A secondary battery capacity fade monitoring method for consumer products

Abstract

This invention discloses a method for monitoring the capacity decay of secondary batteries in consumer products, relating to the field of battery testing technology. The method includes applying a symmetrical triangular wave current to the battery each time it is charged and its state of charge reaches a starting threshold. This current first rises linearly to a preset positive current value and then linearly decreases to zero, with the rise and fall times being equal. During the application of this current, the battery terminal voltage is continuously collected to obtain a voltage sequence. The sampling time corresponding to the maximum voltage value is found from the voltage sequence. The transition time between the rising and falling phases of the symmetrical triangular wave current is obtained. This invention achieves stable, accurate, and online monitoring of the capacity decay of secondary batteries in consumer products without relying on complete charge-discharge cycles or interfering with normal user operation, significantly improving the robustness, environmental adaptability, and safety of the detection.

Description

Technical Field

[0001] This invention belongs to the field of battery testing technology, specifically a method for monitoring the capacity decay of secondary batteries in consumer products. Background Technology

[0002] The secondary batteries used in consumer electronics products experience capacity decay and performance degradation during cyclic use, directly affecting battery life and safety. Existing battery capacity decay monitoring methods mostly rely on full-charge capacity calculation, coulomb integration, and internal resistance detection for state assessment. These methods require a complete charge-discharge cycle and cannot be quickly completed during normal charging. They are also susceptible to fluctuations in charging current, changes in ambient temperature, and interference from electrical loads.

[0003] Some methods detect the internal characteristics of batteries using pulse current and AC impedance, but most employ symmetrical excitation and global statistical analysis, failing to delve into the time-series characteristics of polarization response. During battery aging, electrochemical polarization and concentration polarization change simultaneously. Traditional methods only focus on voltage amplitude or impedance magnitude, failing to distinguish the dynamic process of polarization change and easily leading to misjudgments.

[0004] Meanwhile, existing methods lack collaborative judgment of multiple feature points during a single charge, making it impossible to achieve continuous, stable, and high-precision battery health status monitoring without interrupting the charging process. This makes it difficult to meet the health status monitoring needs of lightweight, real-time, and highly reliable consumer electronics products. Summary of the Invention

[0005] The purpose of this invention is to provide a method for monitoring the capacity decay of secondary batteries in consumer products, so as to solve the problems mentioned in the background art.

[0006] A method for monitoring the capacity degradation of secondary batteries in consumer products includes: Each time the battery is charged and its state of charge reaches the initial threshold, the current charging process is paused, and a symmetrical triangular wave current is applied to the battery. This current first rises linearly to a preset positive current value and then falls linearly to zero, with the rise and fall times being equal. The rise and fall times are determined based on the battery's rated capacity C and dynamic internal resistance Rdyn, specifically: rise time = k × (Rdyn × C), where k is a constant between 0.01 and 0.1, and both the rise and fall times are not less than 50 milliseconds and not more than 2000 milliseconds. During the application of this current, the battery terminal voltage is continuously sampled to obtain a voltage sequence; Find the sampling time corresponding to the maximum voltage value in the voltage sequence; Obtain the transition time between the rising and falling phases of the symmetrical triangular wave current; Compare the sampling time corresponding to the maximum voltage value with the handover time to calculate the lag time length; The length of the current lag time is compared with the initial lag time in the initial healthy state and the historical lag time of the most recent preset number of charging cycles to determine the degree of increase in the lag time. The battery health status information is output based on the degree of growth.

[0007] By actively applying a symmetrical triangular wave current and extracting the lag time between the voltage peak and the current waveform transition, the battery capacity decay is transformed from traditional amplitude measurement to time-domain relative position measurement. This lag time is directly related to aging mechanisms such as diffusion rate decrease and SEI film thickening and exhibits a stable and monotonic relationship. At the same time, the symmetrical waveform itself forms a common-mode interference self-reference, which is naturally immune to temperature drift, contact potential, etc. Monitoring can be completed within a fixed window of each charge without complete charge and discharge.

[0008] In some possible implementations, before applying the symmetrical triangular wave current, the battery management system first performs a short-time pulse test to adjust the positive current value of the symmetrical triangular wave current, specifically including: The battery management system controls the charger to output a short pulse current of fixed duration; The battery management system collects the battery terminal voltage at the end of the short-time pulse current. The battery management system reads the battery open-circuit voltage before the short-time pulse current begins. The battery management system calculates the voltage difference between the terminal voltage and the open-circuit voltage; The battery management system divides the voltage difference by the amplitude of the short-time pulse current to obtain the dynamic internal resistance value of the battery. The battery management system adjusts the positive current value of the symmetrical triangular wave current by the same ratio according to the ratio between the dynamic internal resistance value and the preset standard internal resistance value, so that the product of the adjusted positive current value and the dynamic internal resistance value is equal to the preset constant power value.

[0009] By adopting the above scheme, the dynamic internal resistance of the battery is obtained in real time through short-time pulse testing, and the amplitude of the triangular wave current is adjusted according to the ratio of the internal resistance to the standard internal resistance to keep the excitation power constant. This ensures that batteries with different aging levels produce polarization response under the same electrical stress, and the change in hysteresis time is only caused by capacity decay. This eliminates the influence of internal resistance differences on the measurement and achieves consistency of detection sensitivity throughout the entire life cycle.

[0010] In some possible implementations, after applying the symmetrical triangular wave current, the battery management system also performs the following steps: The battery management system waits for a resting time equal to the total duration of the symmetrical triangular wave current. The battery management system controls the charger to apply a current waveform to the battery that is opposite in phase to the symmetrical triangular wave current. This opposite current waveform first decreases and then increases. The current at the beginning of the decreasing phase is zero and the current at the end is the preset positive current value. The current at the beginning of the increasing phase is the preset positive current value and the current at the end is zero. The duration of the decreasing phase and the increasing phase are equal. During the application of the current waveform with the opposite phase, the battery management system acquires the battery terminal voltage and finds the sampling time corresponding to the minimum voltage value. The battery management system obtains the transition time between the falling and rising phases of the current waveform with opposite phases. The battery management system calculates the lead time length between the sampling time corresponding to the minimum voltage value and the handover time. The battery management system compares the lag time length with the lead time length and takes the larger of the two as the final lag time length for subsequent degradation judgment.

[0011] By adopting the above scheme, a positive triangular wave is first applied, and then a triangular wave with opposite phase is applied after resting. The lag time of the polarization establishment process and the lead time of the polarization decay process are obtained respectively. The larger of the two values ​​is taken as the final feature. This can effectively offset the initial polarization imbalance caused by the different states of charge and resting time of the battery before the test, greatly improve the repeatability and robustness of the measurement, and avoid misjudgment caused by polarization state fluctuations in a single measurement.

[0012] In some possible implementations, the battery management system dynamically changes the rise and fall times of the symmetrical triangular wave current during the next charge based on the most recently measured final lag time, specifically including: When the most recently measured final lag time is greater than 0.8 times the current rise time, the battery management system will increase the rise and fall times of the symmetrical triangular wave current for the next charge by a fixed step. When the most recently measured final lag time is less than 0.2 times the current rise time, the battery management system will reduce the rise and fall times of the symmetrical triangular wave current for the next charge by a fixed step. The battery management system will use the adjusted duration as the new fixed duration for subsequent charging until the ratio of the lag time to the rise time remains between 0.2 and 0.8 times in three consecutive charging cycles.

[0013] Using the above scheme, the rise and fall durations of the next excitation are dynamically adjusted based on the most recently measured lag time. When the lag time is close to 0.8 times the current duration, the duration is extended; when it is less than 0.2 times, the duration is shortened, until the ratio of lag time to rise duration stabilizes between 0.2 and 0.8 times in three consecutive charging cycles. This ensures that the voltage peak always falls within the excitation window throughout the entire aging cycle, avoiding signal loss in the later stages of aging and preventing efficiency waste caused by excessively long excitation in the early stages, thus achieving a dynamic balance between detection accuracy and efficiency.

[0014] In some possible implementations, the battery management system coordinates the determination of the current lag time with the current decay rate during the constant-voltage charging phase that naturally occurs during charging, specifically including: After completing the symmetrical triangular wave current measurement and obtaining the final lag time length, the battery management system continues to execute the normal charging process. When the charging enters the constant voltage stage, the battery management system records the current value within the first preset time window after the start of the constant voltage stage and calculates the rate of decrease of the current value within the window. The battery management system compares the final hysteresis length with the descent rate: When the final hysteresis time length increases and the rate of decrease decreases, the battery management system determines that the capacity decay is normal electrochemical aging. When the final lag time increases while the descent rate remains unchanged or increases, the battery management system determines that there is an abnormal fault and outputs a warning signal. When the final lag time remains constant while the rate of decrease decreases, the battery management system determines that the battery has reached the end of its lifespan and outputs a prompt message to replace the battery.

