An online detection method and system for aging of a power battery cell

Abstract

The application relates to the technical field of battery state detection, and discloses an aging online detection method and system for power battery cells, which comprises the following steps: determining an adaptive voltage step and a phase change active voltage window based on voltage resolution and environmental noise; collecting operation data to determine a pseudo-steady segment, performing charge accumulation during the period, checking the charge amount when the voltage change reaches the step, calculating discrete differential capacity and characteristic voltage; mapping the two to a two-dimensional statistical matrix to perform value increment; traversing the matrix to locate a probability extreme point and comparing the probability extreme point with an initial benchmark to generate a result. The system comprises a parameter acquisition and configuration module, a state slice positioning module, an interlocking verification and extraction module, a statistical matrix mapping module and an aging degree output module. The application integrates fragmented data through a statistical matrix to exclude dynamic interference, extracts features without continuous charging and discharging, and improves the accuracy of online detection of the aging state of the power battery cells.

Description

Technical Field

[0001] This invention relates to the field of battery state detection technology, specifically to an online aging detection method and system for power battery cells. Background Technology

[0002] Power battery cells are the fundamental energy storage units of new energy devices, primarily relying on the electrochemical reversible reactions of their internal active materials to convert electrical energy into chemical energy. During long-term charging, discharging, and use, irreversible physical losses occur within the cell materials, leading to a continuous decrease in usable capacity and thus aging. To understand the internal health status of the cells, prevent potential safety hazards, and provide data support for fault prediction and health management of the entire device, it is necessary to collect physical quantity data during daily operation and implement online monitoring of relevant status indicators.

[0003] Existing detection methods typically collect full-range operating data during constant low-current charging and discharging to construct a continuous curve showing the change in capacity with voltage. These methods mainly calculate the overall capacity degradation ratio by analyzing the coordinate positions of specific characteristic peaks on the complete curve and comparing their offset with the characteristic peaks of the initial factory state.

[0004] This calculation method relies on a charging and discharging process with stable current and a complete voltage range. In actual operation, random load changes can introduce dynamic interference and hardware background noise into the collected data, easily distorting the characteristic curves obtained by conventional continuous difference calculations. Furthermore, daily operating conditions often consist of short-range segments that are difficult to establish a coherent time sequence. When directly extracting features from these noisy, fragmented data, existing processing methods struggle to effectively remove interference from current fluctuations, and truncation errors are easily introduced during the data discretization and gridding process, leading to deviations in the final obtained feature extremum coordinates and compromising the accuracy of detection results under complex operating conditions.

[0005] Therefore, this invention proposes an online aging detection method and system for power battery cells to address the shortcomings of existing technologies. Summary of the Invention

[0006] To address the shortcomings of existing technologies, this invention provides an online aging detection method and system for power battery cells, which solves the problem that the internal aging degree of power battery cells is difficult to accurately analyze when performing online detection due to the superposition and interference of fragmented operating data under actual working conditions.

[0007] To achieve the above objectives, the present invention provides the following technical solution: an online aging detection method for power battery cells, comprising the following steps: The minimum effective voltage resolution and ambient noise floor amplitude of the analog-to-digital converter are obtained. Based on the minimum effective voltage resolution and ambient noise floor amplitude, an adaptive voltage step size is determined, and a phase-change active voltage window is set. Real-time acquisition of operational physical quantity data of power battery cells in online operation state, determination of pseudo steady state segment based on the operational physical quantity data, and setting of initial voltage reference and initial charge accumulation value when the pseudo steady state segment is determined to be entered; During the duration of the pseudo-steady state segment, an updated charge accumulation value is obtained by performing a charge accumulation operation on the initial charge accumulation value based on the running physical quantity data. When the absolute change of the terminal voltage in the running physical quantity data relative to the initial voltage reference is greater than or equal to the adaptive voltage step size, a charge quantity verification judgment is performed using the updated charge accumulation value. If the verification condition of the charge quantity verification judgment is met, the discrete differential capacity and characteristic voltage are calculated, and the initial voltage reference and the updated charge accumulation value are reset. Determine whether the characteristic voltage is within the phase transition active voltage window. If the characteristic voltage is within the phase transition active voltage window, map the characteristic voltage and the discrete differential capacity to the corresponding grid coordinates of a preset two-dimensional statistical matrix and perform an auto-increment operation on the count value corresponding to the grid coordinates. The preset two-dimensional statistical matrix after the count value increment operation is traversed, and the probability ridge extreme point is located based on the maximum count value in the preset two-dimensional statistical matrix. The drift of the probability ridge extreme point is compared with the preset early life reference extreme point to generate the aging detection result for indicating the power battery cell.

[0008] Preferably, the steps of obtaining the minimum effective voltage resolution of the analog-to-digital converter and the ambient noise floor amplitude, determining the adaptive voltage step size based on the minimum effective voltage resolution and the ambient noise floor amplitude, and setting the phase transition active voltage window include: The reference voltage reference value and the conversion sampling bit number of the analog-to-digital converter are read, and the reference voltage reference value is divided by the full-scale digital step number determined based on the conversion sampling bit number to obtain the minimum effective voltage resolution; When the external high-voltage relay is in the open state and the loop current collected by the externally connected current sensor is continuously zero, a set number of power battery cell terminal voltage data are continuously collected. Extract the maximum and minimum voltage values ​​from the terminal voltage data of the power battery cells, and subtract the minimum voltage value from the maximum voltage value to obtain the ambient background noise amplitude. The ambient background noise amplitude is divided by the minimum effective voltage resolution and rounded up. The rounded result is then incremented by one as the adaptive step coefficient. The adaptive voltage step size is obtained by multiplying the adaptive step coefficient by the minimum effective voltage resolution. The chemical system parameters of the power battery cell are retrieved, and based on the chemical system parameters, the lower and upper voltage limits of the phase change active voltage window, which includes the aging characteristic range, are set.

[0009] Preferably, the step of real-time acquisition of operational physical quantity data of the power battery cells in online operation, determining pseudo-steady-state segments based on the operational physical quantity data, and setting an initial voltage reference and an initial charge accumulation value when determining that the pseudo-steady-state segment has been entered includes: The terminal voltage and circuit current of the power battery cell are collected in real time at fixed sampling time intervals as the operating physical quantity data, and the operating physical quantity data are stored in the order of sampling time to establish a sliding time window; The current change rate of the loop current is obtained by subtracting the loop current at the latest moment from the loop current at the oldest moment within the sliding time window, and then dividing by the time length of the sliding time window. Determine whether the absolute value of the current change rate is less than or equal to a preset current fluctuation threshold, and determine whether the absolute value of the loop current is greater than or equal to a preset effective operating current lower limit; when the absolute value of the current change rate is less than or equal to the preset current fluctuation threshold, and the absolute value of the loop current is greater than or equal to the preset effective operating current lower limit, determine that the power battery cell has entered the pseudo-steady state segment. The terminal voltage at which the pseudo-steady state segment is entered is assigned to the initial voltage reference, and the initial charge accumulation value is assigned to zero.

[0010] Preferably, the updated charge accumulation value is obtained by performing a charge accumulation operation on the initial charge accumulation value based on the operational physical quantity data. When the absolute change of the terminal voltage in the operational physical quantity data relative to the initial voltage reference is greater than or equal to the adaptive voltage step size, the step of performing charge quantity verification and determination using the updated charge accumulation value includes: During the duration of the pseudo-steady state segment, the fixed sampling time interval is multiplied by the loop current and converted to ampere-hours to obtain the single charge quantity. The single charge quantity is then used to perform a successive accumulation operation on the initial charge accumulation value to obtain the updated charge accumulation value. The instantaneous voltage difference is obtained by subtracting the initial voltage reference from the current terminal voltage, and the absolute value of the instantaneous voltage difference is extracted as the absolute change. When the absolute change is greater than or equal to the adaptive voltage step size, the absolute value of the updated charge accumulation value is extracted as a verification comparison value, and it is determined whether the verification comparison value is greater than or equal to the preset lower limit threshold of charge integration, so as to complete the charge quantity verification judgment.

[0011] Preferably, the step of calculating the discrete differential capacity and characteristic voltage, and resetting the initial voltage reference and the updated charge accumulation value if the verification conditions of the charge quantity verification judgment are met includes: When the verification comparison value is greater than or equal to the preset lower limit threshold of charge integral, it is confirmed that the verification condition of the charge quantity verification judgment is met. The discrete differential capacity is obtained by dividing the verification comparison value when the verification condition is met by the adaptive voltage step size; the discrete differential capacity is calculated according to the formula: ; In the formula, For discrete difference capacity; The verification comparison value when the verification conditions are met; For adaptive voltage step size; The characteristic voltage is obtained by adding the current terminal voltage to the initial voltage reference and then dividing by two. The current terminal voltage is reassigned to the initial voltage reference, and the updated charge accumulation value is reassigned to zero to complete the operation of resetting the initial voltage reference and the updated charge accumulation value.

[0012] Preferably, the step of mapping the characteristic voltage and the discrete differential capacity to the corresponding grid coordinates of a preset two-dimensional statistical matrix and performing an auto-increment operation on the count value corresponding to the corresponding grid coordinates if the characteristic voltage is within the phase transition active voltage window includes: If the characteristic voltage is within the phase transition active voltage window, the characteristic voltage is subtracted from the lower voltage limit to obtain the characteristic voltage difference, and the discrete differential capacity is subtracted from the preset capacity grid lower limit to obtain the capacity difference. The characteristic voltage difference is divided by the adaptive voltage step size and rounded down to generate a horizontal axis storage index. The capacity difference is divided by the preset capacity grid resolution and rounded down to generate a vertical axis storage index. By concatenating the horizontal axis storage index with the vertical axis storage index, the corresponding grid coordinates in the preset two-dimensional statistical matrix are located, and the count value corresponding to the grid coordinates is incremented.

