A method for testing battery modules of new energy vehicles

By collecting voltage recovery and current surge responses of battery modules to calculate the polarization equivalent time constant, constructing a degradation acceleration factor and updating it in reverse, the problem of error amplification and early warning lag in the health status prediction of new energy vehicle battery modules is solved, and accurate lifetime estimation and timely early warning are achieved.

CN122283506BActive Publication Date: 2026-11-10ZHEJIANG YUNZHONGHE INTELLIGENT CONTROL TECHNOLOGY CO LTD
View PDF 3 Cites 0 Cited by

Patent Information

Application Number
CN202610565076.1
Authority / Receiving Office
CN · China
Patent Type
Patents(China)
Current Assignee / Owner
Filing Date
2026-04-27
Publication Date
2026-11-10
Estimated Expiration
2046-04-27

AI Technical Summary

Technical Problem

In existing technologies, the health status prediction of new energy vehicle battery modules suffers from problems such as conflict between linear extrapolation and nonlinear degradation mechanisms, amplification of errors due to open-loop iteration, and delayed early warning timing. This results in the predicted rate of health decline continuously lagging behind the actual decay, with errors amplified step by step, and the predicted remaining time deviating significantly from the true value, making it difficult to serve as an effective basis for maintenance decisions.

Method used

By collecting the voltage recovery time series of the battery module during the resting and relaxation phase after charging and the current mutation response during the discharge phase, the polarization equivalent time constant series is calculated as health reference data. A degradation acceleration factor is constructed, and the factor is updated in reverse using the deviation between the actual observed value and the predicted value, so as to realize the tracking and correction of the actual degradation process of the battery.

Benefits of technology

It improves the accuracy of remaining service life estimation and the timeliness of early warning, reduces the interference of dynamic noise on health status assessment, suppresses the successive accumulation of errors in open-loop iteration, and ensures that early warning information is issued in time before the battery rapidly degrades.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN122283506B_ABST
    Figure CN122283506B_ABST
Patent Text Reader

Abstract

This invention discloses a method for testing battery modules in new energy vehicles, belonging to the field of battery module fault detection and health management technology. This invention extracts the polarization equivalent time constant from voltage relaxation and current abrupt changes as health reference data, constructs a degradation acceleration factor based on its historical evolution, and then updates this acceleration factor using the deviation between actual observed values ​​and predicted values. This achieves the tracking and correction of the predicted trajectory to the actual degradation process of the battery, improving the accuracy of remaining service life estimation and the timeliness of early warning. It solves the problems in existing technologies such as the conflict between linear extrapolation and nonlinear degradation mechanisms, error amplification due to open-loop iteration, and delayed early warning timing.
Need to check novelty before this filing date? Find Prior Art

Description

Technical Field

[0001] This invention relates to the field of battery module fault detection and health management technology, and in particular to a method for detecting battery modules for new energy vehicles. Background Technology

[0002] With the large-scale application of new energy vehicles, battery modules, as the core energy source of the entire vehicle, are directly affected by the accurate detection of their health status, which in turn impacts the vehicle's range, operational safety, and lifespan. In the full lifecycle management of battery modules, real-time acquisition of operating parameters such as voltage, current, and temperature, and prediction of health degradation trends, has become a core function of the Battery Management System (BMS). However, actual operating conditions are complex and variable, and battery aging exhibits significant nonlinear and stage-specific characteristics. How to extract stable health characteristics from dynamic noise and accurately predict remaining lifespan remains a key technical challenge in the field of battery fault prediction and health management.

[0003] In related technologies, Chinese invention patent CN118330474B discloses a method for detecting battery modules of new energy vehicles. The steps include: collecting the operating parameters of the battery module at fixed intervals, calculating the average growth rate of the accumulated values ​​of each parameter, adding the current value to the average growth rate and inputting it into the Transformer model to obtain the predicted health level, repeating the addition until the predicted health level is lower than the lower limit, obtaining the expected remaining time based on the product of the number of additions and the collection interval, and comparing it with the user's input expected usage time to determine whether to issue a prompt.

[0004] However, this method has the following problems: First, the linear extrapolation using the historical global average growth rate conflicts with the actual multi-stage nonlinear degradation mechanism of the battery, causing the predicted rate of decline in health to lag behind the actual degradation. Second, the prediction error at each step under the open-loop iterative structure cannot be corrected, and the error is amplified step by step as the number of superpositions increases. Third, the error amplification directly causes the predicted remaining time to deviate significantly from the true value, and the warning timing lags behind the actual degradation process of the battery, making it difficult to serve as an effective basis for maintenance decisions. Summary of the Invention

[0005] This application provides a new energy vehicle battery module detection method, which solves the problems in the prior art such as the conflict between linear extrapolation and nonlinear degradation mechanism, error amplification caused by open-loop iteration, and delayed early warning timing. It realizes the tracking and correction of the predicted trajectory to the actual degradation process of the battery, and improves the accuracy of remaining service life estimation and the timeliness of early warning.

