Energy storage cabinet efficient thermal management method and system based on battery state and health management

Abstract

This invention relates to the field of energy storage systems and electrochemical power sources, specifically to a high-efficiency thermal management method and system for energy storage cabinets based on battery state and health management. The method includes: collecting cell temperature, voltage, and current to obtain first data; calculating internal resistance parameters and capacity decay parameters to generate health state parameters; estimating heat generation power and forming a spatial distribution based on the first data and health state parameters; clustering the spatial distribution to obtain partitioning results; establishing a linear programming model based on the partitioning results; solving for the partitioned cooling resource allocation under at least one constraint of a total airflow limit and a total medium flow limit; and adjusting the cooling actuators accordingly to achieve partitioned cooling. This invention can reduce hotspot risk and reduce ineffective cooling energy consumption.

Description

Technical Field

[0001] This invention relates to the field of energy storage systems and electrochemical power sources, specifically to an efficient thermal management method and system for energy storage cabinets based on battery status and health management. Background Technology

[0002] Under high-rate charging and discharging, fluctuating ambient temperatures, and long-term cyclic operating conditions, differences in internal resistance and capacity between individual battery cells inevitably occur in energy storage cabinets, leading to localized heat generation and uneven temperature rise. Thermal management, as a critical aspect of ensuring safety and lifespan, directly impacts the available power, maintenance costs, and accident risks of energy storage systems. Existing solutions often employ uniform air supply or fixed-flow cooling for the entire cabinet, with control based primarily on temperature thresholds. This makes it difficult to adjust cooling based on individual cell health differences, easily resulting in persistent hot spots and over-cooling, leading to increased energy consumption, accelerated degradation, and the accumulation of thermal runaway risks. Summary of the Invention

[0003] This invention provides an efficient thermal management method and system for energy storage cabinets based on battery status and health management, which is used to at least solve the problem of how to achieve zoned thermal management of energy storage cabinets based on individual cell health differences and heat generation distribution under conditions of limited cooling capacity.

[0004] In a first aspect, the present invention provides an efficient thermal management method for an energy storage cabinet based on battery status and health management, the method comprising: The temperature, voltage, and current data of multiple battery cells in the energy storage cabinet are acquired to obtain the first data. Based on the first data, the internal resistance parameters and capacity decay parameters of each battery cell are calculated to generate health status parameters. Based on the first collected data and health status parameters, the heat generation power of each battery cell is estimated, the spatial distribution of heat generation power is generated, and the heat generation power spatial distribution is clustered and grouped to obtain the partitioning results. A linear programming model is established based on the zoning results. The allocation of cooling resources for each zone is obtained under the constraint of cooling capacity. The cooling capacity constraint includes at least one of the upper limit of total cooling air volume and the upper limit of total cooling medium flow. The cooling actuator is adjusted according to the allocation of cooling resources for each zone to achieve zoned cooling.

[0005] In one possible implementation, calculating the internal resistance parameter of a battery cell based on the first acquired data includes: selecting a stable range before the current data changes and a stable range after the current data changes, calculating the average voltage data and the average current data within the stable range respectively, and calculating the internal resistance parameter based on the ratio of the difference between the average voltage data and the difference between the average current data.

[0006] In one possible implementation, calculating the capacity decay parameter of a battery cell based on the first collected data includes: integrating the current data over time within a preset charge / discharge range to obtain the available capacity, and determining the capacity decay parameter based on the available capacity and the rated capacity, wherein the rated capacity is the nominal capacity of the battery cell.

[0007] In one possible implementation, estimating the heat generation power of a single battery cell based on the first collected data and health status parameters includes: calculating the resistive heating power based on current data and internal resistance parameters, and performing temperature compensation on the internal resistance parameters based on temperature data to obtain the heat generation power.

[0008] In one possible implementation, estimating the heat generation power also includes: correcting the resistance heating power based on a linear regression model to obtain the heat generation power, wherein the inputs of the linear regression model include internal resistance parameters and capacity decay parameters, and the model parameters of the linear regression model are determined based on historical operating data.

[0009] In one possible implementation, clustering based on the spatial distribution of heat generation power includes: obtaining battery cell installation location data; when clustering based on the spatial distribution of heat generation power, using both the battery cell installation location data and the spatial distribution of heat generation power as clustering inputs, and using K-means clustering to obtain the partitioning results.

[0010] In one possible implementation, the objective of the linear programming model is to minimize cooling resource consumption. The constraints of the linear programming model include cooling capacity constraints and the requirement that the estimated temperature of each zone does not exceed a preset temperature threshold. The estimated temperature of each zone is determined based on the spatial distribution of heat generation power and the allocation of cooling resources to the candidate zones.