[0013] Preferably, the step of locating the probabilistic ridge extreme point based on the maximum count value in the preset two-dimensional statistical matrix includes: When the preset evaluation period is reached, the preset two-dimensional statistical matrix after the count value increment operation is traversed to obtain the maximum count value, and the specific grid coordinates corresponding to the maximum count value, as well as the horizontal axis storage index and the vertical axis storage index, are extracted. Using the horizontal and vertical storage indices corresponding to the specific grid coordinates, extreme point voltages and extreme point capacities are generated; the extreme point voltages and extreme point capacities are then calculated according to the formulas. ; In the formula, The voltage at the extreme point; Store an index for the horizontal axis corresponding to a specific grid coordinate; For adaptive voltage step size; This is the lower limit of the phase transition active voltage window; ; In the formula, Capacity at extreme points; Store an index for the vertical axis corresponding to a specific grid coordinate; The preset capacity grid resolution; This is the preset lower limit of the capacity grid; The extreme point voltage and the extreme point capacity together form a two-dimensional physical coordinate pair, and the two-dimensional physical coordinate pair is determined as the extreme point of the probability ridge.

[0014] Preferably, the step of comparing the drift of the probability ridge extreme point with the preset initial lifetime reference extreme point includes: Extract the pre-stored preset initial lifetime reference extreme point, which includes the initial extreme point voltage and the initial extreme point capacity. The voltage drift is calculated by subtracting the voltage at the extreme point of the probability ridge extreme point from the voltage at the initial extreme point. The capacity drift is calculated by subtracting the capacity of the extreme point in the probability ridge extreme point from the capacity of the initial extreme point.

[0015] Preferably, the step of generating aging test results for indicating the power battery cell includes: Using the extreme boundary parameters pre-stored in the model parameter set, the voltage drift and the capacity drift are subjected to maximum-minimum normalization preprocessing. The normalized voltage drift and the normalized capacity drift are input into a preset aging state mapping model to perform forward inference, and a floating-point value representing the ratio of the actual usable capacity of the power battery cell to the factory rated capacity is output. The floating-point value is defined as the health status assessment value. The health status assessment value is determined as the aging test result used to indicate the power battery cell.

[0016] This invention also provides an online aging detection system for power battery cells, comprising: The parameter acquisition and configuration module is used to acquire the minimum effective voltage resolution and ambient noise floor amplitude of the analog-to-digital converter, determine the adaptive voltage step size based on the minimum effective voltage resolution and ambient noise floor amplitude, and set the phase transition active voltage window. The state slice positioning module is used to collect the operating physical quantity data of the power battery cells in the online operating state in real time, determine the pseudo steady state segment based on the operating physical quantity data, and set the initial voltage reference and initial charge accumulation value when the pseudo steady state segment is determined to be entered. The interlocking verification and extraction module is used to perform a charge accumulation operation on the initial charge accumulation value based on the running physical quantity data during the duration of the pseudo steady state segment to obtain an updated charge accumulation value. When the absolute change of the terminal voltage in the running physical quantity data relative to the initial voltage reference is greater than or equal to the adaptive voltage step size, the updated charge accumulation value is used to perform a charge quantity verification judgment. If the verification condition of the charge quantity verification judgment is met, the discrete differential capacity and characteristic voltage are calculated, and the initial voltage reference and the updated charge accumulation value are reset. The statistical matrix mapping module is used to determine whether the characteristic voltage is within the phase transition active voltage window. If the characteristic voltage is within the phase transition active voltage window, the characteristic voltage and the discrete differential capacity are mapped to the corresponding grid coordinates of a preset two-dimensional statistical matrix, and the count value corresponding to the corresponding grid coordinates is incremented. The aging measurement output module is used to traverse the preset two-dimensional statistical matrix after the count value increment operation, locate the probability ridge extreme point based on the maximum count value in the preset two-dimensional statistical matrix, compare the drift of the probability ridge extreme point with the preset early life reference extreme point, and generate the aging detection result for indicating the power battery cell.

[0017] This invention provides a method and system for online aging detection of power battery cells. It has the following beneficial effects: 1. This invention extracts an adaptive voltage step size by combining analog-to-digital converter resolution and environmental noise, and performs charge verification by extracting pseudo-steady-state segments from the operating data, thus eliminating interference from dynamic operating conditions on data acquisition. This method enables the power battery cells to effectively filter out transient noise and extract accurate capacity difference values ​​during actual operation. This overcomes the problem of conventional calculations being susceptible to fluctuation interference, improves the accuracy of subsequent online detection of aging evolution states, and lays a reliable data foundation for power battery fault prediction and health management.

[0018] 2. This invention achieves statistical analysis without relying on continuous charge-discharge timing by mapping characteristic voltage and discrete differential capacity to a two-dimensional statistical matrix and incrementing the count value. This step directly integrates short-range data generated during the operation of the power battery cell, solving the problem of being unable to extract complete aging characteristics due to the randomness of the charge-discharge interval. This method adapts to fragmented, random operating modes, ensuring the effective execution of the online detection process.

[0019] 3. This invention locates probabilistic extreme points by traversing a two-dimensional statistical matrix, performs reverse compensation restoration by combining the physical step size and lower limit parameters, and compares the result with an early-life baseline to calculate the health status. This compensation operation eliminates the quantization truncation error caused by discretization processing, making the characteristic coordinate restoration of the power battery cell more accurate. The internal aging state is analyzed based on the drift of the characteristic extreme points, effectively reducing the computational load and meeting the processing requirements of online detection. Attached Figure Description

[0020] Figure 1 This is a diagram illustrating the architecture of the online aging detection system for power battery cells according to the present invention. Figure 2 This is a flowchart of the online aging detection method for power battery cells according to the present invention; Figure 3 This is a logical diagram illustrating the initialization of hardware boundary parameters in this invention. Figure 4 This is a logical diagram illustrating the effective working condition segment and the setting of the reference in this invention. Figure 5 This is a logical diagram illustrating the physical constraint interlocking verification and discrete feature extraction of the present invention. Figure 6 This is a logical schematic diagram illustrating the voltage window filtering and two-dimensional statistical matrix mapping implementation of the present invention; Figure 7 This is a logical diagram illustrating the analytical probability extreme points and the measurement of aging drift in this invention; Figure 8 This is a bar chart comparing the aging assessment errors of different groups in this invention.

[0021] Among them, 10 is the parameter acquisition and configuration module; 20 is the state slice positioning module; 30 is the interlock verification and extraction module; 40 is the statistical matrix mapping module; and 50 is the aging measurement output module. Detailed Implementation

[0022] The technical solutions in the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. 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.

[0023] Reference Figure 1 The present invention provides an online aging detection system for power battery cells. The system includes a parameter acquisition and configuration module 10, a state slice positioning module 20, an interlock verification and extraction module 30, a statistical matrix mapping module 40, and an aging measurement output module 50.

[0024] The online aging detection system for the power battery cells is deployed in the microcontroller of the battery management system. The microcontroller uses a digital logic processor with a processing core and an on-chip bus architecture; the microcontroller is electrically connected to the front-end battery acquisition circuit via a communication bus; the front-end battery acquisition circuit integrates an analog-to-digital converter for acquiring the terminal voltage of the power battery cells; a current sensor is externally connected to the microcontroller for synchronously acquiring the loop current; and non-volatile memory is integrated internally or externally mounted on the microcontroller for storing two-dimensional statistical matrix data generated during long-term operation.

[0025] The parameter acquisition and configuration module 10 is connected to the analog-to-digital converter and the internal storage area of ​​the microcontroller, respectively. It is used to acquire the minimum effective voltage resolution of the analog-to-digital converter and the ambient noise floor amplitude, determine the adaptive voltage step size based on the minimum effective voltage resolution and the ambient noise floor amplitude, and set the phase transition active voltage window.

[0026] The data input terminal of the state slice positioning module 20 is connected to the front-end sensor acquisition channel to collect the operating physical quantity data of the power battery cell in the online operating state in real time, determine the pseudo steady state segment based on the operating physical quantity data, and set the initial voltage reference and initial charge accumulation value when the pseudo steady state segment is determined.

[0027] The data input terminal of the interlocking verification and extraction module 30 receives the variables transmitted by the state slice positioning module 20 and the parameters output by the parameter acquisition and configuration module 10. During the duration of the pseudo-steady state segment, it performs a charge accumulation operation on the initial charge accumulation value based on the running physical quantity data to obtain the updated charge accumulation value. When the absolute change of the terminal voltage in the running physical quantity data compared with the initial voltage reference is greater than or equal to the adaptive voltage step size, the updated charge accumulation value is used to perform charge quantity verification and judgment. If the verification conditions of charge quantity verification and judgment are met, the discrete differential capacity and characteristic voltage are calculated, and the initial voltage reference and the updated charge accumulation value are reset.

[0028] The input of the statistical matrix mapping module 40 receives the calculation results output by the interlocking verification and extraction module 30, which is used to determine whether the characteristic voltage is within the phase transition active voltage window. If the characteristic voltage is within the phase transition active voltage window, the characteristic voltage and discrete differential capacity are mapped to the corresponding grid coordinates of the preset two-dimensional statistical matrix, and the count value corresponding to the grid coordinates is incremented.

[0029] The aging measurement output module 50 is used to read the two-dimensional statistical matrix stored in the non-volatile memory, and traverse the two-dimensional statistical matrix after the count value increment operation. Based on the maximum count value in the two-dimensional statistical matrix, the extreme point of the probability ridge is located. The extreme point of the probability ridge is compared with the preset early life reference extreme point by the drift amount to generate the aging test result for indicating the power battery cell.