[0006] This application provides a method for testing battery modules in new energy vehicles, including: S1. The voltage recovery time series of the battery module during the resting relaxation phase after charging is collected by the sensor, and the polarization equivalent time constant series is calculated by combining the current change response during the discharge phase. The polarization equivalent time constant series is used as health reference data to characterize the current aging state. S2, extract the historical evolution characteristics of the health reference data, construct a degradation acceleration factor based on the historical evolution characteristics, use the degradation acceleration factor to correct the historical capacity retention rate of the battery module, obtain the current health status assessment value, and complete the preliminary status determination of fault prediction and health management. S3, establish a multi-stage trend extrapolation trajectory based on the current health status assessment value, collect the polarization equivalent time constant sequence of the next period to calculate the actual health observation value, and extract the deviation sequence between the actual health observation value and the predicted value in the multi-stage trend extrapolation trajectory; S4. The deviation sequence is used as a feedback adjustment quantity and input to the degradation acceleration factor for dynamic updating. The updated degradation acceleration factor is used to regenerate the corrected health status prediction trajectory. Based on the time when the corrected health status prediction trajectory reaches the failure threshold, the expected remaining service life warning information is output.

[0007] Furthermore, the step of acquiring the voltage recovery time series of the battery module during the resting relaxation phase after charging via sensors, and calculating the polarization equivalent time constant series by combining it with the current surge response during the discharge phase, includes: When the charging current of the battery module is detected to drop to zero, a timer is triggered, and the terminal voltage value during the resting relaxation phase is continuously recorded at a fixed sampling frequency to form a voltage recovery time series. Extract the step current amplitude at the end of the most recent discharge phase of the battery module and the instantaneous voltage drop amplitude corresponding to the step current amplitude. Divide the instantaneous voltage drop amplitude by the step current amplitude to obtain the ohmic internal resistance. The polarization voltage difference is calculated using the difference between the steady-state voltage and the initial power-off voltage in the voltage recovery time series, and the polarization equivalent time constant series is calculated based on the ratio of the ohmic internal resistance to the polarization voltage difference.

[0008] Furthermore, the use of the polarization equivalent time constant sequence as health reference data characterizing the current aging state includes: The polarization equivalent time constant sequences calculated under different charging cycles are arranged in chronological order. The polarization equivalent time constant sequence is smoothed by moving average filtering to remove outlier data points that deviate from the moving average value by more than two standard deviations, and the smoothed sequence is retained as a health reference data characterizing the current aging state.

[0009] Furthermore, the historical evolution characteristics of the extracted health reference data include: The difference between the health reference data and two adjacent charging cycles is calculated to obtain a constant rate of change sequence; By summing up all positively increasing values ​​in the cumulative constant rate of change sequence, the cumulative degradation index is obtained; Dividing the cumulative degradation index by the total number of corresponding charging cycles yields the average polarization growth slope, which serves as the historical evolution characteristic of the health reference data.

[0010] Furthermore, the construction of the degradation acceleration factor based on historical evolution characteristics includes: The baseline polarization growth slope under the preset initial state; Calculate the ratio of the average polarization growth slope to the baseline polarization growth slope; The ratio is transformed by natural logarithm, and the result of the natural logarithmic transformation is multiplied by a preset decay sensitivity coefficient to obtain the degradation acceleration factor used to characterize the rate of battery performance decline.

[0011] Furthermore, the step of correcting the historical capacity retention rate of the battery module using a degradation acceleration factor to obtain the current health status assessment value includes: Obtain the initial nominal capacity of the battery module at the time of manufacture and the recorded historical charge and discharge throughput; Calculate the base capacity decay using historical charge and discharge throughput. Multiply the base capacity decay by the degradation acceleration factor to obtain the corrected capacity decay. The current health status assessment value is obtained by subtracting the corrected capacity decay from the initial nominal capacity and then dividing by the initial nominal capacity.

[0012] Furthermore, the establishment of a multi-stage trend extrapolation trajectory based on the current health status assessment value includes: Use the current health status assessment value as the starting node; Using the fixed increment of historical charge and discharge throughput as the horizontal axis step size, the degradation acceleration factor is converted into the vertical axis decrease amount under each horizontal axis step size. Based on the decrease in the vertical axis, the predicted health status values ​​at each future prediction node are calculated sequentially, and all the predicted health status values ​​are connected to form a multi-stage trend extrapolation trajectory.

[0013] Furthermore, the step of collecting the polarization equivalent time constant sequence for the next period to calculate the actual health observation value, and extracting the deviation sequence between the actual health observation value and the predicted value in the multi-stage trend extrapolation trajectory, includes: After the battery module completes the next full charge-discharge cycle, the acquisition program for voltage recovery time series and current sudden change response is triggered again to generate the latest polarization equivalent time constant. Following the same operational rules used to calculate the current health status assessment value, the latest polarization equivalent time constant is converted into the actual health observation value; In the multi-stage trend extrapolation trajectory, find the predicted value at the time node corresponding to the next complete charge-discharge cycle, subtract the predicted value from the actual health observation value to obtain the single prediction deviation, and record the single prediction deviation of multiple cycles as a deviation sequence.

