Battery energy storage system power distribution method and system, medium and computer equipment

Abstract

The invention belongs to the technical field of battery energy storage systems. The invention provides a power distribution method and system for a battery energy storage system, a medium and computer equipment, when the target power of the energy storage system is not zero and the target power is greater than or equal to the sum of the maximum charging and discharging power of all energy storage units, all the energy storage units operate at the maximum charging and discharging power; when the target power of the energy storage system is not zero and the target power is smaller than the sum of the maximum charging and discharging power of all the energy storage units, determining the number of required energy storage units according to the target power and the maximum charging and discharging power of each energy storage unit, and selecting the energy storage units to be used according to an energy storage unit priority rule; performing global optimization on the target power distribution to obtain an optimal or approximately optimal power distribution value of each energy storage unit to be used; according to the invention, the SOC balance degree of the battery energy storage system is improved, and the risk of over-charging and over-discharging is reduced.

Description

Technical Field

[0001] The present invention relates to the technical field of battery energy storage systems, and in particular to a power distribution method, system, medium and computer device for a battery energy storage system. Background Art

[0002] The statements in this section merely provide background art related to the present invention and do not necessarily constitute prior art.

[0003] With its fast power regulation ability, high-precision control performance and flexible charge and discharge characteristics, the Battery Energy Storage System (BESS) has become an important technical means to solve the problem of new energy power generation volatility. When the output of new energy fluctuates greatly, the battery energy storage system can adjust the power output through rapid response charge and discharge, smooth the fluctuations of new energy power generation, reduce the fluctuation range of the power grid frequency and voltage, thereby improving the consumption capacity of new energy and the stability of the power system. However, the current battery energy storage system still has the following problems:

[0004] (1) The problems of inconsistent state of health and unbalanced state of charge. Due to factors such as minor differences in the battery manufacturing process, different aging speeds during long-term operation, and uneven charge and discharge cycle times, the state of health (SOH) and state of charge (SOC) of each single battery inside the energy storage unit will show inconsistent phenomena. This difference will accumulate and amplify, resulting in problems such as uneven power distribution, difficult thermal management, and reduced system efficiency during charge and discharge of the energy storage unit; (2) The problem of insufficient traditional power distribution strategies. Some existing power distribution strategies only simply consider the SOC of the energy storage unit, or estimate the SOH at the unit level, without deeply considering the differences between the optimal / worst single batteries inside the unit, and also lack real-time and accurate state of health indicators to support, which may lead to overuse of the units with poor state of health, causing them to deteriorate rapidly. Summary of the Invention

[0005] In order to solve the problems of uneven power distribution, aggravated capacity decline and shortened system life caused by inconsistent state of health (SOH) and state of charge (SOC) of single batteries in the battery energy storage system (BESS), the present invention provides a power distribution method, system, medium and computer device for a battery energy storage system, which considers the differences between the optimal and worst batteries inside the unit and the mutual influence of SOH and SOC between units. When distributing power, it can actively reduce the deep use of the units with poor state of health and large internal difference degree, improve the SOC balance degree of the battery energy storage system, and reduce the risk of overcharge and over-discharge.

[0006] To achieve the above object, the present invention adopts the following technical solutions:

[0007] In a first aspect, the present invention provides a method for power distribution of a battery energy storage system.

[0008] A method for power distribution of a battery energy storage system includes the following processes:

[0009] Obtain the state parameters of the battery cells of each energy storage unit, extract key indirect indicators for characterizing the battery health condition according to the state parameters, and determine the battery health state quantization index according to the key indirect indicators;

[0010] When the target power of the energy storage system is not zero and the target power is greater than or equal to the sum of the maximum charge and discharge powers of all energy storage units, all energy storage units operate at the maximum charge and discharge power;

[0011] When the target power of the energy storage system is not zero and the target power is less than the sum of the maximum charge and discharge powers of all energy storage units, determine the number of energy storage units required according to the target power and the maximum charge and discharge powers of each energy storage unit, and select the energy storage units to be used in combination with the battery health state quantization index according to the energy storage unit priority rule;

[0012] Perform global optimization on the target power distribution to obtain the optimal or approximately optimal power distribution values of the energy storage units to be used.

[0013] As a further limitation of the first aspect of the present invention, when the target power of the energy storage system is zero, there is no current power demand, no charge and discharge operation is required, and no power distribution is performed.

[0014] As a further limitation of the first aspect of the present invention, determine the possible number of energy storage units to be put into use according to the comparison between the target power and the maximum charge and discharge powers of each energy storage unit.

[0015] As a further limitation of the first aspect of the present invention, selecting the energy storage units to be used according to the energy storage unit priority rule includes:

[0016] Construct a priority sorting function expression: F1 = af1 + bf2, where a and b are weights, f1 is a sorting judgment function considering only the SOC of the energy storage unit, and f2 is a priority sorting function considering the differences in the health states of the batteries in the energy storage unit, the health states of the optimal and worst single cells, and the inconsistency with the SOC;

[0017] When the state of charge of the energy storage unit exceeds the upper limit of the best cycle interval, F1 takes positive infinity and discharges first;

[0018] When the state of charge of the energy storage unit is lower than the lower limit of the best cycle interval, F1 takes negative infinity and charges first;

[0019] When the state of charge of the energy storage unit is within the optimal cycle range, and when the energy storage unit is discharging, the larger F1 is, the higher the priority for discharge; when the energy storage unit is charging, the smaller F1 is, the higher the priority for charging.

[0020] As a further limitation of the first aspect of the present invention,

[0021] Among them, S OC.max and S OC.min is the maximum and minimum value of the optimal cycle interval, S OC.ref is the optimal cycle midpoint value, S OC.i (t) is the SOC of the i-th energy storage unit at the current moment t.

[0022] As a further limitation of the first aspect of the present invention,

[0023]

[0024] Among them, τ is the charge and discharge judgment flag, which is set to 1 when the energy storage unit participates in the discharge action and is set to -1 when the energy storage unit participates in the charge action. HI，i is the extreme health status of the i-th unit group, ΔS OC，i (t) is the difference in SOC between the worst single cell and the best single cell in the i-th energy storage unit at time t, ΔS OC.lim is the SOC imbalance limit value.

[0025] As a further limitation of the first aspect of the present invention, in order to optimize the power of each energy storage unit, power allocation is performed with the objective function of minimizing the level of inconsistency in the internal state of charge of each energy storage unit, and an AGPSO algorithm is used for iterative solution to obtain the optimal or approximately optimal power allocation value of each energy storage unit to be used.

[0026] In a second aspect, the present invention provides a power distribution system for a battery energy storage system.

