Power distribution method and device for energy storage battery pack and electronic equipment

Abstract

The application discloses a power distribution method and device of an energy storage battery pack and electronic equipment. The method comprises the following steps: performing performance clustering on each single battery of the energy storage battery pack to obtain a plurality of battery clustering clusters; performing virtual parallel connection on the single batteries in each cluster in the plurality of battery clustering clusters respectively to obtain equivalent batteries corresponding to the plurality of battery clustering clusters respectively; determining single power configurations corresponding to each single battery of the energy storage battery pack based on a target optimization function and operation constraints corresponding to the plurality of equivalent batteries respectively; and performing power distribution on each single battery according to the corresponding single power configuration. The application solves the technical problem that, in the related art, when power distribution is performed on each single battery of a large-scale energy storage battery pack, the power distribution is unreasonable, and local overload or unbalanced heating exists when each single battery cooperatively operates.

Description

Technical Field

[0001] This invention relates to the field of batteries, and more specifically, to a power distribution method, apparatus, and electronic device for an energy storage battery pack. Background Technology

[0002] With the widespread application of energy storage technology in electrified transportation, smart grids, and renewable energy grid integration, large-scale battery energy storage systems (BESS) have become core equipment for driving energy transformation. To ensure the safe operation of BESS, it is necessary to rationally allocate power among the individual cells of a large-scale energy storage battery pack. However, in related technologies, there are technical problems with unreasonable power allocation among the individual cells in a large-scale energy storage battery pack, leading to localized overload or uneven heating during the coordinated operation of the individual cells.

[0003] There is currently no effective solution to the above problems. Summary of the Invention

[0004] This invention provides a power distribution method, apparatus, and electronic device for energy storage battery packs, to at least solve the technical problem in the related art where unreasonable power distribution occurs when power is distributed among individual cells in a large-scale energy storage battery pack, resulting in local overload or uneven heating of individual cells during coordinated operation.

[0005] According to one aspect of the present invention, a power allocation method for an energy storage battery pack is provided, comprising: clustering the individual cells of the energy storage battery pack according to their performance to obtain multiple battery clusters; virtually parallelizing the individual cells within each cluster to obtain equivalent cells corresponding to the multiple battery clusters, wherein the virtual parallelization is performed by modeling the individual cells within each cluster without changing the actual physical connection relationship of the individual cells; determining the individual cell power configuration corresponding to each individual cell of the energy storage battery pack based on an objective optimization function and the operating constraints corresponding to the multiple equivalent cells, wherein the objective optimization function is an optimization function aimed at minimizing the sum of the power loss degree, state of charge deviation, and temperature deviation of the multiple equivalent cells, the state of charge deviation is the deviation between the corresponding equivalent cell and the average state of charge of the multiple equivalent cells, and the temperature deviation is the deviation between the corresponding temperature and the average temperature of the multiple equivalent cells; and allocating power to each individual cell according to the corresponding individual cell power configuration.

[0006] Optionally, the power loss level of any target equivalent battery among the plurality of equivalent batteries is determined in the following manner: determining the internal resistance difference index between the reference battery and other batteries in the cluster, wherein the other batteries in the cluster are the individual batteries in each cluster other than the reference battery; determining the equivalent internal resistance with respect to the target equivalent battery based on the reference internal resistance corresponding to the reference battery, the individual internal resistance corresponding to the other batteries in the cluster, and the internal resistance difference index corresponding to the other batteries in the cluster; and determining the power loss level of the target equivalent battery based on the equivalent internal resistance and the charging and discharging current of the target equivalent battery.

[0007] Optionally, the average state of charge of the plurality of equivalent batteries is determined by: determining the average remaining charge and average equivalent capacitance of the plurality of equivalent batteries; and determining the average state of charge based on the average remaining charge and average equivalent capacitance of the plurality of equivalent batteries.

[0008] Optionally, the step of clustering the individual cells of the energy storage battery pack to obtain multiple battery clusters includes: determining the individual cell feature vector corresponding to each individual cell of the energy storage battery pack, wherein the corresponding individual cell feature vector is used to characterize the electrothermal coupling characteristics of the corresponding individual cell from multiple predetermined dimensions, including a state of charge dimension, a temperature dimension, and a resistance dimension; determining the coupling coefficient corresponding to each of the multiple predetermined dimensions, wherein the corresponding coupling coefficient is used to characterize the importance of the corresponding predetermined dimension in characterizing the electrothermal coupling characteristics of the corresponding individual cell; and performing performance clustering on the individual cells of the energy storage battery pack based on the individual cell feature vectors corresponding to each individual cell of the energy storage battery pack and the coupling coefficients corresponding to the multiple predetermined dimensions to obtain multiple battery clusters.

[0009] Optionally, the step of clustering the individual cells of the energy storage battery pack based on their respective individual feature vectors and coupling coefficients corresponding to the multiple predetermined dimensions to obtain multiple battery clusters includes: determining multiple central cells for a given iteration according to the execution order of multiple iterations in the clustering process, wherein each individual cell includes the multiple central cells; determining the difference index between the multiple central cells and each individual cell based on their respective individual feature vectors and coupling coefficients corresponding to the multiple predetermined dimensions; determining the cluster for the corresponding iteration based on the difference index between the multiple central cells and each individual cell, wherein the cluster for the corresponding iteration includes individual cells whose difference index with the corresponding central cell is less than or equal to a difference threshold, until a target condition is met to obtain multiple battery clusters, wherein the target condition includes the completion of the multiple iterations in the clustering process.

[0010] Optionally, determining the individual power configuration of each cell in the energy storage battery pack based on the objective optimization function and the operating constraints corresponding to the multiple equivalent cells includes: determining the equivalent open-circuit voltage corresponding to the multiple equivalent cells; and determining the individual power configuration of each cell in the energy storage battery pack based on the objective optimization function, the operating constraints corresponding to the multiple equivalent cells, and the equivalent open-circuit voltage.

[0011] Optionally, determining the individual power configuration of each cell in the energy storage battery pack based on the objective optimization function and the operating constraints and equivalent open-circuit voltages corresponding to the multiple equivalent cells includes: determining the equivalent power configuration corresponding to each of the multiple equivalent cells based on the objective optimization function and the operating constraints corresponding to the multiple equivalent cells; and determining the individual power configuration corresponding to each cell in the energy storage battery pack based on the equivalent power configurations corresponding to the multiple equivalent cells and the individual power allocation strategy, wherein the individual power allocation strategy includes any of the following: a power quota-based allocation strategy, wherein the power quota includes equal power quotas; a power loss-based allocation strategy, wherein the power loss-based allocation strategy is an allocation strategy aimed at minimizing the power loss of the energy storage battery pack; or an operating characteristic-based allocation strategy, wherein the operating characteristic-based allocation strategy is an allocation strategy aimed at minimizing the state deviation between the operating state and a predetermined state of the energy storage battery pack.

[0012] Optionally, the operating constraints corresponding to the plurality of equivalent cells include: current constraints, temperature constraints, state of charge constraints, operating power constraints, and power quota constraints.

[0013] According to one aspect of the present invention, a power allocation device for an energy storage battery pack is provided, comprising: a first determining module, configured to perform performance clustering on each individual cell of the energy storage battery pack to obtain multiple battery clusters; a second determining module, configured to virtually connect the individual cells within each cluster of the multiple battery clusters in parallel to obtain equivalent cells corresponding to each of the multiple battery clusters, wherein the virtual parallel connection is performed by modeling the individual cells within each cluster in parallel without changing the actual physical connection relationship of the individual cells; a third determining module, configured to determine the individual cell power configuration corresponding to each individual cell of the energy storage battery pack based on an objective optimization function and the operating constraints corresponding to the multiple equivalent cells, wherein the objective optimization function is an optimization function aimed at minimizing the sum of the power loss degree, state of charge deviation, and temperature deviation of the multiple equivalent cells, the state of charge deviation is the deviation between the corresponding equivalent cell and the average state of charge of the multiple equivalent cells, and the temperature deviation is the deviation between the corresponding temperature and the average temperature of the multiple equivalent cells; and a fourth determining module, configured to allocate power to each individual cell according to the corresponding individual cell power configuration.

[0014] According to one aspect of the present invention, an electronic device is provided, comprising: a processor; and a memory for storing processor-executable instructions; wherein the processor is configured to execute the instructions to implement the power distribution method of the energy storage battery pack described in any of the preceding claims.

[0015] In this embodiment of the invention, by clustering the individual cells of the energy storage battery pack to obtain multiple battery clusters, individual cells with similar characteristics can be classified, reducing the complexity of subsequent power allocation. By performing virtual parallel connection on the individual cells within each cluster to obtain corresponding equivalent cells, the multi-cell system can be equivalent to a small number of equivalent units for unified scheduling. This ensures the accuracy and reliability of equivalent modeling and state estimation while simplifying the calculation and control process. The objective optimization function aims to minimize the power loss, state of charge deviation, and temperature deviation of each equivalent cell, simultaneously taking into account energy consumption suppression, charge balance, and thermal consistency management. Thus, by combining the objective optimization function and the operating constraints corresponding to multiple equivalent cells, the power configuration of individual cells is determined, achieving precise and reasonable power allocation of the energy storage battery pack. While meeting safe operating conditions, this reduces system losses, improves charge balance and temperature consistency, and avoids local overload or uneven heating caused by differences between individual cells. This solves the technical problem in related technologies where unreasonable power allocation occurs when allocating power to individual cells in large-scale energy storage battery packs, leading to local overload or uneven heating during the coordinated operation of individual cells. Attached Figure Description

[0016] The accompanying drawings, which are included to provide a further understanding of the invention and form part of this application, illustrate exemplary embodiments of the invention and, together with their description, serve to explain the invention and do not constitute an undue limitation thereof. In the drawings:

[0017] Figure 1 This is a flowchart of a power distribution method for an energy storage battery pack according to an embodiment of the present invention;

[0018] Figure 2 This is a schematic diagram of the equivalent model construction in an optional embodiment of the present invention;

[0019] Figure 3 This is a schematic diagram of the internal resistance model in an optional embodiment of the present invention;

[0020] Figure 4 This is a flowchart of the optimal power allocation process for large-scale energy storage battery packs based on clustering, which is an optional embodiment of the present invention.

[0021] Figure 5 This is a schematic diagram of the circuit structure of a large-scale battery energy storage system in an optional embodiment of the present invention;

[0022] Figure 6 This is a schematic diagram of the power allocation process for a large-scale energy storage battery pack based on clustering in an optional embodiment of the present invention.