[0014] Furthermore, the step of using the deviation sequence as a feedback adjustment quantity to be input back into the degradation acceleration factor for dynamic updating includes: Extract the latest value from the deviation sequence and the difference between the latest value and the value of the previous period; The proportional compensation amount is obtained by multiplying the latest value in the deviation sequence by the proportional adjustment coefficient. The differential compensation amount is obtained by multiplying the difference between the latest value and the value of the previous cycle by the differential adjustment coefficient. The proportional compensation amount and the differential compensation amount are added together to form the feedback adjustment amount, and the original degradation acceleration factor value is added to the feedback adjustment amount to generate the dynamically updated degradation acceleration factor.

[0015] Furthermore, the step of regenerating the corrected health status prediction trajectory using the updated degradation acceleration factor, and outputting the estimated remaining service life warning information based on the time when the corrected health status prediction trajectory reaches the failure threshold, includes: Using the actual health observation value as the new starting point, the decline amount of each future prediction node is recalculated using the updated degradation acceleration factor to generate the corrected health status prediction trajectory. Locate the target node in the corrected health status prediction trajectory where the predicted health status value drops to a preset failure threshold; Calculate the total number of horizontal coordinate steps between the target node and the current node, multiply the total number of horizontal coordinate steps by the average time taken per cycle to obtain the estimated remaining lifespan. If the estimated remaining lifespan is less than the preset safety threshold duration, output a warning message containing a suggested replacement prompt through the display panel.

[0016] One or more technical solutions provided in the embodiments of this application have at least the following technical effects or advantages: This application provides a method for detecting battery modules in new energy vehicles. The method involves using sensors to collect voltage recovery time series during the post-charging relaxation phase and current surge response during the discharge phase, calculating a polarization equivalent time constant series as health reference data. Historical evolution characteristics of this health reference data are extracted to construct a degradation acceleration factor, and the historical capacity retention rate is corrected to obtain the current health status assessment value. Based on the current health status assessment value, a multi-stage trend extrapolation trajectory is established. The polarization equivalent time constant series for the next cycle is collected to calculate the actual health observation value, and the deviation sequence between the observation value and the predicted value is extracted. This deviation sequence is used as a feedback adjustment quantity and input back into the degradation acceleration factor for dynamic updating. The updated degradation acceleration factor is used to regenerate the corrected health status prediction trajectory, and the estimated remaining service life warning information is output based on the time it takes for the trajectory to reach the failure threshold.

[0017] In this process, by extracting the polarization equivalent time constant from the voltage relaxation response and the current sudden change response, the original voltage and current signals are transformed into physical characteristic quantities that are directly related to the internal electrochemical polarization state of the battery, thereby reducing the interference of dynamic noise on health status assessment during driving conditions.

[0018] Furthermore, by calculating the historical evolution characteristics of the polarization equivalent time constant sequence and constructing a degradation acceleration factor, the description of the degradation rate no longer depends on a fixed average increment, but is automatically adjusted according to the actual aging process of the battery, thereby matching the nonlinear change law of the battery from gradual degradation to accelerated degradation.

[0019] Furthermore, by using the deviation between the actual health observation and the predicted value as a feedback adjustment to reverse the degradation acceleration factor, the predicted trajectory is calibrated based on real measurement data after each cycle, which suppresses the successive accumulation of errors in the open-loop iteration. Finally, the deviation between the expected remaining lifespan and the actual remaining lifespan is controlled within an acceptable range, and the early warning information can be issued in time before the battery enters the rapid degradation stage. Attached Figure Description

[0020] Figure 1 This is a flowchart of a new energy vehicle battery module testing method provided in an embodiment of this application. Detailed Implementation

[0021] This application provides a new energy vehicle battery module detection method that solves the problems in the prior art, such as the conflict between linear extrapolation and nonlinear degradation mechanisms, error amplification caused by open-loop iteration, and delayed early warning timing. By extracting the polarization equivalent time constant from voltage relaxation and current mutation as health reference data, and constructing a degradation acceleration factor based on its historical evolution, and then using the deviation between the actual observed value and the predicted value to update the acceleration factor in reverse, the predicted trajectory can be used to follow and correct the actual degradation process of the battery, thereby improving the accuracy of remaining service life estimation and the timeliness of early warning.

[0022] To better understand the above technical solutions, the following will provide a detailed explanation of the technical solutions in conjunction with the accompanying drawings and specific implementation methods.

[0023] like Figure 1 As shown in the figure, this application provides a method for testing new energy vehicle battery modules, including: S1. The voltage recovery time series of the battery module during the resting and relaxation phase after charging is collected by the sensor, and the polarization equivalent time constant series is calculated by combining the current change response during the discharge phase. The polarization equivalent time constant series is used as health reference data to characterize the current aging state.

[0024] Furthermore, the step of acquiring the voltage recovery time series of the battery module during the resting relaxation phase after charging via sensors, and calculating the polarization equivalent time constant series by combining it with the current surge response during the discharge phase, includes: When the charging current of the battery module is detected to drop to zero, a timer is triggered, and the terminal voltage value during the resting relaxation phase is continuously recorded at a fixed sampling frequency to form a voltage recovery time series. Extract the step current amplitude at the end of the most recent discharge phase of the battery module and the instantaneous voltage drop amplitude corresponding to the step current amplitude. Divide the instantaneous voltage drop amplitude by the step current amplitude to obtain the ohmic internal resistance. The polarization voltage difference is calculated using the difference between the steady-state voltage and the initial power-off voltage in the voltage recovery time series, and the polarization equivalent time constant series is calculated based on the ratio of the ohmic internal resistance to the polarization voltage difference.

