Charging and discharging control method and system for energy storage battery system

Abstract

According to the charge and discharge control method and system for the energy storage battery system, a state space model is constructed based on a Thevenin equivalent circuit theory, voltage, current and temperature data collected in real time are fused, and the charge state and the available power state of each battery cluster are estimated with high precision by adopting an extended Kalman filtering algorithm; according to the operation mode, a dynamic sorting algorithm is adopted to calculate the charging and discharging dynamic weight of each battery cluster, the power of the low-SOC battery cluster is preferentially improved in the charging mode, and the power of the high-SOC battery cluster is preferentially called in the discharging mode; in combination with a multi-objective optimization function which aims at improving economical efficiency and power grid supporting capacity, target power of each battery cluster is determined through iterative allocation under the condition that condition constraints are met, and independent control is executed by a string converter. Accurate sensing of the battery state and dynamic optimization and balance control of the charging and discharging strategy are achieved, the system operation economical efficiency and the power grid supporting capacity are remarkably improved, and the service life of the battery is remarkably prolonged.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of power system energy storage and battery management, in particular to a kind of energy storage battery system charge-discharge control method and system. BACKGROUND

[0002] With the increasing penetration of renewable energy (such as solar and wind) in the power grid, its inherent intermittency and volatility pose significant challenges to grid stability (such as load balancing and voltage stability). Energy storage systems, as a key regulating means, effectively smooth fluctuations and shave peaks by charging during low load periods and discharging during peak periods, which are crucial for maintaining stable operation of the power grid.

[0003] However, in a complex dynamic grid environment, there are two major challenges to achieving precise and efficient charge-discharge control of energy storage systems: 1) Insufficient state estimation accuracy: The core state parameters of the battery in the energy storage system, especially the state of charge (SOC), have strong nonlinear characteristics. The state estimation method based on traditional Kalman filtering widely used in existing technologies has limited accuracy in dealing with such nonlinear dynamic parameters, making it difficult to provide reliable real-time state information.

[0004] 2) Poor adaptability of control strategies: Many existing control schemes use fixed rules or simple threshold strategies for charge-discharge decision-making. Such strategies are difficult to respond in real time and flexibly to rapid fluctuations in grid load, changes in electricity prices, and uneven SOC within the battery pack, resulting in poor timing, low efficiency, and suboptimal economics for charging and discharging, and even exacerbating battery aging or affecting grid support effectiveness.

[0005] Therefore, it is a technical problem for those skilled in the art to provide an energy storage battery system charge-discharge control method and system to solve the above problems. SUMMARY

[0006] To solve the technical problems of insufficient state estimation accuracy and poor adaptability of control strategies in the prior art, the present application aims to provide an energy storage battery system charge-discharge control method. This method has clear logic and can break through the bottleneck of existing technologies by constructing a complete technical chain of high-precision state estimation, dynamic weight decision, and multi-objective optimization control, thereby significantly improving the operational economics, support for grid stability, and battery life of energy storage systems in complex grid environments.

[0007] The technical solution provided by the present application is as follows: An energy storage battery system charge-discharge control method, comprising the following steps: construct a state space model, acquire voltage, current, temperature measurement values of the energy storage battery system in real time according to the state space model, and acquire state of charge and available power state of each battery cluster by using an extended Kalman filtering algorithm; According to the state of charge and the available power state of each battery cluster, the dynamic weight of the charge and discharge of each battery cluster in different modes is calculated by using a dynamic sorting algorithm. According to the preset multi-objective optimization function and the dynamic weight of the charge and discharge of each battery cluster, the target charge and discharge power of each battery cluster is determined under the condition of meeting the preset condition constraint. According to the target charge and discharge power of each battery cluster, the independent charge and discharge control of each battery cluster is performed by using a string type energy storage converter.

[0008] Preferably, the state space model is constructed based on the Thevenin equivalent circuit theory, and specifically: ; ; wherein, is a system state vector at a moment, is a control input vector containing voltage, current and temperature measurement values, is a process noise, is a nonlinear state transition function, is an observation vector, is an observation noise, is an observation function. Preferably, the state of charge and the available power state of each battery cluster are acquired by using the extended Kalman filtering algorithm, including the following steps:

[0009] state estimation is performed on the state space model by using the extended Kalman filtering to obtain a state estimation value; The state of charge and the available power state of each battery cluster are acquired by using the extended Kalman filtering and the state estimation value. Preferably, the state estimation is performed on the state space model by using the extended Kalman filtering to obtain a state estimation value, including the following steps:

[0010] Prediction step: based on the system state transition matrix and the process noise covariance matrix, a state prediction value and a prediction covariance matrix are calculated, and the formula is: ; ; wherein, is a first state estimation value, is a first covariance matrix, is a state transition matrix,​ It is the covariance matrix of the process noise; Update steps: Based on the observation matrix and the observation noise covariance matrix, calculate the Kalman gain, and fuse the observations including the state of charge and the available power state. Update the first estimate to the second state estimate and update the first covariance matrix to the second covariance matrix, as shown in the formula: ; ; ; in, For Kalman gain, For the observation matrix, It is the observation noise covariance matrix. This is the second state estimate. The second covariance matrix, It is an identity matrix.

[0011] Preferably, the step of calculating the dynamic charging and discharging weights of each battery cluster under different modes using a dynamic sorting algorithm based on the state of charge and available power state of each battery cluster includes the following steps: Based on the operating instructions of the energy storage system, determine the type of operating mode currently in which the energy storage system is located; If the current energy storage system is in charging mode, the battery clusters are sorted in ascending order of their state of charge, and the battery clusters with lower states of charge that are ranked higher are given higher charging priority. If the current energy storage system is in discharge mode, the battery clusters are sorted in descending order of their state of charge, and the battery clusters with higher state of charge at the top of the list are given higher discharge priority. Based on the sorting priority, the available power state of each battery cluster, and the preset weighting formula, the dynamic weight of charging and discharging of each battery cluster is calculated, wherein the battery cluster with higher priority is assigned a higher power allocation weight. The preset weight formula is specifically as follows: ; in, For the first The dynamic weight of each battery cluster For the target state of charge, To obtain the first digit in real time using the extended Kalman filter algorithm Real-time state of charge of each battery cluster For the first The rated capacity of each battery cluster, This is the temperature influence coefficient. For the first Real-time temperature of each battery cluster For ideal operating temperature.

[0012] Preferably, after determining the type of the current operating mode of the energy storage system based on the operating instructions of the energy storage system, the following steps are further included: If the current energy storage system is in standby mode, each battery cluster is controlled to perform low-power energy transfer at a preset low ratio of the system's rated power. The balancing current used for energy transfer is controlled within a preset current range to achieve autonomous balancing of the state of charge among the battery clusters. If the current energy storage system is in fault mode, the status of each battery cluster is monitored in real time. When the difference in state of charge between battery clusters is found to be greater than the first preset threshold, a system alarm is issued. When the difference in state of charge between battery clusters is greater than the second preset threshold, the battery cluster with abnormal state of charge is disconnected and enters the derating operation state. Wherein, the first preset threshold is less than the second preset threshold.

