed calculation path, and a zero current cause flag is written.

[0229] Furthermore, the demand gap reverse-engineering compartmentalized current orchestration algorithm configures the charging cutoff charge state for the dynamic charging mode record: when the dynamic charging mode is a fast energy replenishment mode or a regular charging mode, the swappable charge state is written into the charging cutoff charge state; when the dynamic charging mode is a derating charging mode, the derating charging cutoff charge state registered in the cabinet configuration table is written into the charging cutoff charge state.

[0230] When the dynamic charging mode is the maintenance charging mode, the maintenance charging cut-off charge status registered in the cabinet configuration table will be written into the charging cut-off charge status.

[0231] When the dynamic charging mode is delayed charging mode or non-charging mode, no new charging cutoff state is written, and the charging cutoff state field is marked as not to be updated in this cycle.

[0232] Furthermore, the demand gap reverse-engineering compartment current orchestration algorithm configures the dynamic charging mode record with a delayed start: when the queue group is marked as a delayed charging group, the delayed start time is generated sequentially according to the queue position identifier, so that each battery compartment in the delayed charging group starts up in staggered peaks within the current control cycle.

[0233] When the fast charging conflict flag is a delayed conflict flag, the start time of the next control cycle after the current control cycle is written into the delayed start time.

[0234] When the sub-compartment configuration object does not have a delayed conflict flag, a thermal diffusion isolation flag, or a zero current cause flag, the delayed start time will be written to the immediate execution flag.

[0235] when When the value is zero, the estimated battery swapping time is not calculated, and the estimated battery swapping time field of the sub-compartment configuration object is marked as not to be updated in this cycle. After completing the above writing, the dynamic charging mode, actual sub-compartment charging current, charging cutoff state of charge, delayed start time, estimated battery swapping time, and energy replenishment reason are written to the corresponding sub-compartment configuration object to obtain the energy replenishment control parameter record.

[0236] For sub-compartment configuration objects that are not written into the aforementioned energy replenishment execution compartment set but have charging permission, they are entered into the non-urgent control path for maintenance charging or delayed charging configuration, and non-urgent control parameter records are obtained. Specifically, according to the battery compartment identifier, the charging permission status, current limit status, fast charging conflict flag, thermal diffusion isolation flag, and zero current reason flag of each sub-compartment configuration object in the non-urgent control path are read one by one: when the charging permission status is the allowed charging status and the current limit status is not the zero current status and does not have a zero current reason flag, the maintenance charging mode is written, and the maintenance current limit and maintenance charging cut-off charge status registered in the cabinet configuration table are written into the corresponding fields;

[0237] When there is a delayed conflict flag, thermal diffusion isolation flag, or zero current cause flag, write the delayed charging mode, and write the time after adding the product of the start time of the next control cycle and the group position in the queue position identifier multiplied by the non-urgent peak shift step size registered in the cabinet configuration table into the delayed start time.

[0238] When the charging permission status is disabled, write to non-charging mode.

[0239] After completing the above writing, the maintenance charging mode, maintenance current limit, maintenance charging cutoff state of charge, delayed charging mode, delayed start time, non-charging mode and non-urgent reason mark are written to the corresponding compartment configuration object to obtain the non-urgent control parameter record.

[0240] After obtaining the records of replenishment control parameters and non-urgent control parameters, write them into the corresponding sub-compartment configuration objects according to the battery compartment identifier.

[0241] Finally, the compartment control parameter records are encapsulated into instructions to generate compartment charging control instructions. Specifically, the compartment control parameter records are batch-encapsulated according to the cabinet identifier and control cycle identifier; a compartment-level control instruction is generated for each compartment control parameter record according to the battery compartment identifier. The compartment-level control instruction includes the dynamic charging mode, actual compartment charging current, charging cutoff state of charge, delayed start time, estimated battery swapping time, profile lock record, and control reason flag. After all battery compartments with compartment control parameter records have generated compartment-level control instructions in the current control cycle, each compartment-level control instruction is encapsulated into a compartment charging control instruction, and the compartment charging control instruction is written into the execution status retrieval input buffer as the input for the execution status retrieval of the compartment charging control instruction in the next step.

[0242] For a compartment configuration object that enters the derating calculation path, the cabinet controller reads the current allowable current limit, derating current status, current charge status and queue position identifier of the compartment configuration object, writes the charging mode into the derating charging mode, writes the current limit value into the lower value between the current allowable current limit and the derating current value registered in the cabinet configuration table, writes the charging cut-off charge status into the swappable charge status threshold, and writes the delayed start time into the start time of the current control cycle to obtain the derating control parameter record.

[0243] For a compartment configuration object entering the delayed calculation path, the cabinet controller reads the thermal diffusion isolation status, delayed conflict flag, and current charge state of the compartment configuration object, writes the charging mode into the delayed charging mode, writes the current limit value into zero, writes the charging cut-off charge state into the swappable charge state threshold, and writes the delayed start time into the start time of the next control cycle. When the compartment configuration object has a thermal diffusion isolation status, the delayed start time is written into the start time of the control cycle after the thermal diffusion isolation is released, thus obtaining the delayed control parameter record.