[0015] By employing the above scheme, the length of the hysteresis time and the current decay rate during the constant voltage charging stage are analyzed together. Based on different combinations of the changing directions of the two, normal electrochemical aging, internal abnormal faults, and the end of the lifespan can be accurately distinguished. That is, an increase in hysteresis time and a decrease in decay rate indicate normal aging, an increase in hysteresis time and no change or increase in decay rate indicate abnormal faults, and no change in hysteresis time and a decrease in decay rate indicate the end of the lifespan. This achieves multi-state identification capabilities that cannot be achieved by a single feature, and improves the accuracy of fault warning.

[0016] In some possible implementations, the battery management system dynamically adjusts the execution frequency of subsequent triangular wave current measurements based on the health status assessment results of the three most recent charges, specifically including: The battery management system stores the health status assessment results of the three most recent charges, with each assessment result indicating normal aging, abnormal fault, or end of life. When the results of the last three determinations are all normal aging, the battery management system will change the execution interval of the triangular wave current measurement from every charge to once every three charges. When an abnormal fault or end of life is detected in the three most recent judgment results, the battery management system will revert to performing triangular wave current measurement for each charge. If the results of three consecutive assessments after recovery are all normal aging, the battery management system will readjust the measurement interval to once every three charges.

[0017] Based on the health status judgment results of the three most recent charges, the execution interval of subsequent triangular wave measurements is dynamically adjusted. When the three consecutive measurements are normal aging, the measurement interval is changed from each charge to once every three charges, which greatly reduces system resource consumption and battery cycle loss. Once an abnormal fault or end of life occurs, the measurement is immediately restored to be executed on every charge, ensuring that abnormal states are captured in a timely manner. This achieves intelligent adaptation that saves resources in normal states and responds quickly in abnormal states, thus optimizing the long-term sustainability of the detection strategy.

[0018] In some possible implementations, the battery management system first obtains the current battery temperature before applying the symmetrical triangular wave current, and adjusts the rise and fall durations of the symmetrical triangular wave current based on a comparison between the temperature and a preset standard temperature. Specifically, this includes: The battery management system reads the temperature sensor readings from inside the battery to obtain the current temperature value; When the current temperature is lower than the standard temperature, the battery management system will extend the rise and fall time of the symmetrical triangular wave current proportionally, with the extension ratio being greater as the temperature decreases. When the current temperature is higher than the standard temperature, the battery management system will shorten the rise and fall time of the symmetrical triangular wave current proportionally, and the higher the temperature, the greater the shortening ratio. The battery management system uses the corrected duration as the actual duration of this measurement, and compares the current lag time with the historical lag time at the same temperature after the measurement is completed, in order to eliminate the influence of temperature on the lag time.

[0019] This invention reads the real-time temperature of the battery, proportionally extending the triangular wave duration when the temperature is below the standard temperature and proportionally shortening the duration when the temperature is above the standard temperature. In subsequent comparisons, it only compares with historical data at the same temperature, thereby completely eliminating the influence of temperature on polarization response speed. Even in low or high temperature environments, the hysteresis time can still accurately reflect the degree of battery aging, breaking through the limitation of traditional methods that are only applicable to room temperature and greatly expanding the application scenarios of consumer products.

[0020] In some possible implementations, the battery management system synchronously records the current charging strategy parameters at each measurement and uses only historical data under the same charging strategy parameters when comparing lag time lengths, specifically including: Before applying the symmetrical triangular wave current, the battery management system reads the current charging strategy parameters, including the current value during the constant current charging phase and the charging cutoff voltage value. The battery management system associates and stores the current charging strategy parameters with the measured lag time length. When comparing the current lag time with the historical lag time, the battery management system selects those historical lag time lengths that are the same as the current charging strategy parameters from the historical sequence, and only uses these selected historical data for comparison. When the number of historical data selected is less than the preset minimum number, the battery management system uses the initial lag time length as a benchmark and marks the current charging strategy parameters as the new benchmark state.

[0021] Using the above scheme, the current constant current charging current value and charging cutoff voltage value are recorded before each measurement, and these parameters are associated with the measured lag time and stored. When making historical comparisons, only historical data with the same parameters as the current charging strategy are selected for comparison, thereby eliminating the interference of different charging strategies on the polarization response. When there is insufficient historical data with the same parameters, the current parameters are automatically marked as the new reference state, ensuring that the system can adapt to various charging protocols and guaranteeing the longitudinal consistency of aging trend judgment.

[0022] In some possible implementations, the battery management system also detects whether the battery has been replaced during each charge and clears the historical lag time length sequence when a battery replacement is detected, specifically including: At the start of each charge, the battery management system reads the unique identification code inside the battery and compares the identification code read this time with the identification code stored during the previous charge. When the current identification code is different from the previous identification code, the battery management system determines that the battery has been replaced and deletes all historical data in the historical lag time length sequence. The battery management system stores the lag time length measured during this charge as the new initial lag time length and resets the growth rate corresponding to this lag time length to zero. After clearing historical data, the battery management system resumes performing triangular wave current measurement on each charge until a preset amount of historical data is accumulated again.

[0023] It should be noted that at the start of each charge, the unique identification code inside the battery is read and compared with the previously stored identification code. If an inconsistency is detected, it is determined that the battery has been replaced. The historical lag time series is immediately cleared, and the lag time measured this time is used as the new initial benchmark and the growth rate is reset to zero. At the same time, the measurement is performed on every charge until enough historical data is accumulated. This completely avoids cross-battery data contamination and ensures that the health status assessment is always for the currently used battery. It is suitable for common scenarios such as users replacing batteries or devices swapping batteries.

[0024] In some possible implementations, the battery management system also detects the type of charger used during each charge and adjusts the baseline value of the hysteresis length based on the charger type, specifically including: When charging starts, the battery management system reads the charger's output capability parameters, including maximum output current and maximum output voltage, through a handshake protocol. The battery management system classifies the charger into fast charging type or normal charging type based on the output capability parameters. When the charger type is fast charging, the battery management system multiplies the measured lag time by a preset fast charging correction factor before comparing the lag time length. This fast charging correction factor is less than one. When the charger type is normal charging, the battery management system does not correct for the lag time length; The battery management system uses the corrected lag time length to compare with historical data in order to eliminate the influence of different charger output characteristics on the measurement results.

[0025] By reading the charger's maximum output current and output voltage through the handshake protocol during charging startup, the charger is classified as either fast charging or normal charging type. For fast charging, the measured lag time is multiplied by a preset correction coefficient of less than one to calibrate the system deviation caused by the fast charging output characteristics. For normal charging, no correction is made, thereby eliminating the influence of different charger output characteristics on polarization response measurement. This makes the monitoring results under fast charging and normal charging scenarios uniform and comparable, improving the reliability of detection in complex usage environments.

[0026] The technical solutions provided by the embodiments of this disclosure have at least the following beneficial effects: This invention transforms battery capacity degradation from traditional amplitude measurement to time-domain relative position measurement by actively applying a symmetrical triangular wave current and extracting the lag time between the voltage peak and the current waveform transition. This lag time is monotonically correlated with the degree of degradation and is naturally immune to common-mode interference such as temperature and contact potential. Combined with short-time pulse testing, the excitation amplitude is adaptively adjusted to ensure that batteries of different ages are subjected to the same electrical stress. Initial polarization imbalance is eliminated by bidirectional excitation with opposite phases. The triangular wave duration is dynamically adjusted to ensure that the signal is measurable throughout the entire life cycle. It is coordinated with the current degradation rate in the constant voltage stage to accurately distinguish between normal aging, abnormal faults, and end-of-life. At the same time, it integrates temperature compensation, charging strategy parameter matching, battery replacement identification, charger type correction, and adaptive measurement frequency adjustment. Thus, it achieves stable, accurate, and online monitoring of the capacity degradation of secondary batteries in consumer products without relying on complete charge and discharge cycles or interfering with normal user operation. Attached Figure Description

[0027] Figure 1 This is a schematic diagram of the method of the present invention. Detailed Implementation

[0028] The technical solution of the present invention will be clearly and completely described below with reference to the embodiments. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0029] Please see Figure 1 This application provides a method for monitoring the capacity degradation of secondary batteries in consumer products, comprising: Each time the battery is charged and its state of charge reaches the initial threshold, the battery management system first controls the charger to pause the current constant current or constant voltage charging output. Then, the charger applies a symmetrical triangular wave current to the battery independently. This current first rises linearly to a preset positive current value and then falls linearly to zero, with the rise and fall times being equal. After the excitation ends, the battery management system resumes the original charging process.