[0027] A power distribution system for a battery energy storage system, comprising:

[0028] A quantitative indicator acquisition unit is configured to: acquire state parameters of battery cells of each energy storage unit, extract key indirect indicators for characterizing battery health status based on the state parameters, and determine a quantitative indicator of battery health status based on the key indirect indicators;

[0029] The first processing unit is configured to: when the target power of the energy storage system is not zero and the target power is greater than or equal to the sum of the maximum charge and discharge powers of all energy storage units, operate all energy storage units at the maximum charge and discharge powers;

[0030] A second processing unit, configured to: when the target power of the energy storage system is not zero and the target power is less than the sum of the maximum charge-discharge powers of all energy storage units, determine the required number of energy storage units according to the target power and the maximum charge-discharge powers of each energy storage unit, and select the energy storage units to be used in combination with the battery health status quantification index according to the energy storage unit priority rule;

[0031] A power distribution unit, configured to: perform global optimization on the target power distribution to obtain the optimal or approximately optimal power distribution values of the energy storage units to be used.

[0032] In a third aspect, the present invention provides a computer device, including: a processor and a computer-readable storage medium;

[0033] The processor is adapted to execute a computer program;

[0034] The computer-readable storage medium stores a computer program, and when the computer program is executed by the processor, it implements the battery energy storage system power distribution method as described in the first aspect of the present invention.

[0035] In a fourth aspect, the present invention provides a computer-readable storage medium, which stores a computer program, and the computer program is adapted to be loaded and executed by a processor to implement the battery energy storage system power distribution method as described in the first aspect of the present invention.

[0036] Compared with the prior art, the beneficial effects of the present invention are:

[0037] 1. The present invention innovatively proposes a battery energy storage system power distribution method, which effectively improves the system consistency. Considering the differences between the optimal and worst batteries inside the unit and the mutual influence of SOH and SOC between units, it can actively reduce the deep use of units with poor health status and large internal difference degrees during power distribution, significantly improve the system SOC balance degree, and reduce the risk of overcharge and over-discharge.

[0038] 2. The present invention innovatively proposes a battery energy storage system power distribution method, which improves the accuracy of health status assessment. Through the combination of AGPSO and HMKELM, it can not only obtain a higher prediction accuracy of the monomer health status with less training data, but also adaptively select the appropriate kernel function ratio weight to avoid falling into local optimum.

[0039] 3. The present invention innovatively proposes a battery energy storage system power distribution method, which extends the battery life and improves the economic benefits. Compared with the traditional distribution strategy that only considers the size of SOC or the strategy that does not refine the monomer health differences, the present invention effectively delays the capacity attenuation speed of the energy storage system through refined distribution and differential control, and significantly improves the battery utilization efficiency and economy.

[0040] 4. The invention innovatively proposes a power distribution method for a battery energy storage system, which improves the global convergence speed. The AGPSO algorithm adopts an autonomous group cooperation mechanism, and the particles in the group have higher diversity and autonomy, which can avoid falling into local optima. When applied to the multi-constraint non-linear optimization problem of BESS power distribution, it has a fast calculation speed and good real-time performance.

[0041] Advantages of additional aspects of the present invention will be partly given in the following description, partly will become obvious from the following description, or will be understood through the practice of the present invention. BRIEF DESCRIPTION OF THE DRAWINGS

[0042] The accompanying drawings forming a part of this specification are used to provide a further understanding of the present invention. The schematic embodiments and descriptions thereof of the present invention are used to explain the present invention and do not constitute an improper limitation to the present invention.

[0043] Figure 1 The battery capacity degradation zone line provided for an exemplary embodiment of the present invention;

[0044] Figure 2 The typical grid-connected system structure provided for an exemplary embodiment of the present invention;

[0045] Figure 3 The f1 function value change zone line under different SOCs provided for an exemplary embodiment of the present invention;

[0046] Figure 4 The calculation example solution flow chart provided for an exemplary embodiment of the present invention;

[0047] Figure 5 The schematic diagram of the power distribution system of the battery energy storage system provided for an exemplary embodiment of the present invention;

[0048] Figure 6 The schematic diagram of the computer device provided for an exemplary embodiment of the present invention. DETAILED DESCRIPTION OF THE EMBODIMENTS

[0049] The present invention will be further described below in conjunction with the accompanying drawings and embodiments.

[0050] It should be noted that the following detailed description is exemplary and is intended to provide further explanation of the present invention. Unless otherwise specified, all technical and scientific terms used herein have the same meaning as commonly understood by those of ordinary skill in the technical field to which the present invention belongs.

[0051] The present invention proposes a power distribution method for a battery energy storage system. First, the technical terms and related concepts involved in this solution will be briefly introduced, where:

[0052] The Battery Energy Storage System (BESS) is a system that uses batteries to store electrical energy. It can store electrical energy when the power demand is low and release electrical energy when the power demand is high, in order to improve the stability and reliability of the power system. It is usually composed of battery modules, a Battery Management System (BMS), a Power Conversion System (PCS), an Energy Management System (EMS), and other auxiliary equipment. Several key technologies of the battery energy storage system: The Battery Management System (BMS) is responsible for monitoring parameters such as the voltage, current, and temperature of the battery, preventing overcharging, over-discharging, overheating, etc., and improving the battery life and safety. The Energy Management System (EMS) optimizes the operation of the energy storage system through intelligent algorithms, coordinates the charge and discharge strategies, improves the economy and reliability, and realizes intelligent scheduling. The Power Conversion System (PCS) is responsible for the conversion between direct current (DC) and alternating current (AC), including functions such as inversion and rectification, to ensure the compatibility and stable operation of the power grid and the energy storage system.

[0053] The Autonomous Groups Particle Swarm Optimization (AGPSO) is an improved Particle Swarm Optimization (PSO) algorithm. It divides the particle swarm into multiple autonomous sub-groups (Autonomous Groups) and introduces different search strategies within and between each sub-group to improve the global search ability and local search accuracy of the algorithm. Several key characteristics of the autonomous groups particle swarm algorithm are as follows:

[0054] Multi-group collaborative search: Divide the overall particle swarm into multiple autonomous sub-groups, each sub-group evolves independently, and information exchange is carried out through a certain mechanism to enhance population diversity and avoid premature convergence;

[0055] Adaptive parameter adjustment: Different sub-groups can adopt different parameter update rules, inertia weights, acceleration coefficients, etc. to achieve a dynamic balance between global exploration and local exploitation;

[0056] Personalized search strategy: Each sub-group can adopt different optimization strategies based on its evolutionary state, such as introducing mutation operations, dynamically adjusting the search space, etc., to improve the adaptability of the algorithm;

[0057] High convergence speed and high accuracy: Through the autonomous mechanism and multi-group collaboration, the algorithm is more likely to jump out of the local optimum in complex optimization problems and improve the global convergence performance.