[0244] For a compartment configuration object that enters a non-charging control path, the cabinet controller reads the charging permission status and current charge status of the compartment configuration object, writes the charging mode to the non-charging mode, writes the current limit value to zero, writes the charging cut-off charge status to the current charge status, and writes the delayed start time to a null value to obtain the non-charging control parameter record.

[0245] The cabinet-side controller writes the energy replenishment control parameter record, non-urgent control parameter record, derating control parameter record, delayed control parameter record and non-charging control parameter record into the parameter field of the same sub-compartment control parameter record according to the battery compartment identifier, and writes the control cycle identifier, cabinet identifier, battery compartment identifier, battery identifier, image lock record and queue position identifier into the index field of the same sub-compartment control parameter record to obtain the sub-compartment control parameter record.

[0246] In one embodiment, the demand gap reverse-engineering compartmentalized current orchestration algorithm further includes:

[0247] In actual operation, the riders' battery swapping rhythm is not evenly distributed, but exhibits a tidal characteristic, concentrated around the morning, noon, and evening peak hours, and sparse at other times. When riders arrive in large numbers, the number of swappable batteries in the cabinet can rapidly drop from meeting the swapping schedule to a shortage state within minutes, while several battery compartments in the cabinet are about to reach the swappable charge state. If the fast charging configuration is directly driven by the original shortage obtained by "target available battery number minus current swappable battery number" in the existing technology, the aforementioned battery compartments that are about to reach the swappable charge state will not be identified in the shortage determination. This will result in the cabinet repeatedly increasing the fast charging load on other battery compartments during peak charging periods, further compressing the dispatchable capacity in the cabinet.

[0248] The demand gap reverse-engineering compartment current orchestration algorithm introduces a safe energy replenishment time window during the gap generation stage. The battery compartments whose expected swappable time is within the sum of the current control cycle start time and the safe energy replenishment time window are deducted in advance. This results in the effective energy replenishment gap perceived by the time window, so that the cabinet will only enter a new fast charging configuration when the effective energy replenishment gap is still greater than zero, thus avoiding repeated fast charging drive for battery compartments that are close to being reached.

[0249] When the current control cycle is the first control cycle after the cabinet controller starts, or when the previous control cycle did not save the estimated swappable time, the cabinet controller generates an initial estimated swappable time based on the current state of charge, swappable state of charge threshold, battery rated capacity, current allowable current and current charging voltage, and uses the initial estimated swappable time as the basis for determining the effective energy replenishment gap deduction item;

[0250] If the current state of charge, current allowable current, or current charging voltage is missing, causing the initial estimated swappable time to fail to be generated, the battery compartment will be written into the estimated time missing record, and the battery compartment will be considered as unable to naturally achieve a swappable state of charge within the safe recharge time window.

[0251] In response to the operational scenario where multiple battery compartments within the charging cabinet exhibit differences in state of charge, health, and thermal condition during peak charging periods at charging stations, the traditional method of selecting charging candidates based on state of charge difference or urgency presents two typical problems: First, battery compartments with low state of charge but low health and high risk envelope are continuously selected as fast-charging candidates, further accelerating aging; second, when the state of charge is close to the swappable boundary but the effective charging gap is still large, these battery compartments are not given sufficient priority, resulting in a mismatch between the station's charging rhythm and the charging rhythm within the cabinet.

[0252] The demand gap reverse-engineering sub-warehouse current orchestration algorithm introduces a risk attenuation factor corresponding to the risk envelope value generated by the single-warehouse constraint profile locking algorithm in the previous step into the energy replenishment selection score, so that the higher the risk envelope value, the lower the energy replenishment selection score.

[0253] Simultaneously, a gap amplification factor generated by the effective energy replenishment gap is introduced, so that the larger the effective energy replenishment gap and the closer the battery compartment is to achieving a swappable charge state, the more amplified it is in the energy replenishment selection score. As a result, during the peak energy replenishment period at the site, the cabinet will not apply fast charging load to high-risk battery compartments, and can prioritize the release of battery compartments that are close to achieving a swappable charge state at the peak moment, so that the energy replenishment rhythm in the cabinet is aligned with the battery swapping rhythm at the site on the control cycle scale.

[0254] For scenarios where multiple current limits are applied to the battery compartment within the cabinet in the aforementioned steps, including battery-side chargeability constraints, thermal coupling conflict cluster staggering algorithms, and single-compartment constraint profiling locking algorithms, the demand gap-based compartment current orchestration algorithm no longer uses a single upper limit as the sole basis for fast charging current. Instead, it takes the minimum of the back-calculated current from the target battery swap completion time, the charging current limit registered by the battery-side chargeability constraints, the thermal safety current upper limit generated by the risk envelope value, and the intra-cluster charging current upper limit generated by the thermal coupling conflict cluster staggering algorithm, and then modulates it with the fast charging conflict release status indication to obtain the actual compartment charging current. When the back-calculated current is lower than the other upper limits, the cabinet executes the minimum fast charging intensity required at the target battery swap completion time to avoid applying current to the battery that exceeds the energy replenishment rhythm. When any safety-type upper limit is lower than the back-calculated current, the cabinet executes according to that upper limit, ensuring that the station's energy replenishment rhythm is not transmitted as a concession to the single-compartment safety boundary.