[0023] Figure 7 This is a schematic diagram of battery clustering in a large-scale battery energy storage system according to an optional embodiment of the present invention;

[0024] Figure 8 This is a schematic diagram of the principle architecture of a large-scale battery energy storage system in an optional embodiment of the present invention;

[0025] Figure 9 This is a structural block diagram of a power distribution device for an energy storage battery pack according to an embodiment of the present invention. Detailed Implementation

[0026] To enable those skilled in the art to better understand the present invention, the technical solutions of the present invention will be clearly and completely described below with reference to the accompanying drawings of the embodiments of the present invention. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort should fall within the scope of protection of the present invention.

[0027] It should be noted that the terms "first," "second," etc., in the specification, claims, and accompanying drawings of this invention are used to distinguish similar objects and are not necessarily used to describe a specific order or sequence. It should be understood that such data can be interchanged where appropriate so that the embodiments of the invention described herein can be implemented in orders other than those illustrated or described herein. Furthermore, the terms "comprising" and "having," and any variations thereof, are intended to cover a non-exclusive inclusion; for example, a process, method, system, product, or apparatus that comprises a series of steps or units is not necessarily limited to those steps or units explicitly listed, but may include other steps or units not explicitly listed or inherent to such processes, methods, products, or apparatus.

[0028] Example 1

[0029] According to an embodiment of the present invention, an embodiment of a power distribution method for an energy storage battery pack is provided. It should be noted that the steps shown in the flowchart in the accompanying drawings can be executed in a computer system such as a set of computer-executable instructions. Furthermore, although a logical order is shown in the flowchart, in some cases, the steps shown or described may be executed in a different order than that shown here.

[0030] Figure 1 This is a flowchart of a power distribution method for an energy storage battery pack according to an embodiment of the present invention, such as... Figure 1 As shown, the method includes the following steps:

[0031] S102, cluster the performance of each individual cell in the energy storage battery pack to obtain multiple battery clusters;

[0032] Optionally, the performance of each individual cell in the energy storage battery pack is clustered to obtain multiple battery clusters, including:

[0033] Each individual cell in the energy storage battery pack is identified by determining its corresponding feature vector. This feature vector characterizes the electrothermal coupling characteristics of the individual cell across multiple predetermined dimensions, including state of charge, temperature, and resistance. Coupling coefficients corresponding to these predetermined dimensions are then determined, representing the importance of each dimension in characterizing the electrothermal coupling characteristics of the individual cell. Based on the individual feature vectors and coupling coefficients for each predetermined dimension, the individual cells in the energy storage battery pack are clustered according to their performance, resulting in multiple battery clusters.

[0034] This involves energy storage battery packs, which are battery packs comprising multiple individual cells connected in series and parallel for energy storage and release. Specifically, this energy storage battery pack can be a large-scale energy storage battery pack, comprising a large number of individual cells, and is a large battery system capable of storing and stably outputting high-power energy.

[0035] This involves individual cells, which are the electrochemical units that make up energy storage battery packs. Their state of charge, temperature, internal resistance, and other characteristics affect the overall operational safety and balance of the energy storage battery pack.

[0036] This involves performance clustering, which is a process of grouping cells with similar characteristics into the same group based on the performance differences of individual cells (such as electrothermal characteristics), in order to reduce the optimization complexity of large-scale battery systems.

[0037] This involves battery clusters, which are groups of batteries with highly similar internal individual cell characteristics (such as electrothermal characteristics) obtained after performance clustering. These clusters serve as the basic units for subsequent equivalent modeling and hierarchical power allocation.

[0038] This involves a single-cell feature vector, which is a multi-dimensional data set used to characterize the operating state of a single cell. Specifically, it can reflect the electrothermal coupling characteristics of a single cell from multiple preset dimensions.

[0039] This involves predetermined dimensions, which are observational dimensions used to describe battery characteristics. Specifically, these include the state of charge dimension, temperature dimension, and resistance dimension, which are the core basis for quantifying the electrothermal coupling characteristics of the battery.

[0040] This includes the state of charge dimension, which is used to characterize the ratio of the current remaining charge of a single battery cell to its rated capacity, reflecting the current charge level and charge / discharge capability of the battery.

[0041] This includes a temperature dimension, which is used to characterize the current operating temperature of a single battery cell, reflect the battery's thermal state and safety risks, and indicate the thermal equilibrium and thermal protection characteristics of the single battery cell.

[0042] This involves the resistance dimension, which is used to characterize the internal ohmic resistance of a single cell and directly determines the magnitude of battery power loss and heat generation.

[0043] This involves the electrothermal coupling characteristic, which is the law of mutual influence and coupling between the electrical and thermal characteristics of a battery. That is, current and internal resistance generate heat loss, and temperature changes in turn change the internal resistance and electrical performance.

[0044] This involves a coupling coefficient, which is a weighting coefficient assigned to a predetermined dimension. It is used to characterize the importance and / or influence of the state of charge, temperature, and internal resistance in describing the electrothermal coupling characteristics of the battery, making the clustering more consistent with the actual operating mechanism of the battery.

[0045] For example, based on the state of charge of each individual cell (referred to as "cell") in an energy storage battery pack (specifically a large-scale energy storage battery pack). ),temperature( ), internal resistance ( Constructing single-unit feature vectors .

[0046] Among them, the state of charge ( ) corresponds to the dimension of charge state, temperature ( Corresponding to the temperature dimension, internal resistance ( The corresponding resistance dimension is shown. State of charge (SOC) represents the ratio of the battery's remaining charge to its rated capacity at a given moment, characterizing the battery's current remaining charge level.

[0047] By determining the individual feature vectors of each cell in an energy storage battery pack (such as a large-scale energy storage battery pack), the electrothermal coupling state of each cell can be fully quantified from three core dimensions: state of charge, temperature, and internal resistance, accurately reflecting the performance differences between cells. By determining the corresponding coupling coefficients, the importance of different characteristic dimensions in clustering can be distinguished, making the grouping results more in line with the actual needs of battery loss, balancing, and safety control. Based on the individual feature vectors and coupling coefficients, performance clustering of large-scale energy storage battery packs can divide massive, highly heterogeneous individual cells into several battery clusters with consistent internal characteristics, achieving system dimensionality reduction and simplification. This provides a clear and consistent grouping basis for subsequent virtual parallel connection, equivalent modeling, and optimal power allocation, significantly improving the optimization efficiency and control accuracy of large-scale systems.

[0048] Optionally, based on the individual feature vectors corresponding to each cell in the energy storage battery pack and the coupling coefficients corresponding to multiple predetermined dimensions, performance clustering is performed on each individual cell in the energy storage battery pack to obtain multiple battery clusters. This includes: determining multiple central cells for each iteration in the clustering process according to the execution order of multiple iterations, wherein each individual cell includes multiple central cells; determining the difference index between each central cell and each individual cell based on the individual feature vectors corresponding to each individual cell and the coupling coefficients corresponding to multiple predetermined dimensions; determining the cluster for each iteration based on the difference index between each central cell and each individual cell, wherein the cluster for each iteration includes individual cells whose difference index with the corresponding central cell is less than or equal to a difference threshold, until a target condition is met to obtain multiple battery clusters. The target condition includes the completion of multiple iterations in the clustering process, as shown in the formula:

[0049]

[0050] in, Indicated by attribution variable and cluster centroid (That is, the electrothermal coupling centroid of the cluster) is the optimization variable, and the objective function value is minimized, which is to minimize the weighted sum of the distances between individual cells within the cluster and the cluster centroid (i.e., the central cell). ; This represents the total number of individual cells in the energy storage battery pack. The total number of clusters (battery clusters) obtained after clustering. ; Let be the attribute variable for the i-th individual cell to the j-th cluster (i.e., the cluster corresponding to the j-th central cell). This is a binary variable, representing only whether the cell belongs to or not. Specifically, when the difference index is greater than the difference threshold, ... This indicates that the element does not belong to the j-th cluster. This occurs when the difference index is less than or equal to the difference threshold. This indicates that the element belongs to the j-th cluster. The coupling coefficient corresponding to the charged state dimension; The coupling coefficient corresponding to the temperature dimension; This represents the coupling coefficient corresponding to the resistance dimension; The state of charge (SOC) of the i-th cell represents the remaining charge level of the cell. The state of charge of the central cell (i.e., the centroid) of the j-th cluster (i.e., the battery cluster); The temperature of the i-th individual cell (specifically, the operating temperature); The temperature (specifically, the operating temperature) of the central cell (i.e., the centroid) of the j-th cluster (i.e., the battery cluster). Let be the internal resistance of the i-th cell, which characterizes the electrical characteristics and loss level of the cell. Let be the internal resistance of the central cell (i.e., the centroid) of the j-th cluster (i.e., the battery cluster); Indicates the difference in state of charge; Indicates the temperature difference index; This represents the power consumption difference index.

[0051] in:

[0052] The difference indices between multiple central cells and individual cells include the state of charge difference index, temperature difference index, and power consumption difference index.

[0053] This represents the coupling coefficient, specifically the weight of electrothermal characteristics, prioritizing the consistency between internal resistance and temperature, and adapting to aggregate modeling.

[0054] , This indicates that cell i belongs to the j-th cluster. This indicates that cell i does not belong to the j-th cluster.

[0055] The formula for determining the multiple central cells at the corresponding iteration number is as follows:

[0056]

[0057] in, For the first The central cell (i.e., the centroid) at the next iteration. For the current iteration, , corresponding to the first step in the clustering process The next iteration; For the first In the next iteration, the variable indicating the affiliation of the i-th individual cell to the j-th cluster; This is a single-unit feature vector.

[0058] That is, in the iterative process (specifically the alternating iterative process), the individual cells are first assigned to the nearest centroid (i.e., the central cell), and then the centroid is updated according to the electrothermal characteristics until the assignment is stable (i.e., the target condition is met).

[0059] Before clustering the individual cells of the energy storage battery pack according to their performance to obtain multiple battery clusters, the process includes using gap statistics to determine the target number of clusters k (i.e., the optimal number of clusters), where k... The optimal number of clusters (n) balances the electrothermal consistency within the cluster with computational efficiency, providing the best grouping basis for subsequent aggregation modeling. In other words, the optimal number of clusters, determined by gap statistics, is the number of clusters best suited for the current large-scale energy storage battery packs. It represents the number of clusters that achieves the best balance between electrothermal characteristic consistency within the cluster and system computational efficiency. Gap statistics is a mathematical method used to automatically determine the optimal number of clusters. By comparing the differences between the actual data distribution and the reference distribution, it objectively selects the number of groups with the best clustering effect. Electrothermal consistency within a cluster refers to the similarity in electrothermal characteristics such as SOC, temperature, and internal resistance among individual cells within the same cluster after clustering. Higher consistency leads to higher accuracy in subsequent equivalent modeling.

