Intelligent charging method for storage battery pack

Abstract

The invention provides an intelligent charging method for a storage battery pack, and the method comprises the steps: obtaining the real-time voltage and current data of each single battery in the battery pack, carrying out the segmented sampling processing of a collected time sequence signal, analyzing the voltage recovery curve shape and the internal resistance change rate, and obtaining the capacity low feature distribution; according to the low-capacity characteristic distribution, applying current stimulation with a preset pulse frequency range and pulse duration to the low-capacity single battery, monitoring capacity amplification and a recovery time window after charging, and determining a capacity recovery rate; according to an attenuation type classification result, configuring high-frequency activation configuration for the single batteries classified as an active substance dormant state, and configuring low-energy protection configuration for the single batteries classified as a permanent loss state to obtain differentiated energy management configuration; and according to the balanced energy distribution path, obtaining overall voltage fluctuation data of the battery pack, dynamically correcting the energy input intensity and the distribution proportion, judging whether the battery pack reaches a balanced state, and obtaining optimized energy management configuration.

Description

Technical Field

[0001] This invention relates to the field of information technology, and in particular to a smart charging method for a battery pack. Background Technology

[0002] During the long-term cyclic use of battery packs, especially in practical applications such as electric vehicles or energy storage systems, it is frequently observed that the capacity of some individual cells is significantly lower than the average level. This low capacity signal has become a common phenomenon. However, the degradation mechanisms hidden behind this signal are complex and diverse, mainly divided into two types: reversible dormancy of active materials and irreversible permanent loss of active materials. Reversible dormancy often stems from the temporary loss of electrochemical activity of active materials during battery usage intervals or under specific operating conditions, but it is possible to reawaken and restore capacity contribution through appropriate stimulation. Irreversible loss, on the other hand, involves material structure damage, permanent depletion of lithium reserves, or accumulation of by-reaction products, leading to the complete loss of function of active materials. The two types of degradation are highly similar in appearance, such as both exhibiting low open-circuit voltage or insufficient discharge capacity, making them difficult to distinguish directly through conventional voltage or current detection. This poses a significant challenge to practical energy management. If a cell with irreversible permanent loss is mistakenly identified as being in a reversible dormant state, applying stimulation measures during charging may cause a sharp increase in the cell's internal resistance and an abnormal rise in local temperature, further accelerating material degradation and even shortening the overall lifespan of the entire battery pack. Conversely, if a reversibly dormant cell is mistakenly identified as permanently damaged, and a conservative low-energy protection mode is directly adopted, while avoiding risk, potential recoverable capacity will be permanently forfeited, leading to a premature decline in the overall usable capacity of the battery pack and a significant reduction in energy utilization efficiency. In actual operation, such misjudgments in electric vehicle battery packs after long-distance driving directly affect driving range and system economy. Furthermore, even if attempts are made to use the capacity recovery magnitude after specific charging to help distinguish the type of degradation, the reliability of this method is difficult to guarantee. This is because the selection of parameters for the recovery stimulus, such as frequency or duration, and the timing window for subsequent capacity recovery observation, significantly affect the strength and stability of the recovery signal. Inappropriate parameter combinations may produce excessively strong recovery performance on truly dormant cells or weak false signals on damaged cells, thus rendering the preset distinction threshold ineffective. In actual charge-discharge cycles of the battery pack, this parameter sensitivity leads to frequent switching of energy allocation strategies between different modes, either resulting in excessive stimulation leading to heat accumulation risks or excessive conservatism causing capacity waste, ultimately making it difficult to establish a stable and consistent discrimination criterion. In summary, the inherent complexity and external similarities of degradation types make initial identification extremely difficult, while the extreme dependence of auxiliary discrimination methods on parameters further amplifies their unreliability. These two factors reinforce each other, creating a technical contradiction that current charging management systems struggle to overcome. This contradiction directly restricts the actual performance and lifespan potential of battery packs under long-term cycling, becoming a pressing practical problem that needs to be solved in the context of the widespread application of power batteries. Summary of the Invention

[0003] This invention provides a smart charging method for a battery pack, mainly comprising:

[0004] Real-time voltage and current data of each individual cell in the battery pack are acquired. The acquired time-series signal is processed by segmented sampling. The shape of the voltage recovery curve and the rate of change of internal resistance are analyzed to obtain the low capacity characteristic distribution.

[0005] Based on the low capacity characteristic distribution, a current stimulation with a preset pulse frequency range and pulse duration is applied to the low capacity single cell, and the capacity increase and recovery time window after charging are monitored to determine the capacity recovery rate.

[0006] The degradation type of a single cell is classified by comparing the capacity recovery rate with a preset recovery rate threshold. If the capacity recovery rate is higher than the preset recovery rate threshold, it is classified as a dormant state of active material; otherwise, it is classified as a permanent loss state, thus obtaining the degradation type classification result.

[0007] Based on the classification results of the degradation type, high-frequency activation configuration is configured for single cells classified as active material dormancy state, and low-energy protection configuration is configured for single cells classified as permanent loss state, thus obtaining differentiated energy management configurations.

[0008] By configuring differentiated energy management, energy is allocated to the battery pack. If a single cell is classified as a dormant state of active material, the frequency of activation energy input is increased. If a single cell is classified as a permanently depleted state, the energy allocation ratio is reduced and the current is limited to determine a balanced energy allocation path.

[0009] Based on the balanced energy distribution path, obtain the overall voltage fluctuation data of the battery pack, dynamically correct the energy input intensity and distribution ratio, determine whether the battery pack has reached a balanced state, and obtain the optimized energy management configuration.

[0010] By optimizing the energy management configuration to perform battery pack energy distribution operations, and monitoring the capacity increase after charging, if the capacity increase exceeds the preset range, the shape of the voltage recovery curve and the rate of change of internal resistance are analyzed back to form a closed loop for identifying the attenuation type.

[0011] Furthermore, the process of acquiring real-time voltage and current data for each individual cell within the battery pack, performing segmented sampling processing on the acquired time-series signals, analyzing the shape of the voltage recovery curve and the rate of change of internal resistance, and obtaining the low-capacity characteristic distribution includes:

[0012] The system acquires real-time voltage and current data of each individual cell in the battery pack, collects time-series signals according to a preset sampling frequency, and segments the collected time-series signals using a sliding window method to obtain a segmented dataset containing voltage and current sequences.

[0013] For the voltage and current sequences in the segmented dataset, the voltage change value and current change value are extracted using the differential calculation method. The internal resistance value of each time period is calculated by dividing the voltage difference by the current difference. The voltage recovery curve is fitted using the least squares method, and the curve slope and curvature feature parameters are extracted. If the rate of change of internal resistance exceeds the preset rate of change threshold, the single cell is marked as a capacity abnormal single cell, and a capacity abnormality mark set is obtained.