[0025] In this embodiment, the voltage recovery time series of the battery module during the resting relaxation phase after charging is acquired by sensors, and the polarization equivalent time constant series is calculated by combining the current surge response during the discharge phase. Specifically, this includes: After detecting the stop signal issued by the charging control command of the battery management system, the terminal voltage data is collected at a fixed frequency of once per second for a period of twenty minutes.

[0026] At this point, the step current amplitude at the end of the most recent discharge before charging of the battery module is extracted. and instantaneous voltage drop .

[0027] Ohmic resistance The calculation formula is as follows: The above formula is derived from a modification of Ohm's law, the basic law of circuits, and is used to isolate the voltage drop of the purely resistive component inside the battery.

[0028] Subsequently, the steady-state voltage at the end of the 20-minute time series was extracted from the voltage recovery time series. Initial power-off voltage at the moment the charging current is cut off Calculate the polarization voltage difference : .

[0029] Therefore, the polarization equivalent time constant Based on the charging and discharging response principle of a first-order RC circuit, the formula is derived as follows: ; In this formula, This represents the amplitude of the discharge step current. This represents the instantaneous voltage drop. For ohmic internal resistance, For steady-state voltage, This is the initial power-off voltage. This represents the polarization voltage difference.

[0030] It needs to be explained that the polarization equivalent time constant The derivation is based on a first-order RC equivalent circuit model. In this model, the battery polarization characteristics are determined by the polarization internal resistance. and polarization capacitor Parallel description, the polarization equivalent time constant is defined as follows: During the resting and relaxation phase, the polarized capacitor discharges through its internal polarization resistance, and the terminal voltage changes from... The index rose to voltage difference It equals the initial voltage drop across the polarization resistance, i.e. .therefore At the same time, Ohm's internal resistance Due to instantaneous voltage drop With step current The ratio is obtained. Polarization equivalent time constant. satisfy ,and The value cannot be directly measured, but it can be obtained through the time response characteristics of a first-order RC circuit. It can also be expressed as the time constant required to reach a steady-state value during the voltage recovery process. Specifically, the voltage recovery curve... ,in That is, the time constant to be determined, where e is the natural constant. Expanding the exponential term and ignoring higher-order terms, we can approximate it as follows: This approximation holds true when the polarization voltage is much smaller than the terminal voltage; under these conditions, the error in a real battery is less than 5%.

[0031] Furthermore, using the polarization equivalent time constant sequence as health reference data characterizing the current aging state includes: The polarization equivalent time constant sequences calculated under different charging cycles are arranged in chronological order. The polarization equivalent time constant sequence is smoothed by moving average filtering to remove outlier data points that deviate from the moving average value by more than two standard deviations, and the smoothed sequence is retained as a health reference data characterizing the current aging state.

[0032] In this embodiment, the polarization equivalent time constant sequence is used as health reference data characterizing the current aging state, specifically including: History The polarization equivalent time constant values ​​obtained in the second iteration are sorted by time. To eliminate measurement noise, a moving average filter is applied to the data points.

[0033] It should be explained that the moving average smoothing filter uses a sliding window method for calculation. The window length is fixed at the data points of the most recent 5 cycles. For the current... Take 1 data point to The arithmetic mean of 5 points (if the first and last points are insufficient, use symmetrical points) is calculated as the moving average for that point. The standard deviation is calculated based on data within the same window. The removal criterion is: if the absolute value of the difference between the current data point and the moving average is greater than twice the standard deviation within the window, it is considered an outlier and replaced by the moving average. Because the window slides with the data points, the removal criterion for each data point depends only on its neighboring data. Therefore, outliers in the early stable phase will not be missed due to an excessively large global standard deviation, and the sensitivity in the later fluctuating phase will not decrease. The window length of 5 is an empirical value and can be adjusted according to the total number of sampling cycles, but at least 3 valid points must be ensured within the window.

[0034] S2, extract the historical evolution characteristics of the health reference data, construct a degradation acceleration factor based on the historical evolution characteristics, use the degradation acceleration factor to correct the historical capacity retention rate of the battery module, obtain the current health status assessment value, and complete the preliminary status determination of fault prediction and health management.

[0035] Furthermore, the extraction of historical evolution features from the health reference data includes: The difference between the health reference data and two adjacent charging cycles is calculated to obtain a constant rate of change sequence; By summing up all positively increasing values ​​in the cumulative constant rate of change sequence, the cumulative degradation index is obtained; Dividing the cumulative degradation index by the total number of corresponding charging cycles yields the average polarization growth slope, which serves as the historical evolution characteristic of the health reference data.

[0036] In this embodiment, extracting the historical evolution features of the health reference data specifically includes: Calculate the rate of change of constant of the polarization equivalent time constant between adjacent cycles. The formula is: ; in Represents the current loop count. This represents the polarization equivalent time constant of the current cycle k. This represents the polarization equivalent time constant of the previous cycle k-1, where k is the cycle number.

