A method for dynamic optimization of full-charge capacity of lithium batteries under unbalanced conditions

Abstract

This invention relates to the field of charging technology and specifically discloses a method for dynamically optimizing the full-charge capacity of lithium batteries under unbalanced conditions. The method includes: after the current battery pack completes constant-voltage charging to the cutoff current, dynamically adjusting the resting time based on the average temperature and cycle number of the current battery pack, performing resting to eliminate polarization effects, and obtaining stable open-circuit voltages for each individual cell; obtaining the state of charge of the smallest capacity cell by querying a pre-established open-circuit voltage-state-of-charge characteristic table based on the obtained lowest cell voltage; generating a segmented correction formula based on segmented thresholds and gain coefficients dynamically determined using a large number of full-charge capacity correction examples; updating the full-charge capacity using the segmented correction formula and the state of charge of the smallest capacity cell to obtain the dynamic optimization result of the full-charge capacity; effectively improving the accuracy of full-charge capacity calculation, better reflecting the true full-charge capacity of lithium batteries under unbalanced conditions, extending battery life, and improving battery efficiency.

Description

Technical Field

[0001] This invention relates to the field of charging technology, and in particular to a method for dynamically optimizing the full charge capacity of a lithium battery under unbalanced conditions. Background Technology

[0002] In today's energy sector, lithium batteries, with their high energy density and long cycle life, are widely used in electric vehicles, portable electronic devices, and energy storage systems. However, in practical applications of lithium-ion battery packs, due to the discreteness of manufacturing processes (such as deviations in electrode material ratios and electrolyte injection amounts), variations in operating environments (such as uneven temperature distribution and inconsistent charging / discharging currents), and inconsistent aging degradation), imbalances in parameters such as capacity and internal resistance inevitably occur between individual battery cells. This imbalance severely affects the performance and lifespan of lithium battery packs. For example, in electric vehicles, battery imbalance can lead to shortened driving range, longer charging times, and even compromised driving safety; in energy storage systems, imbalance reduces overall system efficiency and increases maintenance costs. Therefore, accurately estimating the full-charge capacity of lithium batteries under imbalanced conditions and dynamically optimizing it is crucial. This not only improves the efficiency of the battery pack and extends its lifespan but also enhances the reliability and stability of the entire system. With the continuous expansion of lithium battery application scenarios and the increasing demands for energy efficiency, researching dynamic optimization methods for the full-charge capacity of lithium batteries under imbalanced conditions has broad development prospects. Accurate capacity estimation and dynamic optimization help to achieve rational allocation of battery resources and promote the development of related industries in a more efficient and sustainable direction.

[0003] However, existing methods for calculating full-charge capacity have several shortcomings. Traditional full-charge capacity calculation models mainly rely on the number of charge-discharge cycles and battery temperature parameters, which cannot accurately reflect the actual capacity characteristics under unbalanced conditions. At the end of the charge-discharge process, the voltage difference between individual cells can cause the lowest-capacity cell to not reach the full-charge voltage threshold when the battery pack as a whole reaches the charging cutoff condition. Existing methods struggle to address this situation and cannot accurately estimate capacity by introducing a correlation analysis between individual cell voltage differences and actual capacity. This results in insufficient accuracy in capacity estimation when the battery pack is unbalanced, making it impossible to achieve precise dynamic correction of the full-charge capacity (FCC), thereby affecting the reliability and accuracy of the battery management system and failing to meet the performance and lifespan requirements of lithium batteries in practical applications.

[0004] Therefore, this invention proposes a dynamic optimization method for the full charge capacity of lithium batteries under unbalanced conditions. Summary of the Invention

[0005] This invention provides a dynamic optimization method for the full-charge capacity of lithium batteries under unbalanced conditions. After the current battery pack completes constant-voltage charging to the cutoff current, the resting time is dynamically adjusted based on the average temperature and cycle number of the current battery pack to eliminate polarization effects and obtain stable open-circuit voltages for each individual cell. This method can more accurately consider the characteristics of the battery under different usage stages and environments, laying the foundation for accurate subsequent battery state assessment. Based on the lowest cell voltage obtained after resting, the open-circuit voltage-state-of-charge characteristic table is consulted to obtain the state of charge of the smallest capacity cell. In this way, the key parameters of the worst-performing cell in the battery pack can be quickly and accurately determined, providing an important reference for capacity optimization. A large number of full-charge capacity correction examples are used to dynamically determine the segmented threshold and gain coefficient, thereby generating a segmented correction formula. This formula fully incorporates actual usage data, making it more closely reflect the actual characteristics of the battery. By using a segmented correction formula and the state of charge of the smallest single cell to update the full-charge capacity, dynamic optimization results of the full-charge capacity are obtained, which effectively improves the accuracy of the full-charge capacity calculation. This better reflects the true full-charge capacity of lithium batteries under unbalanced conditions, which is conducive to more rational use and management of lithium batteries, extending battery life and improving battery efficiency.

[0006] This invention provides a method for dynamically optimizing the full-charge capacity of a lithium battery under unbalanced conditions, comprising:

[0007] After the current battery pack completes constant voltage charging to the cutoff current, the resting time is dynamically adjusted and a resting treatment is performed based on the average temperature and cycle number of the current battery pack to eliminate polarization effect and obtain stable open circuit voltage of each cell.

[0008] Based on the lowest single-cell voltage obtained after static treatment, the state of charge of the smallest capacity single cell is obtained by querying the pre-established open-circuit voltage-state of charge characteristic table.

[0009] The segmented correction formula is generated based on the segmented threshold and gain coefficient dynamically determined using a large number of full-charge capacity correction instances.

[0010] The full-charge capacity is updated by using a segmented correction formula and the state of charge of the smallest capacity cell, and the dynamic optimization result of the full-charge capacity is obtained.

[0011] Preferably, the resting time is dynamically adjusted based on the current average temperature and cycle number of the battery pack, including:

[0012] The product of the difference between the current average temperature of the battery pack and the preset temperature threshold and the first preset coefficient is taken as the first resting time of the current battery pack.

[0013] The product of the ratio of the number of cycles to the preset number of cycles and the second preset coefficient is taken as the second resting time of the current battery pack;

[0014] The current resting time of the current battery pack is determined based on the first resting time and the second resting time of the current battery pack.

[0015] Preferred options also include:

[0016] The number of cycles is accumulated in real time. When the accumulated number of cycles reaches an integer multiple of 50, the open-circuit voltage-state-of-charge characteristic table calibration process is triggered.

[0017] The preferred open-circuit voltage-state-of-charge characteristic table calibration procedure includes:

[0018] By comparing the capacity retention rate and internal resistance growth rate of each monomer, the monomer with the best health status is determined as the benchmark monomer.

[0019] Perform a preset number of complete charge-discharge cycles on the selected reference cells, simultaneously record the open-circuit voltage values ​​under different states of charge, and take the arithmetic mean of the open-circuit voltage values ​​for each state of charge interval to obtain the state of charge-open-circuit voltage calibration dataset of the reference cells.

[0020] Based on the state-of-charge-open-circuit voltage calibration dataset and cycle count-capacity retention decay curve of the reference cell, the current open-circuit voltage-state-of-charge characteristic table is corrected to obtain the latest open-circuit voltage-state-of-charge characteristic table.

[0021] Preferably, the monomer with the best health status is determined as the benchmark monomer by comparing the capacity retention rate and internal resistance growth rate of each monomer, including:

[0022] Collect the current full charge capacity and current internal resistance of all cells in the current battery pack, and retrieve the initial parameters of all cells in the current battery pack, including nominal capacity and initial internal resistance.

[0023] The capacity retention rate of each cell is determined based on its current full-charge capacity and nominal capacity.

[0024] The internal resistance growth rate of each cell is determined based on the current internal resistance and the initial internal resistance of each cell.

[0025] The health score of each cell is calculated based on its capacity retention rate and internal resistance growth rate.

[0026] The monomer with the highest health score among all monomers is used as the baseline monomer.

[0027] Preferably, based on the state-of-charge-open-circuit voltage calibration dataset and the cycle count-capacity retention decay curve of the reference cell, the current open-circuit voltage-state-of-charge characteristic table is revised to obtain the latest open-circuit voltage-state-of-charge characteristic table, including:

[0028] Generate the cycle count-capacity retention decay curve of the reference monomer based on historical data of the reference monomer;

[0029] The state of charge-open circuit voltage calibration dataset of the reference cell is compared with the open circuit voltage value corresponding to the same state of charge in the current open circuit voltage-state of charge characteristic table, and the deviation is calculated.

[0030] If all deviations show a linear trend, then the open-circuit voltage value corresponding to each state of charge in the current open-circuit voltage-state of charge characteristic table is linearly offset to obtain the latest open-circuit voltage-state of charge characteristic table.

[0031] If all deviations do not show a linear trend, then only replace the open-circuit voltage value in the current open-circuit voltage-state-of-charge characteristic table that exceeds the preset deviation value to obtain the latest open-circuit voltage-state-of-charge characteristic table.