[0013] Preferably, the step of determining the target charge / discharge power of each battery cluster based on a preset multi-objective optimization function and the dynamic charge / discharge weights of each battery cluster, under preset constraints, includes the following steps: Based on a preset multi-objective function, with the goal of synergistically improving the operating economy of the energy storage system and the grid support capability, the total charging and discharging power demand of the system during the optimization period is solved. Based on the total charging and discharging power demand of the system and the dynamic charging and discharging weights of each battery cluster, the target charging and discharging power of each battery cluster is determined through an iterative allocation algorithm, under the constraints of battery cluster safety limits, power capacity of the energy storage converter, and power over-limit handling mechanism.

[0014] Preferably, the multi-objective function is specifically: ; in, , , These are the weighting coefficients. To optimize the time period, and They are respectively Discharge power and charging power at any given time and for The electricity purchase price and electricity sales price at any given time. for State of charge at time t, For the target SOC value, The net power exchanged between the energy storage system and the grid. This is the reference power for grid demand.

[0015] Preferably, the power overrun processing mechanism comprises the following steps: When any battery cluster reaches its power upper limit during the power distribution process, the power of the battery cluster is fixed as the upper limit value; The battery cluster is excluded from the current power distribution process, and the remaining power to be distributed is redistributed to the remaining battery clusters based on the updated dynamic weights until the power distribution is completed or all battery clusters reach the power upper limit.

[0016] A power storage battery system charge-discharge control system comprises: A parameter acquisition module is configured to construct a state space model, acquire real-time voltage, current and temperature measurement values of the power storage battery system according to the state space model, and acquire state of charge (SOC) and available power state of each battery cluster by using an extended Kalman filter algorithm; A dynamic weight acquisition module is configured to calculate and acquire charge-discharge dynamic weights of each battery cluster in different modes by using a dynamic sorting algorithm according to the SOC and available power state of each battery cluster; A target charge-discharge power determination module is configured to determine target charge-discharge power of each battery cluster under a preset condition constraint according to a preset multi-objective optimization function and the charge-discharge dynamic weights of each battery cluster; A charge-discharge control module is configured to independently control charge and discharge of each battery cluster by using a string-type energy storage converter according to the target charge-discharge power of each battery cluster.

[0017] The power storage battery system charge-discharge control method provided by the application first constructs a state space model based on Thevenin equivalent circuit theory, and fuses real-time acquired voltage, current and temperature data to estimate the SOC and available power state of each battery cluster with high precision by using an extended Kalman filter algorithm; then, the charge-discharge dynamic weights of each battery cluster are calculated by using a dynamic sorting algorithm according to the operation mode, the power of a low SOC battery cluster is preferentially improved in the charging mode, and the power of a high SOC battery cluster is preferentially called in the discharging mode; finally, the target power of each battery cluster is determined by iteration distribution under the condition constraint of battery cluster safety limit and converter capacity in combination with a multi-objective optimization function aiming at improving economy and power grid support capability, and independent control is performed by a string-type converter. The application realizes accurate perception of battery state, dynamic optimization and balanced control of charge-discharge strategy, and significantly improves the system operation economy, power grid support capability and battery life.

[0018] The application realizes systematic and closed-loop intelligent control by constructing a complete technical chain composed of state estimation, dynamic sorting, multi-objective optimization and independent control. The technical effects are realized in multiple levels and in cooperation, and are embodied in the following aspects: High-precision system state perception was achieved: by constructing a state-space model based on Thevenin equivalent circuit theory and using the extended Kalman filter algorithm to fuse multi-source real-time data, the SOC and SOP of each battery cluster can be estimated more accurately, providing a reliable data foundation for subsequent optimization control and effectively avoiding the risks of battery overcharging and over-discharging.

[0019] Intelligent and adaptive power allocation is achieved: by introducing dynamic sorting algorithms and multi-objective optimization functions, the charging and discharging strategy can dynamically respond to system status and external demands. It not only actively maintains the SOC balance among battery clusters by "charging to compensate for low levels and discharging to utilize high levels," thereby improving system capacity utilization and extending battery life, but also optimizes charging and discharging timing to improve operational economy while meeting grid support requirements (such as frequency regulation and backup).

[0020] The system's safety and robustness are ensured by establishing multiple constraints, including battery cluster safety limits, PCS capabilities, and power limit overrun handling mechanisms. An iterative algorithm is employed in power allocation to ensure that the equipment operates within a safe range under any conditions. When the capabilities of some battery clusters are limited, the system can automatically reconfigure and derate through the aforementioned limit overrun handling mechanism, guaranteeing the system's continuous service capability and overall resilience.

[0021] Precise independent control is achieved: By adopting a string energy storage converter architecture, "one-to-one" independent control of each battery cluster is realized, which enables the aforementioned optimization decisions to be executed precisely, avoiding the problem of limiting the performance of the entire system due to individual "weak" battery clusters, and maximizing the potential of the energy storage system.

[0022] In summary, this invention organically integrates high-precision state perception, intelligent decision optimization, and precise independent execution into a closed-loop control system. This not only overcomes the bottlenecks of existing technologies in state estimation, strategy flexibility, and execution accuracy, but also achieves a synergistic leap in multiple key performance indicators such as system security, operational economy, grid support capability, and battery life through the synergistic enhancement effect within the technical solution. This provides core technical support for the efficient, reliable, and economical operation of energy storage systems in complex grid environments.

[0023] The present invention also provides a charging and discharging control system for an energy storage battery system. Since it belongs to the same technical concept as the method and solves the same technical problem, it should have the same beneficial effects, and will not be described in detail here. Attached Figure Description

[0024] To more clearly illustrate the technical solutions in the embodiments of this application or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments recorded in this application. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0025] Figure 1 This is a flowchart of a charging and discharging control method for an energy storage battery system provided in an embodiment of the present invention; Figure 2 This is a flowchart for obtaining the state of charge and available power state of each battery cluster according to an embodiment of the present invention; Figure 3 This is a flowchart of step A1 provided in an embodiment of the present invention; Figure 4 This is a flowchart of step S2 provided in an embodiment of the present invention; Figure 5 This is a flowchart of step S3 provided in an embodiment of the present invention; Figure 6 This is a schematic diagram of the structure of a charging and discharging control system for an energy storage battery system provided in an embodiment of the present invention. Detailed Implementation

[0026] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. 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 are within the scope of protection of the present invention.

[0027] The embodiments of this invention are written in a progressive manner.