[0255] In the case of a scenario where the actual charging current of the battery compartment inside the cabinet may drop to zero instantaneously during fast charging due to voltage sampling fluctuations, thermal protection boosting, or peak switching of conflict clusters, if the compartment-level control command is still configured according to zero current in the traditional way, it will cause the abnormal record of "issued but current is zero" to appear repeatedly in the subsequent execution status recovery stage, interfering with the feedback status judgment.

[0256] When the actual charging current of a battery compartment is zero, the demand gap reverse-engineering current orchestration algorithm will no longer keep the battery compartment in the energy replenishment calculation path. Instead, it will switch it to the delayed calculation path and write a zero current cause flag. The subsequent delayed start configuration stage will generate a delayed start time according to the queue position identifier or the start time of the next control cycle, so that the battery compartment can be added to the fast charging candidate after the conditions are restored, avoiding repeated sampling anomalies caused by instantaneous zero current.

[0257] In response to the operational scenario where the estimated battery swapping time after fast charging in the cabinet deviates from the rider's arrival rhythm, and this deviation is affected by both the peak-shifting result of the thermally coupled conflict cluster and the queue position, the demand gap reverse-engineering compartment current orchestration algorithm, in the stage of generating the estimated swappable time, in addition to obtaining the basic estimated swappable time based on the rated capacity, current state of charge, swappable charge state of charge, charging efficiency coefficient, current charging voltage, and actual compartment charging current, also superimposes a peak-shifting delay term generated by the peak-shifting result of the thermally coupled conflict cluster and a queue position delay term generated by the queue position identifier, so that the estimated swappable time truly reflects the actual reachability of the battery compartment under the peak-shifting within the cluster and queue grouping.

[0258] When the estimated swappable time is backfilled into the next control cycle as an effective energy replenishment gap deduction item, a closed loop that can stably converge during the control cycle is formed between gap generation, energy replenishment selection scoring, and reverse current calculation, avoiding capacity fluctuations caused by repeated expansion of fast charging configuration at the cabinet during peak energy replenishment periods.

[0259] By generating time-window perception gaps, risk-sensitive energy replenishment selection scoring, closing the actual compartment charging current with the minimum upper limit in the above scenarios, transitioning to delayed paths during instantaneous zero current, and generating the expected swappable time based on cluster peak shifting and queue position perception, the demand gap reverse-engineering compartment current orchestration algorithm transforms the post-event current limiting control originally applied by the cabinet end according to the threshold of the whole cabinet or a single compartment into pre-event gap reverse-engineering and multi-constraint closure completed before the instruction is issued. The allocation result is written into the compartment charging control instruction in the form of compartment control parameter records, which serves as the input for execution status retrieval, feedback status judgment, and control output encapsulation in subsequent steps. Thus, in the actual operation of concurrent charging in multiple compartments within the cabinet, it ensures the accurate response of the station's energy replenishment rhythm, the stable maintenance of the single compartment safety boundary, and the maximum release of the cabinet's energy replenishment capacity.

[0260] The execution status of the compartment charging control command is collected to obtain charging execution feedback data;

[0261] The execution status feedback refers to the process by which the cabinet controller, after issuing the sub-compartment charging control command, completes the strategy version binding through the edge control and cloud strategy collaboration algorithm, then initiates status reading commands to each charging control node according to the node address table, and associates and merges the actual execution fields returned by each charging control node with the corresponding compartment issuance record and command execution benchmark record.

[0262] After generating the compartment charging control command, the cabinet controller reads the corresponding compartment charging control command for the current control cycle from the execution status feedback input buffer. Then, according to the battery compartment identifier, it sequentially reads the control cycle identifier, cabinet identifier, battery compartment identifier, battery identifier, profile lock record, dynamic charging mode, actual compartment charging current, charging cutoff state of charge, delayed start time, estimated battery swapping time, and control reason flag for each compartment-level control command. For compartment-level control commands lacking a control cycle identifier, cabinet identifier, battery compartment identifier, or profile lock record, an instruction exception record is written with an exception reason flag, and the command is not included in the strategy version binding for this control cycle. For compartment-level control commands with complete fields, the command is included in the strategy version binding.

[0263] Then, the edge control and cloud strategy collaboration algorithm is invoked to bind the strategy version of the compartment charging control commands. The edge control and cloud strategy collaboration algorithm uses the inconsistency in four dimensions—strategy version identifier, strategy summary, profile locking record, and control cycle identifier—as criteria to perform a quantitatively measurable collaborative evaluation of each compartment-level control command between the cloud and the edge. Based on the evaluation results, the compartment-level control command is handled in a tiered manner: if all dimensions are consistent, it is issued with all its parameters; only when the profile dimension is consistent but other dimensions have slight inconsistencies is it issued with reduced rates or maintenance methods to avoid a collective shutdown at the counter during cloud strategy version switching; if none of the above conditions are met, it is not issued, and version rollback synchronization is enabled as needed.