[0058] The Hybrid Multi-Kernel Extreme Learning Machine (HMKELM) is an improved Extreme Learning Machine (ELM). By introducing the Multi-Kernel Learning (MKL) method, it combines multiple kernel functions (such as linear kernel, Gaussian kernel, polynomial kernel, etc.) to enhance the learning ability and generalization performance of the model. This method is suitable for dealing with non-linear, high-dimensional, and large-scale data problems. Several key characteristics of the hybrid multi-kernel extreme learning machine specifically include:

[0059] Hybrid Multi-Kernel: Combining the advantages of multiple kernel functions, adaptively allocating the weights of different kernel functions to improve the non-linear mapping ability;

[0060] ELM Characteristics: ELM significantly improves the training speed by randomly initializing the hidden layer parameters and using the least squares method to solve the output weights;

[0061] High Generalization Ability: Through the hybrid multi-kernel strategy, it improves the adaptability of the model on different types of datasets and avoids the limitations of a single kernel function;

[0062] High Computational Efficiency: Compared with traditional neural networks or kernel methods, ELM combined with multi-kernel learning can still maintain a low computational complexity and is suitable for large-scale data processing;

[0063] No Need to Tune Hidden Layer Weights: ELM only needs to optimize the output weights, avoiding the gradient descent process in neural network training and improving the convergence speed.

[0064] The State of Health (SOH) of a battery is an indicator that measures the current performance of the battery relative to its initial performance, usually expressed as a percentage (%). SOH reflects the degree of battery degradation and directly affects key characteristics such as its available capacity, internal resistance, and power performance. It is an important monitoring parameter in applications such as energy storage systems, electric vehicles, and power grid frequency modulation. Several key characteristics of the state of health of the battery specifically include:

[0065] Battery Performance Evaluation Index: SOH is mainly used to measure the degree of battery aging and determines the service life and usability of the battery;

[0066] Influence of Multiple Factors: The SOH of the battery is affected by factors such as the number of charge-discharge cycles, temperature, depth of discharge (DOD), and discharge rate;

[0067] Nonlinear Decay: The state of health of the battery usually goes through three stages: slow decay - accelerated decay - failure, and the decay rate is affected by the usage environment and strategy;

[0068] Evaluation, Prediction, and Management: SOH estimation can be used to evaluate and predict the remaining useful life (RUL) of batteries, optimize charge and discharge strategies, and improve battery utilization efficiency;

[0069] Different Technical Paths: Different types of batteries (such as lithium-ion batteries, sodium-ion batteries, lead-acid batteries, etc.) have different SOH degradation mechanisms, and targeted modeling and monitoring are required.

[0070] Power Allocation refers to the rational distribution of limited power resources to different subsystems or components according to different requirements and constraints in scenarios such as multi-input multi-output (MIMO) communication systems, power systems, wireless networks, battery energy storage systems (BESS), etc., in order to optimize performance, improve efficiency, and reduce losses. Several key characteristics of power allocation specifically include:

[0071] Optimize Resource Utilization: Reasonably allocate power to maximize the energy utilization rate of the system and reduce unnecessary power consumption and losses;

[0072] Dynamic Adjustment: Adjust the power allocation strategy according to environmental changes (such as load demand, channel state, battery SOH / SOC level) to ensure the stable operation of the system;

[0073] Multi-Objective Optimization: Weigh factors such as performance, power consumption, safety, service life, etc., and formulate an optimal power allocation plan;

[0074] Adaptive Regulation: Combine AI, machine learning, or intelligent algorithms (such as reinforcement learning, genetic algorithms) to achieve real-time optimization and improve the robustness of the system;

[0075] Wide Range of Application Scenarios: Cover multiple fields such as wireless communication, power systems, battery energy storage (BESS), electric vehicles, and microgrids.

[0076] On the one hand, the present invention proposes a battery health state evaluation model (AGPSO-HMKELM) based on AGPSO and HMKELM, specifically including:

[0077] First, collect historical or real-time charge and discharge data of each single battery in the battery energy storage system, and extract and preprocess various indirect indicators with a relatively high degree of relevance to the health state during the charge and discharge process (such as constant current charging time, average discharge voltage, temperature peak time, equal voltage drop discharge time, etc.) to construct a corresponding feature dataset;

[0078] Secondly, based on the extracted feature dataset, an HMKELM model is constructed. This model combines local kernels (such as the RBF kernel) and global kernels (such as the Polynomial kernel, Sigmoid kernel, etc.) according to certain weights, so as to take into account the local and global non-linear mapping capabilities and achieve a more accurate fitting of the non-linear decay characteristics of the battery health state.

[0079] Thirdly, AGPSO is used to optimize the key parameters of the HMKELM model (including the width parameter of the kernel function, the regularization coefficient, and the multi-kernel weights, etc.). Through the cooperation and adaptive mechanism of the autonomous population in AGPSO, the optimal parameter combination of the HMKELM model can be efficiently found in a large parameter search space, thereby improving the fitting accuracy and stability of the model for the battery health state.

[0080] Then, after completing the training and parameter optimization of the HMKELM model, for each battery cell, input the charge and discharge characteristic indicators of its latest period of time. The battery health state quantification index (CHI) is calculated through the optimized HMKELM model, and further, according to the CHI range between the optimal and the worst batteries inside the unit, the degree of health state inconsistency inside the unit is measured.

[0081] Finally, store or transfer the above battery health state quantification results (CHI) to the subsequent scheduling and control links, providing reference and input for the power distribution strategy based on the health state. Through this evaluation model, the states of each battery cell in the system can be monitored in real time or periodically, helping to detect and suppress the development of inconsistencies in a timely manner and delaying the overall decay of the energy storage system.

[0082] In the present invention, long-term cyclic charge and discharge data of lithium-ion batteries with model numbers B0005, B0006, and B0007 (from a battery database with usage authorization) are collected in the laboratory or on-site, including the capacity decay conditions under different operating cycles. Indirect health indicators such as the average discharge voltage ADV, constant current charging time CCT, peak temperature time PPT, equal voltage drop discharge time VFT, etc. are extracted and labeled with the known SOH (obtained from the discharge capacity / rated capacity) to construct a training sample set. The battery capacity degradation zone line is as Figure 1 shown. As the charge and discharge cycle times increase, these lithium batteries have experienced different degrees of performance decline, but it is not a monotonic downward trend. It is found that during the capacity decay process of lithium batteries, the phenomenon of capacity regeneration occurs in the batteries, which leads to a non-linear relationship between the battery decay and the charge and discharge cycle times.

