Combined estimation method of soc and soh and electronic device

Abstract

This application relates to the field of battery management technology and discloses a joint estimation method and electronic device for SOC and SOH. The joint estimation method for SOC and SOH calculates a first SOC estimate corresponding to the current time by obtaining the SOC estimate value of the previous cycle, the available capacity of the solid-state battery, and the coulombic efficiency. Furthermore, it calculates a first SOH estimate value based on an adaptive dynamic weighted fusion model. Based on the first SOH estimate value, it updates the available capacity and coulombic efficiency of the solid-state battery to obtain a second SOC estimate value. At the same time, based on the first SOC estimate value, it determines the battery operating range in which the first SOC estimate value is located to obtain the second SOH estimate value. This application can achieve bidirectional coupling and mutual correction of SOC and SOH, thereby improving the estimation accuracy of SOC and SOH.

Description

Technical Field

[0001] This application relates to the field of battery management technology, and in particular to a method and electronic device for jointly estimating SOC and SOH. Background Technology

[0002] Solid-state batteries, such as all-solid-state batteries, rely on high-density solid electrolytes to achieve lithium-ion conduction. They have outstanding advantages such as high safety, high voltage resistance, high energy density, long cycle life, wide temperature range operation, and no risk of leakage. They are the core iterative technology direction for the next generation of high-end power batteries and large-scale energy storage batteries.

[0003] Unlike traditional liquid lithium batteries, which exhibit liquid wetting, fluid mass transfer, and single-stage polarization electrochemical characteristics, all-solid-state batteries employ a solid-solid close-contact structure without liquid fluid wetting effects. This fundamentally alters the battery's polarization response characteristics, voltage thermodynamic changes, power-off relaxation recovery characteristics, and long-term aging failure mechanisms. Traditional BMS state estimation algorithms adapted for liquid batteries suffer from fundamental incompatibility and cannot be directly reused.

[0004] It is understandable that the State of Charge (SOC) and State of Health (SOH) of a battery are the two core underlying state variables of a battery management system. These two variables are strongly coupled and work together to determine the control accuracy and safety reliability of the battery system. Specifically, the accuracy of SOC directly controls real-time charge and discharge current limiting and dynamic power output constraints; while the accuracy of SOH directly determines the accuracy of battery life assessment, secondary utilization determination, and safety failure early warning strategies.

[0005] Currently, SOC and SOH estimations typically employ independent estimation architectures. For example, SOC estimation often relies on ampere-hour integration with fixed capacity and coulombic efficiency, supplemented by simple voltage correction, without incorporating real-time battery aging conditions to adjust the basic calculation parameters. SOH estimation, on the other hand, is often based on fitting single impedance characteristics or capacity decay data separately, without considering the battery's state characteristics during its real-time operating range. Because these separate SOC and SOH estimation architectures fail to consider the strong coupling characteristics unique to all-solid-state batteries, estimation errors are prone to accumulating, leading to insufficient estimation accuracy. Summary of the Invention

[0006] The embodiments of this application aim to provide a joint estimation method and electronic device for SOC and SOH, so as to improve the estimation accuracy of SOC and SOH.

[0007] The embodiments of this application provide the following technical solutions: On one hand, embodiments of this application provide a joint estimation method for SOC and SOH, applied to solid-state batteries, the method comprising: Obtain the estimated SOC value, available capacity of the solid-state battery, and coulombic efficiency of the previous cycle; Based on the SOC estimate from the previous cycle, the available capacity of the solid-state battery, and the coulombic efficiency, calculate the first SOC estimate for the current moment. The bulk aging component, the interface comprehensive aging component, and the ion diffusion aging component are obtained. The ion diffusion aging component is calculated from the initial diffusion impedance amplitude and the temperature-corrected diffusion impedance amplitude. Each aging component corresponds to an aging component weight. The first SOH estimate is calculated based on the adaptive dynamic weighted fusion model. The adaptive dynamic weighted fusion model is used to calculate the weighted sum obtained by multiplying each aging component by its corresponding aging component weight, and the weighted sum is used as the first SOH estimate. Based on the first SOH estimate, the available capacity and coulombic efficiency of the solid-state battery are updated, and based on the updated available capacity and coulombic efficiency, the second SOC estimate is obtained. Based on the SOC estimate, the battery operating range in which the SOC estimate is located is determined to update the aging component weights. These weights are then substituted into the adaptive dynamic weighted fusion model to obtain the second SOH estimate.

[0008] In this technical solution, SOC estimation is completed by obtaining the previous cycle's SOC estimate, available capacity, and coulombic efficiency. Bulk aging components, interface aging components, and ion diffusion aging components are also obtained, and an adaptive dynamic weighted fusion model is used to calculate the SOH estimate. Furthermore, the SOH estimate is used to update the available capacity and coulombic efficiency of the solid-state battery, updating the SOC estimate. Simultaneously, the weights corresponding to each aging component are dynamically updated based on the battery's operating range, iteratively obtaining a second SOH estimate. This achieves bidirectional closed-loop coupled estimation of SOC and SOH. On one hand, the aging state is used to correct the SOC calculation parameters, suppressing SOC drift caused by battery aging. On the other hand, the weighted fusion of multiple aging components can avoid result jumps caused by single impedance characteristics being affected by operating conditions and temperature disturbances, improving the robustness of SOH estimation. The bidirectional constraint closed loop of periodic iteration can suppress the unidirectional accumulation and amplification of errors, adapting to the continuously evolving aging behavior throughout the solid-state battery's entire life cycle, and comprehensively improving the estimation accuracy and long-term operational stability of the joint SOC and SOH estimation of the solid-state battery across the entire aging stage.

[0009] In some embodiments, Based on the SOC estimate from the previous cycle, the available capacity of the solid-state battery, and the coulombic efficiency, calculate the first SOC estimate for the current moment, including: Based on the SOC estimate from the previous cycle, determine the battery operating range in which the SOC estimate falls. Based on the battery's operating range, determine the SOC estimation strategy corresponding to the estimated SOC value; Based on the available capacity and coulombic efficiency of solid-state batteries, the first estimated SOC value at the current moment is calculated using a SOC estimation strategy.

[0010] In some embodiments, Based on the battery's operating range, determine the SOC estimation strategy corresponding to the estimated SOC value, including: Obtain the first mapping relationship, wherein the first mapping relationship includes the correspondence between at least two battery operating ranges and at least two SOC estimation strategies, and the battery operating ranges and SOC estimation strategies correspond one-to-one; Based on the first mapping relationship, and considering the battery operating range within which the estimated SOC value lies, the SOC estimation strategy corresponding to the estimated SOC value is determined. The battery operating range includes a first operating range, and the SOC estimation strategy includes a first estimation strategy. The first operating range corresponds to the first estimation strategy. The first estimation strategy includes: calculating the first SOC estimate corresponding to the SOC estimate based on the dynamic Coulomb integral formula, which includes:

[0011] in, Here is the estimated SOC value at time t. Let SOC be the estimated value of the previous time step at time t. For Coulomb efficiency, Let be the current at time t. For a fixed sampling period, This represents the available capacity of the solid-state battery.

[0012] In some embodiments, The method also includes: Obtain the current polarity of the solid-state battery. The current polarity includes charging condition or discharging condition. The charging condition corresponds to the charging OCV-SOC lookup table, and the discharging condition corresponds to the discharging OCV-SOC lookup table. The charging OCV-SOC lookup table is used to store the correspondence between the reference open circuit voltage and the standard SOC correction value under the charging condition. The discharging OCV-SOC lookup table is used to store the correspondence between the reference open circuit voltage and the standard SOC correction value under the discharging condition. Based on the current current polarity, determine the corresponding first OCV-SOC lookup table; Based on the SOC estimate of the previous cycle, consult the first OCV-SOC lookup table to determine the theoretical open-circuit voltage corresponding to the SOC estimate of the previous cycle. Obtain the steady-state open-circuit voltage of the solid-state battery; Calculate the voltage difference between the steady-state open-circuit voltage and the theoretical open-circuit voltage based on the steady-state open-circuit voltage; Based on the voltage difference, the local slope corresponding to the estimated SOC value is obtained from the first OCV-SOC lookup table. The voltage difference is divided by the local slope to obtain the SOC correction amount. Based on the SOC correction amount, determine the corrected first SOC estimate, where the corrected first SOC estimate = first SOC estimate + SOC correction amount.

[0013] In some embodiments, The method also includes: Calculating the steady-state open-circuit voltage of a solid-state battery includes: If the solid-state battery meets the steady-state resting conditions, then obtain the short-time relaxation voltage timing data for a preset duration. Based on the short-time relaxation voltage time series data, the steady-state open-circuit voltage is obtained iteratively using a pre-built relaxation fitting model; The steady-state resting conditions include: the normalized absolute value of the current of the solid-state battery is less than a preset current threshold, and the duration for which the normalized absolute value is less than the preset current threshold is greater than a preset time threshold.

[0014] In some embodiments, Relaxation fitting models include:

[0015] in, This represents the real-time terminal voltage at time t in the short-time relaxation voltage time series data. The steady-state open-circuit voltage, For the interface fast polarization amplitude coefficient, This represents the amplitude coefficient of the slow diffusion polarization in the solid phase. The interface polarization adaptive dynamic time constant. The adaptive dynamic time constant for diffusion polarization; Based on short-time relaxation voltage time-series data and a pre-built relaxation fitting model, the steady-state open-circuit voltage is iteratively obtained, including: Construct a residual objective function, where the residual objective function is used to calculate the residual, which is the sum of the squares of the differences between the real-time terminal voltage in the short-time relaxation voltage time series data and the fitted open-circuit voltage; The steady-state open-circuit voltage, interface fast polarization amplitude coefficient, solid-phase slow diffusion polarization amplitude coefficient, interface polarization adaptive dynamic time constant, and diffusion polarization adaptive dynamic time constant are updated synchronously and iteratively. In each iteration, the residual is calculated in real time; If the residual is less than the preset convergence threshold, or the number of iterations is greater than the maximum number of iterations threshold, then the iteration process ends and the converged steady-state open-circuit voltage is obtained.

[0016] In some embodiments, The battery operating range includes a second operating range, and the SOC estimation strategy includes a second estimation strategy. The second operating range corresponds to the second estimation strategy. The second estimation strategy includes: Obtain the interface impedance amplitude, interface impedance phase, diffusion impedance amplitude, and real-time cell temperature; Obtain a pre-calibrated multidimensional feature mapping function, which includes multiple sets of test data and their corresponding impedance mapping SOC values. Each set of test data includes interface impedance amplitude, interface impedance phase, diffusion impedance amplitude, and real-time cell temperature. The interface impedance amplitude, interface impedance phase, diffusion impedance amplitude, and real-time cell temperature are input into a multi-dimensional feature mapping function, and the first impedance mapping SOC value is obtained through interpolation. The SOC value mapped by the first impedance is weighted and fused with the SOC estimate at the current moment calculated by the dynamic Coulomb integral formula to obtain the fused SOC estimate.

[0017] In some embodiments, The battery operating range includes a third operating range, and the SOC estimation strategy includes a third estimation strategy. The third operating range corresponds to the third estimation strategy. The third estimation strategy includes:

[0018] in, The SOC estimate is calculated based on the third estimation strategy. For smoothing weighting coefficients, The SOC estimate is calculated based on the first estimation strategy. The SOC estimate is calculated based on the second estimation strategy.

[0019] In some embodiments, Based on the adaptive dynamic weighted fusion model, the first SOH estimate is calculated, including: The bulk aging component, the interface comprehensive aging component, and the ion diffusion aging component are input into the adaptive dynamic weighted fusion model to obtain the SOH estimate; The adaptive dynamic weighted fusion model includes:

[0020] in, This is an estimated value for SOH. As the weight of the aging component, This refers to the aging component of the substrate. Assign weights to the overall aging components of the interface. For the overall aging components of the interface, The weighting of the ion diffusion aging component. This is the ion diffusion aging component. .

[0021] In some embodiments, The method also includes: Construct an interface damage cumulative factor model; The cumulative damage factor of the interface is calculated based on the interface damage cumulative factor model. The interface damage cumulative factor model includes:

[0022] in, Accumulate the damage factor for the interface at the current moment. This represents the cumulative interface damage factor from the previous cycle. This is the initial interface impedance magnitude. This represents the relative change in interface impedance in real time. , This is the temperature-corrected interfacial impedance amplitude. These are the phase-weighted calibration coefficients. This refers to the interface phase attenuation. , This represents the initial interface impedance phase. This represents the phase of the interface impedance after temperature correction. For time period.

[0023] In some embodiments, The method also includes: Calculate the real-time relative change in interface impedance, including: Obtain the first interface impedance timing, wherein the first interface impedance timing is the temperature-corrected interface impedance timing; Construct a first time window, where the first time window corresponds to the duration of the first window; Based on the timing of the first interface impedance, calculate the short-term steady-state reference impedance corresponding to the first time window; Calculate the short-time reversible disturbance component based on the short-term steady-state reference impedance; The short-time reversible disturbance component is filtered to obtain the timing sequence of the second interface impedance. Construct a second time window, where the second time window corresponds to the duration of the second window, and the duration of the second window is longer than the duration of the first window; The timing sequence of the second interface impedance is sampled to obtain the long-term irreversible aging impedance component. Based on the long-term irreversible aging impedance component, the relative change in real-time interface impedance is calculated, including:

[0024] in, This represents the relative change in interface impedance in real time. The interface impedance amplitude is the value of the long-term irreversible aging impedance component in the kth calculation period. This represents the initial interface impedance amplitude of the impedance component over long-term irreversible aging.

[0025] In some embodiments, Before obtaining the timing sequence of the first interface impedance, the method also includes: The impedance amplitude and impedance phase corresponding to each frequency point are collected to determine the ohmic impedance amplitude and ohmic impedance phase corresponding to the first frequency band; the interface impedance amplitude and interface impedance phase corresponding to the second frequency band; and the diffusion impedance amplitude and diffusion impedance phase corresponding to the third frequency band. Temperature corrections are applied to the ohmic impedance amplitude, interface impedance amplitude, and diffusion impedance amplitude to obtain the temperature-corrected ohmic impedance amplitude, temperature-corrected interface impedance amplitude, and temperature-corrected diffusion impedance amplitude.

[0026] In some embodiments, Temperature corrections are applied to the ohmic impedance amplitude, interface impedance amplitude, and diffusion impedance amplitude to obtain the temperature-corrected ohmic impedance amplitude, temperature-corrected interface impedance amplitude, and temperature-corrected diffusion impedance amplitude, including: Obtain the temperature coupling coefficient and real-time cell surface temperature; Calculate the temperature deviation, where temperature deviation = real-time cell surface temperature - reference calibration temperature; Based on the temperature coupling coefficient and the real-time cell surface temperature, the ohmic impedance amplitude, interface impedance amplitude, and diffusion impedance amplitude are temperature-corrected using the impedance normalization correction formula. The resulting temperature-corrected ohmic impedance amplitude, interface impedance amplitude, and diffusion impedance amplitude are then obtained. The impedance normalization correction formula includes:

[0027] in, This is the impedance amplitude after temperature correction. This is the impedance amplitude before temperature correction. The temperature coupling coefficient is... For real-time cell temperature, The reference calibration temperature.