[0030] Reference Figure 2 This invention provides an online aging detection method for power battery cells. The online aging detection method for power battery cells is executed based on the aforementioned online aging detection system for power battery cells and includes the following steps: S1, obtain the minimum effective voltage resolution and ambient noise floor amplitude of the analog-to-digital converter, determine the adaptive voltage step size based on the minimum effective voltage resolution and ambient noise floor amplitude, and set the phase transition active voltage window.

[0031] S2 collects real-time physical quantity data of the power battery cells in online operation, determines pseudo-steady-state segments based on the physical quantity data, and sets the initial voltage reference and initial charge accumulation value when the pseudo-steady-state segment is determined.

[0032] S3. During the pseudo-steady state segment, the initial charge accumulation value is updated by performing a charge accumulation operation based on the running physical quantity data. When the absolute change of the terminal voltage in the running physical quantity data compared to the initial voltage reference is greater than or equal to the adaptive voltage step size, the updated charge accumulation value is used to perform charge quantity verification. If the verification condition of the charge quantity verification is met, the discrete differential capacity and characteristic voltage are calculated, and the initial voltage reference and the updated charge accumulation value are reset.

[0033] S4. Determine whether the characteristic voltage is within the phase transition active voltage window. If the characteristic voltage is within the phase transition active voltage window, map the characteristic voltage and discrete differential capacity to the corresponding grid coordinates of the preset two-dimensional statistical matrix and perform an auto-increment operation on the count value corresponding to the grid coordinates.

[0034] S5: Traverse the two-dimensional statistical matrix after the count value increment operation, locate the probability ridge extreme point based on the maximum count value in the two-dimensional statistical matrix, compare the drift of the probability ridge extreme point with the preset initial life benchmark extreme point, and generate the aging test result for indicating the power battery cell.

[0035] To further clarify the implementation of each technical aspect of the present invention, the following will provide a detailed description of the implementation of each functional module involved above and its internal processing flow.

[0036] Reference Figure 3 In this embodiment, step S1 includes the following sub-steps.

[0037] S101 reads the minimum effective voltage resolution of the analog-to-digital converter in the front-end battery acquisition circuit.

[0038] The parameter acquisition and configuration module 10 accesses the hardware configuration register of the analog-to-digital converter in the front-end battery acquisition circuit via the microcontroller's internal communication bus. The module 10 reads the reference voltage base value and the conversion sampling bit depth recorded in the hardware configuration register, and divides the reference voltage base value by the full-scale digital step number determined based on the conversion sampling bit depth to obtain the voltage resolution value. This voltage resolution value is defined as the minimum effective voltage resolution. The minimum effective voltage resolution characterizes the analog voltage span corresponding to a single digital change in the acquisition hardware.

[0039] In one embodiment, if the reference voltage value configured in the front-end battery acquisition circuit is 5V and the conversion sampling bit depth is 12 bits, then the full-scale digital step count is 2^12 minus 1, which is 4095. The calculated minimum effective voltage resolution is 5 / 4095 ≈ 0.0012V. This minimum effective voltage resolution value is used to characterize the smallest voltage change that the acquisition hardware can recognize.

[0040] S102, obtain the ambient background noise amplitude under static conditions.

[0041] When the external high-voltage relay where the power battery cell is located is in the open state, and the loop current collected by the externally connected current sensor remains zero, the parameter acquisition and configuration module 10 triggers the analog-to-digital converter to continuously collect a set number of power battery cell terminal voltage data at a preset sampling rate. In one embodiment, the preset sampling rate is configured to 1000Hz, and the set number is 1000 sample points, so that the sampled data covers at least 1 second of terminal voltage fluctuation process.

[0042] The parameter acquisition and configuration module 10 iterates through the aforementioned power battery cell terminal voltage data, extracts the maximum and minimum voltage values, and subtracts the minimum voltage value from the maximum voltage value to obtain the voltage difference. This voltage difference is determined as the ambient background noise amplitude. The ambient background noise amplitude characterizes the degree of random fluctuation in voltage measurement caused by circuit board distributed parameters and external electromagnetic coupling under static operating conditions where no electrochemical reaction occurs. By measuring the ambient background noise amplitude in a static environment, the voltage fluctuation range of the current acquisition channel can be obtained, providing a basis for subsequently setting anti-interference thresholds.

[0043] S103 divides the ambient background noise amplitude by the minimum effective voltage resolution and rounds it up. The rounded result is then incremented by 1 as the adaptive step coefficient.

[0044] The parameter acquisition and configuration module 10 calculates the adaptive step coefficient according to the following formula: ; In the formula, For adaptive step coefficients; This refers to the ambient background noise amplitude. Minimum effective voltage resolution; This is the symbol for the round-up operation.

[0045] Dividing the ambient noise floor amplitude by the minimum effective voltage resolution yields the discrete quantization order corresponding to the ambient noise floor amplitude. Rounding up is used to cover the quantization range corresponding to noise fluctuations; incrementing the rounded result by one adds a margin of one least significant bit to this quantization range. The resulting adaptive step coefficient is used to determine the subsequent adaptive voltage step size, reducing the impact of underlying acquisition noise on voltage change determination.

[0046] S104 multiplies the adaptive step coefficient by the minimum effective voltage resolution to obtain the adaptive voltage step size.

[0047] The parameter acquisition and configuration module 10 calculates the adaptive voltage step size according to the following formula: ; In the formula, This is for adaptive voltage step size.

[0048] The adaptive voltage step size is jointly determined by the ambient noise floor amplitude and the minimum effective voltage resolution of the analog-to-digital converter. In subsequent data processing, the adaptive voltage step size is used to determine whether the voltage at the battery cell terminals has undergone an effective change, thereby reducing the threshold judgment deviation caused by changes in the external electromagnetic environment when using a fixed empirical constant.

[0049] S105 retrieves the chemical system parameters of the power battery cell and sets the lower and upper voltage limits of the phase change active voltage window, which includes the aging characteristic range, based on the chemical system parameters.

[0050] The parameter acquisition and configuration module 10 reads the microcontroller's non-volatile memory to obtain the pre-written positive electrode material crystal structure identifier and negative electrode material crystal structure identifier. The positive electrode material crystal structure identifier and negative electrode material crystal structure identifier together constitute the chemical system parameters. The parameter acquisition and configuration module 10 uses the chemical system parameters as the addressing index to perform a lookup in a preset electrochemical characteristic mapping table.

[0051] In one embodiment, the preset electrochemical characteristic mapping table is generated based on the technical specifications or datasheets provided by the cell manufacturers for power battery cells with different chemical systems at the time of manufacture. The technical specifications or datasheets include incremental capacity curves (IC) or differential voltage curves (DV) under standard operating conditions, and specify the voltage ranges where the characteristic peaks characterizing material phase transitions are located. The parameter acquisition and configuration module 10 extracts the voltage ranges specified in the technical specifications or datasheets as the boundary values ​​of the characteristic voltage ranges, establishes a correspondence between the parameters of each chemical system and the corresponding boundary values ​​of the characteristic voltage ranges, generates the preset electrochemical characteristic mapping table, and then stores it in the microcontroller.

[0052] The parameter acquisition and configuration module 10 retrieves the boundary values ​​of the characteristic voltage range corresponding to the current chemical system parameters from a preset electrochemical characteristic mapping table. Under actual operating conditions, the power battery cell undergoes a lattice phase transition within a specific charge-discharge range. The voltage within this range exhibits identifiable characteristics as the amount of charge changes. Extracting data from this range helps in assessing the aging state. In one embodiment, for lithium iron phosphate batteries, the characteristic range of their charge-discharge platform is concentrated between 3.1V and 3.4V.

[0053] The characteristic voltage range boundary values ​​include a lower endpoint threshold and an upper endpoint threshold. The parameter acquisition and configuration module 10 sets the lower endpoint threshold as the lower voltage limit and the upper endpoint threshold as the upper voltage limit. The numerical range between the lower voltage limit and the upper voltage limit is defined as the phase transition active voltage window.

[0054] By setting a phase change active voltage window, the parameter acquisition and configuration module 10 limits the processing range of subsequent steps to the operating data segment of the power battery cell containing aging characteristics, thereby reducing the microcontroller's parsing and calculation of data in the non-phase change voltage range.

[0055] Reference Figure 4 In this embodiment, step S2 includes the following sub-steps.

[0056] S201 collects the terminal voltage and loop current of the power battery cell in real time at a fixed sampling time interval as operating physical quantity data, establishes a sliding time window, and calculates the rate of change of the loop current within the sliding time window.

[0057] The state slice positioning module 20 continuously reads the terminal voltage and loop current of the power battery cell through the front-end sensor acquisition channel of the microcontroller, according to a fixed sampling time interval pre-written to the timer interrupt. The terminal voltage and loop current together constitute the operating physical quantity data. In one embodiment, the fixed sampling time interval is set to 100 milliseconds to balance the computational load of the microcontroller and the sampling accuracy of the physical quantity.

[0058] The state slice positioning module 20 allocates a contiguous address space in the microcontroller's random access memory to construct a first-in-first-out (FIFO) data queue. This data queue stores the physical quantity data in the order of sampling time and forms a sliding time window in the time dimension. The length of the sliding time window is determined by the fixed sampling time interval and the preset depth of the data queue.

[0059] The state slice positioning module 20 calculates the current change rate according to the following formula: ; In the formula, The rate of change of current; The loop current is the current at the latest moment within the sliding time window; The loop current is the oldest time within the sliding time window; The duration of the sliding time window.

[0060] The calculated rate of change of current is used to characterize the magnitude of change of loop current within the sliding time window.