[0032] Preferably, the segmented correction formula is generated based on the segmented threshold and gain coefficient dynamically determined using a large number of full-charge capacity correction instances, including:

[0033] Obtain a preset number of full charge capacity correction instances of the same scale as the current battery pack. Each full charge capacity correction instance includes the state of charge of the smallest capacity cell at the time of correction, the corrected full charge capacity, and the actual charge and discharge capacity. The correction deviation rate of each full charge capacity correction instance is determined based on the corrected full charge capacity and the actual charge and discharge capacity in each full charge capacity correction instance.

[0034] Based on a preset percentage interval, the state of charge of the smallest capacity cell in all full-charge capacity correction instances is divided into preset intervals, and the variance of the correction deviation rate of the full-charge capacity correction instances corresponding to the state of charge of all the smallest capacity cells in each interval is calculated as the deviation variance of each interval.

[0035] Calculate the variance difference of each group of adjacent intervals, and take the endpoints of the adjacent intervals of the corresponding group whose variance difference is not less than a preset multiple of the average variance difference of all groups of adjacent intervals as candidate threshold points.

[0036] Select the segmented threshold from all candidate threshold points;

[0037] The gain coefficient is dynamically fitted based on the segmented threshold and a large number of full-charge capacity correction examples with uniform capacity retention rate distribution, and a segmented correction formula is generated based on the segmented threshold and the gain coefficient.

[0038] Preferably, segmented thresholds are selected from all candidate threshold points, including:

[0039] Using each candidate threshold point as the dividing criterion, the state of charge of the smallest capacity cell in all full-charge capacity correction instances is divided into one group of state of charge not exceeding the corresponding candidate threshold point and another group of state of charge exceeding the corresponding candidate threshold point. The sum of the variances of the two groups of state of charge at each candidate threshold point is taken as the inferiority of the corresponding candidate threshold point.

[0040] The candidate threshold with the lowest inferiority among all candidate threshold points is used as the segment threshold.

[0041] Preferably, the gain coefficient is dynamically fitted based on segmented thresholds and a large number of fully charged capacity correction instances with uniform capacity retention distribution, including:

[0042] Based on the capacity retention rate range of each aging level, a large number of full-charge capacity correction instances with uniform capacity retention rate distribution are divided into correction instance groups under each aging level.

[0043] Calculate the theoretical deviation for each full-charge capacity correction instance without introducing a gain factor;

[0044] In each aging level correction group, all correction instances in which the state of charge of the smallest capacity cell does not exceed the corresponding segment threshold are taken as the first correction instance set, and all correction instances in which the state of charge of the smallest capacity cell exceeds the corresponding segment threshold are taken as the second correction instance set.

[0045] Based on the relationship between the gain coefficient and the base coefficient and the aging effect coefficient when the state of charge of the minimum capacity cell does not exceed the corresponding segment threshold, and the theoretical deviation of all full-charge capacity correction instances in the first correction instance set under each aging level without introducing the gain coefficient, the relationship between the gain coefficient and the capacity retention rate when the state of charge of the minimum capacity cell does not exceed the corresponding segment threshold under each aging level is fitted.

[0046] Based on the relationship between the gain coefficient and the base coefficient and the aging effect coefficient when the state of charge of the minimum capacity cell does not exceed the corresponding segment threshold, and the theoretical deviation of all full-charge capacity correction instances in the second correction instance set under each aging level without introducing the gain coefficient, the relationship between the gain coefficient and the capacity retention rate when the state of charge of the minimum capacity cell exceeds the corresponding segment threshold under each aging level is fitted.

[0047] Based on the capacity retention rate of the smallest capacity cell in the current battery pack and the relationship expression between all gain coefficients and capacity retention rate, all gain coefficients in the piecewise correction formula are determined.

[0048] Preferably, all gain coefficients in the piecewise correction formula are determined based on the capacity retention rate of the smallest capacity cell in the current battery pack and the relationship expression between all gain coefficients and capacity retention rate, including:

[0049] Substitute the capacity retention rate of the smallest capacity cell in the current battery pack into the relationship expression between the gain coefficient and the capacity retention rate when the state of charge of the smallest capacity cell does not exceed the corresponding segment threshold, and the relationship expression between the gain coefficient and the capacity retention rate when the state of charge of the smallest capacity cell exceeds the corresponding segment threshold, respectively, to obtain the gain coefficient when the state of charge of the smallest capacity cell does not exceed the corresponding segment threshold and the gain coefficient when the state of charge of the smallest capacity cell exceeds the corresponding segment threshold.

[0050] The beneficial effects of this invention compared to existing technologies are as follows: After the current battery pack completes constant voltage charging to the cutoff current, the resting time is dynamically adjusted based on the average temperature and cycle number of the current battery pack to eliminate polarization effects and obtain stable open-circuit voltages for each individual cell. This allows for more accurate consideration of battery characteristics under different usage stages and environments, laying the foundation for accurate subsequent battery state assessment. Based on the lowest individual cell voltage obtained after resting, the open-circuit voltage-state-of-charge characteristic table is consulted to obtain the state of charge of the smallest capacity cell. This method allows for quick and accurate determination of key parameters of the worst-performing cell in the battery pack, providing an important reference for capacity optimization. Numerous full-charge capacity correction examples are used to dynamically determine segmented thresholds and gain coefficients, thereby generating segmented correction formulas that fully incorporate actual usage data, making the correction formulas more closely reflect the true characteristics of the battery. The full-charge capacity is updated using the segmented correction formula and the state of charge of the smallest capacity cell, obtaining dynamically optimized full-charge capacity results. This effectively improves the accuracy of full-charge capacity calculation, better reflecting the true full-charge capacity of lithium batteries under unbalanced conditions. This facilitates more rational use and management of lithium batteries, extends battery life, and improves battery efficiency.

[0051] Other features and advantages of the invention will be set forth in the description which follows, and will be apparent in part from the description, or may be learned by practicing the invention. The objects and other advantages of the invention may be realized and obtained by means of the structures particularly pointed out in this application.

[0052] The technical solution of the present invention will be further described in detail below with reference to the accompanying drawings and embodiments. Attached Figure Description

[0053] The accompanying drawings are provided to further illustrate the invention and form part of the specification. They are used in conjunction with embodiments of the invention to explain the invention and do not constitute a limitation thereof. In the drawings:

[0054] Figure 1 This is a flowchart of the dynamic optimization method for full-charge capacity of lithium battery under unbalanced state in an embodiment of the present invention.

[0055] Figure 2 This is a schematic diagram of the process for determining the settling time in an embodiment of the present invention;

[0056] Figure 3 This is a schematic diagram of the open-circuit voltage-state-of-charge characteristic table calibration process in an embodiment of the present invention. Detailed Implementation

[0057] The preferred embodiments of the present invention will be described below with reference to the accompanying drawings. It should be understood that the preferred embodiments described herein are for illustration and explanation only and are not intended to limit the present invention.

[0058] like Figure 1 As shown, this invention provides an implementation method for dynamically optimizing the full-charge capacity of a lithium battery under unbalanced conditions, comprising:

[0059] After the current battery pack completes constant voltage charging to the cutoff current, the resting time is dynamically adjusted and a resting treatment is performed based on the average temperature and cycle number of the current battery pack to eliminate polarization effect and obtain stable open circuit voltage of each cell.

[0060] Based on the lowest single-cell voltage obtained after static treatment, the state of charge of the smallest capacity single cell is obtained by querying the pre-established open-circuit voltage-state of charge characteristic table.

[0061] The segmented correction formula is generated based on the segmented threshold and gain coefficient dynamically determined using a large number of full-charge capacity correction instances.

[0062] The full-charge capacity is updated by using a segmented correction formula and the state of charge of the smallest capacity cell, and the dynamic optimization result of the full-charge capacity is obtained.

[0063] In this embodiment, the battery pack completes constant-voltage charging to the cutoff current: During the charging process, the lithium battery pack first uses constant-current charging to increase the battery voltage. Once the voltage reaches a set constant-voltage value, it switches to constant-voltage charging mode. As charging progresses, the charging current gradually decreases until it drops to a pre-set cutoff current value. At this point, the constant-voltage charging of the battery pack is considered complete. This is a crucial node in the battery charging process, marking the end of the regular charging phase and preparing for subsequent precise assessment of the battery's state and capacity optimization. For example, when charging an electric vehicle battery pack, this state is reached when the charging current drops to a cutoff current such as 0.05C (where C is the battery's rated capacity).

[0064] In this embodiment, the average temperature and cycle count of the current battery pack are as follows: The average temperature reflects the current thermal environment of the battery pack, because temperature has a significant impact on the performance and aging process of lithium batteries. The charge-discharge characteristics, internal resistance, and other parameters of the battery will differ at different temperatures. The cycle count records the number of complete cycles from full charge to full discharge and back to full charge. As the cycle count increases, the battery will gradually age, and performance parameters such as capacity and internal resistance will also change.

