Battery cell balance decision-making method and equipment of energy storage system, medium and program product

Abstract

The invention discloses an energy storage system cell equalization decision-making method, equipment, a medium and a program product, and relates to the field of battery detection equipment. In the method, when it is monitored that a target cell unit triggers a preset equalization condition, historical operation data is acquired, and based on the data, a long-term aging rate representing a long-term health risk of a cell is calculated, a power fluctuation characteristic of a system application scene is identified, and a recoverable electric quantity of a current equalization income is quantified. According to the method, the balancing task in a high-power application scene can be preferentially processed; and for other scenes, the risk urgency obtained by mapping the long-term aging rate is combined with the recoverable electric quantity to generate a comprehensive priority, and a final equilibrium decision is formed according to the comprehensive priority. By implementing the technical scheme provided by the invention, limited balance resources can be applied to the battery cells with high risk and high income, so that the overall operation efficiency and economical efficiency of the energy storage system are improved while the system safety is guaranteed.

Description

TECHNICAL FIELD

[0001] The present application relates to the field of battery detection equipment, and in particular to a method and device for equalizing decision of an energy storage system, a medium and a program product. BACKGROUND

[0002] Lithium batteries are a kind of rechargeable batteries, which are favored by users due to their relatively light weight, high energy density, fast charging speed and other advantages. However, in a lithium iron phosphate battery pack, due to the limitations of the manufacturing process, even if the cells are produced in the same batch, there will still be slight differences in key performance parameters such as capacity, internal resistance, and self-discharge rate. These initial differences will be continuously amplified during the long-term cycle charging and discharging process of the energy storage system, eventually leading to a serious imbalance in the state of charge (SOC) of each single cell. This imbalance will limit the actual available capacity of the battery pack and accelerate the overall performance degradation of the battery pack, thereby shortening its service life.

[0003] In related technologies, by monitoring the voltage of all single cells in the battery pack in real time, when it is found that the voltage of one or more cells is significantly higher than that of other cells, the single cell will be restored to a preset consistency range through cell equalization technology. Cell equalization is a process of using specific technical means to make the SOC (State of Charge) of all single cells in the battery pack balanced. This technology alleviates the system performance problems caused by voltage inconsistency to some extent.

[0004] However, since the energy storage system is a long-term and dynamic aging complex system, the performance of the internal cells will evolve over time. Related equalization evaluation methods often only use the end-of-charge cutoff voltage as the basis for judgment. This judgment of equalization demand relying on a single voltage data ignores the differences in long-term degradation trends of different cells, and this inaccurate equalization demand judgment will affect the stability of the energy storage system. SUMMARY

[0005] The present application provides a method and device for equalizing decision of an energy storage system, a medium and a program product, which determines the priority of the equalization task by adapting to the application scenario, to improve the accuracy of the equalization demand judgment of the energy storage system.

[0006] In a first aspect, the present application provides a method for equalization decision of an energy storage system cell. The energy storage system comprises a plurality of cell units. The method comprises: when a running parameter of a target cell unit triggers a preset equalization condition, obtaining historical running data of the target cell unit in the energy storage system in a preset historical period, the historical running data comprising voltage and current corresponding to a timestamp, the current being used to determine a dynamic state and a static state of the target cell unit; after extracting a plurality of conversion events of the target cell unit from the dynamic state to the static state from the historical running data, calculating historical polarization voltage values of the plurality of conversion events and forming a polarization voltage value time sequence, the historical polarization voltage value being a difference between voltage values of the target cell unit before and after the conversion event; performing linear regression on the polarization voltage value time sequence, and calculating a slope as a long-term polarization growth rate of the target cell unit; based on the historical running data, calculating a standard deviation of current values of a battery pack to which the target cell unit belongs in a dynamic state to obtain a power fluctuation characteristic value; if the power fluctuation characteristic value is higher than a preset power application threshold, determining an equalization task of the target cell unit as a primary priority; if the power fluctuation characteristic value is not higher than the power application threshold, calculating a recoverable amount of electricity of the target cell unit based on a preset voltage- state of charge curve; multiplying an urgency coefficient mapped from the long-term polarization growth rate and the recoverable amount of electricity to obtain a product, and determining a priority of the equalization task of the target cell unit in equalization tasks of the energy storage system based on the product, and generating an equalization decision.

[0007] By using the above technical solution, the polarization growth rate reflecting the long-term aging risk and the recoverable amount of electricity quantifying the immediate benefit are calculated, and the two are combined to determine the equalization priority, and the power fluctuation characteristic is identified to prioritize the high-power scenario. The equalization decision can direct the limited equalization resources to the cells with high risk and high value, improving the scientificity of the decision and the resource allocation efficiency.

[0008] In combination with some embodiments of the first aspect, in some embodiments, after the step of performing linear regression on the polarization voltage value time sequence to calculate the slope as the long-term polarization growth rate of the target cell unit, the method further comprises: intercepting a predetermined number of recent conversion events from the polarization voltage value time sequence to form a recent observation window; performing linear regression on the recent observation window to calculate a slope as a recent polarization growth rate of the target cell unit; calculating a deterioration trend factor as a ratio of the recent polarization growth rate to the long-term polarization growth rate; if the deterioration trend factor is within a preset stable interval close to 1, mapping an urgency coefficient based on the long-term polarization growth rate; if the deterioration trend factor is greater than an upper limit of the stable interval, determining a product of the deterioration trend factor and a basic urgency coefficient as the urgency coefficient after mapping the basic urgency coefficient based on the long-term polarization growth rate.

