A lithium battery peak charge-discharge test method and system based on dynamic dichotomy

Abstract

The application discloses a lithium battery peak charging and discharging test method and system based on dynamic dichotomy, which comprises the following steps: initialization configuration before test, completing system self-checking on the lithium battery to be tested and recording initial reference data; obtaining the initial reference data through dynamic dichotomy iterative adjustment, iteratively adjusting the lithium battery charging and discharging current or power, calling a double-factor adaptive step correction algorithm in each iteration to correct the next iteration step, and completing each round of charging and discharging parameter adjustment and action execution. Advantageous effects: the method avoids the problems of low efficiency when far away from the peak value and easy overshoot when approaching the peak value of the traditional fixed step iteration, thereby guaranteeing the convergence speed and significantly improving the precision of peak locking.

Description

Technical Field

[0001] This invention relates to the field of storage batteries, and in particular to a peak charge and discharge test method and system for lithium batteries based on the dynamic dichotomy method. Background Technology

[0002] As a core energy storage unit, the peak charge and discharge capacity of lithium batteries directly determines the performance boundaries and safety margins of power and energy storage systems under extreme operating conditions. Accurately and efficiently identifying the peak charge and discharge boundaries of lithium batteries is crucial for battery selection, system matching, and safety management.

[0003] Currently, testing technologies for lithium battery charge and discharge performance mainly focus on two aspects: First, comprehensive evaluation of charge and discharge quality, which involves collecting multi-dimensional data such as voltage, current, and temperature, and calculating deviation values ​​based on scenario weights to determine the battery's quality qualification. However, this approach does not involve accurately locking the peak charge and discharge boundaries. Second, ensuring communication stability during the testing process, which involves building test engineering, simulating bus interaction, and adding anomaly monitoring programs to solve problems such as message timeouts and data anomalies. The focus is on ensuring the smooth execution of the testing process, rather than exploring the battery's extreme performance.

[0004] Traditional peak power testing methods typically employ a fixed-step iterative approach to adjust charge and discharge parameters, relying on a single voltage or current metric to determine the peak value. This approach has significant drawbacks: firstly, the fixed step size cannot be dynamically adapted to the battery's real-time state, resulting in low iteration efficiency and large peak value locking deviations; secondly, the judgment based on a single electrical metric is too one-sided, lacking in-depth analysis of the battery's internal electrochemical mechanisms, such as the inability to quantify polarization states and lithium plating risk. Furthermore, traditional methods often require interrupting the charge and discharge process for impedance detection, reducing testing efficiency and failing to capture real-time changes in the battery's electrochemical characteristics during charge and discharge. Moreover, the lack of multi-dimensional safety constraint mechanisms poses safety hazards when approaching extreme limits. Summary of the Invention

[0005] Purpose of the invention: The purpose of this invention is to solve the technical problems in the prior art and provide a peak charge and discharge test method and system for lithium batteries based on the dynamic dichotomy method.

[0006] Technical solution: In the first aspect, this application proposes a peak charge-discharge test method for lithium batteries based on the dynamic dichotomy method, including the following steps:

[0007] Step S1, Pre-test initialization configuration: Complete the system self-test of the lithium battery under test and record the initial benchmark data;

[0008] Step S2: Obtain initial baseline data and adjust the lithium battery charging and discharging current or power iteratively through dynamic binary iterative adjustment. In each iteration, the dual-factor adaptive step size correction algorithm is called to correct the step size of the next iteration, thus completing the adjustment of charging and discharging parameters and the execution of actions in each round.

[0009] Step S3: After the charging and discharging parameters are stabilized in each round, perform a wideband online impedance scan of the lithium battery with a single scan duration of 8 and 10 ms, extract key impedance parameters, identify the real-time position of characteristic frequency points, decouple the relaxation time distribution of the impedance spectrum, and call the characteristic frequency drift polarization quantization algorithm to calculate the polarization state quantization value and the DRT lithium plating risk quantization algorithm to calculate the lithium plating risk quantization value respectively.

[0010] Step S4: Integrate the real-time battery voltage, impedance change rate, polarization state quantification value and lithium plating risk quantification value, call the quadruple constraint peak comprehensive judgment algorithm for comprehensive judgment, compare the judgment result with the preset stable range, and determine whether to terminate the current iteration, shrink the step size or continue the iteration.

[0011] Step S5: Repeat steps S2 to S4. When the judgment result is stable within the preset range for multiple consecutive test cycles and there are no sudden changes in the various state parameters of the battery, lock the peak charge and discharge boundary, output the relevant parameters and store them in association, and establish a database linking peak characteristics and internal electrochemical mechanisms.

[0012] Preferably, S1 includes:

[0013] Electrical connection and operating condition calibration, setting parameters such as charging and discharging cutoff voltage, various constraint thresholds, wideband impedance scanning frequency band, dynamic binary initial iteration step size and algorithm coefficients;

[0014] The charging and discharging cutoff voltage is set according to the nominal voltage of the lithium battery under test, with the upper limit charging voltage being 1.05 to 1.10 times the nominal voltage and the lower limit discharging voltage being 0.80 to 0.85 times the nominal voltage.

[0015] The various constraint thresholds specifically include: impedance change rate threshold of 3% / s to 8% / s, characteristic frequency drift rate threshold of 50Hz / s to 150Hz / s, and characteristic peak area change threshold of lithium plating-related relaxation time of 8% / s to 12% / s.

[0016] The wideband impedance scanning frequency band is set to 10Hz to 1kHz;

[0017] The initial iteration step size of the dynamic binary search is 10% to 20% of the rated current or power of the lithium battery under test, and the target iteration accuracy is 0.5% to 1% of the rated current or power.

[0018] The initial reference data includes the initial characteristic frequency and the initial DRT characteristic peak area, which are collected after the lithium battery under test has undergone three pre-charge-discharge cycles to a stable state, and the average value of the three collections is taken as the initial reference data.

[0019] Preferably, the dynamic binary iterative adjustment in step S2 includes:

[0020] A constant current charge-discharge mode is adopted, and the duration of each round of charge-discharge is set to 5 to 10 seconds;

[0021] The criterion for adaptive step size adjustment is: when the iteration interval shrinks to less than 1 / 10 of the initial iteration step size, the step size is adjusted to 1 / 5 to 1 / 10 of the initial step size;

[0022] Continue iterating when the iteration interval is greater than 1 / 10 of the initial iteration step size;

[0023] The battery surface temperature is monitored in real time during the iteration process. When the temperature exceeds 40°C, the initial step size is reduced by 30%.

[0024] Preferably, the mathematical expression for the two-factor adaptive step size correction algorithm in step S2 is:

[0025] ;

[0026] in, The step size is the binary search step size in the (n+1)th iteration. The step size for the nth iteration is the binary step size. This is the normalized value of the real-time impedance change rate. This is the normalized value of the characteristic frequency drift rate. , For safety weighting coefficients, and .

[0027] Preferably, the specific steps of broadband impedance scanning and decoupling in step S3 include:

[0028] After each round of charging and discharging parameters are adjusted and stabilized for 2 to 3 seconds, a wideband impedance scan is started. The duration of a single scan is controlled to be 8 to 10 ms. During the scan, the charging and discharging current or power remains unchanged and the charging and discharging process is not interrupted.