[0028] In some embodiments, Based on the first SOC estimate, the battery operating range in which the first SOC estimate falls is determined to update the aging component weights, including: Obtain the benchmark weight combination corresponding to each battery working range. The benchmark weight combination includes the benchmark weight of the bulk aging component, the benchmark weight of the interface comprehensive aging component, and the benchmark weight of the ion diffusion aging component. Based on the first SOC estimate, determine the first benchmark weight combination corresponding to the battery operating range in which the first SOC estimate is located; Based on the first SOH estimate, the aging stage of the solid-state battery is determined, including the early aging stage, the middle aging stage, and the late aging stage. Based on the aging stage of the solid-state battery, the first benchmark weight combination is adjusted to obtain the updated aging component weights.

[0029] In some embodiments, Based on the first SOH estimate, the available capacity and coulombic efficiency of the solid-state battery are updated, including: Based on the rated capacity and SOH estimate of the solid-state battery, the updated usable capacity is obtained, including:

[0030] in, For the updated available capacity, This refers to the rated capacity of solid-state batteries. This is an estimated value for SOH; Obtain the coulomb efficiency reference table, which includes multiple sets of cell temperature, current ratio, and current polarity correspondences; The current cell temperature, current rate, and current polarity are obtained, and the reference coulomb efficiency is obtained through interpolation based on the coulomb efficiency reference table. Based on the first SOH estimate, the aging correction factor is calculated, including:

[0031] in, This is the aging correction factor. This is the calibration coefficient for interface aging degradation. This is the first estimated SOH value; The updated coulomb efficiency is calculated based on the baseline coulomb efficiency and the aging correction factor, including:

[0032] in, For the updated Coulomb efficiency, As a benchmark coulomb efficiency, This is the aging correction factor.

[0033] On the other hand, embodiments of this application provide an electronic device, including: At least one processor; At least one memory for storing at least one program; When at least one program is executed by at least one processor, the at least one processor performs the method described above.

[0034] On the other hand, embodiments of this application provide a non-volatile computer-readable storage medium storing a processor-executable program, which, when executed by a processor, is used to perform the above-described method.

[0035] The beneficial effects of this application's embodiments: Unlike existing technologies, this application's embodiments provide a joint estimation method for SOC and SOH, applied to solid-state batteries. The method includes: obtaining the estimated SOC value of the previous cycle, the available capacity of the solid-state battery, and the coulombic efficiency; calculating the first estimated SOC value corresponding to the current moment based on the estimated SOC value of the previous cycle, the available capacity of the solid-state battery, and the coulombic efficiency; obtaining the bulk aging component, the interface comprehensive aging component, and the ion diffusion aging component, wherein the ion diffusion aging component is calculated from the initial diffusion impedance amplitude and the temperature-corrected diffusion impedance amplitude, and each aging component corresponds one-to-one with... The aging component weights are calculated; based on the adaptive dynamic weighted fusion model, a first SOH estimate is calculated, wherein the adaptive dynamic weighted fusion model is used to calculate the weighted sum obtained by multiplying each aging component by its corresponding aging component weight, and the weighted sum is used as the first SOH estimate; based on the first SOH estimate, the available capacity and coulombic efficiency of the solid-state battery are updated, and a second SOC estimate is obtained based on the updated available capacity and coulombic efficiency; based on the first SOC estimate, the battery operating range in which the SOC estimate is located is determined, the aging component weights are updated, and the results are substituted into the adaptive dynamic weighted fusion model to obtain the second SOH estimate.

[0036] By acquiring the SOC estimate, available capacity, and coulombic efficiency of the solid-state battery from the previous cycle, a first SOC estimate corresponding to the current moment is calculated. Furthermore, based on an adaptive dynamic weighted fusion model, a first SOH estimate is calculated. Further, based on the first SOH estimate, the available capacity and coulombic efficiency of the solid-state battery are updated to obtain a second SOC estimate. Simultaneously, based on the first SOC estimate, the battery operating range in which the first SOC estimate is located is determined to obtain the second SOH estimate. This application can achieve bidirectional coupling and mutual correction of SOC and SOH, improving the estimation accuracy of SOC and SOH. Attached Figure Description

[0037] One or more embodiments are illustrated by way of example with reference numerals in the accompanying drawings. These illustrations do not constitute a limitation on the embodiments. Elements with the same reference numerals in the drawings are denoted as similar elements. Unless otherwise stated, the figures in the drawings are not to be limited by scale.

[0038] Figure 1 This is a schematic flowchart of a joint estimation method for SOC and SOH provided in an embodiment of this application; Figure 2 This is provided by the embodiments of this application. Figure 1 A detailed flowchart of step S102 in the process; Figure 3 This is provided by the embodiments of this application. Figure 1 A detailed flowchart of step S104 in the process; Figure 4 This is a schematic diagram of a process for determining a corrected first SOC estimate provided in an embodiment of this application; Figure 5 This is a schematic flowchart of a method for calculating steady-state open-circuit voltage provided in an embodiment of this application; Figure 6 This is provided by the embodiments of this application. Figure 1 A detailed flowchart of step S105 in the process; Figure 7 This is provided by the embodiments of this application. Figure 1 A detailed flowchart of step S106 in the process; Figure 8 This is a schematic diagram of a process for optimizing the second SOC estimate and the second SOH estimate provided in an embodiment of this application; Figure 9 This is a schematic diagram of the overall process of a joint estimation method for SOC and SOH provided in an embodiment of this application; Figure 10 This is a schematic diagram of the structure of an electronic device provided in an embodiment of this application.

[0039] Explanation of icon numbers: 100. Electronic device; 101. Processor; 102. Memory. Detailed Implementation

[0040] The present application will now be described in detail with reference to specific embodiments. These embodiments will help those skilled in the art to further understand the present application, but do not limit the present application in any way. It should be noted that those skilled in the art can make several modifications and improvements without departing from the concept of the present application. These all fall within the protection scope of the present application.

[0041] To make the objectives, technical solutions, and advantages of this application clearer, the following detailed description is provided in conjunction with the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are merely illustrative and not intended to limit the scope of this application.

[0042] It should be noted that, unless there is a conflict, the various features in the embodiments of this application can be combined with each other, all of which are within the protection scope of this application. Furthermore, although functional modules are divided in the device schematic diagram and a logical order is shown in the flowchart, in some cases, the steps shown or described can be executed in a different order than the module division in the device or the order in the flowchart. In addition, the terms "first," "second," and "third" used herein do not limit the data or execution order, but only distinguish identical or similar items with essentially the same function and effect.

[0043] Unless otherwise defined, all technical and scientific terms used in this specification have the same meaning as commonly understood by one of ordinary skill in the art to which this application belongs. The terminology used in this specification is for the purpose of describing particular embodiments only and is not intended to limit the scope of this application. The term "and / or" as used in this specification includes any and all combinations of one or more of the associated listed items.

[0044] Furthermore, the technical features involved in the various embodiments of this application described below can be combined with each other as long as they do not conflict with each other.

[0045] Currently, the common approach is to use an independent estimation architecture for SOC and SOH. For example, SOC estimation often relies on the ampere-hour integration method with fixed capacity and fixed coulombic efficiency, supplemented by simple voltage correction, without combining the real-time aging state of the battery to correct the basic calculation parameters; SOH estimation is often based on a single impedance characteristic or capacity decay data for individual fitting, without considering the state characteristics of the battery in the real-time operating range.

[0046] However, since the independent estimation architecture of SOC and SOH does not take into account the strong coupling characteristics unique to all-solid-state batteries: the core of solid-state battery aging is the degradation of interface impedance, and the increase in impedance will directly change the battery polarization response, coulombic efficiency, and effective capacity benchmark, thus continuously introducing systematic biases into SOC estimation; conversely, long-term ampere-hour integral error of SOC will accumulate and interfere with capacity benchmark calibration, which in turn will cause SOH estimation drift.

[0047] Therefore, the traditional independent estimation mode cannot achieve two-way error correction. The errors of the two methods are superimposed and continuously accumulate, resulting in insufficient estimation accuracy of the battery.

[0048] Based on this, embodiments of this application provide a joint estimation method for SOC and SOH to improve the estimation accuracy of SOC and SOH.

[0049] The following specific examples illustrate the joint estimation method for SOC and SOH provided in this application.

[0050] Please see Figure 1 , Figure 1 This is a flowchart illustrating a joint estimation method for SOC and SOH provided in an embodiment of this application.

[0051] The joint estimation method of SOC and SOH is applied to solid-state batteries, such as all-solid-state batteries. Specifically, the execution subject of the joint estimation method of SOC and SOH is one or at least two processors of an electronic device.

[0052] In some embodiments, the electronic device includes a battery management system (BMS), which includes: (1) Voltage sampling circuit, used to collect the single-cell terminal voltage of solid-state battery in real time.

[0053] (2) Current sampling circuit, used to collect the charging and discharging current of solid-state battery in real time.

[0054] (3) Temperature sensor, used to collect the real-time temperature of the solid-state battery cell.

[0055] (4) Electrochemical Impedance Spectroscopy (EIS) detection module, used for high-frequency scanning to obtain multi-band impedance amplitude and impedance phase, and to analyze multi-dimensional characteristics such as ohmic impedance, interfacial impedance, and diffusion impedance.

[0056] like Figure 1 As shown, the joint estimation method for SOC and SOH includes the following steps S101-S106: Step S101: Obtain the estimated SOC value, available capacity of the solid-state battery, and coulombic efficiency of the previous cycle.

[0057] Specifically, it reads the battery's state of charge (SOC), real-time available capacity, and coulombic efficiency from the previous cycle's convergence output.

[0058] As can be understood, State of Charge (SOC) refers to the percentage of a battery's current remaining charge relative to its total usable capacity. Usable capacity refers to the maximum amount of charge a solid-state battery can actually charge or discharge. Coulomb efficiency (η) is a core parameter characterizing the effective utilization of charge during battery charging and discharging; it refers to the ratio of the actual amount of charge released during discharge to the actual amount of charge stored during charging.

[0059] It should be noted that the SOC value is an instantaneous state that fluctuates in real time; for example, it changes in real time during charging, discharging, and resting. Usable capacity, on the other hand, is a slow-moving variable and an aging parameter; for example, it does not change with a single charge-discharge cycle but only degrades over long-term battery aging, determined by the SOH and decreasing slowly.

[0060] In the embodiments of this application, the SOC value, the available capacity of the solid-state battery, and the coulombic efficiency mentioned above are the core basic parameters for SOC estimation in the current cycle. This ensures the continuity of the algorithm's time-series iterations, avoids independent and fragmented single-cycle operations, and provides reliable initial data support for subsequent accurate SOC calculations.

[0061] Step S102: Calculate the first SOC estimate corresponding to the current moment based on the SOC estimate of the previous cycle, the available capacity of the solid-state battery, and the coulombic efficiency.

[0062] Specifically, the SOC value of the previous cycle is used to determine the battery operating range in which the SOC value is located, and then the SOC estimation strategy corresponding to the SOC value is determined. Using the SOC estimation strategy, based on the SOC value of the previous cycle and combined with the available capacity and coulombic efficiency of the solid-state battery, the first SOC estimate value corresponding to the current moment is calculated.

[0063] Step S103: Obtain the bulk aging component, the interface comprehensive aging component, and the ion diffusion aging component, wherein each aging component corresponds to an aging component weight.

[0064] It should be noted that different aging components correspond to different aging processes, for example: (1) The body aging component corresponds to the ohmic aging of the battery body. The body aging component is calculated based on the decay change of the battery ohmic impedance and is used to accurately characterize the basic aging degree of the solid-state battery cell body.

[0065] (2) The interface comprehensive aging component corresponds to the solid-solid interface aging. The interface comprehensive aging component is calculated by combining the interface impedance change and the interface cumulative damage factor. It is specifically adapted to the unique aging characteristics of the solid-solid contact interface of solid-solid batteries and is used to characterize the battery performance degradation caused by solid-solid interface contact deterioration, interface polarization intensification, interface phase decay, and cycle cumulative damage.

[0066] (3) The ion diffusion aging component corresponds to ion diffusion aging. The ion diffusion aging component is calculated based on the dynamic change of solid phase diffusion impedance and is used to characterize the degree of decay of lithium ion diffusion capability inside solid-state battery.

[0067] Step S104: Calculate the first SOH estimate based on the adaptive dynamic weighted fusion model.

[0068] Specifically, an adaptive dynamic weighted fusion model is used to perform a weighted summation operation on the three types of aging components. Specifically, the weighted sum is obtained by multiplying each aging component by its corresponding aging component weight and then summing the results. This weighted sum is used as the first SOH estimate.

[0069] In the embodiments of this application, by fusing multiple aging components to offset the estimation error caused by temperature and operating condition disturbances of a single feature, the robustness and accuracy of solid-state battery aging state identification can be improved.

[0070] Step S105: Based on the first SOH estimate, update the available capacity and coulombic efficiency of the solid-state battery, and obtain the second SOC estimate based on the updated available capacity and coulombic efficiency.

[0071] Specifically, based on the rated capacity and SOH estimate of the solid-state battery, the updated usable capacity is calculated. Then, based on the coulombic efficiency benchmark table, the benchmark coulombic efficiency is determined. Finally, based on the benchmark coulombic efficiency and the aging correction factor, the updated coulombic efficiency is calculated. Next, based on the updated usable capacity and coulombic efficiency, the second SOC estimate is obtained. It can be understood that the second SOC estimate is used to characterize the SOC estimate at the current moment.

[0072] By combining real-time calculated SOH estimates, the available capacity and coulombic efficiency of the battery are dynamically corrected, replacing the traditional fixed parameter configuration. Based on the updated available capacity and coulombic efficiency, the SOC estimates are re-optimized, which can solve the problems of SOC baseline drift and error accumulation caused by battery aging in traditional algorithms, and achieve accurate correction of the energy estimation under aging conditions.

[0073] Step S106: Based on the first SOC estimate, determine the battery operating range in which the first SOC estimate is located, update the aging component weight, and substitute it into the adaptive dynamic weighted fusion model to obtain the second SOH estimate.

[0074] Specifically, by determining the battery operating range where the SOC estimate falls, and based on the baseline weight combination corresponding to each battery operating range, the aging stage of the solid-state battery is determined using the first SOH estimate. The baseline weight combination is then adjusted to obtain the updated aging component weights. These updated aging component weights are then substituted into the adaptive dynamic weighted fusion model to obtain the second SOH estimate. It can be understood that the second SOH estimate is used to characterize the SOH estimate at the current moment.