[0028] This invention provides a charging and discharging control method for an energy storage battery system. It primarily addresses the technical problems in existing technologies, such as insufficient accuracy in estimating the state parameters of nonlinear energy storage battery systems, rigid traditional charging and discharging strategies, inability to respond to real-time grid load and electricity price fluctuations, low capacity utilization due to unbalanced state of charge (SOC), accelerated lifespan degradation, and lag in system dynamic response.

[0029] like Figure 1 As shown, a charging and discharging control method for an energy storage battery system includes the following steps: S1. Construct a state-space model. Based on the state-space model, obtain the voltage, current, and temperature measurements of the energy storage battery system in real time, and use the extended Kalman filter algorithm to obtain the state of charge and available power state of each battery cluster. S2. According to the state of charge and available power state of each battery cluster, the dynamic ranking algorithm is used to calculate the charging and discharging dynamic weight of each battery cluster in different modes; S3. According to the preset multi-objective optimization function and the charging and discharging dynamic weight of each battery cluster, the target charging and discharging power of each battery cluster is determined under the condition of meeting the preset condition constraint; S4. According to the target charging and discharging power of each battery cluster, the independent charging and discharging control of each battery cluster is carried out through the group string energy storage converter.

[0030] The state space model in step S1 is a mathematical framework for describing and analyzing dynamic systems. It uses a set of internal variables called state variables to completely characterize the state of the system at any time, and describes how these state variables evolve over time through a set of equations, as well as their relationship with the external input and output (observations) of the system; Extended Kalman filter is an optimal state estimator suitable for nonlinear systems. Its core task is to estimate the internal state of the system that cannot be directly measured in real time and with high precision from noisy and incomplete observation data when the behavior of the system itself or the observation process is not a simple linear relationship; The dynamic ranking algorithm in step S2 is an intelligent decision-making algorithm for real-time calculation of the charging and discharging priority of each battery cluster in the energy storage system. Its core idea is to abandon the fixed and pre-set charging and discharging sequence, and instead dynamically and automatically calculate and update the "priority score" (i.e. dynamic weight) of each battery cluster according to the real-time running state of the system (such as SOC, SOP, temperature, etc.) and multi-objective optimization requirements, and then allocate power accordingly; The dynamic weight of charging and discharging is a quantitative priority value calculated by the dynamic ranking algorithm. It represents the priority degree and proportion share of a certain battery cluster relative to other battery clusters in the system at a certain time; In step S3, the multi-objective optimization function is a mathematical expression used to quantitatively evaluate the overall performance of a decision on multiple, often conflicting objectives. This function is a scalar function that combines multiple core optimization objectives (economy, grid support, battery health) into a single numerical value. The goal of the control strategy is to find a set of charging and discharging power instructions that minimize (or maximize) the value of this function, thereby achieving optimal or suboptimal control of the system's overall performance; The condition constraint refers to the physical limits, safety boundaries and device capacity restrictions that must be strictly followed in the optimization process of determining the target charging and discharging power of each battery cluster. These constraints form the "safety barrier" of the control strategy, ensuring that the power instructions generated by the optimization algorithm are not only superior in performance, but also physically achievable and safe. Any optimal solution that violates these constraints will be considered infeasible and discarded; The string energy storage converter in step S4 is a distributed, modular power conversion system. Its core architectural feature is "one cluster, one PCS", meaning that each independent battery cluster in the system is equipped with a dedicated, lower-power energy storage converter, which independently controls its charging and discharging.

[0031] Steps S1 to S4 detail the specific implementation of the charging and discharging control method for the energy storage battery system. They involve constructing a state-space model based on Thevenin's equivalent circuit theory, integrating real-time voltage, current, and temperature data, and employing an extended Kalman filter algorithm to accurately estimate the state of charge and available power state of each battery cluster. A dynamic sorting algorithm is used to calculate the dynamic charging and discharging weights of each battery cluster according to the operating mode. Combined with a multi-objective optimization function aimed at improving economic efficiency and grid support capabilities, and under the constraints of battery cluster safety limits and converter capacity, the target power of each battery cluster is determined through iterative allocation, and independent control is performed by the string converter.

[0032] Preferably, the state-space model is constructed based on Thevenin's equivalent circuit theory, specifically as follows: ; ; in, for The system state vector at time t. This is a control input vector that includes measured values ​​of voltage, current, and temperature. For process noise, It is a nonlinear state transition function. For the observation vector, To observe the noise, For observation functions.

[0033] The above content describes the specific implementation details of the state-space model in step S1. By introducing the Thevenin equivalent circuit theory, a fundamental improvement is achieved: accurate characterization of the nonlinear dynamics of the battery is realized; and the estimation accuracy of the extended Kalman filter algorithm is significantly improved.

[0034] By establishing a high-fidelity, nonlinear mathematical model based on physical mechanisms, an accurate predictive foundation is provided for the extended Kalman filter algorithm. This fundamentally overcomes the inherent defects of traditional state estimation methods, such as large accumulated errors and reliance on static operating conditions, and achieves real-time, high-precision, online estimation of battery state of charge and available power state. This achievement constitutes the indispensable and most fundamental data foundation for the entire invention to realize intelligent optimization control, improve system economy, safety, and grid support capabilities.

[0035] like Figure 2As shown, preferably, the extended Kalman filter algorithm is used to obtain the state of charge and available power state of each battery cluster, including the following steps: A1. The extended Kalman filter is used to estimate the state of the state-space model to obtain the state estimate; A2. By using extended Kalman filtering and state estimation, the state of charge and available power state of each battery cluster are obtained.

[0036] Steps A1 and A2 are the specific implementation details of obtaining the state of charge and available power state of each battery cluster in step S1. They involve using the extended Kalman algorithm to perform state estimation on the spatial model, obtaining state estimates, and then further obtaining the state of charge and available power state of each battery cluster.

[0037] In one embodiment, step A2 specifically involves: Extract the required state parameters from the updated state vector.

[0038] Charged state: the first element of the state vector.

[0039] Available power state: Based on the updated state vector, internal resistance, and battery voltage / current limits, the maximum acceptable charging power and maximum output discharge power of the battery in the current state are estimated in real time by looking up tables or calculations. like Figure 3 As shown, preferably, the extended Kalman filter is used to estimate the state of the state-space model to obtain the state estimate value, including the following steps: B1. Prediction Steps: Based on the system state transition matrix and process noise covariance matrix, calculate the predicted state value and the prediction covariance matrix, using the following formula: ; ; in, It is the first state estimate. It is the first covariance matrix. It is the state transition matrix. It is the state transition matrix The transpose of the matrix, It is the covariance matrix of the process noise; B2. Update Steps: Based on the observation matrix and the observation noise covariance matrix, calculate the Kalman gain, fuse the observations including the state of charge and the available power state, update the first estimate to the second state estimate, and update the first covariance matrix to the second covariance matrix. The formula is as follows: ; ; ; in, For Kalman gain, For the observation matrix, It is the observation matrix The transpose of the matrix, It is the observation noise covariance matrix. This is the second state estimate. The second covariance matrix, It is an identity matrix.