[0009] By adopting the technical solution, the deterioration trend factor is formed by calculating the ratio of the short-term and long-term aging rates, and when the factor shows accelerated deterioration, it is multiplied by the base urgency coefficient to amplify the risk level. This way enables the decision to capture and quantify the sudden worsening trend of the battery health status, making up for the deficiency of relying only on the long-term average rate to respond to sudden risks.

[0010] In combination with some embodiments of the first aspect, in some embodiments, the method further comprises: if the deterioration trend factor is less than the lower limit of the stable interval, mapping the long-term polarization growth rate to obtain an urgency coefficient.

[0011] By adopting the technical solution, it is ensured that when the battery aging rate slows down or improves, the decision can return to the evaluation of the long-term stable trend, avoiding unnecessary conservative judgments due to short-term fluctuations.

[0012] In combination with some embodiments of the first aspect, in some embodiments, the step of performing linear regression on the polarization voltage value time series to obtain the long-term polarization growth rate of the target battery cell unit specifically comprises: using a preset robust regression algorithm to preliminarily fit the polarization voltage value time series to obtain an inner point set consistent with the main trend; and performing linear regression on the inner point set to obtain the long-term polarization growth rate of the target battery cell unit.

[0013] By adopting the technical solution, outliers in the data are identified and separated in advance by using a robust regression algorithm, and linear regression is only performed on the inner point set consistent with the main trend. This method effectively avoids the interference of abnormal data points caused by measurement errors or special working conditions on the calculation result of the long-term aging rate, so that the calculated growth rate can more truly reflect the inherent aging law of the battery, thereby improving the accuracy and reliability of risk assessment.

[0014] In combination with some embodiments of the first aspect, in some embodiments, after the step of performing linear regression on the polarization voltage value time series to obtain the long-term polarization growth rate of the target battery cell unit, the method further comprises: counting the number of data points contained in the polarization voltage value time series; and if the number of data points is less than a preset minimum regression data point number, determining that the long-term polarization growth rate is unreliable, and assigning a preset default value to the urgency coefficient of the target battery cell unit.

[0015] By adopting the technical solution, a data quality verification link is introduced into the decision-making process, effectively preventing the distortion of the regression result due to sparse historical data, and thus producing false risk assessment, ensuring that even in the case of poor data, the decision-making system can still operate stably and give a reasonable assessment.

[0016] In some embodiments of the first aspect, in the step of determining the priority of the balancing task of the target battery cell in the balancing tasks of the energy storage system based on the product of the urgency coefficient obtained through the long-term polarization growth rate mapping and the recoverable electric quantity, and generating the balancing decision, the method specifically comprises: obtaining a total number of times of executing the balancing operation of the target battery cell in a preset balancing period as a historical balancing frequency based on a historical balancing record of the target battery cell in the preset balancing period; if the historical balancing frequency is higher than a preset balancing dependency threshold, determining a ratio of the preset balancing dependency threshold to the historical balancing frequency as a benefit decay factor; multiplying the benefit decay factor and the recoverable electric quantity to obtain a modified recoverable electric quantity; and determining the priority of the balancing task of the target battery cell in the balancing tasks of the energy storage system based on the product of the urgency coefficient obtained through the long-term polarization growth rate mapping and the recoverable electric quantity, and generating the balancing decision.

[0017] By using the above technical solution, the balancing priority of the battery cell that may have internal defects and has a lower balancing benefit is dynamically lowered based on the historical cost-benefit of the balancing operation. Therefore, continuous resource waste on the low-efficiency target can be avoided, the overall allocation efficiency of the balancing resource is optimized, and the long-term economic efficiency of the energy storage system is improved.

[0018] In some embodiments of the first aspect, in the step of determining the priority of the balancing task of the target battery cell in the balancing tasks of the energy storage system based on the product of the urgency coefficient obtained through the long-term polarization growth rate mapping and the recoverable electric quantity, and generating the balancing decision, the method specifically comprises: obtaining a total number of times of executing the balancing operation of the target battery cell in a preset balancing period as a historical balancing frequency based on a historical balancing record of the target battery cell in the preset balancing period; if the historical balancing frequency is higher than a preset balancing dependency threshold, determining a ratio of the preset balancing dependency threshold to the historical balancing frequency as a benefit decay factor; multiplying the benefit decay factor and the recoverable electric quantity to obtain a modified recoverable electric quantity; and determining the priority of the balancing task of the target battery cell in the balancing tasks of the energy storage system based on the product of the urgency coefficient obtained through the long-term polarization growth rate mapping and the recoverable electric quantity, and generating the balancing decision.

[0019] By using the above technical solution, the problem that one of the risk and benefit dimensions may occupy too large a weight in calculating the priority score due to the inconsistent data scales of the two dimensions is solved. The priority ranking result is more reasonable.

[0020] In a second aspect, an apparatus is provided, which comprises one or more processors and a memory; the memory is coupled to the one or more processors, and is configured to store computer program codes, the computer program codes comprising computer instructions, which are invoked by the one or more processors to cause the apparatus to perform the method described in the first aspect and any possible implementation manner of the first aspect.

[0021] In a third aspect, a computer program product comprising instructions which, when executed on an apparatus, cause the apparatus to carry out the method described in the first aspect and any possible implementation manner of the first aspect.

[0022] In a fourth aspect, a computer-readable storage medium comprising instructions which, when executed on an apparatus, cause the apparatus to carry out the method described in the first aspect and any possible implementation manner of the first aspect.