[0075] In the embodiments of this application, the estimation accuracy of SOC and SOH can be improved by bidirectional coupling and mutual correction of SOC and SOH.

[0076] Please refer to the following: Figure 2 , Figure 2This is provided by the embodiments of this application. Figure 1 A detailed flowchart of step S102 in the process.

[0077] like Figure 2 As shown, step S102: Based on the SOC estimate from the previous cycle, the available capacity of the solid-state battery, and the coulombic efficiency, calculate the first SOC estimate for the current moment, including the following steps S121-S123: Step S121: Determine the battery operating range in which the SOC estimate falls based on the SOC estimate from the previous cycle.

[0078] Specifically, the battery operating range includes a first operating range, a second operating range, and a third operating range. The SOC value of the first operating range is [0, a] or [d, 1]; the SOC value of the second operating range is [b, c]; and the SOC value of the third operating range is [a, b] or [c, d]. a is the first SOC threshold, b is the second SOC threshold, c is the third SOC threshold, and d is the fourth SOC threshold. Furthermore, 0 < a < b < c < d < 1.

[0079] Assuming a=28%, b=32%, c=68%, d=72%, then: the first working interval is [0%, 28%] or [72%, 100%], the second working interval is [32%, 68%], and the third working interval is [28%, 32%] or [68%, 72%].

[0080] It is understandable that the first SOC threshold, the second SOC threshold, the third SOC threshold, and the fourth SOC threshold can be set according to specific needs, and no restrictions are imposed here.

[0081] Step S122: Determine the SOC estimation strategy corresponding to the SOC estimate based on the battery operating range.

[0082] Specifically, based on the battery's operating range, the SOC estimation strategy corresponding to the estimated SOC value is determined, including: Obtain the first mapping relationship, wherein the first mapping relationship includes the correspondence between at least two battery operating ranges and at least two SOC estimation strategies, and the battery operating range and the SOC estimation strategy correspond one-to-one.

[0083] Based on the first mapping relationship, and considering the battery operating range in which the SOC estimate is located, the SOC estimation strategy corresponding to the SOC estimate is determined.

[0084] Specifically, the first mapping relationship includes at least two battery operating intervals, each battery operating interval corresponding to a SOC estimation strategy. The battery operating interval corresponds to the range of SOC values, and the range of each battery operating interval is [0,1], i.e. [0%,100%].

[0085] The following examples illustrate the three battery operating ranges and three SOC estimation strategies provided in the embodiments of this application, for example: (1) First working interval and first estimation strategy Specifically, the SOC value of the first working interval is in the range of [0,a] or [d,1], for example: [0%,28%] and [72%,100%].

[0086] Understandably, the first operating range includes the low battery charge range and the high battery charge range. Within this range, the solid-state battery exhibits sufficient solid-solid interface polarization, sensitive open-circuit voltage changes, and small impedance fluctuations. This results in high accuracy of dynamic coulomb integration and slow error accumulation, making it suitable for calculating SOC using pure dynamic coulomb integration.

[0087] Specifically, the first estimation strategy includes: calculating the first estimated SOC value corresponding to the SOC value based on the dynamic Coulomb integral formula, which includes:

[0088] in, Here is the estimated SOC value at time t. Let SOC be the estimated value of the previous time step at time t. For Coulomb efficiency, Let be the current at time t. For a fixed sampling period, This represents the available capacity of the solid-state battery.

[0089] It is understandable that the SOC estimate at time t is the SOC estimate obtained by the first estimation strategy.

[0090] (2) Second working interval and second estimation strategy Specifically, the SOC value of the second working interval is in the range of [b,c], for example: [32%,68%].

[0091] Understandably, the second operating range is the main intermediate SOC range during most of the battery's daily operation. In this range, the OCV curve of the solid-state battery is extremely flat and the voltage change is very weak, so relying solely on voltage correction has extremely low accuracy; however, the battery's interface impedance, diffusion impedance, and temperature characteristics in this range are highly sensitive to SOC, and the multidimensional impedance characteristics can accurately infer the SOC state.

[0092] Specifically, the second estimation strategy includes the following steps ①-④: Step ①: Obtain the interface impedance amplitude, interface impedance phase, diffusion impedance amplitude, and real-time cell temperature.

[0093] Specifically, during battery operation, voltage, current, and cell temperature are collected in real time. The voltage is the terminal voltage of a single battery cell, the current is the charging and discharging circuit current, and the cell temperature is the surface temperature of the cell, providing a complete and continuous data source for all subsequent algorithm calculations.

[0094] For example, after the battery is powered on and operating normally, the battery management system continuously collects three types of core timing data in real time: individual cell terminal voltage sequence. U t Charging and discharging circuit current sequence I t Cell surface temperature sequence T t The data acquisition range comprehensively covers all working scenarios, including battery discharge, low-current resting, fast charging, slow charging, extreme high and low temperature conditions, and dynamic load fluctuations, ensuring uninterrupted and complete algorithm input. The sampling frequencies for voltage, current, and temperature can be uniformly set to 100Hz, with voltage sampling accuracy ≤ ±0.001V, current sampling accuracy ≤ ±0.01A, and temperature sampling accuracy ≤ ±0.5℃.

[0095] Furthermore, after obtaining the voltage, current, and cell temperature, data noise reduction processing can be performed on the voltage, current, and cell temperature, for example: A moving average filtering algorithm (e.g., 10-point or 20-point window) is used to synchronously smooth and reduce noise in the three time-series data of voltage, current, and temperature. The filtering calculation formula is as follows:

[0096] in, The data at time t is the filtered data. For time t, the first i There are N original sampled data points, where N is the number of sliding window points.

[0097] By simultaneously smoothing and reducing noise in voltage, current, and cell temperature, filtered voltage, filtered current, and filtered cell temperature are obtained. This effectively filters out high-frequency random noise and instantaneous spike interference, while preserving the true low-frequency change trend of the battery's electrochemical state.

[0098] Furthermore, after obtaining the filtered voltage, filtered current, and filtered cell temperature, it is determined whether the filtered voltage, filtered current, and filtered cell temperature meet the steady-state operating conditions. If the steady-state operating conditions are met, the solid-state battery is determined to be in an effective steady-state operating condition; if the steady-state operating conditions are not met, the solid-state battery is determined to be in an invalid operating condition. The steady-state operating conditions include: the real-time current rate of the solid-state battery is ≤ a preset rate threshold; the cell surface temperature is within a preset temperature range; there is no battery fault alarm; and there is no overcharge or over-discharge protection trigger.

[0099] In this embodiment of the application, the preset magnification threshold can be set according to specific needs, for example, set to 0.5C. Similarly, the preset temperature range can also be set according to specific needs, for example, set to -40℃~85℃.

[0100] Furthermore, after determining that the solid-state battery is in an effective steady-state condition, an imperceptible online EIS multi-frequency excitation and amplitude-phase two-dimensional feature acquisition are performed to simultaneously acquire impedance amplitude and impedance phase dual-dimensional features, thereby achieving refined identification of micro-interface aging.

[0101] Specifically, a small sinusoidal AC excitation signal is periodically applied to both ends of the solid-state battery through the battery management system (BMS). The excitation parameters are set as follows: the excitation voltage amplitude is 5mV~10mV, which is far below the battery polarization threshold and does not affect the normal operation of the battery; the excitation frequency scanning range is 10mHz~1MHz; the duration of a single full-band scan is 2s~5s, and after the scan is completed, the battery sleeps for 30s and is triggered cyclically, without interfering with the normal charging and discharging operation of the battery.

[0102] For any excitation angular frequency, the system synchronously acquires the battery response voltage and response current, calculates the complex impedance, and further solves for the impedance amplitude and impedance phase.

[0103] Furthermore, the impedance amplitude and impedance phase at each frequency point are simultaneously collected, and combined with the electrochemical mechanism of the all-solid-state battery, a precise three-stage segmentation is achieved. The aging mechanism corresponding to each frequency band is unique and has no overlap. For example: Before obtaining the timing of the first interface impedance, the impedance amplitude and impedance phase corresponding to each frequency point are collected to determine the ohmic impedance amplitude and ohmic impedance phase corresponding to the first frequency band; as well as the interface impedance amplitude and interface impedance phase corresponding to the second frequency band; and the diffusion impedance amplitude and diffusion impedance phase corresponding to the third frequency band. For example: if the first frequency band is a high-frequency band, the second frequency band is a mid-frequency band, and the third frequency band is a low-frequency band, then: (1) For the high frequency range (10kHz~1MHz), the ohmic impedance amplitude R0 and ohmic impedance phase Φ0 are extracted to characterize the structural losses of the battery tabs, electrode substrates, and solid electrolyte body, i.e., the magnitude of ohmic losses of the cell body, current collector, and solid electrolyte body.

[0104] (2) For the mid-frequency range (100Hz~10kHz), extract the interface impedance amplitude R. ct Phase Φ with interface impedance ct It is used to characterize the aging state and contact performance of the solid-solid electrode-electrolyte interface, that is, the magnitude of the charge transfer impedance of the solid-solid electrode-electrolyte interface.

[0105] (3) For the low frequency range (10mHz~100Hz), extract the diffusion impedance amplitude R. d Phase Φ with diffusion impedance d It is used to characterize the solid-phase transport capability of lithium ions inside solid active materials, that is, the magnitude of the impedance corresponding to the solid-phase diffusion process of lithium ions inside active materials.

[0106] It is understood that the impedance measured by electrochemical impedance spectroscopy (EIS) is a complex impedance, which includes both real and imaginary parts. In the embodiments of this application, the amplitude refers to the modulus (modulus value (|Z|)) of the corresponding complex impedance. For example, the ohmic impedance amplitude refers to the modulus of the ohmic complex impedance obtained by analysis in the high-frequency band, the interface impedance amplitude refers to the modulus of the interface charge transfer complex impedance obtained by analysis in the mid-frequency band, and the diffusion impedance amplitude refers to the modulus of the lithium-ion solid-phase diffusion complex impedance obtained by analysis in the low-frequency band.

[0107] Furthermore, temperature correction is applied to the ohmic impedance amplitude, interface impedance amplitude, and diffusion impedance amplitude to obtain the temperature-corrected ohmic impedance amplitude, temperature-corrected interface impedance amplitude, and temperature-corrected diffusion impedance amplitude.

[0108] Specifically, temperature correction is applied to the ohmic impedance amplitude, interface impedance amplitude, and diffusion impedance amplitude to obtain the temperature-corrected ohmic impedance amplitude, temperature-corrected interface impedance amplitude, and temperature-corrected diffusion impedance amplitude, including: Obtain the temperature coupling coefficient and real-time cell surface temperature; Calculate the temperature deviation, where temperature deviation = real-time cell surface temperature - reference calibration temperature; Based on the temperature coupling coefficient and the real-time cell surface temperature, the ohmic impedance amplitude, interface impedance amplitude, and diffusion impedance amplitude are temperature-corrected using the impedance normalization correction formula. The resulting temperature-corrected ohmic impedance amplitude, interface impedance amplitude, and diffusion impedance amplitude are then obtained. The impedance normalization correction formula includes:

[0109] in, This is the impedance amplitude after temperature correction. This is the impedance amplitude before temperature correction. The temperature coupling coefficient is... For real-time cell temperature, The reference calibration temperature.

[0110] It is understandable that the reference calibration temperature is the standard temperature used when performing EIS impedance calibration tests, such as 20℃, 25℃, 30℃, etc.; the real-time cell temperature is the cell surface temperature obtained by real-time sampling, for example, the real-time cell temperature is 24℃ under normal operating conditions.

[0111] It should be noted that the temperature coupling coefficient reflects the sensitivity of the corresponding impedance amplitude to temperature changes. Ohmic impedance, interface impedance, and diffusion impedance have different temperature sensitivities and each has its own independent temperature coupling coefficient. For example, the temperature coupling coefficient for ohmic impedance is 0.0042 / ℃, the temperature coupling coefficient for interface impedance is 0.0085 / ℃, and the temperature coupling coefficient for diffusion impedance is 0.0110 / ℃.

[0112] It should be noted that the above values ​​are only used to explain the technical solution of this application and are not intended to limit the scope of protection of this application.

[0113] In the embodiments of this application, after temperature correction, the reversible drift component caused by temperature can be eliminated, and only the effective features corresponding to battery structural degradation and permanent interface aging are retained, which is beneficial to provide a clean input for subsequent high-precision state estimation.

[0114] Step 2: Obtain the pre-calibrated multidimensional feature mapping function, which includes multiple sets of test data and their corresponding impedance mapping SOC values. Each set of test data includes the interface impedance amplitude, interface impedance phase, diffusion impedance amplitude, and real-time cell temperature.

[0115] Specifically, the multidimensional feature mapping function is a high-precision mapping model calibrated offline in the early stages and embedded in the algorithm program, used to establish the nonlinear correlation between "impedance-temperature-SOC". During the calibration process, multiple sets of test data were constructed through extensive experimental tests covering the entire temperature range, the entire SOC range, and different aging states. Each set of test data includes the interface impedance amplitude, interface impedance phase, diffusion impedance amplitude, real-time cell temperature, and the corresponding impedance-mapped SOC value, which is the standard true SOC value.

[0116] In the embodiments of this application, by adapting the multidimensional feature mapping function to the characteristics of the flat voltage plateau and voltage indistinguishability in the middle SOC range of solid-state batteries, a precise nonlinear mapping from multidimensional electrochemical characteristics to SOC state can be achieved, making up for the defects of traditional voltage correction failure.

[0117] Step 3: Input the interface impedance amplitude, interface impedance phase, diffusion impedance amplitude, and real-time cell temperature into the multidimensional feature mapping function, and obtain the first impedance mapping SOC value through interpolation.

[0118] Specifically, the real-time collected interface impedance amplitude, interface impedance phase, diffusion impedance amplitude, and real-time cell temperature are input into a pre-calibrated multidimensional feature mapping function. Since the real-time collected feature parameters are difficult to perfectly match the calibration sample data, the algorithm uses high-precision multidimensional interpolation to fit and solve the calibration sample dataset, and outputs the first impedance mapping SOC value corresponding to the current battery state.

[0119] Understandably, this SOC value is calculated entirely based on the internal electrochemical impedance kinetics of the battery, and is not affected by the flattening of the voltage plateau. It can accurately reflect the true remaining charge state in the middle working range of the solid-state battery.

[0120] Step 4: Weight and fuse the first impedance-mapped SOC value with the SOC estimate at the current moment calculated by the dynamic Coulomb integral formula to obtain the fused SOC estimate.

[0121] Understandably, the SOC estimate after fusion is the SOC estimate obtained through the second estimation strategy.