[0030] It should be noted that the core electrochemical mechanism of battery aging is manifested in the thickening of the SEI film, the decrease in lithium-ion diffusion rate, and the increase in charge transfer impedance. These changes directly lead to a slower response speed of the battery to current excitation.

[0031] Based on this mechanism, this invention chooses to capture aging signals through timing characteristics rather than amplitude characteristics, and the design of symmetrical triangular wave current is precisely to amplify this time domain difference.

[0032] Specifically, the state of charge (SOC) is the ratio of the battery's current remaining capacity to its rated capacity. It is obtained by the battery management unit through real-time acquisition of terminal voltage, charging current, and temperature, using an ampere-hour integration method combined with open-circuit voltage lookup table correction. The integration time step is 100 milliseconds, the open-circuit voltage lookup table interval is 1% SOC, and the output result is a percentage value, rounded to one decimal place.

[0033] The initial threshold, serving as the excitation trigger point, is selected during the charging plateau period to avoid interference from abnormal polarization. The initial threshold is set according to the type of battery cathode material: for lithium cobalt oxide batteries or ternary lithium batteries, the initial threshold is set to 40%~60%; for lithium iron phosphate batteries, the initial threshold is set to 70%~80%. In this embodiment, a lithium cobalt oxide battery is used as an example, and it is fixed at 50%. The battery management unit compares the real-time state of charge value with the initial threshold every 100 milliseconds. When the real-time value is greater than or equal to 50%, an excitation signal is immediately triggered and output.

[0034] By applying excitation during the charging plateau period, baseline interference caused by drastic changes in charging voltage can be avoided, resulting in a purer polarization response and facilitating accurate extraction of timing features. This design not only avoids interference but also overcomes the limitation of existing technologies requiring separate time for battery testing through synchronous detection during charging. It achieves health status monitoring without extending the total charging time or affecting the user experience, thus improving the practicality and user acceptance of the detection.

[0035] The battery management system also detects whether the battery has been replaced during each charge, and clears historical data and re-establishes the baseline when battery replacement is detected.

[0036] At the start of each charge, the battery management system (BMS) reads the unique identifier stored inside the battery and compares it with the identifier stored during the previous charge. This unique identifier is a unique identifier permanently stored in the battery's internal storage unit at the factory. Each identifier corresponds one-to-one with a single battery cell, serving as a unique basis for distinguishing different batteries and ensuring the uniqueness and accuracy of battery identification. The BMS reads the unique identifier from the battery's internal storage unit via the SMBus communication protocol. The identifier is stored at a fixed byte address and read as a hexadecimal string.

[0037] When the current identification code is inconsistent with the previous identification code, the battery management system determines that the battery has been replaced and deletes all historical data in the historical lag time sequence. It should be noted that there are fundamental differences in the initial polarization characteristics, factory parameters and historical aging states of different batteries. If the original historical data is used for comparison, it will lead to serious deviations in the health status judgment. Therefore, historical data must be completely cleared when the battery replacement is detected to avoid cross-battery data interference.

[0038] The battery management system stores the lag time length measured during this charge as the new initial lag time length and resets the growth rate corresponding to this lag time length to zero. Using the lag time length measured for the first time for a new battery as a new benchmark can establish a comparison basis that fits the actual state of the current battery for subsequent health status assessments, ensuring that the calculation of the growth rate starts from the initial state and avoiding distortion of assessment results due to benchmark misalignment.

[0039] After clearing historical data, the battery management system resumes performing triangular wave current measurement on every charge until a preset amount of historical data is accumulated again. Understandably, the preset amount is set to three sets. During the period when new batteries lack historical data, maintaining a high frequency of detection on every charge can quickly accumulate effective data of the battery under the current usage environment and charging strategy, ensuring that the health status assessment can be restored to stability and reliability as soon as possible, while also connecting with the logic of subsequent dynamic adjustment of measurement frequency.

[0040] By identifying the battery's unique identifier during the charging start-up phase to determine if the battery has been replaced, and promptly clearing historical data, re-establishing the initial baseline, and maintaining high-frequency detection after replacement, interference from historical data across batteries can be avoided. This ensures that the health status assessment always matches the actual state of the currently used battery, improving the system's adaptability and accuracy in battery replacement scenarios. Furthermore, this approach overcomes the limitation of existing technologies that assume the battery remains unchanged. It not only solves the problem of misjudgment after battery replacement but also, through the combined strategy of rapid baseline reconstruction and high-frequency detection, enables the system to quickly restore stable and accurate assessment capabilities shortly after battery replacement.

[0041] The battery management system also detects the type of charger used during each charge and adjusts the baseline value of the lag time length according to the charger type.

[0042] During charging startup, the battery management system (BMS) reads the charger's output capability parameters, including maximum output current and maximum output voltage, via a handshake protocol. This handshake protocol is a standard communication process between the BMS and the charger, allowing direct access to the charger's rated output parameters without additional hardware detection, ensuring convenient and accurate parameter reading. The BMS categorizes the charger based on whether the maximum output current is greater than or equal to 3A and the maximum output voltage is greater than or equal to 5V.

[0043] The battery management system classifies chargers into fast charging or normal charging types based on their output capability parameters. It should be noted that a charger is classified as fast charging when the maximum output current is greater than or equal to 3A and the maximum output voltage is greater than or equal to 5V; otherwise, it is classified as normal charging. This classification rule is pre-stored in the non-volatile memory of the battery management system and can be flexibly adjusted according to the charging protocols of different electronic products to ensure compatibility with actual charging scenarios. After classification, the battery management system marks and caches the current charger type for subsequent lag time correction judgments.

[0044] When the charger type is fast charging, the battery management system multiplies the measured lag time by a preset fast charging correction factor before comparing the lag time. This fast charging correction factor is less than one. It is understandable that fast charging chargers have a faster output current rise rate and different output ripple characteristics than ordinary chargers, which will cause the initial establishment speed of the battery polarization response to be slightly faster, resulting in a relatively smaller measured lag time.

[0045] The fast charging correction factor is dynamically selected based on the charger protocol type: 0.82 for QC protocol, 0.85 for PD protocol, and 0.88 for VOOC protocol. The correction factor for each protocol is obtained through pre-calibration tests in the laboratory on chargers using different fast charging protocols. Specifically, the lag time is measured on the same battery using both a standard charger and chargers for each fast charging protocol. The measurement value under the standard charger is divided by the measurement value under the fast charger to obtain the correction factor, which is stored in a lookup table of the battery management system. If a fast charging protocol not found in the lookup table is detected, 0.85 is used by default as the correction factor.

[0046] When the charger type is normal charging, the battery management system does not correct for the lag time. It should be noted that the output characteristics of normal charging chargers are relatively stable and consistent with the benchmark test conditions preset by the battery management system. Therefore, no additional correction is required, and the original measurement value can be directly retained to ensure comparability with historical data.

[0047] The battery management system compares the corrected lag time with historical data to eliminate the influence of different charger output characteristics on the measurement results. By adopting the above classification correction method, the lag time of different charger types can be unified to the same benchmark system, ensuring that the health status assessment is based on a consistent judgment standard regardless of whether the user uses a fast charger or a regular charger. This avoids misjudgment of aging trends due to charger replacement and improves the reliability of the system in complex usage scenarios.

[0048] Identify the charger type through the handshake protocol in the charging startup stage, and correct the length of the hysteresis time for fast charging types, which can eliminate the interference of different charger output characteristics on the polarization response measurement, ensure the consistency and accuracy of the detection results in cross-charger usage scenarios, and further improve the practicality and adaptability of the system.

[0049] Before applying a symmetric triangular wave current, the battery management system first obtains the current temperature of the battery, and corrects the rise and fall durations of the symmetric triangular wave current according to the comparison result between the temperature and the preset standard temperature.

[0050] The battery management system reads the value of the temperature sensor inside the battery to obtain the current temperature value; it can be understood that the temperature sensor is directly set on the surface of the battery cell or inside the module, which can collect temperature data reflecting the working state of the battery body in real time and avoid interference from ambient temperature fluctuations on the measurement results.