[0040] The system state transition matrix in step B1 is a Jacobian matrix that describes how the system state vector linearly evolves around the current operating point within a sampling period. Its function is to linearly propagate the uncertainty of the state estimate from the previous moment (represented by the error covariance matrix) to the current moment in the "prediction step." The process noise covariance matrix is ​​a matrix used to quantify the uncertainty of the state equation model itself. It represents the errors present in the state prediction process, which mainly originate from: model inaccuracy: Thevenin equivalent circuit model is a simplification of complex electrochemical systems and cannot accurately describe all the dynamics of the battery. Unmodeled dynamics: Some minor or difficult-to-model physical / chemical processes are ignored. External random disturbances: Unpredictable small external disturbances. The observation matrix in step B2 is a matrix that maps the system state space to the observation space, and it is also a Jacobian matrix. It precisely describes how the internal state of the system affects the observations we can actually measure. The observation noise covariance matrix is ​​a matrix used to quantify the uncertainty or noise level of sensor observations. It represents the errors present in the measurements read from the sensors, which mainly come from: the accuracy limitations of the sensors themselves: for example, the inherent measurement errors, nonlinearities, and zero-point drift of voltage and current sensors; environmental electromagnetic interference: high-frequency noise introduced by the acquisition system; and quantization errors of analog-to-digital conversion. The Kalman gain is a dynamic weighting matrix that optimally balances and integrates the "model-based predictions" and the "sensor-based observations" in each estimation cycle.

[0041] Steps B1 to B2 detail the implementation of step A1. This algorithm is not a simple mathematical calculation, but rather the core engine for achieving high-precision state estimation through dynamic, adaptive, and optimal means. It doesn't simply rely on model predictions or actual measurements; instead, it uses the Kalman gain as a "smart arbitrator" to dynamically and optimally determine whether to place more weight on the model or the measurements in the current estimation, thus obtaining a more accurate result than relying solely on either. When the model prediction is highly reliable but the measurement noise is significant: the calculation formula reduces the Kalman gain, and the algorithm tends to trust the model's prediction during updates, making only minor corrections using the measured values. When the model prediction has high uncertainty but the measurements are very accurate: the calculation formula increases the Kalman gain. The algorithm tends to trust the new observations during updates, making significant corrections to the predictions using the measured values. This dynamic adjustment makes the state estimation method highly robust, adapting to changes in model parameters and noise characteristics under different operating conditions and aging stages of the battery.

[0042] The above scheme constitutes an "optimal state estimation engine" capable of dynamically balancing model prediction and real-time measurement, and possessing self-uncertainty assessment and correction capabilities. Through its inherent adaptive mechanism, this engine fundamentally guarantees the real-time performance and high accuracy of SOC and SOP estimates, providing the most critical and reliable technical guarantee for overcoming the inherent defects of traditional methods and realizing subsequent intelligent optimization control. It is the core algorithm that enabled the success of the entire invention.

[0043] like Figure 4 As shown, preferably, based on the state of charge and available power state of each battery cluster, a dynamic sorting algorithm is used to calculate and obtain the dynamic charging and discharging weights of each battery cluster under different modes, including the following steps: C1. Determine the current operating mode of the energy storage system based on its operating instructions; C21. If the current energy storage system is in charging mode, the battery clusters are sorted in ascending order of their state of charge, and the battery clusters with lower states of charge that are ranked higher are given higher charging priority. C22. If the current energy storage system is in discharge mode, the battery clusters are sorted in descending order of state of charge, and the battery clusters with higher state of charge at the top of the list are given higher discharge priority. C3. Based on the sorting priority, the available power state of each battery cluster and the preset weight formula, calculate the dynamic weight of charging and discharging for each battery cluster, where higher priority battery clusters are assigned higher power allocation weights. The preset weight formula is as follows: ; in, For the first The dynamic weight of each battery cluster target state of charge, real-time state of charge of the i-th battery cluster, rated capacity of the i-th battery cluster, temperature influence coefficient, real-time temperature of the i-th battery cluster, ideal working temperature.

[0044] The ranking priority in step C3 refers to the order of priority obtained by each battery cluster after all battery clusters in the system are ranked according to a preset optimization target (such as balancing, efficiency, safety) in a certain operating mode through a dynamic ranking algorithm; the charging and discharging dynamic weight is a quantitative, unitless value that represents the priority or proportion of the i-th battery cluster in the current time relative to all other battery clusters in the system, and a higher value means that the battery cluster should be given a larger charging and discharging power instruction at the current time; The real-time state of charge refers to the ratio of the remaining available capacity to the current maximum available capacity of the battery or battery cluster at any current time; the temperature influence coefficient is a greater-than-zero adjustment parameter used to quantify and adjust the punishment of the temperature deviation on the final power distribution priority of the battery cluster in the dynamic weight calculation; the ideal working temperature refers to the temperature point or the center value of the temperature interval at which the battery can exhibit the best overall performance (including capacity, efficiency, power capability, and life); Steps C1 to C3 are specific implementation details of step S2, and through steps C21 and C22, the same ultimate goal of SOC balancing is achieved in different modes using different strategies. When the system is charging, higher charging power is preferentially allocated to low-SOC battery clusters to accelerate their energy recovery. This can quickly narrow the energy gap between low-SOC battery clusters and high-SOC battery clusters, actively leveling the inter-cluster differences, avoiding the "bucket effect", and improving the overall available capacity of the system; when the system is discharging, higher discharging power is preferentially allocated to high-SOC battery clusters, allowing low-SOC battery clusters to rest. This delays the deep discharge of low-SOC battery clusters and protects them, while also consuming the energy of high-SOC clusters, which also helps to bring the SOC of all battery clusters closer to the middle, achieving balancing; the weight formula in step C3 is the point of sublimation of the technical effect. It not only considers SOC, but also achieves multi-objective collaborative optimization, strengthens the SOC balancing driving force, embeds temperature protection and life extension mechanisms, and ensures the feasibility of the decision.

[0045] ​​​​The above scheme converts the accurate state perception results into an intelligent power distribution strategy that can not only actively achieve energy balance among battery clusters, but also simultaneously consider battery temperature safety and life extension, and strictly follow the power capability of the battery itself, through the decision chain of "pattern recognition → dynamic sequencing → multi-factor weight calculation". It makes the control of the energy storage system leap from passive, rigid, and single-target to active, flexible, and multi-target collaborative advanced energy management, which is the key intelligent decision core for improving the economy, safety, and life of the system.