[0122] Specifically, the SOC value mapped by the first impedance is weighted and fused with the SOC estimate at the current moment calculated using the dynamic Coulomb integral formula, to obtain the fused SOC estimate. For example: The fusion calculation formula includes:

[0123] in, This is the estimated SOC value after fusion. This is the estimated SOC value at the current moment, calculated using the dynamic Coulomb integral formula. This is the first impedance mapping SOC value. These are the weighting coefficients for the Coulomb integral, and their values ​​range from 0 to 1. <w<1。

[0124] Understandably, the Coulomb integral weighting coefficient can be set according to specific needs. For example, under steady-state conditions, the Coulomb integral weighting coefficient can be set to 0.4; under high-current disturbance conditions, it can be adaptively increased to prioritize timing smoothness.

[0125] By rationally configuring the weighting coefficients of the Coulomb integral, the time-series smoothness of the Coulomb integral is preserved, and the long-term cumulative error of the Coulomb integral is corrected in real time using impedance mapping SOC. This effectively solves the technical problem of low accuracy and poor stability of single algorithm estimation in the middle SOC range of solid-state batteries, and finally outputs a high-precision and high-stability fused SOC estimate.

[0126] (3) Third working interval and third estimation strategy Specifically, the SOC value of the third working interval is in the range of [a,b] or [c,d], for example: [28%,32%] or [68%,72%].

[0127] Understandably, the third working interval is a transitional interval between the first and second working intervals, and it is necessary to use the first and second estimation strategies to ensure that the estimation results are continuous, smooth, and without abrupt changes.

[0128] Specifically, the third estimation strategy includes:

[0129] in, The SOC estimate is calculated based on the third estimation strategy. For smoothing weighting coefficients, The SOC estimate is calculated based on the first estimation strategy. The SOC estimate is calculated based on the second estimation strategy.

[0130] In this embodiment, the smoothing weight coefficient changes linearly and gradually with the position of SOC within the transition interval. The value range is [0,1]. For example, in the interval [28%, 32%], the closer the SOC is to 28%, the higher the SOC becomes. The closer the SOC is to 1, the more likely the SOC estimate calculated based on the first estimation strategy will be accepted; the closer the SOC is to 32%, the more likely it is to be accepted. The closer the value is to 0, the more the SOC estimate calculated based on the second estimation strategy is accepted, thus achieving a flexible and seamless switching between the two algorithms and ensuring that the SOC curve is continuous without abrupt changes.

[0131] Step S123: Based on the available capacity and coulombic efficiency of the solid-state battery, calculate the first estimated SOC value corresponding to the current moment using the SOC estimation strategy.

[0132] It should be noted that all three estimation strategies rely on real-time updated available capacity and coulomb efficiency. In the first estimation strategy, available capacity and coulomb efficiency are directly substituted into the dynamic coulomb integral formula to complete the calculation. In the second estimation strategy, available capacity and coulomb efficiency are used to calculate the coulomb product branch. In the third estimation strategy, the SOC estimate is based on the value calculated using the first estimation strategy. SOC estimate calculated based on the second estimation strategy The values ​​are calculated by the first and second estimation strategies respectively, so the capacity and coulomb efficiency can be used to indirectly participate in the calculation of the third estimation strategy.

[0133] Therefore, SOC estimation uses the actual parameters of the battery under its current aging state throughout the process, rather than the fixed parameters from the factory, which can reduce the estimation deviation caused by aging.

[0134] Please refer to the following: Figure 3 , Figure 3 This is provided by the embodiments of this application. Figure 1 A detailed flowchart of step S104 is shown.

[0135] like Figure 3 As shown, step S104: Based on the adaptive dynamic weighted fusion model, the first SOH estimate is calculated, including the following steps S141-S142: Step S141: Obtain the bulk aging component, the interface comprehensive aging component, and the ion diffusion aging component.

[0136] Specifically, the component of aging. This reflects the aging degree of the ohmic impedance of the battery cell itself, and is used to characterize the aging and deterioration degree of the ohmic components of the solid-state battery cell. The cell aging component is calculated from the initial ohmic impedance amplitude and the temperature-corrected ohmic impedance amplitude. Specifically:

[0137] in, This is the temperature-corrected ohmic impedance amplitude. This is the initial ohmic impedance magnitude.

[0138] Specifically, the overall aging component of the interface This method considers both interface impedance rise and cumulative interface damage over cycles to characterize the overall aging and degradation of the solid-solid contact interface in solid-state batteries. It takes into account both the interface impedance rise effect and the irreversible interface damage accumulated during charge-discharge cycles. Specifically, the overall interface aging component is calculated from the initial interface impedance amplitude and the temperature-corrected interface impedance amplitude.

[0139] in, As the cumulative damage factor at the interface, The damage attenuation coefficient is... This is the temperature-corrected interfacial impedance amplitude. This is the initial interface impedance magnitude.

[0140] It should be noted that the damage attenuation coefficient It is a constant obtained through offline calibration of multiple cycles of solid-state battery aging, used to characterize the influence weight of the cumulative interface damage factor D on the overall interface aging component, such as the damage decay coefficient. =0.0025.

[0141] Specifically, the ion diffusion aging component is used to characterize the degree of degradation of the solid-state diffusion capability of lithium ions in the active material inside the solid-state battery. The ion diffusion aging component is calculated from the initial diffusion impedance amplitude and the temperature-corrected diffusion impedance amplitude. Specifically:

[0142] in, This is the ion diffusion aging component. This is the temperature-corrected diffusion impedance amplitude. This is the initial diffusion impedance amplitude.

[0143] It should be noted that the initial ohmic impedance amplitude, initial interfacial impedance amplitude, and initial diffusion impedance amplitude are all fixed reference parameters obtained by offline calibration of the solid-state battery using EIS electrochemical impedance spectroscopy at the reference calibration temperature under brand new factory conditions. These parameters are pre-stored in the non-volatile memory of the battery management system. The algorithm execution process directly reads this set of reference parameters and does not require real-time measurement during operation.

[0144] In some embodiments, the method further includes: calculating the cumulative damage factor of the interface, including: Construct an interface damage cumulative factor model; The cumulative damage factor of the interface is calculated based on the interface damage cumulative factor model. The interface damage cumulative factor model includes:

[0145] in, Accumulate the damage factor for the interface at the current moment. This represents the cumulative interface damage factor from the previous cycle. This is the initial interface impedance magnitude. This represents the relative change in interface impedance in real time. , This is the temperature-corrected interfacial impedance amplitude. These are the phase-weighted calibration coefficients. This refers to the interface phase attenuation. , This represents the initial interface impedance phase. This represents the phase of the interface impedance after temperature correction. For time period.

[0146] It should be noted that the phase weighting calibration coefficients These are dimensionless constants obtained from offline aging test calibration, such as: as an exemplary embodiment, phase-weighted calibration coefficients. Pick Initial interface impedance magnitude Take 80mΩ as the initial interface impedance phase. Take 0.42 rad, time period Take 1 second. This set of values ​​is only used to explain the technical solution of this application and is not intended to limit the scope of protection of this application.

[0147] Example calculation: Cumulative interface damage factor from the previous cycle =80; The current temperature-corrected interface impedance amplitude is 92mΩ, and the temperature-corrected interface impedance phase is 0.30rad. The calculated real-time relative change in interface impedance is 12mΩ, and the interface phase attenuation is 0.12rad; Substituting into the interface damage accumulation factor model, the damage increment for this period is 0.1572, and the updated current-time interface accumulation damage factor is... =80.1572.

[0148] It should be noted that the above values ​​are only used to explain the technical solution of this application and are not intended to limit the scope of protection of this application.

[0149] Furthermore, the method also includes: calculating the relative change in real-time interface impedance, including the following steps (1)-(8): Step (1): Obtain the first interface impedance timing, wherein the first interface impedance timing is the temperature-corrected interface impedance timing.

[0150] Specifically, the original interface impedance timing sequence is corrected point by point using the impedance normalization correction formula to eliminate temperature coupling interference, resulting in the temperature-corrected interface impedance timing sequence, which is denoted as the first interface impedance timing sequence.

[0151] Step (2): Construct the first time window, where the first time window corresponds to the first window duration.

[0152] Specifically, a short, narrow sliding time window is set as the first-level filtering statistical window (first time window). The first window has a short duration, for example, 50 seconds, to capture rapid impedance changes caused by short-term battery operating fluctuations and instantaneous polarization. This window is used to accurately distinguish between instantaneous disturbances and the steady-state impedance baseline, providing a time-series statistical interval for extracting reversible disturbance components.

[0153] Step (3): Based on the timing of the first interface impedance, calculate the short-term steady-state reference impedance corresponding to the first time window.

[0154] Specifically, within each sliding first time window, the mean or steady-state fitting calculation is performed on the first interface impedance time series data within the first time window to obtain the short-term steady-state reference impedance under the current short-term window. This reference impedance represents the steady-state baseline of the interface impedance after eliminating instantaneous jumps under the current short-term operating state of the battery. It can be finely adjusted in real time to follow the short-term operating state of the battery and serves as a comparison benchmark for identifying disturbance components.

[0155] Step (4): Calculate the short-time reversible disturbance component based on the short-term steady-state reference impedance.

[0156] Specifically, the difference between the real-time impedance data within the first time window and the short-term steady-state reference impedance is calculated to extract the fluctuation component that deviates from the steady-state baseline, which is the short-term reversible disturbance component.

[0157] Specifically, the difference between the real-time impedance data within the first time window and the short-term steady-state reference impedance is calculated using the following formula:

[0158] in, It is a short-time reversible perturbation component. This represents the number of sampling points corresponding to the first time window. The impedance amplitude at time t in the first time window. This is the short-term steady-state reference impedance.

[0159] Step (5): Filter the short-time reversible disturbance component to obtain the second interface impedance timing.

[0160] Step (6): Construct a second time window, where the second time window corresponds to the duration of the second window, and the duration of the second window is greater than the duration of the first window.

[0161] Specifically, a second sliding time window with a long duration and wide span is constructed, and the duration of the second window is significantly longer than that of the first window. For example, the duration of the second window is set to 2000 seconds.

[0162] Step (7): Sample the timing of the second interface impedance to obtain the long-term irreversible aging impedance component.

[0163] Specifically, within the second time window, steady-state sampling and trend fitting are performed on the denoised second interface impedance time series to extract the slow, monotonic, and irreversible impedance rise trend in the time series, thus obtaining the long-term irreversible aging impedance component, for example:

[0164] in, This is the impedance component of long-term irreversible aging. This represents the number of sampling points corresponding to the second time window. The impedance amplitude at time t in the second time window. This is the initial reference impedance of the solid-state battery cell at the factory.

[0165] Step (8): Based on the long-term irreversible aging impedance component, calculate the real-time relative change in interface impedance, including:

[0166] in, This represents the relative change in interface impedance in real time. The interface impedance amplitude is the value of the long-term irreversible aging impedance component in the kth calculation period. This represents the initial interface impedance amplitude of the impedance component over long-term irreversible aging.

[0167] It should be noted that, This represents the interface impedance amplitude in the kth calculation cycle of the long-term irreversible aging impedance component, where k is the algorithm iteration calculation cycle number; the calculation cycle number k increases by 1 after each round of interface impedance timing processing. This represents the initial interface impedance amplitude of the long-term irreversible aging impedance component. It is the interface impedance amplitude at the initialization moment of the long-term irreversible aging impedance tracking logic, i.e., the calculation cycle corresponding to k=0. This value is the tracking reference zero point obtained by sampling and recording when the BMS starts this aging tracking algorithm, which is different from the initial interface impedance amplitude obtained by offline calibration at the factory. . These are fixed constants pre-stored in the condition of a brand new battery cell when it leaves the factory. It depends on the actual aging state of the battery at the time the algorithm starts.

[0168] Step S142: Input the bulk aging component, the interface comprehensive aging component, and the ion diffusion aging component into the adaptive dynamic weighted fusion model to obtain the first SOH estimate.

[0169] The adaptive dynamic weighted fusion model includes:

[0170] in, This is an estimated value for SOH. As the weight of the aging component, This refers to the aging component of the substrate. Assign weights to the overall aging components of the interface. For the overall aging components of the interface, The weighting of the ion diffusion aging component. This is the ion diffusion aging component. ; in,

[0171] in, This is the temperature-corrected ohmic impedance amplitude. This is the initial ohmic impedance magnitude; in,

[0172] in, As the cumulative damage factor at the interface, The damage attenuation coefficient is... This is the temperature-corrected interfacial impedance amplitude. This is the initial interface impedance magnitude; in,

[0173] in, This is the temperature-corrected diffusion impedance amplitude. This is the initial diffusion impedance amplitude.

[0174] Please refer to the following: Figure 4 , Figure 4 This is a schematic diagram of a process for determining a corrected first SOC estimate provided in an embodiment of this application.

[0175] like Figure 4 As shown, the process for determining the corrected first SOC estimate includes the following steps S401-S407: Step S401: Obtain the current polarity of the solid-state battery.

[0176] Understandably, solid-state batteries exhibit a hysteresis effect in their relaxation open-circuit voltage after charging and discharging, meaning the OCV (Open Voltage Capacity) at the end of charging and the OCV at the end of discharging do not coincide. Therefore, by determining the current polarity—which includes whether the battery was in a charging or discharging state before the current rest period—the battery management system reads the charging and discharging currents within a certain timeframe prior to the rest period trigger. If the current before the rest period is the charging current, the current polarity is determined to be in a charging state; if the current before the rest period is the discharging current, the current polarity is determined to be in a discharging state.

[0177] Among them, the charging condition corresponds to the charging OCV-SOC lookup table, and the discharging condition corresponds to the discharging OCV-SOC lookup table. The charging OCV-SOC lookup table is used to store the correspondence between the reference open circuit voltage and the standard SOC correction value under the charging condition, and the discharging OCV-SOC lookup table is used to store the correspondence between the reference open circuit voltage and the standard SOC correction value under the discharging condition.

[0178] Step S402: Determine the corresponding first OCV-SOC lookup table based on the current current polarity.

[0179] Specifically, in charging mode, the charging OCV-SOC reference table is selected as the first OCV-SOC reference table; in discharging mode, the discharging OCV-SOC reference table is selected as the first OCV-SOC reference table. That is, the charging OCV-SOC reference table is selected for charging mode, and the discharging OCV-SOC reference table is selected for discharging mode, which is denoted as the first OCV-SOC reference table, in order to eliminate the correction error caused by charging and discharging hysteresis.

[0180] Step S403: Based on the SOC estimate of the previous cycle, consult the first OCV-SOC lookup table to determine the theoretical open-circuit voltage corresponding to the SOC estimate of the previous cycle.

[0181] Specifically, the SOC estimate from the previous iteration is used as the lookup index. A precise search is then performed in the first matched OCV-SOC lookup table to obtain the corresponding standard reference open-circuit voltage for that SOC state, which is the theoretical open-circuit voltage. For example, if the SOC estimate from the previous cycle is 50%, an interpolation lookup in the discharge OCV-SOC lookup table yields a theoretical open-circuit voltage of 3.380V.