[0014] Based on the slope and curvature characteristic parameters of each individual in the capacity anomaly label set, the voltage recovery curve is classified by shape using the K-means clustering method. The degree of feature difference between categories is judged by calculating the Euclidean distance between the center points of each category. The distribution density of low-capacity individuals is judged by calculating the ratio of the number of individuals in each category to the corresponding feature space volume, thus forming a low-capacity feature distribution containing category labels and density values.

[0015] Furthermore, the step of applying a preset pulse frequency range and pulse duration current stimulation to low-capacity individual cells based on the low-capacity characteristic distribution, monitoring the capacity increase and recovery time window after charging, and determining the capacity recovery rate includes:

[0016] Based on the category label and density value in the low capacity feature distribution, pulse current stimulation is applied to the single cell that is classified as having abnormal capacity. The pulse frequency starts from the lower limit of the preset range and increases. After each increase, the pulse frequency is applied for a preset number of cycles. The voltage response curve at each frequency is recorded. The frequency domain features of the voltage response are extracted by Fourier transform, and the frequency corresponding to the maximum response amplitude is determined as the optimal activation frequency.

[0017] Using the optimal activation frequency, the pulse duration is set to a preset ratio of the charge-discharge cycle. A square wave pulse current is applied, the open-circuit voltage change during the pulse interval is monitored, and the time required for polarization elimination is recorded.

[0018] After polarization elimination is completed, a constant current charging test is performed, the capacity accumulation curve during the charging process is recorded, and the charging capacity value is calculated.

[0019] The capacity increase percentage is calculated by comparing the current charging capacity value with the initial capacity measured before applying the pulse stimulation. At the same time, the time interval from the start of the pulse stimulation to the capacity stabilization is recorded as the recovery time window. The single recovery rate is obtained by dividing the capacity increase percentage by the recovery time window. The capacity recovery rate is determined by taking the average value after repeating the test a preset number of times.

[0020] Furthermore, the degradation type of a single battery cell is classified by comparing its capacity recovery rate with a preset recovery rate threshold. If the capacity recovery rate is higher than the preset recovery rate threshold, it is classified as a dormant state of active materials; otherwise, it is classified as a permanent loss state. The degradation type classification results include:

[0021] The capacity recovery rate is compared with a preset recovery rate threshold. If the capacity recovery rate is higher than the preset recovery rate threshold, the single cell is marked as a dormant active material state; otherwise, it is marked as a permanent loss state.

[0022] Based on the markers, for dormant cells, their capacity recovery rate and recovery time window are recorded; for permanently damaged cells, their current remaining capacity percentage and internal resistance change trend are recorded, forming a decay type classification result that includes state category and characteristic parameters.

[0023] Furthermore, based on the degradation type classification results, a high-frequency activation configuration is configured for individual cells classified as having a dormant active material state, and a low-energy protection configuration is configured for individual cells classified as having a permanent loss state, resulting in differentiated energy management configurations, including:

[0024] Based on the attenuation type classification results, a high-frequency activation configuration is set for individual cells classified as active material dormant state, and the pulse activation frequency is set to the highest value within the preset range. A low-energy protection configuration is set for individual cells classified as permanent loss state, the charging current is limited to a preset proportion of the rated current, and the charging cut-off voltage is reduced by a preset amount to obtain the initial management parameter set.

[0025] The initial management parameter set is used to assign a higher charging current limit and a higher temperature allowable range to dormant cells than to permanently damaged cells, and to assign a lower charging current limit and a lower temperature protection threshold to permanently damaged cells than to dormant cells, thus forming a differentiated energy management configuration that includes charging parameters and protection thresholds.

[0026] Furthermore, the differentiated energy management configuration implements energy allocation for the battery pack. If a single cell is classified as being in a dormant state of active materials, the activation energy input frequency is increased; if a single cell is classified as being in a state of permanent loss, the energy allocation ratio is reduced and the current is limited, thus determining a balanced energy allocation path, including:

[0027] Through the differentiated energy management configuration, the degradation type identifier of each individual battery is read, the activation energy input frequency of the individual battery in the active material dormant state is set to the upper limit of the preset frequency range, the charging power weight is allocated to the individual battery in the permanent loss state, the energy allocation ratio of the individual battery in the permanent loss state is set to the preset percentage of the nominal value, and the charging current is limited to not exceed the safety threshold.

[0028] The charging control of each cell in the battery pack is carried out by polling. Cells in dormant state receive a longer charging period in each charging cycle, while cells in loss state receive a shorter charging period. The input power of each cell is dynamically adjusted by a power divider.

[0029] Calculate the difference between the current state of charge and the target state of charge of each cell. If the difference exceeds the preset deviation threshold, adjust the charging period length and power allocation weight of the corresponding cell. Through iterative adjustment, make the state of charge of each cell more consistent, establish the energy transmission mapping relationship from the charging source to each cell, and determine the balanced energy allocation path.

[0030] Furthermore, the step of obtaining overall battery pack voltage fluctuation data based on the balanced energy distribution path, dynamically correcting the energy input intensity and distribution ratio, determining whether the battery pack has reached a balanced state, and obtaining an optimized energy management configuration includes:

[0031] Based on the balanced energy distribution path, the real-time voltage value of each cell in the battery pack is obtained, the voltage difference between adjacent cells and the standard deviation of the entire pack voltage are calculated, and the change sequence of the standard deviation over time is recorded.

[0032] When the voltage standard deviation exceeds the preset threshold, the energy input intensity of the low-voltage cells is increased proportionally and the energy allocation ratio of the high-voltage cells is reduced according to the degree to which the voltage of each cell deviates from the average value, so as to obtain the corrected energy allocation parameters.

[0033] Charging control is performed using the corrected energy distribution parameters, and the changing trend of the voltage standard deviation is continuously monitored. If the voltage standard deviation remains within the allowable range for a consecutive preset sampling period and the difference between adjacent sampling points is less than a preset threshold, the battery pack is determined to have reached a balanced state. The current energy input intensity and distribution ratio are recorded as optimized energy management configuration.

[0034] Furthermore, the step of optimizing energy management configuration to perform battery pack energy distribution operations, monitoring the capacity increase after charging, and if the capacity increase exceeds a preset range, retrospectively analyzing the shape of the voltage recovery curve and the rate of change of internal resistance to form a closed loop for attenuation type identification, includes:

[0035] The battery pack energy distribution operation is performed through the optimized energy management configuration, and the charging current, voltage timing data and cumulative capacity of each cell are recorded in real time. After charging is completed, the capacity increase percentage of each cell is calculated.