[0083] Extreme learning machine is improved from a single-hidden-layer feedforward neural network. Its structure is simpler and more efficient than that of traditional neural networks, and the algorithm performance is improved. To further improve the learning and generalization ability of the extreme learning machine, the idea of kernel function is introduced to improve a kernel extreme learning machine. The kernel function has a strong non-linear mapping ability, making the learning ability of the kernel extreme learning machine (KELM) more stable and the generalization ability improved. Therefore, the correct selection of the kernel function can effectively improve the performance of KELM. Commonly used kernel functions include radial basis function (RBF), polynomial kernel function (Poly), linear kernel function, and two-layer neural network kernel function (sigmoid), etc. They can be divided into local kernels and global kernels. Local kernels only have locality, and global kernels only have globality. This results in that no single kernel function can simultaneously have local and global characteristics. To obtain a kernel function with both local and global characteristics, a multi-kernel mixture function is adopted. As a typical local kernel function, RBF has strong learning ability but weak generalization ability. Poly is a global kernel function with weak learning ability but strong generalization ability. Sigmoid can be used to process specific data. The attenuation process of battery health is non-linear due to various factors and can usually be described by the Sigmoid function model. The Sigmoid function increases the non-linear factors of the KELM model to make the network more powerful, process complex data, and represent the complex non-linear function mapping between input and output. Therefore, to further improve the learning and generalization ability of KELM, the RBF, Poly, and sigmoid kernel functions are combined into a mixed multi-kernel function with different weights, and the expressions are (1) and (2):

[0084]

[0085] Among them, x represents the input sample vector, x i represents the i-th sample vector, c represents the constant term in the polynomial kernel function, which is used to adjust the characteristics of the polynomial kernel, d represents the order of the polynomial kernel function, which determines the highest degree of the polynomial, σ represents the width parameter of the radial basis function (RBF) kernel, which controls the local influence range of the RBF kernel, α represents the proportional parameter of the Sigmoid kernel function, which affects the shape of the Sigmoid function, and w1, w2, and w3 respectively represent the weights of the polynomial kernel, radial basis function kernel, and Sigmoid kernel in the mixed kernel function, which are used to adjust the contribution degree of each kernel function in the mixed kernel.

[0086] According to the kernel matrix defined by the Mercer condition, we have (3)-(5):

[0087] Ω ELM =HH T =h(x)h(x i )=K(x, xi ) (3);

[0088]

[0089] f HMKELM (x) = h(x) * β = H * β (5);

[0090] Wherein, Ω ELM represents the kernel matrix related to the Extreme Learning Machine (ELM), which is used to describe the similarity between input samples, H represents the output matrix of the hidden layer of the ELM, h(x) represents the output vector of the input vector x in the hidden layer, h(x i ) represents the output vector of the input vector x i in the hidden layer, K(x, x i ( represents the kernel function, which represents the similarity measure between the input vector x and the input vector x i ), β represents the output weight vector of the ELM model, which is used to map the hidden layer output to the output layer, E represents the identity matrix, C represents the regularization coefficient, which is used to control the complexity of the model and prevent overfitting, T represents the target output vector, f HMKELM (x) represents the output function of the Hybrid Kernel Extreme Learning Machine (HMKELM) model, which is used to predict the output of the input vector x.

[0091] From equations (1)-(5), the final output of HMKELM is:[[]]

[0092]

[0093] Wherein, K(x, x1), K(x, x2) … K(x, x n ) represent the kernel function values between the input vector x and each vector x1, x2 and x n in the training sample set, Ω ELM represents the kernel matrix related to the ELM, and C represents the regularization coefficient.

[0094] There are many parameters in the HMKELM model. The optimal values of the parameter settings determine the accuracy of the HMKELM model for fitting health indicators. Using experimental methods such as cross-validation is inefficient and has low accuracy. Therefore, AGPSO is used to optimize the values of the regularization coefficient C, the width parameter σ of the RBF kernel, the kernel parameters c and d of the Poly kernel, the kernel function ratio parameter α of the sigmoid kernel, and the hybrid weights of the multi-kernel functions in the HMKELM model to find the optimal parameters suitable for evaluating the battery health state.

[0095] On the other hand, the present invention proposes a BESS power distribution method based on AGPSO. Specifically, it includes the following process:[[]]

[0096] First, based on the power imbalance demand on the grid side or the microgrid side (such as the difference between the output of wind power and photovoltaic power and the load), and the actual operation requirements of the energy storage system, determine the target charge-discharge power that needs to be borne by the energy storage system within this scheduling period;

[0097] Secondly, consider the safe operation boundaries and constraints of the energy storage system itself, including the upper and lower limits of the SOC of each energy storage unit, the rated power limit, the minimum power requirement, and the additional safety limits brought by the degradation of the health state, and establish a unified power distribution constraint model;

[0098] Thirdly, based on the battery health state quantification results obtained in the first aspect (i.e., the worst battery CHI inside the unit and the CHI range difference between monomers), and combined with the SOC levels and health degradation conditions of each unit, construct a set of priority ranking rules for the energy storage units. For units with poor health states or large internal CHI differences, moderately reduce their opportunities for deep charge and discharge; for units with good health states and small internal differences, allow them to bear more charge-discharge power, which can effectively inhibit the further deterioration of the inconsistency inside the energy storage units;

[0099] Then, use the Autonomous Group Particle Swarm Optimization (AGPSO) to globally optimize the actual power distribution of each energy storage unit under the above priority rules and power constraints. AGPSO enables the particle swarm to maintain diversity, avoid falling into local optima, and quickly converge to the global optimum or approximate optimum solution through the autonomy and co-evolution mechanism of the group, thereby realizing the real-time or periodic update of the power distribution scheme for the multi-unit BESS;

[0100] Finally, issue the power scheduling instructions output by AGPSO to the control layer of each energy storage unit, and execute the optimized power charge and discharge distribution in each scheduling period. At the same time, monitor the changes in the health state and SOC level of the key battery units in real time, and continue to update the health state evaluation results and power distribution strategies when the next scheduling period arrives, forming a collaborative closed-loop control of the health state and power distribution to extend the overall service life of the energy storage system and maintain good consistency and performance of the system.