[0011] In one possible implementation, adjusting the cooling actuator according to the zoned cooling resource allocation includes: the cooling actuator includes at least one of a fan, a duct damper, a valve, and a pump; when the zoned cooling resource allocation includes cooling airflow, determining the fan speed setpoint and the duct damper opening setpoint and performing adjustment; when the zoned cooling resource allocation includes cooling medium flow rate, determining the valve opening setpoint and the pump speed setpoint and performing adjustment.

[0012] In one possible implementation, the method of the present invention further includes: acquiring temperature data of multiple battery cells again after adjusting the cooling actuator to obtain second acquisition data; determining the temperature difference between intervals based on the second acquisition data; and re-performing clustering and solving the linear programming model if the temperature difference between intervals is greater than a preset temperature difference threshold.

[0013] Secondly, the present invention provides a high-efficiency thermal management system for energy storage cabinets based on battery status and health management, for implementing a high-efficiency thermal management method for energy storage cabinets based on battery status and health management, the system comprising: The data acquisition module is used to acquire temperature, voltage and current data of multiple battery cells in the energy storage cabinet, generate first acquisition data, and calculate the internal resistance parameters and capacity decay parameters of each battery cell based on the first acquisition data to generate health status parameters. The heat load estimation module is used to estimate the heat generation power of each battery cell based on the first collected data and health status parameters, generate the spatial distribution of heat generation power, and perform clustering and grouping based on the spatial distribution of heat generation power to obtain the partitioning results. The resource allocation control module is used to establish a linear programming model based on the partitioning results, solve for the partitioned cooling resource allocation under cooling capacity constraints, where the cooling capacity constraints include at least one of the upper limit of total cooling air volume and the upper limit of total cooling medium flow rate, and adjust the cooling actuator according to the partitioned cooling resource allocation to realize partitioned cooling.

[0014] Compared with the prior art, the advantages and beneficial effects of the present invention are as follows: By jointly acquiring temperature, voltage, and current data, and calculating internal resistance and capacity decay parameters to generate health status parameters, quantifiable input for individual unit differences is achieved. By estimating heat generation power based on health status parameters and generating a spatial distribution of heat generation power, a calculable characterization of heat load distribution is achieved. By clustering and grouping the spatial distribution of heat generation power to obtain partitioning results, executable partitioning control boundaries are realized. By introducing cooling capacity constraints into a linear programming model and solving for the partitioned cooling resource allocation, optimal resource allocation under capacity-constrained conditions is achieved. By adjusting cooling actuators according to the allocation amount, partitioned directional cooling is realized, thereby suppressing hot spots, reducing ineffective cooling energy consumption, and improving operational safety margins. Attached Figure Description

[0015] Figure 1 This is a schematic diagram of the execution flow of the method of the present invention; Figure 2 This is a structural block diagram of the system of the present invention. Detailed Implementation

[0016] The embodiments of the present disclosure will now be described with reference to the accompanying drawings. However, it should be understood that these descriptions are exemplary only and are not intended to limit the scope of the disclosure. In the following detailed description, numerous specific details are set forth to provide a thorough understanding of the embodiments of the present disclosure for ease of explanation. However, it will be apparent that one or more embodiments may be practiced without these specific details. Furthermore, descriptions of well-known structures and techniques are omitted in the following description to avoid unnecessarily obscuring the concepts of the present disclosure.

[0017] The terminology used herein is for the purpose of describing particular embodiments only and is not intended to limit this disclosure. The terms “comprising,” “including,” etc., as used herein indicate the presence of the stated features, steps, operations, and / or components, but do not exclude the presence or addition of one or more other features, steps, operations, or components.

[0018] All terms used herein (including technical and scientific terms) have the meanings commonly understood by those skilled in the art, unless otherwise defined. It should be noted that the terms used herein are to be interpreted in a manner consistent with the context of this specification, and not in an idealized or overly rigid way.

[0019] During the operation of an energy storage cabinet, battery status typically refers to the characterization of the current operating state of a single battery cell, including at least directly measurable operating quantities such as temperature, voltage, and current, and can further reflect operating characteristics such as state of charge and power load changes. Health management, on the other hand, focuses on the quantification of long-term performance changes in battery cells and provides support for maintenance decisions. It commonly uses changes in internal resistance and available capacity as key indicators to characterize the degree of aging, differentiated degradation trends, and the accumulation of potential risks. Battery status provides real-time information on "what is happening now," while health management provides evolutionary information on "what has become in the long term." Together, they determine the actual heat generation level and thermal sensitivity boundary of a battery cell. Introducing battery status information and health management information into thermal management control allows cooling decisions to shift from a single temperature response to proactive allocation based on heat sources and health differences, thus providing a reliable input basis for zoned cooling and optimized resource allocation in energy storage cabinets.