[0264] Specifically, the algorithm first reads the image lock record, control cycle identifier, and cabinet identifier from the compartment-level control instructions one by one according to the battery compartment identifier. It then reads the edge execution version identifier from the edge policy registration table and sends a version query request carrying the cabinet identifier, control cycle identifier, and image lock record to the cloud policy interface through the gateway communication module. When the cloud policy version identifier and cloud policy summary returned by the cloud policy interface are received within the version query timeout window registered in the cabinet configuration table, the above two items are written into the cloud policy return area. When no return data is received within the version query timeout window, the cloud policy version identifier and cloud policy summary retained during the last successful synchronization are read from the edge policy registration table and written into the cloud policy return area. A cloud return missing flag is also written, which serves as the basis for determining the inconsistency value of the policy summary dimension in the subsequent multi-dimensional consistency distance generation.

[0265] Simultaneously, according to the arrangement order of the battery compartment identifiers, the dynamic charging mode, actual compartment charging current, charging cutoff state of charge, delayed start time, estimated swappable time, and profile locking record of each compartment-level control command in the compartment charging control command are extracted one by one. After concatenating the fields in the above order, the same summary calculation function described in the single-compartment constraint profile locking algorithm is called to calculate the summary of the concatenated result, obtaining the edge-end strategy summary. The digest calculation function is represented in this embodiment. The SHA-256 algorithm is used, and the output bit width is 256 bits. The digest calculation function uses the same algorithm and the same output bit width in the version identifier generation, edge policy digest calculation and cloud policy digest calculation of this application, so that the digests are consistent in format and length, which facilitates the segmented comparison of policy digests in subsequent multidimensional consistency distance.

[0266] After obtaining the cloud policy version identifier, cloud policy summary, edge execution version identifier, and edge policy summary, the multidimensional consistency distance is generated as follows:

[0267] The inconsistency level of the strategy version identifier dimension is determined by taking a value of one when the cloud strategy version identifier and the edge execution version identifier are not equal, and zero when they are equal. The inconsistency level of the strategy summary dimension is determined by comparing the cloud strategy summary and the edge strategy summary in segments according to the byte segment length registered in the counter configuration table, and taking the ratio of the number of mismatched segments to the total number of segments as the inconsistency level of this dimension. The byte segment length is registered by the counter configuration table according to the integer division value of the output bit width of the summary algorithm. The inconsistency level of the profile locking record dimension is determined by taking a value of one when the profile locking record carried in the warehouse-level control instruction is not equal to the profile locking record of the corresponding battery warehouse in the single warehouse chargeable constraint set, and zero when they are equal. The inconsistency level of the control cycle identifier dimension is determined by taking a value of one when the control cycle identifier carried in the warehouse-level control instruction is not equal to the control cycle identifier of the current execution status backtracking, and zero when they are equal.

[0268] After obtaining the inconsistencies in the four dimensions mentioned above, the inconsistencies in the strategy version identifier dimension, strategy summary dimension, profile lock record dimension, and control cycle identifier dimension are multiplied by the corresponding weights registered in the cabinet configuration table, respectively. Then, the four products are added together to obtain the multidimensional consistency distance of the battery compartment within the current control cycle. The sum of the weights of the four dimensions is equal to one, and the value of each weight is not less than zero, ensuring that the value of the multidimensional consistency distance always falls between zero and one.

[0269] After obtaining the multidimensional consistency distance, the edge control and cloud strategy collaborative algorithm performs three-level hierarchical processing on the warehouse-level control command: when the multidimensional consistency distance is zero, a full execution state is written, so that the warehouse-level control command is issued according to its dynamic charging mode, actual compartment charging current, charging cut-off charge state, delayed start time, and expected battery swapping time; when the multidimensional consistency distance is greater than zero but does not exceed the hierarchical execution threshold registered in the counter configuration table and the inconsistency of the profile locking record dimension is zero, a degraded execution state is written, so that the warehouse-level control command is issued according to the de-rated or maintenance method; except for the above situations, a rejection execution state is written, and the warehouse-level control command does not enter the warehouse issuance record generation process.

[0270] The hierarchical execution threshold is registered in the cabinet configuration table according to the value range of the multidimensional consistency distance, and the value is no greater than one.

[0271] For warehouse-level control instructions whose strategy coordination gating result is "reject execution", the following three conditions are used to jointly determine whether to trigger version rollback synchronization: the cloud strategy version identifier is higher than the edge execution version identifier; the time interval from the previous successful synchronization to the current control cycle does not exceed the arbitration time window registered in the counter configuration table; and the cloud strategy summary of the current control cycle is equal to the cloud strategy summary of the previous control cycle.

[0272] Both the cloud-based policy version identifier and the edge execution version identifier are registered in the cabinet configuration table as monotonically increasing integers, with larger version identifier values ​​indicating newer versions. When all three conditions are met, the edge terminal updates its execution version identifier according to the cloud-based policy version identifier and writes the updated edge execution version identifier into the edge policy registration table; if any one condition is not met, the edge terminal retains its local execution version identifier.

