An energy equalization system, method and apparatus for energy storage battery clusters

Abstract

The application discloses a kind of energy storage battery cluster equalization energy calculation system, method and device, wherein the system comprises: multidimensional look-up table module, for according to the current acquisition battery monomer's working condition parameter, in multidimensional coupling look-up table model, the open-circuit voltage and direct current resistance corresponding to current working condition parameter are calculated by multilinear interpolation algorithm;Health state online identification module is used to calculate the maximum available capacity and health state of each battery monomer based on the open-circuit voltage difference of adjacent static section and the period coulomb integration;Adaptive fusion estimation module simultaneously runs current integration method and open-circuit voltage difference method two estimation paths, and the variance of two-path output sequence is adaptively assigned fusion weight, and the monomer equalization capacity is calculated;Decoupling aggregation module is used to output unbalanced series available energy, active equalization recoverable energy, passive equalization total dissipation energy and available energy index after equalization based on all monomer equalization capacity vectors and maximum available capacity vectors of cluster.

Description

Technical Field

[0001] This application relates to the field of battery cluster management, and more specifically, to a system, method, and apparatus for calculating the balanced energy of an energy storage battery cluster. Background Technology

[0002] A typical energy storage battery cluster consists of several battery packs connected in series. Each pack contains several battery cells connected in series, such as a lithium iron phosphate (LFP) energy storage cluster with 17 packs, each containing 14 cells, for a total of 238 strings. Due to differences in manufacturing processes, uneven operating temperatures, different charge / discharge rates, and inconsistent aging rates, the voltage, state of charge, and capacity of the individual cells within the battery cluster will gradually differ. This inconsistency will limit the usable capacity of the entire battery cluster to the worst-performing cell. Accurately calculating the balancing energy allows for a quantitative assessment of the potential energy that can be released or transferred through active or passive balancing operations, enabling the battery management system to accurately determine the actual benefits of the current balancing operation.

[0003] Currently, existing technologies for calculating equipotential energy mainly rely on simple threshold judgments based on differences in cell voltage or state of charge, and use ampere-hour integration to estimate the transferable capacity, or directly use open-circuit voltage-state of charge lookup tables to roughly estimate the equipotential capacity. However, these methods generally use simplified one-dimensional or two-dimensional models, failing to fully consider the multidimensional coupling effects of temperature, rate of charge, and degree of aging on the battery's open-circuit voltage and DC internal resistance. This leads to a significant increase in the estimation error of internal resistance voltage drop at low temperatures, high-rate charge and discharge, or in the later stages of battery aging, resulting in serious distortion in the calculation of equipotential energy. Summary of the Invention

[0004] To address the aforementioned technical problems, this application discloses a system, method, and apparatus for calculating the equilibrable energy of an energy storage battery cluster. By constructing a multi-dimensional coupled electrochemical model of electro-thermal-aging and introducing a four-way decoupling and aggregation mechanism addressing individual cell deviation, bottleneck constraints, equilibration benefits, and equilibration costs, the accurate calculation of the equilibrable energy of the battery cluster is achieved. Specifically, the technical solution of this application is as follows: In a first aspect, this application discloses an energy balancing calculation system for energy storage battery clusters, comprising: The multidimensional lookup table module is configured to pre-build a multidimensional coupled lookup table model; the multidimensional coupled lookup table model consists of a three-dimensional open-circuit voltage lookup table and a four-dimensional DC internal resistance lookup table; it is also configured to calculate the open-circuit voltage and the DC internal resistance corresponding to the current operating parameters in the multidimensional coupled lookup table model based on the currently collected operating parameters of the battery cell through a multilinear interpolation algorithm. The online health status identification module is configured to calculate the maximum usable capacity and health status of each battery cell based on the open-circuit voltage difference between adjacent resting periods and the coulomb integral during the period. The adaptive fusion estimation module is configured to run two estimation paths simultaneously, namely the current integration method and the open-circuit voltage difference method, and adaptively allocates fusion weights according to the variance of the output sequences of the two paths to calculate the average charge per unit. The decoupling aggregation module is configured to output multiple energy indicators based on the average charge vector of all cells in the cluster and the maximum available capacity vector. These indicators include unbalanced series available energy, actively balanced recoverable energy, passively balanced total dissipated energy, and available energy after balancing.

[0005] In some implementations, in the multidimensional lookup table module, the lookup dimensions of the three-dimensional open-circuit voltage lookup table are state of charge, temperature, and health status; the lookup dimensions of the four-dimensional DC internal resistance lookup table are the state of charge, the temperature, the current ratio, and the health status. The multilinear interpolation algorithm includes either a quadlinear interpolation algorithm or a trilinear interpolation algorithm.

[0006] In other embodiments, the online health status identification module specifically includes: The open-circuit voltage estimation unit takes the average voltage of multiple sampling points at the end of two adjacent stationary segments as the estimated open-circuit voltage value within the identified stationary segments. The state of charge calculation unit obtains the state of charge of the battery cells at the end of the last two adjacent resting periods by reverse lookup using the multidimensional coupled lookup table model, and calculates the state of charge difference. The equivalent charge calculation unit performs coulomb integration on the cluster current between the resting sections to obtain the equivalent cumulative charge or discharge; and calculates the maximum usable capacity of the battery cell by dividing the cumulative charge by the state of charge difference. The health status calculation unit divides the maximum available capacity by the rated capacity to obtain the health status of the battery cell.

[0007] In other embodiments, the energy storage battery cluster equalization calculation system further includes a temperature gradient compensation module. The temperature gradient compensation module is configured to construct a temperature gradient compensation formula to compensate for the minimum value of the maximum usable capacity of each battery cell, thereby obtaining the equivalent usable capacity of the battery pack after decoupling the temperature effect. The temperature gradient compensation formula is the minimum value of the maximum usable capacity of each battery cell in the battery pack plus a temperature coefficient compensation term. The temperature coefficient compensation term is proportional to the average temperature inside the battery pack minus the lowest temperature, and is calibrated by an offline constant temperature step experiment.

[0008] In other embodiments, the adaptive fusion estimation module specifically includes: The first estimation path unit is used to run the current integration method, and obtain the state of charge of the battery cell at the end of the charging and discharging process by looking up the table in reverse through the multidimensional coupled lookup table model, and calculate the average charge of the first cell. The second estimation path unit is used to run the open-circuit voltage difference method simultaneously. By performing IR drop stripping on the terminal voltage, the true open-circuit voltage estimate is obtained. Then, the state of charge of the battery cell at the end of the charging and discharging process is obtained by looking up the table, and the average charge of the second cell is calculated. The weighted fusion unit is used to evaluate the standard deviation of the output sequences of the two paths within the current time window and calculate the fusion weight of the two paths according to the inverse variance allocation method; the average charge of the first unit and the average charge of the second unit are weighted and summed according to the fusion weight to obtain the estimated value of the average charge of the unit after fusion.