[0060] The above steps can be implemented using the k-means clustering algorithm. However, the standard k-means clustering algorithm only groups by distance and does not consider the battery's electrothermal coupling characteristics and the requirements for cluster modeling. When using the k-means clustering algorithm, it is necessary to add electrothermal characteristic weights (i.e., coupling coefficients) and cluster consistency constraints to ensure that the electrothermal characteristics of individual cells within a cluster are highly consistent, supporting subsequent high-precision cluster modeling. The k-means algorithm is a classic unsupervised machine learning algorithm used to divide data objects into k clusters based on feature similarity.

[0061] This involves the number of iterations, which is the sequential iteration number executed by the clustering algorithm during the grouping optimization process. This number is used to progressively update the central cells and cluster partitions, improving the accuracy of the clustering results. For example, multiple iterations, including the first iteration, the second iteration, etc., each correspond to one iteration.

[0062] This involves a central cell, which is the single cell (i.e., the centroid) selected as the benchmark for the cluster in the corresponding iteration. It serves as a reference for other cells within the cluster, characterizing the typical electrothermal coupling characteristics of the current cluster.

[0063] This includes a difference index, which is an indicator calculated based on the individual cell feature vector and coupling coefficient. It is used to quantify the degree of difference in electrothermal characteristics between the central cell and each individual cell and is the core basis for determining the cell's affiliation.

[0064] This involves a difference threshold, which is a preset threshold for the degree of difference. It is used to determine whether a single cell belongs to the cluster represented by the corresponding central cell, so as to ensure the consistency of the electrothermal characteristics of the cells within the cluster.

[0065] This involves target conditions, which are the criteria for terminating the clustering process. These include the completion of the clustering iterations and the fact that the change index of the resulting multiple battery clusters is less than the change threshold (i.e., each battery cluster is in a stable state). These conditions are used to ensure the integrity of the clustering process and the stability of the grouping results.

[0066] By introducing coupling coefficients corresponding to each predetermined dimension during the clustering iteration process to calculate the difference index between the central battery and individual cells, clustering can be based on the electrothermal coupling characteristics of the attached battery, rather than simply on feature distance. This ensures that individual cells within the same cluster maintain a high degree of consistency in key dimensions such as state of charge, temperature, and internal resistance, avoiding the problem of excessive differences in characteristics within the cluster due to unweighted grouping. This improves the adaptability of clustering results to subsequent aggregation modeling and enables accurate and stable performance grouping of large-scale energy storage battery packs.

[0067] S104, Virtually parallel connect the individual cells in each of the multiple battery clusters to obtain the equivalent cells corresponding to the multiple battery clusters respectively. The virtual parallel connects the individual cells in each cluster in a virtual parallel connection by modeling without changing the actual physical connection relationship of the individual cells.

[0068] For example, Figure 2 This is a schematic diagram of the equivalent model construction in an optional embodiment of the present invention, such as... Figure 2 As shown, based on the clustering results, individuals in the same cluster are virtually connected in parallel.

[0069] This involves battery clusters, which are groups of batteries with highly similar internal individual electrothermal characteristics obtained after performance clustering. These clusters are the basic units for virtual parallel connection and equivalent modeling.

[0070] This involves virtual parallel connection, which is a processing method that treats individual cells within the same cluster as a parallel structure through modeling (specifically mathematical modeling) without changing the actual physical connection relationship of the individual cells.

[0071] This involves an equivalent cell, which is an equivalent model unit obtained through virtual parallel modeling that can comprehensively represent the electrical and thermal characteristics of all individual cells within the corresponding cluster.

[0072] This involves actual physical connections, which are the actual hardware wiring methods between individual batteries. Virtual parallel connections do not modify these connections in any way.

[0073] This involves modeling, which is the process of establishing a mathematical model based on the electrical and thermal characteristics of the battery, used to aggregate multiple individual cells into a single equivalent unit to simplify system analysis and optimization calculations.

[0074] By performing virtual parallel processing on individual cells within each battery cluster, the characteristics of multiple cells in the same cluster can be aggregated into a single equivalent representation without changing the hardware connection. The equivalent cell can uniformly reflect the overall electrical and thermal behavior of all cells in the corresponding cluster, which greatly reduces the model dimension and number of optimization variables of large-scale energy storage battery packs, ensuring that subsequent power allocation calculations are efficient and feasible. At the same time, it preserves the complete expression of the electrothermal coupling characteristics of the cluster, providing an accurate and concise equivalent model basis for inter-cluster convex optimization allocation.

[0075] S106. Based on the objective optimization function and the operating constraints corresponding to multiple equivalent batteries, determine the power configuration of each individual cell in the energy storage battery pack.

[0076] The objective optimization function is an optimization function that aims to minimize the sum of the power loss of multiple equivalent cells, the state of charge deviation, and the temperature deviation. The state of charge deviation is the deviation between the corresponding equivalent cell and the average state of charge of multiple equivalent cells, and the temperature deviation is the deviation between the corresponding temperature and the average temperature of multiple equivalent cells.

[0077] Before determining the power configuration of each cell in the energy storage battery pack based on the objective optimization function and the operating constraints corresponding to multiple equivalent cells, the process includes: constructing equivalent electrothermal aggregation models corresponding to multiple battery clusters. The equivalent electrothermal aggregation models include electrical aggregation models and thermal aggregation models. The electrical aggregation model is used to characterize the electrical characteristics of the corresponding equivalent cells, including the equivalent cell capacity, equivalent current, equivalent voltage (which can specifically be the equivalent open-circuit voltage), equivalent state of charge, equivalent internal resistance, and equivalent power consumption. The thermal aggregation model is used to characterize the thermal characteristics of the corresponding equivalent cells, including the equivalent temperature and thermal dynamic characteristics.

[0078] This involves an objective optimization function, which is an optimization calculation function constructed with the goal of minimizing the sum of the power loss, state-of-charge deviation and temperature deviation of the equivalent battery, and is used to achieve the global optimal solution for cluster-level power allocation.

[0079] This includes the degree of power loss, which is the amount of energy loss caused by the internal resistance of the equivalent battery during charging and discharging, and directly reflects the operating efficiency and heat generation of the energy storage battery pack.

[0080] This involves the state of charge deviation, which is the difference between the state of charge of a single equivalent cell and the average state of charge of all equivalent cells, and is used to quantify the degree of charge balance of the battery pack.

[0081] This involves temperature deviation, which is the difference between the temperature of a single equivalent cell and the average temperature of all equivalent cells, used to characterize the uniformity of thermal distribution in the battery pack.

[0082] This involves operational constraints, which are the limited conditions that the equivalent battery must meet within the safe and stable operating range, including boundary limits such as current, temperature, state of charge, and power.

[0083] This involves the configuration of individual cell power, which is the charging and discharging power command allocated to each individual battery cell after optimized calculation. It is the direct basis for achieving precise control at the individual cell level.

[0084] This involves the equivalent electrothermal aggregation model, which is a unified mathematical model based on equivalent cells. It includes an electrical aggregation model and a thermal aggregation model, and is used to characterize the electrical and thermal dynamic characteristics of the clusters as a whole.

[0085] This involves an electrical aggregation model, which is a mathematical model used to describe the electrical characteristics of an equivalent battery, such as capacity, current, voltage, state of charge, internal resistance, and power consumption, supporting accurate calculation of electrical behavior.

[0086] This involves a thermal polymerization model, which is a mathematical model used to characterize the equivalent temperature and thermal dynamics changes, and can reflect the cluster heating, heat dissipation and temperature evolution process.

[0087] By constructing an equivalent electrothermal aggregation model, the multi-unit characteristics of a single cluster can be integrated into unified equivalent unit characteristics, fully preserving the electrical and thermal coupling relationship. By constructing an objective optimization function with the goal of minimizing power loss, state of charge deviation, and temperature deviation, the system loss reduction, power balance, and thermal consistency improvement can be achieved simultaneously. By solving the objective optimization function in combination with the corresponding operating constraints of the equivalent battery, the optimal power configuration of each individual battery can be obtained under the premise of meeting safe operating conditions, ensuring that the large-scale energy storage battery pack operates efficiently, in a balanced and stable manner.

[0088] For example, the equivalent electrothermal aggregation model is a cluster-level electrothermal aggregation model, that is, a cluster-level electrothermal coupling mathematical model (based on the clustering results, the individual units in the same cluster are virtually connected in parallel to construct a cluster-level electrothermal coupling mathematical model). By using a single cluster model to equivalently represent the electrical characteristics and thermal dynamics of all individual units in the cluster, the model dimensionality is greatly reduced, while retaining the electrothermal coupling characteristics, achieving a seamless connection between "clustering and grouping → aggregation modeling → optimization allocation". The electrothermal consistency of the cluster directly determines the accuracy of the aggregation model, and thus determines the optimization effect of subsequent power allocation.

[0089] Specifically, the clustering output cluster partitioning results (i.e., multiple battery clusters) and the data of the cluster centroid (i.e., the central battery) (including average SOC, average temperature, equivalent internal resistance, etc.) can be used as input parameters for the cluster-level electrothermal aggregation model.

[0090] For the electrical aggregation model:

[0091] The equivalent battery capacity (i.e., the total capacity of the cluster) is expressed as:

[0092]

[0093] in, Let be the equivalent cell capacity of the j-th equivalent cell; Let i be the single cell capacity of the i-th cell within the j-th cell cluster; Let be the number of individual cells within the j-th battery cluster.

[0094] The equivalent current (i.e., the total current of the cluster) is expressed as:

[0095]

[0096] in, Let be the equivalent current of the j-th equivalent cell; Let be the single-cell current of the i-th cell within the j-th cell cluster.

[0097] Equivalent state of charge The dynamic equation, also known as the cluster (i.e., the SOC dynamic equation), is expressed as:

[0098]

[0099] Specifically, .

[0100] in, Let be the rate of change of the equivalent state of charge of the j-th equivalent cell, used to quantify the overall charging and discharging speed of the cluster.

[0101] By performing equivalent aggregation on the open-circuit voltage, internal resistance, and internal resistance of the converter (DC / DC) of each battery cluster, the overall efficiency of the battery cluster is achieved. The individual electrical models are reduced to a single cluster model, preserving the coupling relationship between internal resistance, open-circuit voltage, and power loss, with an accuracy error of less than 3% and a reduction in computational load. This provides precise electrical input for inter-cluster optimization.

[0102] For the temperature-dependent polymerization model:

[0103] The equivalent temperature (cluster average temperature) is expressed as:

[0104]

[0105] in, Let be the equivalent temperature of the j-th equivalent cell; Let be the temperature of the i-th individual cell within the j-th cell cluster.

[0106] Thermal dynamic characteristics, cluster thermal dynamic model equations, the formula is:

[0107]

[0108] in, Let be the equivalent total mass of the j-th equivalent cell, and let be the total thermal inertia of the j-th cell cluster. This refers to the specific heat capacity of a single cell. Let be the equivalent internal resistance of the j-th equivalent cell; Let be the rate of temperature change of the j-th equivalent cell at time t; The ambient temperature; Let be the equivalent convective thermal resistance of the j-th equivalent cell (i.e., the cluster convective thermal resistance).