[0029] The characteristic frequency points specifically include: the frequency corresponding to the imaginary minimum point of the impedance spectrum, the phase angle, the frequency corresponding to the 45° point, and the characteristic peak frequency of the time constant of the charge transfer process;

[0030] The relaxation time distribution decoupling employs the non-negative least squares method, and the decoupling frequency range corresponds to a relaxation time of 10. 、3 Up to 10 2The total impedance is decoupled into four independent sub-processes: the ohmic process, the SEI film process, the charge transfer process, and the solid-state diffusion process; among them, the relaxation time interval associated with the lithium plating process is 10 seconds. 、2 Up to 10 1 In seconds, the area of ​​the characteristic peak within the interval is calculated in real time as the area of ​​the characteristic peak of the real-time lithium plating-related DRT.

[0031] The key impedance parameters include internal ohmic resistance, charge transfer resistance, and impedance phase angle.

[0032] Preferably, in step S3, the mathematical expression of the characteristic frequency drift polarization quantization algorithm is:

[0033] ;

[0034] in, The quantization value for the polarization state. For real-time characteristic frequencies, The initial characteristic frequency, This refers to the frequency band correction factor;

[0035] The mathematical expression for the DRT lithium plating risk quantification algorithm is as follows:

[0036] ;

[0037] in, This is a quantitative value for the risk of lithium plating. To determine the area of ​​the characteristic peak of DRT associated with real-time lithium plating. This represents the area of ​​the initial DRT characteristic peak. These are the mechanism weighting coefficients.

[0038] Preferably, the mathematical expression for the quadruple constraint peak comprehensive determination algorithm in step S4 is:

[0039] ;

[0040] in, As the peak value comprehensive determination factor, It is the voltage stability factor. , , , To constrain the weighting coefficients, This represents the real-time impedance change rate. The quantization value for the polarization state. This is a quantitative value for the risk of lithium plating, and .

[0041] Preferably, the voltage stability factor The calculation method is as follows:

[0042] ;

[0043] in, This refers to the real-time terminal voltage of the battery. The charge / discharge cutoff voltage is set in the initial configuration before testing. The voltage is determined to be in a stable state at that time.

[0044] Preferably, the step of comprehensive determination of the four constraints in step S4 includes:

[0045] After each round of iterative adjustment, impedance scanning and data processing is completed, a comprehensive judgment is made within 10ms, and the judgment delay is ≤15ms;

[0046] When the peak comprehensive determination factor When the preset stable range is exceeded, the binary step size is reduced to 50% to 70% of the current step size, and the charging and discharging current or power is rolled back to the parameter value of the previous iteration. After the rollback is completed, the system stabilizes for 3 to 5 seconds before starting the next iteration.

[0047] The four quantitative parameters—impedance change rate, polarization state quantification value, lithium plating risk quantification value, and peak comprehensive judgment factor—are compared in real time with the corresponding preset thresholds set in the initial configuration before the test. When any quantitative parameter exceeds the threshold twice consecutively, step size shrinkage and iteration rollback are triggered.

[0048] Secondly, this application also proposes a peak charge-discharge test system for lithium batteries based on the dynamic binary search method, comprising:

[0049] Electrical access and calibration module, used for pre-test initialization configuration: completes system self-test of the lithium battery under test and records initial baseline data;

[0050] The dynamic iterative adjustment module is used to obtain initial reference data and perform iterative adjustment of the lithium battery charging and discharging current or power through dynamic binary iterative adjustment. In each iteration, the dual-factor adaptive step size correction algorithm is called to correct the step size of the next iteration, thus completing the adjustment of charging and discharging parameters and the execution of actions in each round.

[0051] The wideband impedance scanning and decoupling module is used to perform wideband online impedance scanning of the lithium battery with a single scan duration of 8 or 10 ms after the parameters of each round of charging and discharging are adjusted and stabilized. It extracts key impedance parameters, identifies the real-time position of characteristic frequency points, decouples the relaxation time distribution of the impedance spectrum, and calls the characteristic frequency drift polarization quantization algorithm to calculate the polarization state quantization value and the DRT lithium plating risk quantization algorithm to calculate the lithium plating risk quantization value respectively.

[0052] The quadruple constraint comprehensive judgment module is used to integrate the real-time battery voltage, impedance change rate, polarization state quantification value and lithium plating risk quantification value, call the quadruple constraint peak comprehensive judgment algorithm to make a comprehensive judgment, compare the judgment result with the preset stable range, and determine whether to terminate the current iteration, shrink the step size or continue the iteration.

[0053] The peak locking and data storage module is used to repeat steps S2 to S4. When the judgment result is stable within the preset range for multiple consecutive test cycles and there are no sudden changes in the various state parameters of the battery, the peak charge and discharge boundary is locked, relevant parameters are output and associated with storage, and a database linking peak characteristics and internal electrochemical mechanisms is established.

[0054] Beneficial effects:

[0055] This invention combines a dynamic bisection method with a two-factor adaptive step size correction algorithm, using the real-time impedance change rate and characteristic frequency drift rate as input factors for step size correction. This allows the iteration step size to be dynamically adapted to the real-time changes in the internal electrochemical state of the battery, avoiding the problems of low efficiency when the traditional fixed step size iteration is far from the peak and easy overshoot when approaching the peak. Thus, it significantly improves the accuracy of peak locking while ensuring the convergence speed.

[0056] This invention achieves real-time, non-disruptive capture of the dynamic electrochemical characteristics of the battery by initiating a wideband online impedance scan within 2 to 3 seconds after the charging and discharging parameters have been stabilized in each round, controlling the single scan duration to 8 to 10 ms, and without interrupting the charging and discharging process during the scan. Furthermore, by decoupling the relaxation time distribution using the non-negative least squares method, the total impedance is decomposed into four independent sub-processes: the ohmic process, the SEI film process, the charge transfer process, and the solid-phase diffusion process. The relaxation time interval associated with lithium plating is locked, quantifying the polarization state and lithium plating risk at the mechanistic level. This provides a deeper mechanistic basis for peak value determination that cannot be provided by traditional single electrical indicators, making the determination results more scientific and accurate.

[0057] This invention constructs a comprehensive judgment system integrating four constraints: voltage stability factor, impedance change rate, polarization state quantification value, and lithium plating risk quantification value, and sets a judgment delay as low as 10ms, thereby achieving multi-dimensional, low-latency collaborative evaluation of battery peak state. At the same time, by setting multiple safety mechanisms such as step size contraction, parameter rollback, and continuous over-limit triggering, it can quickly identify risks and actively avoid them when approaching the limit boundary, effectively solving the safety hazards that may be caused by the single judgment dimension and lack of active safety protection in traditional solutions.

[0058] This invention establishes a database linking peak characteristics and electrochemical mechanisms by locking the peak charge and discharge boundaries and then storing the peak parameters in association with the internal electrochemical mechanism data. This enables the test results to not only serve the boundary determination of a single test, but also to provide systematic data support for subsequent battery research and development, operating condition matching optimization, and safety management strategy formulation, thus greatly expanding the application value of test data. Attached Figure Description

[0059] Figure 1 A schematic diagram of the method framework for this invention is provided.