[0273] After the tiered handling and version rollback synchronization are completed, edge-end version registration is performed on warehouse-level control instructions whose strategy coordination gating results are in full execution or downgraded execution status. Edge-end version registration refers to writing the corresponding battery compartment identifier, control cycle identifier, profile lock record, cloud strategy version identifier, edge execution version identifier, and strategy coordination gating result of the warehouse-level control instruction into the edge-end strategy registration table. Each written record is the edge-end strategy registration record. The edge-end strategy registration record and the strategy version binding record together constitute the version binding registration for the battery compartment within this control cycle.

[0274] The warehouse-level control instructions corresponding to the full execution status are written into the instruction waiting area, and the cabinet identifier, control cycle identifier, battery compartment identifier, profile lock record, cloud strategy version identifier, edge execution version identifier, cloud strategy summary, edge strategy summary, strategy collaboration gating result and version comparison time are sequentially written into the same version registration record and marked as version consistent status.

[0275] The position-level control instructions corresponding to the downgraded execution state are written to the instruction-pending area and marked as downgraded execution state, and the version registration record is marked as a version to be synchronized with the profile. The position-level control instructions corresponding to the rejected execution state are written to the version to be synchronized record, profile inconsistency record, or control cycle inconsistency record according to the inconsistency dimension. After completing the above writing, the strategy version binding record is obtained.

[0276] After the strategy version binding record is written, the corresponding record is retrieved in the strategy version binding record by the battery compartment identifier: first, the candidate version registration record is read from the strategy version binding record using the battery compartment identifier as the primary key, and then the profile lock record in the candidate version registration record is checked to see if it is consistent with the profile lock record carried by the compartment-level control instruction corresponding to the battery compartment in the instruction area to be issued.

[0277] When consistent, the dynamic charging mode, actual sub-compartment charging current, charging cut-off charge state, delayed start time, expected swappable time and control reason flag in the warehouse-level control instruction are written into the distribution field. The cloud strategy version identifier, edge execution version identifier, strategy collaboration gating result and profile locking record in the strategy version binding record are written into the version field and written into the pending distribution status to obtain the warehouse distribution record.

[0278] In case of inconsistency, the position-level control instruction will be written to the position binding exception record, and no position issuance record will be generated.

[0279] Next, based on the battery compartment identifier, the dynamic charging mode, actual compartment charging current, charging cutoff state of charge, delayed start time, estimated swappable time, and control reason flag are read from the compartment issuance records one by one, and these are written into the charging mode to be executed, the charging current to be executed, the charging cutoff state of charge to be executed, and the start time to be executed, respectively. If the estimated swappable time exists in the compartment issuance record, the estimated swappable time is written into the execution end time; if the estimated swappable time does not exist in the compartment issuance record, the charging cutoff state of charge is written into the execution end condition. After completing the above writing, the charging mode to be executed, the charging current to be executed, the charging cutoff state of charge to be executed, the start time to be executed, the execution end time or the execution end condition, the control reason flag, the profile lock record, and the strategy version binding status are written into the same record to obtain the instruction execution baseline record.

[0280] The instruction execution baseline record is encapsulated into a data frame based on the battery compartment identifier. This data frame contains the cabinet identifier, control cycle identifier, battery compartment identifier, required charging mode, required charging current, required charging cutoff charge state, required start time, and image lock record. The data frame is then sent to the corresponding charging control node according to the node address in the node address table, and the system waits for the charging control node to return an instruction reception confirmation record. Battery compartments that receive an instruction reception confirmation record within the confirmation time window registered in the cabinet configuration table are written into the execution status acquisition queue and a successful delivery marker is added. Battery compartments that do not receive an instruction reception confirmation record are written into a delivery error record, which includes the battery compartment identifier, control cycle identifier, image lock record, and delivery timeout marker.

[0281] The execution status acquisition queue initiates execution status acquisition sequentially according to the battery compartment identifier. Specifically, the node address of the corresponding charging control node is read from the node address table according to the battery compartment identifier, and a status read command is sent to the charging control node. The status read command is written into the control cycle identifier, battery compartment identifier, profile lock record, and status read field. The charging control node returns the actual charging mode, actual charging current, actual start time, current state of charge, current voltage, current temperature, charging operation status, and node alarm status based on the status read field. The returned data is written into the same record along with the control cycle identifier, cabinet identifier, battery compartment identifier, battery identifier, profile lock record, and data reception time identifier to obtain the execution status record. When writing the execution status record, a degraded execution status is written for battery compartments with a strategy coordination gating result of degraded execution status; a deferred execution status is written for battery compartments with a delayed charging mode; a non-charging execution status is written for battery compartments with a non-charging mode; and a status acquisition missing mark is written for battery compartments where the status read command was not received within the status read time window registered in the cabinet configuration table.