[0036] If the capacity increase exceeds the preset range, the voltage time series data of the corresponding unit is extracted from the recorded data, the voltage recovery curve is reconstructed using the curve fitting method, and the slope change characteristics and internal resistance change rate sequence of the curve are calculated.

[0037] If the characteristic parameter baseline value recorded when the cell was initially classified as attenuation type is compared with the baseline value, and the deviation from the baseline value exceeds the threshold, the attenuation type identification of the cell is corrected, and its charging current limit and activation frequency parameter are adjusted accordingly to form a closed loop for attenuation type identification.

[0038] The technical solutions provided by the embodiments of the present invention may include the following beneficial effects:

[0039] This invention discloses an intelligent charging method for battery packs. Addressing the overall performance degradation caused by differences in the capacity decay of individual cells within a battery pack, the method collects real-time voltage and current data of individual cells, analyzes voltage recovery curves and internal resistance change rates, accurately identifies low-capacity characteristic distributions, and applies pulse stimulation to low-capacity cells to assess their recovery rate. The degradation type is then classified as either dormant or permanently degraded. Based on the classification results, this invention configures differentiated energy management strategies. For dormant cells, high-frequency activation energy input is increased; for permanently degraded cells, the allocation ratio is reduced and current is limited, achieving a balanced energy distribution path. Simultaneously, the input intensity is dynamically adjusted to ensure the battery pack reaches a balanced state, ultimately forming a closed loop for degradation type identification and energy management. Through precise classification and dynamic adjustment, this invention significantly improves the overall capacity utilization and lifespan of the battery pack and optimizes energy distribution efficiency. Attached Figure Description

[0040] Figure 1 This is a flowchart of a smart charging method for a battery pack according to the present invention. Detailed Implementation

[0041] The technical solutions of the present invention will be clearly and thoroughly described below with reference to the accompanying drawings. The described embodiments are merely some embodiments of the present invention.

[0042] like Figure 1 This embodiment of a smart charging method for a battery pack may specifically include:

[0043] S101. Acquire real-time voltage and current data of each individual cell in the battery pack, perform segmented sampling processing on the acquired timing signal, analyze the shape of the voltage recovery curve and the rate of change of internal resistance, and obtain the low capacity characteristic distribution.

[0044] Real-time voltage and current data of each individual cell in the battery pack are acquired. Time-series signals are collected at a preset sampling frequency. The acquired time-series signals are segmented using a sliding window method, with each segment lasting one-tenth of a charge-discharge cycle, resulting in a segmented dataset containing voltage and current sequences. For the voltage and current sequences in the segmented dataset, a differential calculation method is used to extract the voltage change values ​​between adjacent time points. The internal resistance value for each time period is calculated by dividing the voltage difference by the current difference. The voltage recovery curve is fitted using the least squares method, and the slope and curvature feature parameters of the curve are extracted. If the rate of change of internal resistance exceeds a preset rate of change threshold, the cell is marked as a capacity abnormality cell, resulting in a capacity abnormality label set. Based on the slope and curvature feature parameters of each cell in the capacity abnormality label set, the voltage recovery curve is classified by shape using the K-means clustering method. The distribution density of capacity abnormality cells is determined by calculating the average variance D of the points within each category, where D is the sum of the squared distances from all points in each category to the center point divided by the number of points, forming a capacity abnormality feature distribution containing category labels and density values.

[0045] In one implementation, when acquiring real-time voltage and current data for each individual cell within the battery pack, the acquisition system measures the terminal voltage and charge / discharge current of each cell using a high-precision voltage sensor and a Hall effect current sensor, respectively. The sampling frequency is set to 10 times per second to ensure the capture of the battery's dynamic characteristics during charge / discharge transitions. The sliding window method uses a window width set to one-tenth of the charge / discharge cycle, with a window step size half the window width, achieving partial overlap between data segments and avoiding the loss of boundary information.

[0046] Specifically, the voltage and current sequences in the segmented dataset undergo time alignment to ensure a strict correspondence between voltage and current values ​​at the same moment. The differential calculation method employs a forward differential approach to calculate the voltage change ΔU and current change ΔI between adjacent sampling points. The internal resistance value R is obtained by dividing ΔU by ΔI. This dynamic internal resistance reflects the battery's impedance characteristics under specific operating conditions and better reflects the battery's true state than static internal resistance.

[0047] It should be noted that in the process of fitting the voltage recovery curve using the least squares method, a quadratic polynomial is chosen as the fitting function. The polynomial coefficients are determined by minimizing the sum of squared errors between the measured voltage value and the fitted value. The slope of the curve is obtained by taking the first derivative of the fitting function, while the curvature is calculated using the second derivative. When the rate of change of internal resistance exceeds a preset threshold, it indicates a significant change in the electrochemical characteristics of the single cell, requiring close monitoring of its capacity decay. This identification method based on dynamic changes in internal resistance can detect abnormal cells in the early stages of battery performance degradation.

[0048] In one possible implementation, the K-means clustering method categorizes anomalous cells based on their voltage recovery characteristics into three classes: fast recovery, slow recovery, and no recovery. The clustering process first randomly selects three initial cluster centers. Then, based on the slope and curvature characteristics of each cell, it calculates the distance from each cluster center and assigns the cell to the nearest cluster. The cluster center positions are iteratively updated until convergence.

[0049] Preferably, the calculation of Euclidean distance comprehensively considers two dimensions: slope difference and curvature difference. The distance value reflects the degree of feature difference between different categories. The low-capacity feature distribution includes the number of individuals in each category, the coordinates of the category center, and the degree of dispersion within each category. This information provides a quantitative basis for the subsequent formulation of differentiated charging strategies.

[0050] S102. Based on the low capacity characteristic distribution, apply current stimulation with a preset pulse frequency range and pulse duration to the low capacity single cell, monitor the capacity increase and recovery time window after charging, and determine the capacity recovery rate.

[0051] Based on the category label and density value in the low capacity characteristic distribution, pulse current stimulation is applied to individual batteries classified as having abnormal capacity. The pulse frequency starts from the lower limit of a preset range and increases incrementally. After each increment, the pulse is applied for a preset number of cycles. The voltage response curve at each frequency is recorded. The frequency domain features of the voltage response are extracted using Fourier transform. The frequency corresponding to the maximum response amplitude is taken as the optimal activation frequency, thus obtaining the frequency domain response features. Using the optimal activation frequency from the frequency domain response features, the pulse duration is set to a preset proportion of the charge-discharge cycle. A square wave pulse current is applied, and the pulse amplitude is set to a preset multiple of the rated charging current. The open-circuit voltage change during the pulse interval is monitored. If the open-circuit voltage rises above a preset threshold during the interval, polarization elimination is determined to be complete, and the time required for polarization elimination is recorded. Based on the time required for polarization elimination, a constant current charging test is performed after polarization elimination is completed. The charging current is set to a preset proportion of the rated value, and the capacity accumulation curve during the charging process is recorded. Charging is stopped when the voltage reaches the cutoff voltage, and the capacity value of this charge is calculated to obtain the activated capacity data. The percentage increase in capacity is calculated by comparing the activated capacity data with the initial capacity measured before applying the pulse stimulation. At the same time, the time interval from the start of the pulse stimulation to the stability of the capacity is recorded as the recovery time window. The single recovery rate is obtained by dividing the capacity increase by the recovery time window. The average value is taken after repeating the test a preset number of times to determine the capacity recovery rate.