[0023] It can be understood that the apparatus provided in the second aspect, the computer program product provided in the third aspect, and the computer storage medium provided in the fourth aspect are all used to execute the method provided in the embodiments of the present application. Therefore, the beneficial effects that can be achieved thereby can refer to the beneficial effects in the corresponding method, which will not be described here again.

[0024] The one or more technical solutions provided in the embodiments of the present application have at least the following technical effects or advantages: 1. Since the polarization growth rate reflecting long-term aging risk and the recoverable power quantifying immediate benefits are adopted, and the two are combined to determine the balance priority, and the power fluctuation characteristics are identified to preferentially process high-power scenarios, the balance decision can direct the limited balance resources to the battery cells with high risk and high value, thereby improving the scientificity of the decision and the resource allocation efficiency.

[0025] 2. Since the technical means of introducing the comparison between the recent and long-term aging rates to dynamically amplify the sudden risk is adopted, the technical problem of slow response to the sudden acceleration of the battery cell health state is effectively solved, and the potential high-risk battery cells are identified and warned, thereby greatly improving the operation safety of the energy storage system.

[0026] 3. Since the technical means of introducing the benefit decay factor based on the historical balance frequency to modify the cost-effectiveness of the balance task is adopted, the technical problem of insufficient consideration of the cost-effectiveness of the balance operation, which may cause continuous resource waste, is effectively solved, and the invalid investment on inefficient targets is avoided, the overall allocation efficiency of the balance resources is optimized, and the long-term economic efficiency is improved. BRIEF DESCRIPTION OF DRAWINGS

[0027] Figure 1 is a flowchart of one of the energy storage system cell equalization decision-making methods in the embodiments of the present application; Figure 2 is another flowchart of one of the energy storage system cell equalization decision-making methods in the embodiments of the present application; Figure 3 is a schematic diagram of an exemplary hardware structure of the device in the embodiments of the present application. DETAILED DESCRIPTION

[0028] The terms used in the following embodiments of the present application are only for the purpose of describing specific embodiments and are not intended to be limiting to the present application. As used in the specification and the appended claims of the present application, the singular forms “a,” “an” and “the” are intended to include plural forms as well, unless the context clearly indicates otherwise. It will be further understood that the terms “and / or” as used herein refer to and encompass any or all possible combinations of one or more of the listed items.

[0029] Hereinafter, the terms “first” and “second” are used only for the purpose of description and should not be understood as implying or suggesting 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, and in the description of the embodiments of the present application, the meaning of “a plurality of” is two or more, unless otherwise specified.

[0030] Referring to Figure 1 is a flowchart of one of the energy storage system cell equalization decision-making methods in the embodiments of the present application.

[0031] S101, when the operating parameters of the target cell unit trigger the preset equalization condition, the historical operating data of the target cell unit in the preset historical period in the energy storage system is obtained.

[0032] When the operating parameters of the cell unit (target cell unit) trigger the preset equalization condition, the device will access its internal or external historical database to extract all the historical operating data of the target cell unit in the preset historical period. The obtained data is usually time-stamped serialized data, at least including voltage value and current value, and possibly including temperature and other auxiliary information.

[0033] Specifically, the front-end data acquisition unit of the battery management system (BMS) collects voltage data of all battery cells at a high frequency (e.g., 10 times per second). When the voltage difference of a certain battery cell continuously exceeds a preset threshold (e.g., determined by a sliding time window counter), the device generates an imbalance event. The event is pushed to a message queue, which at least contains the ID of the target battery cell. The back-end analysis service subscribes to this message queue and, upon receiving the event, queries a configured time series database (e.g., InfluxDB or Prometheus) with the battery cell ID and a preset historical period as query conditions, to obtain the required high-precision historical voltage and current data.

[0034] It can be understood that the preset balancing condition is a combination of one or more Boolean conditions. It can be that the absolute value of the difference between the voltage of the target battery cell and the average voltage of the battery pack exceeds 50 mV for more than 5 minutes, or that the difference between the state of charge (SOC) of the target battery cell and the average SOC exceeds 3%, without limitation.

[0035] S102, after extracting multiple transition events of the target battery cell from the dynamic state to the stationary state from the historical running data, calculating the historical polarization voltage values of the multiple transition events and constructing a polarization voltage value time sequence.

[0036] The device extracts multiple transition events of the target battery cell from the dynamic state to the stationary state by scanning the current time sequence of the historical running data, where the dynamic state is when the absolute value of the current is continuously higher than a preset dynamic threshold, and the stationary state is when the absolute value of the current is continuously lower than a preset static threshold. After completing event identification, the device calculates for each transition event, extracts the last dynamic voltage value immediately before the occurrence of the transition event, and the first stationary voltage value when the voltage enters a stable state after the occurrence of the transition event. The voltage difference between the two is the historical polarization voltage value of the transition event. The device arranges the historical polarization voltage values calculated for all identified transition events in chronological order to construct a polarization voltage value time sequence.

[0037] Specifically, the moment when the absolute value of the current of the target cell unit changes from being higher than the dynamic current threshold to being continuously lower than the static current threshold is defined as the end moment of the dynamic state T1, and the terminal voltage V1 at this moment is recorded; starting from the moment T1, the voltage recovery process of the target cell unit is continuously monitored, and the voltage change rate dV / dt is calculated in real time; when the absolute value of the voltage change rate dV / dt is continuously lower than a preset voltage stability slope threshold (for example, 0.1 mV / s) within a preset stable observation window (for example, 10 minutes), it is determined that the electrochemical state inside the cell has reached stability, the moment is defined as the static state stability moment T2, and the open-circuit voltage V2 at this moment is recorded; the difference (V1-V2) between V1 and V2 is taken as the final historical polarization voltage value of the conversion event.