[0046] Preferably, after determining the type of the current operating mode of the energy storage system according to the operating instructions of the energy storage system, the following steps are further included: If the current energy storage system is in standby mode, control each battery cluster to transfer energy at a preset low proportion of the rated power of the system, wherein the balancing current for energy transfer is controlled within a preset current range to achieve self-balancing of the state of charge among the battery clusters; If the current energy storage system is in fault mode, monitor the state of each battery cluster in real time, and when the state of charge difference between the battery clusters is greater than a first preset threshold, issue a system alarm, and when the state of charge difference between the battery clusters is greater than a second preset threshold, remove the battery cluster with abnormal state of charge, and enter a derated operating state; Wherein, the first preset threshold is less than the second preset threshold.

[0047] Wherein, energy transfer specifically refers to the process of transferring energy from one or more battery clusters to another one or more battery clusters through power electronic converters within the energy storage system; the first preset threshold is a preset critical value for triggering a system alarm. It specifically refers to, in fault mode, when the system monitors that the state of charge difference between the battery clusters reaches this value, the system will determine that an abnormal state has occurred and start the alarm mechanism; the second preset threshold is a critical value higher than the first preset threshold for triggering protective isolation actions. It specifically refers to, in fault mode, when the system monitors that the state of charge difference between the battery clusters further deteriorates and reaches this value, the system will determine that the fault has escalated and automatically execute hard protection measures such as removing the abnormal battery cluster; removing the battery cluster with abnormal state of charge refers to, in fault mode, when the system detects that the state of charge of one or more battery clusters deviates significantly from the group (i.e., the SOC difference is greater than the second preset threshold), which may endanger the safety or overall performance of the system, the control action of physically isolating or electrically disconnecting it from the main circuit through hardware switching devices (such as contactors, relays, or power semiconductors); derated operation refers to a protective operation strategy that actively and consciously reduces the upper limit of the overall output / input power and energy capacity of the system in order to maintain the system to continue to provide core services safely and stably after the system fails or part of the capacity is missing (such as the battery cluster being removed).

[0048] The above steps are specific implementation details after step C1. The control strategy of the standby mode enables the system to always maintain a high-precision balancing state, avoids the accumulation and deterioration of unbalanced states, and ensures that the system can work in an "optimal state" (highly consistent SOC of each battery cluster) once a running instruction is received, thereby improving the efficiency and capacity utilization of subsequent charging and discharging and the safety margin of the system. The control strategy of the fault mode establishes a three-level active safety defense system of "early warning-isolation-stability maintenance", realizes smooth transition from fault warning to fault tolerance operation, and achieves the effects of early warning, prevention, prevention of fault expansion, guarantee of core safety, and continuous service capability under fault.

[0049] The control strategies of the standby and fault modes jointly constitute the advanced intelligence and management capability of the energy storage system beyond the basic functions. The standby mode has the effect of forward-looking maintenance, and the system always maintains the best healthy state through "trickle balancing". The fault mode has the effect of defensive protection, and the system ensures the resilience and continuous service capability under abnormal working conditions through "graded safety barriers". The combination of the two significantly improves the overall availability, safety level, and life cycle value of the energy storage system, and upgrades it from a simple energy storage device to an intelligent energy asset with self-maintenance, self-diagnosis, and self-protection.

[0050] In one embodiment, the standby mode specifically performs energy transfer at 0.5% to 1% rated power, and the balancing current is between 1A and 5A.

[0051] The fault mode specifically automatically removes the battery cluster with abnormal state of charge. When the SOC difference is >30% (first preset threshold), an alarm is given, and when the SOC difference is >40% (second preset threshold), the system is cut off, and the capacity is reduced to 70% for operation.

[0052] As shown in Figure 5 Preferably, according to a preset multi-objective optimization function and the charging and discharging dynamic weight of each battery cluster, the target charging and discharging power of each battery cluster is determined under the satisfaction of preset condition constraints, including the following steps: D1. According to the preset multi-objective function, the total charging and discharging demand power of the system in the optimization period is solved, with the goal of synergistically improving the economic efficiency and grid support capability of the energy storage system; D2. According to the total charging and discharging demand power of the system, the charging and discharging dynamic weight of each battery cluster, under the condition constraints of satisfying the safety limit value of the battery cluster, the power capability of the energy storage converter, and the power over-limit processing mechanism, the target charging and discharging power of each battery cluster is determined through an iterative distribution algorithm.

[0053] In Step D1, the energy storage system operation economy refers to maximizing net benefits or minimizing total costs throughout the life cycle of the energy storage system through a series of operation strategies and activities. It focuses on how to make the "charge" and "discharge" behavior of the energy storage bring the maximum economic benefit; the grid support capability refers to the technical ability of the energy storage system to actively help the grid maintain safe, stable and efficient operation in addition to simple charging and discharging; the total charging and discharging demand power refers to the total power value calculated and required to be absorbed or released from the grid port (PCC point) within a specific optimization time interval (such as 15 minutes, 1 hour) in order to achieve the preset comprehensive operation goal of the entire energy storage power station; In Step D2, the battery cluster safety limit refers to the unbreachable boundary value of each key operation parameter set to protect the safety, life and integrity of the battery cluster (a collection of multiple cells connected in series and parallel) during the operation of the energy storage system. These limits are a direct reflection of the physical and chemical properties of the battery and are rigid constraints that usually have the highest priority in control logic; the power overrun processing mechanism is a pre-designed control logic and algorithm that is triggered when the system receives a total charging and discharging demand power instruction and finds that it cannot complete the instruction under the constraints of all "battery cluster safety limits" and "PCS power capability"; the iterative allocation algorithm is a numerical calculation method that solves the optimal solution of each battery cluster power allocation through cyclic calculation and step-by-step approximation under multiple complex constraint conditions.