[0051] When the current temperature value is lower than the standard temperature, the battery management system extends the rise and fall durations of the symmetric triangular wave current in the following proportions: The standard temperature is set at 25°C; if the current temperature T satisfies 10°C ≤ T < 25°C, the duration extension ratio is (25 - T) × α, where α is the temperature correction coefficient. For lithium cobalt oxide batteries, α = 2% / °C, for lithium iron phosphate batteries, α = 1.2% / °C, and for ternary batteries, α = 1.5% / °C; if T < 10°C, the duration extension ratio is fixed at 30%. The extended duration = original duration × (1 + extension ratio / 100); it should be noted that at low temperatures, the migration rate of lithium ions inside the battery decreases, and the polarization establishment and dissipation processes become slower. By extending the excitation duration proportionally, it can ensure that the polarization process is fully completed and prevent the measurement result of the hysteresis time from being too small due to insufficient excitation duration.

[0052] The temperature correction coefficient is obtained by performing constant temperature tests on batteries of the same model within the range of -10°C to 45°C and fitting the relationship curve between the polarization response hysteresis time and temperature, and is pre-stored in the battery management system.

[0053] When the current temperature value is higher than the standard temperature, the battery management system shortens the rise and fall durations of the symmetric triangular wave current in the following proportions: If the current temperature T satisfies 25°C < T ≤ 45°C, the duration shortening ratio is (T - 25) × 1.5%, that is, it shortens by 1.5% for every 1°C increase; if T > 45°C, the duration shortening ratio is fixed at 30%. The shortened duration = original duration × (1 - shortening ratio / 100); It can be understood that at high temperatures, the migration rate of lithium ions inside the battery increases, and the polarization establishment and dissipation processes become faster. By shortening the excitation duration proportionally, it can avoid excessive polarization development and ensure the comparability of the measurement results with those at the standard temperature.

[0054] The battery management system uses the corrected duration as the actual duration of this measurement, and compares the current lag time with the historical lag time at the same temperature after the measurement is completed to eliminate the influence of temperature on the lag time. By adopting the above method, the measurement results at different temperatures can be unified to the same comparison benchmark, eliminating the systematic error caused by temperature changes, and ensuring that the lag time can truly reflect the changes in the battery's own health status.

[0055] By correcting the rise and fall times based on real-time temperature before applying triangular wave excitation and comparing it with historical data at the same temperature, the influence of temperature on polarization response timing measurement can be eliminated, improving the consistency and accuracy of detection results under different temperature environments.

[0056] The battery management system synchronously records the current charging strategy parameters during each measurement and uses only historical data under the same charging strategy parameters when comparing lag time lengths.

[0057] Before applying the symmetrical triangular wave current, the battery management system reads the current charging strategy parameters, including the current value during the constant current charging phase and the charging cutoff voltage value. It is understandable that different constant current charging currents and different cutoff voltages will change the degree of polarization and response speed inside the battery, directly affecting the measurement results of the hysteresis time. Therefore, it is necessary to complete the parameter reading before applying the excitation to ensure that the parameters strictly correspond to the measurement.

[0058] The battery management system associates and stores the current charging strategy parameters with the measured lag time. It should be noted that associative storage means writing the constant current value, charging cut-off voltage value, measured lag time, temperature, and timestamp into the same storage entry to ensure that the same measurement conditions can be completely matched during subsequent screening.

[0059] When comparing the current lag time with historical lag time, the battery management system selects those historical lag times that are identical to the current charging strategy parameters from the historical sequence and uses only these selected historical data for comparison. By comparing historical data under the same charging strategy parameters, polarization interference caused by differences in charging strategies can be eliminated, so that changes in lag time only reflect changes in battery health status, thus improving the reliability of the comparison results.

[0060] When the number of selected historical data is less than the preset minimum, the battery management system uses the initial lag time as the benchmark and marks the current charging strategy parameters as the new benchmark state. It can be understood that the preset minimum number is set to three sets. When there are less than three sets of historical data, the factory initial value is directly used as the benchmark, and the current charging strategy is marked as the new benchmark to ensure that subsequent measurements still have a stable comparison basis.

[0061] By synchronously recording and matching charging strategy parameters during measurement, and comparing lag times only under the same charging conditions, the interference of charging strategy differences on measurement results can be eliminated, further improving the stability and accuracy of battery health status assessment. Not only does it eliminate interference from charging strategies, but through parameter association storage and automatic new benchmark labeling, the system can adapt to the charging protocols of different electronic products, automatically matching without manual configuration.

[0062] Before applying the symmetrical triangular wave current, the battery management system first performs a short-time pulse test to adjust the positive current value of the symmetrical triangular wave current, specifically including: The battery management system controls the charger to output a short pulse current of a fixed duration.

[0063] For example, the short-time pulse current is a constant DC current with an amplitude set to 0.05C. The duration is determined based on the battery's equivalent time constant τ, τ = Rdyn × Cdl, where Cdl is an estimated value of the battery's double-layer capacitance (for consumer product batteries, Cdl is typically in the range of 0.1F to 10F). The fixed duration is no greater than 0.1 × τ, and no less than 10 milliseconds and no more than 100 milliseconds.

[0064] In this embodiment, for a typical mobile phone battery, τ is approximately 500 milliseconds, therefore the fixed duration is set to 50 milliseconds. This pulse has a small amplitude and short duration, will not change the battery's state of charge, and will not have a cumulative effect on subsequent excitations.

[0065] Using short-duration, low-amplitude pulses for internal resistance testing allows for rapid acquisition of the battery's current internal resistance without significantly disturbing its state, ensuring the continuity and stability of subsequent testing procedures. Furthermore, it not only achieves accurate internal resistance measurement but also avoids damage to the battery's SEI film from high-current pulses by selecting small-amplitude, short-duration pulse parameters, thus considering both testing and battery cycle life.

[0066] The battery management system collects the battery terminal voltage at the moment the short-time pulse current ends.

[0067] It should be understood that short-time pulse current has a fixed start time and duration, and its end time can be directly determined by the timing unit inside the battery management system without the need for external detection.

[0068] In this embodiment, the duration of the short-time pulse current is 50 milliseconds. The timing unit starts timing from the pulse output start time, and the pulse ends when the timing value reaches 50 milliseconds. At the pulse end time, the battery management system performs a single sampling of the voltage across the positive and negative terminals of the battery through the analog-to-digital conversion module, uses the sampled voltage value as the terminal voltage after pulse loading, and stores it in a designated cache unit.

[0069] Accurately acquiring the terminal voltage at the end of the pulse can capture the voltage drop generated by the pulse current across the battery's internal resistance to the greatest extent, avoiding measurement errors caused by rapid polarization decay and improving the accuracy of internal resistance calculation.

[0070] The battery management system reads the battery open-circuit voltage before the short-time pulse current begins.

[0071] It should be noted that the open-circuit voltage is the battery's terminal voltage when there is no external current injection and the internal polarization is completely stable. Before outputting the short-time pulse current, the battery management system controls the charger to maintain zero output current for 100 milliseconds, allowing the battery's internal polarization to dissipate fully and the voltage to reach a stable state. The battery management system samples the battery terminal voltage at the last moment before the pulse output, identifies this voltage value as the battery's open-circuit voltage, and caches it in a different storage unit than the aforementioned terminal voltage to avoid data overwriting.

[0072] Allowing the battery to settle to a stable state beforehand and collecting the open-circuit voltage can provide an accurate benchmark for subsequent voltage difference calculations and eliminate the influence of historical polarization residues on this test.

[0073] The battery management system calculates the voltage difference between the terminal voltage and the open-circuit voltage.

[0074] Specifically, the voltage difference is used to characterize the voltage drop across the battery's internal resistance caused by the pulse current. The battery management system retrieves the terminal voltage at the end of the pulse and the open-circuit voltage before the pulse stored in the cache unit, and calculates the voltage difference by subtracting the open-circuit voltage from the terminal voltage. This voltage difference is generated jointly by the battery's ohmic internal resistance and polarization internal resistance under the action of a short-time pulse current, and does not include common-mode interference components such as temperature drift and contact potential.

[0075] By subtracting the terminal voltage from the open-circuit voltage, the influence of the battery electromotive force can be directly eliminated, retaining only the voltage drop component caused by the internal resistance, thus providing clean data for dynamic internal resistance calculation.

[0076] The battery management system divides the voltage difference by the amplitude of the short-time pulse current to obtain the dynamic internal resistance of the battery.