[0060] Figure 2 A schematic diagram of the system structure is provided for this invention. Detailed Implementation

[0061] To make the technical solution of the present invention clearer, the present invention will be further described in detail below with reference to the accompanying drawings and specific embodiments.

[0062] Example 1:

[0063] To make the objectives, technical solutions, and advantages of this invention clearer, the technical solutions in the embodiments of this invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some, not all, of the embodiments of this invention. All other embodiments obtained by those skilled in the art based on the embodiments of this invention without inventive effort are within the scope of protection of this invention. Unless otherwise defined, the technical or scientific terms used herein should have the ordinary meaning understood by those skilled in the art. The terms "including" and similar expressions used herein mean that the element or object preceding the term covers the element or object listed after the term and its equivalents, but do not exclude other elements or objects.

[0064] In response to the problems existing in the current technology, such as Figure 1 As shown, a peak charge / discharge test method for lithium batteries based on dynamic bisection is proposed. This method achieves accurate, safe, and efficient locking of the peak charge / discharge boundary of lithium batteries by constructing a closed-loop test path of "initialization, iterative adjustment, mechanism decoupling, comprehensive judgment, and peak locking". The method includes the following steps:

[0065] Step S1: Initialize the configuration before testing. Complete the system self-test of the lithium battery under test and record the initial benchmark data.

[0066] This step is the starting point of the entire testing process. The initial baseline data is recorded by collecting and calibrating the key electrochemical characteristic parameters of the battery when it is in a stable state. This baseline data will serve as the reference point for subsequent steps to judge changes in battery state, calculate polarization, and assess lithium plating risk. It is understandable that the accuracy and representativeness of the initial baseline data directly affect the accuracy of the entire testing process; therefore, it is usually necessary to collect the data after the battery has undergone several pre-cycles and reached thermal equilibrium and electrochemical stability.

[0067] In some specific embodiments, S1 includes:

[0068] Electrical connection and operating condition calibration, setting parameters such as charging and discharging cutoff voltage, various constraint thresholds, wideband impedance scanning frequency band, dynamic binary initial iteration step size and algorithm coefficients;

[0069] The charge / discharge cutoff voltages are set based on the nominal voltage of the lithium battery under test. The upper limit of the charging voltage is 1.05 to 1.10 times the nominal voltage, and the lower limit of the discharging voltage is 0.80 to 0.85 times the nominal voltage. This range is based on the electrochemical window characteristics of lithium batteries: if the upper limit of the charging voltage exceeds 1.10 times the nominal voltage, the risk of electrolyte oxidation and decomposition increases significantly, while if it is below 1.05 times, it may not be able to fully cover the battery's true peak capacity; if the lower limit of the discharging voltage is below 0.80 times the nominal voltage, the risk of dissolution of the negative electrode copper current collector increases, and if it is above 0.85 times, the discharge may terminate prematurely, resulting in an underestimation of the peak power boundary. Therefore, this voltage window achieves a balance between safety and test adequacy.

[0070] The various constraint thresholds specifically include: an impedance change rate threshold of 3% / s to 8% / s, a characteristic frequency drift rate threshold of 50Hz / s to 150Hz / s, and a lithium plating-related relaxation time characteristic peak area change threshold of 8% / s to 12% / s. These thresholds correspond to different safety boundaries within the battery. The impedance change rate threshold is used to monitor the abrupt change rate of the battery's internal ohmic resistance and charge transfer resistance. When anomalies such as micro-short circuits or electrolyte drying occur inside the battery, the impedance change rate will rise sharply. The characteristic frequency drift rate threshold is used to capture the shift in the characteristic frequency of electrode reaction kinetics, which is directly related to the deterioration of the electrode polarization state. The lithium plating-related relaxation time characteristic peak area change threshold is specifically for lithium plating risk. During lithium plating, the deposition of metallic lithium on the negative electrode surface introduces a new relaxation time characteristic peak, and the growth rate of its peak area is the most direct quantitative characterization of lithium plating risk.

[0071] The wideband impedance scanning frequency is set from 10Hz to 1kHz. This band covers the characteristic response frequencies of the main electrochemical processes inside the lithium battery: the low-frequency band around 10Hz mainly reflects the solid-phase diffusion process, the mid-frequency band from 100Hz to several hundred Hz corresponds to the charge transfer process and the SEI film process, and the high-frequency band around 1kHz mainly reflects ohmic conduction characteristics. Setting the scanning frequency within this range allows for the complete capture of the entire impedance chain from ohmic conduction to solid-phase diffusion without significantly increasing the scanning time.

[0072] The initial iteration step size for the dynamic binary search is 10% to 20% of the rated current or power of the lithium battery under test, with a target iteration accuracy of 0.5% to 1% of the rated current or power. The choice of the initial step size directly affects the balance between iteration convergence speed and approximation accuracy. If the initial step size is too large, although the initial convergence is fast, it is prone to repeated overshoot when approaching the peak value; if the initial step size is too small, there will be too many iterations and the testing time will be too long. The initial step size range of 10% to 20% is an optimal range verified by a large number of experiments, which can reserve sufficient adjustment space for subsequent adaptive reduction of the step size while ensuring the convergence speed. The target iteration accuracy of 0.5% to 1% defines the final resolution of peak locking, which is sufficient to meet the requirements for determining the peak power or current boundary in engineering applications.

[0073] The initial reference data includes initial characteristic frequencies and initial DRT characteristic peak areas, which are collected after the lithium battery under test has undergone three pre-charge-discharge cycles to a stable state. The average of the three collections is taken as the initial reference data. Using the method of averaging after three pre-cycles effectively eliminates random errors that may exist in a single measurement and inconsistencies in the initial state of the battery. Pre-charge-discharge cycles bring the battery to thermal equilibrium and electrochemical stability. The characteristic frequencies and DRT characteristic peak areas collected at this time can accurately reflect the baseline electrochemical characteristics of the battery under normal operating conditions, providing a reliable reference origin for calculating the polarization state quantification value and the lithium plating risk quantification value in subsequent steps.

[0074] Step S2: Obtain initial baseline data and adjust the lithium battery charging and discharging current or power iteratively through dynamic binary iterative adjustment. In each iteration, the dual-factor adaptive step size correction algorithm is called to correct the step size of the next iteration, thus completing the adjustment of charging and discharging parameters and the execution of actions in each round.

[0075] This step is the core driving force of the testing process. Its key difference from traditional fixed-step iteration lies in the collaborative mechanism of "dynamic binary search" and "two-factor adaptive optimization." In traditional schemes, the iteration step size is set to a fixed value before the test begins, making it impossible to adjust according to the real-time state changes of the battery as it approaches its limits. This results in slow convergence when far from the peak value and repeated overshooting due to excessively large step sizes when approaching the peak value. The dynamic binary search method used in this step divides and shrinks the adjustment range of each iteration, allowing the search range to quickly focus near the peak value. Simultaneously, the two-factor adaptive step size correction algorithm called in each iteration dynamically adjusts the step size for the next round based on real-time feedback from the battery's internal electrochemical state—such as impedance change trends and characteristic frequency drift. This ensures the step size remains large when the battery state is stable to accelerate convergence, and automatically shrinks when the battery state undergoes drastic changes for precise approximation, thus achieving a dynamic balance between convergence speed and approximation accuracy.