[0037] The cumulative degradation index is obtained by summing the constant rates of change that are greater than zero. Then the average polarization growth slope was calculated. : ; in, This represents the total number of loops.

[0038] It needs to be explained that the rate of change of the constant The calculated result may be negative, indicating a decrease in the equivalent polarization time constant between adjacent cycles. In actual battery aging, the polarization time constant typically increases monotonically; negative values ​​are mainly caused by measurement noise or fluctuations in operating conditions. Therefore, only positively increasing values ​​(greater than zero) are accumulated. Obtain cumulative degradation index Ignore negative fluctuations to avoid underestimating the overall degradation trend due to temporary declines. Total number of cycles. This represents the total number of complete charge-discharge cycles completed to date, counted from the battery's first use. Average polarization growth slope. The unit is the same as the polarization equivalent time constant (seconds), representing the average increase of the polarization time constant in each cycle.

[0039] Furthermore, the construction of the degradation acceleration factor based on the historical evolution characteristics includes: The baseline polarization growth slope under the preset initial state; Calculate the ratio of the average polarization growth slope to the baseline polarization growth slope; The ratio is transformed by natural logarithm, and the result of the natural logarithmic transformation is multiplied by a preset decay sensitivity coefficient to obtain the degradation acceleration factor used to characterize the rate of battery performance decline.

[0040] In this embodiment, a degradation acceleration factor is constructed based on the historical evolution characteristics, specifically including: Retrieve the reference polarization growth slope stored in the pre-stored memory Degradation acceleration factor constructed by natural logarithm mapping The formula is: ; in The preset attenuation sensitivity coefficient is usually set to a fixed value based on the factory calibration test.

[0041] It needs to be explained that the natural logarithmic transformation is used to transform the ratio Mapped to the logarithmic domain, the battery degradation rate exhibits an exponential relationship with the polarization growth slope. A preset degradation sensitivity coefficient... Calibration was achieved through accelerated aging tests before shipment: several battery modules from the same batch were subjected to cyclic aging under standard operating conditions, and the capacity retention rate and polarization equivalent time constant were recorded at different cycle counts. Using as the independent variable and the actual capacity decay factor as the dependent variable, the following fit was obtained: Once calibrated, this coefficient is a fixed value and is written into the non-volatile memory of the battery management system, requiring no online adjustment.

[0042] Furthermore, the step of using the degradation acceleration factor to correct the historical capacity retention rate of the battery module to obtain the current health status assessment value includes: Obtain the initial nominal capacity of the battery module at the time of manufacture and the recorded historical charge and discharge throughput; Calculate the base capacity decay using historical charge and discharge throughput. Multiply the base capacity decay by the degradation acceleration factor to obtain the corrected capacity decay. The current health status assessment value is obtained by subtracting the corrected capacity decay from the initial nominal capacity and then dividing by the initial nominal capacity.

[0043] In this embodiment, the historical capacity retention rate of the battery module is corrected using the degradation acceleration factor to obtain the current health status assessment value, specifically including: Retrieve initial nominal capacity and historical charge / discharge throughput Through the basic conversion coefficient Calculate the basic capacity decay, and then combine it with the degradation acceleration factor. Calculate the corrected capacity decay : ; Current health status assessment value The calculation formula is: ; This value represents the proportion of the current available capacity to the initial nominal capacity.

[0044] It needs to be explained that the historical charge / discharge throughput... This is accumulated by the current integration module in the battery management system, with the integration starting at the moment the battery is first put into use. Basic conversion coefficient. The physical meaning of the basic conversion factor is the ideal capacity decay corresponding to a unit throughput (e.g., ampere-hour), determined through standard cycle life testing at the factory. Specifically, it is obtained by cyclic aging of new battery modules of the same model at standard temperature (25℃) and standard rate (0.5C charge / discharge), recording the actual capacity after each cycle, fitting a linear relationship between capacity decay and cumulative throughput, and the slope of this relationship is the basic conversion factor. Degradation Accelerator Multiply This is equivalent to scaling the ideal decay rate by the current polarization growth slope. Health status assessment value. The value range is [0,1].

[0045] S3, establish a multi-stage trend extrapolation trajectory based on the current health status assessment value, collect the polarization equivalent time constant sequence of the next period to calculate the actual health observation value, and extract the deviation sequence between the actual health observation value and the predicted value in the multi-stage trend extrapolation trajectory.

[0046] Furthermore, the step of establishing a multi-stage trend extrapolation trajectory based on the current health status assessment value includes: Use the current health status assessment value as the starting node; Using the fixed increment of historical charge and discharge throughput as the horizontal axis step size, the degradation acceleration factor is converted into the vertical axis decrease amount under each horizontal axis step size. Based on the decrease in the vertical axis, the predicted health status values ​​at each future prediction node are calculated sequentially, and all the predicted health status values ​​are connected to form a multi-stage trend extrapolation trajectory.

[0047] In this embodiment, a multi-stage trend extrapolation trajectory is established based on the current health status assessment value, specifically including: by Starting from this point, for the future... Each prediction node has a predicted health status value. for: ; in The throughput increment is a fixed horizontal axis step size.