[0065] In this embodiment, the resting time refers to the length of time the battery pack needs to rest after completing constant voltage charging to the cutoff current. This time is not a fixed value, but is dynamically adjusted according to the current average temperature and cycle count of the battery pack. The purpose of resting is to allow the battery to reach a more stable state internally, so as to obtain parameters such as the open-circuit voltage of individual battery cells more accurately. It is an important time parameter in the entire dynamic optimization process of full charge capacity.

[0066] In this embodiment, a resting treatment is used to eliminate polarization effects: During the charging and discharging process of a lithium battery, the current passing through the electrodes and electrolyte causes uneven ion concentration distribution on and inside the electrodes, resulting in polarization. This affects the accurate measurement of the battery voltage. The resting treatment involves allowing the battery to remain in a static state without charging or discharging after charging is complete. This allows the ion concentration on and inside the electrodes to redistribute and tend towards equilibrium, thereby eliminating the polarization effect. This ensures that the measured battery voltage more accurately reflects the actual state of the battery, providing the necessary conditions for subsequently obtaining the open-circuit voltage accurately.

[0067] In this embodiment, stable open-circuit voltages of each individual cell are obtained: after a period of rest to eliminate polarization effects, the voltage value of each battery cell in an open-circuit (i.e., not connected to an external circuit) state is measured. These open-circuit voltage values ​​can reflect information such as the current state of charge of the battery cell. Furthermore, because the internal state of the battery is stabilized after resting, the open-circuit voltages obtained at this time are relatively accurate and reliable, providing crucial data for further determining the state of charge of the smallest capacity cell in the battery pack and for subsequent optimization calculations of the full-charge capacity.

[0068] In this embodiment, the lowest single-cell voltage is the smallest value among the stable open-circuit voltages of all individual cells. Because of the imbalance among cells in the battery pack, the cell corresponding to the lowest voltage is often the one with relatively small capacity or poor health. By focusing on this voltage value, the cell in the worst condition can be quickly located, providing an important reference for subsequent accurate evaluation of the overall performance and full-charge capacity of the battery pack.

[0069] In this embodiment, a pre-established open-circuit voltage-state-of-charge characteristic table is used. This table is obtained through experiments or statistical analysis of a large amount of data. It records the state of charge (SOC, State of Charge, which represents the percentage of the battery's remaining capacity) of a single lithium battery cell under different open-circuit voltages.

[0070] In this embodiment, based on the lowest single-cell voltage obtained after the resting process, the state of charge of the smallest capacity single cell is obtained by querying a pre-established open-circuit voltage-state of charge characteristic table: after obtaining the lowest single-cell voltage after the resting process, the current state of charge of the smallest capacity single cell in the battery pack is determined by using the pre-established open-circuit voltage-state of charge characteristic table and looking up the state of charge value corresponding to the lowest single-cell voltage.

[0071] In this embodiment, the full-charge capacity correction example refers to a specific instance where the full-charge capacity is corrected during actual use of the lithium battery. Each example includes information such as the state of charge of the smallest capacity cell at the time of correction, the corrected full-charge capacity, and the actual charge / discharge capacity.

[0072] In this embodiment, the segmentation threshold is a key value obtained through analysis of a large number of full-charge capacity correction examples. This segmentation threshold is used to divide the state of charge range of the smallest capacity cell so that different gain coefficients can be used to correct the full-charge capacity for different ranges, making the calculation of the full-charge capacity more consistent with the actual characteristics of the battery.

[0073] In this embodiment, the gain coefficient is a coefficient obtained by fitting based on segmented thresholds and a large number of full-charge capacity correction examples with uniform capacity retention. The gain coefficient is used to adjust the calculation of full-charge capacity, making the calculation result closer to the actual full-charge capacity of the battery, thereby improving the accuracy of full-charge capacity calculation.

[0074] In this embodiment, the segmented correction formula is generated using segmented thresholds and gain coefficients dynamically determined from a large number of full-charge capacity correction examples. This formula calculates the full-charge capacity correction based on different gain coefficients for the state of charge (SCC) of the smallest capacity cell, according to different ranges (divided by segmented thresholds). It comprehensively considers the characteristics of the battery under different aging levels and SCCs, and compared to traditional fixed-mode calculation methods, it more accurately reflects the true full-charge capacity of lithium batteries under unbalanced conditions, providing a core calculation method for dynamic optimization of full-charge capacity. For example:

[0075]

[0076] In the formula, For the updated full charge capacity, This is the full charge capacity before the update. The state of charge of the smallest capacity cell. The gain coefficient when the state of charge of the smallest capacity cell does not exceed the corresponding segment threshold. This is the gain coefficient when the state of charge of the smallest capacity cell exceeds the corresponding segment threshold.

[0077] In this embodiment, a segmented correction formula and the state of charge (SOC) of the smallest capacity cell are used to update the full-charge capacity, resulting in a dynamically optimized full-charge capacity. The obtained SOC of the smallest capacity cell is substituted into the segmented correction formula, and calculations are performed according to parameters such as the gain coefficient corresponding to the SOC range in the formula. This corrects and updates the original full-charge capacity, resulting in a dynamically optimized result that better reflects the actual full-charge capacity of the lithium battery under the current unbalanced state. This result more accurately reflects the actual chargeable capacity of the battery, helping to use and manage lithium batteries more rationally. For example, in electric vehicles, this optimization result can be used to more accurately estimate the driving range, extend battery life, and improve battery efficiency.

[0078] like Figure 2 As shown, in order to comprehensively consider the current temperature and cycle count of the battery pack and accurately determine the resting time that can effectively eliminate polarization effects, it is further proposed to dynamically adjust the resting time based on the current average temperature and cycle count of the battery pack, including:

[0079] The product of the difference between the current average temperature of the battery pack and the preset temperature threshold (e.g., 25 degrees Celsius) and the first preset coefficient (e.g., 0.1 min / ℃) is taken as the first resting time of the current battery pack.

[0080] The product of the ratio of the number of cycles to the preset number of cycles (e.g., 1000) and the second preset coefficient (e.g., 1 min) is used as the second resting time for the current battery pack.

[0081] The current resting time of the current battery pack is determined based on the first resting time and the second resting time of the current battery pack.

[0082] In this embodiment, the first preset coefficient is a pre-set numerical constant used to adjust the influence of the difference between the average temperature of the battery pack and the preset temperature threshold on the first resting time during the calculation of the battery pack resting time. It is an empirical value or a coefficient optimized through experimental data, and its specific value depends on factors such as the type and characteristics of the battery and the actual application scenario. The purpose is to make the resting time calculated based on temperature factors more consistent with the actual needs of the battery to eliminate polarization effects.

[0083] In this embodiment, the product of the difference between the current average temperature of the battery pack and a preset temperature threshold, and a first preset coefficient, is used as the first resting time for the current battery pack. This is a method for calculating a portion of the resting time based on the current average temperature of the battery pack. The preset temperature threshold is a set reference temperature value. Subtracting the current average temperature of the battery pack from this threshold yields a difference that reflects the degree of deviation between the current temperature and the reference temperature. This difference is then multiplied by the first preset coefficient, and the result is the first resting time. This is done because the polarization effect of the battery is greatly affected by temperature, and the time required to eliminate the polarization effect varies at different temperatures. This calculation method allows for dynamic adjustment of a portion of the resting time based on the actual temperature conditions to more accurately meet the requirements for eliminating the polarization effect. For example, if the preset temperature threshold is 25°C, the current average temperature of the battery pack is 30°C, and the first preset coefficient is 2, then the first resting time is (30-25)×2=10 minutes.

[0084] In this embodiment, the second preset coefficient is also a predetermined numerical constant used to adjust the effect of the number of cycles on the second resting time when calculating the resting time. Similar to the first preset coefficient, it is also obtained based on battery characteristics, usage experience, or experimental optimization. Its value is related to the battery's aging characteristics, cycle life, etc., aiming to make the resting time calculated based on the number of cycles better suited to the battery's need to eliminate polarization effects at different stages of use.

[0085] In this embodiment, the product of the ratio of the number of cycles to the preset number of cycles and the second preset coefficient is used as the second resting time for the current battery pack. This calculation method takes into account the influence of the number of battery cycles on the resting time. The preset number of cycles is a standard number of cycles used as a reference. Comparing the current number of cycles of the battery pack with the preset number of cycles yields a ratio that reflects the degree of the battery's current cycle state relative to the reference state. Multiplying this ratio by the second preset coefficient gives the second resting time. Since the internal structure and performance of the battery change as the number of cycles increases, the time for eliminating polarization effects may also change. This calculation allows for dynamic adjustment of another part of the resting time based on the battery's usage stage. For example, if the preset number of cycles is 100, the current number of cycles is 50, and the second preset coefficient is 3, then the second resting time is (50 / 100) × 3 = 1.5 minutes.

[0086] In this embodiment, the current resting time of the current battery pack is determined based on the first resting time and the second resting time of the current battery pack: for example, if the first resting time is 10 minutes and the second resting time is 1.5 minutes, then the current resting time is the sum of the two, that is, 11.5 minutes.