[0076] In some specific embodiments, the dynamic binary iterative adjustment in step S2 includes:

[0077] A constant current charge-discharge mode is adopted, and the duration of each round of charge-discharge is set to 5 to 10 seconds;

[0078] The criterion for adaptive step size adjustment is as follows: when the iteration interval shrinks to less than 1 / 10 of the initial iteration step size, the step size is adjusted to 1 / 5 to 1 / 10 of the initial step size. The engineering logic of this phased adjustment strategy is as follows: in the early stage of iteration, the current search interval is relatively wide, and the battery state is still far from the peak boundary. At this time, maintaining a large step size can quickly narrow the search range and improve convergence efficiency. When the search interval shrinks to less than 1 / 10 of the initial step size, it means that the current parameters have approached the peak. If a large step size is continued to be used at this time, it is very easy to exceed the true peak boundary due to excessive adjustment amplitude in a single step, resulting in overshoot and oscillation. Switching the step size to 1 / 5 to 1 / 10 of the initial step size is equivalent to increasing the adjustment resolution by 5 to 10 times in the approximation stage, so that the iteration process can converge gradually "closely" to the true limit boundary of the battery.

[0079] Continue iterating when the iteration interval is greater than 1 / 10 of the initial iteration step size;

[0080] During the iteration process, the battery surface temperature is monitored in real time. When the temperature exceeds 40°C, the initial step size is reduced by 30%. Temperature is a key indicator of the safety status of lithium batteries. As the battery approaches its peak charge / discharge boundary, internal polarization intensifies, the Joule heating effect strengthens, and the surface temperature gradually rises. 40°C is a crucial safety warning threshold—beyond this temperature, the rate of internal side reactions accelerates significantly, increasing the risk of SEI film decomposition and electrolyte consumption. Actively reducing the step size by 30% at this point slows down the approach to the limit boundary, giving the battery more time for heat dissipation and recovery, thus allowing the test to continue approaching the peak while ensuring safety, rather than simply terminating the test. This "deceleration without stopping" strategy achieves a fine trade-off between safety and testing objectives.

[0081] In some specific embodiments, the mathematical expression of the two-factor adaptive step size correction algorithm in step S2 is:

[0082] ;

[0083] in, The step size is the binary search step size in the (n+1)th iteration. The step size for the nth iteration is the binary step size. This is the normalized value of the real-time impedance change rate. This is the normalized value of the characteristic frequency drift rate. , For safety weighting coefficients, and .

[0084] The core idea of ​​this algorithm is to use the two most sensitive electrochemical state indicators inside the battery—impedance change rate and characteristic frequency drift rate—as input factors for step size correction, so that the iteration step size can respond to changes in battery state in real time.

[0085] Specifically, the normalized value of the real-time impedance change rate This is a dimensionless quantity obtained by dividing the impedance change rate measured in the current round by the impedance change rate threshold set in the initial configuration. For example, if the current impedance change rate is 4% / s, and the initial impedance change rate threshold is 8% / s, then... =0.5. Similarly, the normalized value of the characteristic frequency drift rate. This is a dimensionless quantity obtained by dividing the characteristic frequency drift rate measured in the current round by the initially set characteristic frequency drift rate threshold. For example, if the current characteristic frequency drift rate is 75 Hz / s, and the initially set threshold is 150 Hz / s, then... =0.5. Through normalization, two physical quantities with different dimensions are mapped to the same scale, enabling weighted fusion in the formula.

[0086] Safety weighting coefficient , The contribution weights of impedance change rate and characteristic frequency drift rate to step size correction are determined respectively. + The constraint of 1 ensures that the total amount of the correction factor will not exceed 1, meaning the step size will not be corrected to a negative value or excessively reduced. In practical applications, , The value can be configured according to the battery type and testing requirements. For example, for power batteries where power density is the primary indicator, impedance changes provide a more direct indication of peak boundaries. Set it to 0.6 to 0.7. The corresponding values ​​are set to 0.4 to 0.3; for energy-type batteries where energy density is the primary indicator, the characteristic frequency drift is more sensitive to the polarization state, and can be... Set it to 0.6 to 0.7. The corresponding values ​​are set to 0.4 to 0.3. This configurable weighting mechanism allows the algorithm to adapt to the differences in electrochemical characteristics among different types of lithium batteries.

[0087] When the battery state is stable and far from the limit boundary and All are relatively small, correction items Close to 0 ≈ The step size remains essentially constant, and the iteration converges relatively quickly. As the battery gradually approaches the peak boundary, internal polarization intensifies, and the rate of impedance change and characteristic frequency drift begin to increase. and As it increases, the correction term increases. Compared to Significantly reduced size, with automatically narrowing iteration step size, achieving fine approximation. When the battery state undergoes abrupt changes—such as a sharp increase in lithium plating risk leading to a surge in impedance change rate—[this is particularly effective]. If the value approaches or exceeds 1, the correction term increases significantly, the step size is greatly compressed, and it may even trigger the step size contraction and rollback mechanism in subsequent step S4, thus forming the first line of defense at the algorithm level. This embodiment constructs a complete iterative control chain from parameter benchmark setting and process control to algorithm correction through the synergistic cooperation of the aforementioned initialization parameter configuration, phased step size control strategy, and two-factor adaptive step size correction algorithm. The initialization configuration provides a quantified safety boundary and accuracy target for the iterative process; the phased step size control strategy implements the switching logic of "coarse adjustment" and "fine adjustment" at the macroscopic level; and the two-factor adaptive algorithm realizes the real-time response of the step size to the battery's electrochemical state at the microscopic level. These three elements work in a progressive manner to jointly ensure the efficiency and safety of the peak approximation process.

[0088] Step S3: After the charging and discharging parameters are stabilized in each round, perform a wideband online impedance scan of the lithium battery with a single scan duration of 8 and 10 ms, extract key impedance parameters, identify the real-time position of characteristic frequency points, decouple the relaxation time distribution of the impedance spectrum, and call the characteristic frequency drift polarization quantization algorithm to calculate the polarization state quantization value and the DRT lithium plating risk quantization algorithm to calculate the lithium plating risk quantization value respectively.

[0089] This step enables real-time, non-disruptive capture of the battery's internal electrochemical characteristics, serving as a crucial bridge between external electrical parameter adjustment and quantitative assessment of internal mechanisms. Traditional impedance detection schemes typically require interrupting the charge-discharge process and placing the battery in a static state before scanning. This not only reduces testing efficiency but, more importantly, fails to capture the battery's true electrochemical state during dynamic charge-discharge processes—because the polarization state and ion concentration distribution within the battery have already undergone relaxation recovery after static conditions. This step controls the single scan duration to an extremely short time of 8 to 10 milliseconds, while maintaining a constant charge-discharge current or power during the scan. This allows impedance measurement to be completed without interrupting the charge-discharge process, truly achieving "online" capture of the battery's electrochemical characteristics under dynamic operating conditions. Building upon this, through relaxation time distribution decoupling technology, the macroscopic total impedance is decomposed into independent sub-processes corresponding to different physicochemical processes. This allows for the differentiation of the contributions of ohmic conduction, SEI film interface, charge transfer, and solid-phase diffusion at the mechanistic level, further quantifying polarization state and lithium plating risk, and providing deeply meaningful characteristic parameters for subsequent comprehensive judgment.