[0048] It needs to be explained that the horizontal axis step size is fixed. The value is consistent with the unit of the cumulative throughput, usually set to 1 ampere-hour or the average capacity corresponding to one complete charge and discharge cycle. The vertical axis represents the decrease in throughput. Indicates each increase The amount of power throughput, the amount by which the predicted health status decreases. This decrease remains constant at each step, thus forming a linear extrapolation trajectory. If the degradation acceleration factor... If updated, the slope of the regenerated trajectory will also change. The multi-stage trend extrapolation trajectory in this step is only used as an initial prediction and will be adjusted later through feedback correction. Prediction node number Starting from 1 and incrementing, each node corresponds to one time. The cumulative total.

[0049] Furthermore, the step of collecting the polarization equivalent time constant sequence for the next period to calculate the actual health observation value, and extracting the deviation sequence between the actual health observation value and the predicted value in the multi-stage trend extrapolation trajectory, includes: After the battery module completes the next full charge-discharge cycle, the acquisition program for voltage recovery time series and current sudden change response is triggered again to generate the latest polarization equivalent time constant. Following the same operational rules used to calculate the current health status assessment value, the latest polarization equivalent time constant is converted into the actual health observation value; In the multi-stage trend extrapolation trajectory, find the predicted value at the time node corresponding to the next complete charge-discharge cycle, subtract the predicted value from the actual health observation value to obtain the single prediction deviation, and record the single prediction deviation of multiple cycles as a deviation sequence.

[0050] In this embodiment, the polarization equivalent time constant sequence of the next period is collected to calculate the actual health observation value, and the deviation sequence between the actual health observation value and the predicted value in the multi-stage trend extrapolation trajectory is extracted, specifically including: The latest data will be collected in the next cycle to obtain the actual health observation values. .

[0051] Extracting single prediction bias : ; in This is the time node number corresponding to this cycle on the trajectory.

[0052] It needs to be explained that the actual health observation values The calculation process and the current health status assessment value The process is completely identical: the polarization equivalent time constant after the next complete charge-discharge cycle is collected to obtain the degradation acceleration factor of that cycle, and then combined with the cumulative throughput up to that cycle. Calculate .

[0053] Notice This corresponds to the actual measured health status, while This is the estimated health status of the cycle from the previous prediction. Subtracting the two values ​​yields the single prediction bias. .like A positive value indicates that the actual health status is higher than the predicted value, meaning the decline is slower than expected; if... A negative value indicates that the actual decay is faster than expected. The deviation values ​​from multiple cycles are stored in chronological order as a deviation sequence for feedback adjustment.

[0054] S4, the deviation sequence is used as a feedback adjustment quantity and input to the degradation acceleration factor for dynamic updating. The updated degradation acceleration factor is used to regenerate the corrected health status prediction trajectory, and the estimated remaining service life warning information is output according to the time when the corrected health status prediction trajectory reaches the failure threshold.

[0055] Furthermore, the step of using the deviation sequence as a feedback adjustment quantity to be input back into the degradation acceleration factor for dynamic updating includes: Extract the latest value from the deviation sequence and the difference between the latest value and the value of the previous period; The proportional compensation amount is obtained by multiplying the latest value in the deviation sequence by the proportional adjustment coefficient. The differential compensation amount is obtained by multiplying the difference between the latest value and the value of the previous cycle by the differential adjustment coefficient. The proportional compensation amount and the differential compensation amount are added together to form the feedback adjustment amount, and the original degradation acceleration factor value is added to the feedback adjustment amount to generate the dynamically updated degradation acceleration factor.

[0056] In this embodiment, the deviation sequence is used as a feedback adjustment quantity and input back into the degradation acceleration factor for dynamic updating, specifically including: Deviation value for the current period n Deviation value from the previous period n-1 Calculate the feedback adjustment amount : in This is the proportional adjustment coefficient. This is the differential adjustment coefficient.

[0057] Dynamically updated degradation acceleration factor for: .

[0058] It needs to be explained that the proportional adjustment coefficient and differential adjustment coefficient The parameters are obtained through standard PID controller tuning methods. Specifically, this method is used in battery module bench aging tests, and the deviation sequence is recorded. Changes over time. Using the Ziegler-Nichols method: Let... gradually increase Record the critical gain at the point when the system output exhibits constant amplitude oscillation. and oscillation period Then, according to the empirical formula... , Set an initial value. A simple trial-and-error method can also be used: start with a smaller value. Begin by observing the deviation. If the convergence is slow, increase the value appropriately if the decrease occurs within several iterations. If overshoot oscillation occurs, then reduce... Or increase These coefficients are fixed after factory calibration and do not require online adaptive adjustment.

[0059] Furthermore, the step of regenerating the corrected health status prediction trajectory using the updated degradation acceleration factor, and outputting an estimated remaining service life warning based on the time when the corrected health status prediction trajectory reaches the failure threshold, includes: Using the actual health observation value as the new starting point, the decline amount of each future prediction node is recalculated using the updated degradation acceleration factor to generate the corrected health status prediction trajectory. Locate the target node in the corrected health status prediction trajectory where the predicted health status value drops to a preset failure threshold; Calculate the total number of horizontal coordinate steps between the target node and the current node, multiply the total number of horizontal coordinate steps by the average time taken per cycle to obtain the estimated remaining lifespan. If the estimated remaining lifespan is less than the preset safety threshold duration, output a warning message containing a suggested replacement prompt through the display panel.

