Real-time power prediction method for flow batteries

Abstract

This application provides a real-time power prediction method for a flow battery, comprising: during the charging process of the flow battery, obtaining the current maximum rechargeable power of the flow battery according to a power prediction model; obtaining a first parameter, the first parameter including the total battery voltage and the battery state of charge; and reducing the current maximum rechargeable power according to the first parameter. This application obtains the current maximum rechargeable power of the flow battery according to the power prediction model and reduces the current maximum rechargeable power according to the first parameter. On the one hand, this improves the accuracy of power prediction during the charging process of the flow battery; on the other hand, it can reduce the current maximum rechargeable or dischargeable power in advance before entering the constant voltage charge / discharge mode, avoiding the problem of prolonged side reactions in the battery module due to long-term operation in the constant voltage charge / discharge mode, which could lead to hydrogen evolution, oxygen evolution, and other issues that reduce the battery's lifespan.

Description

Technical Field

[0001] This application mainly relates to the field of flow battery technology, and specifically to a method for real-time power prediction of a flow battery. Background Technology

[0002] During the charging and discharging process of a flow battery, it undergoes a transition from constant power charging / discharging to constant voltage charging / discharging. The maximum chargeable power or maximum dischargeable power of a flow battery exhibits a non-linear relationship with the charging / discharging time. With long-term charge-discharge cycles, the charging and discharging power also exhibits complex non-linear decay characteristics, causing the aforementioned non-linear relationship to change.

[0003] Current Battery Management Systems (BMS) primarily use fixed curve models to predict the power output of flow batteries during charging and discharging. This method has the following technical problems: 1) It cannot update the curve model in real time based on time-varying factors such as battery degradation and temperature changes, thus failing to accurately describe the aforementioned nonlinear relationships. This results in a significant deviation between the predicted power (maximum chargeable power or maximum dischargeable power) and the actual power. 2) The battery operates in constant voltage charge and discharge mode for extended periods, leading to prolonged side reactions in the battery module and increasing the risk of hydrogen and oxygen evolution, thereby reducing battery lifespan. Summary of the Invention

[0004] This application provides a real-time power prediction method for flow batteries to solve the aforementioned technical problems, enabling real-time and accurate prediction of the maximum chargeable power or maximum dischargeable power of the flow battery under constant voltage conditions. The technical solution adopted by this application to solve the aforementioned technical problems is a real-time power prediction method for flow batteries. The method includes: during the charging process of the flow battery, obtaining the current maximum chargeable power of the flow battery according to a power prediction model; obtaining a first parameter, the first parameter including the total battery voltage and the battery state of charge; and reducing the current maximum chargeable power according to the first parameter.

[0005] In one embodiment of this application, reducing the current maximum chargeable power according to the first parameter includes: in response to the total battery voltage being greater than the upper limit threshold of the total battery voltage and the duration reaching a first preset duration, subtracting a first preset value from the current maximum chargeable power.

[0006] In one embodiment of this application, the first parameter further includes liquid circuit voltage or battery module voltage, and the step of reducing the current maximum chargeable power according to the first parameter further includes: in response to the liquid circuit voltage being greater than a preset threshold for liquid circuit voltage and lasting for a second preset duration, or the battery module voltage being greater than a preset threshold for battery module voltage and lasting for a third preset duration, the current maximum chargeable power is reduced by the second preset value.

[0007] In one embodiment of this application, the second preset duration is less than the first preset duration, and the third preset duration is less than the first preset duration.

[0008] In one embodiment of this application, reducing the current maximum chargeable power according to the first parameter includes: in response to the battery state of charge being greater than the upper limit threshold of the battery state of charge, the current maximum chargeable power is zero.

[0009] To address the aforementioned technical problems, this application also provides a real-time power prediction method for a flow battery. The method includes: during the discharge process of the flow battery, obtaining the current maximum discharge power of the flow battery according to a power prediction model; obtaining a second parameter, the second parameter including the total battery voltage and the battery state of charge; and reducing the current maximum discharge power according to the second parameter.

[0010] In one embodiment of this application, reducing the current maximum discharge power according to the second parameter includes: in response to the total battery voltage being less than the lower limit threshold of the total battery voltage and the duration reaching a fourth preset duration, subtracting a third preset value from the current maximum discharge power.

[0011] In one embodiment of this application, the second parameter further includes the liquid circuit voltage, and the step of reducing the current maximum discharge power according to the second parameter further includes: in response to the liquid circuit voltage decreasing at a rate greater than a decreasing rate threshold and lasting for a duration of a fifth preset duration, the current maximum discharge power is reduced by a fourth preset value.

[0012] In one embodiment of this application, the fifth preset duration is less than the fourth preset duration.

[0013] In one embodiment of this application, reducing the current maximum discharge power according to the second parameter includes: in response to the battery state of charge being less than the battery state of charge lower limit threshold, the current maximum discharge power is zero.

[0014] In one embodiment of this application, the power prediction model includes:

[0015] ,

[0016] Where a, b, and c are all parameters to be identified. The maximum chargeable power or the maximum dischargeable power is the current maximum chargeable power, and SOC is the state of charge of the battery.

[0017] In one embodiment of this application, the battery state of charge is calculated using the following formula:

[0018] ,

[0019] in, Let be the state of charge of the battery at time t. Let be the open-circuit voltage of the flow battery at time t. Let i be a coefficient, where i is an integer from 0 to N, and N is a positive integer greater than or equal to 1.

[0020] In one embodiment of this application, the method further includes: acquiring target data, the target data including the current maximum chargeable power or the current maximum dischargeable power under a preset battery state of charge; and updating the parameters to be identified in the power prediction model based on the target data.

[0021] In one embodiment of this application, updating the parameters to be identified in the power prediction model based on the target data includes updating the parameters to be identified using the following formulas:

[0022] ;

[0023] ;

[0024] ;

[0025] in, The parameter vector is composed of the parameters to be identified. , Forgetting factor, Let K(k) be the target data at time k, K(k) be the gain matrix at time k, P(k) be the covariance matrix at time k, and ϕ(k) be the data vector of SOC at time k. =[SOC 2 ,SOC,1].

[0026] In one embodiment of this application, after obtaining the target data, the method further includes: compensating the target data according to the current operating flow of the battery system, including: in response to the current operating flow of the battery system being greater than the theoretical operating flow corresponding to the current state of charge of the battery, the target data is subtracted by a fifth preset value; in response to the current operating flow of the battery system being less than the theoretical operating flow corresponding to the current state of charge of the battery, the target data is increased by a sixth preset value.

[0027] In one embodiment of this application, the target data includes multiple current maximum chargeable power or current maximum dischargeable power of a preset battery state of charge. After obtaining the target data, the method further includes: assigning weights to the target data, wherein the weight of the target data at the current moment is greater than the weight of the target data at the previous moment.