[0090] In some specific embodiments, the specific steps of broadband impedance scanning and decoupling in step S3 include:

[0091] After each round of charging and discharging parameters are adjusted and stabilized for 2 to 3 seconds, a wideband impedance scan is started. The duration of a single scan is controlled to be 8 to 10 ms. During the scan, the charging and discharging current or power remains unchanged and the charging and discharging process is not interrupted.

[0092] The characteristic frequency points specifically include: the frequency corresponding to the imaginary minimum point of the impedance spectrum, the phase angle, the frequency corresponding to the 45° point, and the characteristic peak frequency of the time constant of the charge transfer process;

[0093] These three characteristic frequency points each correspond to different electrochemical processes within the battery and have clear physical significance. The frequency corresponding to the imaginary minimum point of the impedance spectrum is represented on the Nyquist plot as the region near the intersection of the impedance curve and the real axis. This frequency mainly reflects the ohmic conductivity of the battery—including the combined contributions of electrolyte ion conduction, electrode electronic conduction, and contact resistance at various interfaces. A shift in this frequency often indicates electrolyte consumption or contact degradation. The frequency corresponding to the 45° phase angle is represented on the Bode plot as the position where the phase angle curve crosses at 45°. This frequency is usually located in the transition region between the charge transfer and diffusion processes and is extremely sensitive to changes in electrode reaction kinetics. When electrode polarization intensifies, this characteristic frequency will shift significantly towards lower frequencies. The characteristic peak frequency of the time constant of the charge transfer process is the frequency corresponding to the characteristic peak identified within the relaxation time interval of the charge transfer sub-process after decoupling through the relaxation time distribution. It directly characterizes the kinetic rate of the charge transfer reaction of lithium ions at the electrode / electrolyte interface and is a core indicator for quantifying polarization state. It is understood that in other embodiments, depending on the battery system and testing requirements, other characteristic frequency points with clear physical meaning can be selected, such as the characteristic peak frequency of the time constant of the SEI film process or the characteristic frequency of the solid-phase diffusion process, to further enrich the dimensions of mechanism decoupling.

[0094] The relaxation time distribution decoupling employs the non-negative least squares method, and the decoupling frequency range corresponds to a relaxation time of 10. 、3 Up to 10 2 The total impedance is decoupled into four independent sub-processes: the ohmic process, the SEI film process, the charge transfer process, and the solid-state diffusion process; among them, the relaxation time interval associated with the lithium plating process is 10 seconds. 、2 Up to 10 1In s seconds, the characteristic peak area within the interval is calculated in real time as the characteristic peak area of ​​the DRT associated with real-time lithium plating. After the charging and discharging parameters are adjusted, the ion concentration distribution and electrode polarization state inside the battery require a short transition time to reach a new quasi-steady state. The stabilization waiting period of 2 to 3 seconds is precisely to ensure that the battery is in a relatively stable electrochemical state during impedance measurement, avoiding transient disturbances during parameter adjustment that could contaminate the impedance data. Compressing the single scan time to 8 to 10 ms utilizes the rapid frequency sweep technique of electrochemical impedance spectroscopy measurement—by injecting a set of tiny AC excitation signals with frequencies from 10 Hz to 1 kHz into the battery in a very short time and simultaneously acquiring the response signal, the impedance spectrum of the entire frequency band can be plotted within a millisecond time window. Because the scan time is extremely short, the charging and discharging current or power remains constant during this period, and the macroscopic charging and discharging state of the battery is not interrupted. Therefore, this scanning method is called "online" scanning. This is fundamentally different from the traditional "offline" method that requires disconnecting the battery from the charging and discharging circuit, letting it stand, and then measuring the impedance. Online scanning captures the impedance characteristics of the battery under real charging and discharging dynamic conditions, rather than the relaxed state characteristics after resting and recovery. The latter often loses key information about the polarization state and concentration gradient during the charging and discharging process.

[0095] The core idea of ​​relaxation time distribution decoupling lies in the fact that the total impedance of a battery is a linear superposition of multiple electrochemical sub-processes with different time constants, each corresponding to a specific relaxation time. By mathematically inverting the total impedance spectrum into a relaxation time distribution function, the sub-processes that originally overlapped in the frequency domain can be separated in the time domain. The reason for using non-negative least squares instead of ordinary Fourier transform or Tikhonov regularization is that the relaxation time distribution function of the electrochemical process must physically satisfy the non-negativity constraint—the contribution of each sub-process to the total impedance can only be positive, and a negative impedance contribution is impossible. Non-negative least squares achieves a good balance between fitting accuracy and physical interpretability, avoiding the generation of physically meaningless negative peaks.

[0096] The key impedance parameters include internal ohmic resistance, charge transfer resistance, and impedance phase angle.

[0097] In some specific embodiments, in step S3, the mathematical expression of the characteristic frequency drift polarization quantization algorithm is:

[0098] ;

[0099] in, The quantization value for the polarization state. For real-time characteristic frequencies, The initial characteristic frequency, This refers to the frequency band correction factor;

[0100] Increased battery polarization leads to changes in the characteristic time constant of electrode reaction kinetics, resulting in a drift in the characteristic frequency. When the battery is in normal operation, the real-time characteristic frequency... With initial characteristic frequency Basically the same, Approaching zero indicates that the polarization degree has not changed significantly. As the battery gradually approaches the peak boundary and polarization intensifies, the characteristic frequency shifts towards lower frequencies. Less than , The value becomes negative, and the larger the absolute value, the more severe the polarization. The function of the frequency band correction coefficient λ is to normalize the sensitivity of the characteristic frequency drift in different frequency bands—since the characteristic frequencies of high-frequency and low-frequency bands have different sensitivities to polarization, the introduction of the λ coefficient can normalize the sensitivity of the calculated frequency drift in different frequency bands. They are comparable. In practical applications, the value of λ can be calibrated according to the frequency band where the selected characteristic frequency point is located. For example, for the charge transfer characteristic frequency in the mid-frequency band, λ is usually taken as a value between 1.0 and 1.5.

[0101] The mathematical expression for the DRT lithium plating risk quantification algorithm is as follows:

[0102] ;

[0103] in, This is a quantitative value for the risk of lithium plating. To determine the area of ​​the characteristic peak of DRT associated with real-time lithium plating. This represents the area of ​​the initial DRT characteristic peak. These are the mechanism weighting coefficients.