[0052] In one implementation, when applying pulsed current stimulation to a single cell with abnormal capacity based on the category label and density value in the low-capacity characteristic distribution, the pulse frequency is increased in a logarithmic step manner. The initial frequency is set to 0.1Hz, and each increment increases by a factor of 1.5 until an upper limit of 100Hz is reached. At each frequency point, the number of cycles for which the pulsed current is continuously applied is dynamically adjusted according to the frequency: 5 cycles for the low-frequency range, 10 cycles for the mid-frequency range, and 20 cycles for the high-frequency range, to ensure a stable voltage response characteristic.

[0053] Specifically, the voltage response curve is recorded using a high-speed data acquisition card, with a sampling rate set to more than 100 times the pulse frequency to ensure the capture of detailed changes in the voltage response. The Fourier transform employs a Fast Fourier Transform (FFT) algorithm to convert the time-domain voltage response signal to the frequency domain. In the frequency domain, the amplitude and phase information of the fundamental frequency component, as well as the amplitudes of each harmonic component, are extracted. When the fundamental frequency response amplitude reaches its maximum value, the corresponding activation frequency is the optimal activation frequency for that single cell. This frequency domain analysis method can identify the sensitivity of the internal electrochemical reactions of the battery to different frequency stimuli; the optimal activation frequency typically corresponds to the characteristic frequencies of ion migration and charge transfer processes within the battery.

[0054] It should be noted that the pulse duration setting needs to comprehensively consider both activation effect and the risk of heat accumulation. If the pulse duration is too short, the active material will not be fully activated; if the duration is too long, it will cause the internal battery temperature to rise, accelerating side reactions. Extensive experimental verification has shown that a pulse duration of 5% to 10% of the charge-discharge cycle is reasonable. The square wave pulse has a fixed duty cycle of 50%, meaning the pulse application time is equal to the interval time. This symmetrical pulse helps reduce polarization accumulation inside the battery.

[0055] Preferably, polarization elimination is determined based on the dynamic changes in open-circuit voltage. During pulse intervals, concentration polarization and electrochemical polarization within the battery gradually dissipate, manifested as a slow rise in open-circuit voltage. By monitoring the rate of change of open-circuit voltage in real time, polarization is considered essentially eliminated when the rate of change falls below a set threshold. The time required for polarization elimination typically ranges from a few seconds to tens of seconds, and is closely related to the battery's internal structure and electrolyte characteristics. The polarization elimination time after each pulse is recorded to form time-series data for evaluating the battery's dynamic response characteristics.

[0056] In one possible implementation, the constant current charging test is initiated immediately after polarization elimination to prevent further changes in the battery state. The charging current is selected to be 30% to 50% of the rated charging current, which effectively activates dormant active materials without causing excessive thermal stress. During charging, the battery terminal voltage and charged capacity are recorded every 10 seconds to form a capacity accumulation curve. When the terminal voltage reaches the charging cutoff voltage or the charging time exceeds a preset upper limit, charging is stopped and the charging capacity value is recorded. The activated capacity data includes multiple parameters such as charging capacity, charging time, and average charging voltage, comprehensively reflecting the battery's capacity recovery status.

[0057] For example, calculating the capacity recovery rate requires establishing a clear benchmark. Before applying pulse stimulation, the battery's initial capacity is obtained through standard charge-discharge testing and used as a comparison baseline. The capacity increase percentage is equal to the difference between the activated capacity and the initial capacity divided by the initial capacity, then multiplied by 100%. The recovery time window is timed from the first application of pulse stimulation until the deviation of three consecutive charge capacity test results is less than 2%. This time window reflects the time required for the battery to reach a steady state from the start of activation.

[0058] Understandably, the single-cycle recovery rate is obtained by dividing the capacity increase by the recovery time window, expressed as a percentage per hour. To improve measurement accuracy, the entire test process is repeated 3 to 5 times, with sufficient rest time between each test to allow the battery to recover to a stable state. If the deviation of the recovery rate from multiple tests is within a preset range, the arithmetic mean is taken as the final capacity recovery rate. Furthermore, the capacity recovery rate directly reflects the degree of activation of dormant active materials in a single cell. Batteries with a recovery rate higher than 15% indicate significant reversible capacity loss and are suitable for high-frequency activation measures; batteries with a recovery rate between 5% and 15% have a moderate degree of dormancy and require a gentle activation strategy; batteries with a recovery rate lower than 5% have essentially irreversible active material loss and should employ a protective charging strategy to prevent further degradation.

[0059] For example, in the practical application of electric vehicle power battery packs, the capacity recovery rate determined by the above method can provide the battery management system with accurate individual cell state information, realize differentiated charging control, and extend the overall service life of the battery pack.

[0060] S103. By comparing the capacity recovery rate with the preset recovery rate threshold, the degradation type of the single cell is classified. If the capacity recovery rate is higher than the preset recovery rate threshold, it is classified as the active material dormant state; otherwise, it is classified as the permanent loss state, and the degradation type classification result is obtained.

[0061] The capacity recovery rate is compared with a preset recovery rate threshold. If the capacity recovery rate is higher than the preset recovery rate threshold, the single cell is marked as being in a dormant state of active material. If the capacity recovery rate is lower than or equal to the preset recovery rate threshold, it is marked as being in a permanently degraded state, thus obtaining the single cell state label. Based on the single cell state label, for dormant cells, its capacity recovery rate value and recovery time window obtained from previous tests are recorded. For permanently degraded cells, its current remaining capacity percentage and the internal resistance change trend calculated from historical data are recorded, forming a degradation type classification result that includes state category and characteristic parameters.

[0062] In one implementation, the comparison between the capacity recovery rate and a preset recovery rate threshold employs a segmented threshold determination method. The preset recovery rate threshold is set according to the battery type and operating conditions; the threshold for lithium iron phosphate batteries is typically set to 8%, and for ternary lithium batteries, it is set to 10%. When the capacity recovery rate is higher than the corresponding threshold, it indicates that there are reactivatable dormant active materials inside the battery, and this is marked as a dormant state of the active materials; conversely, if the rate is lower, it is considered that the active materials have suffered irreversible loss, and this is marked as a permanent loss state.