[0282] After obtaining the execution status record, records with the same control cycle identifier, battery compartment identifier, and image lock record are retrieved from the instruction execution baseline record according to the battery compartment identifier. For battery compartments with corresponding records, the execution status record and the instruction execution baseline record are written into the same associated cache. The actual charging mode and the required charging mode, the actual charging current and the required charging current, the actual start time and the required start time, the current state of charge and the required charging cutoff state, and the node alarm status and the required termination condition are written into the corresponding comparison fields to obtain the execution status associated record. For battery compartments without corresponding records, the execution status record is written into the unassociated execution record and a non-association reason flag is added. For battery compartments with abnormal issuance records, downgraded execution status, delayed execution status, non-charging execution status, or missing status acquisition flags, the corresponding status is written into the execution status associated record.

[0283] Finally, a collection and sorting batch is established according to the control cycle identifier and cabinet identifier, and a feedback merging record is established within the collection and sorting batch according to the battery compartment identifier as the primary key; the warehouse distribution record corresponding to the same battery compartment is written into the distribution field of the feedback merging record, the instruction execution benchmark record is written into the benchmark field, the execution status record is written into the execution field, and the execution status association record is written into the association field, and the merging is aligned according to the above four fields.

[0284] After merging, abnormal issuance records, unassociated execution records, and missing status acquisition markers are written to the corresponding abnormal fields according to the battery compartment identifier. The strategy version binding status, instruction issuance status, actual charging mode, actual charging current, actual start-up time, current state of charge, current temperature, node alarm status, execution abnormality marker, and data reception time identifier are written to the feedback field. When multiple execution status records exist in the same battery compartment, the execution status record with the later reception time is retained according to the data reception time identifier, and the remaining execution status records are written to the duplicate record index. After completing the above processing, charging execution feedback data is obtained and written to the feedback status determination input buffer as input for the next step of determining the feedback status of the charging execution feedback data.

[0285] The charging execution feedback data is used to determine the feedback status and obtain the charging feedback determination result.

[0286] The feedback status determination refers to the process of comparing the actual execution field corresponding to each battery compartment in the charging execution feedback data with the field to be executed in the corresponding compartment-level control instruction in the compartment charging control instruction, and classifying each battery compartment into one of the following states according to the deviation type: normal execution state, derating trigger state, delayed trigger state, charging stop state, or abnormal state.

[0287] The charging execution feedback data refers to the data set obtained by merging the data according to the battery compartment identifier after the execution status is collected. Each battery compartment corresponds to one feedback record. The feedback record includes control cycle identifier, cabinet identifier, battery compartment identifier, battery identifier, profile lock record, strategy version binding status, instruction issuance status, actual charging mode, actual charging current, actual start time, current state of charge, current temperature, node alarm status, execution anomaly marker, and data reception time identifier.

[0288] The charging feedback determination result refers to the data set obtained by encapsulating the battery compartment identifier after the feedback status determination. Each battery compartment corresponds to a compartment feedback determination record. The compartment feedback determination record includes control cycle identifier, cabinet identifier, battery compartment identifier, battery identifier, image lock record, feedback status record, feedback deviation record, feedback status flag, auxiliary status field, and abnormal reason field.

[0289] After receiving the charging execution feedback data, the cabinet controller reads the charging execution feedback data corresponding to the current control cycle from the feedback status determination input buffer, and reads the charging execution feedback data according to the battery compartment identifier to obtain the feedback status record.

[0290] Then, the execution deviation of the feedback status record and the corresponding compartment charging control command is extracted to obtain the feedback deviation record.

[0291] Then, the feedback deviation records are classified into states to obtain feedback state labels.

[0292] Finally, the feedback status markers are summarized to obtain the charging feedback determination result, and the charging feedback determination result is written into the control output encapsulation input buffer as the input for the control output encapsulation of the charging feedback determination result in the next step.

[0293] Specifically, the cabinet controller reads the required charging mode, actual charging mode, required charging current, actual charging current, required delayed start time, actual start time, current state of charge, current temperature, node alarm status, and command issuance status from the feedback deviation record, and generates a feedback status flag according to the following rules: when the actual charging mode is consistent with the required charging mode, the actual charging current falls within the allowable deviation range corresponding to the required charging current, the current temperature does not trigger a temperature over-limit alarm, and the node alarm status is empty, the feedback status flag is written to the normal execution status;

[0294] When the actual charging current is lower than the required charging current and the node alarm status does not trigger a prohibited charging alarm, the feedback status flag is written to the derating trigger status; when the required delayed start time has been reached, the actual charging current is still zero and the charging permission status is not the prohibited charging status, the feedback status flag is written to the delayed trigger status.

[0295] When the node alarm status is temperature over-limit, voltage over-limit, insulation abnormality, or the charging permission status is charging prohibited, the feedback status flag will be written to the stopped charging status.

[0296] When the feedback status record is missing the control cycle identifier, battery compartment identifier, actual charging current, or actual charging mode, the feedback status mark will be written to determine the abnormal status, and the missing field name will be written to the abnormal reason field.

[0297] Combination Figure 2 The above-described comparison chart of the battery charging control method for battery swapping cabinets based on the Internet of Things and existing technologies shows that the black bars represent the technical effects of the present invention, while the gray bars represent existing technologies. Figure 2 It can be seen that the present invention is superior to the prior art to a certain extent.