[0020] like Figure 1 As shown, an efficient thermal management method for energy storage cabinets based on battery state and health management is disclosed, the method comprising: The temperature, voltage, and current data of multiple battery cells in the energy storage cabinet are acquired to obtain the first data. Based on the first data, the internal resistance parameters and capacity decay parameters of each battery cell are calculated to generate health status parameters. Each battery cell within the energy storage cabinet is equipped with a temperature acquisition point, and its operating voltage and current are acquired through voltage and current acquisition circuits. During data acquisition, the channels are synchronized in time, and the data is aggregated according to the battery cell number to form the initial data set. Based on this initial data set, the internal resistance and capacity decay parameters of each battery cell are calculated. These parameters are then combined according to the battery cell number to obtain the health status parameters, which serve as input for subsequent thermal management calculations.

[0021] The calculation of the internal resistance parameter of a single battery cell based on the first collected data includes: selecting a stable range before the current data changes and a stable range after the current data changes; calculating the average voltage data and the average current data within the stable range respectively; and calculating the internal resistance parameter based on the ratio of the difference between the average voltage data and the difference between the average current data.

[0022] The internal resistance parameter is calculated using a current change event-driven approach. First, changes in the current data are detected in the initial data acquisition. These changes are triggered when the current difference exceeds a preset threshold, which can be set proportionally to the rated current.

[0023] For each point of change, two stable intervals are searched forward and backward. A stable interval must have a duration that meets a preset time limit, and the current fluctuation amplitude within the interval must not exceed a preset fluctuation threshold. Voltage data must also meet the same stability requirements to avoid errors caused by transient noise. After the stable intervals are determined, the average voltage and current data are calculated within the stable interval before the current change, and the average voltage and current data are calculated within the stable interval after the current change.

[0024] Then, the difference between the average values ​​of the voltage data and the average value of the current data are calculated, and the internal resistance parameter is calculated accordingly. The core expression for calculating the internal resistance parameter is:

[0025] Where is the internal resistance parameter; is the average difference of voltage data; and is the average difference of current data. To ensure the usability of the calculation results, when the absolute value of is less than the minimum current difference threshold, the internal resistance parameter is not output for this event, and the system waits for the next current change event.

[0026] To reduce random errors from single events, internal resistance parameters from multiple events can be accumulated within a preset time window, and the final internal resistance parameter can be obtained using the median or weighted average. The weights can be determined based on the length and fluctuation amplitude within the stable interval; the longer the stable interval and the smaller the fluctuation, the greater the weight. Internal resistance parameters can also be compensated for using temperature data. Temperature compensation is implemented using a lookup table, where entries are obtained from the factory calibration of the battery cells. The input is the temperature range, and the output is the internal resistance correction coefficient. The final internal resistance parameter is written into the health status parameters according to the battery cell number and stored in association with the corresponding timestamp for easy use in subsequent heat generation estimation.

[0027] The calculation of the capacity decay parameters of a single battery cell based on the first collected data includes: integrating the current data over time within a preset charge-discharge range to obtain the usable capacity, and determining the capacity decay parameters based on the usable capacity and the rated capacity, wherein the rated capacity is the nominal capacity of the single battery cell.

[0028] The capacity decay parameter is calculated using a preset charge / discharge range as the boundary. This preset range can be defined by either a charging start event and a charging end event, or a discharging start event and a discharging end event. The event source can be a status flag from the battery management system, or it can be determined by a combination of current direction and voltage change trends.

[0029] To reduce boundary uncertainties, both the start and end points of the charge / discharge interval must meet stability conditions. These stability conditions include current variation amplitudes being less than a preset threshold and durations reaching a preset duration. After determining the charge / discharge interval, the current data within the interval is integrated over time according to the sampling period to obtain the available capacity. Time integration is implemented using discrete accumulation. The current value at each sampling point is multiplied by the sampling period to obtain the capacity increment, and then the capacity increments within the interval are summed to obtain the available capacity.

[0030] To avoid symbolic ambiguity, charging current and discharging current are standardized to the same notation before calculation. Capacity degradation parameters are determined by usable capacity and rated capacity. Rated capacity uses the nominal capacity of the individual battery cell, derived from the cell's specifications or manufacturer's parameters. Capacity degradation parameters can be recorded in differential or proportional form. The differential form represents the absolute decrease in usable capacity, while the proportional form represents the relative degree of degradation. To improve consistency, only one form is consistently used within the same system.

[0031] The capacity decay parameter can be updated with a minimum update cycle to avoid fluctuations caused by frequent updates within a short period. If abnormal data exists within the preset charge / discharge range, such as temperature data exceeding the safe range or voltage data showing sampling omissions, the capacity decay parameter will not be updated, and the reason for the anomaly will be recorded. Coulomb efficiency correction can also be incorporated into the calculation of usable capacity. The coulomb efficiency correction coefficient can be obtained by looking up a table according to the temperature range and rate range; the table entries are generated from historical operational calibrations. The final capacity decay parameter is written into the health status parameter by the battery cell number, and together with the internal resistance parameter, constitutes the health status parameter entry for that battery cell.