[0028] In one embodiment of this application, the power prediction model includes: a data acquisition layer for acquiring the target data; a data processing layer for processing the target data; a model iteration layer for updating the parameters to be identified in the power prediction model based on the target data; and a control output layer for enabling the battery management system to acquire the current maximum chargeable power and / or the current maximum dischargeable power based on the power prediction model, and to provide the current maximum chargeable power and / or the current maximum dischargeable power to the energy storage converter and / or the energy management system for power plant scheduling.

[0029] The real-time power prediction method and system of this application can obtain the current maximum chargeable power of the flow battery according to the power prediction model, and reduce the current maximum chargeable power according to the first parameter. On the one hand, it improves the accuracy of power prediction during the charging process of the flow battery. On the other hand, it can reduce the current maximum chargeable power in advance before entering the constant voltage charging mode, thereby delaying the time for the flow battery voltage to reach the voltage threshold for the transition from constant power charging to constant voltage charging. That is, it delays the time for the flow battery to enter the constant voltage charging mode, thereby shortening the running time of the flow battery in constant voltage charging. This avoids the problem of long-term operation of the battery in constant voltage charging mode, which leads to long duration of side reactions in the battery module, resulting in hydrogen evolution, oxygen evolution, and other issues that reduce the battery's service life. Attached Figure Description

[0030] To make the above-mentioned objectives, features and advantages of this application more apparent and understandable, the specific embodiments of this application will be described in detail below with reference to the accompanying drawings, wherein:

[0031] Figure 1 It is a charging and discharging operation flowchart;

[0032] Figure 2 This is a flowchart of a real-time power prediction method during the charging process according to an embodiment of the present application.

[0033] Figure 3 This is a flowchart of a real-time power prediction method during the discharge process according to an embodiment of the present application.

[0034] Figure 4 This is a graph showing the relationship between SOC and power in a real-time power prediction method according to an embodiment of this application.

[0035] Figure 5 This is a simplified block diagram of the power prediction model in a real-time power prediction method according to an embodiment of this application;

[0036] Figure 6 This is an interactive flowchart of the power prediction model update in a real-time power prediction method according to an embodiment of this application;

[0037] Figure 7 This is a flowchart illustrating the reduction of the current maximum chargeable power in a real-time power prediction method according to an embodiment of this application;

[0038] Figure 8 This is a flowchart illustrating the reduction of the current maximum amplifiable power in a real-time power prediction method according to an embodiment of this application;

[0039] Figure 9 This is a charging and discharging flowchart of a real-time power prediction method according to an embodiment of this application;

[0040] Figure 10 This is a flowchart of a real-time power prediction method according to another embodiment of this application. Detailed Implementation

[0041] To make the above-mentioned objectives, features and advantages of this application more apparent and understandable, the specific embodiments of this application will be described in detail below with reference to the accompanying drawings.

[0042] Many specific details are set forth in the following description in order to provide a full understanding of this application. However, this application may also be implemented in other ways different from those described herein, and therefore this application is not limited to the specific embodiments disclosed below.

[0043] As illustrated in this application, unless the context clearly indicates otherwise, the words "a," "an," "an," and / or "the" do not specifically refer to the singular and may also include the plural. Generally speaking, the terms "comprising" and "including" only indicate the inclusion of explicitly identified steps and elements, which do not constitute an exclusive list, and the method or apparatus may also include other steps or elements.

[0044] Flowcharts are used in this application to illustrate the operations performed according to the embodiments of this application. It should be understood that the preceding or following operations are not necessarily performed in exact order. Instead, various steps can be processed in reverse order or simultaneously. Furthermore, other operations may be added to these processes, or one or more steps may be removed from these processes.

[0045] Within the flow battery industry, most manufacturers follow the same battery system charging and discharging process as the newly released national standard GB / T 33339-2025, "Test Methods for Vanadium Redox Flow Batteries," published in 2025. Currently, the mainstream flow battery manufacturers use the following overall charging and discharging process: the flow battery is first charged and discharged at rated power (constant power charging and discharging mode). Once the charging and discharging power boundary conditions are met (i.e., the total voltage of the battery system reaches the constant voltage charging voltage or constant voltage discharging voltage), it then enters the constant voltage charging and discharging mode.

[0046] Figure 1 A flow chart illustrating the charging and discharging operation of a flow battery is shown. Figure 1As shown, after the flow battery charging and discharging begins, in step S11: the total battery voltage (BATV) is detected. During the charging phase, proceed to step S12: determine if BATV is greater than the constant voltage charging set value. If yes, proceed to step S13; otherwise, proceed to step S18. Step S13: The PCS switches from constant power charging to constant voltage charging. That is, the PCS activates the constant voltage charging mode limit, and at this time, the PCS no longer executes the power commands issued by the Energy Management System (EMS) and instead executes the constant voltage charging mode. Until the constant voltage charging is completed and the battery is fully charged, proceed to step S14. Step S14: The BMS reports that the battery is fully charged. During the discharging phase, proceed to step S15: determine if BATV is less than the constant voltage discharging set value. If yes, proceed to step S16; otherwise, proceed to step S18. Step S16: The PCS switches from constant power discharging to constant voltage discharging. The PCS activates the constant voltage discharging mode limit, and at this time, the PCS no longer executes the power commands issued by the EMS and instead executes the constant voltage discharging mode. The constant voltage discharge continues until the battery is completely discharged, then step S17 is executed. Step S17: The BMS reports that the battery is empty. Step S18: When the constant voltage charging setpoint > BATV > constant voltage discharging setpoint, the PCS maintains the current active power command. When the BMS reports that the battery is fully charged in step S14 or reports that the battery is empty in step S17, the charging and discharging behavior ends. Afterwards, the EMS monitors the BMS telemetry and telecontrol data and no longer responds to power transmission. Some electrochemical energy storage power stations require the EMS to monitor the status of the BMS and adjust the power transmission of the PCS according to the status of the BMS.

[0047] While the above-described charging and discharging process allows the battery to operate at maximum power throughout the entire process, shortening the charging and discharging time from 0% to 100% and reducing the power consumption of auxiliary equipment (including the BMS) of the battery module, the following problems still exist: 1) In constant voltage charging and discharging mode, the DC active power of the battery module can only be limited by the battery voltage in the constant power-to-constant voltage mode of the PCS to output power, and it is impossible to accurately predict the maximum chargeable or dischargeable DC active power of the battery system at this moment, which is inconvenient for power station operation and grid dispatch. For flow battery energy storage projects, whether it is new energy distribution storage (generation side) or independent energy storage power station (grid side), the energy storage power station is required to pass the "GB / T 36547-2024 Technical Specification for Electrochemical Energy Storage Power Station Grid Connection" and "GB / T 36548-2024 Test Specification for Electrochemical Energy Storage Power Station Grid Connection". Both of the above-mentioned specifications have assessment items and requirements for the active power control accuracy and control deviation of electrochemical energy storage power stations. For example, for energy storage power stations participating in the electricity spot market, the power grid requires them to report their self-planned output every 15 minutes, that is, to report the maximum chargeable and maximum dischargeable power of the energy storage power station every 15 minutes. Furthermore, the grid dispatching system will also assess the active power control deviation based on the power reported by the energy storage power station. The assessment standard is that the actual power output of the energy storage power station should not exceed ±2% of the reported plan; 2) When the battery operates at a high charging or discharging power in constant voltage charging and discharging mode, the side reactions of the battery module will last longer, easily leading to hydrogen and oxygen evolution, etc. Concentration polarization will occur at the end of constant voltage discharge, causing inconsistent voltage in the stack modules, thereby reducing the battery's lifespan.