[0087] To ensure the accuracy of the open-circuit voltage-state-of-charge characteristic table and to allow it to be updated according to battery usage, further measures are proposed, including:

[0088] The number of cycles is accumulated in real time. When the accumulated number of cycles reaches an integer multiple of 50 (e.g., 50, 100, 150, etc.), the open circuit voltage-state of charge characteristic table calibration process is triggered.

[0089] In this embodiment, real-time cumulative cycle count refers to the number of charge-discharge cycles completed by the lithium battery pack, continuously tracked and recorded by the Battery Management System (BMS). Each complete process of the battery pack going from a fully charged state, through discharge until depletion, and then recharged back to a fully charged state is counted as one cycle. The BMS continuously monitors the battery status during battery use, and increments the cycle count by 1 each time such a cycle is completed. For example, in the usage scenario of electric vehicles, the BMS continuously accumulates the cycle count of the battery pack. When the accumulated cycle count reaches a certain value, such as an integer multiple of 50, it can trigger operations such as the open-circuit voltage-state-of-charge characteristic table calibration process to calibrate the relevant characteristic tables according to the current aging level of the battery, thereby more accurately assessing the battery status and optimizing the calculation of full-charge capacity.

[0090] like Figure 3 As shown, in order to select the cell with the best health condition and obtain calibration data through multiple charge-discharge cycles, thereby correcting the open-circuit voltage-state-of-charge characteristic table, a calibration process for the open-circuit voltage-state-of-charge characteristic table is further proposed, including:

[0091] By comparing the capacity retention rate and internal resistance growth rate of each monomer, the monomer with the best health status is determined as the benchmark monomer.

[0092] Perform a preset number of complete charge-discharge cycles on the selected reference cells, simultaneously record the open-circuit voltage values ​​under different states of charge, and take the arithmetic mean of the open-circuit voltage values ​​for each state of charge interval to obtain the state of charge-open-circuit voltage calibration dataset of the reference cells.

[0093] Based on the state-of-charge-open-circuit voltage calibration dataset and cycle count-capacity retention decay curve of the reference cell, the current open-circuit voltage-state-of-charge characteristic table is corrected to obtain the latest open-circuit voltage-state-of-charge characteristic table.

[0094] In this embodiment, the state of charge (SOC) range is defined as follows: To facilitate the analysis and management of the battery's SOC, the entire SOC range is divided into multiple smaller intervals. SOC represents the percentage of remaining battery capacity relative to its rated capacity, typically ranging from 0% to 100%. Dividing the SOC range allows for a more detailed study of the battery's characteristics at different charge levels. For example, the 0%-100% SOC range can be further divided into intervals such as 0-10%, 10%-20%, and 20%-30%, each representing the battery's state within a specific charge range.

[0095] In this embodiment, the arithmetic mean of the open-circuit voltage values ​​for each state of charge (SOC) interval is taken to obtain the SOC-open-circuit voltage calibration dataset for the reference cell. During a preset number of complete charge-discharge cycles on the selected reference cell, the open-circuit voltage values ​​under different SOC states are recorded simultaneously. Since the open-circuit voltage values ​​recorded within each SOC interval may fluctuate, an arithmetic mean is calculated for all recorded open-circuit voltage values ​​within each SOC interval to obtain more representative data. The average open-circuit voltage values ​​corresponding to each SOC interval are then grouped together to form the SOC-open-circuit voltage calibration dataset for the reference cell. For example, SOC=100% corresponds to OCV1, SOC=90% corresponds to OCV2, ..., SOC=0% corresponds to OCV10. This dataset reflects the relatively stable open-circuit voltage characteristics of the reference cell under different SOC states, providing accurate and reliable data for subsequent correction of the open-circuit voltage-SOC characteristic table, helping to improve the accuracy of the characteristic table, and thus more accurately determining the battery's SOC through open-circuit voltage.

[0096] In this embodiment, the cycle count-capacity retention rate decay curve is plotted with the number of battery cycles on the x-axis and the capacity retention rate on the y-axis. The capacity retention rate refers to the ratio of the battery's current fully charged capacity to its initial nominal capacity, reflecting the degree to which the battery retains its capacity relative to its initial capacity after multiple cycles. As the number of cycles increases, various irreversible chemical reactions and physical changes occur inside the battery, leading to a gradual decrease in capacity and a decline in the capacity retention rate. The cycle count-capacity retention rate decay curve is generated by analyzing and processing historical data from a benchmark cell (usually a relatively stable and representative cell in the battery pack). This curve visually demonstrates the trend of battery capacity decay with increasing cycle count, helping to understand the aging process of the battery.

[0097] To select the cell with the best health status as the calibration benchmark cell by calculating capacity retention rate, internal resistance growth rate, and obtaining a health score, a further proposal is made to determine the cell with the best health status as the benchmark cell by comparing the capacity retention rate and internal resistance growth rate of each cell, including:

[0098] Collect the current full charge capacity and current internal resistance of all cells in the current battery pack, and retrieve the initial parameters of all cells in the current battery pack, including nominal capacity and initial internal resistance.

[0099] The capacity retention rate of each cell is determined based on its current full-charge capacity and nominal capacity.

[0100] The internal resistance growth rate of each cell is determined based on the current internal resistance and the initial internal resistance of each cell.

[0101] The health score of each cell is calculated based on its capacity retention rate and internal resistance growth rate.

[0102] The monomer with the highest health score among all monomers is used as the baseline monomer.

[0103] In this embodiment, a single cell refers to an individual battery unit that makes up a lithium battery pack. Due to factors such as manufacturing processes and usage environments, the performance parameters of individual cells in a lithium battery pack may differ, and the characteristics of these cells collectively affect the performance of the entire battery pack. For example, in a large lithium battery pack for an electric vehicle, hundreds or thousands of such cells may be combined in series, parallel, or other ways. Monitoring and analyzing the performance parameters of each cell helps to comprehensively understand the state of the battery pack, especially when the cells are in an unbalanced state, where the differences between cells have a significant impact on the performance of the battery pack.

[0104] In this embodiment, the current full-charge capacity and current internal resistance are defined as follows: Current full-charge capacity refers to the maximum amount of energy a single battery cell can hold at the time of measurement, reflecting the battery's current actual energy storage capacity. As the battery is used and ages, the current full-charge capacity gradually decreases. Current internal resistance refers to the resistance value exhibited by the battery to current flow in its current state. The magnitude of internal resistance affects the battery's charging and discharging efficiency, heat generation, and output voltage stability. Measuring the current full-charge capacity and current internal resistance of a single cell allows for real-time monitoring of individual battery performance changes and is an important basis for assessing battery health and optimizing full-charge capacity calculations. For example, a new battery's current full-charge capacity is close to its nominal capacity, and its internal resistance is relatively low; however, after multiple charge-discharge cycles, the current full-charge capacity decreases, and the internal resistance may increase.

[0105] In this embodiment, nominal capacity and initial internal resistance are used: Nominal capacity is the amount of electricity a battery can store under ideal conditions (such as standard temperature, charge / discharge rate, etc.). It is the rated capacity value set during battery manufacturing and is usually used as a reference standard for measuring battery capacity. Initial internal resistance is the internal resistance of a brand new, unused battery, representing its resistive characteristics in its initial state. These two parameters are inherent properties of the battery at their initial stage. Comparing them with the current fully charged capacity and current internal resistance clearly shows the performance changes of the battery during use. For example, by comparing the current fully charged capacity with the nominal capacity, the degree of battery capacity degradation can be understood; by comparing the current internal resistance with the initial internal resistance, the increase in internal resistance can be determined, thereby assessing the aging state of the battery.

[0106] In this embodiment, the capacity retention rate of each cell is determined based on its current full-charge capacity and nominal capacity. Capacity retention rate is an important indicator for assessing battery health. It is calculated by dividing the current full-charge capacity of each cell by its nominal capacity; the resulting ratio is the capacity retention rate of that cell. For example, if a cell has a nominal capacity of 100 Ah and its current full-charge capacity is measured to be 80 Ah, then the capacity retention rate of that cell is 80 Ah ÷ 100 Ah = 0.8 (or 80%). A higher capacity retention rate indicates less capacity decay during use and a relatively better battery health. By calculating the capacity retention rate of each cell, the capacity retention of each cell in the battery pack can be compared intuitively, providing a basis for selecting the cells with the best health.

[0107] In this embodiment, the internal resistance growth rate of each cell is determined based on its current and initial internal resistance. The internal resistance growth rate reflects the increase in battery internal resistance over time or cycle count. The calculation involves subtracting the initial internal resistance from the current internal resistance of each cell to obtain the growth value, and then dividing this growth value by the initial internal resistance to obtain the internal resistance growth rate. For example, if a cell has an initial internal resistance of 0.05Ω and a current internal resistance measurement of 0.06Ω, then the internal resistance growth rate is (0.06Ω - 0.05Ω) ÷ 0.05Ω = 0.2 (or 20%). A higher internal resistance growth rate indicates a faster increase in battery internal resistance, which may mean more irreversible changes have occurred inside the battery, affecting battery performance. Calculating the internal resistance growth rate of each cell helps to understand the trend of internal resistance changes in each cell and also provides important information for assessing the health status of the cells.