[0032] Based on the first collected data and health status parameters, the heat generation power of each battery cell is estimated, the spatial distribution of heat generation power is generated, and the heat generation power spatial distribution is clustered and grouped to obtain the partitioning results. Based on the initial data collection and health status parameters, the heat generation power of each battery cell is estimated, and the heat generation power is mapped according to the battery cell number and installation location to form a spatial distribution of heat generation power. Subsequently, clustering is performed based primarily on the spatial distribution of heat generation power, dividing multiple battery cells into several zones, resulting in zoning results. The zoning results are used to reflect the differences in heat load in different areas within the energy storage cabinet and serve as input for subsequent cooling resource allocation and zoning control.

[0033] The estimation of the heat generation power of a single battery cell based on the first collected data and health status parameters includes: calculating the resistive heating power based on current data and internal resistance parameters, and performing temperature compensation on the internal resistance parameters based on temperature data to obtain the heat generation power.

[0034] In one embodiment, the heat generation power estimation is based on resistive heating. First, current and temperature data are extracted from the initial data collection by cell number, and internal resistance parameters are extracted from the health status parameters. To avoid instantaneous noise amplification, the current data is averaged using a sliding window, with the window length determined by the sampling period and the control period.

[0035] The resistance heating power is then calculated, and temperature compensation is introduced to correct the internal resistance parameter, making the heat generation power closer to the actual state under the current temperature conditions. The core calculation expression for resistance heating power is:

[0036] Where is the resistance heating power; is the average current value of the current data within the window; and is the internal resistance parameter.

[0037] Temperature compensation is achieved using a lookup table. The input to the lookup table is the temperature range corresponding to the temperature data, and the output is the internal resistance correction coefficient. The internal resistance parameter after compensation is obtained by multiplying the internal resistance parameter by the internal resistance correction coefficient. The compensated internal resistance parameter is then substituted into the aforementioned expression to update the resistive heating power, thus obtaining the heat generation power. To ensure feasibility, the temperature range and internal resistance correction coefficient are derived from the factory calibration data of the individual battery cells or the calibration data before system commissioning. The calibration process can be achieved by applying a preset current under different temperature conditions and recording the voltage response.

[0038] If the temperature data exceeds the calibration coverage range, it is extrapolated using the correction factor of the most recent temperature range, and the extrapolation is recorded for subsequent maintenance. If the current data is close to zero, the heat generation power is set to zero or a preset minimum value to avoid meaningless small fluctuations from entering subsequent clustering. The output of heat generation power is stored using the battery cell number as an index and associated with a timestamp to form a heat load sequence that updates over time, providing a basis for spatial distribution construction.

[0039] Estimating the heat generation power also includes: correcting the resistance heating power based on a linear regression model to obtain the heat generation power, wherein the inputs of the linear regression model include internal resistance parameters and capacity decay parameters, and the model parameters of the linear regression model are determined based on historical operating data.

[0040] In one embodiment, the heat generation power estimation incorporates linear regression correction based on the resistive heating power to absorb the residual effects of heating deviation caused by capacity decay and internal resistance estimation error. The input to the linear regression model consists of internal resistance parameters and capacity decay parameters, and the output is the correction amount of heat generation power to resistive heating power.

[0041] The parameters of the linear regression model are determined using historical operating data. This historical operating data includes at least the current data, temperature data, internal resistance parameters, capacity decay parameters, and the cooling execution status corresponding to time for the same energy storage cabinet under different load conditions.

[0042] To obtain supervisory parameters usable for regression, a period of relatively stable cooling performance was selected from historical data, and the trend of temperature data change was used as the correction basis: when the resistive heating power is similar but the rate of temperature rise shows a systematic difference, this difference is attributed to the additional heating or heat dissipation efficiency changes brought about by the health state parameters, forming the regression objective. The core expression of linear regression is:

[0043] Where is the heat generation power; is the resistance heating power; is the internal resistance parameter; is the capacity decay parameter; is a constant term; and and are regression coefficients. The regression coefficients are fixed after offline training and can also be updated according to a preset period.

[0044] During updates, historical data is first cleaned to remove samples with missing temperature data, frequent switching of cooling execution states, and abnormal fluctuations in current data, thus avoiding the introduction of uncontrollable disturbances. To control overfitting, the range of regression coefficient values ​​is limited during training, and a minimum sample size threshold is set. If the sample size in the current running stage is insufficient for updates, the regression coefficients from the previous version are used. When calculating heat generation power, a baseline value is first obtained based on the resistance heating power, and then substituted into the regression expression to obtain the corrected heat generation power.

[0045] If the correction result is negative, it is truncated to zero to ensure physical rationality. The corrected heat generation power is still output according to the battery cell number and synchronized with the timestamp for subsequent spatial distribution and clustering input, ensuring that the mapping from health status parameters to heat generation power is clear and reproducible throughout the entire chain.