[0048] In addition, traditional methods for predicting maximum chargeable power or maximum dischargeable power rely on the fixed curve model of the BMS. The power prediction process cannot adapt to time-varying factors such as battery degradation and temperature changes, resulting in a large deviation between the predicted results and the actual power.

[0049] To address the aforementioned technical problems, this application proposes a real-time power prediction method for flow batteries. This method is applicable to real-time power prediction before entering the constant voltage stage in the charge and discharge modes of the aforementioned flow batteries.

[0050] Figure 2 A flowchart illustrating a real-time power prediction method according to an embodiment of this application is shown, illustrating the real-time power prediction method during the charging process. Figure 2 As shown, the real-time power prediction method for the charging process in this embodiment includes:

[0051] Step S21: During the charging process of the flow battery, obtain the current maximum chargeable power of the flow battery according to the power prediction model;

[0052] Step S22: Obtain the first parameter, which includes the total battery voltage and the battery state of charge; and

[0053] Step S23: Reduce the current maximum chargeable power according to the first parameter.

[0054] Figure 3 A flowchart illustrating a real-time power prediction method during the discharge process, according to an embodiment of this application, is shown. Figure 3 As shown, the real-time power prediction method for the discharge process in this embodiment includes:

[0055] Step S31: During the discharge process of the flow battery, obtain the current maximum discharge power of the flow battery according to the power prediction model;

[0056] Step S32: Obtain the second parameter, which includes the total battery voltage and the battery state of charge; and

[0057] Step S33: Reduce the current maximum amplifiable power according to the second parameter.

[0058] It should be noted that, in one embodiment, the real-time power prediction method can be used simultaneously during the charging and discharging processes, in which case the method includes steps S21-S23 and S31-S33. The power prediction method of this application can be used as needed during the charging and / or discharging processes of the flow battery.

[0059] This application obtains the current maximum chargeable power of the flow battery based on a power prediction model and reduces it according to a first parameter; and / or, obtains the current maximum dischargeable power of the flow battery based on a power prediction model and reduces it according to a second parameter. This improves the accuracy of power prediction during the charging and discharging process. Furthermore, it reduces the current maximum chargeable or dischargeable power before entering the constant-voltage charging and discharging mode, thus delaying the battery voltage from reaching the voltage threshold for transitioning from constant-power to constant-voltage charging and discharging. This delays the time it takes for the flow battery to enter the constant-voltage charging and discharging mode, thereby shortening its operating time and preventing prolonged operation in this mode, which can lead to prolonged side reactions in the battery module, resulting in hydrogen and oxygen evolution and reduced battery life. Additionally, by reducing the current maximum dischargeable power before entering the constant-voltage discharge mode, it also avoids concentration polarization caused by high-power discharge at the end of the constant-voltage discharge phase, preventing voltage inconsistencies in the stack modules.

[0060] In some embodiments, the power prediction model in steps S21 and S31 is represented by the following formula (1):

[0061] (1)

[0062] Where a, b, and c are all parameters to be identified. This refers to the current maximum chargeable power or the current maximum dischargeable power. SOC refers to the battery's state of charge.

[0063] In some embodiments, SOC is calculated using the following formula (2):

[0064] (2)

[0065] Among them, SOC (t) Let SOC and OCV be the values ​​at time t. (t) Let be the open-circuit voltage of the flow battery at time t. Let i be a coefficient, i be a constant from 0 to N, and N be a positive integer greater than or equal to 1.

[0066] In some embodiments, The values ​​of N can be set according to the actual curve fitting effect in formula (2). In some embodiments, N=5.

[0067] In some embodiments, the real-time power prediction method of this application further includes: acquiring target data, which includes the current maximum chargeable power or the current maximum dischargeable power under a preset SOC; and updating the parameters to be identified in the power prediction model based on the target data. By updating the parameters to be identified in the power prediction model based on the target data, the prediction method of this application can correct the power prediction model according to the actual battery data under different environments (including battery degradation, temperature changes, etc.), further improving the accuracy of power prediction.

[0068] In some embodiments, during the constant-voltage charging phase of the battery, the preset state of charge (SOC) ranges from 70% to 95%, exemplarily including 70%, 75%, 80%, 85%, 90%, and 95%. During the constant-voltage discharging phase of the battery, the preset SOC ranges from 5% to 30%, exemplarily including 5%, 10%, 15%, 20%, 25%, and 30%. In some embodiments, the preset SOC is determined based on the inflection point of the battery power characteristics. For example, during the charging phase, when the battery's SOC reaches 80% and begins to show a significant power reduction, the SOC boundary condition is 80%, and six points evenly distributed within the 80%~100% (constant-voltage charging) range are selected as preset SOCs. In some embodiments, the number of preset SOCs can be set according to actual needs, with several preset SOCs evenly distributed within the constant-voltage charging range. In some embodiments, target data is stored using a queue. For example, a queue (FIFO) is configured for each preset SOC to store the 10 most recently obtained valid current maximum chargeable power or current maximum dischargeable power under the preset SOC. For example, when SOC=70%, a one-dimensional array [P0,P1,P2…,P9] of SOC_70_Power

[10] will be configured to store 10 valid maximum discharge power values. The initial value of the array is set by the BMS at the factory, and is generally selected as the value under the preset SOC under the battery charge and discharge characteristic curve. When the SOC reaches SOC=70% in each discharge process, a new target data (current maximum discharge power) will be cached. After the target data is determined to be valid, the P0 value will be discharged and the data will be placed into the array as P10. The one-dimensional array of SOC_70_Power

[10] will then become [P1,P2,P2…,P10]. When the data is invalid, it will be discarded and will not be entered into the array of SOC_70_Power

[10] .

[0069] In some embodiments, after obtaining the target data, the method further includes: determining the validity of the target data and filtering outliers in the target data. For example, the latest obtained target data can be compared with the average value of the current array. If the rate of change of the latest obtained target data exceeds 30%, it is considered outlier data and will be discarded.

[0070] In some embodiments, after acquiring the target data, the method further includes: compensating the target data according to the current operating flow of the battery system, including: in response to the current operating flow of the battery system being greater than the theoretical operating flow corresponding to the current SOC, decreasing the target data by a fifth preset value; and in response to the current operating flow of the battery system being less than the theoretical operating flow corresponding to the current SOC, increasing the target data by a sixth preset value.