[0061] S202, determine whether the absolute value of the current change rate is less than or equal to the preset current fluctuation threshold, and determine whether the absolute value of the loop current is greater than or equal to the preset effective operating current lower limit.

[0062] The state slice positioning module 20 performs absolute value calculation on the calculated current change rate to obtain the absolute value of the current change rate, and compares the absolute value of the current change rate with the preset current fluctuation threshold.

[0063] The preset current fluctuation threshold and the preset effective operating current lower limit are calculated based on the rated capacity ratio specified in the power battery cell's manufacturer's specifications. In one embodiment, the rated capacity of the power battery cell is CN, where CN is in Ah; the preset current fluctuation threshold is set to 0.05 × CN, in A / s; and the preset effective operating current lower limit is set to 0.1 × CN, in A.

[0064] The state slice positioning module 20 performs absolute value calculation on the loop current at the current moment to obtain the absolute value of the loop current, and determines whether the absolute value of the loop current is greater than or equal to the preset effective lower limit of the operating current.

[0065] S203: When the absolute value of the current change rate is less than or equal to the current fluctuation threshold, and the absolute value of the loop current is greater than or equal to the lower limit of the effective operating current, the power battery cell is determined to have entered a pseudo-steady state segment.

[0066] The state slice positioning module 20 combines the above two judgment conditions through a logical AND operation; when the absolute value of the current change rate is less than or equal to the current fluctuation threshold, and the absolute value of the loop current is greater than or equal to the lower limit of the effective operating current, the state slice positioning module 20 determines that the power battery cell has entered a pseudo steady state segment.

[0067] If the absolute value of the current change rate is greater than the preset current fluctuation threshold, or the absolute value of the loop current is less than the preset effective operating current lower limit, the state slice positioning module 20 determines that the power battery cell is not in a pseudo steady state segment, discards the operating physical quantity data in the current sliding time window, and continues to wait for the next acquisition and judgment at a fixed sampling time interval.

[0068] When the state slice positioning module 20 has determined that the power battery cell has entered the pseudo-steady state segment, but the subsequent sampling period no longer meets the pseudo-steady state segment determination conditions, the state slice positioning module 20 outputs a flag bit to exit the pseudo-steady state segment, clears the contents of the register used to record the initial voltage reference, and clears the variable used to record the initial charge accumulation value to zero.

[0069] The pseudo-steady-state segment characterizes the operating condition range where the rate of change of the circuit current is small and the circuit current reaches the lower limit of the effective operating current. Based on the above logic, the state slice positioning module 20 excludes the operating condition range corresponding to frequent fluctuations in external load, as well as the operating condition range corresponding to quiescent or small dark current.

[0070] S204 assigns the terminal voltage at the time of entering the pseudo-steady state segment to the initial voltage reference and assigns the initial charge accumulation value to zero.

[0071] When the conditions for entering the pseudo-steady state segment are met, the state slice positioning module 20 extracts the voltage value of the power battery cell at the current sampling time and writes the voltage value into the register used to record the reference state as the initial voltage reference.

[0072] Meanwhile, the state slice positioning module 20 allocates a double-precision floating-point variable in the memory variable area to record the amount of charge, and assigns 0 to the double-precision floating-point variable; the 0 recorded by the double-precision floating-point variable at the current moment is determined as the initial charge accumulation value.

[0073] The initial voltage reference and the initial charge accumulation value are used together as the starting parameters for calculating charge accumulation and voltage offset within the current pseudo-steady-state segment. This initialization operation is designed to reduce the impact of historical operating conditions prior to entering the current pseudo-steady-state segment on subsequent feature extraction.

[0074] Reference Figure 5 In this embodiment, step S3 includes the following sub-steps.

[0075] S301 During the duration of the pseudo-steady-state segment, the fixed sampling time interval is multiplied by the loop current and converted into ampere-hours to obtain the single charge quantity. The single charge quantity is then used to perform a successive accumulation operation on the initial charge accumulation value to obtain the updated charge accumulation value.

[0076] When the confirmation state slice positioning module 20 continuously outputs the flag indicating a pseudo-steady-state segment, the interlock verification and extraction module 30 extracts the loop current value within the current acquisition cycle. The interlock verification and extraction module 30 calculates the single charge amount according to the following formula: ; In the formula, This refers to the amount of charge in a single charge. This represents the current loop current. The sampling time interval is fixed.

[0077] When the fixed sampling time interval is used in the calculation in seconds, and The product of the two is divided by 3600 to unify the single charge quantity and the updated charge accumulation value into the Ah dimension, and to unify the subsequent discrete differential capacity and capacity grid parameters into the Ah / V dimension. The calculated single charge quantity represents the amount of charge flowing into or out of the power battery cell within the fixed sampling time interval.

[0078] After acquiring the single charge quantity, the interlocking verification and extraction module 30 adds the single charge quantity to a double-precision floating-point variable that carries the initial charge accumulation value. The latest value recorded by this double-precision floating-point variable after the single charge accumulation is defined as the updated charge accumulation value. The above successive accumulation operation is performed continuously according to the sampling period within the pseudo-steady-state segment to record the cumulative charge transfer amount since the initial voltage reference setting.

[0079] S302, subtract the initial voltage reference from the current terminal voltage to obtain the instantaneous voltage difference, and extract the absolute value of the instantaneous voltage difference as the absolute change.

[0080] The interlocking verification and extraction module 30 extracts the latest terminal voltage from the real-time operating physical quantity data and reads the initial voltage reference recorded in the register. The interlocking verification and extraction module 30 performs calculations according to the following formula: ; In the formula, It is an absolute change; This represents the current terminal voltage; As the initial voltage reference; This is the absolute value operator.

[0081] The instantaneous voltage difference is obtained by subtracting the terminal voltage from the initial voltage reference. Taking the absolute value of this instantaneous voltage difference yields the offset of the terminal voltage relative to the initial voltage reference. This absolute change characterizes the magnitude of the terminal voltage change in the power battery cell since entering the pseudo-steady-state segment or since the last initial voltage reference reset.

[0082] S303 triggers charge quantity verification when the absolute change is greater than or equal to the adaptive voltage step size.

[0083] After calculating the absolute change each time, the interlocking verification and extraction module 30 compares the absolute change with the adaptive voltage step size determined in the previous step. When the absolute change is greater than or equal to the adaptive voltage step size, it indicates that the change in the battery cell terminal voltage relative to the initial voltage reference has exceeded the judgment scale determined by the ambient background noise amplitude and the minimum effective voltage resolution. This terminal voltage change can be used as a trigger condition for further judging the charge transfer situation. At this time, the interlocking verification and extraction module 30 triggers the charge quantity verification judgment.

[0084] When the absolute change is less than the adaptive voltage step size, it indicates that the current voltage change is still within the range of the adaptive voltage step size. The interlocking verification and extraction module 30 does not trigger the charge quantity verification judgment and continues to perform physical quantity data acquisition and charge accumulation operations in the pseudo steady state segment.

[0085] S304. Extract the absolute value of the updated charge accumulation value as the verification comparison value, and determine whether the verification comparison value is greater than or equal to the preset lower limit threshold of charge integration, so as to complete the charge quantity verification judgment.

[0086] After the charge quantity verification is triggered, the interlock verification and extraction module 30 extracts the absolute value of the double-precision floating-point variable that carries the current updated charge accumulation value to obtain the verification comparison value.

[0087] In one embodiment, the preset lower limit threshold for charge integration is determined by multiplying the rated capacity parameter specified in the battery cell's manufacturer's specifications by a proportional coefficient. This proportional coefficient is set to 0.0001, and the unit of the preset lower limit threshold for charge integration is Ah. The corresponding preset lower limit threshold for charge integration is pre-written into a read-only memory.

[0088] The interlocking verification and extraction module 30 compares the verification comparison value with a preset lower limit threshold for charge integration. This comparison is used to determine whether the power battery cell synchronously undergoes charge transfer that meets the lower limit requirement when the terminal voltage reaches the adaptive voltage step size, thereby reducing misjudgments caused by connection fluctuations, instantaneous load changes, or polarization voltage changes.

[0089] S305, when the verification comparison value is greater than or equal to the lower limit threshold of charge integral, the verification condition for charge quantity verification is confirmed to be met.

[0090] If the verification comparison value is greater than or equal to the preset lower limit threshold of charge integration, it indicates that there is a charge transfer amount that meets the lower limit requirement within the sampling interval corresponding to the current voltage change. The interlock verification and extraction module 30 confirms that the current data meets the verification conditions for charge quantity verification.

[0091] If the verification comparison value is less than the preset lower limit threshold of charge integration, it indicates that the charge transfer amount corresponding to the current voltage change is insufficient. The interlock verification and extraction module 30 confirms that the current data does not meet the verification conditions for charge quantity verification. At this time, the interlock verification and extraction module 30 discards the current updated charge accumulation value, skips the calculation step of discrete differential capacity, and performs a reset operation on the initial voltage reference and the updated charge accumulation value to reduce the impact of insufficient charge transfer data on subsequent discrete differential capacity extraction.

[0092] S306 divides the verification comparison value when the verification condition is met by the adaptive voltage step size to obtain the discrete differential capacity, and adds the current terminal voltage to the initial voltage reference and divides it by two to obtain the characteristic voltage.

[0093] After confirming that the verification conditions are met, the interlock verification and extraction module 30 extracts the currently retained valid parameters and performs differential operations. Specifically, the interlock verification and extraction module 30 calculates the discrete differential capacity according to the following formula: ; In the formula, For discrete difference capacity; The verification comparison value when the verification conditions are met; This is for adaptive voltage step size.