[0109] The formula for determining it is:

[0110]

[0111] in, Let be the mass of the i-th individual cell within the j-th cell cluster.

[0112] The formula for determining it is:

[0113]

[0114] in, The convective heat transfer coefficient between a single battery cell and the environment; Let be the surface area of ​​the i-th individual cell within the j-th cell cluster.

[0115] By thermally aggregating each battery cluster—that is, coupling the heat generation and heat dissipation of individual cells within a cluster to the overall thermal dynamics of the cluster—we can accurately correlate electrical losses with temperature rise. This supports temperature balance and safety constraints, preventing individual cell overheating, while significantly reducing the computational dimensionality of the thermal model. Electrical power loss can be directly used as a heat source term in the thermal model, achieving a closed-loop coupling of electrical characteristics → heat loss → temperature dynamics, laying the foundation for subsequent coupling optimization.

[0116] Optionally, the power loss level of any target equivalent battery among multiple equivalent batteries is determined using the following method: Determine the internal resistance difference index between the reference battery and other batteries within the cluster, where other batteries within the cluster are the individual cells in each cluster excluding the reference battery; Based on the reference internal resistance corresponding to the reference battery, the individual internal resistances corresponding to the other batteries within the cluster, and the internal resistance difference index corresponding to the other batteries within the cluster, determine the equivalent internal resistance to the target equivalent battery; Based on the equivalent internal resistance and the charging / discharging current of the target equivalent battery, determine the power loss level of the target equivalent battery, used to measure the power loss caused by internal resistance consumption during operation, and to assess system operating efficiency and heat generation level.

[0117] This involves a target equivalent cell, which is any equivalent cell selected from multiple equivalent cells for calculating the degree of power loss. It is the object of the calculation of equivalent internal resistance and power loss.

[0118] This involves a reference cell, which is a single cell selected from the corresponding cluster and used as a benchmark for calculating internal resistance, to provide a unified reference standard for solving the equivalent internal resistance.

[0119] This involves other cells within the cluster, which are all individual cells in the corresponding cluster except for the reference cell, and are the constituent units participating in the equivalent internal resistance calculation.

[0120] This includes the internal resistance difference index, which is an indicator used to quantify the degree of internal resistance difference between the reference cell and other cells in the cluster, and can reflect the dispersion of the internal resistance of individual cells within the cluster.

[0121] This involves a reference internal resistance, which is the internal resistance corresponding to a reference battery and serves as the benchmark parameter for calculating the equivalent internal resistance.

[0122] This involves the internal resistance of individual cells, which is the internal resistance of each cell within the cluster and directly affects battery loss and heat generation.

[0123] This involves the equivalent internal resistance, which is the internal resistance of the target equivalent battery. It is a comprehensive parameter that can characterize the internal resistance characteristics of the target equivalent battery as a whole, calculated based on the reference internal resistance, the internal resistance of individual cells, and the internal resistance difference index.

[0124] This involves the charging and discharging current, which is the operating current flowing through the target equivalent battery during operation. It is a key electrical parameter for calculating power loss (including charging current and discharging current, corresponding to charging and discharging conditions, respectively).

[0125] By calculating the internal resistance difference index between the reference battery and other batteries in the cluster, the degree of difference in the internal resistance of individual cells within the cluster can be accurately quantified, providing a reliable basis for difference correction in the calculation of equivalent internal resistance. By combining the reference internal resistance, individual cell internal resistance, and internal resistance difference index to solve for the equivalent internal resistance, the internal resistance characteristics of multiple cells within the cluster can be reasonably aggregated into a unified equivalent parameter, ensuring that the equivalent model truly reflects the overall electrical characteristics of the cluster. Thus, based on the equivalent internal resistance and charging / discharging current, the degree of power loss can be determined, accurately quantifying the internal resistance heating and energy loss of the equivalent battery, providing a real basis for loss calculation for optimized allocation, and improving the rationality of power allocation and system operating efficiency.

[0126] Figure 3 This is a schematic diagram of the internal resistance model in an optional embodiment of the present invention, such as... Figure 3 As shown, the electrical dynamic characteristics of any target equivalent cell, such as the j-th cell cluster (i.e., the j-th equivalent cell), are described by the internal resistance (Rint) model.

[0127] Based on the reference internal resistance of the reference cell, the individual cell internal resistances of other cells within the cluster, and the internal resistance difference index of other cells within the cluster, the equivalent internal resistance of the target equivalent cell is determined using the following formula:

[0128]

[0129] in, The equivalent internal resistance of the target equivalent cell (specifically the j-th equivalent cell) is also the equivalent internal resistance of the j-th cell cluster (representing the cluster equivalent internal resistance). For the first The internal resistance of a single cell; This is the internal resistance of the converter (also known as the series resistance of the DC / DC converter). The converter is used to regulate the power of the corresponding single battery cell. The converter and the single battery cell are connected in series. The number of individual cells within the j-th battery cluster; Let be the internal resistance of the i-th individual cell in the j-th battery cluster; This is the internal resistance difference index.

[0130] Among them, the reference internal resistance of the reference battery can be used as the first... The internal resistance of a single cell is represented by its resistance.

[0131] Based on the equivalent internal resistance and the charging and discharging current of the target equivalent battery, the power loss level of the target equivalent battery (i.e., cluster power loss) is determined by the following formula:

[0132]

[0133] in, This indicates the degree of power loss of the target equivalent cell (i.e., the j-th equivalent cell); This represents the charging and discharging current of the target equivalent battery at time t.

[0134] Optionally, based on the objective optimization function and the operating constraints corresponding to the multiple equivalent batteries, the power configuration of each individual cell in the energy storage battery pack is determined, including: determining the equivalent open-circuit voltage corresponding to the multiple equivalent batteries; and determining the power configuration of each individual cell in the energy storage battery pack based on the objective optimization function, the operating constraints corresponding to the multiple equivalent batteries, and the equivalent open-circuit voltage.

[0135] This involves the equivalent open-circuit voltage, which is the open-circuit voltage corresponding to the equivalent cell. It is obtained by aggregating the open-circuit voltages of each individual cell within the corresponding cluster and is used to characterize the potential characteristics of the equivalent cell.

[0136] By determining the corresponding equivalent open-circuit voltage for each equivalent cell, the electrical potential characteristics of the equivalent cell can be fully characterized. By combining the equivalent open-circuit voltage with the objective optimization function and operating constraints, a global optimal solution can be achieved under the premise of meeting the safety boundary. This ensures that the output single-cell power configuration takes into account both minimum loss, balanced charge, and consistent temperature. Based on the single-cell power configuration determined by the above parameters, the actual electrical characteristics of each single cell can be accurately matched, avoiding single-cell overload or uneven heating, and achieving efficient, stable, and consistent operation of large-scale energy storage battery packs.

[0137] The objective function is:

[0138]

[0139] in, This is the equilibrium penalty coefficient corresponding to the state of charge, which is also the SOC equilibrium weight coefficient. This is the equilibrium penalty coefficient corresponding to temperature, also known as the temperature consistency weighting coefficient. Let be the state of charge deviation of the j-th equivalent cell (i.e., the energy / SOC equilibrium slack variable). Let be the temperature deviation of the j-th equivalent cell (i.e., the temperature equilibrium slack variable). This is the integral term for the prediction time domain (from the current time t to the future time t+H, where H is the length of the prediction time domain).

[0140] Compared to the traditional approach of minimizing power loss, the objective optimization function is based on a cluster-level electrothermal coupling model. It minimizes the total electrothermal coupling loss of the cluster within the rolling time domain [t, t+H) while penalizing SOC / temperature imbalance, thus achieving convex optimization allocation between clusters based on the electrothermal coupling aggregation model.

[0141] in, The formula for determining it is:

[0142]

[0143] in, The remaining energy of the j-th equivalent battery (the total electrical energy stored at time t); Let be the equivalent capacitance of the j-th equivalent cell. , This is the scaling factor; For the first The remaining energy of an equivalent battery (the total electrical energy stored at time t). For the first The equivalent capacitance of an equivalent cell; Let be the allowable energy balance deviation threshold for the j-th equivalent battery; This represents the total number of multiple battery clusters (i.e., the total number of multiple equivalent batteries).

[0144] When the energy / SOC deviation between clusters (between different clusters) exceeds the allowable value If the value is greater than 0, it is penalized in the objective function and forced back to equilibrium.

[0145] in, The formula for determining it is:

[0146]

[0147] in, The average temperature of all equivalent cells; The threshold for allowable temperature equalization deviation.

[0148] When the temperature difference exceeds the limit, A value greater than 0 is penalized by the objective optimization function to ensure temperature balance.

[0149] The following formula can be used to determine this. :

[0150]

[0151] in, Let be the state of charge of the j-th equivalent cell.

[0152] Optionally, the average state of charge (SOC) of multiple equivalent cells is determined as follows: The average remaining charge and average equivalent capacitance of the multiple equivalent cells are determined; based on the average remaining charge and average equivalent capacitance of the multiple equivalent cells, the average SOC is determined using the following formula:

[0153]

[0154] in, This represents the average state of charge.

[0155] This involves the average state of charge (SBC), which is the overall average SBC of all equivalent cells and serves as a unified benchmark for measuring the degree of balance in charge among equivalent cells.

[0156] This includes the average remaining energy, which is the arithmetic mean of the current remaining energy of all equivalent batteries, used to characterize the overall energy reserve level of the battery pack.

[0157] This includes the average equivalent capacitance, which is the arithmetic mean of the equivalent capacitances of all equivalent cells and is used to characterize the average level of the overall charge storage capacity of the battery pack.

[0158] By determining the average remaining capacity and average equivalent capacitance of multiple equivalent cells, a unified reference quantity characterizing the overall energy reserve and storage capacity of the battery pack can be obtained. Based on the average remaining capacity and average equivalent capacitance, the average state of charge can be calculated, resulting in an objective and stable global energy balance benchmark. This ensures that the calculation results of the state of charge deviation between equivalent cells are true and reliable, providing an accurate basis for the balance evaluation of the objective optimization function and improving the accuracy and stability of energy balance control during power distribution.

[0159] The discretized form of the objective function is as follows:

[0160]

[0161] in, Let j be the state variable vector of the j-th equivalent battery. Indicates transpose; The total number of time-domain steps (total length) for prediction. Let be the power loss level of the j-th equivalent cell at time t; The penalty weight is specifically the SOC equilibrium penalty weight. Let be the SOC equilibrium relaxation variable of the j-th equivalent cell at time t, representing the state of charge deviation; This is the penalty weight, specifically the penalty weight for temperature consistency; Let be the temperature uniformity relaxation variable of the j-th equivalent cell at time t, representing the temperature deviation.