[0063] Specifically, for dormant cells, the key is the battery number, and the value includes the capacity recovery percentage, the recovery time window length, and the voltage plateau difference before and after activation. For permanently damaged cells, the following are recorded: the percentage of current remaining capacity relative to nominal capacity, the increase factor of internal resistance relative to initial internal resistance, and the decrease in charging cutoff voltage.

[0064] Preferably, the attenuation type classification result is represented by a status identifier code, with a value of 1 for dormant state and a value of 0 for permanent loss state, forming a binary status sequence, which facilitates rapid identification and processing by the battery management system.

[0065] S104. Based on the classification results of the decay type, configure a high-frequency activation configuration for single cells classified as active material dormancy state, and configure a low-energy protection configuration for single cells classified as permanent loss state, thus obtaining differentiated energy management configurations.

[0066] Based on the degradation type classification results, a high-frequency activation configuration is set for individual cells classified as active material dormant state, with the pulse activation frequency set to the highest value within a preset range and the charging current set to a preset multiple of the rated current. A low-energy protection configuration is set for individual cells classified as permanently damaged state, with the charging current limited to a preset percentage of the rated current and the charging cut-off voltage lowered by a preset amount, resulting in an initial management parameter set. Using this initial management parameter set, dormant state cells are assigned a higher upper limit to the charging current and a higher temperature allowable range than permanently damaged state cells; permanently damaged state cells are assigned a lower upper limit to the charging current and a lower temperature protection threshold than dormant state cells, forming a differentiated energy management configuration that includes charging parameters and protection thresholds.

[0067] In one implementation, when configuring differentiated energy management parameters based on the degradation type classification results, a high-frequency activation configuration is used for individual cells in a dormant active material state. The pulse activation frequency is set within the range of 10Hz to 50Hz, selecting a frequency value close to the upper limit. The charging current is set to 1.2 to 1.5 times the rated current, activating the dormant active material through high-frequency pulses and a larger current. The allowable temperature range is set to 0°C to 45°C, providing a wide operating temperature window.

[0068] Specifically, low-energy protection is implemented for cells in a permanently damaged state. The charging current is strictly limited to 0.3 to 0.5 times the rated current to prevent high current from accelerating material degradation. The charging cut-off voltage is lowered by 50mV to 100mV to reduce the stress of high voltage on damaged materials. The temperature protection threshold is tightened to 10°C to 35°C to prevent abnormal temperature from triggering side reactions.

[0069] Preferably, a differentiated charging control table can be established using an initial management parameter set and implemented with a lookup table structure, allowing for quick indexing of corresponding charging parameters based on the individual battery number. The control table includes parameters such as charging current upper limit, temperature protection threshold, charging cut-off voltage, and pulse frequency, enabling precise management of individual cells with different degradation types and extending the overall lifespan of the battery pack.

[0070] S105. Through differentiated energy management configuration, energy is allocated to the battery pack. If a single cell is classified as a dormant state of active material, the activation energy input frequency is increased. If a single cell is classified as a permanently damaged state, the energy allocation ratio is reduced and the current is limited to determine a balanced energy allocation path.

[0071] By configuring differentiated energy management, the degradation type identifier of each individual battery cell is read. For cells in a dormant state, the activation energy input frequency is set to the upper limit of a preset frequency range, and a higher charging power weight is assigned to cells in a permanently damaged state. For cells in a permanently damaged state, the energy allocation ratio is set to a preset percentage of the nominal value, and the charging current is limited to a safety threshold, resulting in an initial energy allocation table. Based on the initial energy allocation table, charging control of each cell in the battery pack is performed using a polling method. Dormant cells receive a longer charging period in each charging cycle, while damaged cells receive a shorter charging period. The input power of each cell is dynamically adjusted through a power divider, and the actual energy value received by each cell is recorded to obtain a real-time energy flow matrix. Using the real-time energy flow matrix, the difference between the current state of charge and the target state of charge of each cell is calculated. If the difference exceeds a preset deviation threshold, the charging period length and power allocation weight of the corresponding cell are adjusted. Through iterative adjustment, the state of charge of each cell tends to be consistent, obtaining dynamic equilibrium control parameters. The dynamic equalization control parameters establish an energy transmission mapping relationship from the charging source to each cell. Cells in dormant state adopt a high-current charging path, while cells in loss state adopt a current-limited charging path. Energy distribution is optimized based on charging efficiency to determine the balanced energy distribution path.

[0072] In one implementation, when allocating battery pack energy through differentiated energy management configuration, the degradation type identifier of each individual cell is first read from the memory of the battery management system. For cells identified as being in a dormant state of active materials, the activation energy input frequency is set to a higher value within the range of 20Hz to 50Hz, typically 40Hz. The charging power weight is set to 1.5, indicating that the cell receives 1.5 times the standard power in the power allocation. For cells in a permanently degraded state, the energy allocation ratio is limited to 30% to 50% of the nominal value, and the charging current is strictly controlled below 0.5C, where C represents the battery's rated capacity. The initial energy allocation table is stored using a two-dimensional array structure, with row indices corresponding to cell numbers and columns containing parameters such as charging power weight, current limit, and activation frequency.

[0073] Specifically, the polling charging control mechanism is based on the time-slice allocation principle. The entire charging cycle is divided into several basic time slices, each with a length of 100 milliseconds. Dormant cells are allocated 60% to 70% of the total time slices within a charging cycle, while depleted cells are allocated only 20% to 30%. Within each time slice, the power divider dynamically adjusts the output power according to a preset power weight. The power divider uses pulse-width modulation (PWM) control, achieving precise power control by adjusting the duty cycle. When a dormant cell is in a charging time slice, the power divider outputs a higher charging power; when switching to a depleted cell, it automatically reduces the output power and activates current-limiting protection.

[0074] It should be noted that the real-time energy flow matrix is ​​an N×M matrix structure, where N represents the number of individual cells in the battery pack, and M represents the types of monitored parameters. Each row of the matrix corresponds to an individual cell, and the columns record parameters such as input current, input voltage, cumulative charge capacity, instantaneous power, and energy conversion efficiency for that cell at the current moment. The matrix data is updated 10 times per second, acquiring the real-time electrical parameters of each cell through a high-speed data acquisition module. The energy conversion efficiency is calculated as the ratio of output energy to input energy, reflecting the energy loss during the charging process. The cumulative charge capacity is obtained by integrating the charging current over time and is used to evaluate the actual capacity recovery of each cell.