[0071] In the above embodiments, the current operating flow rate of the battery system typically refers to the total flow rate of the electrolyte circulation system during battery system operation, including the electrolyte flow rate. The theoretical value of the operating flow rate is the applicable operating flow rate for a flow battery under a preset state of charge. In the above embodiments, this application compensates for the target data based on the current operating flow rate of the battery system. When the operating flow rate is greater than the theoretical value, the target data is decreased; when the operating flow rate is less than the theoretical value, the target data is increased. This compensates for the deviation in the obtained current maximum chargeable power or current maximum dischargeable power caused by the deviation between the actual operating flow rate and the theoretical value, improving the accuracy of the target data of the flow battery system under different flow rate conditions.

[0072] It should be understood that when the actual operating flow rate deviates from the theoretical optimal flow rate (whether it is too high or too low), it may lead to a distortion in the battery state assessment. Through a flow rate compensation mechanism: if the flow rate is higher than the theoretical value, the target data is appropriately lowered; if the flow rate is lower than the theoretical value, the target data is appropriately increased, ultimately making the system's output chargeable and dischargeable power closer to the actual usable capacity.

[0073] For example, when SOC=70%, the one-dimensional array corresponding to SOC_70_Power

[10] is suitable for a battery system operating flow rate of 40 m³ / h. When the actual operating flow rate of the battery system changes to 48 m³ / h, the new target data is obtained as P11. flow=48 At this point, because the operating traffic is higher than the theoretical value, P11 will be... flow=48 Low compensation, so that P11 flow=48 Convert the battery system's operating flow rate to P11, which is 40 m³ / h, and then sort it into the array.

[0074] In some embodiments, the fifth and sixth preset values ​​can be set according to the characteristics of the actual flow battery.

[0075] In some embodiments, the target data can also be compensated and corrected based on the relationship between the charge / discharge capacity of the flow battery and the electrolyte temperature.

[0076] In some embodiments, the target data includes multiple current maximum chargeable power or current maximum dischargeable power under a preset SOC. After obtaining the target data, the method further includes: assigning weights to the target data, wherein the weight of the target data at the current moment is greater than the weight of the target data at the previous moment.

[0077] In the above embodiments, for the charging phase, the target data at the current moment is the target data for this charging phase under the preset SOC, and the target data at the previous moment is the target data for the last charging phase under the preset SOC. Similarly, for the discharging phase, the target data at the current moment is the target data for this discharging phase under the preset SOC, and the target data at the previous moment is the target data for the last discharging phase under the preset SOC. Compared to the target data at the previous moment, the target data at the current moment is more recent data. By assigning higher weight to the more recent data in the charging and discharging phases, the time decay effect is reflected, so as to reflect the changes in battery state in real time and avoid errors caused by the lag of historical data.

[0078] In some embodiments, weighting can be achieved by adding time-related weights to the data when storing it in the array, for example, new data weight = 1, old data weight = 0.5 × time difference.

[0079] In some embodiments, updating the parameters to be identified in the power prediction model based on the target data includes updating the parameters to be identified by simultaneously using the following formulas (3)-(5):

[0080] (3)

[0081] (4)

[0082] (5)

[0083] in, A parameter vector consisting of the parameters to be identified. , Forgetting factor, Let K(k) be the target data at time k, K(k) be the gain matrix at time k, P(k) be the covariance matrix at time k, and ϕ(k) be the data vector of SOC at time k. =[SOC 2 ,SOC,1].

[0084] In the above embodiments, the FFRLS (Forgetting Factor RLS) algorithm, i.e., recursive least squares with a forgetting factor, is used to update the parameters to be identified in real time. The forgetting factor λ is a constant between 0 and 1 (typically 0.95~0.99). λ is introduced when updating the covariance matrix (P(k)) to address the problem of the traditional RLS algorithm's excessively long update time and the presence of zero regions in the P(k) matrix, effectively diminishing the influence of older data.

[0085] In some embodiments, =0.98.

[0086] Figure 4The diagram illustrates the relationship between State of Charge (SOC) and power in a real-time power prediction method according to an embodiment of this application. The first curve 41 has SOC on the horizontal axis and the current maximum chargeable power (in kW) on the vertical axis; the second curve 42 has SOC on the horizontal axis and the current maximum dischargeable power (in kW). Figure 5 As shown, during the charging phase, since the current maximum chargeable power is negative, the higher the SOC, the greater the current maximum chargeable power. During the discharging phase, the current maximum dischargeable power is positive, and the higher the SOC, the greater the current maximum dischargeable power. Therefore, the first curve 41 during the charging phase is an upward-opening curve, and the second curve 42 during the discharging phase is a downward-opening curve. Therefore, in some embodiments, during the discharging phase, the parameter update also needs to satisfy the physical constraint of a<0 in formula (1), that is, to ensure that the curve of SOC and power during the discharging phase opens downward. During the charging phase, the parameter update also needs to satisfy the physical constraint of a>0 in formula (1), that is, to ensure that the curve of SOC and power during the discharging phase opens upward.

[0087] In some embodiments, the current maximum chargeable power and the current maximum dischargeable power are both positive values, and both the charging and discharging phases must satisfy the physical constraint of a<0 in formula (1).

[0088] Figure 5 A simplified block diagram of the power prediction model in a real-time power prediction method according to an embodiment of this application is shown. Figure 5 As shown, in some embodiments, the power prediction model 50 includes: a data acquisition layer 51 for acquiring target data; a data processing layer 52 for processing the target data; a model iteration layer 53 for updating the parameters to be identified in the power prediction model based on the target data; and a control output layer 54 for enabling the BMS controller to acquire the current maximum chargeable power and / or the current maximum dischargeable power based on the power prediction model, and to provide the current maximum chargeable power and / or the current maximum dischargeable power to the energy storage converter and / or the energy management system for power plant scheduling.

[0089] In some embodiments, the data processing layer 52 is used to assign weights to the target data or compensate the target data according to fluid characteristics (i.e., the current operating flow rate or electrolyte temperature of the battery system as described above).

[0090] In some embodiments, the power prediction model 50 includes the following core modules: a data management module responsible for data acquisition, storage, and preprocessing; a parameter identification module that updates model parameters online based on the FFRLS algorithm; a power prediction module that predicts power based on the current SOC and model parameters; a safety monitoring module that ensures the predicted power is within a safe range; and a degradation compensation module that reduces model parameters based on the number of battery cycles to compensate for the effects of battery aging.