[0054] Steps D1 to D2 are specific implementation details of step S3. By using a "preset multi-objective function", the economic objectives such as participating in peak-valley arbitrage, reducing electricity costs, participating in electricity market transactions, etc. and the grid support objectives such as smoothing renewable energy fluctuations, participating in frequency modulation, providing backup capacity, stabilizing local voltage, etc. can be considered at the same time. This enables the energy storage system to upgrade from a simple "electricity cost management tool" to an "active support asset" of the power grid, creating direct economic benefits and providing auxiliary service value, improving asset utilization and return on investment; Since the solution is obtained within an "optimized time period" (such as the next 15 minutes or 1 hour), it means that the controller will develop a globally optimal charging and discharging plan based on the prediction of future electricity prices, loads, and new energy output. Avoiding "short-sighted" behavior. For example, it will not empty all the electricity because the current electricity price is high, but will reserve a portion of the capacity to respond to subsequent more important frequency modulation tasks or higher electricity prices, thereby maximizing total revenue over the entire time period; The dynamic weight of each battery cluster in step D2 effectively solves the "barrel effect" between battery clusters. Avoiding the problem of some battery clusters being overused and prematurely decommissioned while other battery clusters are underutilized, thereby significantly extending the service life of the entire energy storage system and improving the economic efficiency of the entire life cycle; The algorithm strictly allocates within the "battery cluster safety limits" (such as maximum charging and discharging current, voltage, and temperature window) and the "power capacity of the energy storage converter" (the rated power and overload capacity of the PCS), fundamentally eliminating safety issues such as overcharging, overdischarging, and overheating caused by improper power allocation, ensuring that the equipment always operates within the safety zone, and improving the reliability and safety of the system; When the total power instruction cannot be fully executed due to the safety limits of one or more battery clusters: Power exceeding limit handling mechanism: A contingency plan will be initiated, such as prioritizing discharge power (for critical loads) or reducing total power by weight proportion. Iterative allocation algorithm: dynamically and repeatedly calculates to approach the total demand power as much as possible or redistribute the remaining power to other battery clusters that can execute it under all constraints, greatly enhancing the adaptability and robustness of the system. Even in the case of partial battery cluster performance degradation or exit from operation, the system can continue to work in a "reduced capacity" mode rather than completely shut down, ensuring service continuity.

[0055] The control strategy designed in the above scheme transforms a potentially "extensive" energy storage system into an "intelligent agent" with fine management, intelligent decision-making, and high resilience, which is an important technical path for promoting the large-scale and high-quality development of energy storage systems.

[0056] Preferably, the multi-objective function is specifically: ; Wherein, , , These are the weighting coefficients. To optimize the time period, and They are respectively Discharge power and charging power at any given time, and for The electricity purchase price and electricity sales price at any given time. for State of charge at time t, For the target SOC value, The net power exchanged between the energy storage system and the grid. This is the reference power for grid demand.

[0057] The technical advantage of the aforementioned multi-objective function lies in achieving coordinated, adaptive, and multi-dimensional optimization of energy storage systems in complex grid environments. It doesn't simply mechanically stack three objectives; rather, it organically integrates and dynamically balances operational objectives belonging to different dimensions (economics, grid support, and equipment health) that often conflict with each other through a unified mathematical framework. Specifically, it systematically and simultaneously considers the three key dimensions of economic benefits, grid support, and battery health through a unified objective function. This avoids short-sighted "one-sided" behavior from a system architecture perspective, achieving a leap in overall operational efficiency. (Weighting coefficients) , , The dynamic adjustability is a prominent and substantial feature of this solution. It makes the optimization strategy no longer static and rigid, but rather capable of adaptive adjustment based on external market conditions (electricity price), real-time grid status (frequency, voltage), and internal equipment health (SOH). This endows the energy storage system with intelligent decision-making capabilities and robustness in the face of uncertainty and changing operating conditions.

[0058] This is further reflected in: through dynamic adjustment This invention enables energy storage systems to adaptively respond to real-time fluctuations in the electricity market. During periods of high electricity price volatility or significant forecast uncertainty, the system automatically reduces the weight of economic objectives to mitigate potential risks; conversely, when market conditions are favorable, it increases the weight to maximize returns. This represents an upgrade from a static, passive economic model to a dynamic, proactive risk-return management strategy, enhancing the economic resilience and long-term profitability stability of energy storage systems in real-world market environments. Through weighting coefficients Through dynamic adjustment, this invention achieves intelligent coordination between the grid support strength and the operational safety boundary of the energy storage device itself. When the grid experiences significant frequency deviations or voltage fluctuations, the system can automatically adjust... The power support demand of the power grid is preferentially met, and meanwhile, by comprehensively considering the current operation capability (such as the maximum charging and discharging power and the temperature) of the equipment, it is ensured that the support behavior will not cause the equipment to be overloaded, thereby effectively prolonging the service life of the energy storage equipment under the premise of guaranteeing the safe and stable operation of the power grid, and solving the contradiction between the support intensity and the equipment safety. By introducing the SOC target value and its deviation penalty, the battery health management is "embedded" in the real-time control algorithm rather than being remedied afterwards. More importantly, the weight coefficient γ can be dynamically adjusted according to the actual health state (SOH) of the battery. For example, for an aged battery with a low SOH, the system can automatically increase the value of γ to implement a more conservative and protective charging and discharging strategy, thereby differentially prolonging the effective life of each battery cluster. This realizes the transition from "one-size-fits-all" extensive management to "state-based predictive fine management", significantly improving the life cycle economy and long-term operation reliability of the energy storage system.

[0059] Preferably, the power overrun processing mechanism comprises the following steps: When any battery cluster reaches its power upper limit in the power allocation process, the power of the battery cluster is fixed as the upper limit value; The battery cluster is excluded from the current power allocation process, and the remaining to-be-allocated power is redistributed to the remaining battery clusters based on the updated dynamic weight until the power allocation is completed or all battery clusters reach the power upper limit.

[0060] The specific implementation process of the power overrun processing mechanism includes: triggering and locking: when any battery cluster reaches its power upper limit in the power allocation process, the power of the battery cluster is immediately fixed as the upper limit value. Exclusion and reconstruction: the battery cluster that has reached the full load is excluded from the current allocation list. Iteration and redistribution: the "remaining to-be-allocated power" is redistributed to the remaining battery clusters based on the "updated dynamic weight". Termination condition: the above steps are executed in a loop until the power is fully allocated or all battery clusters reach their power upper limit.

[0061] The power overrun processing mechanism, under the premise of absolutely guaranteeing the safety of the equipment, dynamically and maximally utilizes the remaining capacity of the system through iterative redistribution to complete the total power instruction with the optimal strategy, thereby greatly improving the reliability, availability and economy of the energy storage system.

[0062] In one embodiment, the proportional allocation method of the charging and discharging process is an energy management strategy suitable for systems that need to balance system balance and operation efficiency, and performs excellently especially in a battery group with gradient utilization or a battery cluster with large differences in aging degree. This method realizes the optimization of state balance among batteries and overall system efficiency by dynamically adjusting the charging and discharging power allocation proportion.

[0063] The nonlinear function is constructed for the first battery unit charging process ; wherein, is a charging adjustment factor (typically in the range of 1.5-2.0), which determines the sensitivity of the system to the imbalance of SOC, and the greater the value, the more significant the charging tilt for the low-SOC battery.

[0064] discharging process ; wherein, is a discharging adjustment factor (typically in the range of 1.0-1.2), which can achieve a good balance between SOC balancing and efficiency.