[0104] The risk of lithium plating is directly related to the rate of increase in the characteristic peak area within the lithium plating-related relaxation time interval. (Initial DRT characteristic peak area) This is a baseline value obtained by averaging data collected during pre-charge and discharge cycles under healthy battery conditions. It reflects the battery's inherent impedance contribution within this relaxation time interval under normal operating conditions. Real-time lithium plating is correlated with the characteristic peak area of ​​the DRT. This is the current value calculated online in each iteration. When lithium plating has not occurred in the battery, and Basically unchanged. Approaching zero. When lithium plating begins, the additional impedance contribution introduced by lithium metal deposition leads to an increase in the area of ​​the characteristic peak in this range. Exceed , The value becomes positive, and the larger the value, the higher the risk of lithium plating. The mechanism weighting coefficient μ is used to adjust the sensitivity of the quantification value of lithium plating risk. Its value can be configured according to the tolerance of the battery system to lithium plating. For battery systems with low tolerance to lithium metal plating, μ can be taken as a larger value to enhance the warning sensitivity; for systems with a certain tolerance to lithium plating, μ can be taken as a smaller value to avoid oversensitivity.

[0105] Step S4: Integrate the real-time battery voltage, impedance change rate, polarization state quantification value and lithium plating risk quantification value, call the quadruple constraint peak comprehensive judgment algorithm for comprehensive judgment, compare the judgment result with the preset stable range, and determine whether to terminate the current iteration, shrink the step size or continue the iteration.

[0106] This step constructs a multi-dimensional, multi-source information fusion judgment system, breaking through the limitations of traditional schemes that rely solely on single voltage or current indicators for peak value determination. During the process of a lithium battery approaching its peak charge / discharge boundary, minute changes in voltage often lag behind the deterioration of its internal electrochemical state—for example, even when the risk of lithium plating has significantly increased, the terminal voltage may not yet show obvious anomalies. The four dimensions integrated in this step characterize the battery's real-time state from different perspectives: voltage reflects macroscopic electrical characteristics, impedance change rate reflects the dynamic evolution of the battery's internal conduction characteristics, polarization state quantification reflects the degree to which the electrode reaction deviates from equilibrium, and lithium plating risk quantification directly points to the most dangerous safety hazards. The four-constraint comprehensive judgment algorithm weighted and fused the information from these four dimensions to generate a comprehensive judgment factor. By comparing it with a preset stable range, it can more comprehensively and sensitively determine whether the current iteration has reached the battery's true limit, thereby making scientific decisions to continue iteration, reduce the step size for finer approximation, or terminate the iteration.

[0107] In some specific embodiments, the mathematical expression for the quadruple constraint peak comprehensive determination algorithm in step S4 is:

[0108] ;

[0109] in, As the peak value comprehensive determination factor, It is the voltage stability factor. , , , To constrain the weighting coefficients, This represents the real-time impedance change rate. The quantization value for the polarization state. This is a quantitative value for the risk of lithium plating, and .

[0110] This weighted fusion formula integrates state information from four dimensions into a dimensionless comprehensive judgment factor. Each dimension characterizes the real-time state of the battery as it approaches the peak boundary from a different perspective. Voltage stability factor Real-time impedance change rate reflects the degree of deviation of the battery terminal voltage from the set cutoff voltage, and is an indicator of whether the battery is approaching its limit from a macroscopic electrical characteristic perspective. The rate of change of impedance reflects the dynamic evolution of conductive properties such as internal ohmic resistance and charge transfer resistance within the battery. When abnormalities occur within the battery, such as electrolyte depletion, micro-short circuits, or electrode structure degradation, the rate of impedance change increases significantly. Polarization state quantification value. This reflects the degree to which the electrode reaction deviates from thermodynamic equilibrium. The more severe the polarization, the greater the lithium-ion concentration gradient at the electrode surface, the further the electrode potential deviates from the equilibrium potential, and the closer the battery is to its mass transfer limit. Lithium plating risk quantification value. This directly points to the most dangerous safety hazard—the irreversible deposition of metallic lithium on the negative electrode surface. An increase in this indicator means the battery has entered a dangerous zone where permanent damage may occur. The weighted fusion of four dimensions makes the comprehensive judgment factor... It can reflect the true limiting state of a battery more comprehensively and sensitively than any single indicator.

[0111] In some specific embodiments, the voltage stability factor The calculation method is as follows:

[0112] ;

[0113] in, This refers to the real-time terminal voltage of the battery. The charge / discharge cutoff voltage is set in the initial configuration before testing. The voltage is determined to be in a stable state at that time.

[0114] when When the deviation is ≤0.02, it means that the deviation between the real-time terminal voltage and the cutoff voltage does not exceed 2% of the cutoff voltage. At this point, the battery can be considered to be in a stable state in terms of voltage and has not yet reached the voltage limit boundary. A value exceeding 0.02 indicates that the battery terminal voltage has significantly deviated from the set value, possibly rapidly approaching or already reaching its voltage limit, requiring close attention in the comprehensive judgment process. The 0.02 threshold is an optimal value verified through extensive experimentation—a threshold that is too high will lead to insufficient judgment sensitivity, potentially failing to trigger an alert even when a significant voltage anomaly has occurred; a threshold that is too low may result in frequent false alarms due to normal measurement noise or minor fluctuations, affecting the smoothness of the testing process. Of course, in other embodiments, depending on the battery system and testing accuracy requirements, this threshold can also be adjusted to a nearby value such as 0.015 or 0.025, as long as it meets the reasonable judgment requirements for voltage stability.

[0115] In some specific embodiments, the step of comprehensive determination of the four constraints in step S4 includes:

[0116] After each round of iterative adjustment, impedance scanning and data processing, a comprehensive judgment is completed within 10ms, with a judgment delay of ≤15ms. The design of completing the judgment within 10ms and the overall delay not exceeding 15ms ensures that the entire chain from data acquisition to judgment command output is completed within an extremely short time window, enabling the test system to respond to changes in battery status at a near real-time speed.

[0117] When the peak comprehensive determination factor When the battery exceeds the preset stable range, the step size is reduced to 50% to 70% of the current step size, and the charging / discharging current or power is rolled back to the parameter values ​​of the previous iteration. After the rollback is completed, the battery stabilizes for 3 to 5 seconds before starting the next iteration. This stabilization waiting period allows time for the polarization state, ion concentration distribution, and temperature field inside the battery to recover to a relatively balanced state, ensuring that the next iteration starts from a stable and controllable starting point, rather than continuing rashly before the disturbance of the previous iteration has subsided. This strategy plays a dual role of safety buffer and state reset when approaching the limit boundary.

[0118] Four quantification parameters—impedance change rate, polarization state quantification value, lithium plating risk quantification value, and peak value comprehensive judgment factor—are compared in real time with their corresponding preset thresholds set in the pre-test initialization configuration. When any quantification parameter exceeds the threshold twice consecutively, step size shrinkage and iteration rollback are triggered.

[0119] This mechanism, triggering a safety retreat after two consecutive exceedances, constitutes the final safety line of defense in the entire system. The reason for requiring two consecutive exceedances, rather than a single exceedance, is to avoid false triggers caused by transient noise interference or accidental measurement fluctuations—a single exceedance might be a transient disturbance in the measurement system, while two consecutive exceedances have a higher confidence level, indicating that the battery state has indeed experienced a persistent anomaly. Furthermore, this mechanism covers every one of the four quantitative parameters, meaning that regardless of which dimension's indicator first issues a danger signal, a safety retreat can be triggered independently without waiting for a comprehensive judgment factor. It also exceeds the limit. This design, which involves independent monitoring of multiple parameters and responds immediately when any parameter exceeds the limit, forms a redundant safety monitoring network, avoiding safety blind spots caused by the failure of a single judgment channel.