[0077] Specifically, the dynamic internal resistance is the ratio of the voltage difference to the short-time pulse current amplitude. The calculation formula is: Dynamic internal resistance = Voltage difference ÷ Short-time pulse current amplitude. The result is in milliohms and rounded to the nearest integer. This dynamic internal resistance value reflects the sum of the battery's current ohmic impedance and polarization impedance in real time, providing a basis for subsequent amplitude adjustments.

[0078] By calculating the dynamic internal resistance in real time, the actual impedance of the battery under the current temperature, aging state, and charging state can be accurately reflected, rather than using the assumption of fixed internal resistance, thus improving the matching degree of subsequent excitation parameters.

[0079] The battery management system adjusts the positive current value of the symmetrical triangular wave current by the same ratio according to the ratio between the dynamic internal resistance value and the preset standard internal resistance value, so that the product of the adjusted positive current value and the dynamic internal resistance is equal to the preset constant power value.

[0080] Specifically, the preset standard internal resistance value is the DC internal resistance measured after discharging the battery at 0.1C current for 10 seconds under the conditions of 100% health state, temperature 25°C, and 50% state of charge. This value is measured once by the battery management system at the factory and stored in non-volatile memory. The preset constant power value is a fixed excitation power preset by the system, which is set to 0.5W in this embodiment (for mobile phone batteries) or scaled proportionally according to the battery's rated capacity. The battery management system divides the standard internal resistance value by the dynamic internal resistance value to obtain a scaling factor, and then multiplies the initial triangular wave amplitude value by this scaling factor to obtain the adjusted positive current value. In this way, it can be ensured that batteries with different aging levels are tested under the same excitation power, avoiding excessively large or small voltage responses due to differences in internal resistance, and improving the comparability of timing characteristics.

[0081] By adaptively adjusting the excitation amplitude based on dynamic internal resistance, polarization responses can be generated at similar excitation power regardless of battery age or internal resistance, making the timing characteristics of batteries in different states directly comparable. Furthermore, not only does it achieve adaptive matching of excitation parameters, but the constant power excitation design also avoids detection failures caused by applying excessively high voltage to aging batteries or excessively low current to new batteries. This ensures the system maintains stable detection sensitivity throughout the entire battery lifespan, extending the effective application period of the technology.

[0082] The symmetrical triangular wave current is a unidirectional linearly changing signal, generated as follows: The microcontroller inside the battery management system outputs a voltage-type triangular wave reference signal through a digital-to-analog converter module. This signal is converted into a corresponding triangular wave charging current by a voltage-controlled current source circuit and superimposed on the constant current output of the charger; alternatively, the battery management system directly controls the charger's digital control interface to update the current setpoint every millisecond, causing the charger's output current to rise and then fall linearly. The current change slope during the rising phase is equal to the preset positive current value divided by the rising time, and the current change slope during the falling phase is equal to the negative preset positive current value divided by the rising time. In this embodiment, both the rising and falling times are set to 200 milliseconds, with a total excitation time of 400 milliseconds. The positive current value, adjusted through short-time pulse testing, ensures stable excitation power, effectively stimulating a voltage response without perceptibly affecting the normal battery charging process.

[0083] During the application of this current, the battery terminal voltage is continuously sampled to obtain a voltage sequence.

[0084] Understandably, the extraction of time-series features depends on the complete voltage response curve, so the data acquisition must strictly cover the entire excitation period.

[0085] In this embodiment, voltage acquisition is performed by the analog-to-digital conversion module of the battery management unit, with a sampling frequency set to 1 kHz, meaning sampling occurs every millisecond. The acquisition start signal is triggered synchronously with the triangular wave current output signal, and the acquisition stop signal is triggered synchronously with the triangular wave zero-reset signal.

[0086] During each data acquisition, the analog-to-digital converter converts the analog voltage signal into a 16-bit digital signal, reads the current time from the timing unit with the current start time as the zero point, and stores the digital voltage value and the key-value pair of the sampling time in the buffer.

[0087] After the data acquisition is completed, the data in the buffer are sorted in ascending order of the sampling time to form a voltage sequence. The sequence length is fixed at 401 data points, corresponding to a 400-millisecond excitation duration and a 1 kilohertz sampling frequency.

[0088] High-frequency synchronous sampling can completely restore the continuous process of voltage change with current excitation, avoid peak position shift due to sparse sampling, and ensure the accuracy of time series feature extraction.

[0089] Find the sampling time corresponding to the maximum voltage value in the voltage sequence.

[0090] In one implementation, a traversal comparison method is used to locate the peak value. First, all digital voltage values ​​are extracted from the voltage sequence, the maximum value variable is initialized to the first voltage value, and the sampling time corresponding to that voltage value is recorded synchronously.

[0091] Then, following the sequence, each subsequent voltage value is compared with the maximum value variable. If the current voltage value is greater than the maximum value variable, the maximum value variable is updated to the current voltage value, and the corresponding sampling time is updated synchronously. After traversing all 401 data points, the maximum voltage value is obtained.

[0092] Then, the voltage sequence is traversed again, and all sampling times when the voltage value is equal to the maximum value are recorded to form a set of times. If there is only one time in the set, then that time is taken as the sampling time corresponding to the maximum voltage value; if there are multiple times, then the median of these times (the median value is taken when the number is odd, and the average of the two median values ​​is taken when the number is even) is taken as the sampling time corresponding to the maximum voltage value.

[0093] The target time is set to zero at the current initiation point, in milliseconds, and accurate to 1 millisecond. Essentially, it is the time node when the battery polarization response reaches its peak, and it is the core feature point for capturing differences in battery aging.

[0094] By traversing and comparing to determine the peak voltage time, the time point of strongest polarization response can be located stably and reliably, providing an accurate basis for lag time calculation.

[0095] Obtain the transition time between the rising and falling phases of the symmetrical triangular wave current.

[0096] It should be understood that the transition moment between the rising and falling phases refers to the instant when the symmetrical triangular wave current changes from a linear rise to a linear fall. This moment is also the moment when the symmetrical triangular wave current reaches the preset positive current value, which is an inherent node of the excitation waveform.

[0097] The handover time can be directly determined by the excitation output timing, without the need for voltage or current sampling detection. In this embodiment, the rise time of the symmetrical triangular wave current is preset to two hundred milliseconds. The timing unit starts timing from the current initiation time, and when the timing value reaches two hundred milliseconds, it is the handover time between the rise and fall phases.

[0098] The battery management unit directly reads the current value from the timing unit to obtain the handover time. This time is set with the current initiation time as the zero point, in milliseconds, and with an accuracy of 1 millisecond. This handover time is used for subsequent comparison with the sampling time corresponding to the maximum voltage value, serving as a built-in benchmark for determining whether the voltage response is lagging.

[0099] By directly using the inherent nodes of the excitation waveform as the reference time, there is no need to rely on detection and calculation, it is not affected by noise, and the reference stability is extremely high, which can improve the reliability of the lag time calculation.

[0100] Compare the sampling time corresponding to the maximum voltage value with the handover time to calculate the lag time length.

[0101] Furthermore, the hysteresis time refers to the time offset between the sampling time corresponding to the maximum voltage and the transition time, used to characterize the delay in the battery polarization response. The hysteresis time is obtained by subtracting the transition time from the sampling time corresponding to the maximum voltage, and the result is calculated in milliseconds, rounded to the nearest integer.

[0102] If the calculated result is less than or equal to zero milliseconds, it indicates that the voltage peak does not lag behind the current peak, and the battery polarization response speed is within the normal range. If the calculated result is greater than zero milliseconds, then this value is the actual lag time of the voltage response relative to the current excitation. The longer the lag time, the slower the polarization establishment speed inside the battery, and the more obvious the degradation of electrochemical kinetic characteristics.

[0103] By directly calculating the time difference between the peak moment and the baseline moment, the abstract polarization response speed can be transformed into a specific, quantifiable time value, enabling precise differentiation of battery aging levels. This effect not only achieves quantitative differentiation of aging levels but also, through the intuitive indicator of time difference, allows the system to identify latent battery degradation at an early stage. These latent degradations are difficult to detect in traditional capacity testing, thus enabling early warning of battery failures and reducing the safety risks caused by sudden battery failures.

[0104] After applying a symmetrical triangular wave current, the battery management system also performs the following steps: The battery management system waits for a settling time equal to the total duration of the symmetrical triangular wave current.

[0105] It should be noted that the resting time is used to eliminate residual polarization from the previous excitation and to avoid interference between the responses of the two excitations. In this embodiment, the total duration of the symmetrical triangular wave is four hundred milliseconds, so the resting time is also set to four hundred milliseconds. During the resting period, the battery management system controls the charger output current to remain at zero and continuously monitors the battery terminal voltage until the voltage fluctuation between two adjacent samples is less than 1 millivolt, confirming that the polarization is completely stable.