[0108] In this embodiment, a health score is calculated for each cell based on its capacity retention rate and internal resistance growth rate. To comprehensively assess the health status of each cell, these two key indicators are combined to calculate the health score. The specific calculation method may involve assigning certain weights to the capacity retention rate and internal resistance growth rate, and then calculating according to the corresponding formulas. For example, assuming the weight of the capacity retention rate is 0.6 and the weight of the internal resistance growth rate is 0.4, and a cell has a capacity retention rate of 0.8 and an internal resistance growth rate of 0.2, the health score calculation formula is: Health Score = 0.8 × 0.6 + (1 - 0.2) × 0.4 = 0.64. The health score obtained in this way can more comprehensively reflect the health status of the cell.

[0109] To address the deviation between the open-circuit voltage-open-circuit voltage calibration dataset and the current characteristic table, and to specifically correct the open-circuit voltage-open-circuit voltage characteristic table, a further proposal is made to revise the current open-circuit voltage-open-circuit voltage characteristic table based on the open-circuit voltage-open-circuit voltage calibration dataset and the cycle number-capacity retention decay curve of a reference cell, resulting in the latest open-circuit voltage-open-circuit voltage characteristic table, including:

[0110] Generate the cycle count-capacity retention decay curve of the reference monomer based on historical data of the reference monomer;

[0111] The state of charge-open circuit voltage calibration dataset of the reference cell is compared with the open circuit voltage value corresponding to the same state of charge in the current open circuit voltage-state of charge characteristic table, and the deviation is calculated.

[0112] If all deviations show a linear trend, then the open-circuit voltage value corresponding to each state of charge in the current open-circuit voltage-state of charge characteristic table is linearly offset to obtain the latest open-circuit voltage-state of charge characteristic table.

[0113] If all deviations do not show a linear trend, then only replace the open-circuit voltage value in the current open-circuit voltage-state-of-charge characteristic table that exceeds the preset deviation value to obtain the latest open-circuit voltage-state-of-charge characteristic table.

[0114] In this embodiment, a cycle count-capacity retention rate decay curve for a reference cell is generated based on historical data of the reference cell. The reference cell is typically a representative and relatively stable cell selected from the battery pack. By collecting capacity retention rate data of this reference cell at different cycle counts, a curve is plotted using mathematical methods (such as curve fitting) with cycle count as the x-axis and capacity retention rate as the y-axis. This curve depicts the decay trend of the reference cell's capacity retention rate as the cycle count increases. For example, the curve is generated by recording the capacity retention rate of the reference cell at different cycle counts, such as the 10th, 20th, and 30th cycles, through multiple experiments.

[0115] In this embodiment, the state-of-charge (POC)-open-circuit voltage (OPV) calibration dataset of the reference cell is compared with the open-circuit voltage value corresponding to the same POC in the current open-circuit voltage-POC characteristic table, and the deviation is calculated. The POC-POC calibration dataset of the reference cell is obtained by performing a preset number of complete charge-discharge cycles on the reference cell and taking the arithmetic mean of the open-circuit voltage values ​​for each POC interval. The open-circuit voltage value corresponding to each POC in this dataset is subtracted from the open-circuit voltage value for the same POC in the current open-circuit voltage-POC characteristic table, and the difference is the deviation. For example, if the open-circuit voltage is 3.8V when the POC is 50% in the calibration dataset, while the open-circuit voltage corresponding to 50% POC in the characteristic table is 3.75V, then the deviation is 3.8V - 3.75V = 0.05V.

[0116] In this embodiment, it is determined whether all deviations exhibit a linear trend: the calculated deviations for each state of charge are analyzed to observe whether these deviations show linear changes. A linear trend means that as the state of charge changes, the change in deviation can be approximated by a straight line, i.e., there is a linear relationship between the deviation and the state of charge. Determining whether the deviations are linear helps determine what method to use to correct the open-circuit voltage-state of charge characteristic table. If the deviations show a linear trend, it indicates that the deviation in the characteristic table may be due to some systematic factor, and a linear correction method is suitable; if they do not show a linear trend, it indicates that the deviation may be local and irregular, requiring different correction strategies. This can be determined by visually observing whether the relationship between the deviation and the state of charge is approximately a straight line, or by using statistical analysis methods (such as linear regression analysis).

[0117] In this embodiment, a linear offset correction is performed on the open-circuit voltage value corresponding to each state of charge in the current open-circuit voltage-state of charge characteristic table to obtain the latest open-circuit voltage-state of charge characteristic table. When it is determined that all deviations show a linear trend, it means that there is a linear deviation related to the state of charge in the current open-circuit voltage-state of charge characteristic table as a whole. To make the characteristic table more consistent with the actual situation, a uniform linear adjustment is performed on the open-circuit voltage value corresponding to each state of charge in the characteristic table. Specifically, based on the linear relationship between the deviation and the state of charge, a linear offset is determined, and then this linear offset is added to (or subtracted from) the open-circuit voltage value corresponding to each state of charge to obtain the corrected open-circuit voltage value, forming the latest open-circuit voltage-state of charge characteristic table. For example, if the linear offset is determined to be 0.05V, then the open-circuit voltage value when the state of charge is 30% in the original characteristic table is 3.6V, which becomes 3.6V + 0.05V = 3.65V after correction. A similar correction is performed on the open-circuit voltage values ​​corresponding to all states of charge to obtain the latest characteristic table.

[0118] In this embodiment, only the open-circuit voltage values ​​in the current open-circuit voltage-state-of-charge characteristic table that exceed a preset deviation are replaced to obtain the latest open-circuit voltage-state-of-charge characteristic table. When the deviation is not linear, it indicates that the deviation in the characteristic table is not systemic but local and irregular. In this case, a preset deviation is set as a criterion. Only when the deviation for a certain state of charge exceeds this preset value is the open-circuit voltage value corresponding to that state of charge in the current open-circuit voltage-state-of-charge characteristic table replaced. The specific replacement value can be the open-circuit voltage value of the corresponding state of charge in the calibration dataset, or adjusted according to a certain algorithm. This corrects only the parts with large deviations, correcting obvious errors in the characteristic table while preserving the reasonable parts of the original characteristic table to the greatest extent, making the corrected characteristic table more consistent with the actual open-circuit voltage-state-of-charge relationship of the battery. For example, if the preset deviation is 0.1V and the deviation for a certain state of charge is 0.15V, exceeding the preset value, then the open-circuit voltage value corresponding to that state of charge in the characteristic table is replaced with the value in the calibration dataset.

[0119] To determine the segmentation threshold and gain coefficient by analyzing a large number of full-charge capacity correction examples, and thus generate a segmentation correction formula, a further proposal is made to generate the segmentation correction formula based on the segmentation threshold and gain coefficient dynamically determined using a large number of full-charge capacity correction examples, including:

[0120] Obtain a preset number (e.g., 1000) of full-charge capacity correction instances of the same scale as the current battery pack. Each full-charge capacity correction instance contains the state of charge of the smallest capacity cell at the time of correction, the corrected full-charge capacity, and the actual charge and discharge capacity. The correction deviation rate of each full-charge capacity correction instance is determined based on the corrected full-charge capacity and the actual charge and discharge capacity in each full-charge capacity correction instance.

[0121] Based on a preset percentage interval, the state of charge of the smallest capacity cell in all full-charge capacity correction instances is divided into preset intervals, and the variance of the correction deviation rate of the full-charge capacity correction instances corresponding to the state of charge of all the smallest capacity cells in each interval is calculated as the deviation variance of each interval.

[0122] Calculate the variance difference of each group of adjacent intervals, and take the endpoints of the corresponding groups of adjacent intervals that are not less than a preset multiple (e.g., 1.5 times) of the average variance difference of all groups of adjacent intervals as candidate threshold points.

[0123] Select the segmented threshold from all candidate threshold points;

[0124] The gain coefficient is dynamically fitted based on the segmented threshold and a large number of full-charge capacity correction examples with uniform capacity retention rate distribution, and a segmented correction formula is generated based on the segmented threshold and the gain coefficient.

[0125] In this embodiment, a battery pack of the same size as the current battery pack refers to other battery packs that have the same or similar characteristics as the battery pack currently being processed in terms of specifications, model, capacity, number of cells, and connection method. Using full-charge capacity correction examples of battery packs of the same size is because battery packs of the same size have a high degree of similarity in performance and aging patterns, and the patterns reflected in these correction examples are more valuable for reference for the current battery pack. For example, if the current battery pack is a 48V electric vehicle battery pack composed of 18650 lithium-ion battery cells, then a battery pack of the same size is another electric vehicle battery pack that is also composed of 18650 lithium-ion battery cells, has similar voltage and capacity, and has the same connection method.