[0091] In some embodiments, to ensure the safe operation of the system, the following boundary conditions are set in the safety monitoring module: 1) Voltage boundary conditions include: Single stack voltage (CSV): charging ≤ 80.6V, discharging ≥ 50V; BATV: within the range of charging cutoff voltage and discharging cutoff voltage, i.e., 312*Num~484*Num; 2) SOC boundary conditions include: Power prediction effective range: for example, SOC>70% (charging), SOC<30% (discharging), where the power prediction effective range is determined according to the inflection point of the battery power characteristics; Charge / discharge termination conditions: SOC=100% (charging), SOC=0% (discharging); 3) Power boundary conditions include: Maximum power limit: the current maximum discharge power or the current maximum charge power does not exceed 105% of the rated power; Power change rate: the change in the current maximum discharge power or the current maximum charge power per minute does not exceed 20% of the rated power; 4) Fluid parameter boundary conditions include: Electrolyte temperature: 5℃-45℃; Electrolyte flow rate: Flow_Stack_Min*Stack_Num~ Flow_Stack_Min*Stack_Num, corresponding to 32~48 m³ / h for a 500kW system; recommended lower limit of fuel cell stack flow rate (Flow_Stack_Min); recommended upper limit of fuel cell stack flow rate (Flow_Stack_Min); number of fuel cell stacks in the same liquid path (Stack_Num). 5) Safety protection includes: voltage consistency: the deviation of single stack voltage from the average value does not exceed ±5%; power prediction deviation: the deviation from the actual maximum power does not exceed ±2%.

[0092] Figure 6 The diagram illustrates the interactive flowchart of power prediction model update in a real-time power prediction method according to an embodiment of this application.

[0093] like Figure 6 As shown, the power prediction model update process includes the following steps S61 to S69.

[0094] Step S61: Reach the SOC collection point. That is, the BMS detects that the current SOC is the preset SOC and proceeds to step S62.

[0095] Step S62: Send SOC / power data. That is, the data acquisition layer 51 collects the target data and SOC value and sends them to the model iteration layer 53.

[0096] Step S63: Update data storage. That is, model iteration layer 53 obtains new target data and updates the current data storage.

[0097] Step S64: Check update conditions. That is, the model iteration layer 53 checks whether the update conditions of the power prediction model 50 are met. If the update conditions are met, step S65 is executed; otherwise, step S69 is executed. The update conditions are: the new target data is valid or the new valid data has undergone compensation, weighting, or other processing.

[0098] Step S65: Constraint curve fitting. That is, the model iteration layer 53 refits the new power prediction model curve based on the target data and the FFRLS algorithm.

[0099] Step S66: Update the attenuation model. That is, the model iteration layer 53 updates the parameters to be identified in the power prediction model based on the new power prediction model curve from step S67.

[0100] Step S67: Send new parameters. That is, the model iteration layer 53 sends the parameters to be identified obtained in step S66 to the control output layer 54.

[0101] Step S68: Update the control curve model. That is, the control output layer 54 sends the updated parameters to be identified (i.e., the updated power prediction model) from step S66 to the BMS.

[0102] Step S69: Maintain current parameters. That is, the model iteration layer 53 does not change the parameters to be identified in the power prediction model and sends the parameters to be identified to the BMS.

[0103] In the above embodiment, when the current maximum chargeable power or current maximum dischargeable power predicted by the updated power prediction model is less than that of the original power prediction model, the updated power prediction model is used as an attenuation model. This attenuation model will be used as a new power prediction model for the next power prediction to meet the requirement of reducing the predicted current maximum chargeable power or current maximum dischargeable power.

[0104] It should be noted that the power prediction model and the process of updating the power prediction model described in the above embodiments are applicable to power prediction in both the charging and discharging stages. Subsequent embodiments will describe the power prediction methods for the charging and discharging stages separately.

[0105] Next, the power prediction method applicable to the charging phase will be explained in detail. In some embodiments, the step S23 of reducing the current maximum chargeable power according to the first parameter includes: in response to the total battery voltage being greater than the upper limit threshold of the total battery voltage for a duration of a first preset time, the current maximum chargeable power is subtracted from the first preset value. By dynamically monitoring the total battery voltage and its duration, when the total battery voltage exceeds the upper limit threshold for a duration of a first preset time, the maximum chargeable power is automatically reduced (subtracted from the first preset value). This allows the maximum chargeable power to be reduced before entering the constant voltage charging mode, thereby delaying the time for the flow battery voltage to reach the voltage threshold for the transition from constant power charge / discharge to constant voltage charge / discharge. In other words, it delays the time for the flow battery to enter the constant voltage charge / discharge mode, thereby shortening the operating time of the flow battery in constant voltage charge / discharge mode. This avoids the problem of prolonged side reactions in the battery module due to long-term operation in constant voltage charge / discharge mode, which can lead to hydrogen evolution, oxygen evolution, and other issues that reduce the battery's lifespan.

[0106] In some embodiments, the first preset duration is 30s-60s. In some embodiments, the first preset duration is set according to the characteristics of the flow battery. For example, if the flow battery is highly sensitive to voltage changes, that is, the total battery voltage of the flow battery changes rapidly during charging, and the total battery voltage can be detected more frequently, then the first preset duration can be set to be shorter.

[0107] In some embodiments, the upper limit threshold of the total battery voltage is 0 to 10V lower than the maximum value of the total battery voltage operating range.

[0108] In some embodiments, the first preset value is 1% to 2% of the rated power of the battery system.

[0109] In some embodiments, the first parameter further includes the liquid circuit voltage or the battery module voltage. Reducing the current maximum chargeable power based on the first parameter further includes: in response to the liquid circuit voltage exceeding a preset threshold for a duration of a second preset time, or the battery module voltage exceeding a preset threshold for a third preset time, subtracting a second preset value from the current maximum chargeable power. By dynamically monitoring the liquid circuit voltage and the battery module voltage, when either the liquid circuit voltage or the battery module voltage exceeds its preset threshold for a certain period of time, the current maximum chargeable power is automatically reduced (subtracting the second preset value). This allows for a reduction in the maximum chargeable power before entering constant voltage charging.

[0110] In some embodiments, the preset threshold for the liquid circuit voltage is 0 to 1.5V above the maximum value of the battery module voltage in its operating range, and the preset threshold for the battery module voltage is 0 to 3V above the maximum value of the battery module voltage in its operating range.

[0111] In some embodiments, the second preset duration and the third preset duration can be set according to the characteristics of the flow battery. In some embodiments, the second preset duration and the third preset duration are 15s-30s. In some embodiments, the second preset duration and the third preset duration can be the same or different.

[0112] In some embodiments, the second preset duration is shorter than the first preset duration, and the third preset duration is shorter than the first preset duration. In the above embodiments, the sensitivity of the total battery voltage to changes with charging and discharging is lower than the sensitivity of the liquid circuit voltage and the battery module voltage to changes with charging and discharging; therefore, the detection of the liquid circuit voltage and the battery module voltage is more frequent.

[0113] In some embodiments, the second preset value is less than the first preset value. It should be understood that when the total battery voltage is closer to the upper limit threshold of the total battery voltage, the flow battery will also enter constant voltage charging, therefore, the maximum chargeable power needs to be reduced further.

[0114] In some embodiments, reducing the current maximum chargeable power according to the first parameter includes: in response to the SOC being greater than the SOC upper limit threshold, the current maximum chargeable power is zero.

[0115] In some embodiments, the SOC upper limit threshold is set based on the charge / discharge capability of the flow battery.