[0046] Clustering based on the spatial distribution of heat generation power includes: obtaining data on the installation location of individual battery cells; when clustering based on the spatial distribution of heat generation power, the data on the installation location of individual battery cells and the spatial distribution of heat generation power are used together as clustering inputs, and K-means clustering is used to obtain the partitioning results.

[0047] In one embodiment, clustering grouping uses the spatial distribution of heat generation power as the primary input, and incorporates battery cell installation location data to form a joint clustering input, ensuring that the partitioning results take into account both heat load differences and spatial feasibility. The battery cell installation location data comes from the energy storage cabinet structural configuration, including the battery cell's layer, column, or tray number, and can also be represented by discrete location indices in the energy storage cabinet coordinate system.

[0048] When generating the spatial distribution of heat generation power, the heat generation power of each battery cell is bound to the battery cell installation location data to form a spatially ordered heat load distribution sequence. Before clustering, the heat generation power is normalized to avoid clustering bias caused by differences in dimensions; the installation location data is discretely encoded and scaled so that location differences are reflected as proximity constraints in clustering.

[0049] The clustering algorithm uses K-means clustering. The number of clusters is determined based on the number of independently controllable partitions of the cooling actuator, such as the number of independent air supply partitions that can be formed by the duct damper or the number of independent loops that can be formed by the cooling medium distribution. During the initialization phase, several battery cells are spatially distributed as initial cluster centers to avoid cluster centers being concentrated in the same local area.

[0050] During the iteration phase, each battery cell is assigned to the cluster center closest to its joint features, and the cluster centers are updated based on the assignment results until the change in cluster centers is less than a preset threshold or the number of iterations reaches the upper limit. To improve executability, spatial continuity constraints are added during the iteration process: when a battery cell and its adjacent battery cells in the same layer are assigned to different clusters for a long time and the difference in heat generation power is less than a preset difference threshold, they are preferentially merged and adjusted according to the principle of spatial adjacency to reduce partition fragmentation.

[0051] After clustering, the system outputs partitioning results, including partition numbers and a set of battery cell numbers within each partition. It also outputs the representative heat generation power for each partition for subsequent cooling resource allocation. If the number of battery cells in a partition falls below a minimum threshold, that partition is merged into its nearest neighbor partition, and the representative heat generation power is recalculated to ensure stable and usable partitioning results. The output partitioning results are consistently mapped to the battery cell installation location data to ensure accurate positioning of the corresponding cooling execution area for subsequent control.

[0052] A linear programming model is established based on the zoning results. The allocation of cooling resources for each zone is obtained under the constraint of cooling capacity. The cooling capacity constraint includes at least one of the upper limit of total cooling air volume and the upper limit of total cooling medium flow. The cooling actuator is adjusted according to the allocation of cooling resources for each zone to achieve zoned cooling.

[0053] After the zoning results are generated, the process enters the zoning cooling resource allocation and execution control stage. First, based on the zoning results, the cooling objects are modeled into zones, and the heat load characterization of each zone is given in conjunction with the spatial distribution of heat generation power. Then, a linear programming model is established, using at least one of the cooling airflow and cooling medium flow rate of each zone as the allocation quantity to be solved, and the zoning cooling resource allocation quantity is obtained under cooling capacity constraints. The cooling capacity constraint includes at least an upper limit of total cooling airflow or an upper limit of total cooling medium flow rate. After the solution is completed, the zoning cooling resource allocation quantity is converted into control setpoints for the cooling actuators and issued for execution, realizing zoning cooling.

[0054] The goal of the linear programming model is to minimize cooling resource consumption. The constraints of the linear programming model include cooling capacity constraints and the requirement that the estimated temperature of each zone does not exceed a preset temperature threshold. The estimated temperature of each zone is determined based on the spatial distribution of heat production power and the allocation of cooling resources to the candidate zones.

[0055] In the implementation of the linear programming model, the objective is to minimize cooling resource consumption while ensuring that the estimated temperature of each zone does not exceed a preset temperature threshold. To facilitate engineering implementation, the estimated temperature of each zone is determined jointly by the spatial distribution of heat generation power and the allocation of cooling resources to candidate zones.

[0056] The specific procedure is as follows: First, calculate the estimated reference temperature for each zone based on the spatial distribution of heat generation power. This estimated reference temperature reflects the temperature level under the current load without increasing cooling resources. Then, establish a cooling effect coefficient for each zone to describe the impact of cooling resources on the estimated temperature. The cooling effect coefficient is derived from commissioning or operational calibration.

[0057] Commissioning calibration involves conducting a test on a specific zone with a slight increase in airflow or flow rate under constant load, recording the temperature drop, and converting it into the temperature change per unit airflow or flow rate. Operational calibration utilizes historical operating data to statistically analyze the correlation between airflow or flow rate changes and temperature changes over a period of stable cooling performance, generating updated values ​​for the zone's cooling effect coefficient.