[0126] In this embodiment, the state of charge (SOC) of the smallest capacity cell at the time of correction, the corrected full-charge capacity, and the actual charge / discharge capacity are crucial parameters in each full-charge capacity correction instance. The SOC of the smallest capacity cell at the time of correction represents the state of charge of the smallest capacity cell in the battery pack when the full-charge capacity correction operation is performed. The corrected full-charge capacity is the value obtained after adjusting the original full-charge capacity using a certain method or strategy. The actual charge / discharge capacity is the amount of electricity actually charged and discharged by the battery pack in a complete charge / discharge cycle during actual use. By recording these parameters, the relationship between the full-charge capacity correction method and the actual charge / discharge situation can be analyzed under different minimum capacity cell SOCs, providing data support for optimizing the calculation of full-charge capacity. For example, in a certain correction instance, the SOC of the smallest capacity cell at the time of correction is 30%, the corrected full-charge capacity is 90 Ah, and the actual charge / discharge capacity is 85 Ah.

[0127] In this embodiment, the correction deviation rate for each full-charge capacity correction instance is determined based on the corrected full-charge capacity and the actual charge / discharge capacity in each instance. The correction deviation rate measures the accuracy of each full-charge capacity correction. It is calculated by subtracting the actual charge / discharge capacity from the corrected full-charge capacity, and then dividing by the actual charge / discharge capacity. The result is the correction deviation rate for that instance. It reflects the degree of deviation between the corrected full-charge capacity and the actual charge / discharge capacity. The smaller the deviation rate, the closer the corrected full-charge capacity is to the actual situation, and the more accurate the correction method. For example, if the corrected full-charge capacity in a certain correction instance is 90Ah and the actual charge / discharge capacity is 85Ah, then the correction deviation rate = (90-85)÷85≈0.059 (or 5.9%).

[0128] In this embodiment, a preset percentage interval is used to divide the state of charge (SOC) of the smallest capacity cell in all full-charge capacity correction instances into intervals. By setting an appropriate percentage interval, the SOC range can be reasonably divided into multiple intervals, allowing for a more detailed analysis of the distribution characteristics of the full-charge capacity correction deviation rate within different SOC intervals. For example, if the preset percentage interval is 10%, the SOC range of 0-100% will be divided into multiple intervals such as 0-10%, 10%-20%, and 20%-30%.

[0129] In this embodiment, the variance of the correction deviation rate of the full-charge capacity correction instances corresponding to the state of charge of all the smallest capacity cells in each interval is calculated as the deviation variance of each interval: For each divided state of charge interval, the correction deviation rate of all full-charge capacity correction instances in that interval is statistically analyzed. The variance measures the dispersion of these correction deviation rates relative to the mean. The larger the variance, the greater the fluctuation of the correction deviation rate in that interval, and the more dispersed the data; the smaller the variance, the more concentrated the correction deviation rate is around the mean. For example, in the 0-10% state of charge interval, there are 10 correction instances with correction deviation rates of 3%, 5%, 4%, 6%, 3%, 4%, 5%, 4%, 3%, and 5%, respectively. First, the mean is calculated as (3+5+4+6+3+4+5+4+3+5)÷10=4.2%, and then the variance is calculated to obtain the deviation variance of that interval.

[0130] In this embodiment, the variance difference between adjacent intervals is calculated: the variance differences between two adjacent state-of-charge intervals are subtracted, and the resulting difference reflects the degree of change in the full-charge capacity correction deviation rate characteristic between adjacent intervals. By analyzing these differences, it is possible to identify the boundaries between which state-of-charge intervals show a significant change in the full-charge capacity correction deviation rate characteristic. These points of change are crucial for determining the segmentation threshold. For example, if the variance of the 0-10% interval is 0.0005 and the variance of the 10%-20% interval is 0.0008, then their variance difference = 0.0008 - 0.0005 = 0.0003.

[0131] In this embodiment, the endpoints of adjacent intervals in a group whose variance difference is not less than a preset multiple of the average variance difference of all adjacent intervals are considered candidate threshold points. First, the average variance difference of all adjacent intervals is calculated, and then a preset multiple (e.g., 1.5 times) is set. For each group of adjacent intervals, if their variance difference is not less than this preset multiple of the average, then the endpoints of these adjacent intervals are considered candidate threshold points. These candidate threshold points represent states of charge points where the full-charge capacity correction deviation rate characteristic changes significantly, and may be key nodes for segmenting the state of charge range. The final segmentation thresholds will be further selected from these candidate threshold points. For example, if the average variance difference of all adjacent intervals is 0.0002, the preset multiple is 1.5, and the variance difference of a certain adjacent interval is 0.0005, which is greater than 1.5 times 0.0002 (0.0003), then the endpoints of this adjacent interval are considered candidate threshold points.

[0132] In this embodiment, a large number of full-charge capacity correction examples with uniform capacity retention rate distribution refers to a collection of numerous full-charge capacity correction examples in which the battery capacity retention rate is relatively evenly distributed across various possible value ranges. Using a large number of full-charge capacity correction examples with uniform capacity retention rate distribution comprehensively covers full-charge capacity correction under various battery aging conditions, thereby making the gain coefficient determined based on these examples and the generated segmented correction formula more universal and accurate, and more adaptable to batteries in different aging states. For example, among the collected full-charge capacity correction examples, there are a certain number of examples for each percentage point of capacity retention rate from 80% to 100%, thus ensuring uniform distribution.

[0133] To select the segmented threshold from candidate threshold points that minimizes the sum of variances of instances of full-charge capacity correction, a further step is proposed to select segmented thresholds from all candidate threshold points, including:

[0134] Using each candidate threshold point as the dividing criterion, the state of charge of the smallest capacity cell in all full-charge capacity correction instances is divided into one group of state of charge not exceeding the corresponding candidate threshold point and another group of state of charge exceeding the corresponding candidate threshold point. The sum of the variances of the two groups of state of charge at each candidate threshold point is taken as the inferiority of the corresponding candidate threshold point.

[0135] The candidate threshold with the lowest inferiority among all candidate threshold points is used as the segment threshold.

[0136] In this embodiment, using each candidate threshold point as a dividing criterion, the state of charge of the smallest capacity cell in all full-charge capacity correction instances is divided into one group of state of charge not exceeding the corresponding candidate threshold point and another group of state of charge exceeding the corresponding candidate threshold point:

[0137] After determining the candidate threshold points, each candidate threshold point is used as a dividing line to further screen for the most suitable segmentation threshold. For example, if the state of charge (SOC) value corresponding to a certain candidate threshold point is 80%, then among all full-charge capacity correction instances, those with an SOC of less than or equal to 80% at the time of correction are grouped into one group, and those with an SOC greater than 80% are grouped into another. This is done to analyze the distribution characteristics of full-charge capacity correction instances from the perspective of different SOC ranges, observe the difference in the full-charge capacity correction deviation rate on both sides of the candidate threshold point, and thus evaluate the rationality of using the candidate threshold point as a segmentation threshold.

[0138] In this embodiment, the sum of the variances of the two sets of charge states at each candidate threshold point is taken as the inferiority of the corresponding candidate threshold point:

[0139] For each candidate threshold point, the two states of charge (SOCs) are divided into two groups, and their respective variances are calculated (as mentioned earlier, variances are obtained by calculating the mean and variance of the correction deviation rates of the full-charge capacity correction instances within each group). The variances of these two SOCs are then summed, and the result is defined as the inferiority score of the candidate threshold point. Inferiority can be understood as an indicator of the "poorness" of a candidate threshold point as a segmentation threshold. A smaller inferiority score indicates that the dispersion and fluctuation of the full-charge capacity correction deviation rates of the two SOCs divided by this candidate threshold point are relatively small overall, meaning that the candidate threshold point is more likely to be a suitable segmentation point, making the full-charge capacity correction based on this segmentation more reasonable and accurate. For example, if a candidate threshold point divides two SOCs, one with a variance of 0.0006 and the other with 0.0004, then the inferiority score of this candidate threshold point is 0.0006 + 0.0004 = 0.001. The point with the lowest inferiority will be selected from all candidate threshold points as the final segmentation threshold.

[0140] To classify full-charge capacity correction instances based on capacity retention rate, a relationship expression between the gain coefficient and capacity retention rate under different conditions is fitted using theoretical deviation. Furthermore, a dynamic fitting of the gain coefficient based on a segmented threshold and a large number of full-charge capacity correction instances with uniform capacity retention rate distribution is proposed, including:

[0141] Based on the capacity retention rate range of each aging level, a large number of full-charge capacity correction instances with uniform capacity retention rate distribution are divided into correction instance groups under each aging level.

[0142] Calculate the theoretical deviation for each full-charge capacity correction instance without introducing a gain factor;

[0143] In each aging level correction group, all correction instances in which the state of charge of the smallest capacity cell does not exceed the corresponding segment threshold are taken as the first correction instance set, and all correction instances in which the state of charge of the smallest capacity cell exceeds the corresponding segment threshold are taken as the second correction instance set.