[0075] Preferably, the calculation of the state of charge (SOC) difference uses a composite method combining the ampere-hour integral method and the open-circuit voltage method. The target SOC is uniformly set to 80% to avoid overcharging and damage to the battery. When the SOC difference of a single cell exceeds a 5% deviation threshold, an adjustment mechanism is triggered. During the adjustment process, if the SOC of a single cell is lower than the target value, its charging period length is increased by one basic time slice each time; if it is higher than the target value, the charging period is reduced or charging is paused. The power allocation weight adjustment uses a proportional-integral (PI) control algorithm, calculating the adjustment amount based on the magnitude and rate of change of the deviation. The iterative adjustment process continues until the SOC deviations of all cells converge to the allowable range; the parameter set obtained at this point is the dynamic equalization control parameter.

[0076] In one possible implementation, the energy transfer mapping relationship is established considering the differences in charging characteristics of cells with different degradation types. Dormant cells, because their internal active materials are in an activated state, can withstand larger charging currents, thus employing a high-current charging path. This path is achieved by connecting multiple charging channels in parallel, each configured with an independent power switch and current detection circuit, with charging currents reaching 2C to 3C. For damaged cells, since the active materials have undergone irreversible loss, high-current charging would accelerate degradation; therefore, a current-limiting charging path is adopted, limiting the charging current to below 0.3C through a series current-limiting resistor or constant current source circuit.

[0077] For example, charging efficiency optimization is achieved by real-time monitoring of power loss in each charging path. The system calculates the difference between the input power and output power for each path to obtain the power loss value. When the power loss of a path exceeds a preset threshold, the system automatically adjusts the operating parameters of that path, such as reducing the charging current or changing the pulse frequency, to improve charging efficiency. Furthermore, the final determination of the balanced energy distribution path also considers the influence of temperature. Each individual battery cell is equipped with a temperature sensor to monitor the battery surface temperature in real time. When a cell's temperature is detected to rise too rapidly, the system automatically reduces the charging power of that cell to prevent thermal runaway.

[0078] For example, in the practical application of electric vehicle power battery packs, the above-mentioned balanced energy distribution path can achieve precise energy management of large battery packs containing hundreds of cells, effectively extending the cycle life of the battery pack and improving the driving range and reliability of the whole vehicle.

[0079] S106. Based on the balanced energy distribution path, obtain the overall voltage fluctuation data of the battery pack, dynamically correct the energy input intensity and distribution ratio, determine whether the battery pack has reached a balanced state, and obtain the optimized energy management configuration.

[0080] Based on the balanced energy distribution path, the real-time voltage values ​​of each cell in the battery pack are acquired through the voltage acquisition module. The voltage difference between adjacent cells and the standard deviation of the overall voltage are calculated, where the standard deviation reflects the dispersion of the voltage distribution. The change sequence of the standard deviation over time is recorded to obtain the battery pack voltage fluctuation characteristic data. Using the voltage fluctuation characteristic data, when the voltage standard deviation exceeds a preset threshold of 50 mV, the energy input intensity of low-voltage cells is increased proportionally and the energy distribution ratio of high-voltage cells is reduced according to the degree of deviation of each cell's voltage from the average value. Abrupt changes are eliminated by performing moving average processing on five consecutive adjustment values ​​to obtain the corrected energy distribution parameters. Charging control is performed using the corrected energy distribution parameters, and the changing trend of the voltage standard deviation is continuously monitored. If the voltage standard deviation remains within the allowable range of 30 mV for 10 consecutive sampling periods and the difference between adjacent sampling points is less than the preset threshold of 2 mV, the battery pack is determined to have reached a balanced state. The current energy input intensity and distribution ratio are recorded as a specific component of the optimized energy management configuration.

[0081] In one implementation, when acquiring overall battery pack voltage fluctuation data based on the balanced energy distribution path, the voltage acquisition module synchronously acquires the terminal voltage of all individual cells at a sampling frequency of 100Hz. The voltage difference between adjacent cells is obtained through pairwise comparisons, forming a difference sequence. The voltage standard deviation is calculated using statistical methods: first, the average voltage of all cells is calculated; then, the sum of squares of the deviations of each cell voltage from the average value is calculated, divided by the number of cells, and the square root is taken to obtain the standard deviation. This standard deviation directly reflects the dispersion of voltage distribution within the battery pack; a larger standard deviation indicates poorer voltage consistency.

[0082] Specifically, voltage fluctuation characteristic data comprises two parts: instantaneous standard deviation and historical standard deviation series. The instantaneous standard deviation reflects the voltage distribution at the current moment, while the historical standard deviation series records the trajectory of standard deviation changes over a past period. By analyzing the time series of standard deviations, the periodic characteristics and trends of voltage fluctuations can be identified. When the standard deviation exceeds a preset threshold of 50mV, it indicates a significant imbalance within the battery pack, requiring the activation of the energy distribution adjustment mechanism.

[0083] It should be noted that the correction of energy input intensity and allocation ratio adopts a differentiated adjustment method. For cells with voltage below the average value, the greater the deviation, the greater the increase in energy input intensity, with the increase being directly proportional to the deviation. For cells with voltage above the average value, their energy allocation ratio is reduced accordingly. Multi-point moving average processing is achieved by averaging five consecutive adjustment values. This smoothing process can eliminate sudden changes in adjustment values ​​caused by measurement noise or instantaneous disturbances, avoiding system oscillations. The corrected energy allocation parameters include the charging current setpoint, charging time allocation, and pulse frequency parameters for each cell.

[0084] Preferably, a dual-condition judgment mechanism is used to determine the battery pack's equilibrium state. The first condition is that the voltage standard deviation must remain within the allowable range of 30mV, and the second condition is that the change in standard deviation between adjacent sampling points is less than 2mV. Only when both conditions are met within 10 consecutive sampling periods is the battery pack considered to have reached an equilibrium state. The recorded energy input intensity and distribution ratio at this time constitute an optimized energy management configuration, which can achieve balanced management of the battery pack while ensuring charging efficiency.

[0085] For example, in large-scale battery pack applications in energy storage power stations, this dynamic optimization method can achieve coordinated charging control of hundreds of individual cells, keeping voltage inconsistencies within a small range and effectively improving the overall performance and lifespan of the energy storage system.

[0086] S107. By optimizing the energy management configuration, the battery pack energy distribution operation is executed, and the capacity increase after charging is monitored. If the capacity increase exceeds the preset range, the shape of the voltage recovery curve and the rate of change of internal resistance are analyzed back to form a closed loop for identifying the attenuation type.