[0058] In the linear programming model, the zone temperature estimate is expressed in linear form as the zone baseline temperature estimate minus the cooling airflow contribution and the cooling medium flow rate contribution. This ensures the model maintains its linear form and can be directly solved by commonly used solvers. The core expression of the linear programming model can be written as:

[0059] Wherein, is the number of zones; is the cooling air volume of zone 1; is the cooling medium flow rate of zone 2; is the upper limit of total cooling air volume; is the upper limit of total cooling medium flow rate; is the estimated baseline temperature of zone 1; is the air volume cooling effect coefficient of zone 1; is the flow rate cooling effect coefficient of zone 1; is the preset temperature threshold; is the air volume consumption weight; is the flow rate consumption weight.

[0060] If the system only has air cooling or only liquid cooling, then one type of variable can be fixed to zero while retaining the other constraints. To ensure solution stability, minimum and maximum values ​​can be set for air volume and flow rate, with the boundaries derived from the operating range of the fan, valve, and pump. The partitioned cooling resource allocation output is updated according to the control cycle and mapped consistently with the partitioned results to avoid mismatch of control objectives caused by partition changes.

[0061] Adjusting the cooling actuator according to the zoned cooling resource allocation includes: the cooling actuator includes at least one of a fan, a duct damper, a valve, and a pump; when the zoned cooling resource allocation includes cooling airflow, determining the fan speed setting value and the duct damper opening setting value and performing adjustment; when the zoned cooling resource allocation includes cooling medium flow rate, determining the valve opening setting value and the pump speed setting value and performing adjustment.

[0062] The cooling actuator is adjusted based on the zone's cooling resource allocation as input, and outputs a directly executable setpoint. The cooling actuator includes at least one or more of the following: fan, duct damper, valve, and pump. Adjustment of cooling airflow is typically achieved through a combination of fan speed and duct damper opening. First, the zone's air supply demand is determined based on the zone's cooling airflow; then, the fan speed setpoint is calculated using the fan performance curve and duct pressure loss curve.

[0063] Subsequently, the duct damper opening setpoint is calculated based on the zoned air supply requirements to ensure that different zones receive the same airflow ratio as the target. The fan performance curve and duct pressure loss curve can be obtained from the equipment's factory parameters or corrected through on-site calibration. Adjustment of the cooling medium flow rate is typically achieved through a combination of pump speed and valve opening. First, the zoned flow rate requirement is determined based on the zoned cooling medium flow rate; then, the pump speed setpoint is calculated based on the pump performance curve.

[0064] Subsequently, the valve opening setpoint is calculated based on the zone's flow requirements to ensure that the zone's flow meets the target allocation. The correspondence between valve opening and flow rate can be obtained through a one-time calibration and corrected using flow sensor data during operation. To avoid frequent actuator vibration, the setpoint is issued using a limiting and speed limiting strategy. The limiting is used to ensure that the fan speed, damper opening, valve opening, and pump speed do not exceed safe ranges.

[0065] Speed ​​limiting is used to restrict the variation range between adjacent control cycles, preventing temperature fluctuations caused by sudden changes in airflow and flow rate. After execution, the allocation of cooling resources and setpoints for this cycle are recorded and archived along with the partitioning results for subsequent calibration and backtracking.

[0066] The method of the present invention further includes: after completing the adjustment of the cooling actuator, acquiring temperature data of multiple battery cells again to obtain second acquisition data; determining the temperature difference between intervals based on the second acquisition data; and if the temperature difference between intervals is greater than a preset temperature difference threshold, re-executing clustering and re-executing the linear programming model solution.

[0067] In one embodiment, to ensure the closed-loop effectiveness of zoned cooling in actual operation, temperature data from multiple battery cells are acquired again after the cooling actuator adjustment is completed, forming a second set of collected data. This second set of collected data is still grouped by battery cell number, and the zoned temperature statistics for each zone are calculated based on the zoned results.

[0068] The zone temperature statistics can be the maximum temperature within the zone, the average temperature within the zone, or a combination of both. The system consistently uses one of these methods. Then, the temperature difference between zones is calculated based on the zone temperature statistics. The temperature difference between zones can be defined as the difference between the highest and lowest zone temperature statistics, or as the maximum difference between adjacent zones.

[0069] The temperature difference between zones is compared with a preset temperature difference threshold. The preset temperature difference threshold is determined by the thermal management objectives and may be given in conjunction with the allowable temperature difference of individual battery cells. If the temperature difference between zones is not greater than the preset temperature difference threshold, the current zone result and the current zone cooling resource allocation are maintained, and the normal update of the next control cycle begins. If the temperature difference between zones is greater than the preset temperature difference threshold, a recalculation process is triggered.