[0182] Step S404: Obtain the steady-state open-circuit voltage of the solid-state battery.

[0183] Specifically, the steady-state open-circuit voltage is obtained iteratively based on the short-time relaxation voltage time series data and a pre-built relaxation fitting model.

[0184] Step S405: Based on the steady-state open-circuit voltage, calculate the voltage difference between the steady-state open-circuit voltage and the theoretical open-circuit voltage.

[0185] Specifically, the voltage difference = steady-state open-circuit voltage - theoretical open-circuit voltage. For example: if the calculated steady-state open-circuit voltage is 3.392V and the theoretical open-circuit voltage is 3.380V, then the voltage difference = steady-state open-circuit voltage - theoretical open-circuit voltage = 3.392V - 3.380V = 0.012V.

[0186] Step S406: Based on the voltage difference and the first OCV-SOC lookup table, obtain the local slope corresponding to the estimated SOC value, divide the voltage difference by the local slope, and obtain the SOC correction amount.

[0187] Specifically, the OCV-SOC lookup table is a set of points consisting of multiple reference open-circuit voltages and standard SOC correction values. The corresponding nonlinear curve can be obtained through the OCV-SOC lookup table. Since the OCV-SOC lookup table corresponds to a nonlinear curve, the voltage at different SOC points changes at different rates with SOC, so a globally fixed conversion factor cannot be used. The SOC correction is obtained by calculating the local slope of the nonlinear curve corresponding to the current estimated SOC value in the OCV-SOC lookup table, and dividing the voltage difference by the local slope.

[0188] For example: Based on the discrete point set of the first OCV-SOC lookup table, at the location of the estimated SOC value of the previous period, the local slope of the OCV-SOC curve is obtained. The physical meaning of this local slope is: at this SOC point, the change in open-circuit voltage corresponding to each unit change in SOC.

[0189] In this embodiment, the local slope can be solved using methods such as the central difference method and interpolation. Preferably, the local slope is calculated using the central difference method, which includes: obtaining a set of adjacent calibration sampling points before and after the previous sampling point and performing a difference operation to obtain the local slope corresponding to the SOC estimate, wherein the local slope corresponding to the SOC estimate is = (reference open-circuit voltage of the next sampling point - reference open-circuit voltage of the previous sampling point) / (SOC correction value of the next sampling point - SOC correction value of the previous sampling point).

[0190] Step S407: Determine the corrected first SOC estimate based on the SOC correction amount, wherein the corrected first SOC estimate = first SOC estimate + SOC correction amount.

[0191] For example, assuming the battery was in discharge condition before being left to stand, the discharge OCV-SOC reference table is selected as the first OCV-SOC reference table; the estimated value of the first SOC in the previous cycle is 48.5%, and the corresponding theoretical open-circuit voltage is obtained by interpolation from the table as 3.380V; after relaxation fitting, the current steady-state open-circuit voltage of the solid-state battery is obtained as 3.392V; the voltage difference is calculated to be +0.012V; at this SOC point, the local slope of the OCV-SOC curve is read as 0.08V per unit SOC; the SOC correction is calculated to be +1.5%; adding the SOC correction (+1.5%) to the estimated value of the first SOC (48.5%), the corrected estimated value of the first SOC is 50.0%.

[0192] It should be noted that the above values ​​are only used to explain the technical solution of this application and are not intended to limit the scope of protection of this application.

[0193] In this embodiment of the application, the SOC estimate is corrected by the SOC correction amount, which can eliminate the systematic estimation error caused by the solid-solid interface polarization hysteresis of the all-solid-state battery and improve the estimation accuracy of the SOC estimate.

[0194] Please refer to the following: Figure 5 , Figure 5 This is a schematic diagram of a process for calculating steady-state open-circuit voltage provided in an embodiment of this application.

[0195] like Figure 5 As shown, the process for calculating the steady-state open-circuit voltage includes the following steps S501-S502: Step S501: If the solid-state battery meets the steady-state resting conditions, then obtain the short-time relaxation voltage timing data for a preset duration.

[0196] The steady-state resting conditions include: the normalized absolute value of the current of the solid-state battery is less than a preset current threshold, and the duration for which the normalized absolute value is less than the preset current threshold is greater than a preset time threshold.

[0197] Specifically, the steady-state resting conditions need to be met simultaneously: the normalized absolute value of the current of the solid-state battery is less than a preset current threshold; and the duration of this state is greater than a preset time threshold.

[0198] Specifically, the normalized absolute value of the current is the ratio value obtained by dividing the measured current by the rated capacity of the solid-state battery. Calculating the normalized absolute value of the current includes: obtaining the rated capacity of the solid-state battery, taking the absolute value of the measured current value, and dividing the absolute value of the current value by the rated capacity of the solid-state battery to obtain the normalized absolute value of the current.

[0199] In the embodiments of this application, the preset current threshold and the preset time threshold can be set according to specific needs. For example, the preset current threshold is set to 0.05C and the preset time threshold is set to 3s. These values ​​are only used to explain the technical solution of this invention and are not intended to limit the scope of protection of this application.

[0200] After the steady-state resting conditions are met, the terminal voltage timing data is collected within a preset time period, i.e., short-time relaxation voltage timing data. For example, if the preset time period is 60 seconds, the short-time voltage timing data from 0 to 60 seconds after the resting period is extracted as a fitting sample. This time window can completely cover the entire process of rapid relaxation of the interface and slow relaxation of the solid phase in the all-solid-state battery, without the need for a long resting period of several hours.

[0201] For example, the preset current threshold is 0.02C, the preset time threshold is 5s, and the preset duration for acquiring the short-time relaxation voltage sequence is 60s.

[0202] Taking a solid-state battery with a rated capacity of 20Ah as an example: if the absolute value of the measured current is less than 0.4A, and this low current state is maintained for more than 3 seconds, then the steady-state resting condition is met; the terminal voltage time series data for the next 60 seconds from the start of the resting is extracted as the input sample for relaxation fitting. If only a small current appears briefly, but the duration is less than 3 seconds, it is judged as a false resting, and the steady-state open-circuit voltage solution process is not initiated.

[0203] In this embodiment, since it is not necessary to wait for the battery to fully relax to a constant state, the steady-state OCV can be extrapolated using a fitting algorithm, thus shortening the waiting time. Furthermore, if a sudden change in current or a switch in operating conditions occurs during the resting process, the current fitting calculation is immediately terminated, and the valid OCV result from the previous cycle is retained, ensuring data validity and algorithm stability.

[0204] Step S502: Based on the short-time relaxation voltage time series data, the steady-state open-circuit voltage is obtained iteratively based on the pre-built relaxation fitting model.

[0205] Specifically, relaxation fitting models include:

[0206] in, This represents the real-time terminal voltage at time t in the short-time relaxation voltage time series data. The steady-state open-circuit voltage, For the interface fast polarization amplitude coefficient, This represents the amplitude coefficient of the slow diffusion polarization in the solid phase. The interface polarization adaptive dynamic time constant. The adaptive dynamic time constant for diffusion polarization.

[0207] It should be noted that the interface fast polarization amplitude coefficient This represents the voltage deviation amplitude corresponding to the rapid polarization effect at the solid-solid interface, for example, the value range is -0.08V to 0.08V; solid-phase slow diffusion polarization amplitude coefficient This represents the voltage deviation amplitude corresponding to the slow diffusion process of solid-phase lithium ions, for example, the value range is -0.04V to 0.04V; Interface polarization adaptive dynamic time constant This is the adaptive dynamic time constant, a dynamically changing parameter that varies with the SOC and cell temperature of the solid-state battery. It is updated synchronously during the iterative fitting process. For example, the interface polarization adaptive dynamic time constant. The value range is from 5s to 40s; Diffusion polarization adaptive dynamic time constant The adaptive dynamic time constant is a dynamically changing parameter that changes with the SOC and cell temperature of the solid-state battery. It is updated synchronously during the iterative fitting process. For example, the diffusion polarization adaptive dynamic time constant. The value range is from 60s to 300s.

[0208] It should be noted that the above values ​​are only used to explain the technical solution of this application and are not intended to limit the scope of protection of this application.

[0209] Based on the short-time relaxation voltage time series data, and using a pre-built relaxation fitting model, the steady-state open-circuit voltage is obtained iteratively, including the following steps (1)-(4): (1) Construct the residual objective function, where the residual objective function is used to calculate the residual, which is the sum of the squares of the differences between the real-time terminal voltage in the short-time relaxation voltage time series data and the fitted open-circuit voltage.

[0210] Specifically, the residual is the sum of the squares of the differences between the measured terminal voltage and the model-fitted voltage at each sampling point in the time series; the smaller the residual, the better the fitting effect. For example, the residual objective function is:

[0211] Wherein, residual objective function P Used for parameter identification in bi-exponential relaxation models; The measured terminal voltage was collected at time t during the battery's resting phase. This is the model-fitted voltage calculated from the parameters to be identified based on a two-stage relaxation model; the objective function P is the sum of the squares of the voltage residuals at all sampling times from 0 to 60 seconds. A Levenburg-Marquardt nonlinear least squares iterative algorithm is employed, with minimizing the objective function P as the optimization objective. First, the parameters to be identified are initialized, the model-fitted voltage is calculated, and the residual objective function is solved. The open-circuit voltage, relaxation amplitude coefficients, and dynamic time constant are iteratively updated, and the objective function is repeatedly calculated. When the change in the objective function is less than a preset convergence threshold or the maximum number of iterations is reached, the iteration stops, and the optimally identified open-circuit voltage is output.

[0212] (2) Synchronously iteratively update the steady-state open-circuit voltage, the interface fast polarization amplitude coefficient, the solid-phase slow diffusion polarization amplitude coefficient, the interface polarization adaptive dynamic time constant, and the diffusion polarization adaptive dynamic time constant.

[0213] Specifically, the Levenberg-Marquardt nonlinear least squares algorithm is used to synchronously iteratively update the five unknown parameters: steady-state open-circuit voltage, interface fast polarization amplitude coefficient, solid-phase slow diffusion polarization amplitude coefficient, interface polarization adaptive dynamic time constant, and diffusion polarization adaptive dynamic time constant; the objective function is the sum of squares of voltage residuals in the 0~60s static sampling interval.

[0214] First, the five-dimensional parameters are initialized. The initial value of the open-circuit voltage is taken as the measured terminal voltage at the initial resting time, and the initial value of the relaxation time constant is calculated based on the current battery SOH according to the aging evolution formula. In each iteration, the residual between the fitted voltage and all sampling times is calculated based on the current parameter vector, and the Jacobian matrix of the residual with respect to the five-dimensional parameters is solved. The LM linear equation system is constructed, and the correction increments of all five parameters are obtained at one time, and the parameters are updated synchronously. , , , , After the update is complete, , SOH constraint boundary clamping is performed; the new objective function residual is calculated based on the updated parameters, the damping coefficient is dynamically adjusted, and effective iteration steps that reduce the residual are retained. When the parameter increment or the change in the objective function is less than the preset convergence threshold, or the maximum number of iterations is reached, the iteration terminates, and the optimal open-circuit voltage and optimal relaxation parameters after convergence are output; thus, the optimal steady-state open-circuit voltage is output.

[0215] It should be noted that the damping coefficient is an adjustable algorithm parameter in the LM (Levonburg-Marquardt) nonlinear least squares iterative algorithm, with a value greater than or equal to zero. It is used to control the step size of the correction update of the five-dimensional parameters in each iteration, balancing the iteration convergence speed and numerical stability. For example, the initial damping coefficient can be 0.01; during the iteration process, when the residual decreases, the damping coefficient can be reduced to 1 / 3 of its original value; when the residual increases, the damping coefficient can be amplified to 3 times its original value.

[0216] Understandably, a larger damping coefficient suppresses parameter correction step size more effectively, resulting in more conservative iterative updates and reducing the likelihood of parameter overshoot and divergence, but also slows down convergence. Conversely, a smaller damping coefficient allows for larger parameter update step sizes, leading to faster convergence, but also introduces the risk of parameter oscillation and divergence. In this embodiment, the damping coefficient is not fixed before iteration but dynamically adjusted each iteration based on residual changes: if the objective function residual decreases after parameter updates, the iteration step is considered valid, and the damping coefficient is reduced to allow for a larger step size in the next iteration to accelerate convergence; if the residual increases after parameter updates, the iteration step is considered invalid, the parameter update result is discarded, and the damping coefficient is increased while the update step size is reduced to recalculate the parameter increment. By dynamically adjusting the damping coefficient, it is possible to prevent parameters such as the relaxation time constant from falling outside the physically reasonable range in the early stages of iteration and to achieve rapid convergence in the later stages, ultimately obtaining the optimal steady-state open-circuit voltage and the corresponding relaxation model parameters.

[0217] (3) Calculate the residuals in real time during each iteration; (4) If the residual is less than the preset convergence threshold, or the number of iterations is greater than the maximum number of iterations threshold, then the iteration process ends and the converged steady-state open-circuit voltage is obtained.

[0218] Understandably, the preset convergence threshold is the stopping threshold of the residuals, used to determine whether the fitting effect is good enough. For example: the preset convergence threshold is set to... When the sum of squared residuals obtained from the iteration is less than the threshold, it means that the fitting accuracy has met the engineering requirements, and there is no need to continue the iteration. The iteration is terminated directly and the steady-state open-circuit voltage is output.

[0219] Understandably, the maximum number of iterations threshold is a safe upper limit for iterative operations. It is used to prevent the iterative operation from failing to converge and resulting in an infinite loop when there is significant noise in the voltage timing or data anomalies. When the number of iterations reaches this threshold, the iteration is forcibly terminated regardless of whether the residual meets the condition. The set of parameters with the smallest residual during the iteration process is selected as the optimal parameters, and the converged steady-state open-circuit voltage is output. For example, the maximum number of iterations threshold is set to 25 times.

[0220] It should be noted that the above values ​​are only used to explain the technical solution of this application and are not intended to limit the scope of protection of this application. In the embodiments of this application, the steady-state open-circuit voltage (OCV) is solved by using short-time relaxation voltage time series data, which can accurately adapt to the two-stage relaxation characteristics of all-solid-state batteries, namely "fast interface relaxation + slow solid-phase diffusion relaxation", thereby improving the accuracy of the steady-state open-circuit voltage.

[0221] Please refer to the following: Figure 6 , Figure 6 This is provided by the embodiments of this application. Figure 1A detailed flowchart of step S105 in the process.

[0222] like Figure 6 As shown, step S105: Based on the first SOH estimate, update the available capacity and coulombic efficiency of the solid-state battery, including the following steps S151-S155: Step S151: Based on the rated capacity of the solid-state battery and the first SOH estimate, obtain the updated available capacity.