[0162] in, , The charging and discharging power of the j-th equivalent battery (a variable used for optimization control); This indicates the power loss level of the j-th equivalent battery.

[0163] in, The operator is optimized to minimize the objective.

[0164] Based on the above formula, a discretized representation of the convex optimization problem is achieved, where... The method of determination can be with The determination method is the same. The method of determination and The method of determination is the same.

[0165] The above formula, by introducing the remaining energy of the cluster The non-convex power balance constraint is transformed into a convex constraint to achieve convexity processing, ensuring a fast global optimum solution. An energy relaxation (slack) variable is added. and temperature Slack variables Avoiding constraints is not feasible.

[0166] The equilibrium threshold is dynamically adjusted as the slack variable approaches zero. The adaptive update formula is:

[0167]

[0168] in, For the updated ; This represents the energy deviation within the cluster between a single cell and the cluster centroid (central cell) in the k-th battery cluster. This energy deviation is the maximum energy deviation between a single cell and the cluster centroid (central cell). This represents the maximum value of the energy deviation within all battery clusters.

[0169] The adaptive update formula is:

[0170]

[0171] in, For the updated ; This represents the temperature deviation between the individual cells within the k-th cell cluster and the cluster centroid (central cell). This represents the maximum intra-cluster temperature deviation among all battery clusters.

[0172] The formula for determining the equivalent open-circuit voltage (i.e., cluster open-circuit voltage) corresponding to multiple equivalent cells is as follows:

[0173]

[0174] in, Let be the equivalent open-circuit voltage of the j-th equivalent cell; Let be the open-circuit voltage of the i-th individual cell in the j-th cell cluster.

[0175] Based on the objective optimization function and the operating constraints and equivalent open-circuit voltages corresponding to multiple equivalent batteries, the individual power configuration of each cell in the energy storage battery pack is determined. Specifically, this includes: determining the target equivalent current (i.e., the optimal cluster current, which is the current that minimizes the objective optimization function) corresponding to multiple equivalent batteries based on the objective optimization function and the operating constraints corresponding to multiple equivalent batteries; determining the equivalent power configuration of multiple equivalent batteries based on the target equivalent current and equivalent open-circuit voltages corresponding to multiple equivalent batteries; and determining the individual power configuration of each cell in the energy storage battery pack based on the equivalent power configuration of multiple equivalent batteries.

[0176] This involves the target equivalent current, which is also the optimal cluster current. The equivalent battery operating current is the value that minimizes the target optimization function and is a key intermediate variable for achieving optimal allocation.

[0177] This involves equivalent power configuration, which is the optimal charging and discharging power allocated to each equivalent battery based on the target equivalent current and equivalent open-circuit voltage, thereby minimizing the value of the target optimization function.

[0178] By combining the objective optimization function with operational constraints to solve for the target equivalent current, the optimal synergy of system loss, power balance, and thermal consistency can be achieved while meeting safety limits. By determining the equivalent power configuration based on the target equivalent current and equivalent open-circuit voltage, the optimal cluster-level command can be accurately converted into power form. Furthermore, by determining the individual cell power configuration from the equivalent power configuration, the cluster-level power can be reasonably distributed to the individual cell level, ensuring consistent operating status of each cell in a large-scale energy storage battery pack, avoiding overload and unbalanced heating, and improving the overall system efficiency and reliability.

[0179] Among them, based on the target equivalent current and equivalent open-circuit voltage corresponding to each of the multiple equivalent cells, the equivalent power configuration corresponding to each of the multiple equivalent cells is determined, as shown in the formula:

[0180]

[0181] in, Configure the equivalent power for the j-th equivalent cell; Let be the target equivalent current of the j-th equivalent cell.

[0182] Based on the electrothermal polymerization model, it can be seen that , , All three quantities are uniquely determined by the cluster current (equivalent current). That is, the entire link is to minimize the objective function → obtain the optimal current → calculate the cluster power using the equivalent open-circuit voltage → distribute it to individual cells within the cluster. Therefore, in order to minimize the objective function, the optimal cluster current (i.e., the target equivalent current) can be directly solved.

[0183] Optionally, based on the objective optimization function and the operating constraints and equivalent open-circuit voltages corresponding to multiple equivalent batteries, the individual power configuration of each cell in the energy storage battery pack is determined, including: determining the equivalent power configuration corresponding to each of the multiple equivalent batteries based on the objective optimization function and the operating constraints corresponding to each of the multiple equivalent batteries; and determining the individual power configuration corresponding to each cell in the energy storage battery pack based on the equivalent power configuration corresponding to each of the multiple equivalent batteries and the individual power allocation strategy. This involves an individual power allocation strategy, which is a strategy to decompose the cluster-level equivalent power to individual cells to adapt to the control requirements of different scenarios. The individual power allocation strategy includes any of the following:

[0184] The power quota-based allocation strategy, wherein the power quota includes equal power quota, that is, the power quota-based allocation strategy is a strategy of allocating power according to a preset quota, including equal power quota, which is suitable for scenarios where the characteristics of individual units are highly consistent.

[0185] The allocation strategy based on power loss is an allocation strategy that aims to minimize the power loss of the energy storage battery pack, and prioritizes reducing system energy loss.

[0186] The allocation strategy based on operating characteristics is an allocation strategy aimed at minimizing the state deviation between the operating state and the predetermined state of the energy storage battery pack, in order to improve operational consistency and control accuracy.

[0187] By combining the objective optimization function, operational constraints, and equivalent open-circuit voltage to determine the equivalent power configuration, the globally optimal power allocation result at the cluster level can be obtained while meeting the safety boundary. By determining the individual power configuration based on the equivalent power configuration and the selected individual power allocation strategy, the optimal power at the cluster level can be reasonably distributed to the individual level, balancing computational efficiency and control accuracy. By providing a variety of differentiated individual power allocation strategies, it can adapt to different hardware conditions, control accuracy, and efficiency requirements, ensuring that the individual power allocation is more in line with the actual operating characteristics of the battery, avoiding local overload and uneven heating, and improving the stability, efficiency, and balance of large-scale energy storage battery pack operation.

[0188] The power quota-based allocation strategy (aiming to achieve equal power allocation, suitable for scenarios with highly consistent individual units and low computing power) has the following formula:

[0189]

[0190] in, Configure the power of the i-th cell.

[0191] Thus obtain .

[0192] The power loss-based allocation strategy (specifically, internal resistance-weighted allocation, prioritizing power loss reduction to suit loss-sensitive applications) has the following formula:

[0193]

[0194] in, Let be the internal resistance of the i-th individual cell; For the j-th battery cluster, the j-th The internal resistance of a single cell.

[0195] The allocation strategy based on operational characteristics involves constructing a small-scale convex optimization by building a single-cell electrothermal model to minimize the total loss within the cluster while satisfying safety and balance constraints. This is suitable for high-precision scenarios. In other words, a small-scale single-cell optimization is solved again to obtain the optimal current of the single cell, and then the power of the single cell is obtained (i.e., convex optimization allocation within the cluster). This is the same as the method of solving the equivalent power configuration of the equivalent cell for the energy storage battery pack, and will not be elaborated further.

[0196] Among them, the single-unit convex optimization constraints based on the allocation strategy according to operational characteristics include:

[0197] 1) Safety constraints:

[0198]

[0199] in, For single cell batteries The minimum value of the current (specifically the charging and discharging current); For single cell batteries The current; For single cell batteries The maximum value of the current (specifically the charging and discharging current).

[0200]

[0201] in, For single cell batteries The current SOC; Let S be the average SOC of all individual cells within the j-th battery cluster; This is the permissible deviation threshold for SOC within the cluster.

[0202]

[0203] in, For single cell batteries The current temperature; Let be the average temperature of all individual cells within the j-th cell cluster; This is the allowable temperature deviation threshold within the cluster.

[0204] 2) Power quota constraints:

[0205]

[0206] in, For single cell batteries The port output power; For single cell batteries The internal resistance; For the series resistor of the DC / DC converter; The equivalent power configuration for the j-th battery cluster, i.e., the optimal equivalent power; Let be the fixed power loss of the j-th battery cluster.

[0207] To address this, the objective optimization function is based on the collector-cell thermal coupling aggregation model. By minimizing coupling power loss and balancing relaxation variable penalty terms, the optimal charging and discharging current of each cluster is obtained. Then, based on the cluster equivalent open-circuit voltage determined in the cluster electrical aggregation model, the optimal charging and discharging power of each cluster is calculated according to the basic power relationship. Finally, according to the three preset allocation strategies mentioned above, the cluster power quota is allocated to each battery cell within the cluster, completing the final power allocation.

[0208] Optionally, the operating constraints corresponding to the multiple equivalent cells include: current constraints, temperature constraints, state of charge constraints, operating power constraints, and power quota constraints.

[0209] This involves operational constraints, which are the electrical and thermal safety limits that the equivalent battery must meet during charging and discharging to ensure the stable operation of the equivalent battery and individual cells within the cluster.

[0210] This involves current constraints, which limit the upper and lower limits of the equivalent battery charging and discharging current to prevent overcurrent-induced heat generation and damage.

[0211] This involves temperature constraints, which include temperature range restrictions, temperature equilibrium restrictions, and temperature equilibrium restrictions with relaxation variables, used to control the temperature within a safe range and maintain temperature consistency between clusters.

[0212] This involves a temperature balance constraint, which is a constraint that controls the deviation between the equivalent cell temperature and the average temperature of all clusters within an allowable range, in order to maintain thermal equilibrium between clusters.

[0213] This involves temperature equilibrium constraints with relaxation variables. These constraints are temperature limits that introduce relaxation variables to achieve flexible equilibrium, thereby avoiding constraint conflicts and ensuring a smooth equilibrium process.

[0214] This involves state of charge constraints, which include state of charge range restrictions, state of charge balance restrictions, and energy balance restrictions with slack variables, used to prevent overcharging and over-discharging and maintain energy balance.

[0215] This involves a state of charge balance constraint, which is a constraint that controls the deviation between the equivalent battery state of charge and the average state of charge of all clusters within a threshold, in order to achieve consistency in charge levels between clusters.

[0216] This involves energy / SOC equilibrium constraints with relaxation variables. These constraints are charge state restrictions that introduce relaxation variables to achieve flexible equilibrium, thereby improving the feasibility and robustness of optimization.

[0217] This involves the SOC allowable deviation threshold, which is the maximum allowable deviation value of the charge state between clusters. It can be adaptively updated according to the deviation of individual cells within the cluster and is used to dynamically adjust the equalization intensity.

[0218] This involves operating power constraints, including power balance constraints, which are used to ensure that the total power of each equivalent battery meets the total power demand of the system's external output.