[0094] Discrete differential capacity characterizes the charge throughput of a power battery cell within an adaptive voltage step range, without distinguishing between charging and discharging directions, and is used to reflect the differential capacity characteristics within that voltage range.

[0095] Meanwhile, the interlocking verification and extraction module 30 calculates the characteristic voltage according to the following formula: ; In the formula, Characteristic voltage; This represents the current terminal voltage; This serves as the initial voltage reference.

[0096] The current terminal voltage is added to the initial voltage reference and then divided by two to obtain the center voltage value of the voltage variation range. This center voltage value is used as the characteristic voltage to represent the voltage position corresponding to the discrete differential capacity.

[0097] S307 reassigns the current terminal voltage to the initial voltage reference and reassigns the updated charge accumulation value to zero to complete the operation of resetting the initial voltage reference and the updated charge accumulation value.

[0098] After calculating the discrete differential capacity and characteristic voltage, the interlocking verification and extraction module 30 extracts the latest acquired terminal voltage and overwrites the terminal voltage into the register that records the initial voltage reference; at the same time, the interlocking verification and extraction module 30 clears the double-precision floating-point variable that carries the updated charge accumulation value to zero.

[0099] Through the above reset operation, the current terminal voltage is used as the initial voltage reference for judging the next voltage change, and the updated charge accumulation value starts to accumulate again from zero. This allows the interlocking verification and extraction module 30 to segment the terminal voltage change interval according to the adaptive voltage step size during the same pseudo-steady-state segment, and continuously extract the discrete differential capacity and characteristic voltage.

[0100] Reference Figure 6 In this embodiment, step S4 includes the following sub-steps.

[0101] S401, determine whether the characteristic voltage is greater than or equal to the lower limit of the phase transition active voltage window, and determine whether the characteristic voltage is less than or equal to the upper limit of the phase transition active voltage window, so as to confirm whether the characteristic voltage is within the phase transition active voltage window.

[0102] After the interlocking verification and extraction module 30 outputs the discrete differential capacity and characteristic voltage, the statistical matrix mapping module 40 extracts the characteristic voltage and compares it with the boundary parameters of the phase transition active voltage window configured in the previous step.

[0103] When the characteristic voltage is greater than or equal to the lower voltage limit and less than or equal to the upper voltage limit, the statistical matrix mapping module 40 determines that the characteristic voltage is within the phase transition active voltage window. If the characteristic voltage is less than the lower voltage limit or greater than the upper voltage limit, the statistical matrix mapping module 40 determines that the characteristic voltage is not within the phase transition active voltage window.

[0104] For feature voltages not within the phase transition active voltage window, the statistical matrix mapping module 40 discards the current feature voltage and its corresponding discrete differential capacity, and waits for the next set of feature voltages and discrete differential capacities to be input. Through the above judgment, the statistical matrix mapping module 40 limits the subsequent statistical range to the data interval corresponding to the phase transition active voltage window, so as to reduce the occupation of the two-dimensional statistical matrix by data in the non-phase transition voltage interval.

[0105] S402, if the characteristic voltage is within the phase transition active voltage window, subtract the lower limit of the voltage from the characteristic voltage to obtain the characteristic voltage difference, and subtract the preset lower limit of the capacity grid from the discrete differential capacity to obtain the capacity difference.

[0106] After confirming that the characteristic voltage is within the phase transition active voltage window, the statistical matrix mapping module 40 performs coordinate translation calculations. The statistical matrix mapping module 40 subtracts the lower limit of the phase transition active voltage window from the characteristic voltage to obtain the characteristic voltage difference.

[0107] Simultaneously, the statistical matrix mapping module 40 extracts a preset capacity grid lower limit. In one embodiment, the preset capacity grid lower limit is determined based on the standard differential capacity curve provided in the power battery cell's manufacturer's specifications. The minimum differential capacity value within the corresponding phase transition active voltage window in the standard differential capacity curve is set as the preset capacity grid lower limit, which is 0.5 Ah / V. The statistical matrix mapping module 40 subtracts the preset capacity grid lower limit from the discrete differential capacity to obtain the capacity difference.

[0108] Through the above subtraction operation, the characteristic voltage and discrete differential capacity are converted into offsets relative to the lower limit of voltage and the lower limit of capacity grid, respectively, providing a data basis for the subsequent generation of horizontal axis storage index and vertical axis storage index.

[0109] S403: Divide the characteristic voltage difference by the adaptive voltage step size and round down to generate the horizontal axis storage index; divide the capacity difference by the preset capacity grid resolution and round down to generate the vertical axis storage index.

[0110] After obtaining the characteristic voltage difference and capacity difference, the statistical matrix mapping module 40 performs storage index calculation. The statistical matrix mapping module 40 generates the horizontal axis storage index according to the following formula: ; In the formula, Store indexes for the horizontal axis; The characteristic voltage difference; For adaptive voltage step size; This is the symbol for rounding down.

[0111] Simultaneously, the statistical matrix mapping module 40 generates the vertical axis storage index according to the following formula: ; In the formula, Store indexes for the vertical axis; This is the capacity difference; The preset capacity grid resolution; This is the symbol for rounding down.

[0112] The preset capacity grid resolution represents the capacity span value corresponding to each statistical grid on the vertical axis. In one embodiment, the statistical matrix mapping module 40 extracts the maximum differential capacity value and the minimum differential capacity value within the corresponding phase change active voltage window from the power battery cell's specifications and calculates the difference between them; then, it divides the difference by the maximum number of grids allocated by the microcontroller on the vertical axis to obtain the preset capacity grid resolution.

[0113] For example, when the number of grid cells on the vertical axis is set to 128, the preset capacity grid resolution can be set to 0.05 Ah / V. By dividing the characteristic voltage difference and capacity difference by the corresponding resolution parameters and performing a round-down operation, the statistical matrix mapping module 40 converts continuous physical quantities into discrete array indices.

[0114] In one embodiment, before locating a preset two-dimensional statistical matrix using the horizontal and vertical axis storage indices, the statistical matrix mapping module 40 first determines whether the horizontal axis storage index is greater than or equal to 0 and less than the preset number of columns in the two-dimensional statistical matrix, and determines whether the vertical axis storage index is greater than or equal to 0 and less than the preset number of rows in the two-dimensional statistical matrix. If either the horizontal or vertical axis storage index exceeds the corresponding range, the statistical matrix mapping module 40 discards the current characteristic voltage and discrete differential capacity; if both the horizontal and vertical axis storage indices are within the corresponding range, the statistical matrix mapping module 40 continues to perform the corresponding grid coordinate positioning.

[0115] S404: Concatenate the horizontal axis storage index and the vertical axis storage index to locate the corresponding grid coordinates in the preset two-dimensional statistical matrix, and perform an auto-increment operation on the count value corresponding to the grid coordinates.

[0116] After generating the horizontal axis storage index and the vertical axis storage index, the statistical matrix mapping module 40 combines the horizontal axis storage index and the vertical axis storage index into a row and column index, and uses this index to locate the corresponding grid coordinates in the preset two-dimensional statistical matrix.

[0117] In one embodiment, a preset two-dimensional statistical matrix is ​​generated during the system initialization phase. The microcontroller sets the number of matrix columns based on the maximum possible value of the horizontal axis storage index, which is the maximum possible value of the horizontal axis storage index plus 1. This maximum value is determined based on the width of the phase transition active voltage window and the adaptive voltage step size. The microcontroller also sets the number of matrix rows based on the maximum possible value of the vertical axis storage index, which is the maximum possible value of the vertical axis storage index plus 1. Subsequently, the microcontroller allocates a contiguous two-dimensional array space in non-volatile memory corresponding to the number of matrix rows and columns, and initializes all elements in this contiguous two-dimensional array space to 0, thereby generating the preset two-dimensional statistical matrix.

[0118] After locating the corresponding grid coordinates, the statistical matrix mapping module 40 reads the original count value stored in the corresponding grid coordinates, increments the original count value by 1 to obtain the updated count value, and writes the updated count value back to the storage address of the corresponding grid coordinates. Thus, the characteristic voltage and discrete differential capacity are mapped to the corresponding grid coordinates in the preset two-dimensional statistical matrix, and the count value increment operation is completed.

[0119] Locating a continuous two-dimensional array space using row and column indices and performing data read and write operations is a common implementation method in microcontroller data processing. Those skilled in the art can complete the corresponding instruction configuration based on C language or assembly language.

[0120] Through the above statistical mapping method, the statistical matrix mapping module 40 can accumulate the dispersed characteristic voltage and discrete differential capacity into a two-dimensional statistical matrix without relying on continuous long-term charge and discharge conditions, thereby forming a statistical distribution feature that reflects the state of the power battery cell.

[0121] Reference Figure 7 In this embodiment, step S5 includes the following sub-steps.

[0122] S501, when the preset evaluation period is reached, traverse the two-dimensional statistical matrix after the count value increment operation to obtain the maximum count value, and extract the specific grid coordinates, horizontal axis storage index and vertical axis storage index corresponding to the maximum count value.

[0123] The preset evaluation period is used to limit the time or usage conditions for performing aging state detection. In one embodiment, the preset evaluation period is set to a cumulative operation of 30 days or an equivalent full charge-discharge cycle of 50 times. When the actual operation record of the power battery cell reaches the preset evaluation period, the aging measurement output module 50 begins to traverse the two-dimensional statistical matrix and reads all the count values ​​in the two-dimensional statistical matrix after the count value increment operation through nested loop instructions.

[0124] During the traversal, the aging measurement output module 50 compares the currently read count value with the historical maximum count value temporarily stored in a temporary register. This temporary register is initialized to 0 before the traversal begins. When the currently read count value is greater than the historical maximum count value, the aging measurement output module 50 writes the currently read count value into the temporary register and simultaneously records the corresponding horizontal axis storage index and vertical axis storage index.