[0223] Specifically, based on the rated capacity of the solid-state battery and the first estimated state of harm (SOH), the updated usable capacity is obtained, including:

[0224] in, For the updated available capacity, This refers to the rated capacity of solid-state batteries. This is the first estimated SOH value.

[0225] Specifically, by using real-time SOH estimates to correct the available capacity of solid-state batteries, the fixed rated capacity is updated to the current actual available capacity of the battery, so that the denominator parameter of subsequent ampere-hour integration matches the actual aging level of the battery.

[0226] Step S152: Obtain the coulomb efficiency reference table, which includes multiple sets of cell temperature, current ratio, and current polarity correspondences.

[0227] Specifically, the coulomb efficiency benchmark table is an offline calibrated multi-dimensional lookup table. Input dimensions: cell temperature, current rate, current polarity (charging / discharging); output: benchmark coulomb efficiency. This benchmark coulomb efficiency characterizes the coulomb efficiency of a brand new battery under this operating condition. The coulomb efficiency benchmark table is used to provide an original efficiency benchmark without aging and pure operating condition effects, providing a basic benchmark value for subsequent aging correction.

[0228] Step S153: Obtain the current cell temperature, current ratio, and current polarity. Based on the coulomb efficiency reference table, obtain the reference coulomb efficiency through interpolation.

[0229] Specifically, the current cell temperature, current ratio, and current polarity are read in real time. Based on the data in the coulomb efficiency benchmark table, interpolation calculations are performed to obtain the benchmark coulomb efficiency corresponding to the current cell temperature, current ratio, and current polarity.

[0230] Step S154: Calculate the aging correction factor based on the first SOH estimate.

[0231] Specifically, based on the first SOH estimate, the aging correction factor is calculated, including:

[0232] in, This is the aging correction factor. This is the calibration coefficient for interface aging degradation. This is the first estimated SOH value.

[0233] Understandably, the aging correction factor Used for aging degradation correction of the benchmark coulomb efficiency, interface aging degradation calibration coefficient This is a dimensionless constant obtained through offline calibration using multiple cycles of solid-state battery aging tests. It characterizes the strength of the impact of interface aging damage on coulombic efficiency degradation. For example, it is the interface aging degradation calibration coefficient. =0.18.

[0234] Step S155: Calculate the updated coulombic efficiency based on the baseline coulombic efficiency and the aging correction factor, including:

[0235] in, For the updated Coulomb efficiency, As a benchmark coulomb efficiency, This is the aging correction factor.

[0236] In this embodiment, on the one hand, the available capacity is directly corrected by the SOH estimate to solve the problem of the power metering baseline offset caused by aging; on the other hand, the dynamic coulombic efficiency is solved by a two-layer mechanism of "operating condition lookup table benchmark + aging damage correction", which adapts to transient fluctuations in operating conditions and slow aging over the life cycle. Ultimately, the subsequent SOC estimation is based on the current real aging parameters of the battery throughout the entire process, and the SOH aging state is used to reverse correct the SOC accuracy.

[0237] Please refer to the following: Figure 7 , Figure 7 This is provided by the embodiments of this application. Figure 1 A detailed flowchart of step S106 is shown.

[0238] like Figure 7 As shown, step S106: Based on the first SOC estimate, determine the battery operating range in which the first SOC estimate is located, in order to update the aging component weight, including the following steps S161-S164: Step S161: Obtain the benchmark weight combination corresponding to each battery working interval.

[0239] In this process, a set of benchmark weight combinations is assigned to each battery operating range. Each set of benchmark weight combinations contains three fixed benchmark weights: the benchmark weight of the bulk aging component, the benchmark weight of the interface comprehensive aging component, and the benchmark weight of the ion diffusion aging component.

[0240] Step S162: Based on the first SOC estimate, determine the first benchmark weight combination corresponding to the battery operating range where the first SOC estimate is located.

[0241] Specifically, based on the first SOC estimate, the battery operating range in which the first SOC estimate is located is determined, and using the battery operating range in which the first SOC estimate is located, the first benchmark weight combination corresponding to the battery operating range is determined.

[0242] Step S163: Determine the aging stage of the solid-state battery based on the first SOH estimate.

[0243] The aging process includes the early aging stage, the middle aging stage, and the late aging stage.

[0244] Specifically, a first SOH threshold and a second SOH threshold are set, dividing the aging process into three stages: (1) Early aging stage: The estimated SOH value is greater than the first SOH threshold, indicating that the battery body and interface are slightly damaged, and the aging is mainly reflected in the decline of lithium ion diffusion capacity.

[0245] (2) Mid-term aging stage: The second SOH threshold < the estimated SOH value < the first SOH threshold, which indicates that the three types of aging—bulk, interface, and diffusion—develop simultaneously.

[0246] (3) Late aging stage: SOH estimated value < second SOH threshold; solid-solid interface is severely deteriorated, bulk structure damage is prominent, and diffusion decay tends to saturate.

[0247] In the embodiments of this application, the first SOH threshold and the second SOH threshold can be set according to specific needs, for example, the first SOH threshold and the second SOH threshold can be set to 0.9 and 0.7 respectively.

[0248] Step S164: Adjust the first benchmark weight combination according to the aging stage of the solid-state battery to obtain the updated aging component weight.

[0249] Specifically, based on the aging stage of the solid-state battery, the first baseline weight combination is adjusted to obtain the updated aging component weights, including: If the solid-state battery is in the early aging stage, the basis weights of the bulk aging component and the interface comprehensive aging component are reduced, and the basis weights of the ion diffusion aging component are increased to obtain the updated aging component weights. Among them, the estimated SOH value corresponding to the early aging stage is greater than or equal to the first SOH threshold. If the solid-state battery is in the mid-aging stage, the basis weights of the bulk aging component and the ion diffusion aging component are increased, while the basis weight of the interface comprehensive aging component is maintained to obtain the updated aging component weights. Among them, the estimated SOH value corresponding to the mid-aging stage is less than the first SOH threshold and greater than the second SOH threshold. If the solid-state battery is in the late aging stage, the basis weights of the bulk aging component and the interface comprehensive aging component are increased, while the basis weights of the ion diffusion aging component are decreased to obtain the updated aging component weights. Among them, the estimated SOH value corresponding to the late aging stage is less than the second SOH threshold.

[0250] Please see Figure 8 , Figure 8 This is a schematic diagram of a process for optimizing the second SOC estimate and the second SOH estimate provided in an embodiment of this application.

[0251] like Figure 8 As shown, the process for optimizing the second SOC estimate and the second SOH estimate includes the following steps S801-S804: Step S801: Construct a dual-state equation, using the SOC and SOH values ​​as the state variables to be optimized.

[0252] Specifically, the state vector includes: ;in, SOC value This refers to the SOH value.

[0253] Specifically, the two-state equations of state include: ; in, This is the state vector for the current period. This is the mean vector of the SOC-SOH states after convergence in the previous cycle. It is a nonlinear state transfer function. The process noise is used to characterize the model recursion error. Its statistical properties are described by the process noise covariance matrix Q, and it supports cycle-by-cycle adaptive updates.

[0254] Step S802: Construct a multi-source observation equation, using the observed values ​​of steady-state open-circuit voltage, multidimensional impedance characteristics, and interface cumulative aging damage as joint observation inputs.

[0255] Specifically, the multi-source observation equations include:

[0256] in, This is a multi-source observation vector, including observed values ​​of steady-state open-circuit voltage, multidimensional impedance characteristics, and interface cumulative aging damage. The observed values ​​of multidimensional impedance characteristics include interface impedance amplitude, interface impedance phase, and diffusion impedance amplitude, while the observed values ​​of interface cumulative aging damage include the real-time relative change in interface impedance.

[0257] Step S803: Introduce process noise and observation noise, and update the process noise covariance matrix and observation noise covariance matrix in real time adaptively.

[0258] Specifically, process noise and observation noise are introduced to describe the recursive error of the state model and the observation error of the sensor, respectively. The process noise covariance matrix and the observation noise covariance matrix are then adaptively updated periodically based on the real-time operating conditions of the battery and the disturbance state.

[0259] Step S804: Perform unscented transformation, sampling point propagation, and weighted regression calculation to complete the synchronous optimal estimation of the second SOC estimate and the second SOH estimate, and output the third SOC estimate and the third SOH estimate.

[0260] Specifically, unscented transformation, sampling point propagation, and weighted regression calculations are performed to approximate the nonlinear probability distribution and predict the state of the two state variables constituted by the second SOC estimate and the second SOH estimate, achieving simultaneous optimal estimation of both, including: First, an unscented transformation is performed: using the state mean vector composed of the second SOC estimate and the second SOH estimate, along with the corresponding error covariance matrix, as input; for the two-dimensional SOC-SOH state, (2n+1) Sigma sampling points are deterministically generated, and each Sigma sampling point is assigned a corresponding mean weight and covariance weight. This set of Sigma sampling points can equivalently represent the mean and covariance statistical properties of the original state, without needing to perform Taylor linearization approximation on the nonlinear state equation coupled with SOC-SOH, thus avoiding truncation error.

[0261] Secondly, sample point propagation is performed: all Sigma sample points are successively substituted into the two-state nonlinear state equation to complete the nonlinear state recursion propagation, obtaining the predicted Sigma sample point after time-series evolution for each sample point. This process directly uses the original nonlinear state mapping relationship without introducing linear approximation.

[0262] Next, weighted regression calculation is performed: using pre-assigned weights, weighted statistical regression is performed on the predicted Sigma sampling points after all propagation is completed to calculate the mean of the prior predicted state and the prior predicted covariance matrix for this period, thus completing the model-driven state prediction.

[0263] After obtaining the prior predicted state, the multi-source observation residuals are calculated by combining the observed values ​​of steady-state open-circuit voltage, multi-dimensional impedance characteristics, and interface cumulative aging damage. The residual weights are adaptively scaled based on the duration of the operating disturbance, and the observation noise covariance matrix is ​​updated to solve for the adaptive Kalman gain. The prior predicted state is linearly corrected using the Kalman gain and weighted residuals to obtain the posterior optimal state estimate. The third SOC estimate and the third SOH estimate are output simultaneously.

[0264] Understandably, the second SOC estimate and the second SOH estimate are coarse estimation results output by the partitioned SOC estimation strategy and the adaptive dynamic weighted fusion model; the third SOC estimate and the third SOH estimate are the optimal estimation results output after joint optimization by the Adaptive Unscented Kalman Filter (AUKF). Specifically, an unscented transformation is used, and a nonlinear state distribution is fitted through high-precision Sigma sampling points. Through state equation propagation, multi-source observation residual correction, and weighted regression iteration, simultaneous optimal estimation of the two state variables is achieved. Finally, the denoised, converged, and corrected final SOC estimate and final SOH estimate are output, completing the joint optimization.

[0265] Understandably, the Unscented Transform (UT) is the core algorithm of the Unscented Kalman Filter (UKF), which is a nonlinear probability distribution approximation sampling method.

[0266] In this embodiment of the application, the adaptive unscented Kalman filter algorithm includes multiple cycles, and the iterative process of a single cycle includes the following steps (1)-(8): Step (1): Read the optimal state values ​​of SOC, SOH, and error covariance matrix after convergence in the previous cycle, and use them as the initial reference parameters for this iteration; Step (2): Based on the state mean and error covariance matrix of the previous period, generate 2n+1 Sigma sampling points symmetrically; Step (3): Substitute all Sigma sampling points into the dual-state equation to complete the state prediction propagation operation and obtain the predicted state mean and prediction covariance matrix for this iteration. Step (4): Dynamically adjust the process noise covariance matrix and the observation noise covariance matrix according to the real-time operating conditions of the solid-state battery; Step (5): Combine the observations of steady-state open-circuit voltage, multidimensional impedance characteristics, and interface cumulative aging damage, and solve for the state residuals of each observation; calculate the adaptive Kalman gain based on the adaptive scaling of residual weights and observation noise matrix during the operating condition disturbance duration; and use the Kalman gain and weighted residuals to linearly correct the predicted state.

[0267] Specifically, the residual weights and observation noise matrix are adaptively scaled based on the duration of the operating condition disturbance, including the following steps ①-②: Step ①: Adaptively scale the residual weights based on the duration of the operating condition disturbance, including: For short-term disturbance conditions, since instantaneous voltage and impedance fluctuations are pseudo-disturbances and do not represent a shift in the true state, reducing the residual weight weakens the interference of instantaneous observation residuals on state correction, thus preventing the battery state from being skewed or experiencing jumps due to a single disturbance. For long-term steady-state conditions, the observed data are stable and reliable, without random disturbances. By increasing the residual weight, the correction effect of the real-state residual is amplified, accurately offsetting the long-term small drift of the model.

[0268] Step ②: Adaptively update the observation noise covariance matrix R based on the perturbation duration, including: If the duration of the operating condition disturbance is longer (non-steady state), it indicates that the observation noise is large. In this case, the value of the observation noise covariance matrix is ​​adaptively amplified to reduce the confidence of the observation data and avoid overcorrection and tracking distortion under dynamic operating conditions. If the duration of the operating condition disturbance is shorter (steady state), it indicates that the observation noise is smaller. The value of the observation noise covariance matrix is ​​adaptively reduced to improve the confidence of the true observation value and correct the model bias.

[0269] Specifically, calculating the adaptive Kalman gain includes: Based on the adaptively updated prediction covariance matrix and the scaled observation noise matrix R, the adaptive Kalman gain K is calculated in real time. For example, when the disturbance duration is long and the observation noise is large, the adaptive Kalman gain is reduced; when the disturbance duration is short and the observation is stable and reliable, the adaptive Kalman gain is increased.

[0270] Specifically, linear corrections to the predicted state are made using Kalman gain and weighted residuals, including: The original residuals corresponding to the observed values ​​of steady-state open-circuit voltage, multidimensional impedance characteristics, and interface cumulative aging damage are multiplied by adaptive operating condition weights and then weighted and fused to obtain the comprehensive state residual. The integrated state residuals are linearly weighted and corrected using adaptive Kalman gain. The prior predicted state is iteratively updated to obtain the corrected posterior optimal state value, so as to simultaneously optimize the second SOC estimate and the second SOH estimate, and complete the convergence of this round of filtering.

[0271] In this embodiment, by using single-variable closed-loop parameter tuning for the duration of operating condition disturbance, the residual weight, observation noise, and Kalman gain are updated in a linked and adaptive manner, ensuring that SOC and SOH do not jump under dynamic operating conditions and have strong stability. Furthermore, it can improve the estimation accuracy under steady-state operating conditions.

[0272] Step (6): After the linear correction is completed, update the optimal state values ​​of SOC and SOH to obtain the optimal state values ​​of SOH and SOC after iterative updates, and complete the single iteration convergence.