[0070] The recalculation process includes re-performing clustering and re-solving the linear programming model. During re-clustering, the spatial distribution of heat generation power is used, and temperature differences reflected in the second data acquisition can be overlaid to make the partition boundaries more closely match the current heat load distribution. When re-solving the linear programming model, the estimated baseline temperature for each partition is updated, and cooling resources are reallocated under cooling capacity constraints.

[0071] To avoid frequent re-partitioning due to repeated triggering, a minimum trigger interval or a threshold for the number of consecutive exceedances can be set. Re-clustering and re-solving are only performed when the temperature difference between partitions continuously exceeds the preset temperature difference threshold. Through this feedback mechanism, partitioned cooling not only performs feedforward allocation based on heat generation estimation but also corrects for deviations based on actual temperature response, ensuring long-term stable availability of partitioned cooling.

[0072] like Figure 2 As shown, an efficient thermal management system for energy storage cabinets based on battery status and health management is used to implement an efficient thermal management method for energy storage cabinets based on battery status and health management. The system includes: The data acquisition module is used to acquire temperature, voltage, and current data from multiple battery cells within the energy storage cabinet, generating initial data. Based on this initial data, it calculates the internal resistance and capacity decay parameters of each battery cell, generating health status parameters. The data acquisition module consists of a sensor front-end, signal conditioning circuitry, sampling and synchronization unit, and a local communication interface. Temperature acquisition typically uses surface-mount thermistors or digital temperature sensors, placed on the surface of the battery cells or near the busbar, and connected to the sampling channel via an isolated input. Voltage acquisition is achieved through a high-resistance voltage divider network and an isolation amplifier, ensuring electrical isolation between the high-voltage side and the control side. Current acquisition can be achieved using Hall effect current sensors or shunt resistors in conjunction with a differential amplifier, configured with anti-aliasing filtering and overvoltage protection. The sampling and synchronization unit is generally composed of multiple ADCs and sample-and-hold circuits. With clock synchronization, it ensures that multiple channels sample at the same time. The average value, integral and other calculations required to calculate the internal resistance parameters and capacity decay parameters are performed by an embedded processor or a dedicated microcontroller. The calculation results and the original sampled data together form the first acquisition data and health status parameters, which are output to the upper control unit via CAN, RS485 or Ethernet interface.

[0073] The heat load estimation module estimates the heat generation power of each battery cell based on the first acquired data and health status parameters, generates a spatial distribution of heat generation power, and performs clustering based on this spatial distribution to obtain partitioning results. The heat load estimation module is primarily composed of a processing unit, a storage unit, and a data interface unit, typically implemented as a high-performance microcontroller or industrial-grade processor on the main control board. The processing unit is responsible for calculations such as heat generation power calculation, temperature compensation lookup, and clustering based on the first acquired data and health status parameters. The storage unit stores calibration tables, historical operating segments, and model parameters, providing a traceable data foundation for heat generation estimation and clustering inputs. The data interface unit exchanges synchronous sampling results with the data acquisition module and outputs the spatial distribution and partitioning results of heat generation power to the subsequent control module. To adapt to the electromagnetic environment of the energy storage cabinet, this module is typically equipped with power isolation and interface isolation devices, and employs a watchdog timer and fault self-testing circuits to ensure continuous operation.

[0074] The resource allocation control module is used to establish a linear programming model based on the zoning results, and solve for the zoning cooling resource allocation under cooling capacity constraints. These constraints include at least one of the upper limit of total cooling airflow and the upper limit of total cooling medium flow. The module then adjusts the cooling actuators according to the zoning cooling resource allocation to achieve zoning cooling. The resource allocation control module consists of a control processor, actuator drivers, and feedback interfaces. It can be integrated with the heat load estimation module on the same main control board or function as an independent controller node. The control processor establishes and solves the linear programming model, outputs the zoning cooling resource allocation, and converts the allocation into executable control setpoints. The actuator driver side includes at least one of the following: fan drive interface, baffle actuator drive interface, valve actuator drive interface, and pump drive interface. Typical implementations include PWM output, 0-10V analog output, or industrial bus drive. Relays or solid-state switches are also included for start / stop control and safety shut-off. The feedback interface receives fan speed signals, baffle position feedback, valve opening feedback, and flow or differential pressure sensor signals to monitor the zoning cooling execution status and support closed-loop regulation. In terms of hardware protection, resource allocation control modules are typically configured with overcurrent, overtemperature and undervoltage protection, and have a fault degradation strategy interface, such as switching to a preset safe cooling level when the control calculation is abnormal.

[0075] Those skilled in the art will understand that embodiments of the present invention can be provided as methods, systems, or computer program products. Therefore, the present invention can take the form of a completely hardware embodiment, a completely software embodiment, or an embodiment combining software and hardware aspects.