[0116] In some embodiments, the upper limit of SOC is related to the application level of the energy storage power station. For typical parallel operation of multiple energy storage systems, the consistency of charge and discharge depth is considered. For example, the SOC of lithium batteries is generally 5%-95%, sodium batteries are 10%-90%, and flow batteries are 0%-100%. If the energy storage power station only uses flow batteries, the upper limit threshold can be set to 100%. If flow batteries, sodium batteries, and lithium batteries are used in parallel, the upper limit threshold can be set to the upper limit of sodium battery charging, which is 90%.

[0117] Figure 7 A flowchart illustrating the reduction of the current maximum chargeable power in a real-time power prediction method according to an embodiment of this application is shown. Figure 7 As shown, the battery charging phase includes the following steps to reduce the current maximum chargeable power.

[0118] Step S71: Detect whether the current total battery voltage is greater than the upper limit threshold of the total battery voltage. In response to the current total battery voltage being greater than the upper limit threshold of the total battery voltage, proceed to step S72;

[0119] Step S72: Time for 30 seconds. That is, if the current total battery voltage is greater than the upper limit threshold of the total battery voltage for 30 seconds, proceed to step S73, and at the same time, reset the timer to zero to monitor cyclically;

[0120] Step S73: Reduce the current maximum chargeable power by a first preset value. That is, when a pulse signal indicating that the current total battery voltage is greater than the upper limit threshold of the total battery voltage is detected for 30 seconds, the current maximum chargeable power is reduced by a first preset value.

[0121] Step S74: Detect whether the liquid circuit voltage LQV is greater than the liquid circuit voltage overvoltage LV3 protection value (i.e., whether the liquid circuit voltage is greater than the liquid circuit voltage preset threshold), or whether the battery module voltage MDV is greater than the battery module overvoltage LV3 protection value (i.e., whether the battery module voltage is greater than the battery module voltage preset threshold). If the liquid circuit voltage is greater than the liquid circuit voltage preset threshold, or the battery module voltage is greater than the battery module voltage preset threshold, proceed to step S75;

[0122] Step S75: Time for 15 seconds. That is, when the liquid circuit voltage is greater than the preset threshold for liquid circuit voltage for 15 seconds, or when the battery module voltage is greater than the preset threshold for battery module voltage for 15 seconds, proceed to step S76, and at the same time, reset the timer to zero to monitor cyclically;

[0123] Step S76: Decrease the current maximum rechargeable power by a second preset value. That is, when a pulse signal indicating that the liquid circuit voltage is greater than the liquid circuit voltage preset threshold is detected for 15 seconds, the current maximum rechargeable power is decreased by a second preset value.

[0124] Step S77: Detect whether the battery system SOC is greater than the upper limit threshold. In response to the battery system SOC being greater than the upper limit threshold, proceed to step S78.

[0125] Step S78: The battery charge / discharge status (INT) is that the battery is fully charged, and only discharging is allowed. The battery is fully charged flag (BOOL):=TRUE, and the maximum chargeable power of the battery (INT):=0.

[0126] In the above embodiments, steps S71, S74 and S78 can be performed in parallel.

[0127] Next, we will explain the power prediction method applicable to the discharge phase.

[0128] In some embodiments, reducing the current maximum discharge power according to the second parameter includes: in response to the total battery voltage being less than a lower threshold value for a fourth preset duration, subtracting a third preset value from the current maximum discharge power. By dynamically monitoring the total battery voltage (the total voltage of the DC portion of the flow battery system) and its duration, the maximum discharge power is automatically reduced when the total battery voltage is less than the lower threshold value for a fourth preset duration. This reduces the maximum discharge power before entering the constant voltage discharge mode, preventing voltage inconsistencies in the stack modules caused by concentration polarization due to high discharge power at the end of the constant voltage discharge.

[0129] In some embodiments, the fourth preset duration is 30s-60s. In some embodiments, the fourth preset duration is set in the same way as the first preset duration, according to the characteristics of the flow battery. For example, during discharge, the total battery voltage of the flow battery changes faster and needs to be detected more frequently, so the fourth preset duration can be set shorter.

[0130] In some embodiments, the lower limit threshold of the total battery voltage is the minimum value of the total battery voltage operating range increased by 0 to 10V.

[0131] In some embodiments, the fourth preset value is 1% to 2% of the rated power of the battery system.

[0132] In some embodiments, the second parameter further includes the liquid circuit voltage, and reducing the current maximum discharge power according to the second parameter further includes: in response to the liquid circuit voltage decreasing at a rate greater than a decreasing rate threshold and lasting for a fifth preset duration, the current maximum discharge power is reduced by a fourth preset value.

[0133] By reducing the current maximum discharge power based on the rate of decrease of the liquid circuit voltage, the judgment criterion of system concentration polarization amplification (the rate of decrease of the liquid circuit voltage is greater than the rate of decrease threshold) is added to the discharge power prediction. This allows the grid to reduce the discharge power and further avoids the problem of inconsistent stack module voltage caused by concentration polarization due to large discharge power.

[0134] In some embodiments, the fifth preset duration is 3 seconds.

[0135] In some embodiments, the rate of decrease of the liquid circuit voltage exceeding the rate of decrease threshold includes: the rate of decrease of the liquid circuit voltage within 10 seconds is greater than 10 times the rate of decrease of the conventional liquid circuit voltage (i.e., the rate of decrease of the flow battery during constant power discharge).

[0136] In some embodiments, the fourth preset value is 0.5% to 1% of the battery's rated power.

[0137] In some embodiments, the fourth preset value can be reduced according to the voltage characteristics of the concentration polarization of the battery system. For example, the greater the polarization of the stack module voltage, the greater the fourth preset value.

[0138] In some embodiments, the fifth preset duration is shorter than the fourth preset duration. In this embodiment, the sensitivity of the total battery voltage to changes in charge and discharge is lower than the sensitivity of the liquid circuit voltage to changes in discharge, and in order to mitigate the problem of concentration polarization, the liquid circuit voltage is detected more frequently.

[0139] In some embodiments, reducing the current maximum amplifiable power according to the second parameter includes: in response to the SOC being less than the lower limit threshold of SOC, the current maximum amplifiable power is zero.

[0140] In some embodiments, the SOC lower limit threshold is set according to the charge / discharge capability of the flow battery.

[0141] In some embodiments, the lower limit of SOC is related to the application level of the energy storage power station, just like the upper limit threshold. For example, the SOC of lithium batteries is generally 5%-95%, sodium batteries are 10%-90%, and flow batteries are 0%-100%. If the energy storage power station only uses flow batteries, the lower limit threshold can be set to 0%. If flow batteries, sodium batteries, and lithium batteries are used in parallel, the lower limit threshold can be set to 10% of the discharge limit of sodium batteries.

[0142] Figure 8 A flowchart illustrating the reduction of the current maximum amplifiable power in a real-time power prediction method according to an embodiment of this application is shown. Figure 8 As shown, the battery discharge phase includes the following steps to reduce the current maximum discharge power.

[0143] Step S81: Detect whether the total battery voltage BATV is less than the lower limit threshold of the total battery voltage. In response to the current total battery voltage being less than the lower limit threshold of the total battery voltage, proceed to step S82.