[0219] This involves a power balance constraint, which is a constraint that the sum of the output power of each equivalent battery after deducting losses equals the total output power of the system, and is used to maintain the matching of power supply and demand in the system.

[0220] This involves power quota constraints, which include power quota limits under both discharge and charging conditions, to ensure that the power allocated at the cluster level does not exceed the capacity of individual cells.

[0221] This includes power quota constraints under discharge conditions, which stipulate that the cluster power quota shall not exceed the sum of the maximum allowable power of all individual cells within the cluster, in order to prevent individual cell overload during discharge.

[0222] This includes power quota constraints under charging conditions, which stipulate that the cluster charging power quota shall not exceed the sum of the minimum allowable power of all individual cells within the cluster, in order to prevent overcurrent in individual cells during charging.

[0223] Specifically as follows:

[0224] (1) Current constraints include current range constraints (safety constraints), which are expressed as:

[0225]

[0226] in, Let be the minimum value of the equivalent current of the j-th equivalent cell; Let be the equivalent current of the j-th equivalent cell; It represents the maximum value of the equivalent current of the j-th equivalent cell.

[0227] (2) Temperature constraints include:

[0228] 1) Temperature range constraints (safety constraints):

[0229]

[0230] in, Let be the minimum equivalent temperature of the j-th equivalent cell; It represents the maximum value of the equivalent temperature of the j-th equivalent cell.

[0231] 2) Temperature equilibrium constraint:

[0232]

[0233] in, This represents the average temperature across all battery clusters.

[0234] 3) Temperature equilibrium (balance) constraint with slack variables:

[0235]

[0236] Among them, slack variables .

[0237] (3) Charge state constraints include:

[0238] 1) Charge state interval constraints (safety constraints):

[0239]

[0240] in, Let be the minimum equivalent SOC of the j-th equivalent cell; Let SOC be the equivalent state of charge (SOC) of the j-th equivalent cell. It represents the maximum value of the equivalent SOC of the j-th equivalent cell.

[0241] 2) Charge state balance constraints:

[0242]

[0243] in, The equivalent SOC mean for all battery clusters; This is the allowable deviation threshold for SOC.

[0244] in, The update process is as follows:

[0245]

[0246] in, For the updated ; This represents the maximum deviation between the individual SOC and the centroid of the k-th battery cluster.

[0247] 3) Energy / SOC equilibrium constraints with slack variables:

[0248]

[0249] Among them, slack variables .

[0250] (4) Operating power constraints include power balance constraints, expressed as:

[0251]

[0252] in, This refers to the total power that the energy storage battery pack needs to output.

[0253] (5) Power quota constraints include:

[0254] 1) Power quota constraints under discharge conditions:

[0255]

[0256] The power quota of a cluster (i.e., a battery cluster) (determined by the upper limit of current) is equal to the sum of the power quotas of all individual cells in the cluster (determined by the upper limit of current of each individual cell). The power allocated to each battery cluster cannot exceed the sum of the power that all individual cells in the cluster can withstand, thus ensuring that individual cells will not be overloaded from the source.

[0257] in, It represents the maximum current of the i-th individual cell.

[0258] 2) Power quota constraints under charging conditions (power quota is achieved through current quota to ensure that individual cells do not experience overcurrent during charging):

[0259]

[0260] in, Let be the minimum current of the i-th individual cell.

[0261] S108 allocates power to each individual battery cell according to the corresponding individual cell power configuration.

[0262] Through steps S102-S108 above, multiple battery clusters are obtained by clustering the individual cells of the energy storage battery pack. This allows for the classification of individual cells with similar characteristics, reducing the complexity of subsequent power allocation. By performing virtual parallel connection on the cells within each cluster to obtain corresponding equivalent cells, the multi-cell system can be equivalently converted into a small number of equivalent units for unified scheduling. This ensures the accuracy and reliability of equivalent modeling and state estimation while simplifying the calculation and control process. The objective optimization function aims to minimize the power loss, state-of-charge deviation, and temperature deviation of each equivalent cell, simultaneously taking into account... Energy consumption suppression, power balance, and thermal consistency management are achieved by combining the objective optimization function and the operating constraints corresponding to multiple equivalent batteries to determine the power configuration of individual cells. This enables precise and reasonable power allocation of the energy storage battery pack, reducing system losses, improving power balance and temperature consistency while meeting safe operating conditions. It also avoids local overload or uneven heating caused by differences between individual cells, thus solving the technical problem in related technologies where unreasonable power allocation occurs when distributing power among individual cells in large-scale energy storage battery packs, leading to local overload or uneven heating during the coordinated operation of individual cells.

[0263] Based on the above embodiments and optional embodiments, an optional implementation method is provided, which is described in detail below.

[0264] In related technologies, with the widespread application of energy storage technology in electrified transportation, smart grids, and renewable energy grid integration, large-scale battery energy storage systems (BESS) have become core equipment for driving energy transformation. To ensure the safe operation of BESS, it is necessary to rationally allocate power among the individual cells of a large-scale energy storage battery pack. However, in related technologies, there are technical problems with unreasonable power allocation among the individual cells of a large-scale energy storage battery pack, leading to localized overload or uneven heating during the coordinated operation of the individual cells.

[0265] For example, in related technologies, the following problems exist when performing power distribution:

[0266] 1) High computational complexity: Large-scale BESS contains a massive number of battery cells. Traditional cell-level optimization methods need to handle a large number of decision variables and high-dimensional optimization space, which leads to an exponential increase in computational overhead with the number of batteries, making it difficult to meet real-time control requirements.

[0267] 2) Insufficient scalability: Although existing hierarchical control and distributed control methods can reduce the amount of computation to a certain extent, they do not make full use of the similarity of the characteristics of individual battery cells. When the number of batteries increases significantly, there is still a computational bottleneck, which cannot effectively adapt to ultra-large-scale energy storage systems.

[0268] 3) Poor multi-objective coordination: Power allocation needs to take into account multiple objectives such as minimizing power loss, equalizing SOC, and equalizing temperature. Related technologies often focus on optimizing a single objective or ignore the impact of differences in battery characteristics in multi-objective optimization, resulting in a decline in overall control performance.

[0269] 4) Limited model adaptability: Traditional optimization methods rely on accurate modeling at the cell level. The heterogeneity of cell cells in large-scale systems (differences in SOC, temperature, and internal resistance) leads to a sharp increase in model complexity, and it is difficult to balance modeling accuracy and computational efficiency.

[0270] There is currently no effective solution to the above problems.

[0271] In view of this, an optional embodiment of the present invention provides a power distribution method for an energy storage battery pack, which can effectively solve the above-mentioned technical problems.

[0272] Figure 4 This is a flowchart of the optimal power allocation process for large-scale energy storage battery packs based on clustering, as described in an optional embodiment of the present invention. Figure 4 As shown, this paper illustrates the entire process logic of scalable optimal power allocation for large-scale energy storage battery packs based on clustering. It fully presents the closed-loop technical link of "data acquisition - clustering grouping - model building - optimization control - power allocation". The process is centered on closed-loop feedback and is unfolded in the following order according to functional modules:

[0273] (1) State perception and data acquisition: First, the core operating characteristics of the battery cells, such as state of charge (SOC), temperature, and internal resistance, are collected to provide basic data for subsequent clustering and model building;

[0274] (2) Clustering and grouping: The k-means algorithm is used to divide the massive number of individuals into clusters with similar characteristics (the number of clusters k is optimized by gap statistics method), thereby achieving dimensionality reduction from "high-dimensional individual space to low-dimensional cluster space" and solving the computational complexity problem of large-scale systems.

[0275] (3) Cluster-level aggregation modeling: Construct an "electrothermal aggregation model" for each cluster, integrating the equivalent electrical parameters (such as capacity, internal resistance, open circuit voltage) and thermal characteristics (such as temperature dynamics, thermal resistance) of the cluster to form a global characteristic representation of the cluster;

[0276] (4) Inter-cluster convex optimization control: With the goal of "minimizing the total power loss of the cluster", safety constraints such as current, SOC, and temperature and balance constraints are incorporated, and the target power quota of each cluster is solved by convex optimization.

[0277] (5) Adaptive equilibrium adjustment: Based on the degree of deviation of individual characteristics within the cluster, dynamically tighten / relax the equilibrium thresholds of SOC and temperature to achieve progressive optimization from "cluster-level equilibrium to individual-level equilibrium";

[0278] (6) Power allocation scheme selection: Three differentiated intra-cluster power allocation strategies are provided: equal power allocation (adapted to low computing resource scenarios), internal resistance-based allocation (adapted to loss-sensitive scenarios), and single-unit level convex optimization allocation (adapted to high-precision control scenarios), to achieve a balance between algorithm flexibility and control accuracy.

[0279] (7) Execution and feedback: The allocated power command is converted into the drive signal of the DC / DC converter to control the operation of the battery pack, and the individual cell status is updated through real-time data feedback to form a closed-loop optimization.

[0280] Based on the above technical path of "clustering dimensionality reduction - aggregation modeling - hierarchical optimization", it not only realizes the scalable control of large-scale battery packs, but also takes into account the balance of performance and loss optimization. It is a complete logical carrier of scalable optimal power allocation technology, which will be described in detail below.

[0281] S1, Multi-dimensional feature acquisition and clustering of individual battery cells:

[0282] Collect the core features of each individual cell, such as SOC, temperature, and internal resistance, and construct a feature vector;

[0283] The k-means clustering algorithm is used to divide a large number of individuals into a small number of clusters, ensuring that the characteristics within clusters are similar and the differences between clusters are significant.

[0284] The number of clusters is optimized by using gap statistics to balance modeling accuracy and computational efficiency.

[0285] S2, Cluster-level Electrothermal Aggregation Model Construction:

[0286] Electrical Model: By aggregating parameters such as cluster capacity, current, SOC, open-circuit voltage, and internal resistance, an overall electrical dynamics and power loss model of the cluster is established.

[0287] Thermal model: Aggregates parameters such as temperature, mass, and thermal resistance of clusters to characterize the temperature variation law of clusters and provide a basis for temperature equalization control.

[0288] S3, convex optimization of optimal power allocation between clusters:

[0289] With the goal of minimizing total power loss due to cluster electrothermal coupling, safety constraints, balance constraints, and power equilibrium constraints are incorporated.

[0290] By substituting variables, non-convex optimization problems are transformed into convex optimization problems, ensuring efficient solution of the global optimum.

[0291] By introducing the slack variable to handle constraint conflicts and combining it with adaptive equilibrium threshold adjustment, the process of progressively balancing from cluster equilibrium to individual equilibrium is achieved.

[0292] S4, Individual power allocation within the cluster:

[0293] The power quota-based allocation strategy (equal power allocation) is suitable for scenarios where individual unit characteristics are highly consistent and has the highest computational efficiency.