[0273] Step (7): Based on the optimal SOH state value updated in this iteration, correct the available capacity and coulombic efficiency required for the SOC estimation in the next cycle; and based on the battery operating range where the optimal SOC state value updated in this iteration is located, update the aging component weight of the adaptive dynamic weighted fusion model.

[0274] Step (8): After a single algorithm cycle is completed, store the currently converged optimal state values ​​of SOC and SOH as the initial parameters for the next cycle, and continue to run the iterative process in a loop.

[0275] This application employs a multimodal operating condition fault tolerance mechanism to automatically freeze instantaneous abnormal parameters and enable historical steady-state parameters for smooth transition when operating conditions fluctuate drastically, signal-to-noise ratio is too low, or EIS excitation is abnormal.

[0276] Specifically, the multimodal operating condition fault tolerance mechanism pre-defines three types of quantifiable abnormal operating condition modes. Fault tolerance protection is activated when any one of these abnormal operating condition modes is met. The three types of abnormal operating condition modes include: (1) Severe fluctuation mode of operating conditions: Real-time monitoring of the change rate of loop current. When the instantaneous fluctuation rate of current exceeds the normal steady state range, it is judged as a severe dynamic operating condition, which means that the battery is in a high-power transient impact and pulse charging and discharging scenario, the polarization state changes drastically, and the original observation data is distorted.

[0277] (2) Low signal-to-noise ratio signal failure mode: Real-time calculation of the signal-to-noise ratio (SNR) of voltage and current signals. When SNR < 20dB, it is determined that the sampled signal is submerged by electromagnetic noise and operating noise, and the effective features are ineffective. Directly substituting it into the calculation will cause serious errors.

[0278] (3) Abnormal EIS excitation acquisition mode: Real-time verification of the validity of multi-frequency EIS excitation response. When any phenomenon occurs, such as impedance amplitude mutation rate exceeding the threshold, excitation response phase distortion, or frequency band data missing, the current EIS feature acquisition is deemed invalid, and the interface impedance, diffusion impedance, and phase characteristics of the current frame are unavailable.

[0279] Specifically, the multimodal fault tolerance mechanism employs a five-tiered execution logic: "anomaly detection → parameter freezing and locking → smooth steady-state transition → continuous anomaly verification → adaptive recovery." This ensures a complete closed loop within a single algorithm cycle without interrupting the main algorithm's iteration flow. The complete execution process of this multimodal fault tolerance mechanism is as follows: First-order: Real-time determination of abnormal operating conditions At the beginning of each UKF iteration cycle, the above three types of working condition modes are detected synchronously and in parallel, the current working condition disturbance level is quantified, and the data state of this frame is marked as steady-state valid data or abnormal failure data, so as to realize the early identification of anomalies.

[0280] Second-order: Instantaneous abnormal parameter freezing If this frame is determined to be an abnormal operating condition, the update permissions for all current real-time abnormal observation parameters will be immediately frozen. The specific parameters to be frozen include: instantaneous EIS amplitude and phase characteristics, instantaneous temperature correction impedance, instantaneous voltage relaxation fitting OCV, and instantaneous current integral increment.

[0281] By refraining from adopting real-time distorted observations in abnormal frames and forcibly discarding the weight contribution of abnormal data in this frame, abnormal noise can be prevented from being added to the observation equation.

[0282] Third-order: Enable time-series smoothing transition using historical steady-state parameters. After the parameters are frozen, the system automatically retrieves the historical steady-state optimal parameter set that has converged and stabilized in the previous z cycles (e.g., the first 5 cycles), and uses weighted time-series interpolation to generate the virtual steady-state observation value of the current frame, replacing the distorted real-time data to ensure that the algorithm input is continuous and without breaks.

[0283] By enabling the smooth transition of historical steady-state parameters, it is possible to maintain a smooth steady-state output without interruption of the algorithm, sudden changes in parameters, or drift in state under abnormal operating conditions.

[0284] Fourth-order: Dynamic verification of persistent abnormal states The system continuously monitors the operating status in real time, distinguishing between short-term transient anomalies and long-term continuous anomalies. (1) Short-term transient anomaly (duration < 2s): Continue to maintain the freeze + smooth transition strategy to ensure state stability.

[0285] (2) Long-term extreme disturbances (duration ≥ 2s): Adaptively amplify UKF observation noise R, weaken the weight of abnormal observations, strengthen the historical steady-state constraints, and prevent the filter state from diverging.

[0286] Fifth-order: Adaptive fault tolerance removal during condition recovery When the operating conditions return to steady state, and simultaneously the following conditions are met: current fluctuation rate, signal-to-noise ratio, and EIS response quality all return to the normal threshold range, the system gradually unfreezes the system. It uses a weighted gradual change method to smoothly switch back to real-time sampled data from historical steady-state interpolation data, avoiding a secondary jump during recovery and achieving fault-tolerant closed-loop self-healing.

[0287] In this embodiment, a multimodal fault-tolerant mechanism can shield transient abnormal interference without interrupting the algorithm iteration process or disrupting the bidirectional coupling relationship between SOC and SOH, thus ensuring continuous, stable, and seamless operation of the algorithm.

[0288] Please refer to the following: Figure 9 , Figure 9 This is a schematic diagram of the overall process of a joint estimation method for SOC and SOH provided in an embodiment of this application.

[0289] like Figure 9 As shown, the overall flow of the joint estimation method for SOC and SOH includes the following steps S901-S917: Step S901: Power on the battery and start running.

[0290] Specifically, after the solid-state battery is powered on, the battery management system (BMS) is initialized, loading the initial available capacity, initial coulombic efficiency, and other relevant parameters.

[0291] Step S902: Collect voltage, current, and cell temperature.

[0292] Specifically, through sampling circuits and temperature sensors, three basic operating condition data—cell terminal voltage, charging and discharging current, and cell temperature—are collected in real time, providing the original input data source for all subsequent state estimations.

[0293] Step S903: Data noise reduction processing.

[0294] Specifically, noise reduction processing is applied to voltage, current, and cell temperature.

[0295] Step S904: Extract impedance characteristics, including impedance amplitude and impedance phase.

[0296] Specifically, through multi-band impedance scanning, multi-dimensional electrochemical characteristics such as interface impedance amplitude, interface impedance phase, and diffusion impedance amplitude are obtained, providing core feature support for SOC impedance mapping, aging component calculation, and state observation.

[0297] Step S905: Perform temperature correction on the impedance characteristics.

[0298] Specifically, the impedance characteristics are temperature-corrected using the impedance normalization correction formula, resulting in temperature-corrected impedance characteristics such as: temperature-corrected ohmic impedance amplitude, temperature-corrected interface impedance amplitude, and temperature-corrected diffusion impedance amplitude.

[0299] Step S906: Determine that the battery is in a steady-state static condition.

[0300] Specifically, if the solid-state battery meets the steady-state resting conditions, then the battery is determined to be in a steady-state resting condition. The steady-state resting conditions include: the normalized absolute value of the current of the solid-state battery is less than a preset current threshold, and the duration for which the normalized absolute value is less than the preset current threshold is greater than a preset time threshold.

[0301] Step S907: Calculate the steady-state open-circuit voltage.

[0302] Specifically, based on the short-time relaxation voltage time series data, the steady-state open-circuit voltage is obtained iteratively using a pre-built relaxation fitting model.

[0303] Step S908: Determine the battery operating range.

[0304] Step S909: Determine the SOC estimation strategy based on the battery operating range.

[0305] Step S910: Calculate the first SOC estimate based on the SOC estimation strategy.

[0306] Step S911: Determine the long-term irreversible aging impedance component.

[0307] Step S912: Calculate the first SOH estimate based on the adaptive dynamic weighted fusion model.

[0308] Step S913: Calculate the SOH correction amount.

[0309] Step S914: Determine the first estimated SOH value.

[0310] Step S915: Optimize the first SOH estimate and the first SOC estimate based on the adaptive unscented Kalman filter algorithm.

[0311] Step S916: Output the second SOH estimate and the second SOC estimate.

[0312] Step S917: Repeat the iterative process continuously.

[0313] It should be noted that the relevant content of steps S901-S917 above can be referred to the content mentioned in the above embodiments, and will not be repeated here.

[0314] In this embodiment, a joint estimation method for SOC and SOH is provided and applied to solid-state batteries. The method includes: obtaining the estimated SOC value, available capacity, and coulombic efficiency of the solid-state battery from the previous cycle; calculating the first estimated SOC value corresponding to the current moment based on the estimated SOC value, available capacity, and coulombic efficiency of the solid-state battery from the previous cycle; obtaining the bulk aging component, the interface comprehensive aging component, and the ion diffusion aging component, wherein each aging component corresponds to an aging component weight; calculating the first estimated SOH value based on an adaptive dynamic weighted fusion model, wherein the adaptive dynamic weighted fusion model is used to calculate the weighted sum obtained by multiplying each aging component by its corresponding aging component weight, and using the weighted sum as the first estimated SOH value; updating the available capacity and coulombic efficiency of the solid-state battery based on the estimated SOH value, and obtaining the second estimated SOC value based on the updated available capacity and coulombic efficiency; determining the battery operating range in which the estimated SOC value is located based on the first estimated SOC value, updating the aging component weights, and substituting them into the adaptive dynamic weighted fusion model to obtain the second estimated SOH value.

[0315] By acquiring the SOC estimate, available capacity, and coulombic efficiency of the solid-state battery from the previous cycle, the first SOC estimate corresponding to the current moment is calculated. Furthermore, based on the adaptive dynamic weighted fusion model, the first SOH estimate is calculated. Further, based on the SOH estimate, the available capacity and coulombic efficiency of the solid-state battery are updated to update the SOC estimate. Simultaneously, based on the SOC estimate, the battery operating range in which the SOC estimate is located is determined to update the SOH estimate. This application can achieve bidirectional coupling and mutual correction of SOC and SOH, improving the estimation accuracy of SOC and SOH.

[0316] Please see Figure 10 , Figure 10 This is a schematic diagram of the structure of an electronic device provided in an embodiment of this application.

[0317] like Figure 10 As shown, the electronic device 100 includes one or more processors 101 and a memory 102. Wherein, Figure 10 Take a processor 101 as an example.

[0318] Processor 101 and memory 102 can be connected via a bus or other means. Figure 10 Taking the example of a connection between China and Israel via a bus.

[0319] Processor 101 provides computing and control capabilities to control electronic device 100 to perform corresponding tasks, such as controlling electronic device 100 to execute the joint estimation method of SOC and SOH in any of the above method embodiments. This joint estimation method of SOC and SOH is applied to solid-state batteries. The method includes: obtaining the estimated SOC value of the previous cycle, the available capacity of the solid-state battery, and the coulombic efficiency; calculating the first estimated SOC value corresponding to the current moment based on the estimated SOC value of the previous cycle, the available capacity of the solid-state battery, and the coulombic efficiency; and obtaining the bulk aging component, the interface comprehensive aging component, and the ion diffusion aging component, wherein the ion diffusion aging component is corrected by the initial diffusion impedance amplitude and temperature. The diffusion impedance amplitude is calculated, and each aging component corresponds to an aging component weight. Based on the adaptive dynamic weighted fusion model, the first SOH estimate is calculated. The adaptive dynamic weighted fusion model is used to calculate the weighted sum obtained by multiplying each aging component by its corresponding aging component weight, and the weighted sum is used as the first SOH estimate. Based on the SOH estimate, the available capacity and coulombic efficiency of the solid-state battery are updated, and the second SOC estimate is obtained based on the updated available capacity and coulombic efficiency. Based on the first SOC estimate, the battery operating range in which the SOC estimate is located is determined, the aging component weights are updated, and the results are substituted into the adaptive dynamic weighted fusion model to obtain the second SOH estimate.

[0320] By acquiring the SOC estimate, available capacity, and coulombic efficiency of the solid-state battery from the previous cycle, the first SOC estimate corresponding to the current moment is calculated. Furthermore, based on the adaptive dynamic weighted fusion model, the first SOH estimate is calculated. Further, based on the first SOH estimate, the available capacity and coulombic efficiency of the solid-state battery are updated to update the SOC estimate. Simultaneously, based on the SOC estimate, the battery operating range in which the SOC estimate is located is determined to update the SOH estimate. This application can achieve bidirectional coupling and mutual correction of SOC and SOH, improving the estimation accuracy of SOC and SOH.

[0321] Processor 101 can be a general-purpose processor, including a central processing unit (CPU), a network processor (NP), a hardware chip, or any combination thereof; it can also be a digital signal processor (DSP), an application-specific integrated circuit (ASIC), a programmable logic device (PLD), or a combination thereof. The aforementioned PLD can be a complex programmable logic device (CPLD), a field-programmable gate array (FPGA), a generic array logic (GAL), or any combination thereof.

[0322] Memory 102, as a non-transitory computer-readable storage medium, can be used to store non-transitory software programs, non-transitory computer-executable programs, and modules, such as the program instructions / modules corresponding to the joint estimation method of SOC and SOH in the embodiments of this application. Processor 101 can implement the joint estimation method of SOC and SOH in any of the above method embodiments by running the non-transitory software programs, instructions, and modules stored in memory 102. Specifically, memory 102 may include volatile memory (VM), such as random access memory (RAM); memory 102 may also include non-volatile memory (NVM), such as read-only memory (ROM), flash memory, hard disk drive (HDD), solid-state drive (SSD), or other non-transitory solid-state storage devices; memory 102 may also include combinations of the above types of memory.

[0323] Memory 102 may include high-speed random access memory, and may also include non-volatile memory, such as at least one disk storage device, flash memory device, or other non-volatile solid-state storage device. In some embodiments, memory 102 may optionally include memory remotely located relative to processor 101, and these remote memories may be connected to processor 101 via a network. Examples of such networks include, but are not limited to, the Internet, intranets, local area networks, mobile communication networks, and combinations thereof.

[0324] One or more modules are stored in memory 102. When executed by one or more processors 101, they perform the joint estimation method of SOC and SOH in any of the above method embodiments. For example, they perform the methods described above. Figure 1 The steps shown.

[0325] In this embodiment, the electronic device 100 may also have wired or wireless network interfaces, keyboards, and input / output interfaces for input and output. The electronic device 100 may also include other components for implementing device functions, which will not be described in detail here.

[0326] In this application embodiment, the electronic device includes: a battery management system (BMS) for a power battery in an electric vehicle (typically a metal-cased box mounted on the side or top of the battery pack); or a battery cluster management unit (BCU) or battery management host in an energy storage power station; or a battery protection board (intelligent BMS with advanced computing capabilities) in a portable device or special power supply.

[0327] It should be noted that the relevant content of this electronic device can be found in the above embodiments, and will not be repeated here.

[0328] This application also provides a non-volatile computer-readable storage medium, such as a memory including program code, which can be executed by a processor to complete the joint estimation method of SOC and SOH in the above embodiments. For example, the non-volatile computer-readable storage medium may be a read-only memory (ROM), a random access memory (RAM), a compact disc read-only memory (CDROM), magnetic tape, floppy disk, and optical data storage device, etc.