[0101] Establish a typical grid-connected system with a common DC bus as shown in Figure 2 As shown in the figure, the arrow direction is the positive direction of power transmission. After the wind power, photovoltaic power, conventional load, and electric vehicle load are gathered at the DC bus, the total power balance of the grid-connected system can be obtained as:

[0102] P w (t)+P p (t)=P L (t)+P E (t)+P B (t)(7);

[0103] In the formula: P w (t), P p (t), P L (t), P E (t) are the powers of wind power, photovoltaic power, conventional load, and electric vehicle load at a certain moment respectively. By absorbing the unbalanced power, the power fluctuation of the DC bus can be suppressed to meet the grid connection requirements. According to the coordination control technical index regulations including the energy storage system, the power fluctuation rate of the grid connection point within 30 minutes (the ratio of the power fluctuation amplitude of the grid connection point to the total installed capacity of new energy in the microgrid) should not be greater than 7%. The unbalanced power P*(t) of new energy grid connection needs to meet the following limiting conditions:

[0104] P * (t) ≤ (P w (t) + P P (t)) × 7% (8);

[0105] If the unbalanced power of the DC bus does not meet the above requirements, the energy storage system is used to suppress it. The difference between the smoothed new energy grid-connected power and the load is the power P B (t) that needs to be absorbed by the energy storage system after suppression, which is:

[0106] P B (t) = P * (t) - P L (t) - P E (t) (9);

[0107] Among them, the power P * (t) that meets the grid connection requirements can be obtained by smoothing with the recursive average filtering algorithm:

[0108]

[0109] In the formula: N is the number of sliding windows, generally an odd number. The larger N is, the better the smoothing effect, but it will cause the power borne by the energy storage system to increase, make more requirements on the performance and scale of the energy storage system, and cause waste of energy storage resources.

[0110] The present invention constructs the constraint conditions of the battery energy storage system. Specifically, it includes:

[0111] SOC constraint of the energy storage unit:

[0112] S OC.min ≤ S OC.i ≤ S OC.max (11);

[0113] The charge-discharge power constraint of the energy storage unit: The energy storage unit constrains the charge-discharge power according to different safety states, and appropriately increases the minimum charge-discharge power of the energy storage unit to prevent the reduction of the efficiency of the energy storage unit and related electrical equipment:

[0114] P d.min ≤P i ≤t d.max (12);

[0115] t c.min ≤P i ≤P c.max (13);

[0116] Total power constraint of the energy storage system:

[0117]

[0118] Change of state of charge during discharge:

[0119]

[0120] Change of state of charge during charge:

[0121]

[0122] Generally, the charge-discharge efficiency can be simplified as:

[0123]

[0124] Remaining capacity:

[0125] C re =C ni -C i.loss (18);

[0126]

[0127] The present invention constructs a charge-discharge quantity constraint for the energy storage unit. The number of energy storage units participating in charge and discharge should meet the actual power demand considering power loss, that is:

[0128]

[0129] If the actual power demand is greater than the sum of the powers of all energy storage units, it is required that all energy storage units be put into operation.

[0130] The present invention constructs the charging and discharging priority constraints of the energy storage unit. For a battery energy storage unit integrated with multiple single cells, the charging and discharging performance of the unit depends on the single cell with the worst internal health state. Moreover, the attenuation rate of the single cell with a poor health state is faster than that of the good single cell, which will lead to a further increase in the health state gap within the unit. When distributing each battery within the unit group, the average distribution method is often adopted. However, due to the inconsistency of the internal health state, there will be differences in the state of charge, resulting in differences in the charging level and discharging depth of each single cell, further increasing the difference in the health state between single cells, forming a vicious cycle, and then leading to the imbalance of the internal state of charge and affecting the charging and discharging performance of the energy storage unit. Therefore, maintaining the charging and discharging performance of the energy storage unit is not only related to the single cell with the worst internal health state but also depends on the difference between the best and the worst single cells. Therefore, when making the distribution, the use of energy storage units with a large difference in the health state between single cells and those with a low health state should be reduced.

[0131] The internal SOC imbalance of the energy storage unit is mainly peak imbalance and valley imbalance. In peak imbalance, the SOC of the worst battery is too high. If continuous charging occurs, the SOC will exceed the limit or directly cause overcharging of the battery, resulting in a decrease in the power level of the unit. However, discharging can reduce the degree of imbalance and is beneficial for the energy storage unit to maintain good regulation ability. Valley imbalance is the opposite, which is not conducive to continuous discharging but can be charged. If the imbalance is not controlled, it will lead to accelerated deep charging and discharging of the worst single cell and its aging. Therefore, in order to reduce the degree of SOC imbalance within the group, the charging and discharging should be moderately controlled. When there is peak imbalance, the discharging priority of the unit should be appropriately increased, and when there is valley imbalance, it should also be appropriately increased.

[0132] The present invention comprehensively considers the influence of the SOC of the energy storage unit, the health state, and the difference in the health state of the internal batteries of the energy storage unit on the priority ranking of the unit. ΔCHI.i is the difference in the health state between the best single cell battery and the worst single cell battery within the i-th energy storage unit. The battery health state of the energy storage unit is restricted by the worst single cell battery inside it during actual use. Therefore, the CHI of the worst battery of the energy storage unit is used as the CHI of the energy storage unit. Therefore, the priority ranking function expression is F1, and F1 is obtained by weighting the functions f1 and f2. Among them, a and b are the weight values of the two strategies. When the unit discharges, the larger F1 is, the more preferentially it discharges. When the unit charges, the smaller F1 is, the more preferentially it charges. When the SOC of the energy storage unit exceeds the upper limit of the best cycling interval, F1 takes positive infinity and discharges first. When it is lower than the lower limit of the best cycling interval, F1 takes negative infinity and charges first. Within the best cycling interval, it is selected according to the following priority rules:

[0133] F1 = af1 + bf2 (22);

[0134]

[0135] Among them, f1 is the ranking judgment function that only considers the SOC of the energy storage unit. The relationship between the two functions is as follows: Figure 3 As shown, the original function is the inverse hyperbolic tangent function, and the domain of this function is set to [S OC.max , S OC.min ], where S OC.max =0.8, S OC.min =0.2; when the energy storage unit SOC is at the optimal cycle midpoint, that is, S OC.ref = 0.5, the energy storage unit will not cause the energy storage system SOC to be too high or too low regardless of charging or discharging. At this time, the function value is 0, and the impact on priority sorting is small. As the energy storage unit charges and discharges, the SOC will move away from the optimal cycle midpoint, and the function value will increase or decrease rapidly, increasing the impact on priority sorting. This reduces the number of times that energy storage units with too high SOC participate in charging actions and the number of times that energy storage units with too low SOC participate in discharging actions.