[0087] By optimizing the energy management configuration and performing battery pack energy distribution operations, charging control is applied to each cell according to the configured charging parameters. The charging current, voltage timing data, and cumulative capacity of each cell are recorded in real time. After charging is complete, the capacity increase percentage of each cell is calculated to obtain capacity recovery monitoring data. Based on this capacity recovery monitoring data, if the capacity increase exceeds the upper or lower limit of a preset range, the voltage timing data of the corresponding cell is extracted from the recorded data. A curve fitting method is used to reconstruct the voltage recovery curve, calculating the slope change characteristics and curvature distribution of the curve. Simultaneously, the internal resistance change rate sequence is calculated using the ratio of voltage difference to current difference to obtain a set of attenuation characteristic parameters. Using this set of attenuation characteristic parameters, the baseline values ​​of the characteristic parameters recorded when the cell was initially classified as attenuating are compared. If the shape of the voltage recovery curve or the internal resistance change rate deviates from the baseline value by more than a threshold, the attenuation type identification of the cell is corrected, and its charging current limit and activation frequency parameters are adjusted accordingly, forming a closed loop for attenuation type identification that includes monitoring, judgment, backtracking, and correction.

[0088] In one implementation, when performing battery pack energy distribution operations through optimized energy management configuration, the charging control module calls the corresponding charging parameters based on the degradation type identifier of each individual battery cell. For dormant cells, a high-frequency pulse is applied to activate charging; for cells in a degraded state, current-limiting protection charging is employed. The data acquisition system records the current value, voltage value, and timestamp of each cell during the charging process at a sampling interval of 10 milliseconds, forming a complete charging process dataset. After charging is completed, the cumulative capacity is calculated by integrating the current over time and compared with the baseline capacity value before charging to obtain the capacity increase percentage.

[0089] Specifically, the normal range for capacity increase is set at 5% to 20%. When the capacity increase of a single cell is lower than 5% or higher than 20%, a backtracking analysis mechanism is triggered. The system extracts the voltage time-series data of the abnormal cell from the stored charging process dataset, performs cubic polynomial curve fitting using the least squares method, and reconstructs the complete voltage recovery curve. The slope variation of the curve is obtained by taking the first derivative of the fitted function, reflecting the rate of voltage rise; the curvature distribution is calculated using the second derivative, reflecting the degree of curvature of the curve. These characteristic parameters can reveal the dynamic characteristics of the electrochemical reactions inside the battery.

[0090] It should be noted that the calculation of the internal resistance change rate sequence is based on the differential form of Ohm's law. At each sampling point during the charging process, the voltage difference ΔV and current difference ΔI between adjacent points are calculated, and the internal resistance change rate is equal to ΔV divided by ΔI. The internal resistance change rates of all sampling points are arranged in chronological order to form the internal resistance change rate sequence. This sequence reflects the dynamic evolution characteristics of the battery's internal resistance during the charging process and is an important indicator for judging the battery's health status.

[0091] Preferably, the degradation type identification closed loop achieves adaptive correction through comparative analysis. The system saves the baseline values ​​of characteristic parameters for each individual cell at the time of initial classification, including the shape parameters of the standard voltage recovery curve and the range of internal resistance change rate. When the deviation of the new characteristic parameters obtained from the retrospective analysis from the baseline values ​​exceeds 30%, it indicates that the degradation characteristics of the cell have changed significantly. The system automatically corrects the degradation type identification of the cell, changing it from a dormant state to a loss state or vice versa, and adjusts the charging current limit, activation frequency, and other control parameters accordingly to achieve dynamic optimization management.

[0092] For example, in the application of electric vehicle fast charging stations, this closed-loop identification mechanism can detect changes in battery degradation characteristics in a timely manner, avoid overcharging or undercharging problems caused by misjudgment, and improve charging safety and battery life.

[0093] The specific embodiments described above further illustrate the purpose, technical solution, and beneficial effects of the present invention. It should be understood that the above description is only a specific embodiment of the present invention and is not intended to limit the scope of protection of the present invention. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of the present invention should be included within the scope of protection of the present invention.

Claims

1. A method of intelligent charging of a battery pack, characterized by, The method comprises: Obtaining real-time voltage and current data of each single battery in the battery pack, segmenting and sampling the collected time sequence signal, analyzing the voltage recovery curve shape and internal resistance change rate, and obtaining the capacity low characteristic distribution; According to the capacity low characteristic distribution, a current stimulus with a preset pulse frequency range and pulse duration is applied to the capacity low single battery, the capacity increase amplitude and recovery time window after charging are monitored, and the capacity recovery rate is determined; By comparing the capacity recovery rate with the preset recovery rate threshold, the single battery is classified by decay type, if the capacity recovery rate is higher than the preset recovery rate threshold, it is classified as active material dormancy state, otherwise it is classified as permanent loss state, and the decay type classification result is obtained; According to the decay type classification result, the single battery classified as active material dormancy state is configured with high-frequency activation configuration, and the single battery classified as permanent loss state is configured with low-energy protection configuration, and the differential energy management configuration is obtained; Through the differential energy management configuration, the energy distribution of the battery pack is implemented, if the single battery is classified as active material dormancy state, the activation energy input frequency is increased, if the single battery is classified as permanent loss state, the energy distribution proportion is reduced and the current is limited, and the equalization energy distribution path is determined; According to the equalization energy distribution path, the overall voltage fluctuation data of the battery pack is obtained, the energy input intensity and distribution proportion are dynamically corrected, whether the battery pack reaches the equalization state is judged, and the optimized energy management configuration is obtained; Through the optimized energy management configuration, the battery pack energy distribution operation is executed, the capacity increase amplitude after charging is monitored, if the capacity increase amplitude exceeds the preset range, the voltage recovery curve shape and internal resistance change rate are backtracked and analyzed, and the decay type identification closed loop is formed.

2. The battery pack intelligent charging method according to claim 1, wherein The method comprises: Obtaining real-time voltage and current data of each single battery in the battery pack, segmenting and sampling the collected time sequence signal, analyzing the voltage recovery curve shape and internal resistance change rate, and obtaining the capacity low characteristic distribution, comprising: Obtaining real-time voltage and current data of each single battery in the battery pack, collecting time sequence signals according to a preset sampling frequency, and segmenting the collected time sequence signals by a sliding window method to obtain segmented data sets containing voltage sequences and current sequences; For the voltage sequences and current sequences in the segmented data sets, the differential calculation method is used to extract voltage change values and current change values, the internal resistance values in each period are calculated by dividing the voltage difference value by the current difference value, the least square method is used to fit the voltage recovery curve, the curve slope and curvature characteristic parameters are extracted, and if the internal resistance change rate exceeds the preset change rate threshold, the single battery is marked as a capacity abnormal single battery, and a capacity abnormal marker set is obtained; According to the slope and curvature characteristic parameters of each single battery in the capacity abnormal marker set, the K-means clustering method is used to classify the shape of the voltage recovery curve, the Euclidean distance between the center points of each category is calculated to judge the feature difference degree between categories, and the ratio of the number of single batteries in each category to the corresponding feature space volume is calculated to judge the distribution density of the capacity low single battery, and a capacity low characteristic distribution containing category labels and density values is formed.