[0076] The above are merely embodiments of the present invention and are not intended to limit the invention. Various modifications and variations can be made to the present invention by those skilled in the art. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principle of the present invention should be included within the scope of the claims of the present invention.

Claims

1. A high-efficiency thermal management method for energy storage cabinets based on battery state and health management, characterized in that, The method includes: The temperature, voltage, and current data of multiple battery cells in the energy storage cabinet are acquired to obtain the first data. Based on the first data, the internal resistance parameters and capacity decay parameters of each battery cell are calculated to generate health status parameters. Based on the first collected data and the health status parameters, the heat generation power of each battery cell is estimated, a spatial distribution of heat generation power is generated, and clustering and grouping are performed based on the spatial distribution of heat generation power to obtain the partitioning results; A linear programming model is established based on the partitioning results. The partitioning cooling resource allocation is obtained under the cooling capacity constraint. The cooling capacity constraint includes at least one of the upper limit of total cooling air volume and the upper limit of total cooling medium flow. The cooling actuator is adjusted according to the partitioning cooling resource allocation to achieve partitioning cooling.

2. The method according to claim 1, characterized in that, The calculation of the internal resistance parameter of the battery cell based on the first collected data includes: Select a stable interval before the current data changes and a stable interval after the current data changes. Calculate the average voltage data and the average current data within the stable intervals, and calculate the internal resistance parameter based on the ratio of the difference between the average voltage data and the difference between the average current data.

3. The method according to claim 1, characterized in that, The capacity decay parameters of the battery cell are calculated based on the first collected data, including: The available capacity is obtained by integrating the current data over time within a preset charge / discharge range, and the capacity decay parameter is determined based on the available capacity and the rated capacity, wherein the rated capacity is the nominal capacity of a single battery cell.

4. The method according to claim 1, characterized in that, Estimating the heat generation power of the battery cell based on the first collected data and the health status parameters includes: The resistance heating power is calculated based on the current data and the internal resistance parameter, and the internal resistance parameter is temperature compensated based on the temperature data to obtain the heat generation power.

5. The method according to claim 4, characterized in that, The estimated heat generation power also includes: The heat generation power is obtained by correcting the resistance heating power based on a linear regression model, wherein the input of the linear regression model includes the internal resistance parameter and the capacity decay parameter, and the model parameters of the linear regression model are determined based on historical operating data.

6. The method according to claim 1, characterized in that, Clustering based on the spatial distribution of heat generation power includes: Obtain battery cell installation location data; when performing clustering based on the spatial distribution of heat generation power, use both the battery cell installation location data and the spatial distribution of heat generation power as clustering input, and use K-means clustering to obtain the partitioning results.

7. The method according to claim 1, characterized in that, The objective of the linear programming model is to minimize cooling resource consumption. The constraints of the linear programming model include the cooling capacity constraint and the requirement that the estimated temperature of each zone does not exceed a preset temperature threshold. The estimated temperature of each zone is determined based on the spatial distribution of heat generation power and the allocation of cooling resources to the candidate zones.

8. The method according to claim 1, characterized in that, Adjusting the cooling actuator according to the allocated cooling resources for the zone includes: Cooling actuators include at least one of a fan, a duct damper, a valve, and a pump; When the allocation of cooling resources in the partition includes cooling air volume, determine the fan speed setting value and the duct damper opening setting value and perform adjustment; When the allocation of cooling resources in the partition includes the flow rate of the cooling medium, the valve opening setting value and the pump speed setting value are determined and adjusted.

9. The method according to claim 1, characterized in that, The method further includes: After adjusting the cooling actuator, the temperature data of the multiple battery cells are acquired again to obtain the second set of data. The temperature difference between the different zones is determined based on the second collected data; If the temperature difference between the intervals is greater than a preset temperature difference threshold, the clustering and grouping are re-executed and the linear programming model is solved again.

10. A high-efficiency thermal management system for energy storage cabinets based on battery state and health management, used to implement the high-efficiency thermal management method for energy storage cabinets based on battery state and health management as described in any one of claims 1-9, characterized in that, The system includes: The data acquisition module is used to acquire temperature data, voltage data and current data of multiple battery cells in the energy storage cabinet, generate first acquisition data, and calculate the internal resistance parameter and capacity decay parameter of each battery cell based on the first acquisition data to generate health status parameters. The heat load estimation module is used to estimate the heat generation power of each battery cell based on the first collected data and the health status parameters, generate a spatial distribution of heat generation power, and perform clustering and grouping based on the spatial distribution of heat generation power to obtain partitioning results; The resource allocation control module is used to establish a linear programming model based on the partitioning results, solve for the partitioned cooling resource allocation under cooling capacity constraints, wherein the cooling capacity constraints include at least one of the upper limit of total cooling air volume and the upper limit of total cooling medium flow, and adjust the cooling actuator according to the partitioned cooling resource allocation to realize partitioned cooling.

Images