[0136]

[0137] Where: τ is the charge and discharge judgment flag, which is set to 1 when the cell participates in the discharge action and -1 when the cell participates in the charge action:

[0138]

[0139] Among them, f2 is the priority ranking function that takes into account the health status of each battery in the energy storage unit, the difference between the health status of the best and worst cells, and the inconsistency of SOC. Among them, ΔC HI，i is the extreme health status of the i-th unit group, ΔS OC，i (t) is the difference in SOC between the worst single cell and the best single cell in the i-th energy storage unit at time t, ΔS OC.lim is the SOC imbalance limit value. f2 is determined by the energy storage unit C HI，i and ΔC HI，i The composition represents the changes in the SOC inconsistency within the unit caused by the degradation of the battery health state after the energy storage system has been running for a period of time. f2 determines the charging and discharging priority order of the energy storage unit based on the three influencing factors of the internal health state of the energy storage unit, the inconsistency of the health state, and the inconsistency of the SOC.

[0140] When the energy storage unit is discharged (τ=1), when ΔS oC，i (t)<0 and |ΔS oC，i (t)|>ΔS OC.lim When the SOC difference inside the energy storage unit is greater than the set threshold ΔS OC.lim , the unit is in an unbalanced state. At this time, the primary goal is to reduce the imbalance and increase |ΔS oc，i(t) | The discharging priority of the larger unit is to reduce the SOC inconsistency within the energy storage unit. When ΔS OC，i (t) < 0 and |ΔS OC，i (t) | ≤ Δs OC.lim , discharging will reduce the SOC inconsistency within the unit and no imbalance occurs. At this time, f2 is ineffective. When ΔS OC，i (t) ≥ 0, unit discharging will exacerbate the SOC inconsistency within the energy storage unit. Therefore, regardless of whether the energy storage unit is in an imbalanced state, C HI，i with poor performance and ΔC H1，i with a large value should be avoided from discharging.

[0141] When the energy storage unit is in the charging state (τ = -1), when ΔS OC，i (t) > 0 and |ΔS oC，i (t) | > ΔS OC.lim , the SOC difference within the energy storage unit is greater than the set threshold ΔS OC.lim . Then the unit is in an imbalanced state. At this time, with the primary goal of reducing the imbalance, the charging priority of the unit with a larger |ΔS oc，i (t) | is increased to reduce the SOC inconsistency within the energy storage unit. When ΔS OC，i (t) > 0 and |ΔS OC，i (t) | ≤ ΔS oC.lim , charging will reduce the SOC inconsistency within the unit and no imbalance occurs. At this time, f2 is ineffective. When ΔS OC，i (t) ≤ 0, unit charging will exacerbate the SOC inconsistency within the energy storage unit. Therefore, regardless of whether the energy storage unit is imbalanced, C HI，i with poor performance and ΔC HI，i with a large value should be avoided from participating in the charging operation.

[0142] The present invention constructs a power distribution strategy for the energy storage unit. After sorting the priorities of the energy storage units and selecting the units participating in the power regulation operation according to the power demand, power distribution needs to be performed on the finally selected unit combination. Since the SOC inconsistency of the batteries within the energy storage unit cannot be completely eliminated and only has an impact on the operation of the energy storage unit when it is large enough, power distribution is performed with the objective of optimizing the power of each energy storage unit to minimize the level of the state of charge inconsistency ΔS OC，i within each energy storage unit. represents that when the target power of the unit is evenly distributed to each cell, for an energy storage unit with a rated capacity of E N [[ID= forty-six]] performing charging and discharging operations at power P B，i (t) in the t-th scheduling period with an operation time of Δt, it will cause the C HI of the optimal single-cell battery SOC and CHI Difference of the worst single-cell SOC:

[0143]

[0144] where: S oH，i is the battery health state of the i-th energy storage unit, and ΔS oH，i is the battery health state inconsistency level of the i-th energy storage unit, and ΔS oC，i (t - 1) is the value of ΔS in the (t - 1)-th scheduling period. When the ΔS of the energy storage unit is OC，i small, the function value is small; when ΔS OC，i gradually increases, the function value increases. By optimizing the allocated power of each energy storage unit through the AGPSO algorithm, the proportion of the target power borne by the energy storage unit with high ΔS OC，i is restricted, and the further deterioration of the inconsistency of the energy storage unit is slowed down. Through formula (28), the state changes caused by the inconsistency of the battery health state during the charge and discharge process of the energy storage unit can be more accurately described, providing a basis for subsequent power distribution and operation control of the energy storage unit. OC，i As shown in

[0145] For example Figure 4 shown, the present invention gives an example of a specific power distribution scheme, including the following steps:

[0146] Step 1: Obtain the battery state parameters of each energy storage unit, and read the state information of the internal single cells of each energy storage unit from the battery management system (BMS) or other monitoring devices, including basic data such as the current operating temperature, voltage, current, and capacity.

[0147] Step 2: Extract the battery health characteristic indicators, preprocess the above battery state parameters, and extract the key indirect indicators (such as constant current charging time, average discharge voltage, temperature peak time, equal voltage drop discharge time, etc.) used to characterize the battery health status to form the health characteristic input data.

[0148] Step 3: Evaluate the battery health state based on the AGPSO-HMKELM model. Send the health characteristic input data into the hybrid multi-core extreme learning machine (HMKELM) model optimized by the autonomous group particle swarm algorithm (AGPSO) to obtain the health state quantization index CHI of each battery inside each energy storage unit. The overall health state of the energy storage unit (such as the worst battery CHI) and information such as the maximum difference between the single-cell CHIs can be further summarized.

[0149] Step 4: Obtain the target power PB of the energy storage system. According to the power imbalance demand of the external power grid (for example, the suppression amount of the difference between the output of wind power and photovoltaic power and the load), determine the target charge / discharge power PB that the energy storage system needs to bear in this scheduling period.

[0150] Step 5: Determine whether the target power PB is 0.

[0151] If PB = 0, it means there is no current power demand, and no charge and discharge operation is required. The process can directly enter the end step.

[0152] If PB ≠ 0, enter Step 6.

[0153] Step 6: Obtain the key parameters of the energy storage unit. Combining the foregoing evaluation results and the basic operation constraints of each energy storage unit, obtain the unit's: quantified health status index (CHI), current SOC (state of charge), and maximum charge and discharge power limit (pmax).

[0154] Step 7: Compare PB with the sum of the maximum charge and discharge powers of all energy storage units and judge:

[0155]

[0156] If it holds, it means the target power to be satisfied is greater than or equal to the maximum output (or absorption) power of all energy storage units. Then all energy storage units need to operate at the maximum power and enter Step 8; if it does not hold, enter Step 9.