[0144] Based on the relationship between the gain coefficient and the base coefficient and the aging effect coefficient when the state of charge of the minimum capacity cell does not exceed the corresponding segment threshold, and the theoretical deviation of all full-charge capacity correction instances in the first correction instance set under each aging level without introducing the gain coefficient, the relationship between the gain coefficient and the capacity retention rate when the state of charge of the minimum capacity cell does not exceed the corresponding segment threshold under each aging level is fitted.

[0145] Based on the relationship between the gain coefficient and the base coefficient and the aging effect coefficient when the state of charge of the minimum capacity cell does not exceed the corresponding segment threshold, and the theoretical deviation of all full-charge capacity correction instances in the second correction instance set under each aging level without introducing the gain coefficient, the relationship between the gain coefficient and the capacity retention rate when the state of charge of the minimum capacity cell exceeds the corresponding segment threshold under each aging level is fitted.

[0146] Based on the capacity retention rate of the smallest capacity cell in the current battery pack and the relationship expression between all gain coefficients and capacity retention rate, all gain coefficients in the piecewise correction formula are determined.

[0147] In this embodiment, the capacity retention rate range for each aging level is as follows: the capacity retention rate is divided into different ranges, and each range corresponds to an aging level. For example, a capacity retention rate of 90%-100% may be classified as a mild aging level, 70%-90% as a moderate aging level, and below 70% as a severe aging level.

[0148] In this embodiment, based on the capacity retention rate range of each aging level, a large number of full-charge capacity correction instances with uniform capacity retention rate distribution are divided into correction instance groups under each aging level: according to the capacity retention rate range of each aging level determined above, the large number of collected full-charge capacity correction instances with uniform capacity retention rate distribution are classified. For example, for a full-charge capacity correction instance, if the battery capacity retention rate involved is 95%, then the instance is classified into the correction instance group corresponding to the mild aging level.

[0149] In this embodiment, the theoretical deviation of each full-charge capacity correction example without introducing a gain coefficient is calculated. The theoretical deviation refers to the difference between the calculated corrected full-charge capacity and the actual charge / discharge capacity without considering the influence of the gain coefficient on the full-charge capacity correction. The calculation method is to subtract the actual charge / discharge capacity from the corrected full-charge capacity. This theoretical deviation reflects the degree of deviation between the corrected full-charge capacity value and the actual situation without introducing a gain coefficient adjustment, and is used for subsequent analysis of the impact of the gain coefficient on the accuracy of the correction. For example, in a certain full-charge capacity correction example, the corrected full-charge capacity is 92Ah, and the actual charge / discharge capacity is 88Ah. Therefore, the theoretical deviation without introducing a gain coefficient is 92-88=4Ah.

[0150] In this embodiment, based on the relationship between the gain coefficient, the base coefficient, and the aging effect coefficient when the state of charge of the smallest capacity cell does not exceed the corresponding segment threshold, and the theoretical deviation of all full-charge capacity correction instances in the first correction instance set under each aging level without introducing the gain coefficient, a relationship expression between the gain coefficient and the capacity retention rate when the state of charge of the smallest capacity cell does not exceed the corresponding segment threshold is fitted for each aging level: First, there exists an expression describing the relationship between the gain coefficient, the base coefficient, and the aging effect coefficient when the state of charge of the smallest capacity cell does not exceed the corresponding segment threshold (the specific form is determined according to battery characteristics and research). For example, it is set as:

[0151]

[0152] In the formula, The gain coefficient when the state of charge of the smallest capacity cell does not exceed the corresponding segment threshold. Based on the coefficient, The aging effect coefficient. Capacity retention rate;

[0153] For each aging level, the theoretical deviation of each instance in the first corrected instance set corresponding to that aging level (i.e., all corrected instances where the state of charge of the smallest capacity cell does not exceed the corresponding segment threshold) without introducing a gain coefficient is substituted into the above relational expression. Then, using mathematical fitting methods (such as least squares fitting), the intrinsic relationship between the gain coefficient and capacity retention rate is found (i.e., the relationship under this condition is determined). and The value of is used to obtain the relationship expression between the gain coefficient and capacity retention rate under the state of charge condition for each aging level. For example, the low aging group is:

[0154]

[0155] This expression can help determine the appropriate gain coefficient value when the minimum capacity cell state of charge does not exceed the segment threshold at different aging levels, so as to more accurately correct the full charge capacity.

[0156] In this embodiment, based on the relationship between the gain coefficient, the base coefficient, and the aging effect coefficient when the state of charge (SCC) of the minimum capacity cell does not exceed the corresponding segment threshold (the formula is the same as in the previous case), and the theoretical deviation of all full-charge capacity correction instances in the second correction instance set under each aging level without introducing the gain coefficient, a relationship between the gain coefficient and the capacity retention rate when the SCC of the minimum capacity cell exceeds the corresponding segment threshold is fitted to derive the relationship between the gain coefficient, the base coefficient, and the aging effect coefficient under each aging level. Similar to the previous case, the relationship between the gain coefficient, the base coefficient, and the aging effect coefficient when the SCC of the minimum capacity cell does not exceed the corresponding segment threshold is used. For the second correction instance set under each aging level (i.e., all correction instances where the SCC of the minimum capacity cell exceeds the corresponding segment threshold), the theoretical deviation of each instance without introducing the gain coefficient is substituted into this relationship expression. Through mathematical fitting, the relationship between the gain coefficient and the capacity retention rate when the SCC of the minimum capacity cell exceeds the corresponding segment threshold is determined. This allows for accurate determination of the gain coefficient even under different aging levels and when the SCC exceeds the segment threshold, enabling precise correction of the full-charge capacity.

[0157] In this embodiment, the base coefficient is a constant related to the characteristics of the battery itself, playing a fundamental adjusting role in the calculation of the gain coefficient. This coefficient is typically determined based on factors such as battery type, materials, and design, and does not change with the battery's usage status (such as aging level or state of charge). In the relationship expression between the gain coefficient, the base coefficient, and the aging effect coefficient, the base coefficient provides a starting point or basic reference value for the calculation of the gain coefficient, affecting the overall magnitude and trend of the gain coefficient. For example, in a certain gain coefficient calculation formula, the base coefficient may be a fixed value obtained through the analysis of a large amount of experimental data for that type of battery, such as 1.02.

[0158] In this embodiment, the aging effect coefficient reflects the impact of battery aging on the gain coefficient. As the battery ages (reflected by capacity retention), its internal chemical and physical properties change, affecting the correction method for full-charge capacity. The aging effect coefficient is a parameter used to quantify this impact. In the expression relating the gain coefficient to the base coefficient and the aging effect coefficient, the aging effect coefficient works together with the base coefficient to adjust the magnitude of the gain coefficient, allowing it to change reasonably according to the degree of battery aging, thereby more accurately correcting the full-charge capacity under different aging conditions.

[0159] To determine the gain coefficients under different states of charge conditions in the piecewise correction formula based on the capacity retention rate of the smallest capacity cell in the current battery pack, this paper further proposes an expression based on the relationship between the capacity retention rate of the smallest capacity cell in the current battery pack and the capacity retention rate of all gain coefficients to determine all gain coefficients in the piecewise correction formula, including:

[0160] Substitute the capacity retention rate of the smallest capacity cell in the current battery pack into the relationship expression between the gain coefficient and the capacity retention rate when the state of charge of the smallest capacity cell does not exceed the corresponding segment threshold, and the relationship expression between the gain coefficient and the capacity retention rate when the state of charge of the smallest capacity cell exceeds the corresponding segment threshold, respectively, to obtain the gain coefficient when the state of charge of the smallest capacity cell does not exceed the corresponding segment threshold and the gain coefficient when the state of charge of the smallest capacity cell exceeds the corresponding segment threshold.

[0161] In this embodiment, the aim is to determine a suitable gain coefficient for the current battery pack in order to more accurately correct the full-charge capacity. The preceding steps have already yielded expressions relating the gain coefficient to the capacity retention rate under different aging levels, corresponding to when the minimum capacity single-cell state of charge does not exceed and exceeds the segmented threshold.

[0162] To illustrate with a concrete example, suppose the current battery pack has been tested and determined that its smallest capacity cell is at a low aging level, and its capacity retention rate is 60%. Meanwhile, based on previous analysis of numerous full-charge capacity correction examples, two relational expressions for the low aging level were obtained:

[0163]

[0164] At this point, substituting the capacity retention rate of the smallest capacity monomer, 60%, into this expression, we get... The value is 1.0212. By substituting these values ​​into the calculation, the gain coefficients of the smallest capacity cell in the current battery pack under different states of charge are obtained. These gain coefficients will be used in the subsequent piecewise correction formula to accurately correct the full-charge capacity based on the actual state of charge of the smallest capacity cell, so as to obtain a result that is more consistent with the actual full-charge capacity of the battery.