[0060] In this embodiment, the corrected health status prediction trajectory is regenerated using the updated degradation acceleration factor, and the estimated remaining service life warning information is output based on the time when the corrected health status prediction trajectory reaches the failure threshold. Specifically, this includes: use Recalculate the future trajectory. The predicted trajectory has reached the failure threshold. Remaining step size .

[0061] Expected remaining service life The formula is: ; in This represents the average time taken per cycle.

[0062] It needs to be explained that the failure threshold Pre-set according to the safety requirements of the battery module, typically between 0.7 and 0.8, corresponding to a capacity retention rate of 70% to 80%. Average cycle time. The remaining step size is calculated by averaging the time intervals of the 10 most recent complete charge-discharge cycles, as determined by the battery management system, and updated in real time. Starting from the current node, according to the updated degradation acceleration factor The health status prediction value is calculated successively until it first falls below the failure threshold. The number of steps taken. Estimated remaining service life. Units and Same, usually in hours or days. If the remaining useful life is expected... If the time frame is less than a preset safety threshold, a warning message will be output. The preset safety threshold includes the maintenance cycle calculated based on the user's daily mileage. The display panel refers to the vehicle's central control screen or instrument panel, and the warning message will be displayed in text form as "Recommendation to replace the battery module".

[0063] The above formulas are all dimensionless calculations. The formulas are derived from software simulations based on a large amount of collected data to obtain the most recent real-world results. The preset parameters in the formulas are set by those skilled in the art according to the actual situation.

[0064] The above embodiments can be implemented, in whole or in part, by software, hardware, firmware, or any other combination thereof. When implemented using software, the above embodiments can be implemented, in whole or in part, in the form of a computer program product.

[0065] Those skilled in the art will recognize that the modules and algorithm steps of the various examples described in conjunction with the embodiments disclosed herein can be implemented in electronic hardware, or a combination of computer software and electronic hardware. Whether these functions are implemented in hardware or software depends on the specific application and design constraints of the technical solution. Those skilled in the art can use different methods to implement the described functions for each specific application, but such implementation should not be considered beyond the scope of this application.

[0066] In addition, the functional modules in the various embodiments of this application can be integrated into one processing module, or each module can exist physically separately, or two or more modules can be integrated into one module.

[0067] 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.

[0068] In conclusion, the above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of the present invention should be included within the protection scope of the present invention.

Claims

1. A method for testing battery modules of new energy vehicles, characterized in that, Includes the following steps: S1. The voltage recovery time series of the battery module during the resting relaxation phase after charging is collected by the sensor, and the polarization equivalent time constant series is calculated by combining the current change response during the discharge phase. The polarization equivalent time constant series is used as health reference data to characterize the current aging state. S2, extract the historical evolution characteristics of the health reference data, construct a degradation acceleration factor based on the historical evolution characteristics, use the degradation acceleration factor to correct the historical capacity retention rate of the battery module, obtain the current health status assessment value, and complete the preliminary status determination of fault prediction and health management. S3, establish a multi-stage trend extrapolation trajectory based on the current health status assessment value, collect the polarization equivalent time constant sequence of the next period to calculate the actual health observation value, and extract the deviation sequence between the actual health observation value and the predicted value in the multi-stage trend extrapolation trajectory; S4, the deviation sequence is used as a feedback adjustment quantity and input to the degradation acceleration factor for dynamic updating. The updated degradation acceleration factor is used to regenerate the corrected health status prediction trajectory, and the estimated remaining service life warning information is output according to the time when the corrected health status prediction trajectory reaches the failure threshold. A multi-stage trend extrapolation trajectory is established based on the current health status assessment value, including: by Starting from this point, for the future... Each prediction node has a predicted health status value. for: ; in For a fixed horizontal axis step size, the throughput power increment, The current health status assessment value, the basic conversion coefficient. The physical meaning is the ideal capacity decay per unit throughput. The ordinate represents the rate of decrease in degradation. Indicates each increase The amount of power throughput, and the amount of reduction in the predicted health status.

2. The method for testing new energy vehicle battery modules as described in claim 1, characterized in that, The step of acquiring the voltage recovery time series of the battery module during the resting relaxation phase after charging via sensors, and calculating the polarization equivalent time constant series by combining the current surge response during the discharge phase, includes: When the charging current of the battery module is detected to drop to zero, a timer is triggered, and the terminal voltage value during the resting relaxation phase is continuously recorded at a fixed sampling frequency to form a voltage recovery time series. Extract the step current amplitude at the end of the most recent discharge phase of the battery module and the instantaneous voltage drop amplitude corresponding to the step current amplitude. Divide the instantaneous voltage drop amplitude by the step current amplitude to obtain the ohmic internal resistance. The polarization voltage difference is calculated using the difference between the steady-state voltage and the initial power-off voltage in the voltage recovery time series, and the polarization equivalent time constant series is calculated based on the ratio of the ohmic internal resistance to the polarization voltage difference.