[0144] Step S82: Timer for 30 seconds. That is, if the current total battery voltage is lower than the lower limit threshold for 30 seconds, proceed to step S83, and the timer is reset to zero to perform cyclic monitoring.

[0145] Step S83: Reduce the maximum discharge power of the battery by a third preset value. That is, when a pulse signal indicating that the current total battery voltage is lower than the lower limit threshold of the total battery voltage is detected for 30 seconds, the current maximum discharge power is reduced by a third preset value.

[0146] Step S84: Detect that the 10-second drop slope of the liquid circuit voltage is greater than 10 times the slope of a conventional constant power discharge. When the drop rate of the liquid circuit voltage is greater than the drop rate threshold, proceed to step S85. Here, the slope refers to the slope of the curve with time as the x-axis and liquid circuit voltage as the y-axis. A 10-second drop slope of the liquid circuit voltage greater than 10 times the slope of a conventional constant power discharge means that the drop rate of the liquid circuit voltage, as mentioned above, is greater than the drop rate threshold.

[0147] Step S85: Time for 3 seconds. That is, when the rate of decrease of the liquid circuit voltage exceeds the rate of decrease threshold for 3 seconds, proceed to step S86, and at the same time reset the timer to zero for cyclic monitoring.

[0148] Step S86: Decrease the current maximum discharge power of the battery by a fourth preset value. That is, when it is detected that the rate of decrease of the liquid circuit voltage is greater than the rate of decrease threshold for 3 seconds, the current maximum discharge power is decreased by a fourth preset value.

[0149] Step S87: Detect whether the battery system SOC is less than the lower limit set value (lower limit threshold). In response to the battery system SOC being less than the lower limit threshold, proceed to step S88.

[0150] Step S88: The battery charge / discharge state (INT) is that the battery is discharged and only charging is allowed. The battery discharge flag (BOOL) is TRUE and the maximum discharge power of the battery (INT) is 0.

[0151] In the above embodiments, steps S81, S84 and S88 can be performed in parallel.

[0152] Figure 9 A charging and discharging flowchart of a real-time power prediction method according to an embodiment of this application is shown. Figure 9 As shown, in step S91, it is detected that the flow battery energy storage system meets the charge / discharge power boundary conditions. That is, when the BMS of the flow battery detects that the total battery voltage BATV has reached the set value, step S92 is executed. Step S92: The flow battery BMS performs state monitoring and power prediction. When the BMS's state monitoring and power prediction fail, and the total battery voltage continues to rise or fall to the upper or lower limit of the battery system voltage, step S93 is executed. Step S93: The PCS activates constant voltage mode charge / discharge limitation. That is, the PCS switches from constant power mode to constant voltage mode. At this time, the PCS no longer executes the power command issued by the EMS, but instead executes the constant voltage charging / discharging mode. When the BMS reports that the battery is fully charged or discharged, the charging / discharging behavior ends and the process proceeds to step S94. Step S94: The EMS monitors the BMS telemetry and telecontrol data and no longer responds to power commands. Some electrochemical energy storage power stations require the EMS to monitor the state of the BMS and adjust the power command of the PCS according to the state of the BMS.

[0153] like Figure 9 As shown, with Figure 1 Compared to the charging and discharging process in the previous application, this application performs the aforementioned power prediction method before entering constant voltage charging and discharging. This method can accurately predict the current maximum chargeable power and the current maximum dischargeable power, effectively participate in electricity spot market transactions, accurately report the battery system's self-planning, and prevent power deviation assessments. Compared to the 2016 national standard, the above process can reduce the overall charging and discharging time of the battery. Compared to the 2025 national standard, the overall charging and discharging time of the above process is close to that of the 2025 national standard. By monitoring the status of important parameters such as the BMS voltage (CSV, LQV, MDV, BATV), open circuit voltage (OCV), and SOC, the maximum chargeable power and maximum dischargeable power of the battery are reduced in advance before approaching the constant voltage charging and discharging mode. This prevents the generation of side reactions in the stack under long-term constant voltage charging and discharging and solves the problem of inconsistent stack module voltage caused by concentration polarization at the end of constant voltage discharge.

[0154] The constant-voltage charging and discharging power limitation of the flow battery PCS in step S93 is a measure taken after the status monitoring and power calculation of the flow battery BMS in step S92 have been completed. It uses the PCS's own constant power to constant voltage setpoint and the voltage detection of the battery's total voltage (BATV) on the PCS's DC side to perform constant-voltage charging and discharging operation. Although the BMS can no longer accurately report the maximum chargeable and dischargeable power at this time, it can maintain the battery's charging and discharging function. In step S94, the EMS monitors the telemetry and telecontrol data of the BMS and, after completing steps S92 and S93, adjusts the power transmitted from the PCS at the station control level.

[0155] In some embodiments, the power prediction method described above can be performed during both the charging and discharging phases, or it can be performed only during one of the charging and discharging phases.

[0156] In some embodiments, the method for updating the parameters to be identified in the power prediction model (i.e., updating the power prediction model) described above can be used in conjunction with the power prediction method described above. Figure 10 A flowchart of a real-time power prediction method according to another embodiment of this application is shown. Figure 10 As shown, the initial steps include: enabling power prediction, i.e., starting the execution of the real-time power prediction method, followed by:

[0157] Step S101: Call the OCV-SOC state-space function. That is, the formula (2) mentioned above.

[0158] Step S102: Obtain the battery system SOC.

[0159] Step S103: Call the SOC and the power prediction model for the maximum chargeable and dischargeable power. That is, call the SOC in step S102 and the formula (1) mentioned above.

[0160] Step S104: Obtain the current maximum chargeable power and / or the current maximum dischargeable power.

[0161] Step S105: Real-time monitoring of liquid circuit voltage LQV, battery module voltage MDV, and total battery voltage BATV.

[0162] Step S106: The real-time power prediction and iterative algorithm includes: iterating the maximum chargeable power and maximum dischargeable power based on the monitored values. That is, reducing the current maximum chargeable power and / or the current maximum dischargeable power according to the above voltage parameters until the battery is fully charged / discharged, at which point the corresponding values ​​of the current maximum chargeable power and / or the current maximum dischargeable power are set to 0. After the charge / discharge state transitions, return to step S103. If parameter updates are required, execute step S107.

[0163] Step S107: Correct the power prediction model of SOC and maximum chargeable and dischargeable power. That is, update the parameters to be identified in the power prediction model. This includes: Step S1071: Record the maximum chargeable power BCPM and maximum dischargeable power BDPM under the feature SOC (i.e., the preset SOC) and store them in the FIFO queue. That is, obtain the target data as mentioned above. And Step S1072: Construct the FFRLS algorithm to correct the power prediction model of SOC and BCPM, and correct the power prediction model of SOC and BDPM. That is, update the parameters to be identified by using formulas (3)-(5) as mentioned above. After updating the parameters to be identified in the power prediction model, return to step S103.