[0009] Among them, the second single-unit average charge obtained by the open-circuit voltage difference method under steady-state conditions has a higher weight; Under transient operating conditions, the first single-cell average charge calculated by the current integration method has a higher weight.

[0010] Based on any of the above embodiments, optionally, the unbalanced series available energy is the minimum value of the total number of battery cells in the cluster multiplied by the rated voltage of the battery cells and then multiplied by the maximum available capacity of the battery cells in the cluster; used to characterize the most conservative dispatchable energy under the constraint of series bottleneck. The active balancing recoverable energy is the total number of battery cells in the cluster multiplied by the rated voltage of each battery cell, and then multiplied by the maximum value of the average charge of each battery cell in the cluster; it is used to characterize the maximum energy that the active balancing circuit can transport and recover. The total energy dissipated by passive balancing is the rated voltage of the battery cell multiplied by the sum of the average charge of all the battery cells in the entire cluster; it is used to characterize the total heat that needs to be dissipated through the discharge resistor under the passive balancing scheme. The available energy after balancing is the total number of battery cells in the cluster multiplied by the rated voltage of the battery cells, multiplied by the minimum of the maximum available capacity of the battery cells in the cluster, plus the product of the balancing efficiency and the difference between the average and minimum values ​​of the maximum available capacity; it is used to characterize the dispatchable energy after considering the balancing efficiency loss.

[0011] In other embodiments, the energy storage battery cluster balanced energy calculation system further includes an acquisition access layer and a data preprocessing module; The acquisition and access layer is used to obtain timing data from the battery management system, including the voltage matrix of the individual battery cells, cluster current, cluster temperature, and cluster state of charge. The data preprocessing module is used to slice the time-series data by day, and divide the data into charging segment, discharging segment and static segment based on current sign and status bit; at the same time, it filters invalid data based on duration and current amplitude.

[0012] Secondly, this application also discloses a method for calculating the balanced energy of an energy storage battery cluster, wherein the calculation method is implemented based on the calculation system described in any of the above embodiments; and includes the following steps: A multidimensional coupled lookup table model is pre-constructed; the multidimensional coupled lookup table model consists of a three-dimensional open-circuit voltage lookup table and a four-dimensional DC internal resistance lookup table. Based on the currently collected operating parameters of the battery cells, the open-circuit voltage and DC internal resistance corresponding to the current operating parameters are calculated in the multidimensional coupled lookup table model using a multilinear interpolation algorithm. The maximum usable capacity and state of health of each battery cell are calculated based on the open-circuit voltage difference between adjacent resting sections and the coulomb integral during the period. Simultaneously running two estimation paths, the current integration method and the open-circuit voltage difference method, and adaptively allocating fusion weights according to the variance of the output sequences of the two paths, the average charge per unit is calculated. Based on the average charge vector and maximum available capacity vector of all individual cells in the entire cluster, the output includes multiple energy indicators, including unbalanced series available energy, actively balanced recoverable energy, passively balanced total dissipated energy, and available energy after balancing.

[0013] Thirdly, this application also discloses an energy storage battery cluster balancing calculation device, wherein the calculation device is at least the energy storage battery cluster balancing calculation system described in any of the above embodiments.

[0014] Compared with the prior art, this application has at least one of the following beneficial effects: 1. The multidimensional lookup table in this application has scalability. In addition to the default four dimensions of state of charge, temperature, health status, and current ratio, a cycle number dimension can be introduced as needed to form a five-dimensional lookup table. This design enables more refined modeling of non-monotonic aging segments. Although the calibration workload will expand with dimension multiplication, it significantly improves the accuracy of describing complex aging processes.

[0015] 2. The health status identification window in this application is configurable. It defaults to a 15-minute threshold between adjacent static periods and allows for flexible adjustment within a range of 10 to 30 minutes based on on-site conditions. A longer threshold yields more stable identification results but results in slower updates. Users can optimize the selection based on actual data noise levels and update frequency requirements.

[0016] 3. The fusion weighting strategy of this application offers multiple options. In addition to the default field-adaptive weighting based on the standard deviation of the output sequence, dynamic weighting based on Kalman gain or fixed weighting directly specified by the user can also be selected. This flexibility allows the algorithm to adapt to the noise characteristics and operating complexity of different energy storage sites.

[0017] 4. The aggregated index output of this application can be flexibly tailored according to the strategy. In the active balancing scenario, only three indices need to be output: unbalanced series available energy, active balancing recoverable energy, and balancing available energy after balancing. In the passive balancing scenario, only two indices need to be output: unbalanced series available energy and passive balancing total dissipated energy. In the mixed scenario, all four indices are enabled, and the enabling or disabling of each index is uniformly controlled by the configuration file.

[0018] 5. The temperature compensation coefficient of this application can be configured differently according to the battery pack type, and battery packs with different air duct structures can be set with their own compensation coefficient values. This design allows the same algorithm to cover multiple rack configurations without modifying the code, greatly improving the algorithm's versatility and engineering deployment efficiency. Attached Figure Description

[0019] The preferred embodiments will now be described in a clear and easy-to-understand manner, in conjunction with the accompanying drawings, to further explain the above-mentioned characteristics, technical features, advantages, and implementation methods of this application.

[0020] Figure 1 This is a schematic diagram of the overall system architecture of an embodiment of an energy storage battery cluster energy balancing calculation system of this application; Figure 2 A lookup table structure and a schematic diagram of quadlinear interpolation for the multidimensional coupled battery model provided in the embodiments of this application; Figure 3 A schematic diagram of the online health status identification module based on the difference in OCV between adjacent static sections for single-cell level SOH online identification in the embodiments provided in this application; Figure 4A schematic diagram of the adaptive fusion estimation process of the adaptive fusion estimation module using the current integration method and the open-circuit voltage difference method in the embodiments provided in this application; Figure 5 A schematic diagram comparing the four energy outputs and physical semantics of the decoupled aggregation module in the embodiments provided in this application. Detailed Implementation

[0021] In the following description, specific details such as particular system architectures and techniques are set forth for illustrative purposes and not for limitation, in order to provide a thorough understanding of the embodiments of this application. However, those skilled in the art will understand that this application can also be implemented in other embodiments without these specific details. In other instances, detailed descriptions of well-known systems, apparatuses, circuits, and methods have been omitted so as not to obscure the description of this application with unnecessary detail.