[0120] Through a multi-layered approach employing mechanisms such as weighted fusion of the aforementioned four constraints, low-latency judgment, step-size contraction and parameter rollback, and continuous over-limit triggering, this embodiment constructs a complete safety protection system from multi-dimensional perception and rapid judgment to proactive avoidance. This system does not simply terminate the test when a danger is detected; rather, through precise step-size control and state rollback, it achieves a dynamic balance between safety and approaching limits. This allows the testing process to fully exploit the battery's peak capabilities while also promptly stopping and proactively avoiding risks at their initial appearance. This fundamentally solves the safety hazards caused by traditional solutions due to their single judgment dimension, delayed response, and lack of proactive avoidance mechanisms.

[0121] Step S5: Repeat steps S2 to S4. When the judgment result is stable within the preset range for multiple consecutive test cycles and there are no sudden changes in the various state parameters of the battery, lock the peak charge and discharge boundary, output the relevant parameters and store them in association, and establish a database linking peak characteristics and internal electrochemical mechanisms.

[0122] This step is the convergence endpoint of the testing process and the stage for solidifying data value. By repeatedly executing a closed loop of "iterative adjustment, impedance scanning, and comprehensive judgment," the testing system continuously approaches and ultimately locks the peak charge-discharge boundary of the battery. The locking condition requires that the judgment result remain stable within a preset range for multiple consecutive test cycles, and that there are no sudden changes in the various state parameters of the battery. This design avoids false locking caused by accidental fluctuations or instantaneous disturbances, ensuring that the output peak parameters have repeatability and reliability. After locking the peak, this step not only outputs macroscopic boundary parameters such as peak current or power, but also associates and stores the mechanistic data such as polarization state, lithium plating risk, and DRT characteristic peak area collected during this test with the peak parameters, establishing a mapping relationship between peak characteristics and internal electrochemical mechanisms. The construction of this associated database makes the value of the test results go beyond a single boundary judgment itself, providing systematic data support for subsequent battery R&D optimization, operating condition matching strategy formulation, and safety management algorithm development.

[0123] Through the closed-loop testing process consisting of the above five steps, this embodiment organically integrates the rapid convergence characteristics of the dynamic bisection method, the real-time response capability of the dual-factor adaptive step size correction, the non-disruptive capture advantage of wideband impedance online scanning, the mechanism differentiation capability of relaxation time distribution decoupling, and the multi-dimensional safety assessment capability of four-fold constraint comprehensive judgment into a whole, realizing the comprehensive and accurate locking of the peak charge and discharge boundary of lithium battery from macroscopic to microscopic, and from external characteristics to internal mechanisms.

[0124] On the other hand, this application proposes a peak charge-discharge test system for lithium batteries based on the dynamic bisection method, combined with... Figure 2 ,include:

[0125] Electrical access and calibration module 201 is used for pre-test initialization configuration: the lithium battery under test completes system self-test and records initial benchmark data;

[0126] The dynamic iterative adjustment module 202 is used to obtain initial reference data and perform iterative adjustment on the charging and discharging current or power of the lithium battery through dynamic binary iterative adjustment. In each round of iteration, the dual-factor adaptive step size correction algorithm is called to correct the step size of the next round of iteration, thus completing the adjustment of charging and discharging parameters and the execution of actions in each round of charging and discharging.

[0127] The wideband impedance scanning and decoupling module 203 is used to perform wideband online impedance scanning of the lithium battery with a single scan duration of 8 or 10 ms after the parameters of each round of charging and discharging are adjusted and stabilized. It extracts key impedance parameters, identifies the real-time position of characteristic frequency points, decouples the relaxation time distribution of the impedance spectrum, and calls the characteristic frequency drift polarization quantization algorithm to calculate the polarization state quantization value and the DRT lithium plating risk quantization algorithm to calculate the lithium plating risk quantization value respectively.

[0128] The quadruple constraint comprehensive judgment module 204 is used to integrate the real-time battery voltage, impedance change rate, polarization state quantification value and lithium plating risk quantification value, call the quadruple constraint peak comprehensive judgment algorithm to make a comprehensive judgment, compare the judgment result with the preset stable range, and determine whether to terminate the current iteration, shrink the step size or continue the iteration.

[0129] The peak locking and data storage module 205 is used to repeat steps S2 to S4. When the judgment result is stable within the preset range for multiple consecutive test cycles and there are no sudden changes in the various state parameters of the battery, the peak charge and discharge boundary is locked, relevant parameters are output and associated with storage, and a database of correlation between peak characteristics and internal electrochemical mechanism is established.

[0130] The above description is merely a specific implementation of the embodiments of the present invention, but the protection scope of the embodiments of the present invention is not limited thereto. Any changes or substitutions within the technical scope disclosed in the embodiments of the present invention should be covered within the protection scope of the embodiments of the present invention. Therefore, the protection scope of the embodiments of the present invention should be determined by the protection scope of the claims.

Claims

1. A peak charge / discharge test method for lithium batteries based on the dynamic dichotomy method, characterized in that, Includes the following steps: Step S1: Initialize the configuration before testing. Complete the system self-test of the lithium battery under test and record the initial benchmark data. Step S2: Obtain initial baseline data and adjust the lithium battery charging and discharging current or power iteratively through dynamic binary iterative adjustment. In each iteration, the dual-factor adaptive step size correction algorithm is called to correct the step size of the next iteration, thus completing the adjustment of charging and discharging parameters and the execution of actions in each round. Step S3: After the charging and discharging parameters are stabilized in each round, perform a wideband online impedance scan of the lithium battery with a single scan duration of 8 and 10 ms, extract key impedance parameters, identify the real-time position of characteristic frequency points, decouple the relaxation time distribution of the impedance spectrum, and call the characteristic frequency drift polarization quantization algorithm to calculate the polarization state quantization value and the DRT lithium plating risk quantization algorithm to calculate the lithium plating risk quantization value respectively. Step S4: Integrate the real-time battery voltage, impedance change rate, polarization state quantification value and lithium plating risk quantification value, call the quadruple constraint peak comprehensive judgment algorithm for comprehensive judgment, compare the judgment result with the preset stable range, and determine whether to terminate the current iteration, shrink the step size or continue the iteration. Step S5: Repeat steps S2 to S4. When the judgment result is stable within the preset range for multiple consecutive test cycles and there are no sudden changes in the various state parameters of the battery, lock the peak charge and discharge boundary, output the relevant parameters and store them in association, and establish a database linking peak characteristics and internal electrochemical mechanisms.