[0065] It should be noted that the adjustment factor selection of the charging and discharging function proportional distribution method needs to comprehensively consider the system balancing demand, operating efficiency and battery aging characteristics: when the SOC difference between the batteries is large, a larger adjustment factor should be selected to strengthen the balancing effect; however, too high a factor will increase the conduction loss and reduce the system efficiency, so a compromise needs to be made; for a seriously aged battery pack, a gentler distribution strategy should be adopted to reduce the stress impact. This method dynamically adjusts the power distribution through an exponential function, prioritizes charging for low-SOC batteries and prioritizes discharging for high-SOC batteries, while automatically adapting to the dynamic aging differences of the battery pack, optimizing the overall operating efficiency while ensuring balancing.

[0066] The dynamic proportional distribution of the energy storage system is: charging proportional distribution ; discharging proportional distribution ; Through real-time adjustment of the charging and discharging proportion, the lower the SOC of the battery, the higher the charging weight, accelerating the balancing; the higher the SOC of the battery, the higher the discharging weight, delaying the further degradation of the low-SOC battery.

[0067] The dynamic power distribution of the energy storage system can be represented as: When charging, the power allocated to the first energy storage battery cluster is: ; wherein, is the total charging demand power of the system.

[0068] When discharging, the power allocated to the first energy storage battery cluster is: ; wherein, is the total discharging demand power of the system.

[0069] The charging and discharging power distribution of the energy storage system is self-adaptive balanced by dynamically adjusting the weights, and automatically adapts to the SOC difference and aging degree between the batteries; the efficiency optimization relies on the reasonable selection of the charging and discharging factors to balance between the balancing speed and system loss; at the same time, the aging battery is significantly protected, the deep discharge risk is reduced by reducing the discharge load of the low SOC battery, thereby prolonging the overall life. This method is especially suitable for cascade utilization battery pack or energy storage system with different aging degrees, maintains the state balance between the batteries while ensuring the efficient operation of the system. 、

[0070] The algorithm execution process of the charging and discharging power dynamic distribution is as follows: The upper optimizer is a power pre-distribution layer, which realizes fast power pre-distribution based on a priority sorting algorithm, meets the dynamic response demand of the system, and outputs the basic reference power value of each group of string type PCS ; in the priority response logic, the charging scene is sorted in ascending order of SOC, and the low SOC battery is given priority, and the discharging scene is sorted in descending order of SOC, and the high SOC battery is given priority.

[0071] The lower controller is a power dynamic balancing layer, which dynamically adjusts the power based on the charging and discharging function dynamic adjustment mechanism, and the target adjustment power is: ; Among them, the power dynamic correction term can be expressed as: ; By adopting a PI regulator (the proportional coefficient and the integral coefficient are 0.5 and 0.01 respectively) to track the power instruction, closed-loop control is performed to suppress steady-state error.

[0072] When the real-time power is executed, the safety boundary protection logic needs to meet: ; Among them, is the instantaneous power capability based on the state parameters of the energy storage battery cluster SOP, temperature and aging degree, is the temperature adaptive power limit value.

[0073] The energy storage system dynamically switches the control mode according to the real-time operating state to realize adaptive balancing regulation in multiple scenes.

[0074] The charging mode is specifically: preferentially improving the low SOC branch current, up to 120% of the nominal value, and limiting the high SOC branch to 80% of the nominal value.

[0075] ​The discharge mode is specifically: high SOC branch dominant discharge, low SOC branch backup (SOC >= 20% enabled), and the power limit of the battery is 70% of the nominal value when SOH < 80%.

[0076] As shown in Figure 6 A power storage battery system charge-discharge control system comprises: A parameter acquisition module is configured to construct a state space model, acquire voltage, current and temperature measurement values of the power storage battery system in real time according to the state space model, and acquire state of charge and available power state of each battery cluster by using an extended Kalman filtering algorithm. A dynamic weight acquisition module is configured to acquire charge-discharge dynamic weights of each battery cluster in different modes by using a dynamic sorting algorithm according to the state of charge and available power state of each battery cluster. A target charge-discharge power determination module is configured to determine target charge-discharge power of each battery cluster under a preset condition constraint according to a preset multi-objective optimization function and the charge-discharge dynamic weights of each battery cluster. A charge-discharge control module is configured to independently control charge and discharge of each battery cluster by using a group string type energy storage converter according to the target charge-discharge power of each battery cluster.

[0077] The application also discloses a power storage battery system charge-discharge control system, which aims to realize charge-discharge control of the power storage battery system in a modular manner. The system constructs a state space model by using a parameter acquisition module, acquires voltage, current and temperature measurement values of the power storage battery system in real time according to the state space model, and acquires state of charge and available power state of each battery cluster by using an extended Kalman filtering algorithm. The system acquires charge-discharge dynamic weights of each battery cluster in different modes by using a dynamic sorting algorithm according to the state of charge and available power state of each battery cluster through a dynamic weight acquisition module. The system determines target charge-discharge power of each battery cluster under a preset condition constraint according to a preset multi-objective optimization function and the charge-discharge dynamic weights of each battery cluster through a target charge-discharge power determination module. The system independently controls charge and discharge of each battery cluster by using a group string type energy storage converter according to the target charge-discharge power of each battery cluster through a charge-discharge control module.

[0078] One or more embodiments in the application are intended to cover all such alternatives, modifications and variations falling within the broad scope of the application. Therefore, any omissions, modifications, equivalent replacements, improvements, etc. made within the spirit and principles of one or more embodiments in the application should be included in the protection scope of the application.

[0079] Those skilled in the art can understand that all or part of the steps of the above-mentioned method embodiments can be completed by program instructions and related hardware. The aforementioned program instructions can be stored in a computer readable storage medium. When the program instructions are executed, the steps of the above-mentioned method embodiments are executed. The aforementioned storage medium includes a mobile storage device, a read only memory (ROM), a magnetic disc or an optical disc, and various media that can store program codes.

[0080] Hereinafter, the terms "first" and "second" are only used for the purpose of description, and cannot be understood as indicating or implying relative importance or implicitly indicating the number of the indicated technical features. Therefore, the features defined with "first" and "second" can explicitly or implicitly include one or more of the features.

[0081] If a flowchart is used in the present application, the flowchart is used to illustrate the operations performed by the system according to the embodiments of the present application. It should be understood that the preceding or subsequent operations are not necessarily executed in sequence. On the contrary, each step can be processed in reverse order or simultaneously. Meanwhile, other operations can be added to these processes, or one or more steps of the operations can be removed from these processes.

[0082] The above provides a kind of energy storage battery system charge-discharge control method and system provided in the present application in detail.The above description of the disclosed embodiments enables those skilled in the art to implement or use the present application.Modifications of these embodiments will be apparent to those skilled in the art, and the general principles defined herein can be implemented in other embodiments without departing from the spirit or scope of the present application.The present application will not be limited to the embodiments shown herein, but will conform to the widest range consistent with the principles and novel features disclosed herein.