[0022] It should be understood that, when used in this specification and the appended claims, the term "comprising" indicates the presence of the described features, integrals, steps, operations, elements and / or components, but does not exclude the presence or addition of one or more other features, integrals, steps, operations, elements, components and / or sets.

[0023] To keep the drawings concise, each figure only schematically shows the parts relevant to the invention, and these do not represent the actual structure of the product. Furthermore, to facilitate understanding, in some figures, only one of components with the same structure or function is schematically depicted, or only one is labeled. In this document, "one" not only means "only one," but can also mean "more than one."

[0024] It should also be further understood that the term “and / or” as used in this application specification and the appended claims means any combination of one or more of the associated listed items and all possible combinations, and includes such combinations.

[0025] Furthermore, in the description of this application, the terms "first," "second," etc., are used only to distinguish descriptions and should not be construed as indicating or implying relative importance.

[0026] To more clearly illustrate the technical solutions in the embodiments of this application or the prior art, the specific implementation methods of this application will be described below with reference to the accompanying drawings. Obviously, the accompanying drawings described below are only some embodiments of this application. For those skilled in the art, other drawings and other implementation methods can be obtained based on these drawings without creative effort.

[0027] A battery cluster is a basic energy storage unit composed of multiple battery modules connected in series, and it is a key hierarchical structure between the battery stack and individual battery cells. In an energy storage system, battery clusters need to be combined in series and parallel according to actual needs to achieve megawatt-level energy storage capacity. Due to manufacturing process deviations, differences in initial conditions, and differentiation in aging paths during subsequent use, even cells from the same batch will gradually develop differences in parameters such as capacity, internal resistance, and state of charge. This difference can trigger the "weakest link" effect in the battery pack. In a series-connected battery cluster, the weakest individual cell often reaches the charge / discharge cutoff voltage first, forcing the entire battery cluster to stop charging and discharging, resulting in a large amount of usable energy being wasted and unable to be output.

[0028] Therefore, those skilled in the art can accurately quantify how much additional energy constrained by the bottleneck can be released through balancing operations by precisely calculating the balancing energy. Based on the calculation results of the balancing energy, the upper-level system can make the optimal choice among several balancing schemes: choosing active balancing to maximize energy utilization, choosing passive balancing to simplify control costs, or adopting hybrid balancing to achieve a balance between the two. This mechanism upgrades system decision-making from simple voltage balancing or state-of-charge balancing to intelligent decision-making oriented towards maximizing energy.

[0029] Currently, commonly used energy storage cluster balancing calculation methods typically include the following basic steps: a) Modeling the battery open circuit voltage (OCV) and direct current internal resistance (DCIR) as a one-dimensional lookup table relying solely on the state of charge (SOC); b) Using the cluster-level SOC reported by the Battery Management System (BMS) as input, calculating the energy difference between each cell and the reference cell using the current-time integration method; c) Aggregating the cell with the largest deviation within the battery pack to the Pack level, and then aggregating the maximum value of the Pack to the cluster level; d) The remaining energy of the cluster is directly calculated by subtracting the above aggregation result from the rated total energy.

[0030] The existing technology has obvious defects in implementation, including: (1) the temperature effect is ignored. The DCIR of LFP cells can be 3 to 5 times higher at 0°C than at 25°C, and the open circuit voltage also has a drift of several millivolts at different temperatures. One-dimensional lookup table only uses SOC as index, and the internal resistance (IR) under low temperature conditions is estimated to be significantly different from the actual value, which in turn makes the average charge underestimated.

[0031] (2) The effect of charge / discharge rate is ignored. Under high-rate conditions, ohmic polarization and concentration polarization are amplified simultaneously, and the deviation of the terminal voltage from the off-circuit voltage increases significantly with the current. Continuing to use the same OCV meter to check the SOC will produce a systematic deviation.

[0032] (3) The impact of cell aging health status (SOH) is ignored. After thousands of cycles, the DCIR can increase by 30% to 100% and the capacity can decrease by 10% to 20%, and the aging rate of different cells is not consistent. One-dimensional tables cannot reflect the dispersion of cell-level aging, and the differences within the pack are treated as overall errors.

[0033] (4) The algorithm path is too simple. The current-time integration method relies on the complete charge and discharge section and cannot provide an average energy estimate under transient or incomplete operating conditions. In the LFP system with extremely low voltage sensitivity in the plateau section, the open-circuit voltage difference method is also inaccurate without internal resistance compensation.

[0034] (5) Temperature gradient is not attributed separately. Temperature gradient caused by the cooling structure inside the pack will reduce the actual capacity of the pack, but existing methods mix this type of loss with the individual aging loss and express it together, which cannot be effectively used for fault attribution.

[0035] (6) There is a conceptual flaw in the cluster-level energy aggregation formula. The existing formula is to directly subtract "the maximum single-cell energy difference multiplied by the number of cells in the whole cluster and then multiplied by the rated voltage" from the rated total energy. Its physical meaning is neither equal to the actual usable energy under the series short circuit, nor equal to the usable energy after the equalization is completed, nor equal to the energy dissipated by the passive equalization scheme. As a result, the battery management system (BMS) and energy management system (EMS) cannot make quantitative selection of active, passive, and hybrid equalization schemes based on it.

[0036] As the installed capacity of energy storage power stations increases year by year and cell parameters approach the end of their cycle life, the demand for single-cell-level online diagnostics, fine-grained energy equalization assessment, and flexible equalization scheme selection is rapidly growing. The use of multi-dimensional coupled electrochemical models and decoupled polymerization outputs has become a core evolution direction for next-generation BMS. To address the shortcomings of existing technologies, this application proposes a novel method for calculating the equalizable energy of energy storage battery clusters. By constructing a multi-dimensional coupled electrochemical model of electro-thermal-aging and introducing a four-way decoupled polymerization mechanism, it achieves accurate quantification and simultaneous reporting of single-cell deviations, bottleneck constraints, equalization benefits, and equalization costs, thus providing a direct basis for the quantitative selection of equalization schemes.

[0037] The application areas of the technical solutions in this application include, but are not limited to: containerized large-scale energy storage power stations, grid-side frequency regulation and peak shaving energy storage, integrated industrial and commercial photovoltaic-energy storage systems, and wind-solar-energy storage power stations. Specific embodiments are described below.

[0038] Reference manual attached Figure 1 As shown, this application discloses an energy storage battery cluster balanced energy calculation system, which specifically consists of an acquisition and access layer, a data preprocessing module, a multi-dimensional lookup table module, an online health status identification module, a temperature gradient compensation module, an adaptive fusion estimation module, a decoupling aggregation module, and a reporting interface.