[0125] After traversing the two-dimensional statistical matrix, the aging measurement output module 50 checks whether the maximum count value recorded in the temporary register is 0. If the maximum count value is 0, it indicates that no effective feature mapping has been formed within the current preset evaluation period. The aging measurement output module 50 does not output the aging detection result for this time, but retains the count value in the two-dimensional statistical matrix, waiting for subsequent feature data to continue to accumulate.

[0126] If the maximum count value is greater than 0, the aging measurement output module 50 extracts the horizontal and vertical storage indices corresponding to the maximum count value. The horizontal and vertical storage indices together determine the specific grid coordinates where the maximum count value is located, and these specific grid coordinates correspond to the feature region with the highest frequency of occurrence in the two-dimensional statistical matrix.

[0127] S502 utilizes the horizontal and vertical storage indices corresponding to specific grid coordinates, and performs reverse compensation restoration in conjunction with the physical step size and lower limit parameters to generate extreme point voltage and extreme point capacity, thereby determining the probability ridge extreme point.

[0128] After obtaining specific grid coordinates, the aging measurement output module 50 extracts the horizontal and vertical axis storage indices from these coordinates to perform a reverse calculation of the physical coordinates. The aging measurement output module 50 reconstructs the extreme point voltage according to the following formula: ; In the formula, The voltage at the extreme point; Store an index for the horizontal axis corresponding to a specific grid coordinate; For adaptive voltage step size; This is the lower limit of the phase transition active voltage window.

[0129] Simultaneously, the aging measurement output module 50 restores the extreme point capacity according to the following formula: ; In the formula, Capacity at extreme points; Store an index for the vertical axis corresponding to a specific grid coordinate; The preset capacity grid resolution; This is the preset lower limit of the capacity grid.

[0130] In the above restoration process, the starting physical position of the corresponding grid is obtained by multiplying the storage index by the corresponding adaptive voltage step size or the preset capacity grid resolution, and then adding the corresponding lower limit parameter. The constant 0.5 introduced in the formula represents a 50% compensation ratio. and The half-step physical compensation amount represents the corresponding dimension and is used to accurately restore the extreme point voltage and extreme point capacity from the grid starting boundary to the center position of the corresponding grid, so as to eliminate the quantization truncation error caused by the discretization rounding operation.

[0131] The calculated extreme point voltage and extreme point capacity form a two-dimensional physical coordinate pair, which is determined as the probability ridge extreme point. The probability ridge extreme point refers to the center physical coordinate point of the grid corresponding to the maximum count value in the two-dimensional statistical matrix, and is used to characterize the differential capacity feature location that occurs most frequently in the current evaluation period.

[0132] S503 calculates the voltage drift and capacity drift by subtracting the extreme point voltage and capacity from the corresponding parameters in the preset initial lifespan reference extreme point.

[0133] The aging measurement output module 50 reads the initial lifespan reference extreme point pre-stored in the microcontroller's non-volatile memory. The preset initial lifespan reference extreme point includes the initial extreme point voltage and the initial extreme point capacity. This parameter set is the initial differential characteristic peak coordinate constant captured by standard constant current charge-discharge test when the power battery leaves the factory or is initially assembled and rolled off the production line.

[0134] In one embodiment, when the power battery cell is a lithium iron phosphate cell and the phase change active voltage window is set to 3.1V to 3.4V, the initial extreme point voltage is set to 3.25V and the initial extreme point capacity is set to 12.5Ah / V.

[0135] The aging measurement output module 50 subtracts the initial extreme point voltage from the current extreme point voltage to obtain the voltage drift; and subtracts the initial extreme point capacity from the current extreme point capacity to obtain the capacity drift. The above difference process is used to compare the drift between the probability ridge extreme point and the preset initial lifespan reference extreme point.

[0136] Voltage drift is used to characterize the change in phase transition characteristic voltage during the aging process of power battery cells; capacity drift is used to characterize the change in differential capacity peak value during the aging process of power battery cells. The aging measurement output module 50 characterizes the aging state changes of power battery cells together with voltage drift and capacity drift.

[0137] S504 inputs the voltage drift and capacity drift into a preset aging state mapping model to perform forward inference and generate aging test results to indicate the power battery cell.

[0138] The aging measurement output module 50 extracts the calculated voltage drift and capacity drift as input data and submits them to a preset aging state mapping model for calculation. In this embodiment, the preset aging state mapping model adopts a lightweight multilayer perceptron regression model, which includes an input layer, a first fully connected layer, a second fully connected layer, and an output layer connected in sequence. The input layer contains two neuron nodes, which receive the normalized voltage drift and the normalized capacity drift, respectively. The first fully connected layer contains 64 neuron nodes and uses the ReLU activation function. The second fully connected layer contains 32 neuron nodes and uses the ReLU activation function. The output layer contains one neuron node and uses the Sigmoid activation function.

[0139] Before inputting the lightweight multilayer perceptron regression model, the aging measurement output module 50 uses the extreme boundary parameters pre-stored in the model parameter set to perform max-min normalization preprocessing on the voltage drift and capacity drift, mapping them to a numerical range of 0 to 1. In one embodiment, the normalization lower limit for the voltage drift is set to -0.5V, and the normalization upper limit is set to 0.5V; the normalization lower limit for the capacity drift is set to -5.0Ah / V, and the normalization upper limit is set to 0Ah / V.

[0140] The normalized preprocessed feature data undergoes matrix operations of weight multiplication and bias addition layer by layer between the input layer, the first fully connected layer, the second fully connected layer, and the output layer. The first fully connected layer is used for nonlinear feature extraction of the normalized voltage and capacity drift. The second fully connected layer further maps the feature data output from the first fully connected layer. The output layer uses a sigmoid activation function to constrain the output result to a value between 0 and 1. The floating-point value derived from the model output layer represents the ratio of the actual usable capacity of the power battery cell to its factory rated capacity; this floating-point value is defined as a health status assessment value.

[0141] The aging measurement output module 50 determines the health status assessment value as the aging test result of the power battery cell and reports it to the controller for subsequent energy management strategies and fault prediction and health management system calls.

[0142] After completing the model inference, the aging measurement output module 50 clears all count values ​​in the two-dimensional statistical matrix to zero, resets the cumulative state, and enters the next preset evaluation cycle.

[0143] In practical applications, the pre-set aging state mapping model is deployed and run after offline training. The specific training process includes building a pre-training dataset, which covers the full life cycle aging data of at least 500 power battery cells of the same model within an ambient temperature range of 5℃ to 45℃ and a charge / discharge rate range of 0.2C to 1C.

[0144] In an offline environment, periodic charge-discharge cycle aging tests were performed on power battery cells from the same batch but in different aging states to generate a pre-training dataset. Using the current and voltage data recorded during the test, the voltage drift and capacity drift corresponding to each cycle stage were obtained according to the same feature extraction logic described above, and the voltage drift and capacity drift were used as input data for training samples.

[0145] Meanwhile, the actual usable capacity of the power battery cells at each cycle stage is determined using the ampere-hour integration method, and the ratio of the actual usable capacity to the factory rated capacity is used as the corresponding label data.

[0146] In the model training phase, the pre-training dataset is divided into training and test sets in an 8:2 ratio. Mean squared error is used as the loss function to measure the difference between the network's forward output values ​​and the label data. The Adam optimization algorithm is used to iteratively update the weight matrix and bias parameters in the network layers based on the error backpropagation mechanism.

[0147] In one embodiment, the learning rate of the Adam optimization algorithm is set to 0.001, the batch size is set to 32, and the maximum number of training epochs is set to 500. When the mean squared error on the test set is below 0.001 for 10 consecutive training epochs, the training stop condition is triggered, the frozen network parameters are extracted, and they are embedded into the microcontroller's underlying code.

[0148] For forward propagation computation of matrix multiplication and addition and gradient-based backpropagation computation in multilayer perceptrons, those skilled in the art can implement them based on the underlying operators of conventional deep learning frameworks.

[0149] To further illustrate the implementation process and technical effects of this invention, the following description is provided in conjunction with specific application scenarios and experimental data. The specific numerical values, scenario parameters, and comparative experiments described below are only used to explain the implementation principle of this invention and do not limit the scope of protection of this invention.

[0150] For a lithium iron phosphate power battery cell with a rated capacity of 100Ah, the power battery cell is connected to a microcontroller and a front-end battery acquisition circuit to perform an online aging detection method for the power battery cell.

[0151] Phase 1: Initializing Hardware Boundary Parameters The parameter acquisition and configuration module 10 reads that the minimum effective voltage resolution of the analog-to-digital converter is 0.0012V. Under static conditions, 1000 terminal voltage samples of the power battery cells are continuously collected, the maximum and minimum voltage values ​​are extracted, and the ambient background noise amplitude is calculated to be 0.003V.

[0152] The parameter acquisition and configuration module 10 calculates the adaptive step coefficient according to the following formula: ; The adaptive step size is obtained as 4. The parameter acquisition and configuration module 10 calculates the adaptive voltage step size according to the following formula: ; The adaptive voltage step size was 0.0048V.

[0153] The parameter acquisition and configuration module 10 retrieves the chemical system parameters and sets the lower limit of the phase transition active voltage window to 3.1V and the upper limit to 3.4V.

[0154] Phase Two: Locating Effective Operating Condition Segments and Setting Benchmarks The state slice positioning module 20 acquires operational physical quantity data at a fixed sampling time interval of 100 milliseconds. At a certain moment, the loop current within the sliding time window stabilizes at 30A. The state slice positioning module 20 calculates that the absolute value of the current change rate is 0A / s, which is less than the preset current fluctuation threshold (5A / s); the absolute value of the loop current is 30A, which is greater than the preset effective operating current lower limit (10A).