[0298] Example 2, based on Example 1, further explains the battery charging control method for battery swapping cabinets, including the encapsulation of the charging feedback judgment result, the judgment rules for generating the battery swapping cabinet charging control result, and the field writing entry points. The step names, terminology meanings, and data flow are consistent with Example 1. The specific scheme is as follows:

[0299] In one embodiment, the step of encapsulating the control output of the charging feedback determination result to generate the battery swapping cabinet charging control result includes:

[0300] The control output encapsulation refers to the process of organizing the control quantities of each battery compartment in the next control cycle based on the feedback status flag, including the charging mode, current limit, charging cut-off state of charge, and delayed start time, and encapsulating them into the charging control result of the battery swapping cabinet according to the cabinet identifier, control cycle identifier, and battery compartment identifier.

[0301] The charging control result of the battery swapping cabinet refers to the data set obtained after the control output is packaged, the control result batch is established according to the cabinet identifier and the control cycle identifier, and the control output records corresponding to each battery compartment in the current control cycle are written into the same control result package according to the battery compartment identifier. The data set records the feedback status mark, execution result, control quantity type and control quantity of each battery compartment in the current control cycle, and serves as the reference input for the configuration of compartment control parameters in the next control cycle.

[0302] When classifying the feedback deviation records, feedback status markers are generated according to the following rules: Normal execution status is written when the actual charging mode matches the required charging mode, the actual charging current falls within the allowable deviation range corresponding to the required charging current, the current temperature does not trigger a temperature over-limit alarm, and the node alarm status is empty. Derating trigger status is written when the actual charging current is lower than the required charging current and the node alarm status does not trigger a prohibited charging alarm. Delayed trigger status is written when the required delayed start time has elapsed, the actual charging current is still zero, and the charging permission status is not a prohibited charging status. Stopped charging status is written when the node alarm status is a temperature over-limit, voltage over-limit, insulation abnormality, or the charging permission status is a prohibited charging status. Abnormal status is written when the feedback status record lacks a control cycle identifier, battery compartment identifier, actual charging current, or actual charging mode, and the missing field name is written to the abnormal reason field.

[0303] When the same battery compartment simultaneously meets two or more types of feedback status flag writing conditions, the writing status is determined in the order of abnormal status, stopped charging status, delayed trigger status, derating trigger status, and normal execution status, and the remaining hit conditions are recorded as auxiliary status fields in the corresponding feedback status record of the battery compartment.

[0304] The foregoing has shown and described the basic principles, main features, and advantages of the present invention. Those skilled in the art should understand that the present invention is not limited to the above embodiments. The embodiments and descriptions in the specification are merely preferred examples and are not intended to limit the invention. Various changes and modifications can be made to the invention without departing from its spirit and scope, and all such changes and modifications fall within the scope of the present invention as claimed. The scope of protection of the present invention is defined by the appended claims and their equivalents.

Claims

1. A battery charging control method for a battery swapping cabinet based on the Internet of Things, characterized in that, include: Acquire IoT data from the cabinet and process its operational status to obtain the cabinet operational dataset; A single-warehouse refillability constraint is constructed on the aforementioned counter-end operation dataset to obtain a single-warehouse refillability constraint set; The charging priority is sorted according to the charging constraints of the single compartment to obtain the battery charging queue; Configure compartment control parameters for the battery charging queue and generate compartment charging control commands; The execution status of the compartment charging control command is collected to obtain charging execution feedback data; The charging execution feedback data is used to determine the feedback status and obtain the charging feedback determination result. The charging feedback determination result is encapsulated into a control output to generate the charging control result of the battery swapping cabinet.

2. The battery charging control method for the battery swapping cabinet according to claim 1, characterized in that, The single-warehouse refillability constraint is constructed by performing a single-warehouse refillability constraint on the aforementioned counter operation dataset, resulting in a single-warehouse refillability constraint set, including: The battery compartment identifiers of the cabinet-side operation dataset are merged to obtain the battery compartment operation records corresponding to each battery compartment. The battery compartment operation records are split into state sources to obtain battery-side state records, compartment-side state records, and demand-side state records; The rechargeability constraints are extracted from the battery-side state records to obtain the battery-side rechargeability constraints. Thermal coupling constraints are extracted from the warehouse-side state records to obtain warehouse-side thermal coupling constraints; The energy replenishment constraint is extracted from the demand-side state record to obtain the demand-side energy replenishment constraint; Write the battery-side rechargeability constraint, the storage location-side thermal coupling constraint, and the demand-side energy replenishment constraint into the corresponding battery storage operation record to obtain a battery storage constraint profile. The state of the battery compartment constraint profile is written and the version is locked to obtain a single compartment chargeable constraint set.