[0039] In one embodiment of the energy storage battery cluster equalization calculation system provided in this application, an access layer is used to obtain time-series data from the battery management system, including the voltage matrix of the individual battery cells, cluster current, cluster temperature, and cluster state of charge.

[0040] Specifically, the acquisition and access layer obtains time-series data from the field BMS, including individual cell voltage matrices, cluster current, cluster temperature, cluster SOC, and multi-dimensional lookup table data generated by the offline experimental platform. The sampling period is no more than 15 seconds.

[0041] The data preprocessing module slices the time-series data by day, dividing the data into charging, discharging, and resting segments based on the current sign and status bit. It also filters invalid data based on duration and current amplitude.

[0042] Specifically, the collected time-series data is sliced ​​daily and divided into three segments—charging, discharging, and resting—based on the current symbol and status bit. Optionally, invalid data can be filtered based on duration and current amplitude, i.e., invalid segments with a duration of less than 10 minutes or a current amplitude of less than 5% of the rated capacity are filtered out.

[0043] The multidimensional lookup table module is configured to pre-build a multidimensional coupled lookup table model. This model consists of a three-dimensional open-circuit voltage lookup table and a four-dimensional DC internal resistance lookup table. It is also configured to calculate the open-circuit voltage and DC internal resistance corresponding to the currently acquired battery cell operating parameters using a multilinear interpolation algorithm within the multidimensional coupled lookup table model.

[0044] In some implementations, in the multidimensional lookup table module, the lookup dimensions of the three-dimensional open-circuit voltage lookup table are state of charge, temperature, and health status. The lookup dimensions of the four-dimensional DC internal resistance lookup table are the state of charge, temperature, current ratio, and health status.

[0045] For details, please refer to the attached instruction manual. Figure 2 As shown, Figure 2This diagram illustrates the lookup table structure and four-linear interpolation of a multidimensional coupled battery model. The module consists of a three-dimensional open-circuit voltage lookup table and a four-dimensional DC internal resistance lookup table. The three axes of the open-circuit voltage lookup table are SOC, temperature, and SOH. The four axes of the DC internal resistance lookup table are SOC, temperature, current ratio, and SOH. Figure 2 As can be seen, in the offline stage, open-circuit voltage samples were established using 21 SOC points, 7 temperature points, and 6 SOH levels. DC internal resistance samples were established using a 21×7×5×6 combination. During runtime, open-circuit voltage and DC internal resistance were obtained using quadlinear interpolation based on the input coordinates, and nearest-neighbor extrapolation was used for the boundary coordinates.

[0046] The online health status identification module is configured to calculate the maximum usable capacity and health status of each battery cell based on the open-circuit voltage difference between adjacent resting periods and the coulomb integral during the period.

[0047] For details, please refer to the attached instruction manual. Figure 3 As shown, Figure 3 This is a schematic diagram of the single-cell-level SOH online identification process based on the OCV difference between adjacent resting sections in this embodiment. In the identified resting sections, the online health status identification module takes the average voltage of the last five sampling points of two adjacent resting sections as the estimated open-circuit voltages OCV1 and OCV2 at that moment. The single-cell SOC estimates SOC1 and SOC2 are obtained through a multi-dimensional lookup table. The equivalent charge or discharge capacity ΔQ_int is obtained by performing a coulomb integral on the cluster current between the two resting sections. Based on this, the maximum usable capacity and health status of each battery cell are calculated.

[0048] The adaptive fusion estimation module is configured to run two estimation paths simultaneously: the current integration method and the open-circuit voltage difference method. It adaptively allocates fusion weights according to the variance of the output sequences of the two paths to calculate the average charge per unit.

[0049] For details, please refer to the attached instruction manual. Figure 4 As shown, Figure 4 This diagram illustrates the adaptive fusion estimation process of the current integration method and the open-circuit voltage difference method in this embodiment. The adaptive fusion estimation module runs two independent single-unit averageable energy estimation paths simultaneously. Path A is the current integration method: the final-state SOC of the single unit is obtained by looking up a multi-dimensional table, and then converted into the single-unit averageable energy. Path B is the open-circuit voltage difference method: the terminal voltage is stripped by IR drop to obtain the true open-circuit voltage estimate, and then the single-unit SOC is obtained by looking up a table, and converted into the single-unit averageable energy. The module evaluates the noise level of the two paths under the current operating conditions, characterized by the standard deviation, and adaptively allocates fusion weights accordingly, finally outputting the weighted fusion single-unit averageable energy. Under steady-state conditions, the fusion weights are biased towards path B, and under transient conditions, they are biased towards path A, thereby achieving averageable energy estimation under all operating conditions.

[0050] The decoupling aggregation module is configured to output multiple energy indicators based on the average charge vector of all cells in the cluster and the maximum available capacity vector. These indicators include unbalanced series available energy, actively balanced recoverable energy, passively balanced total dissipated energy, and available energy after balancing.

[0051] In this context, "average capacity per cell" refers to the average capacity value that a specific battery cell needs to achieve, expressed as the difference between the cell's current capacity and a baseline capacity. The "average capacity per cell vector" is a multi-dimensional array formed by arranging the average capacity values ​​of all cells within a battery cluster in numerical order. Similarly, "maximum usable capacity" refers to the total capacity a battery cell can release from a fully charged state to a fully discharged state under its current aging condition. The "maximum usable capacity vector" is a multi-dimensional array formed by arranging the maximum usable capacities of all cells within a battery cluster in numerical order.

[0052] For details, please refer to the attached instruction manual. Figure 5 As shown, Figure 5 This is a schematic diagram comparing the four energy outputs and physical semantics of the decoupled aggregation module in this embodiment. Specifically, for the average charge vector and maximum usable capacity vector of all cells in the entire cluster, the output includes at least two or more of the following four energy indicators: unbalanced series usable energy, actively balanced recoverable energy, passively balanced total dissipated energy, and usable energy after balancing.

[0053] Wherein: the unbalanced series available energy is the minimum value of the total number of battery cells in the cluster multiplied by the rated voltage of each battery cell, and then multiplied by the maximum available capacity of each battery cell in the cluster. It is used to characterize the most conservative dispatchable energy under the constraint of a series bottleneck.

[0054] In some implementations, the unbalanced series available energy is equal to the "maximum capacity of the smallest individual cell within the cluster" multiplied by the rated voltage of the individual cell and then multiplied by the total number of cells in the cluster. The unbalanced series available energy directly maps to the most conservative lower limit of dispatchable energy, avoiding scheduling risks caused by overly optimistic expectations of available energy in low-temperature or aging conditions, and is provided by the EMS as the most conservative dispatchable energy for reporting.