2. The peak charge / discharge test method for lithium batteries based on the dynamic dichotomy method according to claim 1, characterized in that, S1 includes: Electrical connection and operating condition calibration, setting parameters such as charging and discharging cutoff voltage, various constraint thresholds, wideband impedance scanning frequency band, dynamic binary initial iteration step size and algorithm coefficients; The charging and discharging cutoff voltage is set according to the nominal voltage of the lithium battery under test, with the upper limit charging voltage being 1.05 to 1.10 times the nominal voltage and the lower limit discharging voltage being 0.80 to 0.85 times the nominal voltage. The various constraint thresholds specifically include: impedance change rate threshold of 3% / s to 8% / s, characteristic frequency drift rate threshold of 50Hz / s to 150Hz / s, and characteristic peak area change threshold of lithium plating-related relaxation time of 8% / s to 12% / s. The wideband impedance scanning frequency band is set to 10Hz to 1kHz; The initial iteration step size of the dynamic binary search is 10% to 20% of the rated current or power of the lithium battery under test, and the target iteration accuracy is 0.5% to 1% of the rated current or power. The initial reference data includes the initial characteristic frequency and the initial DRT characteristic peak area, which are collected after the lithium battery under test has undergone three pre-charge-discharge cycles to a stable state, and the average value of the three collections is taken as the initial reference data.

3. The peak charge / discharge test method for lithium batteries based on the dynamic dichotomy method according to claim 2, characterized in that, The dynamic binary iterative adjustment in step S2 includes: A constant current charge-discharge mode is adopted, and the duration of each round of charge-discharge is set to 5 to 10 seconds; The criterion for adaptive step size adjustment is: when the iteration interval shrinks to less than 1 / 10 of the initial iteration step size, the step size is adjusted to 1 / 5 to 1 / 10 of the initial step size; Continue iterating when the iteration interval is greater than 1 / 10 of the initial iteration step size; The battery surface temperature is monitored in real time during the iteration process. When the temperature exceeds 40°C, the initial step size is reduced by 30%.

4. The peak charge / discharge test method for lithium batteries based on the dynamic dichotomy method according to claim 2, characterized in that, The mathematical expression for the two-factor adaptive step size correction algorithm in step S2 is: ； in, The step size is the binary search step size in the (n+1)th iteration. The step size for the nth iteration is the binary step size. This is the normalized value of the real-time impedance change rate. This is the normalized value of the characteristic frequency drift rate. , For safety weighting coefficients, and .

5. The peak charge / discharge test method for lithium batteries based on the dynamic dichotomy method according to claim 1, characterized in that, The specific steps of broadband impedance scanning and decoupling in step S3 include: After each round of charging and discharging parameters are adjusted and stabilized for 2 to 3 seconds, a wideband impedance scan is started. The duration of a single scan is controlled to be 8 to 10 ms. During the scan, the charging and discharging current or power remains unchanged and the charging and discharging process is not interrupted. The characteristic frequency points specifically include: the frequency corresponding to the imaginary minimum point of the impedance spectrum, the phase angle, the frequency corresponding to the 45° point, and the characteristic peak frequency of the time constant of the charge transfer process; The relaxation time distribution decoupling employs the non-negative least squares method, and the decoupling frequency range corresponds to a relaxation time of 10. 、3 Up to 10 2 The total impedance is decoupled into four independent sub-processes: the ohmic process, the SEI film process, the charge transfer process, and the solid-state diffusion process; among them, the relaxation time interval associated with the lithium plating process is 10 seconds. 、2 Up to 10 1 In seconds, the area of ​​the characteristic peak within the interval is calculated in real time as the area of ​​the characteristic peak of the real-time lithium plating-related DRT. The key impedance parameters include internal ohmic resistance, charge transfer resistance, and impedance phase angle.

6. The peak charge / discharge test method for lithium batteries based on the dynamic dichotomy method according to claim 5, characterized in that, In step S3, the mathematical expression of the characteristic frequency drift polarization quantization algorithm is: ； in, The quantization value for the polarization state. For real-time characteristic frequencies, The initial characteristic frequency, This refers to the frequency band correction factor; The mathematical expression for the DRT lithium plating risk quantification algorithm is as follows: ； in, This is a quantitative value for the risk of lithium plating. To determine the area of ​​the characteristic peak of DRT associated with real-time lithium plating. This represents the area of ​​the initial DRT characteristic peak. These are the mechanism weighting coefficients.

7. The peak charge / discharge test method for lithium batteries based on the dynamic dichotomy method according to claim 1, characterized in that, The mathematical expression for the quadruple constraint peak comprehensive determination algorithm in step S4 is: ； in, As the peak value comprehensive determination factor, It is the voltage stability factor. , , , To constrain the weighting coefficients, This represents the real-time impedance change rate. The quantization value for the polarization state. This is a quantitative value for the risk of lithium plating, and .

8. The peak charge / discharge test method for lithium batteries based on the dynamic dichotomy method according to claim 7, characterized in that, The voltage stability factor The calculation method is as follows: ； in, This refers to the real-time terminal voltage of the battery. The charge / discharge cutoff voltage is set in the initial configuration before testing. The voltage is determined to be in a stable state at that time.

9. The peak charge / discharge test method for lithium batteries based on the dynamic dichotomy method according to claim 7, characterized in that, The steps for comprehensive determination of the four constraints in step S4 include: After each round of iterative adjustment, impedance scanning and data processing is completed, a comprehensive judgment is made within 10ms, and the judgment delay is ≤15ms; When the peak comprehensive determination factor When the preset stable range is exceeded, the binary step size is reduced to 50% to 70% of the current step size, and the charging and discharging current or power is rolled back to the parameter value of the previous iteration. After the rollback is completed, the system stabilizes for 3 to 5 seconds before starting the next iteration. The four quantitative parameters—impedance change rate, polarization state quantification value, lithium plating risk quantification value, and peak comprehensive judgment factor—are compared in real time with the corresponding preset thresholds set in the initial configuration before the test. When any quantitative parameter exceeds the threshold twice consecutively, step size shrinkage and iteration rollback are triggered.

10. A peak charge / discharge testing system for lithium batteries based on the dynamic binary search method, characterized in that, include: Electrical access and calibration module, used for pre-test initialization configuration: completes system self-test of the lithium battery under test and records initial baseline data; The dynamic iterative adjustment module is used to obtain initial reference data and perform iterative adjustment of the lithium battery charging and discharging current or power through dynamic binary iterative adjustment. In each iteration, the dual-factor adaptive step size correction algorithm is called to correct the step size of the next iteration, thus completing the adjustment of charging and discharging parameters and the execution of actions in each round. The wideband impedance scanning and decoupling module is used to perform wideband online impedance scanning of the lithium battery with a single scan duration of 8 or 10 ms after the parameters of each round of charging and discharging are adjusted and stabilized. It extracts key impedance parameters, identifies the real-time position of characteristic frequency points, decouples the relaxation time distribution of the impedance spectrum, and calls the characteristic frequency drift polarization quantization algorithm to calculate the polarization state quantization value and the DRT lithium plating risk quantization algorithm to calculate the lithium plating risk quantization value respectively. The quadruple constraint comprehensive judgment module is used to integrate the real-time battery voltage, impedance change rate, polarization state quantification value and lithium plating risk quantification value, call the quadruple constraint peak comprehensive judgment algorithm to make a comprehensive judgment, compare the judgment result with the preset stable range, and determine whether to terminate the current iteration, shrink the step size or continue the iteration. The peak locking and data storage module is used to repeat steps S2 to S4. When the judgment result is stable within the preset range for multiple consecutive test cycles and there are no sudden changes in the various state parameters of the battery, the peak charge and discharge boundary is locked, relevant parameters are output and associated with storage, and a database linking peak characteristics and internal electrochemical mechanisms is established.

Images