[0155] The state slice positioning module 20 determines that the power battery cell has entered a pseudo-steady state segment. The state slice positioning module 20 assigns the terminal voltage of 3.3000V at the time when it is determined to have entered the pseudo-steady state segment to the initial voltage reference, and assigns the initial charge accumulation value to zero.

[0156] Phase 3: Perform physical constraint interlocking verification and extract discrete features. During the pseudo-steady-state segment, the loop current remains at 30A, with the discharge current in the positive direction in this embodiment. The interlock verification and extraction module 30 calculates the single charge amount every 100 milliseconds using the following formula: ; After 60 sampling cycles, the terminal voltage drops to 3.2952V. The interlock verification and extraction module 30 calculates according to the following formula: ; The extracted absolute change is equal to the adaptive voltage step size (0.0048V), triggering the charge quantity verification and determination.

[0157] After 60 accumulation operations, the updated charge accumulation value is 0.05Ah. The absolute value of the updated charge accumulation value is extracted as the verification comparison value (0.05Ah). This verification comparison value is greater than the preset lower limit threshold of charge integration (0.01Ah), confirming that the verification condition for charge quantity verification is met.

[0158] The interlocking verification and extraction module 30 calculates the discrete differential capacity according to the following formula: ; The interlocking verification and extraction module 30 calculates the characteristic voltage according to the following formula: ; After completing the calculation, the interlock verification and extraction module 30 reassigns 3.2952V to the initial voltage reference and reassigns the updated charge accumulation value to zero to complete the operation of resetting the initial voltage reference and the updated charge accumulation value.

[0159] Phase 4: Implementing voltage window screening and two-dimensional statistical matrix mapping The statistical matrix mapping module 40 determines that the characteristic voltage (3.2976V) is greater than the lower voltage limit (3.1V) and less than the upper voltage limit (3.4V), confirming that the characteristic voltage is within the phase transition active voltage window.

[0160] The statistical matrix mapping module 40 subtracts the lower voltage limit from the characteristic voltage to obtain a characteristic voltage difference of 0.1976V; and subtracts the preset capacity grid lower limit (0.5Ah / V) from the discrete differential capacity to obtain a capacity difference of 9.916Ah / V.

[0161] The preset capacity grid resolution is 0.05 Ah / V. In this embodiment, the number of grid cells on the vertical axis of the two-dimensional statistical matrix is ​​set to 256, so that the vertical axis storage index 198 is within the range of the number of rows in the two-dimensional statistical matrix. The statistical matrix mapping module 40 generates the horizontal axis storage index according to the following formula: ; The statistical matrix mapping module 40 generates the vertical axis storage index according to the following formula: ; The statistical matrix mapping module 40 locates the corresponding grid coordinates of row number 198 and column number 41 in the preset two-dimensional statistical matrix and performs an auto-increment operation on the count value corresponding to the grid coordinates.

[0162] Phase 5: Analyzing probabilistic extreme points and measuring aging drift After the preset evaluation period is reached, the aging measurement output module 50 traverses the two-dimensional statistical matrix after the count value increment operation, obtains the maximum count value, and extracts the specific grid coordinates, horizontal axis storage index, and vertical axis storage index corresponding to the maximum count value. The horizontal axis storage index is read as 41, and the vertical axis storage index is read as 198.

[0163] The aging measurement output module 50 restores the extreme point voltage according to the following formula: ; The aging measurement output module 50 restores the extreme point capacity according to the following formula: ; The aging measurement output module 50 reads the initial extreme point voltage (3.25V) and initial extreme point capacity (12.5Ah / V) from the preset initial lifespan reference extreme point.

[0164] The aging measurement output module 50 calculates the voltage drift: ; The aging measurement output module 50 calculates the capacity drift. ; The aging measurement output module 50 normalizes the voltage drift (0.0492V) and capacity drift (-2.075Ah / V) and inputs them into the preset aging state mapping model to perform forward inference, generating a health status assessment value of 0.882 to indicate the aging test results of the power battery cell. This value represents that the current actual usable capacity is 88.2% of the factory rated capacity.

[0165] Experimental verification and effect comparison: A charge-discharge cycle aging comparative experiment was designed to improve the online aging detection method for power battery cells. Ten power battery cells from the same batch were selected for testing. The experimental environment temperature was controlled at 25℃.

[0166] The experiment employs an independent loop control method. An independent test loop is used to perform the actual usable capacity calibration test. The loop current is precisely controlled and fed back through the closed-loop control circuit within the charge / discharge test equipment, ensuring that the acquisition of the benchmark reference value is unaffected by external load fluctuations. Within the 1500 cycles of the entire lifespan, an actual usable capacity calibration test is performed every 100 cycles: the battery is charged to a preset full charge voltage using a constant loop current of 0.05C using the charge / discharge test equipment, and then discharged to a preset discharge cutoff voltage using a constant loop current of 0.05C after resting. The actual usable capacity of the battery cell at the current cycle stage is calculated cumulatively using the ampere-hour integration method throughout the entire slow charge and discharge process, and this capacity is used as the benchmark reference value.

[0167] The experimental data extraction and processing were divided into three groups: Comparison Group 1: The evaluation results are calculated based on the accumulated current data at a fixed sampling time interval using the traditional ampere-hour integration method. Comparison Group 2: The offline incremental capacity analysis method was used to extract characteristic voltage and discrete differential capacity based on complete constant current charging data and obtain evaluation results; Test group: The aforementioned online aging detection method for power battery cells was implemented. The characteristic voltage and discrete differential capacity were extracted using only the irregular online operation physical quantity data during the discharge stage, and a health status assessment value was generated to indicate the aging detection results of the power battery cells.

[0168] Extract the results data from the three groups, calculate the difference between each data point and the benchmark reference value, and divide by the benchmark reference value to obtain the maximum absolute error and the mean absolute error for each of the three groups.

[0169] Reference Figure 8 The test data distribution is as follows: the maximum absolute error of comparison group 1 is 8.5%, and the average absolute error is 5.2%; the maximum absolute error of comparison group 2 is 2.1%, and the average absolute error is 1.3%; the maximum absolute error of the test group is 2.8%, and the average absolute error is 1.7%.

[0170] Reference Figure 8 The data distribution results show that: The maximum absolute error (2.8%) and average absolute error (1.7%) of the test group were smaller than those of the control group 1 (8.5% and 5.2%, respectively). The traditional ampere-hour integration method used in control group 1 was susceptible to cumulative errors due to current sensor zero-point drift and capacity calibration deviation under dynamic operating conditions. The test group used feature extraction and statistical matrix mapping to reduce the cumulative errors caused by long-term current integration.

[0171] When processing fragmented operational physical quantity data, the error percentage of the test group was similar to that of the control group 2, which relied on constant current test data. Under online operation conditions, it is typically difficult to obtain the continuous standard constant current operating condition data required by control group 2. The online aging detection method for the power battery cells in the test group locates pseudo-steady-state segments in daily fluctuating operating conditions and accumulates discrete differential capacity points into a two-dimensional statistical matrix to reconstruct probability ridge extreme points.

[0172] The data distribution comparison above shows that the online aging detection method for power battery cells has the ability to extract aging characteristic parameters without the need for offline operation and standard constant current charging. Its evaluation error is close to that of the offline analysis method, providing an implementation path for measuring the online health status of power batteries.

[0173] Although embodiments of the invention have been shown and described, it will be understood by those skilled in the art that various changes, modifications, substitutions and alterations can be made to these embodiments without departing from the principles and spirit of the invention, the scope of which is defined by the appended claims and their equivalents.

Claims

1. A method for online aging detection of power battery cells, characterized in that, The method includes: Obtain the minimum effective voltage resolution and ambient noise floor amplitude of the analog-to-digital converter under static operating conditions; The adaptive voltage step size is determined based on the minimum effective voltage resolution and the ambient background noise amplitude, and the phase change active voltage window is set based on the cell chemical system parameters. Real-time acquisition of the terminal voltage and loop current of the power battery cell under online operation, and determination of pseudo-steady state segments based on the terminal voltage and loop current; When determining whether to enter the pseudo-steady state segment, an initial voltage reference and an initial charge accumulation value are set; During the duration of the pseudo-steady state segment, a charge accumulation operation is performed on the initial charge accumulation value to obtain an updated charge accumulation value; When the terminal voltage meets the preset voltage change condition compared to the initial voltage reference, the updated charge accumulation value is used to perform charge quantity verification and determination. If the verification conditions for the charge quantity verification are met, then the discrete differential capacity and characteristic voltage are calculated. The characteristic voltages are filtered based on the phase transition active voltage window, and the filtered characteristic voltages are mapped to a preset two-dimensional statistical matrix along with the discrete differential capacity. Based on the preset two-dimensional statistical matrix, the extreme points of the probability ridge are determined, and the aging detection results are generated according to the drift relationship between the extreme points of the probability ridge and the baseline extreme points at the beginning of the lifespan.