[0055] The recoverable energy of the active balancing circuit is the total number of battery cells in the cluster multiplied by the rated voltage of each battery cell, and then multiplied by the maximum value of the average charge capacity of each battery cell in the cluster. This represents the maximum energy that the active balancing circuit can recover.

[0056] In some implementations, the active balancing recoverable energy is equal to the maximum recoverable charge per unit multiplied by the unit's rated voltage and then by the number of units in the cluster. The active balancing recoverable energy quantifies the maximum theoretical value that can be recovered through energy transfer, reflecting the maximum recoverable energy that the active balancing circuit can transfer excess charge to the weaker unit.

[0057] The total energy dissipated by passive balancing is the rated voltage of the individual battery cell multiplied by the sum of the average charge of all the individual battery cells in the entire cluster. It is used to characterize the total heat that needs to be dissipated through the discharge resistor under the passive balancing scheme.

[0058] In some implementations, the total passive balancing energy dissipation is equal to the sum of the average charge of all cells within the cluster multiplied by the rated voltage of each cell. This reflects the total heat that needs to be dissipated through the bleed resistors under a passive balancing scheme. The total passive balancing energy dissipation reflects the total heat that must be handled using a resistor-based bleed scheme, helping thermal management assess heat dissipation costs and safety.

[0059] The available energy after equalization is calculated as the total number of battery cells in the cluster multiplied by the rated voltage of each battery cell, then multiplied by the minimum of the maximum available capacity of the battery cells in the cluster, plus the product of the equalization efficiency and the difference between the average and minimum maximum available capacity. This value characterizes the dispatchable energy after considering equalization efficiency losses.

[0060] In some implementations, the available energy after equalization is equal to [the minimum single-cell maximum capacity plus equalization efficiency multiplied by (average single-cell maximum capacity minus minimum single-cell maximum capacity)] multiplied by the single-cell rated voltage and then multiplied by the number of cells in the cluster. Its physical meaning is that the equalization circuit only applies efficiency losses to the "surplus portion that needs to be transferred via DC-DC," and the energy not transferred is directly output through the terminals without loss. This reflects the upper limit of the available energy that can actually be output to the grid after considering energy transfer losses. This formula satisfies the following boundary conditions: when the equalization efficiency is 1, it converges to the average capacity (the upper limit of perfect equalization); when the equalization efficiency is 0, it degenerates into the energy of the unbalanced series short circuit, always ensuring that it is not less than the unbalanced series available energy, which conforms to the physical constraint that "equalization can only increase available energy."

[0061] Based on the above embodiments, this application discloses another embodiment of an energy storage battery cluster energy balancing calculation system. The multidimensional lookup table module has the following multidimensional coupled lookup table models for open-circuit voltage and DC internal resistance: (1) (2) Where SOC is the state of charge of a single cell, T is the cell temperature, C_rate is the current rate, and SOH is the state of health of a single cell. The functions f_OCV and f_R are both stored in the form of multidimensional arrays from offline experimental data and evaluated online using multilinear interpolation.

[0062] The specific multilinear interpolation algorithms used include quadlinear interpolation or trilinear interpolation.

[0063] In some alternative implementations, for quadlinear interpolation, let the normalized weights of the input points along the four coordinate directions within the hypercube be α, β, γ, δ ∈ [0, 1], and the general formula for the lookup table output be written as: (3) In the formula, y_ijkl represents the lookup table value corresponding to the sixteen vertices of the hypercube containing the input point.

[0064] In some alternative implementations, when looking up the three-dimensional open-circuit voltage table, the δ term is set to a constant 1, and the summation degenerates into a trilinear interpolation.

[0065] Based on the above embodiments, this application discloses another embodiment of an energy storage battery cluster energy equalization calculation system, wherein the online health status identification module specifically includes: The open-circuit voltage estimation unit takes the average voltage of multiple sampling points at the end of two adjacent stationary segments as the estimated open-circuit voltage value within the identified stationary segments.

[0066] The state of charge (SOC) calculation unit uses the multidimensional coupled lookup table model to reverse look up the SOC of the individual cells at the end of two adjacent resting periods and calculates the SOC difference.

[0067] The equivalent charge calculation unit performs coulomb integration on the cluster current between the resting sections to obtain the equivalent cumulative charge or discharge. The maximum usable capacity of the battery cell is then calculated by dividing the cumulative charge by the state-of-charge difference.

[0068] The health status calculation unit divides the maximum available capacity by the rated capacity to obtain the health status of the battery cell.

[0069] For details, please refer to the attached instruction manual. Figure 3 As shown, the online health status identification process is as follows for each unit of the online health status identification module. The data preprocessing module executes step 1: searching for a target duration where the current during the static period is below a threshold.

[0070] The open-circuit voltage estimation unit performs step 2: sampling the average voltage at the end of the sampling period as the open-circuit voltage estimate. Specifically, the mathematical expression corresponding to the identification process is as follows. For single cell i, the average voltage of the M sampling points at the end of the resting period approximates the open-circuit voltage: (4) In the formula, k = 1, 2 correspond to the end times t_1 and t_2 of the two resting periods. V_i,t is the sampled terminal voltage value of single cell i at time t.

[0071] The charge state calculation unit executes step 3: multidimensional table lookup to obtain the charge state of the individual cell at two time points.

[0072] The equivalent charge calculation unit executes step 4: the coulomb integration of the cluster current yields the equivalent charge for the interval. Specifically, the cumulative charge of the cluster between two resting segments is obtained by the time integration of the cluster current: (5) The open-circuit voltages at two time points are used to obtain the individual cell SOC1^(i) and SOC2^(i) through a three-dimensional lookup table. Then, combined with ΔQ_int, the maximum usable capacity of cell i is obtained. (6) By normalizing the maximum usable capacity of a single cell to the cell's rated capacity C_nom, we can obtain the single-cell SOH: (7) Step 5 of the health status calculation unit: Calculate the maximum capacity of a single unit converted to a health status using a sliding window smoothing. Output the daily updated single unit SOH vector to the downstream module. Specifically, to suppress noise, the EWMA smoothing of the W-day sliding window is used as the final output: (8) This application provides another embodiment of an energy storage battery cluster balanced energy calculation system, which, based on any of the above embodiments, further includes a temperature gradient compensation module.

[0073] The temperature gradient compensation module is configured to construct a temperature gradient compensation formula to compensate for the minimum value of the maximum usable capacity of each battery cell, thereby obtaining the equivalent usable capacity of the battery pack after decoupling the temperature effect.

[0074] The temperature gradient compensation formula is the minimum of the maximum usable capacity of each battery cell in the battery pack plus a temperature coefficient compensation term.

[0075] The temperature coefficient compensation term is proportional to the average temperature inside the battery pack minus the lowest temperature, and is calibrated by an offline constant temperature step experiment.