[0038] S103, linear regression is performed on the polarization voltage value time sequence, and the slope obtained is the long-term polarization growth rate of the target cell unit.

[0039] The device takes the generated polarization voltage value time sequence as input, and the X axis of the sequence is time and the Y axis is the historical polarization voltage value. The device performs linear regression analysis on these data points, finds an optimal straight line to fit the distribution trend of these data points, and determines the average growth amount of the polarization voltage per unit time as the long-term polarization growth rate.

[0040] It can be understood that linear regression can be implemented in various ways: optionally, a robust regression method based on RANSAC (random sample consensus) can be used. This algorithm does not directly use all data points. The system first randomly selects a minimum subset (2 data points for linear fitting) from the polarization voltage value time sequence and fits an initial straight line. Then, the system traverses all data points, calculates their distance to the straight line, and classifies the points with a distance less than a preset threshold as "inliers". The system repeats this process several times and finally selects the set with the most inliers. Finally, the system only uses this largest inlier set to perform the final linear regression by the standard least squares method, and the slope obtained is the long-term polarization growth rate. Optionally, the Huber regression method can be used. This is a regression model that is inherently robust. Unlike the least squares method, which squares all errors, Huber regression uses a square loss for smaller errors (below a certain threshold) and a linear loss for larger errors (considered outliers). This reduces the weight of outliers in the model fitting process. The system directly takes the entire polarization voltage value time sequence as input and applies the Huber regression model for fitting, and the slope obtained is the long-term polarization growth rate.

[0041] In some embodiments, the device does not fit all data points. Instead, it first performs a preliminary analysis using a pre-defined robust regression algorithm to identify and separate outliers that deviate from the main trend of the data. Subsequently, linear regression analysis is performed only on the "interior set" of data points that are highly consistent with the main trend.

[0042] In some embodiments, data loss or missing data may lead to insufficient historical valid data. The device counts the number of data points in the polarization voltage value time series. If the number of data points is less than a preset minimum number of regression data points, the device determines that the long-term polarization growth rate is unreliable. An urgency coefficient for the target cell is assigned a preset default value for subsequent calculations, and the data missing situation is reported.

[0043] S104. Based on historical operating data, calculate the standard deviation of the current value of the battery pack to which the target cell belongs under dynamic conditions to obtain the power fluctuation characteristic value.

[0044] After acquiring historical data of the target battery cell through step S101, the device filters out time periods in a dynamic state from the historical operating data. It extracts the total current value sequence of the entire battery pack within these time periods and performs statistical calculations on this current value sequence. The calculated standard deviation of the current value is used as a power fluctuation feature value, which serves as a quantified scene identification indicator.

[0045] It's important to note that standard deviation is a measure of data dispersion. A large standard deviation means that the current value deviates frequently and significantly from its average value, typically corresponding to power-intensive applications such as grid frequency regulation and smoothing renewable energy fluctuations. These scenarios require batteries to charge and discharge rapidly and at high rates. A small standard deviation, on the other hand, indicates that the current value is relatively stable, often corresponding to capacity-intensive applications such as peak shaving and valley filling, and backup power supplies.

[0046] S105. If the power fluctuation characteristic value is higher than the preset power application threshold, the balancing task of the target cell unit will be determined as the first priority.

[0047] The device compares the power fluctuation characteristic value with a preset power application threshold. If the power fluctuation characteristic value is determined to be higher than the threshold, it is determined that the current target cell is operating in a power scenario with stringent requirements for cell consistency. The device marks the balancing task of the target cell as "first priority" or "highest priority". This task will be inserted at the top of the balancing task queue to ensure that it can be processed as soon as possible.

[0048] It should be noted that the preset power type application threshold is a critical value pre-set according to experience or experiment, which is used to distinguish the power type application scenario and the non-power type application scenario. In the power type application scenario (such as grid frequency modulation, smoothing renewable energy fluctuation), the performance short board of the battery cell (such as too high internal resistance) may be amplified under the impact of fast and large current, so as to cause the voltage to reach the upper limit or lower limit too early, thereby limiting the power output capability of the whole battery pack. In the non-power type application scenario, the current is usually stable and the fluctuation is small, and the system is more focused on the energy throughput (i.e. capacity utilization) rather than the instantaneous power response capability, such as peak shaving and valley filling or as a backup power supply.

[0049] S106, if the power fluctuation characteristic value is not higher than the power type application threshold, the recoverable electric quantity of the target battery cell unit is calculated based on the preset voltage-charge state curve.

[0050] If it is judged that the power fluctuation characteristic value is not higher than the preset power type application threshold, the device determines the SOC value corresponding to the target battery cell voltage and the average voltage based on the preset voltage-charge state (V-SOC) curve by obtaining the current voltage value of the target battery cell and the average voltage value (or a reference reference voltage value) of the battery pack where the target battery cell is located. The difference (SOC value) between the two represents the deviation degree of the battery cell in the electric quantity. The device multiplies the SOC difference value by the rated capacity (or the current actual capacity) of the battery cell to calculate the recoverable electric quantity.

[0051] It should be noted that the preset voltage-charge state (V-SOC) curve is a data model describing the corresponding relationship between the open circuit voltage and the charge state of the battery cell under certain conditions (such as certain temperature, health state).

[0052] S107, after multiplying the urgency coefficient obtained by mapping the long-term polarization growth rate and the recoverable electric quantity to obtain a product, the priority of the balancing task of the target battery cell unit in the balancing task of the energy storage device is determined based on the product, and a balancing decision is generated.