[0106] Setting a settling time that matches the excitation duration ensures that the polarization generated by the previous excitation completely fades, allowing the reverse excitation to obtain an independent and clean response signal, and avoiding mutual interference between the two waveforms.

[0107] The battery management system controls the charger to apply a current waveform to the battery that is opposite in phase to the symmetrical triangular wave current. This opposite current waveform first decreases and then increases. During the decreasing phase, the starting current is zero and the ending current is a preset positive current value. During the increasing phase, the starting current is a preset positive current value and the ending current is zero. The duration of the decreasing phase and the increasing phase are equal.

[0108] Specifically, the reverse waveform has the same parameters as the forward triangular wave, only the timing structure is reversed. It is still a forward charging current and does not include a negative discharge segment. The fall time and rise time are both 200 milliseconds, and the total duration is still 400 milliseconds. The preset positive current value is the same as the value adjusted by dynamic internal resistance to ensure that the excitation power is consistent in the two excitations.

[0109] By employing bidirectional excitation with opposite phases and identical parameters, temporal features can be obtained from both polarization establishment and polarization de-polarization directions, making the detection results more comprehensive and suppressing random errors that may occur in a single direction.

[0110] The innovation of this invention lies in the fact that it not only suppresses random errors, but also distinguishes between reversible and irreversible polarization degradation of the battery through the design of bidirectional polarization feature fusion, avoiding misjudgment caused by reversible fluctuations and greatly improving the anti-interference ability of health status assessment.

[0111] During the application of current waveforms with opposite phase, the battery management system acquires the battery terminal voltage and finds the sampling time corresponding to the minimum voltage value.

[0112] Understandably, the acquisition method is consistent with the positive excitation, with a sampling frequency of 1 kHz, synchronous triggering and synchronous termination, forming a voltage sequence of equal length. A traversal comparison method is used to find the minimum voltage value, and its corresponding sampling time is recorded. This time reflects the location of the voltage valley during the polarization decay process.

[0113] Maintaining acquisition parameters that are completely consistent with the positive stimulus can ensure that the two sets of time series features are acquired under the same conditions, thereby improving the comparability and fusion rationality of lagging and leading features.

[0114] The battery management system obtains the transition time between the falling and rising phases of the current waveform with opposite phases.

[0115] It should be understood that this handover moment is the moment when the reverse waveform changes from falling to rising, that is, the moment when the preset positive current value is reached. Since the falling duration is fixed at two hundred milliseconds, the timing unit starts timing from the start of the reverse waveform, and the handover moment is the time when the timing reaches two hundred milliseconds, with an accuracy of 1 millisecond, which serves as the benchmark for judging the lead relationship.

[0116] Using the same benchmark determination method as the positive excitation method can ensure the uniformity and stability of the two time series calculation systems, and avoid systematic errors caused by inconsistent rules.

[0117] The battery management system calculates that the sampling time corresponding to the minimum voltage value precedes the lead time of the handover time.

[0118] Furthermore, the lead time is the advance of the voltage minimum moment relative to the transition moment of the reverse waveform, used to characterize the speed of polarization decay. The calculation formula is: Lead Time = Transition Moment - Sampling Moment Corresponding to Voltage Minimum. The result is a non-negative value in milliseconds, rounded to the nearest integer. A larger value indicates faster polarization decay and more pronounced degradation of dynamic characteristics.

[0119] By calculating the lead time of the voltage valley relative to the waveform reference, the battery dynamics performance can be quantified from the perspective of polarization decay, forming a complementary feature with the hysteresis time obtained from positive excitation.

[0120] The battery management system compares the lag time length with the lead time length and takes the larger of the two as the final lag time length for subsequent degradation judgment.

[0121] In this embodiment, the lag time reflects the delay in polarization establishment, while the lead time reflects the early polarization decay; both are positively correlated with the degree of aging. Taking the larger value as the final feature can enhance the aging signal, suppress single-measurement noise, and improve the stability and consistency of the detection results.

[0122] Selecting the larger value from the two sets of features as the final result can maximize the preservation of the dynamic degradation characteristics caused by battery aging and avoid the true aging state being masked by single noise or local interference.

[0123] The battery management system dynamically adjusts the rise and fall times of the symmetrical triangular wave current during the next charge based on the most recently measured final lag time. Specifically, this includes: When the most recently measured final lag time is greater than 0.8 times the current rise time, the battery management system will increase the rise and fall times of the symmetrical triangular wave current for the next charge by a fixed step.

[0124] Specifically, the battery management system calculates the ratio of the most recently measured final lag time to the current rise time, based on the condition that the final lag time > 0.8 × the current rise time. If this condition is met, the rise time and fall time used in the next charging process are increased by a fixed step size from the current duration. In this embodiment, the fixed step size is set to 50 milliseconds, ensuring that the rise time and fall time remain equal.

[0125] When the most recently measured final lag time is less than 0.2 times the current rise time, the battery management system will reduce the rise and fall times of the symmetrical triangular wave current for the next charge by a fixed step.

[0126] Specifically, the corresponding judgment condition is: the final lag time length < 0.2 × the current rise time. If the condition is met, the rise time and fall time used for the next charge will be reduced by the same fixed step size of 50 milliseconds based on the current duration.

[0127] The battery management system will use the adjusted duration as the new fixed duration for subsequent charging until the ratio of the lag time to the rise time remains between 0.2 and 0.8 times in three consecutive charging cycles.

[0128] By using the final lag time as feedback, the rise and fall durations of the next excitation are adaptively adjusted, ensuring that the timescale of the triangular wave waveform is always precisely matched with the current polarization response speed of the battery. When the battery is highly aged and the polarization response speed is slow, the excitation duration is automatically extended, providing sufficient time for the complete appearance of the voltage peak. This prevents the polarization response from ending before reaching the peak due to insufficient excitation duration, thus preventing the loss of timing features and ensuring the integrity and accuracy of aging detection. When the battery is in good condition and the polarization response speed is fast, the excitation duration is automatically shortened, reducing invalid excitation time without affecting feature extraction. This reduces the impact of the detection process on the overall charging process, improving charging efficiency and user experience.

[0129] The innovation of this invention lies in the fact that it not only achieves adaptive matching of excitation duration, but also avoids frequent fluctuations in excitation parameters and reduces system computational overhead by locking parameters after three stabilizations. At the same time, it ensures the consistency of detection, optimizes system resource usage while improving detection performance, and achieves a balance between performance and efficiency.

[0130] The battery management system will make a combined judgment based on the final lag time and the current decay rate during the constant voltage charging phase that naturally occurs during the charging process.

[0131] After completing the symmetrical triangular wave current measurement and obtaining the final lag time, the battery management system continues the normal charging process. When charging enters the constant voltage stage, the battery management system waits for the first 2 seconds after the constant voltage starts (avoiding the current jump region in the initial stage of constant voltage), and then records the current value within a continuous Twindow seconds from that moment as the first preset time window. Twindow is determined based on the battery's constant voltage charging time constant τcv: τcv is the time required for the current to decrease from the initial value to 37% during the constant voltage stage. Twindow is between 0.5×τcv and 1×τcv, and is not less than 5 seconds and not more than 30 seconds. In this embodiment, for a typical mobile phone battery, τcv is approximately 15 seconds, therefore Twindow is set to 10 seconds.

[0132] The start time of the window is denoted as tstart, and the end time is denoted as tend = tstart + 10 seconds. The rate of decrease of the current value within this window is calculated as: Rate of decrease = (I(tstart) - I(tend)) / 10 seconds, in milliamperes per second. Where I(tstart) is the current value at the start of the window, and I(tend) is the current value at the end of the window. If the calculated rate of decrease is negative (i.e., the current increases instead of decreasing), the absolute value is taken and marked as abnormal.

[0133] By collecting current and calculating the rate of decrease in the early stage of constant voltage, characteristic information reflecting the internal state of the battery can be obtained without interfering with the normal charging process, providing a reliable basis for subsequent collaborative judgment.

[0134] The battery management system compares the final lag time with the rate of capacity decay: when the final lag time increases and the rate of capacity decay decreases, the battery management system determines that the capacity decay is normal electrochemical aging; when the final lag time increases while the rate of capacity decay remains unchanged or increases, the battery management system determines that there is an abnormal fault and outputs a warning signal; when the final lag time remains unchanged while the rate of capacity decay decreases, the battery management system determines that the battery has entered the end of its life and outputs a prompt to replace the battery.