[0076] Specifically, the temperature gradient compensation module addresses the temperature gradient existing among the individual cells within the pack due to their different locations in the cooling circuit. This module represents the equivalent usable capacity of the pack as "the maximum capacity of the smallest cell within the pack" plus a compensation term proportional to "the average temperature minus the lowest temperature," with the proportionality coefficient obtained through offline calibration. This design explicitly decouples the capacity mismatch caused by temperature gradients from the capacity mismatch caused by cell aging, facilitating fault attribution.

[0077] The explicit expression for the Pack-level equivalent usable capacity C_pack is: (9) Where i iterates through all cells in the Pack, T̄ is the average temperature in the Pack, T_min is the lowest temperature in the Pack, and κ is the temperature gradient sensitivity coefficient (unit Ah / ℃), obtained by fitting from an offline isothermal step experiment. When the temperature in the Pack is completely uniform, the compensation term returns to zero, degenerating into the short-board capacity.

[0078] The calculated equivalent available capacity is used to correct the conversion of the average charge per unit in the open-circuit voltage difference method path.

[0079] The temperature gradient compensation module in this application explicitly decouples the capacity loss caused by temperature gradient due to differences in cooling structure inside the battery pack from the capacity loss caused by cell aging. This allows maintenance personnel to clearly distinguish whether the decrease in pack capacity is due to the health degradation of the cells themselves or the structural reason of excessive temperature difference inside the pack, thus avoiding misjudgment caused by the traditional method of conflating the two types of loss.

[0080] The temperature gradient compensation module of this application significantly improves the estimation accuracy of Pack-level available capacity under low-temperature conditions in winter by introducing a compensation term that is proportional to the difference between the average temperature and the minimum temperature. It reduces the estimation deviation from more than 10% in traditional methods to less than 3%, enabling the energy management system to report dispatchable energy more accurately and avoid wasting energy storage resources due to underestimating capacity or causing dispatch risks due to overestimating capacity.

[0081] Based on the above embodiments, this application discloses another embodiment of an energy storage battery cluster energy equalization calculation system, wherein the adaptive fusion estimation module specifically includes: The first estimation path unit is used to run the current integration method, and obtain the state of charge of the battery cell at the end of the charging and discharging process by looking up the table in reverse through the multidimensional coupled lookup table model, and calculate the average charge of the first cell.

[0082] The second estimation path unit is used to run the open-circuit voltage difference method simultaneously. By performing IR drop stripping on the terminal voltage, the true open-circuit voltage estimate is obtained. Then, the state of charge of the battery cell at the end of the charging and discharging process is obtained by looking up the table, and the average charge of the second cell is calculated.

[0083] The weighted fusion unit is used to evaluate the standard deviation of the output sequences of the two paths within the current time window and calculate the fusion weight of the two paths using the inverse variance allocation method. The average charge of the first individual unit and the average charge of the second individual unit are weighted and summed according to the fusion weight to obtain the estimated value of the average charge of the fused individual unit.

[0084] In some implementations, the second cell-average charge obtained by the open-circuit voltage difference method under steady-state conditions has a higher weight. Under transient conditions, the first cell-average charge calculated by the current integration method has a higher weight.

[0085] In some alternative implementations, when estimating the open-circuit voltage using the open-circuit voltage difference method, it is necessary to remove the internal resistance voltage drop from the terminal voltage. The IR stripping model used is as follows: (10) In the formula, V_term^(i) is the terminal voltage of cell i, I_clu is the cluster current (positive for discharge), and V̂^(i) is the estimated open-circuit voltage of cell i obtained by back-reasoning. The standard deviations of the estimated sequences output by the two paths within the current time window are denoted as σ_A and σ_B, respectively, and the fusion weights are allocated according to the inverse variance: (11) The final average charge estimate for cell i is obtained by weighted fusion of the two paths: (12) Equation (11) satisfies w_A + w_B = 1. Under steady-state conditions, path B has a small variance and a large weight, while under transient conditions, path A has a small variance and a large weight. The algorithm automatically balances the two paths.

[0086] This application provides another embodiment of an energy storage battery cluster equalization energy calculation system, wherein the decoupled aggregation module simultaneously outputs four energy indicators, including unbalanced series available energy, actively balanced recoverable energy, passively balanced total dissipated energy, and available energy after equalization. (See attached specification) Figure 5 As shown.

[0087] Let the total number of cells in the cluster be N, the rated voltage of a cell be U_nom, the maximum usable capacity and average charge of cell i be C_max^(i) and Q_bal^(i) respectively, the average maximum capacity be C̄_max, and the overall DC-DC efficiency of the active balancing circuit be η. Then the analytical expressions for the four outputs are as follows: (13) (14) (15) (16) Equation (16) converges to N·U_nom·C̄_max (the upper limit of perfect equilibrium) when η → 1, and degenerates to Equation (13) when η → 0. It always satisfies E_D ≥ E_A, reflecting the physical constraint that "equilibrium can only increase available energy".

[0088] In some alternative embodiments, the minimum maximum available capacity of monomer i in formulas (13) and (16) is... After temperature gradient compensation, the minimum maximum usable capacity of each individual cell is achieved. Equivalent to the compensated equivalent available capacity .

[0089] This application, for the first time, presents four types of energy information with completely different physical meanings—series bottleneck constraints, active equilibrium benefits, passive equilibrium costs, and equilibrium upper limits considering efficiency losses—to the upper-level system simultaneously. This allows the battery management system and energy management system to no longer rely on a single, ambiguous aggregation result for decision-making. It avoids the crude management approach of traditional methods that rely on experience or simple thresholds for a one-size-fits-all solution.

[0090] Based on the same concept, this application also discloses a method for calculating the equitable energy of an energy storage battery cluster. The method is implemented based on the energy storage battery cluster equitable energy calculation system described in any of the above system embodiments. Specifically, one embodiment of the energy storage battery cluster equitable energy calculation method of this application includes the following steps: A multidimensional coupled lookup table model is pre-constructed. The multidimensional coupled lookup table model consists of a three-dimensional open-circuit voltage lookup table and a four-dimensional DC internal resistance lookup table.

[0091] Timing data is obtained from the battery management system, including the voltage matrix of the individual battery cells, cluster current, cluster temperature, and cluster state of charge.

[0092] The time-series data is sliced ​​daily, and the data is divided into charging, discharging, and resting segments based on the current sign and status bit. Invalid data is also filtered based on duration and current amplitude.

[0093] Based on the currently collected operating parameters of the battery cells, the open-circuit voltage and DC internal resistance corresponding to the current operating parameters are calculated in the multidimensional coupled lookup table model using a multilinear interpolation algorithm.