3. The method for testing new energy vehicle battery modules as described in claim 1, characterized in that, The use of the polarization equivalent time constant sequence as health reference data characterizing the current aging state includes: The polarization equivalent time constant sequences calculated under different charging cycles are arranged in chronological order. The polarization equivalent time constant sequence is smoothed by moving average filtering to remove outlier data points that deviate from the moving average value by more than two standard deviations, and the smoothed sequence is retained as a health reference data characterizing the current aging state.

4. The method for testing new energy vehicle battery modules as described in claim 1, characterized in that, The historical evolution characteristics of the extracted health reference data include: The difference between the health reference data and two adjacent charging cycles is calculated to obtain a constant rate of change sequence; By summing up all positively increasing values ​​in the cumulative constant rate of change sequence, the cumulative degradation index is obtained; Dividing the cumulative degradation index by the total number of corresponding charging cycles yields the average polarization growth slope, which serves as the historical evolution characteristic of the health reference data.

5. The method for testing new energy vehicle battery modules as described in claim 1, characterized in that, The degradation acceleration factor constructed based on historical evolution characteristics includes: The baseline polarization growth slope under the preset initial state; Calculate the ratio of the average polarization growth slope to the baseline polarization growth slope; The ratio is transformed by natural logarithm, and the result of the natural logarithmic transformation is multiplied by a preset decay sensitivity coefficient to obtain the degradation acceleration factor used to characterize the rate of battery performance decline.

6. The method for testing new energy vehicle battery modules as described in claim 1, characterized in that, The process of correcting the historical capacity retention rate of the battery module using a degradation acceleration factor to obtain the current health status assessment value includes: Obtain the initial nominal capacity of the battery module at the time of manufacture and the recorded historical charge and discharge throughput; Calculate the base capacity decay using historical charge and discharge throughput. Multiply the base capacity decay by the degradation acceleration factor to obtain the corrected capacity decay. The current health status assessment value is obtained by subtracting the corrected capacity decay from the initial nominal capacity and then dividing by the initial nominal capacity.

7. The method for testing new energy vehicle battery modules as described in claim 1, characterized in that, The establishment of a multi-stage trend extrapolation trajectory based on the current health status assessment value includes: Use the current health status assessment value as the starting node; Using the fixed increment of historical charge and discharge throughput as the horizontal axis step size, the degradation acceleration factor is converted into the vertical axis decrease amount under each horizontal axis step size. Based on the decrease in the vertical axis, the predicted health status values ​​at each future prediction node are calculated sequentially, and all the predicted health status values ​​are connected to form a multi-stage trend extrapolation trajectory.

8. The method for testing new energy vehicle battery modules as described in claim 1, characterized in that, The process of collecting the polarization equivalent time constant sequence for the next cycle to calculate the actual health observation value, and extracting the deviation sequence between the actual health observation value and the predicted value in the multi-stage trend extrapolation trajectory, includes: After the battery module completes the next full charge-discharge cycle, the acquisition program for voltage recovery time series and current sudden change response is triggered again to generate the latest polarization equivalent time constant. Following the same operational rules used to calculate the current health status assessment value, the latest polarization equivalent time constant is converted into the actual health observation value; In the multi-stage trend extrapolation trajectory, find the predicted value at the time node corresponding to the next complete charge-discharge cycle, subtract the predicted value from the actual health observation value to obtain the single prediction deviation, and record the single prediction deviation of multiple cycles as a deviation sequence.

9. The method for testing new energy vehicle battery modules as described in claim 1, characterized in that, The step of using the deviation sequence as a feedback adjustment quantity to be input back into the degradation acceleration factor for dynamic updating includes: Extract the latest value from the deviation sequence and the difference between the latest value and the value of the previous period; The proportional compensation amount is obtained by multiplying the latest value in the deviation sequence by the proportional adjustment coefficient. The differential compensation amount is obtained by multiplying the difference between the latest value and the value of the previous cycle by the differential adjustment coefficient. The proportional compensation amount and the differential compensation amount are added together to form the feedback adjustment amount, and the original degradation acceleration factor value is added to the feedback adjustment amount to generate the dynamically updated degradation acceleration factor.

10. The method for testing new energy vehicle battery modules as described in claim 1, characterized in that, The process of regenerating the corrected health status prediction trajectory using the updated degradation acceleration factor, and outputting an estimated remaining service life warning based on the time when the corrected health status prediction trajectory reaches the failure threshold, includes: Using the actual health observation value as the new starting point, the decline amount of each future prediction node is recalculated using the updated degradation acceleration factor to generate the corrected health status prediction trajectory. Locate the target node in the corrected health status prediction trajectory where the predicted health status value drops to a preset failure threshold; Calculate the total number of horizontal coordinate steps between the target node and the current node, multiply the total number of horizontal coordinate steps by the average time taken per cycle to obtain the estimated remaining lifespan. If the estimated remaining lifespan is less than the preset safety threshold duration, output a warning message containing a suggested replacement prompt through the display panel.

Citation Information

Patent Citations

  • A new energy vehicle battery module detection method

    CN118330474B

  • Sodium-ion battery state-of-charge and state-of-health joint estimation method

    CN120595133A

  • Battery health state detection method, device and equipment and storage medium

    CN120630018A