3. The battery pack intelligent charging method according to claim 1, wherein The method comprises the following steps: According to the category label and density value in the capacity low characteristic distribution, the single battery classified as capacity anomaly is subjected to pulse current stimulation, the pulse frequency is increased from the lower limit value of the preset range, and the voltage response curve at each frequency is recorded. The frequency corresponding to the maximum response amplitude is determined as the optimal activation frequency by Fourier transform to extract the frequency domain features of the voltage response; The optimal activation frequency is used to set the pulse duration to a preset proportion of the charge and discharge cycle, and a square wave pulse current is applied. The open circuit voltage change during the pulse interval is monitored, and the time required for polarization elimination is recorded; After polarization elimination, a constant current charging test is performed, and the capacity accumulation curve during charging is recorded. The charging capacity value is calculated; The initial capacity measured before applying the pulse stimulation is compared with the charging capacity value to calculate the capacity increase percentage. The time interval from the start of pulse stimulation to the stable capacity is recorded as the recovery time window. The single recovery rate is obtained by dividing the capacity increase percentage by the recovery time window. After repeating the test for a preset number of times, the average value is determined as the capacity recovery rate.

4. The battery pack intelligent charging method according to claim 1, wherein The single battery is classified according to the capacity recovery rate and the preset recovery rate threshold. If the capacity recovery rate is higher than the preset recovery rate threshold, it is classified as active material dormancy state, otherwise it is classified as permanent loss state, and the attenuation type classification result is obtained, which comprises: The capacity recovery rate is compared with the preset recovery rate threshold. If the capacity recovery rate is higher than the preset recovery rate threshold, the single battery is marked as active material dormancy state, otherwise it is marked as permanent loss state; According to the mark, the capacity recovery rate value and the recovery time window of the single battery in dormancy state are recorded, and the current remaining capacity percentage and the internal resistance change trend of the single battery in permanent loss state are recorded, forming the attenuation type classification result containing state category and characteristic parameters.

5. The battery pack intelligent charging method according to claim 1, wherein According to the attenuation type classification result, the single battery classified as active material dormancy state is configured with high-frequency activation configuration, and the single battery classified as permanent loss state is configured with low-energy protection configuration, obtaining the differentiated energy management configuration, which comprises: According to the attenuation type classification result, the single battery classified as active material dormancy state is configured with high-frequency activation configuration, and the single battery classified as permanent loss state is configured with low-energy protection configuration, obtaining the differentiated energy management configuration, which comprises: The initial management parameter set is obtained. The initial management parameter set is used to assign a higher upper limit of charging current to the hibernation state monomer than to the permanent loss state monomer, and a larger temperature allowable range to the hibernation state monomer than to the permanent loss state monomer, and the permanent loss state monomer is assigned a lower upper limit of charging current than the hibernation state monomer and a lower temperature protection threshold than the hibernation state monomer, to form a differentiated energy management configuration containing charging parameters and protection thresholds.

6. The battery pack intelligent charging method according to claim 1, wherein The energy distribution of the battery pack is implemented through the differentiated energy management configuration, if the monomer battery is classified as an active substance hibernation state, the activation energy input frequency is increased, if the monomer battery is classified as a permanent loss state, the energy distribution proportion is reduced and the current is limited, and the equalization energy distribution path is determined, including: Through the differentiated energy management configuration, the attenuation type identification of each monomer battery is read, the activation energy input frequency of the monomer in the active substance hibernation state is set to the upper limit value of the preset frequency range, the charging power weight of the permanent loss state monomer is assigned to be higher than that of the permanent loss state monomer, and the energy distribution proportion of the monomer in the permanent loss state is set to be a preset percentage of the nominal value, and the charging current is limited to be not more than a safety threshold; The charging control of each monomer in the battery pack is performed in a polling manner, the hibernation state monomer obtains a longer charging period in each charging cycle, and the loss state monomer obtains a shorter charging period, and the input power of each monomer is dynamically adjusted through the power distributor; The difference between the current state of charge and the target state of charge of each monomer is calculated, if the difference exceeds a preset deviation threshold, the charging period length and power distribution weight of the corresponding monomer are adjusted, the state of charge of each monomer is made consistent through iterative adjustment, the energy transmission mapping relationship from the charging source to each monomer is established, and the equalization energy distribution path is determined.

7. The battery pack intelligent charging method according to claim 1, wherein According to the equalization energy distribution path, the overall voltage fluctuation data of the battery pack is obtained, the energy input intensity and distribution proportion are dynamically corrected, it is judged whether the battery pack reaches the equalization state, and the optimized energy management configuration is obtained, including: According to the equalization energy distribution path, the real-time voltage values of each monomer in the battery pack are obtained, the voltage difference between adjacent monomers and the standard deviation of the overall voltage are calculated, and the change sequence of the standard deviation with time is recorded; When the voltage standard deviation exceeds a preset threshold, according to the degree of voltage deviation of each monomer from the average value, the energy input intensity of the low-voltage monomer is increased in proportion and the energy distribution proportion of the high-voltage monomer is reduced, and the corrected energy distribution parameter is obtained; Through the corrected energy distribution parameter, charging control is performed, the change trend of the voltage standard deviation is continuously monitored, if the voltage standard deviation remains within the allowable range and the difference between adjacent sampling points is less than a preset threshold for a continuous preset number of sampling periods, it is determined that the battery pack reaches the equalization state, and the current energy input intensity and distribution proportion are recorded as the optimized energy management configuration.

8. The battery pack intelligent charging method of claim 1, wherein, The battery pack energy distribution operation is performed through the optimized energy management configuration, the capacity increment after charging is monitored, if the capacity increment exceeds a preset range, the voltage recovery curve shape and internal resistance change rate are analyzed, and a decay type identification closed loop is formed, including: The battery pack energy distribution operation is performed through the optimized energy management configuration, the charging current, voltage time sequence data and accumulated capacity of each single body are recorded in real time, and the capacity increase percentage of each single body is calculated after the charging is completed; If the capacity increase exceeds the preset range, the voltage time sequence data of the corresponding single body is extracted from the recorded data, the voltage recovery curve is reconstructed by using the curve fitting method, and the slope change characteristics and the internal resistance change rate sequence of the curve are calculated; If the deviation from the reference value exceeds the threshold value, the attenuation type identification of the single body is corrected, and the charging current limit value and the activation frequency parameters are adjusted accordingly to form an attenuation type identification closed loop.

Images