[0094] The maximum usable capacity and state of health of each battery cell are calculated based on the open-circuit voltage difference between adjacent resting sections and the coulomb integral during the period.

[0095] A temperature gradient compensation formula is constructed to compensate for the minimum value of the maximum usable capacity of each battery cell, thereby obtaining the equivalent usable capacity of the battery pack after decoupling the temperature effect.

[0096] Simultaneously running two estimation paths—the current integration method and the open-circuit voltage difference method—and adaptively allocating fusion weights according to the variance of the output sequences of the two paths, the average charge per unit is calculated.

[0097] Based on the average charge vector and maximum available capacity vector of all individual cells in the entire cluster, the output includes multiple energy indicators, including unbalanced series available energy, actively balanced recoverable energy, passively balanced total dissipated energy, and available energy after balancing.

[0098] The following is a specific implementation example: A containerized energy storage power station has a single cluster configuration of 17 Pack × 14 Cell, a single unit rated capacity of 280 Ah, a single unit rated voltage of 3.2 V, a passive balancing discharge current of 100 mA, and a discharge resistance of 32 Ω. An optional configuration is an active balancing DC-DC module, with a single-channel balancing current of 3 A and a typical DC-DC efficiency of 90%.

[0099] Offline Phase: LFP cells from the same batch were placed on a constant temperature and constant current test bench. The open-circuit voltage and DC internal resistance of 21 points were measured at 7 temperature points (-20, -10, 0, 10, 25, 40, 55 ℃), 5 rates (0.1C, 0.3C, 0.5C, 1C, 2C), and 6 SOH levels (100%, 95%, 90%, 85%, 80%, 75%). Based on this, 882 sample points of open-circuit voltage and 4410 sample points of DC internal resistance were generated and written into the Flash storage area of ​​the cluster CMU as calibration files.

[0100] Online Phase: The cluster CMU collects individual cell voltage matrix, cluster current, cluster temperature, and cluster SOC at 15-second intervals, caching them in a 24-hour rolling window. A complete algorithm process is initiated daily at 00:30: First, according to the architecture in Figure 1, five modules are sequentially scheduled: data preprocessing, multi-dimensional table lookup, online SOH identification, temperature gradient compensation, and adaptive fusion estimation. Then, the obtained individual cell averaged charge vector and maximum usable capacity vector are sent to the decoupled aggregation module, outputting four energy indicators, as well as engineering quantities such as balancing time and passive heat dissipation.

[0101] The EMS reports the most conservatively scheduled energy based on the "unbalanced series available energy". The BMS balancing controller compares the "actively balanced recoverable energy" with the "passively balanced total dissipation". When the two are close and active balancing hardware is configured on-site, active balancing is prioritized; otherwise, passive balancing is used. The "available energy after balancing" is provided for reference in the EMS's medium- and long-term scheduling plans.

[0102] On a 17-pack×14-cell×280Ah LFP energy storage cluster, the typical indices obtained by comparing the method of this application with the traditional one-dimensional method on the same time-series dataset are as follows:

[0103] As can be seen, under conditions of -10 ℃ and 85% SOH, the average energy estimation error of this application is reduced by 75% and 74% respectively compared with the existing one-dimensional method. The cluster-level energy index is expanded from a single path to four paths. At the same time, it is the first application to have the ability to identify the single-cell level SOH online in an energy storage BMS.

[0104] On the other hand, this application also discloses an energy storage battery cluster balancing calculation device, the calculation device including at least the energy storage battery cluster balancing calculation system described in any of the above embodiments.

[0105] Optionally, in some implementations, the device hardware layer comprises the following devices: a single-unit voltage sampling chip (AFE sampling front-end, accuracy ±1 mV); a cluster-level current Hall sensor (range coverage ±1.5C, accuracy ±0.5% of full scale); NTC temperature sensors (at least 2 per pack); a BMS main control MCU (clock frequency not less than 200 MHz, RAM not less than 512 KB, Flash not less than 2 MB); and optional DC-DC equalization power converter (active balancing scenario) or bleeder resistor plus MOSFET group (passive balancing scenario).

[0106] Connection method: The individual voltage sampling chip is connected to the BMU in the pack via a daisy-chain bus. The pack BMU is connected to the cluster master controller (CMU) via a CAN bus. The cluster CMU runs the calculation process of this application. The multidimensional lookup table data is stored in Flash as a calibration file, and the running parameters are stored in the erasable parameter area in JSON format, supporting remote upgrades.

[0107] Operation steps: (i) Load parameter configuration and multi-dimensional table lookup after power-on. (ii) Collect individual cell voltage, cluster current, cluster temperature, and cluster SOC at 15-second intervals and cache them in a scrolling window. (iii) Run a complete calculation method once daily at 00:30. (iv) Write the output results to the operation log and simultaneously report to EMS via CAN or Ethernet. (v) The equalization controller selects and executes an active or passive equalization scheme based on the output.

[0108] The energy storage battery cluster balanced energy calculation system, method and apparatus of this application have the same technical concept, and the technical details of the embodiments of the three are applicable to each other. In order to reduce repetition, they will not be described again here.

[0109] Those skilled in the art will clearly understand that, for the sake of convenience and brevity, the above-described division of program modules is merely an example. In practical applications, the above functions can be assigned to different program modules as needed, that is, the internal structure of the device can be divided into different program units or modules to complete all or part of the functions described above. The program modules in the embodiments can be integrated into one processing unit, or each unit can exist physically separately, or two or more units can be integrated into one processing unit. The integrated unit can be implemented in hardware or as a software program unit. Furthermore, the specific names of the program modules are only for easy differentiation and are not intended to limit the scope of protection of this application.

[0110] Although preferred embodiments of this application have been described, those skilled in the art, upon learning the basic inventive concept, can make other changes and modifications to these embodiments. Therefore, the appended claims are intended to be interpreted as including the preferred embodiments as well as all changes and modifications falling within the scope of this application.

Claims

1. A balanced energy calculation system for energy storage battery clusters, characterized in that, include: The multidimensional lookup table module is configured to pre-build a multidimensional coupled lookup table model; The multidimensional coupling lookup table model consists of a three-dimensional open-circuit voltage lookup table and a four-dimensional DC internal resistance lookup table; it is also configured to calculate the open-circuit voltage and the DC internal resistance corresponding to the current operating parameters in the multidimensional coupling lookup table model by using a multilinear interpolation algorithm based on the currently collected operating parameters of the battery cell. The online health status identification module is configured to calculate the maximum usable capacity and health status of each battery cell based on the open-circuit voltage difference between adjacent resting periods and the coulomb integral during the period. The adaptive fusion estimation module is configured to run two estimation paths simultaneously, namely the current integration method and the open-circuit voltage difference method, and adaptively allocates fusion weights according to the variance of the output sequences of the two paths to calculate the average charge per unit. The decoupling aggregation module is configured to output multiple energy indicators based on the average charge vector of all cells in the cluster and the maximum available capacity vector. These indicators include unbalanced series available energy, actively balanced recoverable energy, passively balanced total dissipated energy, and available energy after balancing.