Claims

1. A charging and discharging control method for an energy storage battery system, characterized in that, Includes the following steps: A state-space model is constructed. Based on the state-space model, the voltage, current, and temperature measurements of the energy storage battery system are obtained in real time. The extended Kalman filter algorithm is used to obtain the state of charge and available power state of each battery cluster. Based on the state of charge and available power state of each battery cluster, a dynamic sorting algorithm is used to calculate and obtain the dynamic weights of charging and discharging of each battery cluster under different modes. Based on the preset multi-objective optimization function and the charging and discharging dynamic weights of each battery cluster, the target charging and discharging power of each battery cluster is determined under preset constraints. Based on the target charge and discharge power of each battery cluster, the charging and discharging of each battery cluster is independently controlled by a string energy storage converter.

2. The charging and discharging control method for an energy storage battery system as described in claim 1, characterized in that, The state-space model is constructed based on Thevenin's equivalent circuit theory, specifically as follows: ； ； in, for The system state vector at time t. This is a control input vector that includes measured values ​​of voltage, current, and temperature. For process noise, It is a nonlinear state transition function. For the observation vector, To observe the noise, For observation functions.

3. The charging and discharging control method for an energy storage battery system as described in claim 2, characterized in that, The step of obtaining the state of charge and available power state of each battery cluster using the extended Kalman filter algorithm includes the following steps: An extended Kalman filter is used to estimate the state of the state-space model, and the state estimate is obtained. The state of charge and available power state of each battery cluster are obtained through the extended Kalman filter and the state estimate.

4. The charging and discharging control method for an energy storage battery system as described in claim 3, characterized in that, The step of using extended Kalman filtering to estimate the state space model and obtain the state estimate value includes the following steps: Prediction steps: Based on the system state transition matrix and process noise covariance matrix, calculate the predicted state value and the prediction covariance matrix, using the following formula: ； ； in, It is the first state estimate. It is the first covariance matrix. It is the state transition matrix. It is the covariance matrix of the process noise; Update steps: Based on the observation matrix and the observation noise covariance matrix, calculate the Kalman gain, and fuse the observations including the state of charge and the available power state. Update the first estimate to the second state estimate and update the first covariance matrix to the second covariance matrix, as shown in the formula: ； ； ； in, For Kalman gain, For the observation matrix, It is the observation noise covariance matrix. This is the second state estimate. The second covariance matrix, It is an identity matrix.

5. The charging and discharging control method for an energy storage battery system as described in claim 1, characterized in that, The step of calculating the dynamic charging and discharging weights of each battery cluster under different modes using a dynamic sorting algorithm based on the state of charge and available power state of each battery cluster includes the following steps: Based on the operating instructions of the energy storage system, determine the type of operating mode currently in which the energy storage system is located; If the current energy storage system is in charging mode, the battery clusters are sorted in ascending order of their state of charge, and the battery clusters with lower states of charge that are ranked higher are given higher charging priority. If the current energy storage system is in discharge mode, the battery clusters are sorted in descending order of their state of charge, and the battery clusters with higher state of charge at the top of the list are given higher discharge priority. Based on the sorting priority, the available power state of each battery cluster, and the preset weighting formula, the dynamic weight of charging and discharging of each battery cluster is calculated, wherein higher priority battery clusters are assigned higher power allocation weights. The preset weight formula is specifically as follows: ； in, For the first The dynamic weight of each battery cluster For the target state of charge, To obtain the first digit in real time using the extended Kalman filter algorithm Real-time state of charge of each battery cluster For the first The rated capacity of each battery cluster This is the temperature influence coefficient. For the first Real-time temperature of each battery cluster For ideal operating temperature.

6. The charging and discharging control method for an energy storage battery system as described in claim 5, characterized in that, After determining the current operating mode of the energy storage system based on its operating instructions, the following steps are also included: If the current energy storage system is in standby mode, each battery cluster is controlled to perform low-power energy transfer at a preset low ratio of the system's rated power. The balancing current used for energy transfer is controlled within a preset current range to achieve autonomous balancing of the state of charge among the battery clusters. If the current energy storage system is in fault mode, the status of each battery cluster is monitored in real time. When the difference in state of charge between battery clusters is found to be greater than the first preset threshold, a system alarm is issued. When the difference in state of charge between battery clusters is greater than the second preset threshold, the battery cluster with abnormal state of charge is disconnected and enters the derating operation state. Wherein, the first preset threshold is less than the second preset threshold.

7. The charging and discharging control method for an energy storage battery system as described in claim 1, characterized in that, The step of determining the target charge / discharge power of each battery cluster based on a preset multi-objective optimization function and the dynamic charge / discharge weights of each battery cluster, under preset constraints, includes the following steps: Based on a preset multi-objective function, with the goal of synergistically improving the operating economy of the energy storage system and the grid support capability, the total charging and discharging power demand of the system during the optimization period is solved. Based on the total charging and discharging power demand of the system and the dynamic charging and discharging weights of each battery cluster, the target charging and discharging power of each battery cluster is determined through an iterative allocation algorithm, under the constraints of battery cluster safety limits, power capacity of the energy storage converter, and power over-limit handling mechanism.

8. The charging and discharging control method for an energy storage battery system as described in claim 7, characterized in that, The multi-objective function is specifically: ； in, , , These are the weighting coefficients. To optimize the time period, and They are respectively Discharge power and charging power at any given time, and for The electricity purchase price and electricity sales price at any given time. for State of charge at time t, For the target SOC value, The net power exchanged between the energy storage system and the grid. This is the reference power for grid demand.

9. The charging and discharging control method for an energy storage battery system as described in claim 7, characterized in that, The power over-limit handling mechanism includes the following steps: When any battery cluster reaches its power limit during power distribution, the power of that battery cluster is fixed at the limit value. The battery cluster is excluded from the current power allocation process, and the remaining power to be allocated is redistributed to the other battery clusters based on the updated dynamic weights, until the power allocation is completed or all battery clusters have reached their power limits.

10. A charging and discharging control system for an energy storage battery system, characterized in that, include: The parameter acquisition module is used to construct a state space model, and based on the state space model, to acquire the voltage, current and temperature measurements of the energy storage battery system in real time, and to use the extended Kalman filter algorithm to obtain the state of charge and available power state of each battery cluster. The dynamic weight acquisition module is used to calculate and obtain the dynamic weight of charging and discharging of each battery cluster under different modes based on the state of charge and available power state of each battery cluster using a dynamic sorting algorithm. The target charge / discharge power determination module is used to determine the target charge / discharge power of each battery cluster based on a preset multi-objective optimization function and the charge / discharge dynamic weights of each battery cluster, under preset constraints. The charge / discharge control module is used to independently control the charge / discharge of each battery cluster through a string energy storage converter, based on the target charge / discharge power of each battery cluster.

Images