[0135] By combining the polarization response timing characteristics with the current decay rate during the constant voltage stage, it is possible to accurately distinguish between normal electrochemical aging, internal abnormal faults, and end-of-life states, avoiding misjudgments caused by single-feature judgments and further improving the accuracy and reliability of battery health status assessment.

[0136] The innovation of this invention lies in the fact that it first achieves accurate differentiation of multiple states, and then, through cross-dimensional synergy of polarization features and charging behavior features, it can identify complex faults that cannot be detected by a single feature, thus greatly improving the comprehensiveness of fault diagnosis and providing dual protection for the safe use of batteries.

[0137] The battery management system dynamically adjusts the execution frequency of subsequent triangular wave current measurements based on the health status assessment results of the three most recent charges.

[0138] The battery management system (BMS) stores the health status assessment results of the three most recent charges, with each result indicating normal aging, abnormal fault, or end-of-life. When the three most recent assessment results are all normal aging, the BMS changes the interval for triangular wave current measurement from every charge to once every three charges. If any of the three most recent assessment results indicates an abnormal fault or end-of-life, the BMS reverts to performing triangular wave current measurement on every charge. When the three consecutive assessment results after this reversion are again normal aging, the BMS readjusts the measurement interval to once every three charges. To ensure safety, during charging cycles where measurements are skipped, the BMS continues to monitor the battery temperature: if the detected temperature change exceeds ±5°C compared to the temperature at the time of the most recent measurement, or if the cumulative number of skipped measurements reaches 10 (i.e., 30 consecutive charges without measurement), a triangular wave current measurement is forcibly triggered, regardless of the interval adjustment.

[0139] By adaptively adjusting the execution frequency of triangular wave current measurement based on the results of the three most recent health status assessments, the detection frequency and system resource consumption can be reduced when the battery status is stable, and the detection frequency can be increased when the battery status is abnormal to ensure timely fault identification, thus balancing detection accuracy, fault response speed and system operating efficiency.

[0140] The innovation of this invention lies in its ability to optimize system resource usage. Furthermore, through a three-judgment trigger adjustment strategy, it avoids frequent frequency switching caused by a single abnormal judgment, thereby improving the stability of system operation. At the same time, it promptly increases the detection frequency in the early stages of battery deterioration, achieving intelligent adaptation that saves resources in normal conditions and responds quickly in abnormal conditions.

[0141] The final lag time is compared with the initial lag time in the initial healthy state and the historical lag time of the most recent preset number of charging cycles to determine the degree of increase in lag time.

[0142] In this embodiment, a single measurement value is easily affected by instantaneous operating conditions. By comparing it with the baseline value and historical values ​​in multiple dimensions, the stability of aging judgment can be effectively improved.

[0143] The initial health state is defined as the state of the battery after completing its first full charge-discharge cycle after leaving the factory. The initial hysteresis time is measured during factory calibration according to all the test conditions of this embodiment. After the test is completed, it is stored in the non-volatile memory of the battery management unit as a permanent reference value.

[0144] The most recent preset number of times is set to eight. Eight storage units are reserved in the non-volatile memory of the battery management unit to store the lag time length measured during the last eight complete charging processes, using a rolling update mechanism. Each time a new lag time length is measured, the oldest stored historical value is deleted, and the new value is stored in the storage unit.

[0145] In one implementation, the growth rate is calculated using a relative growth rate. The final lag time is subtracted from the initial lag time, then divided by the initial lag time and converted into a percentage. The result is rounded to two decimal places.

[0146] Let's assume another scenario where a linear fitting correction method can be used to reduce the impact of fluctuations in a single measurement.

[0147] First, the historical lag times and corresponding charge counts for the most recent eight measurements are read from non-volatile memory. Using the charge counts as the x-axis and the lag time as the y-axis, a linear fit is performed using the least squares method to obtain the fitted line equation y = kx + b, where k is the slope and b is the intercept. Substituting the current charge count into this equation yields the predicted lag time. The degree of increase is calculated by dividing the difference between the current measurement and the predicted lag time by the predicted lag time and converting the result to a percentage, rounded to two decimal places.

[0148] By quantifying the absolute degree of aging by comparing it with an initial benchmark, smoothing out individual fluctuations with recent historical data, and highlighting trends through linear fitting, the judgment of the degree of growth can be made more stable, accurate, and in line with the actual aging state.

[0149] It should be noted that before comparing the current lag time with historical lag times, all historical data included in the comparison must be preprocessed according to the same temperature correction rules, charging strategy parameter matching rules, and charger type correction rules as the current measurement to ensure a consistent comparison benchmark. Historical data that has not undergone the same preprocessing will not be included in this comparison.

[0150] This invention improves the stability of the growth rate assessment and, through a linear fitting design that highlights the trend, can predict the accelerated aging trend of the battery in advance, rather than quantifying it only after aging occurs. This provides users with more time to prepare for battery replacement and further enhances the user experience.

[0151] Output battery health status information based on the degree of increase.

[0152] For example, the battery management unit's non-volatile memory pre-stores a health status grading rule table. A growth rate below 10% corresponds to a good status, greater than or equal to 10% but less than 30% corresponds to a fair status, and greater than or equal to 30% corresponds to a poor status.

[0153] During the output process, the battery management system queries the grading rule table based on the calculated growth rate to determine the corresponding health status identifier.

[0154] Subsequently, the health status identifier, the final lag time, the growth rate, the duration of the incentive, and the test timestamp are packaged and sent to the main control chip of the electronic product via the communication interface. The main control chip then displays and stores the data, allowing users to query historical records later.

[0155] By synchronously outputting and storing complete test data and health levels, the battery status can be displayed intuitively to users, and complete data support can be provided for subsequent tracking of aging trends and analysis of abnormal situations.

[0156] Specifically, this invention enables data visualization and traceability, and through its multi-dimensional data packaging and storage design, it provides data support for battery fault tracing and batch quality analysis.

[0157] This invention shifts battery state detection from the traditional amplitude value domain to the relative position in the time domain. It directly reflects the changes in lithium-ion diffusion rate, SEI film thickness, and charge transfer impedance through the voltage peak lag time, forming a stable and clear correlation with capacity decay. The detection sensitivity is higher than that of conventional internal resistance or capacity methods.

[0158] The symmetrical structure of the symmetrical triangular wave can form a self-reference within a single excitation, automatically canceling common-mode interference such as temperature, contact potential, and initial voltage drop, so that the detection results only reflect the electrochemical characteristics of the battery itself, and the anti-interference ability is stronger.

[0159] By identifying the battery's unique identifier during the charging start-up phase and determining whether the battery has been replaced, and promptly clearing historical data and re-establishing the initial baseline after replacement, cross-battery data interference can be completely avoided, ensuring that the health status assessment always matches the current battery and significantly improving the system's accuracy in multi-battery alternation scenarios.

[0160] By identifying the charger type and correcting the lag time for fast charging scenarios, detection bias caused by different charger output characteristics can be eliminated, enabling unified evaluation for both fast charging and regular charging scenarios.

[0161] By adaptively correcting the triangular wave excitation duration at temperature and comparing it with historical data at the same temperature, the influence of temperature on polarization timing measurement can be eliminated, enabling the system to maintain high accuracy even at extreme temperatures, thus overcoming the limitation of traditional methods that are only applicable to room temperature.

[0162] By dynamically measuring the internal resistance in real time and adaptively adjusting the excitation amplitude, constant power excitation is achieved throughout the entire life cycle, avoiding detection failure due to changes in internal resistance caused by battery aging and ensuring stable detection sensitivity throughout the entire process.

[0163] By using bidirectional triangular wave excitation in both positive and negative directions, and integrating polarization to establish two types of characteristics—hysteresis and polarization decay leading—it is possible to distinguish between reversible fluctuations and irreversible aging, significantly reducing the misjudgment rate and improving the robustness of the assessment.

[0164] By combining polarization timing characteristics with the current decay rate during the constant voltage stage, it is possible to accurately distinguish between normal aging, internal abnormal faults, and the end of the battery's lifespan, thereby enabling early warning of faults and improving battery safety.

[0165] By adaptively adjusting the detection frequency based on the health status, resource consumption is reduced when the battery is stable and the detection frequency is increased when abnormal, thus achieving a balance between detection accuracy and system efficiency.

[0166] The above description is merely a specific embodiment of this application, but the scope of protection of this application is not limited thereto. Any variations or substitutions that can be easily conceived by those skilled in the art within the scope of the technology disclosed in this application should be included within the scope of protection of this application. Therefore, the scope of protection of this application should be determined by the scope of the claims.