[0053] After converting the long-term polarization growth rate representing the long-term aging risk of the battery cell into a quantitative urgency coefficient, the urgency coefficient and the recoverable electric quantity are normalized, and the urgency coefficient and the recoverable electric quantity are mapped into a unified preset closed interval. Then, through the product of the normalized urgency coefficient and the normalized recoverable electric quantity, the priority of the balancing task of the target battery cell unit in the balancing task of the energy storage device is determined, and a balancing decision is generated.

[0054] It should be noted that the mapping principle is based on data driving, and the device pre-collects the long-term polarization growth rate of all cells from a large-scale cell group to form a global data set reflecting the health status of the group. The global data set is analyzed by using an unsupervised learning algorithm (such as K-means clustering) to obtain a plurality of risk clusters based on the similarity of the aging rate of the cells. After the clusters are divided, the device determines the boundaries between different risk levels based on the center points of each cluster (i.e., the average aging rate of each group). At the same time, the device assigns an urgency coefficient to the long-term polarization growth rate of different degrees for these risk clusters ranked from low to high through a preset nonlinear function (for example, an exponential function).

[0055] In some embodiments, the device first traverses all currently pending balancing tasks, collects their urgency coefficient set and recoverable capacity set. Then, the device processes the two sets respectively by using the min-max normalization method. The recoverable capacity is also processed in the same way. After normalization, the device multiplies the normalized urgency coefficient of each task with the normalized recoverable capacity to obtain the final priority score. The device sorts all tasks according to this score and submits the sorting result to the balancing execution module.

[0056] In the embodiments of the present application, since the dynamic priority evaluation based on long-term aging risk and immediate balancing value is adopted, the problem of lack of scene adaptability caused by balancing decision-making relying on current voltage, SOC and other instantaneous states is effectively solved, the accuracy of balancing demand judgment of the energy storage device is improved, and the limited operation and maintenance balancing resources can be accurately applied to the key cells, so as to ensure the safety of the device while relieving the difficulty of the shortage of operation and maintenance resources.

[0057] In actual application, when the above-mentioned balancing decision-making method for cells of an energy storage system is executed, the long-term polarization growth rate is a historical average trend, and when the cells accelerate the deterioration due to internal state mutation in the near future, the method cannot timely capture the sharp change of the risk level, and there is a response delay.

[0058] Please refer to Figure 2 , which is another flowchart of the balancing decision-making method for cells of an energy storage system in the embodiments of the present application.

[0059] S201, when the running parameters of the target cell unit trigger the preset balancing condition, the historical running data of the target cell unit in the preset historical period in the energy storage device is obtained.

[0060] S202, after extracting a plurality of conversion events of the target cell unit from the dynamic state to the static state from the historical running data, the historical polarization voltage values of the plurality of conversion events are calculated and a polarization voltage value time sequence is constructed.

[0061] S203, linear regression is performed on the polarization voltage value time sequence, and the slope obtained is the long-term polarization growth rate of the target cell unit.

[0062] Steps S201 to S203 are similar to steps S101 to S103, which will not be described here.

[0063] S204, a predetermined number of recent conversion events are intercepted from the polarization voltage value time sequence to form a recent observation window.

[0064] In the polarization voltage value time sequence generated in step S202, the closest in time to the predetermined number of conversion events and their corresponding historical polarization voltage values are located and extracted, and they together form a recent observation window, which is a subset of the polarization voltage value time sequence.

[0065] Specifically, the device first sorts the complete polarization voltage value time sequence in descending order of timestamp, and then intercepts a predetermined number of data points from the head of the sequence. The device records the specific timestamps of these events occurring during data interception for subsequent time span calculation. The device verifies the integrity of the intercepted data to ensure that each conversion event selected has a valid polarization voltage value. In order to ensure the time continuity of the data, the device checks the time interval between adjacent conversion events, and if it finds that the interval between certain events is abnormal (such as long-term data loss due to device maintenance or failure), it adjusts the range of the observation window accordingly to ensure that the selected data can truly reflect the continuous performance change of the battery.

[0066] In some embodiments, when the number of valid conversion events in the recent period of time is less than the predetermined number, the device automatically calculates the time distribution characteristics of the current available data points, including the average interval and the standard deviation. Based on these statistical characteristics, the device dynamically adjusts the minimum number of data points required to meet the basic requirements of statistical analysis. Specifically, the device calculates the ratio of the time span of the current available data points to the time span of the expected observation window. When the ratio is greater than a preset threshold, the device still constructs the observation window based on the existing data even if the number of data points is less than the predetermined value S205, linear regression is performed on the recent observation window, and the slope obtained is the recent polarization growth rate of the target cell unit.

[0067] The device performs similar operations on the recent observation window as in step S203, taking time as the independent variable and the polarization voltage value as the dependent variable, and fitting an optimal straight line through the least squares method. The slope of the straight line represents the average change in polarization voltage per unit time, and the recent average growth of polarization voltage per unit time is determined as the recent polarization growth rate.

[0068] Specifically, the device first pre-processes the data in the recent observation window. The time stamp is converted into a time difference relative to the start time of the window, and normalized. The polarization voltage value is also normalized. The device uses weighted least squares method for regression analysis, giving higher weight to data points closer in time to the current. During the regression process, the device also calculates the coefficient of determination (R²) to assess the reliability of the regression result. The device also calculates the 95% confidence interval of the regression line to assess the uncertainty range of the recent polarization growth rate.