[0329] This application also provides a computer program product comprising one or more lines of program code stored in a non-volatile computer-readable storage medium. A processor of an electronic device reads the program code from the non-volatile computer-readable storage medium and executes the program code to complete the method steps of the joint estimation method for SOC and SOH provided in the above embodiments.

[0330] Those skilled in the art will understand that all or part of the steps of the above embodiments can be implemented by hardware, or by a program or program code related to hardware. The program can be stored in a computer-readable storage medium, such as a read-only memory, a disk, or an optical disk.

[0331] Through the above description of the embodiments, those skilled in the art can clearly understand that each embodiment can be implemented using software plus a general-purpose hardware platform, or of course, using hardware. Those skilled in the art will understand that all or part of the processes in the above embodiments can be implemented by a computer program instructing related hardware. The program can be stored in a non-volatile computer-readable storage medium. When executed, the program can include the processes of the embodiments of the above methods. The non-volatile computer-readable storage medium can be a magnetic disk, optical disk, read-only memory (ROM), or random access memory (RAM), etc.

[0332] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of this application, and not to limit them; under the concept of this application, the technical features of the above embodiments or different embodiments can also be combined, the steps can be implemented in any order, and there are many other variations as described above in different aspects of this application, which are not provided in detail for the sake of brevity; although this application has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that they can still modify the technical solutions described in the foregoing embodiments, or make equivalent substitutions for some of the technical features; and these modifications or substitutions do not cause the essence of the corresponding technical solutions to deviate from the scope of the technical solutions of the embodiments of this application.

Claims

1. A method for jointly estimating SOC and SOH, characterized in that, Applied to solid-state batteries, the method includes: Obtain the estimated SOC value of the previous cycle, the available capacity of the solid-state battery, and the coulombic efficiency; Based on the SOC estimate of the previous cycle, the available capacity of the solid-state battery, and the coulombic efficiency, calculate the first SOC estimate corresponding to the current moment. The bulk aging component, the interface comprehensive aging component, and the ion diffusion aging component are obtained. The ion diffusion aging component is calculated from the initial diffusion impedance amplitude and the temperature-corrected diffusion impedance amplitude. Each aging component corresponds to an aging component weight. Based on the adaptive dynamic weighted fusion model, the first SOH estimate is calculated, wherein the adaptive dynamic weighted fusion model is used to calculate the weighted sum obtained by multiplying each aging component by its corresponding aging component weight, and the weighted sum is used as the first SOH estimate. Based on the first SOH estimate, the available capacity and coulombic efficiency of the solid-state battery are updated, and based on the updated available capacity and coulombic efficiency, a second SOC estimate is obtained. Based on the first SOC estimate, the battery operating range in which the first SOC estimate is located is determined, the aging component weight is updated, and then substituted into the adaptive dynamic weighted fusion model to obtain the second SOH estimate.

2. The method according to claim 1, characterized in that, The step of calculating the first estimated SOC value at the current moment based on the SOC estimate of the previous cycle, the available capacity of the solid-state battery, and the coulombic efficiency includes: Based on the SOC estimate from the previous cycle, determine the battery operating range in which the SOC estimate falls; Based on the battery operating range, determine the SOC estimation strategy corresponding to the SOC estimate; Based on the available capacity and coulombic efficiency of the solid-state battery, the first estimated SOC value at the current moment is calculated using the SOC estimation strategy.

3. The method according to claim 2, characterized in that, The step of determining the SOC estimation strategy corresponding to the SOC estimate based on the battery operating range includes: Obtain a first mapping relationship, wherein the first mapping relationship includes a correspondence between at least two battery operating ranges and at least two SOC estimation strategies, and the battery operating ranges and the SOC estimation strategies correspond one-to-one; Based on the first mapping relationship, and considering the battery operating range where the estimated SOC value lies, a SOC estimation strategy corresponding to the estimated SOC value is determined, wherein... The battery operating range includes a first operating range, and the SOC estimation strategy includes a first estimation strategy, with the first operating range corresponding to the first estimation strategy. The first estimation strategy includes: calculating a first SOC estimate corresponding to the SOC estimate based on a dynamic Coulomb integral formula, wherein the dynamic Coulomb integral formula includes: in, Here is the estimated SOC value at time t. Let SOC be the estimated value of the previous time step at time t. For Coulomb efficiency, Let be the current at time t. For a fixed sampling period, The available capacity of the solid-state battery is given.

4. The method according to claim 3, characterized in that, The method further includes: Obtain the current polarity of the solid-state battery, wherein the current polarity includes charging condition or discharging condition, the charging condition corresponds to a charging OCV-SOC lookup table, the discharging condition corresponds to a discharging OCV-SOC lookup table, the charging OCV-SOC lookup table is used to store the correspondence between the reference open circuit voltage and the standard SOC correction value under the charging condition, and the discharging OCV-SOC lookup table is used to store the correspondence between the reference open circuit voltage and the standard SOC correction value under the discharging condition; Based on the current current polarity, determine the corresponding first OCV-SOC lookup table; Based on the SOC estimate of the previous cycle, consult the first OCV-SOC lookup table to determine the theoretical open-circuit voltage corresponding to the SOC estimate of the previous cycle. Obtain the steady-state open-circuit voltage of the solid-state battery; Based on the steady-state open-circuit voltage, calculate the voltage difference between the steady-state open-circuit voltage and the theoretical open-circuit voltage; Based on the voltage difference, and using the first OCV-SOC lookup table, the local slope corresponding to the estimated SOC value is obtained. The voltage difference is then divided by the local slope to obtain the SOC correction amount. Based on the SOC correction amount, a corrected first SOC estimate is determined, wherein the corrected first SOC estimate = the first SOC estimate + the SOC correction amount.

5. The method according to claim 4, characterized in that, The method further includes: Calculating the steady-state open-circuit voltage of the solid-state battery includes: If the solid-state battery meets the steady-state resting condition, then short-time relaxation voltage timing data of a preset duration is obtained; Based on the short-time relaxation voltage time series data, the steady-state open-circuit voltage is obtained iteratively using a pre-built relaxation fitting model. The steady-state resting condition includes: the normalized absolute value of the current of the solid-state battery is less than a preset current threshold, and the duration for which the normalized absolute value is less than the preset current threshold is greater than a preset time threshold.

6. The method according to claim 5, characterized in that, The relaxation fitting model includes: in, The real-time terminal voltage at time t in the short-time relaxation voltage time series data. The steady-state open-circuit voltage, For the interface fast polarization amplitude coefficient, This represents the amplitude coefficient of the slow diffusion polarization in the solid phase. The interface polarization adaptive dynamic time constant. The adaptive dynamic time constant for diffusion polarization; The step of iteratively obtaining the steady-state open-circuit voltage based on the short-time relaxation voltage time-series data and a pre-built relaxation fitting model includes: Construct a residual objective function, wherein the residual objective function is used to calculate the residual, which is the sum of the squares of the differences between the real-time terminal voltage and the fitted open-circuit voltage in the short-time relaxation voltage time series data; The steady-state open-circuit voltage, interface fast polarization amplitude coefficient, solid-phase slow diffusion polarization amplitude coefficient, interface polarization adaptive dynamic time constant, and diffusion polarization adaptive dynamic time constant are updated synchronously and iteratively. In each iteration, the residual is calculated in real time; If the residual is less than the preset convergence threshold, or the number of iterations is greater than the maximum number of iterations threshold, then the iteration process ends and the converged steady-state open-circuit voltage is obtained.

7. The method according to claim 3, characterized in that, The battery operating range includes a second operating range, and the SOC estimation strategy includes a second estimation strategy, with the second operating range corresponding to the second estimation strategy. The second estimation strategy includes: Obtain the interface impedance amplitude, interface impedance phase, diffusion impedance amplitude, and real-time cell temperature; Obtain a pre-calibrated multidimensional feature mapping function, wherein the multidimensional feature mapping function includes multiple sets of test data and their corresponding impedance mapping SOC values, and each set of test data includes interface impedance amplitude, interface impedance phase, diffusion impedance amplitude, and real-time cell temperature. The interface impedance amplitude, interface impedance phase, diffusion impedance amplitude, and real-time cell temperature are input into the multidimensional feature mapping function, and the first impedance mapping SOC value is obtained through interpolation. The first impedance-mapped SOC value is weighted and fused with the SOC estimate at the current moment calculated by the dynamic Coulomb integral formula to obtain the fused SOC estimate.

8. The method according to claim 7, characterized in that, The battery operating range includes a third operating range, the SOC estimation strategy includes a third estimation strategy, and the third operating range corresponds to the third estimation strategy; The third estimation strategy includes: in, The SOC estimate is calculated based on the third estimation strategy. For smoothing weighting coefficients, The SOC estimate is calculated based on the first estimation strategy. The SOC estimate is calculated based on the second estimation strategy.

9. The method according to claim 1, characterized in that, The calculation of the first SOH estimate based on the adaptive dynamic weighted fusion model includes: The bulk aging component, the interface comprehensive aging component, and the ion diffusion aging component are input into the adaptive dynamic weighted fusion model to obtain the first SOH estimate. The adaptive dynamic weighted fusion model includes: in, This is the first estimated SOH value. As the weight of the aging component, This is the component of the body's aging process. Assign weights to the overall aging components of the interface. For the overall aging components of the interface, The weighting of the ion diffusion aging component. This is the ion diffusion aging component. .

10. The method according to claim 9, characterized in that, The method further includes: Construct an interface damage cumulative factor model; Based on the aforementioned interface damage accumulation factor model, the interface cumulative damage factor is calculated. The interface damage cumulative factor model includes: in, Accumulate the damage factor for the interface at the current moment. This represents the cumulative interface damage factor from the previous cycle. This is the initial interface impedance magnitude. This represents the relative change in interface impedance in real time. , This is the temperature-corrected interfacial impedance amplitude. These are the phase-weighted calibration coefficients. This refers to the interface phase attenuation. , This represents the initial interface impedance phase. This represents the temperature-corrected interface impedance phase. For time period.

11. The method according to claim 10, characterized in that, The method further includes: The calculation of the relative change in the real-time interface impedance includes: Obtain the first interface impedance timing, wherein the first interface impedance timing is a temperature-corrected interface impedance timing; Construct a first time window, where the first time window corresponds to the duration of the first window; Based on the first interface impedance timing, calculate the short-term steady-state reference impedance corresponding to the first time window; Calculate the short-time reversible disturbance component based on the short-term steady-state reference impedance; The short-time reversible disturbance component is filtered to obtain the second interface impedance timing. Construct a second time window, wherein the second time window corresponds to a second window duration, and the second window duration is longer than the first window duration; The timing sequence of the second interface impedance is sampled to obtain the long-term irreversible aging impedance component. Based on the long-term irreversible aging impedance component, the relative change in real-time interface impedance is calculated, including: in, This represents the relative change in interface impedance in real time. The interface impedance amplitude is the value of the k-th calculation period of the long-term irreversible aging impedance component. The initial interface impedance amplitude of the long-term irreversible aging impedance component.

12. The method according to claim 11, characterized in that, Before obtaining the timing sequence of the first interface impedance, the method further includes: The impedance amplitude and impedance phase corresponding to each frequency point are collected to determine the ohmic impedance amplitude and ohmic impedance phase corresponding to the first frequency band; the interface impedance amplitude and interface impedance phase corresponding to the second frequency band; and the diffusion impedance amplitude and diffusion impedance phase corresponding to the third frequency band. Temperature correction is applied to the ohmic impedance amplitude, interface impedance amplitude, and diffusion impedance amplitude to obtain the temperature-corrected ohmic impedance amplitude, temperature-corrected interface impedance amplitude, and temperature-corrected diffusion impedance amplitude.

13. The method according to claim 12, characterized in that, Temperature correction is applied to the ohmic impedance amplitude, interface impedance amplitude, and diffusion impedance amplitude to obtain temperature-corrected ohmic impedance amplitude, temperature-corrected interface impedance amplitude, and temperature-corrected diffusion impedance amplitude, including: Obtain the temperature coupling coefficient and real-time cell surface temperature; Calculate the temperature deviation, where the temperature deviation = real-time cell surface temperature - reference calibration temperature; Based on the temperature coupling coefficient and the real-time cell surface temperature, the ohmic impedance amplitude, interface impedance amplitude, and diffusion impedance amplitude are respectively temperature-corrected using the impedance normalization correction formula to obtain the temperature-corrected ohmic impedance amplitude, temperature-corrected interface impedance amplitude, and temperature-corrected diffusion impedance amplitude. The impedance normalization correction formula includes: in, This is the temperature-corrected impedance amplitude. This is the impedance amplitude before temperature correction. The temperature coupling coefficient is... For real-time cell temperature, The reference calibration temperature.

14. The method according to claim 9, characterized in that, The step of determining the battery operating range in which the first SOC estimate falls, based on the first SOC estimate, and updating the aging component weights, includes: Obtain the benchmark weight combination corresponding to each battery working range, wherein the benchmark weight combination includes the benchmark weight of the bulk aging component, the benchmark weight of the interface comprehensive aging component, and the benchmark weight of the ion diffusion aging component. Based on the first SOC estimate, determine the first benchmark weight combination corresponding to the battery operating range in which the first SOC estimate is located; Based on the SOH estimate, the aging stage of the solid-state battery is determined, wherein the aging stage includes an early aging stage, a mid-aging stage, and a late aging stage. Based on the aging stage of the solid-state battery, the first benchmark weight combination is adjusted to obtain the updated aging component weights.

15. The method according to claim 1, characterized in that, The step of updating the available capacity and coulombic efficiency of the solid-state battery based on the first SOH estimate includes: Based on the rated capacity of the solid-state battery and the first SOH estimate, the updated usable capacity is obtained, including: in, For the updated available capacity, This refers to the rated capacity of solid-state batteries. This is the first estimated SOH value; Obtain a coulomb efficiency reference table, wherein the coulomb efficiency reference table includes multiple sets of correspondences between cell temperature, current ratio, and current polarity; Obtain the current cell temperature, current rate, and current polarity, and obtain the reference coulomb efficiency through interpolation based on the coulomb efficiency reference table; Based on the first SOH estimate, the aging correction factor is calculated, including: in, This is the aging correction factor. This is the calibration coefficient for interface aging degradation. This is the first estimated SOH value; The updated coulomb efficiency is calculated based on the baseline coulomb efficiency and the aging correction factor, including: in, For the updated Coulomb efficiency, As a benchmark coulomb efficiency, This is the aging correction factor.

16. An electronic device, characterized in that, include: At least one processor; At least one memory for storing at least one program; When at least one of the programs is executed by at least one of the processors, the at least one of the processors implements the method according to any one of claims 1 to 15.

Images