[0164] The basic concepts have been described above. Obviously, for those skilled in the art, the above disclosure is merely illustrative and does not constitute a limitation of this application. Although not explicitly stated herein, those skilled in the art may make various modifications, improvements, and corrections to this application. Such modifications, improvements, and corrections are suggested in this application, and therefore remain within the spirit and scope of the exemplary embodiments of this application.

[0165] Furthermore, this application uses specific terms to describe embodiments of the application. For example, "an embodiment," "one embodiment," and / or "some embodiments" refer to a particular feature, structure, or characteristic related to at least one embodiment of the application. Therefore, it should be emphasized and noted that "an embodiment," "one embodiment," or "an alternative embodiment" mentioned twice or more in different locations in this specification do not necessarily refer to the same embodiment. In addition, certain features, structures, or characteristics in one or more embodiments of the application can be appropriately combined.

[0166] In some embodiments, numbers describing the quantity of components and attributes are used. It should be understood that such numbers used in the description of embodiments are modified in some examples with the terms "approximately," "approximately," or "generally." Unless otherwise stated, "approximately," "approximately," or "generally" indicates that the numbers are allowed to vary by ±20%. Accordingly, in some embodiments, the numerical parameters used in the specification are approximate values, which may be changed according to the characteristics required by individual embodiments. In some embodiments, numerical parameters should take into account specified significant digits and employ a general method of digit reservation. Although the numerical ranges and parameters used to confirm their breadth of range in some embodiments of this application are approximate values, in specific embodiments, such values ​​are set as precisely as feasible.

Claims

1. A method for real-time power prediction of a flow battery, characterized in that, include: During the charging process of the flow battery, the current maximum chargeable power of the flow battery is obtained according to a power prediction model, wherein the power prediction model includes: ， Where a, b, and c are all parameters to be identified. The maximum chargeable power is defined as , and SOC is defined as the battery state of charge. Acquire target data, the target data including the current maximum chargeable power under a preset battery state of charge; Updating the parameters to be identified in the power prediction model based on the target data includes updating the parameters to be identified using the following formulas: ； ； ； in, The parameter vector is composed of the parameters to be identified. , Forgetting factor, Let K be the target data at time k, K(k) be the gain matrix at time k, and P(k) be the covariance matrix at time k. (k) represents the data vector of the preset battery state of charge (SOC) at time k. =[SOC 2 [,SOC,1]; Obtain the first parameter, which includes the total battery voltage and the battery state of charge; and The current maximum chargeable power is reduced based on the first parameter.

2. The method as described in claim 1, characterized in that, The step of reducing the current maximum chargeable power according to the first parameter includes: In response to the total battery voltage being greater than the upper limit threshold of the total battery voltage and lasting for a duration of a first preset duration, the current maximum chargeable power is reduced by the first preset value.

3. The method as described in claim 2, characterized in that, The first parameter also includes the liquid circuit voltage or the battery module voltage, and the step of reducing the current maximum chargeable power according to the first parameter further includes: In response to the liquid circuit voltage being greater than a preset threshold for liquid circuit voltage and lasting for a duration of a second preset duration, or the battery module voltage being greater than a preset threshold for battery module voltage and lasting for a duration of a third preset duration, the current maximum chargeable power is reduced by the second preset value.

4. The method as described in claim 3, characterized in that, The second preset duration is less than the first preset duration, and the third preset duration is less than the first preset duration.

5. The method as described in claim 1, characterized in that, The step of reducing the current maximum chargeable power according to the first parameter includes: In response to the battery state of charge being greater than the upper limit threshold of the battery state of charge, the current maximum chargeable power is zero.

6. A method for real-time power prediction of a flow battery, characterized in that, include: During the discharge process of the flow battery, the current maximum discharge power of the flow battery is obtained according to a power prediction model, wherein the power prediction model includes: ， Where a, b, and c are all parameters to be identified. The maximum discharge power is defined as , and SOC is defined as the battery state of charge. Acquire target data, which includes the current maximum discharge power under a preset battery state of charge; Updating the parameters to be identified in the power prediction model based on the target data includes updating the parameters to be identified using the following formulas: ； ； ； in, The parameter vector is composed of the parameters to be identified. , Forgetting factor, Let K be the target data at time k, K(k) be the gain matrix at time k, and P(k) be the covariance matrix at time k. (k) represents the data vector of the preset battery state of charge (SOC) at time k. =[SOC 2 [,SOC,1]; Obtain the second parameter, which includes the total battery voltage and the battery state of charge; and The current maximum amplifiable power is reduced based on the second parameter.

7. The method as described in claim 6, characterized in that, The step of reducing the current maximum amplifiable power according to the second parameter includes: In response to the total battery voltage being less than the lower limit threshold of the total battery voltage and the duration reaching the fourth preset duration, the current maximum discharge power is reduced by the third preset value.

8. The method as described in claim 7, characterized in that, The second parameter also includes the liquid circuit voltage, and the step of reducing the current maximum discharge power according to the second parameter further includes: In response to the liquid circuit voltage decreasing at a rate greater than the decreasing rate threshold and lasting for a duration of a fifth preset duration, the current maximum discharge power is reduced by a fourth preset value.

9. The method as described in claim 8, characterized in that, The fifth preset duration is less than the fourth preset duration.

10. The method as described in claim 6, characterized in that, The step of reducing the current maximum discharge power according to the second parameter includes: in response to the battery state of charge being less than the battery state of charge lower limit threshold, the current maximum discharge power is zero.

11. The method as described in claim 1 or 6, characterized in that, The battery state of charge is calculated using the following formula: ， in, Let be the state of charge of the battery at time t. Let be the open-circuit voltage of the flow battery at time t. Let i be a coefficient, where i is an integer from 0 to N, and N is a positive integer greater than or equal to 1.

12. The method as described in claim 1 or 6, characterized in that, After acquiring the target data, the process also includes: The target data is compensated based on the current operating flow of the battery system, including: in response to the current operating flow of the battery system being greater than the theoretical operating flow corresponding to the current state of charge of the battery, the target data is subtracted by a fifth preset value; in response to the current operating flow of the battery system being less than the theoretical operating flow corresponding to the current state of charge of the battery, the target data is increased by a sixth preset value.

13. The method as described in claim 1 or 6, characterized in that, The target data includes multiple current maximum chargeable power or current maximum dischargeable power of a preset battery state of charge. After acquiring the target data, the process further includes: Weights are assigned to the target data, wherein the weight of the target data at the current time is greater than the weight of the target data at the previous time.

14. The method as described in claim 1 or 6, characterized in that, The power prediction model includes: The data acquisition layer is used to acquire the target data; The data processing layer is used to process the target data; A model iteration layer is used to update the parameters to be identified in the power prediction model based on the target data; and The control output layer is used to enable the battery management system to obtain the current maximum chargeable power and / or the current maximum dischargeable power according to the power prediction model, and to provide the current maximum chargeable power and / or the current maximum dischargeable power to the energy storage converter and / or the energy management system for power plant scheduling.

Images