[0294] The allocation strategy based on power loss (based on internal resistance) is an allocation strategy that aims to minimize the power loss of the energy storage battery pack, prioritizing the reduction of power loss and adapting to loss-sensitive applications.

[0295] The allocation strategy based on operating characteristics (optimal power allocation) aims to minimize the state deviation between the operating state and the predetermined state of the energy storage battery pack. It achieves high-precision control through single-cell convex optimization and is suitable for scenarios with stringent performance requirements.

[0296] Figure 5 This is a schematic diagram of the circuit structure of a large-scale battery energy storage system in an optional embodiment of the present invention, such as... Figure 5 The diagram illustrates the hierarchical collaborative control architecture of a Battery Energy Storage System (BESS), also known as a Battery Management System (BMS) for large-scale energy storage battery packs. Its core logic involves a three-tiered linkage of "decision layer - execution layer - perception layer" to achieve precise power regulation of battery clusters / cells. The architecture comprises three functional layers:

[0297] (1) BMS decision layer: integrates "optimal power allocation module" and multiple "local controllers". The former is based on clustering algorithm and convex optimization model to complete the global calculation of target power at the battery cluster level; the latter converts cluster-level power instructions into hardware-executable control signals to realize the mapping of algorithm decision to hardware execution.

[0298] (2) Battery Cluster Execution Layer: Adopting a modular design of "single cell - independent DC / DC converter", each cluster corresponds to a set of "Cell-DC / DC" units, which receive control signals from the local controller and realize independent control of the charging and discharging current of individual cells within the cluster through topology adjustment of the DC / DC converter (such as Buck / Boost conversion). It is the hardware carrier of the power distribution strategy. Among them, the "Cell-DC / DC" unit is a unit composed of a single battery cell and an independent DC / DC converter. The Buck conversion is used to charge the battery, and the Boost conversion is used to discharge the battery.

[0299] (3) Measurement and sensing layer: Real-time acquisition of voltage, current and temperature of each battery cell is completed through the measurement bus. On the one hand, it provides state perception data for the BMS decision layer (supporting the input of SOC estimation and optimization algorithm), and on the other hand, it realizes the energy interaction interface between the load / charger and the battery pack.

[0300] This architecture, through layered decoupling of "algorithm-control-hardware", not only adapts to the scalable optimization needs of large-scale battery packs, but also ensures the accuracy of power regulation at the individual cell level. It is the core hardware-software collaborative carrier for achieving battery pack SOC balancing and minimizing losses.

[0301] Figure 6 This is a schematic diagram of a cluster-based power allocation process for large-scale energy storage battery packs in an optional embodiment of the present invention, as shown below. Figure 6 As shown, battery cells with similar characteristics are divided into multiple "similar units" (clusters) through "clustering". Then, "aggregation model construction" is performed on each cluster (using a virtual battery to represent the characteristics of the entire cluster). Next, "inter-cluster optimization control" is completed through "convex optimization" and the reference power value of each cluster is output. Finally, the power of the cluster is allocated to each cell within the cluster to achieve hierarchical power allocation of large-scale battery packs.

[0302] Figure 7 This is a schematic diagram of battery clustering in a large-scale battery energy storage system according to an optional embodiment of the present invention, such as... Figure 7 The diagram shows the clustering of a 400-cell battery energy storage system at k=4. Different clusters are labeled with different colors (a total of four categories: blue, purple, red, and orange). Clusters of different colors correspond to relatively greater differences in battery performance, while batteries within the same cluster are more similar in terms of SOC, temperature, and internal resistance.

[0303] Figure 8 This is a schematic diagram of the principle architecture of a large-scale battery energy storage system in an optional embodiment of the present invention, such as... Figure 8The diagram illustrates the layered architecture of the control and hardware of a large-scale energy storage battery pack BMS: The upper BMS control layer includes a SOC estimation module (collecting cell voltage / current / temperature and calculating SOC), an optimal power allocation module (performing clustering / convex optimization algorithms based on SOC to output the optimal target current for each cell), and a local controller (converting the target current into a pulse width modulation signal); the lower circuit structure layer adopts a modular design of "single cell combined with an independent DC / DC converter," with each cell corresponding to a DC / DC module containing a metal-oxide-semiconductor field-effect transistor (MOSFET), inductor, and capacitor. It adjusts the cell charging and discharging current through pulse width modulation (PWM) signals, while simultaneously collecting cell status data via a measurement bus and feeding it back to the control layer, ultimately forming a closed loop of "state acquisition to algorithm decision-making, and final hardware execution," achieving cell-level power allocation and SOC / temperature balancing for large-scale battery packs.

[0304] The following description, with specific examples, will further illustrate this point.

[0305] Example 1, simulation verification:

[0306] Experimental parameters: 400 high-rate power lithium-ion batteries were selected, with a nominal voltage of 3.6V, a capacity of 2.5Ah, and an internal resistance of 31.3mΩ per cell; the system output power was based on the driving conditions of a dynamometer on urban roads, with a peak charging power of 6kW and a discharging power of 10kW; the optimization time domain H=10s, and the sampling time Δt=1s; the SOC equalization threshold Δq=0.5%, and the temperature equalization threshold ΔT=0.5K; among them, Scheme 1 is an allocation strategy based on power quota (equal power allocation), Scheme 2 is an allocation strategy based on power loss (based on internal resistance allocation), and Scheme 3 is an allocation strategy based on operating characteristics (optimal power allocation).

[0307] Initial conditions: The initial SOC of the monomer follows a uniform distribution of U(0.7,0.75), the temperature follows a uniform distribution of U(301,305)K, and the internal resistance follows a uniform distribution of U(31.3,41.3)mΩ.

[0308] Experimental results:

[0309] Equalization performance: Scheme 3 achieves a SOC equalization time of 700s and a temperature equalization time of 1100s, which is better than Scheme 1 (SOC equalization 1000s, temperature equalization 1400s) and Scheme 2 (SOC equalization 1000s, temperature equalization 1700s).

[0310] Loss performance: The final cumulative power loss is reduced by 4.6% compared with the traditional single-unit level optimization, and Scheme 3 has the lowest loss in the unbalanced stage;

[0311] Computational efficiency: Scheme 1 (15 clusters) has a computation time of only 6.76s, which is 99.43% lower than the traditional single-unit optimization (1200.45s).

[0312] Example 2, experimental verification:

[0313] Experimental setup: 20-cell 4-series 5-parallel (4s5p) battery pack prototype, MOSFET, gate driver, 33μH inductor, main controller, local controller; K-type thermocouples to measure cell temperature, data acquisition card to record data. The K-type thermocouples are temperature sensors that can be used to measure the surface temperature of the batteries.

[0314] Experimental conditions: initial SOC 75%-80%, initial temperature 21.7℃, upper limit of discharge current 5A, optimization step size 30s, experimental duration 30 minutes;

[0315] Experimental results:

[0316] Equalization performance: Scheme 3 achieves SOC equalization time of 750s, while Scheme 1 achieves 900s; temperature deviation is always controlled within the equalization threshold.

[0317] Power allocation: Scheme 3 achieves dynamic power adjustment through optimal allocation within the cluster, and the power of each individual unit tends to be consistent after balancing;

[0318] Reliability: There were no constraint conflicts during the experiment. The slack variable was initially non-zero and gradually converged to zero, verifying the feasibility of convex optimization.

[0319] To address this, k-means clustering was performed based on the SOC, temperature, and internal resistance characteristics of individual battery cells. A cluster-level electro-thermal aggregation model was constructed to characterize the overall dynamic characteristics of the cluster. Optimal power allocation among clusters was achieved through convex optimization, combined with adaptive equalization adjustment to improve control accuracy. Three individual cell power allocation schemes within the cluster were designed to adapt to different scenarios. Validation was performed on a 400-cell simulation system and a 20-cell experimental prototype. The computational overhead was reduced by more than 98% compared to traditional single-cell level optimization, the SOC and temperature equalization speed was improved by more than 30%, and power loss was reduced by 4.6%, providing reliable technical support for the efficient and safe operation of large-scale energy storage battery packs.

[0320] The above optional implementation methods can achieve at least the following beneficial effects:

[0321] (1) Compared with related technologies, this invention transforms the large-scale single-unit optimization problem into a small-scale cluster optimization problem through clustering, reducing the number of computational variables from Xn to Xk(k n), where the number of variables for each individual battery cell is X. The computational overhead is reduced by more than 60% for small-scale systems and by more than 98% for large-scale systems (400 cells), meeting the requirements for real-time control. Moreover, the number of clusters does not increase linearly with the number of cells, and the clustering results are dynamically adjusted according to the battery equilibrium state, which can seamlessly adapt to large-scale BESS from hundreds to tens of thousands of cells.

[0322] (2) Compared with related technologies, this invention accurately captures the overall dynamics through a cluster electrothermal aggregation model and combines adaptive equalization adjustment to achieve synergistic optimization of SOC equalization, temperature equalization and power loss minimization. The equalization speed is improved by more than 30% in 20 individual prototype experiments.

[0323] (3) Compared with related technologies, this invention supports the heterogeneity of battery cells (different types and different decay states), handles constraint conflicts through slack variables, and adapts to parameter fluctuations under complex working conditions; and the three power allocation schemes within the cluster can be flexibly selected according to computing resources and control accuracy requirements, and are applicable to various scenarios such as grid energy storage, electric vehicles, and renewable energy grid connection.

[0324] It should be noted that, for the sake of simplicity, the foregoing method embodiments are all described as a series of actions. However, those skilled in the art should understand that the present invention is not limited to the described order of actions, because according to the present invention, some steps can be performed in other orders or simultaneously. Furthermore, those skilled in the art should also understand that the embodiments described in the specification are preferred embodiments, and the actions and modules involved are not necessarily essential to the present invention.

[0325] Through the above description of the embodiments, those skilled in the art can clearly understand that the methods according to the above embodiments can be implemented by means of software plus necessary general-purpose hardware platforms. Of course, they can also be implemented by hardware, but in many cases the former is a better implementation method. Based on this understanding, the technical solution of the present invention, or the part that contributes to the prior art, can be embodied in the form of a software product. This computer software product is stored in a storage medium (such as ROM / RAM, magnetic disk, optical disk) and includes several instructions to cause a terminal device (which may be a mobile phone, computer, server, or network device, etc.) to execute the methods of the various embodiments of the present invention.

[0326] Example 2

[0327] According to embodiments of the present invention, an apparatus for implementing the power distribution method of the above-described energy storage battery pack is also provided. Figure 9 This is a structural block diagram of a power distribution device for an energy storage battery pack according to an embodiment of the present invention, such as... Figure 9As shown, the device includes: a first determining module 902, a second determining module 904, a third determining module 906, and a fourth determining module 908. The device will be described in detail below.