2. The energy equalization calculation system for energy storage battery clusters as described in claim 1, characterized in that: In the multidimensional lookup table module, the lookup dimensions of the three-dimensional open-circuit voltage lookup table are state of charge, temperature, and health status; the lookup dimensions of the four-dimensional DC internal resistance lookup table are the state of charge, temperature, current ratio, and health status. The multilinear interpolation algorithm includes either a quadlinear interpolation algorithm or a trilinear interpolation algorithm.

3. The energy equalization calculation system for energy storage battery clusters as described in claim 1, characterized in that, The online health status identification module specifically includes: The open-circuit voltage estimation unit takes the average voltage of multiple sampling points at the end of two adjacent stationary segments as the estimated open-circuit voltage value within the identified stationary segments. The state of charge calculation unit obtains the state of charge of the battery cells at the end of the last two adjacent resting periods by reverse lookup using the multidimensional coupled lookup table model, and calculates the state of charge difference. The equivalent charge calculation unit performs coulomb integration on the cluster current between the resting sections to obtain the equivalent cumulative charge or discharge; and calculates the maximum usable capacity of the battery cell by dividing the cumulative charge by the state of charge difference. The health status calculation unit divides the maximum available capacity by the rated capacity to obtain the health status of the battery cell.

4. The energy balancing calculation system for an energy storage battery cluster as described in claim 1, characterized in that, It also includes a temperature gradient compensation module; The temperature gradient compensation module is configured to construct a temperature gradient compensation formula to compensate for the minimum value of the maximum usable capacity of each battery cell, thereby obtaining the equivalent usable capacity of the battery pack after decoupling the temperature effect. The temperature gradient compensation formula is the minimum value of the maximum usable capacity of each battery cell in the battery pack plus a temperature coefficient compensation term. The temperature coefficient compensation term is proportional to the average temperature inside the battery pack minus the lowest temperature, and is calibrated by an offline constant temperature step experiment.

5. The energy equalization calculation system for energy storage battery clusters as described in claim 1, characterized in that, The adaptive fusion estimation module specifically includes: The first estimation path unit is used to run the current integration method, and obtain the state of charge of the battery cell at the end of the charging and discharging process by looking up the table in reverse through the multidimensional coupled lookup table model, and calculate the average charge of the first cell. The second estimation path unit is used to run the open-circuit voltage difference method simultaneously. By performing IR drop stripping on the terminal voltage, the true open-circuit voltage estimate is obtained. Then, the state of charge of the battery cell at the end of the charging and discharging process is obtained by looking up the table, and the average charge of the second cell is calculated. The weighted fusion unit is used to evaluate the standard deviation of the output sequences of the two paths within the current time window and calculate the fusion weight of the two paths according to the inverse variance allocation method; the average charge of the first unit and the average charge of the second unit are weighted and summed according to the fusion weight to obtain the estimated value of the average charge of the unit after fusion.

6. The energy balancing calculation system for an energy storage battery cluster as described in claim 5, characterized in that: Under steady-state conditions, the second single-unit average charge obtained by the open-circuit voltage difference method has a higher weight; Under transient operating conditions, the first single-cell average charge calculated by the current integration method has a higher weight.

7. The energy balancing calculation system for an energy storage battery cluster as described in claim 1, characterized in that: The unbalanced series available energy is the minimum value of the total number of battery cells in the cluster multiplied by the rated voltage of the battery cells and then multiplied by the maximum available capacity of the battery cells in the cluster. Used to characterize the most conservative schedulable energy under series bottleneck constraints; The active equalization recoverable energy is the total number of battery cells in the cluster multiplied by the rated voltage of each battery cell, and then multiplied by the maximum value of the equalizable energy of each battery cell in the cluster. Used to characterize the maximum energy that can be transported and recovered by an active balancing circuit; The passive equalization total dissipation energy is the rated voltage of the battery cell multiplied by the sum of the equalizable energy of all the battery cells in the entire cluster. Used to characterize the total heat that needs to be dissipated through the venting resistor under a passive equilibrium scheme; The available energy after balancing is the total number of battery cells in the cluster multiplied by the rated voltage of the battery cells, multiplied by the minimum of the maximum available capacity of the battery cells in the cluster, plus the product of the balancing efficiency and the difference between the average and minimum values ​​of the maximum available capacity. Used to characterize schedulable energy after taking into account equilibrium efficiency losses.

8. The energy equalization calculation system for energy storage battery clusters as described in claim 1, characterized in that, It also includes a data acquisition and access layer and a data preprocessing module; The acquisition and access layer is used to obtain timing data from the battery management system, including the voltage matrix of the individual battery cells, cluster current, cluster temperature, and cluster state of charge. The data preprocessing module is used to slice the time-series data by day, and divide the data into charging segment, discharging segment and static segment based on current sign and status bit; at the same time, it filters invalid data based on duration and current amplitude.

9. A method for calculating the balanced energy of an energy storage battery cluster, characterized in that, The calculation method is implemented based on the computing system described in any one of claims 1-8; and includes the following steps: A multidimensional coupled lookup table model is pre-constructed; the multidimensional coupled lookup table model consists of a three-dimensional open-circuit voltage lookup table and a four-dimensional DC internal resistance lookup table. Based on the currently collected operating parameters of the battery cells, the open-circuit voltage and DC internal resistance corresponding to the current operating parameters are calculated in the multidimensional coupled lookup table model using a multilinear interpolation algorithm. The maximum usable capacity and state of health of each battery cell are calculated based on the open-circuit voltage difference between adjacent resting sections and the coulomb integral during the period. Simultaneously running two estimation paths, the current integration method and the open-circuit voltage difference method, and adaptively allocating fusion weights according to the variance of the output sequences of the two paths, the average charge per unit is calculated. Based on the average charge vector and maximum available capacity vector of all individual cells in the entire cluster, the output includes multiple energy indicators, including unbalanced series available energy, actively balanced recoverable energy, passively balanced total dissipated energy, and available energy after balancing.

10. A device for calculating the energy balance of an energy storage battery cluster, characterized in that, The computing device includes at least the energy-balanced computing system for energy storage battery clusters as described in any one of claims 1-8.

Images