[0157] Step 8: All energy storage units operate at the maximum charge / discharge power. When the target power exceeds (or equals) the total available power, instruct all energy storage units to participate in the operation according to the maximum charge and discharge capacity, and output the final power distribution scheme. The process ends.

[0158] Step 9: Determine the required number of units according to the target power and the maximum charge and discharge power. When PB is within the total available power, it is not necessary for all units to participate at full power. First, based on the comparison between PB and the pmax of each unit, initially determine the range of the number of energy storage units that may need to be put into use.

[0159] Step 10: Select the units to be used according to the energy storage unit priority rule. Based on the health status CHI obtained in Step 3 and the SOC of each unit, combined with the priority function (such as restricting the deep use of units with poor health status and high internal inconsistency, and giving more usage opportunities to units with good health status and high internal consistency), screen out the set of energy storage units to be put into operation in this scheduling.

[0160] Step 11: Use the AGPSO algorithm to globally optimize the target power distribution. Model the power distribution problem of the selected energy storage units (with the goal of minimizing the internal state of charge inconsistency level ΔS of each energy storage unit under the premise of meeting the SOC upper and lower limits and health status restrictions) as an optimization objective, and use the AGPSO algorithm for iterative solution to obtain the optimal or approximately optimal power distribution values of each unit. OC，i to be minimized as the objective function)

[0161] Step 12: Output the power allocation results of each energy storage unit and issue the power instructions obtained by the AGPSO algorithm to each energy storage unit to complete the power allocation of this scheduling cycle. The process ends here.

[0162] Figure 5 A power distribution system for a battery energy storage system provided by an exemplary embodiment of the present invention is shown, comprising:

[0163] The quantitative indicator acquisition unit 201 is configured to: acquire state parameters of battery cells of each energy storage unit, extract key indirect indicators for characterizing battery health status based on the state parameters, and determine a quantitative indicator of battery health status based on the key indirect indicators;

[0164] The first processing unit 202 is configured to: when the target power of the energy storage system is not zero and the target power is greater than or equal to the sum of the maximum charge and discharge powers of all energy storage units, all energy storage units operate at the maximum charge and discharge power;

[0165] The second processing unit 203 is configured to: when the target power of the energy storage system is not zero and the target power is less than the sum of the maximum charge and discharge powers of all energy storage units, determine the required number of energy storage units based on the target power and the maximum charge and discharge powers of each energy storage unit, and select the energy storage unit to be used according to the energy storage unit priority rule in combination with the battery health status quantitative index;

[0166] The power distribution unit 204 is configured to perform global optimization on the target power distribution to obtain an optimal or approximately optimal power distribution value for each energy storage unit to be used.

[0167] It is understandable that each of the above-mentioned units can be separately or entirely combined into one or several other units to constitute, or a certain unit (or units) thereof can also be further split into multiple smaller units in function to constitute, which can achieve the same operation without affecting the realization of the technical effects of the embodiments of the present application. The above-mentioned units are divided based on logical functions. In practical applications, the function of a unit can also be realized by multiple units, or the function of multiple units can be realized by one unit. In other embodiments of the present application, the system can also include other units. In practical applications, these functions can also be implemented with the assistance of other units, and can be implemented by the collaboration of multiple units.

[0168] According to another embodiment of the present application, the system described in this embodiment can be constructed by running a computer program (including program code) that can execute the respective steps involved in the corresponding method described in Embodiment 1 on a general computing device such as a computer including processing elements and storage elements such as a Central Processing Unit (CPU), a Random Access Memory (RAM), and a Read-Only Memory (ROM), and the power distribution method of the battery energy storage system of the present application can be implemented. The computer program can be recorded on a computer-readable recording medium, loaded into the above computing device through the computer-readable recording medium, and run therein.

[0169] Figure 6 Fig. 4 shows an electronic device provided by an exemplary embodiment of the present invention. The electronic device includes a processor 301, a communication interface 302, and a computer-readable storage medium 303. Among them, the processor 301, the communication interface 302, and the computer-readable storage medium 303 can be connected through a bus or other means.

[0170] Among them, the communication interface 302 is used to receive and send data. The computer-readable storage medium 303 can be stored in the memory of the electronic device. The computer-readable storage medium 303 is used to store a computer program. The computer program includes program instructions. The processor 301 is used to execute the program instructions stored in the computer-readable storage medium 303.

[0171] The processor 301 (or CPU (Central Processing Unit, central processor)) is the computing core and control core of the electronic device, and is adapted to implement one or more instructions, specifically adapted to load and execute one or more instructions to implement the corresponding method flow or corresponding function.

[0172] The processor 301 is configured to execute the following process:

[0173] Obtain the state parameters of the battery cells of each energy storage unit, extract key indirect indicators for characterizing the battery health status according to the state parameters, and determine the battery health status quantization index according to the key indirect indicators;

[0174] When the target power of the energy storage system is not zero and the target power is greater than or equal to the sum of the maximum charge and discharge powers of all energy storage units, all energy storage units operate at the maximum charge and discharge power;

[0175] When the target power of the energy storage system is not zero and the target power is less than the sum of the maximum charge-discharge powers of all energy storage units, determine the number of required energy storage units according to the target power and the maximum charge-discharge powers of each energy storage unit, combine the battery health status quantification index, and select the energy storage units to be used according to the energy storage unit priority rule;

[0176] Perform global optimization on the target power distribution to obtain the optimal or approximately optimal power distribution values of each energy storage unit to be used.

[0177] The present invention also provides a computer-readable storage medium (Memory). The computer-readable storage medium is a memory device in an electronic device and is used to store programs and data. It can be understood that the computer-readable storage medium here can include both the built-in storage medium in the electronic device and, of course, the extended storage medium supported by the electronic device. The computer-readable storage medium provides a storage space, and this storage space stores the processing system of the electronic device.

[0178] Moreover, one or more instructions suitable for being loaded and executed by the processor are also stored in this storage space. These instructions can be one or more computer programs (including program codes). It should be noted that the computer-readable storage medium here can be a high-speed RAM memory or a non-volatile memory, such as at least one disk memory; optionally, it can also be at least one computer-readable storage medium located far from the aforementioned processor.

[0179] In one embodiment, one or more instructions are stored in the computer-readable storage medium; the processor loads and executes one or more instructions stored in the computer-readable storage medium to implement the following process:

[0180] Obtain the state parameters of the battery cells of each energy storage unit, extract the key indirect indicators for characterizing the battery health status according to the state parameters, and determine the battery health status quantification index according to the key indirect indicators;

[0181] When the target power of the energy storage system is not zero and the target power is greater than or equal to the sum of the maximum charge-discharge powers of all energy storage units, all energy storage units operate at the maximum charge-discharge power;

[0182] When the target power of the energy storage system is not zero and the target power is less than the sum of the maximum charge-discharge powers of all energy storage units, determine the number of required energy storage units according to the target power and the maximum charge-discharge powers of each energy storage unit, combine the battery health status quantification index, and select the energy storage units to be used according to the energy storage unit priority rule;

[0183] Perform global optimization on the target power distribution to obtain the optimal or approximately optimal power distribution values of each energy storage unit to be used.