[0069] In some embodiments, when performing linear regression, abnormal data points may interfere with the regression result. For this case, the device uses an outlier detection method based on Cook's distance: first, calculate the Cook's distance for all data points, and mark the points with Cook's distance greater than a preset threshold as potential outliers. The device analyzes the causes of these outliers by checking the operation log of the corresponding time period, and confirms whether there is a special operating condition or device state. For the data points confirmed as outliers, the device reduces their weight in the regression calculation while retaining the original data.

[0070] S206, calculate the deterioration trend factor by dividing the recent polarization growth rate by the long-term polarization growth rate.

[0071] The ratio of the recent polarization growth rate to the long-term polarization growth rate is the deterioration trend factor. The value of the deterioration trend factor quantifies the change trend of the aging rate: if the factor is significantly greater than 1, it indicates that the recent aging rate is much faster than the long-term average, and the battery health is deteriorating rapidly; if the factor is approximately equal to 1, it indicates that the recent trend is consistent with the long-term trend, and the aging state is stable; if the factor is significantly less than 1, it indicates that the recent aging rate is slowing down, or even improving.

[0072] It can be understood that the calculation of the deterioration trend factor can be implemented in various ways: optionally, the device checks the absolute value of the long-term polarization growth rate before calculation, and if it is less than a very small positive number, the deterioration trend factor is directly set to 1, considering it to be in a stable state. Optionally, when both rate values are very small or span multiple orders of magnitude, the device first takes the logarithm of both rates, then calculates the difference. A positive difference indicates acceleration, a negative difference indicates deceleration, and a zero difference indicates stability.

[0073] S207, determine whether the deterioration trend factor is greater than the upper limit of the stable interval.

[0074] Compare the deterioration trend factor with the upper limit of a preset stable interval (for example, the upper limit of the stable interval can be set to 1.2) to identify the battery whose health condition is deteriorating at a rate that exceeds the normal range.

[0075] If the result of the judgment is that the deterioration trend factor is not greater than the upper limit of the stable interval, step S208 is performed. If the result of the judgment is that the deterioration trend factor is greater than the upper limit of the stable interval, step S209 is performed.

[0076] When performing the threshold judgment, the factor value may frequently fluctuate near the threshold, resulting in unstable judgment results. The device can use a hysteresis-based judgment mechanism by setting a trigger threshold higher than the upper limit of the stable interval and a recovery threshold lower than the upper limit. When the factor value first exceeds the trigger threshold, the device marks it as an unstable state; only when the factor value decreases below the recovery threshold and remains for a certain period of time, the device removes the unstable state mark and performs step S208.

[0077] S208, based on the long-term polarization growth rate mapping, obtain the urgency coefficient.

[0078] If the result of the judgment is that the deterioration trend factor is not greater than the upper limit of the stable interval, the long-term polarization growth rate representing the long-term aging risk of the battery cell is mapped to a quantitative urgency coefficient through pre-set experimental data.

[0079] S209, after obtaining the basic urgency coefficient based on the long-term polarization growth rate mapping, the product of the deterioration trend factor and the basic urgency coefficient is determined as the urgency coefficient.

[0080] If the result of the judgment is that the deterioration trend factor is greater than the upper limit of the stable interval, first, the basic urgency coefficient is obtained based on the long-term polarization growth rate mapping, and then the basic urgency coefficient is multiplied by the deterioration trend factor itself to obtain the final urgency coefficient.

[0081] S210, based on historical operation data, calculate the standard deviation of the current value of the battery pack to which the target battery cell belongs in a dynamic state to obtain a power fluctuation characteristic value.

[0082] Step S210 is similar to step S104, which will not be repeated here.

[0083] S211, if the power fluctuation characteristic value is not higher than the power application threshold, the recoverable capacity of the target battery cell is calculated based on the pre-set voltage-charge state curve.

[0084] Step S211 is similar to step S106, which will not be repeated here.

[0085] In some embodiments, if the power fluctuation characteristic value is higher than the pre-set power application threshold, the balancing task of the target battery cell is determined as the primary priority, which will not be repeated here.

[0086] S212, based on the historical balancing records of the target cell unit in a preset balancing period, obtaining the total number of times that the target cell unit is executed balancing operation as a historical balancing frequency.

[0087] After monitoring the operating parameters of the target cell unit triggering the preset balancing condition, the device accesses its internal operation and maintenance log database to specifically query the historical balancing records related to the target cell unit. The time range of the query is limited within a preset balancing period, for example, the past 30 days or 100 complete charge and discharge cycles. The device counts the total number of times that the target cell unit is successfully executed balancing operation within this period.

[0088] Specifically, when performing the counting, the device first determines the time window for counting. Usually, a sliding time window is used to count only the balancing records in the recent period. The device verifies the effectiveness of each balancing record to exclude incomplete or abnormally interrupted balancing operations. During the counting process, the device also records the time distribution characteristics of the balancing operation, including the mean and variance of the operation interval.

[0089] It can be understood that a high historical balancing frequency usually implies that the cell may have a high self-discharge rate, a too fast capacity decay or other internal defects, resulting in the need for continuous external intervention to keep up with the overall state of the battery pack.

[0090] S213, if the historical balancing frequency is higher than the preset balancing dependency threshold, determining the ratio of the preset balancing dependency threshold to the historical balancing frequency as a yield decay factor.

[0091] If the historical balancing frequency is higher than the preset balancing dependency threshold, it is determined that the cell is a balancing-dependent cell, and the calculation of the yield decay factor is started. The calculation method is to take the preset balancing dependency threshold as the numerator and the actual historical balancing frequency as the denominator, and the ratio obtained by dividing the two is the yield decay factor. Since the denominator is greater than the numerator at this time, the factor is a number less than 1, and the higher the historical balancing frequency, the smaller the value of the factor.