[0165] Obviously, those skilled in the art can make various modifications and variations to this invention without departing from its spirit and scope. Therefore, if these modifications and variations fall within the scope of this invention and its equivalents, this invention also intends to include these modifications and variations.

Claims

1. A method for dynamically optimizing the full-charge capacity of a lithium battery under unbalanced conditions, characterized in that, include: After the current battery pack completes constant voltage charging to the cutoff current, the resting time is dynamically adjusted and a resting treatment is performed based on the average temperature and cycle number of the current battery pack to eliminate polarization effect and obtain stable open circuit voltage of each cell. Based on the lowest single-cell voltage obtained after static treatment, the state of charge of the smallest capacity single cell is obtained by querying the pre-established open-circuit voltage-state of charge characteristic table. A piecewise correction formula is generated based on the piecewise threshold and gain coefficient dynamically determined using a large number of full-charge capacity correction examples, including: Obtain a preset number of full charge capacity correction instances of the same scale as the current battery pack. Each full charge capacity correction instance includes the state of charge of the smallest capacity cell at the time of correction, the corrected full charge capacity, and the actual charge and discharge capacity. The correction deviation rate of each full charge capacity correction instance is determined based on the corrected full charge capacity and the actual charge and discharge capacity in each full charge capacity correction instance. Based on a preset percentage interval, the state of charge of the smallest capacity cell in all full-charge capacity correction instances is divided into preset intervals, and the variance of the correction deviation rate of the full-charge capacity correction instances corresponding to the state of charge of all the smallest capacity cells in each interval is calculated as the deviation variance of each interval. Calculate the variance difference of each group of adjacent intervals, and take the endpoints of the adjacent intervals of the corresponding group whose variance difference is not less than a preset multiple of the average variance difference of all groups of adjacent intervals as candidate threshold points. Select the segmented threshold from all candidate threshold points; The gain coefficient is dynamically fitted based on the segmented threshold and a large number of full-charge capacity correction examples with uniform capacity retention rate distribution, and a segmented correction formula is generated based on the segmented threshold and the gain coefficient. The full-charge capacity is updated by using a segmented correction formula and the state of charge of the smallest capacity cell, and the dynamic optimization result of the full-charge capacity is obtained.

2. The method for dynamic optimization of full-charge capacity of lithium batteries under unbalanced conditions according to claim 1, characterized in that, The resting time is dynamically adjusted based on the current average temperature and cycle count of the battery pack, including: The product of the difference between the current average temperature of the battery pack and the preset temperature threshold and the first preset coefficient is taken as the first resting time of the current battery pack. The product of the ratio of the number of cycles to the preset number of cycles and the second preset coefficient is taken as the second resting time of the current battery pack; The current resting time of the current battery pack is determined based on the first resting time and the second resting time of the current battery pack.

3. The method for dynamic optimization of full-charge capacity of lithium batteries under unbalanced conditions according to claim 1, characterized in that, Also includes: The number of cycles is accumulated in real time. When the accumulated number of cycles reaches an integer multiple of 50, the open-circuit voltage-state-of-charge characteristic table calibration process is triggered.

4. The method for dynamic optimization of full-charge capacity of lithium batteries under unbalanced conditions according to claim 3, characterized in that, The calibration procedure for the open-circuit voltage-state-of-charge characteristic table includes: By comparing the capacity retention rate and internal resistance growth rate of each monomer, the monomer with the best health status is determined as the benchmark monomer. Perform a preset number of complete charge-discharge cycles on the selected reference cells, simultaneously record the open-circuit voltage values ​​under different states of charge, and take the arithmetic mean of the open-circuit voltage values ​​for each state of charge interval to obtain the state of charge-open-circuit voltage calibration dataset of the reference cells. Based on the state-of-charge-open-circuit voltage calibration dataset and cycle count-capacity retention decay curve of the reference cell, the current open-circuit voltage-state-of-charge characteristic table is corrected to obtain the latest open-circuit voltage-state-of-charge characteristic table.

5. The method for dynamic optimization of full-charge capacity of lithium batteries under unbalanced conditions according to claim 4, characterized in that, By comparing the capacity retention and internal resistance growth rate of each monomer, the monomer with the best health status is determined as the benchmark monomer, including: Collect the current full charge capacity and current internal resistance of all cells in the current battery pack, and retrieve the initial parameters of all cells in the current battery pack, including nominal capacity and initial internal resistance. The capacity retention rate of each cell is determined based on its current full-charge capacity and nominal capacity. The internal resistance growth rate of each cell is determined based on the current internal resistance and the initial internal resistance of each cell. The health score of each cell is calculated based on its capacity retention rate and internal resistance growth rate. The monomer with the highest health score among all monomers is used as the baseline monomer.

6. The method for dynamic optimization of full-charge capacity of lithium battery under unbalanced conditions according to claim 4, characterized in that, Based on the state-of-charge (POC)-open-circuit voltage (OPV) calibration dataset and the cycle count-capacity retention decay curve of a reference cell, the current OPC-OPV characteristic table is revised to obtain the latest OPC-OPV characteristic table, including: Generate the cycle count-capacity retention decay curve of the reference monomer based on historical data of the reference monomer; The state of charge-open circuit voltage calibration dataset of the reference cell is compared with the open circuit voltage value corresponding to the same state of charge in the current open circuit voltage-state of charge characteristic table, and the deviation is calculated. If all deviations show a linear trend, then the open-circuit voltage value corresponding to each state of charge in the current open-circuit voltage-state of charge characteristic table is linearly offset to obtain the latest open-circuit voltage-state of charge characteristic table. If all deviations do not show a linear trend, then only replace the open-circuit voltage value in the current open-circuit voltage-state-of-charge characteristic table that exceeds the preset deviation value to obtain the latest open-circuit voltage-state-of-charge characteristic table.

7. The method for dynamic optimization of full-charge capacity of lithium batteries under unbalanced conditions according to claim 1, characterized in that, Select segmented thresholds from all candidate threshold points, including: Using each candidate threshold point as the dividing criterion, the state of charge of the smallest capacity cell in all full-charge capacity correction instances is divided into one group of state of charge not exceeding the corresponding candidate threshold point and another group of state of charge exceeding the corresponding candidate threshold point. The sum of the variances of the two groups of state of charge at each candidate threshold point is taken as the inferiority of the corresponding candidate threshold point. The candidate threshold with the lowest inferiority among all candidate threshold points is used as the segment threshold.

8. The method for dynamic optimization of full-charge capacity of lithium batteries under unbalanced conditions according to claim 1, characterized in that, Gain coefficients are dynamically fitted based on segmented thresholds and a large number of fully charged capacity correction examples with uniform capacity retention, including: Based on the capacity retention rate range of each aging level, a large number of full-charge capacity correction instances with uniform capacity retention rate distribution are divided into correction instance groups under each aging level. Calculate the theoretical deviation for each full-charge capacity correction instance without introducing a gain factor; In each aging level correction group, all correction instances in which the state of charge of the smallest capacity cell does not exceed the corresponding segment threshold are taken as the first correction instance set, and all correction instances in which the state of charge of the smallest capacity cell exceeds the corresponding segment threshold are taken as the second correction instance set. Based on the relationship between the gain coefficient and the base coefficient and the aging effect coefficient when the state of charge of the minimum capacity cell does not exceed the corresponding segment threshold, and the theoretical deviation of all full-charge capacity correction instances in the first correction instance set under each aging level without introducing the gain coefficient, the relationship between the gain coefficient and the capacity retention rate when the state of charge of the minimum capacity cell does not exceed the corresponding segment threshold under each aging level is fitted. Based on the relationship between the gain coefficient and the base coefficient and the aging effect coefficient when the state of charge of the minimum capacity cell does not exceed the corresponding segment threshold, and the theoretical deviation of all full-charge capacity correction instances in the second correction instance set under each aging level without introducing the gain coefficient, the relationship between the gain coefficient and the capacity retention rate when the state of charge of the minimum capacity cell exceeds the corresponding segment threshold under each aging level is fitted. Based on the capacity retention rate of the smallest capacity cell in the current battery pack and the relationship expression between all gain coefficients and capacity retention rate, all gain coefficients in the piecewise correction formula are determined.

9. The method for dynamic optimization of full-charge capacity of lithium batteries under unbalanced conditions according to claim 8, characterized in that, Based on the capacity retention rate of the smallest capacity cell in the current battery pack and the relationship expression between all gain coefficients and capacity retention rate, all gain coefficients in the piecewise correction formula are determined, including: Substitute the capacity retention rate of the smallest capacity cell in the current battery pack into the relationship expression between the gain coefficient and the capacity retention rate when the state of charge of the smallest capacity cell does not exceed the corresponding segment threshold, and the relationship expression between the gain coefficient and the capacity retention rate when the state of charge of the smallest capacity cell exceeds the corresponding segment threshold, respectively, to obtain the gain coefficient when the state of charge of the smallest capacity cell does not exceed the corresponding segment threshold and the gain coefficient when the state of charge of the smallest capacity cell exceeds the corresponding segment threshold.

Images