[0184] The present invention also provides a computer program product or a computer program. The computer program product or the computer program includes computer instructions, and the computer instructions are stored in a computer-readable storage medium. A processor of an electronic device reads the computer instructions from the computer-readable storage medium, and the processor executes the computer instructions to enable the electronic device to perform the following processes:

[0185] Obtain the state parameters of each battery cell of each energy storage unit, extract key indirect indicators for characterizing the battery health status according to the state parameters, and determine the battery health status quantification index according to the key indirect indicators;

[0186] When the target power of the energy storage system is not zero and the target power is greater than or equal to the sum of the maximum charge and discharge powers of all energy storage units, all energy storage units operate at the maximum charge and discharge power;

[0187] When the target power of the energy storage system is not zero and the target power is less than the sum of the maximum charge and discharge powers of all energy storage units, determine the number of energy storage units required according to the target power and the maximum charge and discharge powers of each energy storage unit, and select the energy storage units to be used in combination with the battery health status quantification index according to the energy storage unit priority rule;

[0188] Perform global optimization on the target power distribution to obtain the optimal or approximately optimal power distribution values of each energy storage unit to be used.

[0189] The above are only the preferred embodiments of the present invention and are not used to limit the present invention. For those skilled in the art, the present invention can have various changes and modifications. Any modification, equivalent replacement, improvement, etc. made within the spirit and principle of the present invention shall be included in the protection scope of the present invention.

Claims

1. A power distribution method for a battery energy storage system, characterized in that, It includes the following processes: Obtain the state parameters of the battery cells of each energy storage unit, extract key indirect indicators for characterizing the battery health status according to the state parameters, and determine the battery health state quantification index according to the key indirect indicators; When the target power of the energy storage system is not zero and the target power is greater than or equal to the sum of the maximum charge-discharge powers of all energy storage units, all energy storage units operate at the maximum charge-discharge power; When the target power of the energy storage system is not zero and the target power is less than the sum of the maximum charge-discharge powers of all energy storage units, determine the required number of energy storage units according to the target power and the maximum charge-discharge powers of each energy storage unit, and select the energy storage units to be used in combination with the battery health state quantification index according to the energy storage unit priority rule; Perform global optimization on the target power distribution to obtain the optimal or approximately optimal power distribution values of the energy storage units to be used.

2. The method for power distribution of a battery energy storage system according to claim 1, characterized in that When the target power of the energy storage system is zero, there is no current power demand, no charge-discharge operation is required, and no power distribution is performed.

3. The method for power distribution of a battery energy storage system according to claim 1, characterized in that Determine the possible number of energy storage units to be put into use according to the comparison between the target power and the maximum charge-discharge powers of each energy storage unit.

4. The method for power distribution of a battery energy storage system according to claim 1, characterized in that Select the energy storage units to be used according to the energy storage unit priority rule, including: Construct a priority sorting function expression: F1 = af1 + bf2, where a and b are weights, f1 is a sorting judgment function that only considers the SOC of the energy storage unit, and f2 is a priority sorting function that considers the differences in the health status of each battery, the health status of the optimal cell and the worst cell, and the inconsistency between the SOC in the energy storage unit; When the state of charge of the energy storage unit exceeds the upper limit of the best cycle interval, F1 takes positive infinity and discharges first; When the state of charge of the energy storage unit is lower than the lower limit of the best cycle interval, F1 takes negative infinity and charges first; When the state of charge of the energy storage unit is within the best cycle interval, and when the energy storage unit discharges, the larger F1 is, the more priority is given to discharging, and when the energy storage unit charges, the smaller F1 is, the more priority is given to charging.

5. The method for power distribution of a battery energy storage system according to claim 4, characterized in that Among them, S OC.max and S OC.min are the maximum and minimum values of the optimal cycle interval, and S OC.ref is the optimal cycle midpoint value, and S OC.i (t) is the SOC of the i-th energy storage unit at the current time t.

6. The method for power distribution of a battery energy storage system according to claim 4, characterized in that Among them, τ is the charge and discharge judgment flag bit, which is set to 1 when the energy storage unit participates in the discharge action and set to -1 when the energy storage unit participates in the charge action, ΔC HI，i is the extremely poor health state within the i-th unit group, ΔS OC，i (t) is the difference in SOC between the worst single cell and the best single cell in the i-th energy storage unit at time t, ΔS OC.lim is the SOC imbalance limit value.

7. The method for power distribution of a battery energy storage system according to claim 4, characterized in that Optimize the power of each energy storage unit, take the minimum level of inconsistency in the state of charge within each energy storage unit as the objective function for power distribution, and use the AGPSO algorithm for iterative solution to obtain the optimal or approximately optimal power distribution values of the energy storage units to be used.

8. A power distribution system for a battery energy storage system, characterized in that, It includes: A quantization index acquisition unit configured to: obtain the state parameters of the battery cells of each energy storage unit, extract key indirect indicators for characterizing the battery health status according to the state parameters, and determine the battery health state quantification index according to the key indirect indicators; A first processing unit, configured to: when the target power of the energy storage system is not zero and the target power is greater than or equal to the sum of the maximum charge-discharge powers of all energy storage units, all energy storage units operate at the maximum charge-discharge power; A second processing unit, configured to: when the target power of the energy storage system is not zero and the target power is less than the sum of the maximum charge-discharge powers of all energy storage units, determine the required number of energy storage units according to the target power and the maximum charge-discharge power of each energy storage unit, and select the energy storage units to be used in accordance with the energy storage unit priority rule in combination with the battery health status quantization index; A power distribution unit, configured to: perform global optimization on the target power distribution to obtain the optimal or approximately optimal power distribution values of the energy storage units to be used.

9. A computer device, characterized in that, Comprising: A processor and a computer-readable storage medium; The processor is adapted to execute a computer program; The computer-readable storage medium stores a computer program, and when the computer program is executed by the processor, it implements the battery energy storage system power distribution method according to any one of claims 1 to 7.

10. A computer-readable storage medium, characterized in that, The computer-readable storage medium stores a computer program, and the computer program is adapted to be loaded and executed by the processor to implement the battery energy storage system power distribution method according to any one of claims 1 to 7.

Images