[0092] In some embodiments, a dynamic dependency threshold based on population statistics is introduced. The device can perform a statistical analysis of the historical balancing frequency of all cells in the energy storage system periodically (for example, every week), and calculate the average balancing frequency and standard deviation of the entire population. Then, the device can dynamically set the balancing dependency threshold to a value related to the statistical result.

[0093] S214, taking the product of the yield decay factor and the recoverable electric quantity as the corrected recoverable electric quantity.

[0094] S215, multiplying the urgency coefficient obtained through the long-term polarization growth rate mapping by the recoverable power to obtain a product, determining the priority of the balancing task of the target battery cell in the balancing task of the energy storage system based on the product, and generating a balancing decision.

[0095] After obtaining the urgency coefficient through step S208 or step S209, in combination with the recoverable power obtained in step S214, the urgency coefficient and the recoverable power are normalized, and the urgency coefficient and the recoverable power are mapped into a unified preset closed interval. Then, through the product of the normalized urgency coefficient and the normalized recoverable power, the priority of the balancing task of the target battery cell in the balancing task of the energy storage system is determined, and a balancing decision is generated.

[0096] It can be understood that steps S213-S214 realize quantitative evaluation of the input-output ratio of the balancing task by introducing the benefit decay based on the historical balancing frequency. In some embodiments, steps S213-S214 can also not be executed when the historical balancing frequency is not higher than the preset balancing dependency threshold, so that step S215 can be directly executed after step S212 is executed, which is not limited here.

[0097] In the embodiments of the present application, by introducing the comparison between the recent and long-term aging rates, the problem of slow response to sudden accelerated deterioration of the battery health state is effectively solved, and the balancing effect of the energy storage system is improved.

[0098] And by introducing the benefit decay based on the historical balancing frequency, the technical problem of insufficient consideration of the cost-effectiveness of the balancing operation, which may cause resource waste to the heavy and difficult-to-repair battery, is effectively solved, thereby avoiding ineffective balancing of the inefficient battery, optimizing the balancing resource allocation, and improving the overall operation efficiency and economic efficiency of the system.

[0099] The following describes an exemplary device 300 provided by the embodiments of the present application. Figure 3 is an exemplary hardware structure schematic diagram of the device 300 provided by the embodiments of the present application.

[0100] In some embodiments, the device 300 is a computer device or includes a computer device therein. The computer device includes a processor, a memory and a network interface connected by a system bus. The processor of the computer device is configured to provide computing and control capabilities. The memory of the computer device includes a non-volatile storage medium and an internal memory. The non-volatile storage medium stores an operating system, a computer program and a database. The internal memory provides an environment for running the operating system and the computer program in the non-volatile storage medium. The database of the computer device is configured to store data. The network interface of the computer device is configured to communicate with other terminals or servers outside through a network connection. In some embodiments, the network interface can be a wired network interface, and in some embodiments, the network interface can also be a wireless network interface. The computer program is executed by the processor to implement the method in the embodiments of the present application.

[0101] Those skilled in the art can understand that, Figure 3 The structure shown in the above-mentioned embodiments is only a block diagram of part of the structure related to the scheme of the present application, and does not constitute a limitation on the computer device to which the scheme of the present application is applied. The specific computer device can include more or fewer components than those shown in the figure, or combine certain components, or have a different arrangement of components.

[0102] The above-described embodiments are only used to illustrate the technical solutions of the present application, but not limit the technical solutions thereof. Even though the technical solutions of the present application have been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that they can still modify the technical solutions recorded in the foregoing embodiments, or make equivalent replacements to some technical features thereof. Such modifications or replacements do not cause the essence of the corresponding technical solutions to deviate from the scope of the technical solutions of the embodiments of the present application.

[0103] In the above-described embodiments, according to the context, the term "when" can be interpreted as meaning "if" or "after" or "in response to determining" or "in response to detecting". Similarly, according to the context, the phrase "upon determining" or "if detecting (the stated condition or event)" can be interpreted as meaning "if determining" or "in response to determining" or "upon detecting (the stated condition or event)" or "in response to detecting (the stated condition or event)".

[0104] In the above embodiments, all or part of the embodiments can be implemented by software, hardware, firmware or any combination thereof. When implemented by software, all or part of the embodiments can be implemented in the form of a computer program product. The computer program product includes one or more computer instructions. When the computer program instructions are loaded and executed on a computer, all or part of the processes or functions described in the embodiments of the present application are generated. The computer can be a general purpose computer, a special purpose computer, a computer network, or other programmable apparatus. The computer instructions can be stored in a computer readable storage medium or transmitted from one computer readable storage medium to another computer readable storage medium, for example, the computer instructions can be transmitted from one website, computer, server or data center to another website, computer, server or data center through wired (such as coaxial cable, optical fiber, digital subscriber line) or wireless (such as infrared, wireless, microwave, etc.) manner. The computer readable storage medium can be any available medium that can be accessed by a computer or a data storage device such as a server, data center, etc. integrated with one or more available media. The available media can be a magnetic medium (for example, a floppy disk, a hard disk, a magnetic tape), an optical medium (for example, a DVD), or a semiconductor medium (for example, a solid state disk) and the like.

[0105] Those of ordinary skill in the art understand that all or part of the processes in the above embodiments can be implemented by a computer program to instruct the relevant hardware, which can be stored in a computer readable storage medium. The program can include the processes of the above method embodiments when executed. The aforementioned storage medium includes ROM or random access memory (RAM), magnetic disk or optical disk, and various media that can store program codes.