3. The battery charging control method for the battery swapping cabinet according to claim 2, characterized in that, The battery compartment constraint profile is written to a state and locked to a version, including: The single-compartment constraint profile locking algorithm binds the real-time battery status data, historical aging characteristics, compartment thermal coupling relationship, cabinet environment status and site recharge requirements in the battery compartment constraint profile to the same period to obtain the same period profile record. The battery compartment identifier, battery identifier, and control cycle identifier in the same period image record are concatenated to obtain the image identifier field; Version identifiers are generated for the portrait identifier field and the portrait records in the same period to obtain a portrait lock record; The image lock record is written into the corresponding battery compartment constraint image to complete the construction of the single compartment rechargeable constraint set.

4. The battery charging control method for the battery swapping cabinet according to claim 1, characterized in that, The charging priority is sorted according to the charging constraints of the single compartment to obtain the battery charging queue, including: The rechargeable state is filtered by the single-compartment rechargeable constraint set to obtain a sortable battery compartment set; The sortable battery compartment set is marked with a power replenishment eligibility label to obtain a power replenishment candidate compartment set; The set of candidate charging stations is marked with fast charging conflict flags and current limit flags to obtain a set of restricted charging stations; The limited charging bay set is sorted according to charging eligibility, fast charging conflict and current limit to obtain a grouped charging sequence; The grouped charging sequences are encapsulated into a queue to obtain a battery charging queue.

5. The battery charging control method for a battery swapping cabinet according to claim 4, characterized in that, The restricted charging station set is sorted according to charging eligibility, fast charging conflict, and current limitation order, including: The thermally coupled conflict clusters are obtained by dividing the battery compartments with thermal impact correlation in the set of restricted energy replenishment compartments into conflict clusters using the thermally coupled conflict cluster staggering algorithm. The fast charging candidate status of the battery compartment within the thermally coupled conflict cluster is read to obtain the conflict cluster candidate compartment record; Fast charging eligibility is allocated to the candidate warehouse records of the conflict clusters to obtain the conflict cluster staggered peak allocation results; The collision cluster peak shifting results are written into the corresponding group charging sequence to complete the generation of the group charging sequence.

6. The battery charging control method for a battery swapping cabinet according to claim 1, characterized in that, Configure compartment control parameters for the battery charging queue and generate compartment charging control commands, including: The battery charging queue is read to obtain the compartment configuration object; Configure the charging mode, current limit, charging cut-off charge state and delayed start for the compartment configuration object to obtain the compartment control parameter record; The compartment control parameter records are encapsulated into instructions to generate compartment charging control instructions.

7. The battery charging control method for a battery swapping cabinet according to claim 6, characterized in that, Configure the charging mode, current limit, charging cutoff state of charge, and delayed start for the compartment configuration object, including: The energy replenishment execution positions are determined by the demand gap reverse-engineering sub-compartment current orchestration algorithm for the sub-compartment configuration object, resulting in a set of energy replenishment execution positions. Dynamic charging mode configuration is performed on the set of energy replenishment execution positions to obtain dynamic charging mode records; The dynamic charging mode record is configured with current limiting, charging cutoff state of charge and expected swappable time to obtain the energy replenishment control parameter record. For sub-compartment configuration objects that are not written into the aforementioned energy replenishment execution unit set but have charging permission, perform maintenance charging or postpone charging configuration to obtain non-urgent control parameter records; The generation of the compartment control parameter record is completed based on the energy replenishment control parameter record and the non-urgent control parameter record.

8. The battery charging control method for a battery swapping cabinet according to claim 1, characterized in that, The execution status of the compartment charging control commands is sampled to obtain charging execution feedback data, including: The strategy version binding is performed on the compartment charging control command to obtain the strategy version binding record; The position distribution record is generated by binding the strategy version to the record. A baseline record is generated from the aforementioned warehouse allocation record to obtain the instruction execution baseline record; The execution status record is obtained by collecting the status of the instruction execution baseline record; The execution status records are associated with warehouse locations to obtain status association records; The state association records are collected and processed to obtain charging execution feedback data.

9. The battery charging control method for a battery swapping cabinet according to claim 8, characterized in that, The strategy version binding for the compartment charging control commands includes: The consistency of the cloud policy version identifier, cloud policy summary, profile locking record and control cycle identifier carried by the compartment charging control command is compared by the edge control and cloud policy collaborative algorithm to obtain the multi-dimensional consistency distance. Based on the multidimensional consistency distance, the compartment charging control command is divided into full execution state, degraded execution state, or rejection execution state to obtain the strategy version matching result. For the sub-compartment charging control instructions that belong to the full execution state or the degraded execution state in the strategy version matching results, perform edge end version registration to obtain edge end strategy registration records; The edge-end policy registration record is bound to the corresponding battery compartment to obtain the compartment distribution record; The command execution baseline record is generated based on the position issuance record.

10. The battery charging control method for a battery swapping cabinet according to claim 1, characterized in that, The charging execution feedback data is used to determine the feedback status, and a charging feedback determination result is obtained, including: The charging execution feedback data is read to obtain the feedback status record; The execution deviation of the feedback status record and the corresponding compartment charging control command is extracted to obtain the feedback deviation record; The feedback deviation records are categorized by state to obtain feedback state labels; The feedback status markers are summarized to obtain the charging feedback determination result.

Images