2. The online aging detection method for power battery cells according to claim 1, characterized in that, Obtain the minimum effective voltage resolution and ambient noise floor amplitude of the analog-to-digital converter under static operating conditions, including: Read the reference voltage value of the analog-to-digital converter; Read the conversion sampling bit depth of the analog-to-digital converter and determine the full-scale digital steps based on the conversion sampling bit depth; The minimum effective voltage resolution is obtained based on the reference voltage value and the full-scale digital step count; When the external high-voltage relay is in the open state and the loop current collected by the externally connected current sensor is continuously zero, a set number of power battery cell terminal voltage data are continuously collected. Extract the maximum and minimum voltage values ​​from the terminal voltage data of the power battery cells, and subtract the minimum voltage value from the maximum voltage value to obtain the ambient background noise amplitude. The adaptive voltage step size is determined based on the minimum effective voltage resolution and the ambient background noise amplitude, and the phase transition active voltage window is set based on the cell chemical system parameters, including: The ambient background noise amplitude is divided by the minimum effective voltage resolution and rounded up. The rounded result is then incremented by one as the adaptive step coefficient. The adaptive voltage step size is obtained by multiplying the adaptive step coefficient by the minimum effective voltage resolution. The chemical system parameters of the power battery cell are retrieved, and based on the chemical system parameters, the lower and upper voltage limits of the phase change active voltage window, which includes the aging characteristic range, are set.

3. The online aging detection method for power battery cells according to claim 1, characterized in that, Real-time acquisition of the terminal voltage and loop current of the power battery cells under online operation, and determination of pseudo-steady-state segments based on the terminal voltage and loop current, including: The terminal voltage and circuit current of the power battery cell are collected in real time at fixed sampling time intervals as operating physical quantity data, and the operating physical quantity data are stored in the order of sampling time to establish a sliding time window; The current change rate of the loop current is obtained by subtracting the loop current at the latest moment from the loop current at the oldest moment within the sliding time window, and then dividing by the time length of the sliding time window. Determine whether the absolute value of the current change rate is less than or equal to a preset current fluctuation threshold; If the absolute value of the current change rate is less than or equal to the preset current fluctuation threshold, then determine whether the absolute value of the loop current is greater than or equal to the preset effective operating current lower limit. If the absolute value of the loop current is greater than or equal to the preset effective operating current lower limit, then the power battery cell is determined to have entered the pseudo-steady state segment. If the absolute value of the current change rate is greater than the preset current fluctuation threshold, or the absolute value of the loop current is less than the preset effective operating current lower limit, then it is determined that the power battery cell has not entered the pseudo steady state segment.

4. The online aging detection method for power battery cells according to claim 3, characterized in that, When determining whether to enter the pseudo-steady-state segment, an initial voltage reference and an initial charge accumulation value are set, including: The terminal voltage at which the pseudo-steady state segment is determined is assigned to the initial voltage reference. Set the initial charge accumulation value to zero; The step of performing a charge accumulation operation on the initial charge accumulation value during the duration of the pseudo-steady state segment to obtain the updated charge accumulation value includes: During the duration of the pseudo-steady state segment, the fixed sampling time interval is multiplied by the loop current and converted to ampere-hours to obtain the single charge quantity; The initial charge accumulation value is successively accumulated using the single charge amount to obtain the updated charge accumulation value.

5. The online aging detection method for power battery cells according to claim 4, characterized in that, When the terminal voltage meets a preset voltage change condition compared to the initial voltage reference, a charge quantity verification judgment is performed using the updated charge accumulation value, including: The instantaneous voltage difference is obtained by subtracting the initial voltage reference from the current terminal voltage, and the absolute value of the instantaneous voltage difference is extracted as the absolute change. The condition for setting the preset voltage change is that the absolute change is greater than or equal to the adaptive voltage step size. If the terminal voltage does not meet the preset voltage change condition compared to the initial voltage reference, the charge accumulation operation continues. When the terminal voltage meets the preset voltage change condition compared to the initial voltage reference, the absolute value of the updated charge accumulation value is extracted as a verification comparison value; Determine whether the verification comparison value is greater than or equal to a preset lower limit threshold for charge integration to complete the charge quantity verification.

6. The online aging detection method for power battery cells according to claim 5, characterized in that, If the verification conditions for the charge quantity verification are met, then the discrete differential capacity and characteristic voltage are calculated, including: When the verification comparison value is greater than or equal to the preset lower limit threshold of charge integral, it is confirmed that the verification condition of the charge quantity verification judgment is met; when the verification comparison value is less than the preset lower limit threshold of charge integral, it is confirmed that the verification condition of the charge quantity verification judgment is not met. The discrete differential capacity is obtained by dividing the verification comparison value when the verification condition is met by the adaptive voltage step size; the discrete differential capacity is calculated according to the formula: ； In the formula, For discrete difference capacity; The verification comparison value when the verification conditions are met; For adaptive voltage step size; The characteristic voltage is obtained by adding the current terminal voltage to the initial voltage reference and then dividing by two. After calculating the discrete differential capacity and the characteristic voltage, the following is also included: The current terminal voltage is reassigned to the initial voltage reference, and the updated charge accumulation value is reassigned to zero to complete the operation of resetting the initial voltage reference and the updated charge accumulation value.

7. The online aging detection method for power battery cells according to claim 6, characterized in that, The characteristic voltages are filtered based on the phase transition active voltage window, and the filtered characteristic voltages are mapped to a preset two-dimensional statistical matrix along with the discrete differential capacity, including: Determine whether the characteristic voltage is within the phase transition active voltage window; If the characteristic voltage is not within the phase transition active voltage window, then stop performing matrix mapping between the current characteristic voltage and the discrete differential capacity; If the characteristic voltage is within the phase transition active voltage window, the characteristic voltage difference is obtained by subtracting the lower limit of the phase transition active voltage window from the characteristic voltage, and the capacity difference is obtained by subtracting the preset capacity grid lower limit from the discrete differential capacity. The characteristic voltage difference is divided by the adaptive voltage step size and rounded down to generate a horizontal axis storage index. The capacity difference is divided by the preset capacity grid resolution and rounded down to generate a vertical axis storage index. By concatenating the horizontal axis storage index with the vertical axis storage index, the corresponding grid coordinates in the preset two-dimensional statistical matrix are located. The method further includes: performing a count increment operation on the count value corresponding to the corresponding grid coordinates.

8. The online aging detection method for power battery cells according to claim 7, characterized in that, Determining the extreme points of the probability ridge based on the preset two-dimensional statistical matrix includes: When the preset evaluation period is reached, the preset two-dimensional statistical matrix after the count value increment operation is traversed to obtain the maximum count value. Extract the specific grid coordinates corresponding to the maximum count value, as well as the horizontal axis storage index and the vertical axis storage index; Using the horizontal and vertical storage indices corresponding to the specific grid coordinates, extreme point voltages and extreme point capacities are generated; the extreme point voltages and extreme point capacities are then calculated according to the formulas. ； In the formula, The voltage at the extreme point; Store an index for the horizontal axis corresponding to a specific grid coordinate; For adaptive voltage step size; This is the lower limit of the phase transition active voltage window; ； In the formula, Capacity at extreme points; Store an index for the vertical axis corresponding to a specific grid coordinate; The preset capacity grid resolution; This is the preset lower limit of the capacity grid; The extreme point voltage and the extreme point capacity together form a two-dimensional physical coordinate pair, and the two-dimensional physical coordinate pair is determined as the extreme point of the probability ridge.

9. The online aging detection method for power battery cells according to claim 8, characterized in that, Based on the drift relationship between the extreme points of the probability ridge and the baseline extreme points at the beginning of the lifespan, aging detection results are generated, including: Extract the pre-stored initial lifetime reference extreme point, which includes the initial extreme point voltage and the initial extreme point capacity. The voltage drift is calculated by subtracting the voltage at the extreme point of the probability ridge extreme point from the voltage at the initial extreme point. The capacity drift is calculated by subtracting the capacity of the extreme point in the probability ridge extreme point from the capacity of the initial extreme point. Using the extreme boundary parameters pre-stored in the model parameter set, the voltage drift and the capacity drift are subjected to maximum-minimum normalization preprocessing. The normalized voltage drift and the normalized capacity drift are input into a preset aging state mapping model to perform forward inference, and a floating-point value representing the ratio of the actual usable capacity of the power battery cell to the factory rated capacity is output. The floating-point value is defined as the health status assessment value. The health status assessment value is determined as the aging test result of the power battery cell.

10. An online aging detection system for power battery cells, applied to the online aging detection method for power battery cells as described in any one of claims 1-9, characterized in that, include: The parameter acquisition and configuration module is used to acquire the minimum effective voltage resolution and ambient noise floor amplitude of the analog-to-digital converter, determine the adaptive voltage step size based on the minimum effective voltage resolution and ambient noise floor amplitude, and set the phase transition active voltage window. The state slice positioning module is used to collect the operating physical quantity data of the power battery cells in the online operating state in real time, determine the pseudo steady state segment based on the operating physical quantity data, and set the initial voltage reference and initial charge accumulation value when the pseudo steady state segment is determined to be entered. The interlocking verification and extraction module is used to perform a charge accumulation operation on the initial charge accumulation value based on the running physical quantity data during the duration of the pseudo steady state segment to obtain an updated charge accumulation value. When the absolute change of the terminal voltage in the running physical quantity data relative to the initial voltage reference is greater than or equal to the adaptive voltage step size, the updated charge accumulation value is used to perform a charge quantity verification judgment. If the verification condition of the charge quantity verification judgment is met, the discrete differential capacity and characteristic voltage are calculated, and the initial voltage reference and the updated charge accumulation value are reset. The statistical matrix mapping module is used to determine whether the characteristic voltage is within the phase transition active voltage window. If the characteristic voltage is within the phase transition active voltage window, the characteristic voltage and the discrete differential capacity are mapped to the corresponding grid coordinates of a preset two-dimensional statistical matrix, and the count value corresponding to the corresponding grid coordinates is incremented. The aging measurement output module is used to traverse the preset two-dimensional statistical matrix after the count value increment operation, locate the probability ridge extreme point based on the maximum count value in the preset two-dimensional statistical matrix, compare the drift of the probability ridge extreme point with the preset early life reference extreme point, and generate the aging detection result for indicating the power battery cell.

Images