Claims

1. A cell balancing decision-making method for an energy storage system, characterized in that, The energy storage system includes multiple battery cell units, and the method includes: When the operating parameters of the target cell unit are detected to trigger the preset equalization condition, the historical operating data of the target cell unit in the energy storage system within the preset historical period are obtained. The historical operating data includes voltage and current corresponding to the timestamp, and the dynamic state and static state of the target cell unit are determined by the current. After extracting multiple transition events from the dynamic state to the static state of the target cell from the historical operating data, the historical polarization voltage values ​​of the multiple transition events are calculated and a polarization voltage value time series is formed. The historical polarization voltage value is the difference between the voltage values ​​of the target cell before and after the transition event. A linear regression was performed on the time series of the polarization voltage values, and the slope obtained was the long-term polarization growth rate of the target cell unit. Based on the historical operating data, the standard deviation of the current value of the battery pack to which the target cell belongs is calculated under dynamic conditions to obtain the power fluctuation characteristic value. If the power fluctuation characteristic value is higher than the preset power application threshold, then the balancing task of the target cell unit is determined as the first priority. If the power fluctuation characteristic value is not higher than the power application threshold, the recoverable capacity of the target cell unit is calculated based on the preset voltage-state-of-charge curve. After multiplying the urgency coefficient obtained by mapping the long-term polarization growth rate with the recoverable power to obtain a product, the priority of the balancing task of the target cell unit in the balancing task of the energy storage system is determined based on the product, and a balancing decision is generated.

2. The method according to claim 1, characterized in that, After the step of performing linear regression on the time series of the polarization voltage values ​​to calculate the slope as the long-term polarization growth rate of the target cell, the method further includes: A predetermined number of recent transition events are extracted from the time series of polarization voltage values ​​to form a recent observation window; Linear regression was performed on the recent observation window, and the slope obtained was the recent polarization growth rate of the target cell unit. The deterioration trend factor is obtained by calculating the ratio of the recent polarization growth rate to the long-term polarization growth rate; If the deterioration trend factor is within a preset stable range close to 1, then the urgency coefficient is obtained based on the long-term polarization growth rate mapping. If the deterioration trend factor is greater than the upper limit of the stable interval, then after obtaining the basic urgency coefficient based on the long-term polarization growth rate mapping, the product of the deterioration trend factor and the basic urgency coefficient is determined as the urgency coefficient.

3. The method according to claim 2, characterized in that, The method further includes: If the deterioration trend factor is below the lower limit of the stable range, then the urgency coefficient is obtained based on the long-term polarization growth rate mapping.

4. The method according to claim 1, characterized in that, The step of performing linear regression on the time series of the polarization voltage value to calculate the slope as the long-term polarization growth rate of the target cell unit specifically includes: A preset robust regression algorithm is used to perform preliminary fitting on the time series of polarization voltage values ​​to obtain an inlier set that is consistent with the main trend. A linear regression is performed on the set of interior points, and the slope obtained is the long-term polarization growth rate of the target cell unit.

5. The method according to claim 1, characterized in that, After the step of performing linear regression on the time series of the polarization voltage values ​​to calculate the slope as the long-term polarization growth rate of the target cell, the method further includes: Count the number of data points contained in the time series of the polarization voltage values; If the number of data points is less than the preset minimum number of regression data points, the long-term polarization growth rate is determined to be unreliable, and the urgency coefficient of the target cell is assigned a preset default value.

6. The method according to claim 1, characterized in that, The step of multiplying the urgency coefficient obtained by mapping the long-term polarization growth rate with the recoverable energy to obtain a product, and then determining the priority of the balancing task of the target cell unit in the balancing task of the energy storage system based on the product, and generating a balancing decision, specifically includes: Based on the historical equalization records of the target cell unit within a preset equalization period, the total number of times the target cell unit was subjected to equalization operations is obtained as the historical equalization frequency. If the historical equilibrium frequency is higher than the preset equilibrium dependence threshold, the ratio of the preset equilibrium dependence threshold to the historical equilibrium frequency is determined as the revenue decay factor. The product of the revenue decay factor and the recoverable power is used as the corrected recoverable power. After multiplying the urgency coefficient obtained by mapping the long-term polarization growth rate with the recoverable power to obtain a product, the priority of the balancing task of the target cell unit in the balancing task of the energy storage system is determined based on the product, and a balancing decision is generated.

7. The method according to claim 1, characterized in that, The step of multiplying the urgency coefficient obtained by mapping the long-term polarization growth rate with the recoverable energy to obtain a product, and then determining the priority of the balancing task of the target cell unit in the balancing task of the energy storage system based on the product, specifically includes: By normalizing the process, the urgency coefficient obtained by mapping the long-term polarization growth rate is mapped to a preset closed interval to obtain the normalized urgency coefficient. The recoverable power is mapped to the preset closed interval to obtain the normalized recoverable power. Multiply the normalized urgency coefficient by the normalized recoverable power to obtain the priority score; The balancing tasks of the energy storage system are traversed, and the priority sequence of the target battery cell is determined according to the priority score.

8. A device, characterized in that, The device includes: one or more processors and a memory; the memory is coupled to the one or more processors, the memory being used to store computer program code, the computer program code including computer instructions, the one or more processors invoking the computer instructions to cause the device to perform the method as described in any one of claims 1-7.

9. A computer program product containing instructions, characterized in that, When the computer program product is run on the device, the device causes the device to perform the method as described in any one of claims 1-7.

10. A computer-readable storage medium comprising instructions, characterized in that, When the instructions are executed on the device, the device causes the device to perform the method as described in any one of claims 1-7.

Images