Claims

1. A method for monitoring the capacity decay of secondary batteries in consumer products, characterized in that, include: Each time the battery is charged and the state of charge reaches the initial threshold, a symmetrical triangular wave current is applied to the battery. This current first rises linearly to a preset positive current value and then falls linearly to zero. The rise and fall times are equal. During the application of this current, the battery terminal voltage is continuously sampled to obtain a voltage sequence; Find the sampling time corresponding to the maximum voltage value in the voltage sequence; Obtain the transition time between the rising and falling phases of the symmetrical triangular wave current; Compare the sampling time corresponding to the maximum voltage value with the handover time to calculate the lag time length; The length of the current lag time is compared with the initial lag time in the initial healthy state and the historical lag time of the most recent preset number of charging cycles to determine the degree of increase in the lag time. The battery health status information is output based on the degree of growth.

2. The method for monitoring the capacity decay of secondary batteries in consumer products according to claim 1, characterized in that, Before applying the symmetrical triangular wave current, the battery management system first performs a short-time pulse test to adjust the positive current value of the symmetrical triangular wave current, specifically including: The battery management system controls the charger to output a short pulse current of fixed duration; The battery management system collects the battery terminal voltage at the end of the short-time pulse current. The battery management system reads the battery open-circuit voltage before the short-time pulse current begins. The battery management system calculates the voltage difference between the terminal voltage and the open-circuit voltage; The battery management system divides the voltage difference by the amplitude of the short-time pulse current to obtain the dynamic internal resistance value of the battery. The battery management system adjusts the positive current value of the symmetrical triangular wave current by the same ratio according to the ratio between the dynamic internal resistance value and the preset standard internal resistance value, so that the product of the adjusted positive current value and the dynamic internal resistance value is equal to the preset constant power value.

3. The method for monitoring the capacity decay of secondary batteries in consumer products according to claim 1, characterized in that, After applying the symmetrical triangular wave current, the battery management system also performs the following steps: The battery management system waits for a resting time equal to the total duration of the symmetrical triangular wave current. The battery management system controls the charger to apply a current waveform to the battery that is opposite in phase to the symmetrical triangular wave current. This opposite current waveform first decreases and then increases. The current at the beginning of the decreasing phase is zero and the current at the end is the preset positive current value. The current at the beginning of the increasing phase is the preset positive current value and the current at the end is zero. The duration of the decreasing phase and the increasing phase are equal. During the application of the current waveform with the opposite phase, the battery management system acquires the battery terminal voltage and finds the sampling time corresponding to the minimum voltage value. The battery management system obtains the transition time between the falling and rising phases of the current waveform with opposite phases. The battery management system calculates the lead time length between the sampling time corresponding to the minimum voltage value and the handover time. The battery management system compares the lag time length with the lead time length and takes the larger of the two as the final lag time length for subsequent degradation judgment.

4. The method for monitoring the capacity decay of secondary batteries in consumer products according to claim 3, characterized in that, The battery management system dynamically adjusts the rise and fall times of the symmetrical triangular wave current during the next charge based on the most recently measured final lag time. Specifically, this includes: When the most recently measured final lag time is greater than 0.8 times the current rise time, the battery management system will increase the rise and fall times of the symmetrical triangular wave current for the next charge by a fixed step. When the most recently measured final lag time is less than 0.2 times the current rise time, the battery management system will reduce the rise and fall times of the symmetrical triangular wave current for the next charge by a fixed step. The battery management system will use the adjusted duration as the new fixed duration for subsequent charging until the ratio of the lag time to the rise time remains between 0.2 and 0.8 times in three consecutive charging cycles.

5. The method for monitoring the capacity decay of secondary batteries in consumer products according to claim 4, characterized in that, The battery management system coordinates the length of the lag time with the current decay rate during the constant voltage charging phase that naturally occurs during charging, specifically including: After completing the symmetrical triangular wave current measurement and obtaining the final lag time length, the battery management system continues to execute the normal charging process. When the charging enters the constant voltage stage, the battery management system records the current value within the first preset time window after the start of the constant voltage stage and calculates the rate of decrease of the current value within the window. The battery management system compares the final hysteresis length with the descent rate: When the final hysteresis time length increases and the rate of decrease decreases, the battery management system determines that the capacity decay is normal electrochemical aging. When the final lag time increases while the descent rate remains unchanged or increases, the battery management system determines that there is an abnormal fault and outputs a warning signal. When the final lag time remains constant while the rate of decrease decreases, the battery management system determines that the battery has reached the end of its lifespan and outputs a prompt message to replace the battery.

6. The method for monitoring the capacity decay of secondary batteries in consumer products according to claim 5, characterized in that, The battery management system dynamically adjusts the execution frequency of subsequent triangular wave current measurements based on the health status assessment results of the three most recent charges. Specifically, this includes: The battery management system stores the health status assessment results of the three most recent charges, with each assessment result indicating normal aging, abnormal fault, or end of life. When the results of the last three determinations are all normal aging, the battery management system will change the execution interval of the triangular wave current measurement from every charge to once every three charges. When an abnormal fault or end of life is detected in the three most recent judgment results, the battery management system will revert to performing triangular wave current measurement for each charge. If the results of three consecutive assessments after recovery are all normal aging, the battery management system will readjust the measurement interval to once every three charges.

7. The method for monitoring the capacity decay of secondary batteries in consumer products according to claim 6, characterized in that, Before applying the symmetrical triangular wave current, the battery management system first obtains the current battery temperature and, based on a comparison with a preset standard temperature, adjusts the rise and fall times of the symmetrical triangular wave current, specifically including: The battery management system reads the temperature sensor readings from inside the battery to obtain the current temperature value; When the current temperature is lower than the standard temperature, the battery management system will extend the rise and fall time of the symmetrical triangular wave current proportionally, with the extension ratio being greater as the temperature decreases. When the current temperature is higher than the standard temperature, the battery management system will shorten the rise and fall time of the symmetrical triangular wave current proportionally, and the higher the temperature, the greater the shortening ratio. The battery management system uses the corrected duration as the actual duration of this measurement, and compares the current lag time with the historical lag time at the same temperature after the measurement is completed, in order to eliminate the influence of temperature on the lag time.

8. The method for monitoring the capacity decay of secondary batteries in consumer products according to claim 7, characterized in that, The battery management system synchronously records the current charging strategy parameters during each measurement and uses only historical data under the same charging strategy parameters when comparing lag time lengths. Specifically, this includes: Before applying the symmetrical triangular wave current, the battery management system reads the current charging strategy parameters, including the current value during the constant current charging phase and the charging cutoff voltage value. The battery management system associates and stores the current charging strategy parameters with the measured lag time length. When comparing the current lag time with the historical lag time, the battery management system selects those historical lag time lengths that are the same as the current charging strategy parameters from the historical sequence, and only uses these selected historical data for comparison. When the number of historical data selected is less than the preset minimum number, the battery management system uses the initial lag time length as a benchmark and marks the current charging strategy parameters as the new benchmark state.

9. A method for monitoring the capacity decay of secondary batteries in consumer products according to claim 8, characterized in that, The battery management system also checks whether the battery has been replaced during each charge and clears the historical lag time sequence when battery replacement is detected. Specifically, this includes: At the start of each charge, the battery management system reads the unique identification code inside the battery and compares the identification code read this time with the identification code stored during the previous charge. When the current identification code is different from the previous identification code, the battery management system determines that the battery has been replaced and deletes all historical data in the historical lag time length sequence. The battery management system stores the lag time length measured during this charge as the new initial lag time length and resets the growth rate corresponding to this lag time length to zero. After clearing historical data, the battery management system resumes performing triangular wave current measurement on each charge until a preset amount of historical data is accumulated again.

10. A method for monitoring the capacity decay of secondary batteries in consumer products according to claim 9, characterized in that, The battery management system also detects the type of charger used during each charge and adjusts the baseline value of the lag time length according to the charger type, specifically including: When charging starts, the battery management system reads the charger's output capability parameters, including maximum output current and maximum output voltage, through a handshake protocol. The battery management system classifies the charger into fast charging type or normal charging type based on the output capability parameters. When the charger type is fast charging, the battery management system multiplies the measured lag time by a preset fast charging correction factor before comparing the lag time length. This fast charging correction factor is less than one. When the charger type is normal charging, the battery management system does not correct for the lag time length; The battery management system uses the corrected lag time length to compare with historical data in order to eliminate the influence of different charger output characteristics on the measurement results.

Images