[0328] The first determining module 902 is used to perform performance clustering on each individual cell of the energy storage battery pack to obtain multiple battery clusters.

[0329] The second determining module 904 is connected to the first determining module 902 and is used to virtually connect the individual cells in each of the multiple battery clusters in parallel to obtain the equivalent cells corresponding to the multiple battery clusters respectively. The virtual parallel connection is to virtually connect the individual cells in each cluster in parallel by modeling without changing the actual physical connection relationship of the individual cells.

[0330] The third determining module 906, connected to the second determining module 904, is used to determine the power configuration of each cell in the energy storage battery pack based on the objective optimization function and the operating constraints corresponding to the multiple equivalent cells. The objective optimization function is an optimization function that aims to minimize the sum of the power loss degree, state of charge deviation and temperature deviation of the multiple equivalent cells. The state of charge deviation is the deviation between the corresponding equivalent cell and the average state of charge of the multiple equivalent cells, and the temperature deviation is the deviation between the corresponding temperature and the average temperature of the multiple equivalent cells.

[0331] The fourth determining module 908 is connected to the third determining module 906 and is used to allocate power to each individual battery cell according to the corresponding individual cell power configuration.

[0332] It should be noted that the first determining module 902, the second determining module 904, the third determining module 906, and the fourth determining module 908 mentioned above correspond to steps S102 to S108 in the power distribution method for implementing the energy storage battery pack. The multiple modules and the corresponding steps are the same in terms of implementation examples and application scenarios, but are not limited to the content disclosed in the above embodiment 1.

[0333] Example 3

[0334] According to another aspect of the present invention, an electronic device is also provided, comprising: a processor; and a memory for storing processor-executable instructions, wherein the processor is configured to execute instructions to implement the power distribution method of the energy storage battery pack described above.

[0335] Example 4

[0336] According to another aspect of the present invention, a computer-readable storage medium is also provided, which, when the instructions in the computer-readable storage medium are executed by a processor of an electronic device, enables the electronic device to perform the power distribution method of the energy storage battery pack described above.

[0337] The sequence numbers of the above embodiments of the present invention are for descriptive purposes only and do not represent the superiority or inferiority of the embodiments.

[0338] In the above embodiments of the present invention, the descriptions of each embodiment have different focuses. For parts not described in detail in a certain embodiment, please refer to the relevant descriptions of other embodiments.

[0339] In the several embodiments provided in this application, it should be understood that the disclosed technical content can be implemented in other ways. The device embodiments described above are merely illustrative; for example, the division of units can be a logical functional division, and in actual implementation, there may be other division methods. For instance, multiple units or components may be combined or integrated into another system, or some features may be ignored or not executed. Furthermore, the displayed or discussed mutual coupling, direct coupling, or communication connection may be through some interfaces; the indirect coupling or communication connection between units or modules may be electrical or other forms.

[0340] The units described as separate components may or may not be physically separate. The components shown as units may or may not be physical units; that is, they may be located in one place or distributed across multiple units. Some or all of the units can be selected to achieve the purpose of this embodiment according to actual needs.

[0341] Furthermore, the functional units in the various embodiments of the present invention can be integrated into one processing unit, or each unit can exist physically separately, or two or more units can be integrated into one unit. The integrated unit can be implemented in hardware or as a software functional unit.

[0342] If the integrated unit is implemented as a software functional unit and sold or used as an independent product, it can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of the present invention, in essence, or the part that contributes to the prior art, or all or part of the technical solution, can be embodied in the form of a software product. This computer software product is stored in a storage medium and includes several instructions to cause a computer device (which may be a personal computer, server, or network device, etc.) to execute all or part of the steps of the methods described in the various embodiments of the present invention. The aforementioned storage medium includes various media capable of storing program code, such as USB flash drives, read-only memory (ROM), random access memory (RAM), portable hard drives, magnetic disks, or optical disks.

[0343] The above description is only a preferred embodiment of the present invention. It should be noted that for those skilled in the art, several improvements and modifications can be made without departing from the principle of the present invention, and these improvements and modifications should also be considered within the scope of protection of the present invention.

Claims

1. A power distribution method for an energy storage battery pack, characterized in that, include: The performance of each individual cell in the energy storage battery pack is clustered to obtain multiple battery clusters. The individual cells within each of the multiple battery clusters are virtually connected in parallel to obtain equivalent cells corresponding to the multiple battery clusters. The virtual parallel connection is to virtually connect the individual cells within each cluster in parallel through modeling without changing the actual physical connection relationship of the individual cells. Based on the objective optimization function and the operating constraints corresponding to the multiple equivalent batteries, the power configuration of each individual battery in the energy storage battery pack is determined. The objective optimization function is an optimization function that aims to minimize the sum of the power loss, state of charge deviation, and temperature deviation of the multiple equivalent batteries. The state of charge deviation is the deviation between the corresponding equivalent battery and the average state of charge of the multiple equivalent batteries, and the temperature deviation is the deviation between the corresponding temperature and the average temperature of the multiple equivalent batteries. The power of each individual battery cell is allocated according to its corresponding individual cell power configuration.

2. The method according to claim 1, characterized in that, The power loss level of any target equivalent cell among the plurality of equivalent cells is determined using the following method: In the target equivalent cell, the internal resistance difference index between the reference cell and other cells in the cluster is determined, wherein the other cells in the cluster are individual cells in each cluster other than the reference cell. Based on the reference internal resistance corresponding to the reference battery, the individual internal resistances of the other batteries in the cluster, and the internal resistance difference index of the other batteries in the cluster, the equivalent internal resistance of the target equivalent battery is determined. The power loss level of the target equivalent battery is determined based on the equivalent internal resistance and the charging and discharging current of the target equivalent battery.

3. The method according to claim 1, characterized in that, The average state of charge of the plurality of equivalent cells is determined in the following manner: Determine the average remaining capacity and average equivalent capacitance corresponding to the plurality of equivalent batteries; The average state of charge is determined based on the average remaining charge and average equivalent capacitance of the multiple equivalent batteries.

4. The method according to claim 1, characterized in that, The performance of each individual cell in the energy storage battery pack is clustered to obtain multiple battery clusters, including: Determine the individual cell feature vector corresponding to each cell of the energy storage battery pack. The corresponding individual cell feature vector is used to characterize the electrothermal coupling characteristics of the corresponding individual cell from multiple predetermined dimensions, including the state of charge dimension, temperature dimension, and resistance dimension. Determine the coupling coefficients corresponding to the plurality of predetermined dimensions, wherein the corresponding coupling coefficients are used to characterize the importance of the corresponding predetermined dimension to characterizing the electrothermal coupling characteristics of the corresponding single cell. Based on the individual feature vectors corresponding to each cell in the energy storage battery pack, and the coupling coefficients corresponding to the multiple predetermined dimensions, the individual cells in the energy storage battery pack are clustered according to their performance to obtain multiple battery clusters.

5. The method according to claim 4, characterized in that, Based on the individual feature vectors corresponding to each cell in the energy storage battery pack, and the coupling coefficients corresponding to the multiple predetermined dimensions, the individual cells in the energy storage battery pack are clustered according to their performance to obtain multiple battery clusters, including: According to the execution order of multiple iterations in the clustering process, for the corresponding iteration in the multiple iterations, multiple central cells under the corresponding iteration are determined, wherein each individual cell includes the multiple central cells; Based on the individual feature vectors corresponding to each individual cell and the coupling coefficients corresponding to the multiple predetermined dimensions, the difference index between the multiple central cells and each individual cell is determined. Based on the difference index between the multiple central batteries and each individual battery, the clusters under the corresponding iterations are determined. The clusters under the corresponding iterations include individual batteries whose difference index with the corresponding central battery is less than or equal to the difference threshold. This process continues until the target condition is met, resulting in multiple battery clusters. The target condition includes the completion of multiple iterations during the clustering process.

6. The method according to claim 1, characterized in that, The process of determining the individual power configuration of each cell in the energy storage battery pack based on the objective optimization function and the operational constraints corresponding to multiple equivalent batteries includes: Determine the equivalent open-circuit voltage corresponding to each of the plurality of equivalent cells; Based on the objective optimization function, as well as the operating constraints and equivalent open-circuit voltages corresponding to multiple equivalent batteries, the power configuration of each individual cell in the energy storage battery pack is determined.

7. The method according to claim 6, characterized in that, The process of determining the individual power configuration of each cell in the energy storage battery pack based on the objective optimization function, the operating constraints corresponding to multiple equivalent batteries, and the equivalent open-circuit voltage, includes: Based on the objective optimization function and the operational constraints corresponding to the multiple equivalent batteries, the equivalent power configuration corresponding to the multiple equivalent batteries is determined. Based on the equivalent power configuration and single-cell power allocation strategy corresponding to the plurality of equivalent cells, the single-cell power configuration corresponding to each cell of the energy storage battery pack is determined, wherein the single-cell power allocation strategy includes any of the following: A power quota-based allocation strategy, wherein the power quota includes equal power quotas; A power loss-based allocation strategy, wherein the power loss-based allocation strategy is an allocation strategy aimed at minimizing the power loss of the energy storage battery pack; The allocation strategy based on operating characteristics is an allocation strategy aimed at minimizing the state deviation between the operating state and the predetermined state of the energy storage battery pack.

8. The method according to any one of claims 1 to 7, characterized in that, The operational constraints corresponding to the multiple equivalent cells include: current constraints, temperature constraints, state of charge constraints, operating power constraints, and power quota constraints.

9. A power distribution device for an energy storage battery pack, characterized in that, include: The first determining module is used to perform performance clustering on each individual cell of the energy storage battery pack to obtain multiple battery clusters. The second determining module is used to virtually connect the individual cells in each of the multiple battery clusters in parallel to obtain the equivalent cells corresponding to the multiple battery clusters respectively. The virtual parallel connection is to virtually connect the individual cells in each cluster in parallel by modeling without changing the actual physical connection relationship of the individual cells. The third determining module is used to determine the individual power configuration of each cell in the energy storage battery pack based on the objective optimization function and the operating constraints corresponding to the multiple equivalent cells. The objective optimization function is an optimization function that aims to minimize the sum of the power loss degree, state of charge deviation and temperature deviation of the multiple equivalent cells. The state of charge deviation is the deviation between the corresponding equivalent cell and the average state of charge of the multiple equivalent cells. The temperature deviation is the deviation between the corresponding temperature and the average temperature of the multiple equivalent cells. The fourth determining module is used to allocate power to each individual battery cell according to the corresponding individual cell power configuration.

10. An electronic device, characterized in that, include: processor; Memory used to store the processor's executable instructions; The processor is configured to execute the instructions to implement the power distribution method for the energy storage battery pack as described in any one of claims 1 to 8.

Images