A method, apparatus, electronic device and dielectric for predicting the DC internal resistance of a battery.

Abstract

This application provides a method, apparatus, electronic device, and medium for predicting the DC internal resistance of a battery. The method includes: acquiring battery state data, including the battery's state of charge (SOC), temperature, and state of health (SQH) at a target time; and inputting the SOC into a prediction model of the battery to obtain a predicted DC internal resistance value. The prediction model, based on a historical SOC dataset, constructs a correspondence between the battery's DC internal resistance value and its SOC, temperature, and SQH values. This application achieves accurate prediction of the battery's DC internal resistance.

Description

Technical Field

[0001] This application relates to the field of power battery technology, and more specifically, to a method, apparatus, electronic device, and medium for predicting the DC internal resistance of a battery. Background Technology

[0002] As the mainstream electrochemical energy storage device, the performance, safety, and lifespan of lithium-ion batteries are crucial to the success of applications such as electric vehicles and energy storage power stations. One of the core functions of a Battery Management System (BMS) is to accurately assess the battery's state. Direct Current Resistance (DCR), typically defined as the ratio of the total voltage drop across the battery terminals over a specified time under a specific pulse current to that current, is a key parameter characterizing the combined effects of ohmic losses, electrochemical polarization, and concentration polarization within the battery. Accurate DCR values ​​are the physical basis for a BMS to accurately predict the battery's maximum input / output power.

[0003] However, the characterization of DCR is extremely complex. It is strongly influenced by various factors such as State of Charge (SOC), temperature, current rate, and health status. Especially in low-temperature environments, the DCR value will increase exponentially, and the measurement data is often accompanied by strong noise. Its relationship curve with SOC often shows an asymmetric "J" or "L" shape, which poses a huge challenge to accurate modeling and prediction. Summary of the Invention

[0004] In view of this, the purpose of this application is to provide a method, apparatus, electronic device and medium for predicting the DC internal resistance of a battery, which aims to overcome at least one of the above-mentioned defects.

[0005] In a first aspect, this application provides a method for predicting the DC internal resistance of a battery. The method includes: acquiring state data of the battery, the state data including the state of charge, temperature, and health status of the battery at a target time; inputting the state data into a prediction model of the battery to obtain a predicted value of the DC internal resistance of the battery, wherein the prediction model constructs a correspondence between the DC internal resistance value of the battery and the state of charge, temperature, and health status of the battery based on the historical state dataset of the battery.

[0006] In one possible implementation, the prediction model is constructed as follows: A historical state dataset of the battery is obtained, comprising historical DC internal resistance measurements corresponding to multiple states of charge (SOC) values ​​at multiple temperature values ​​and corresponding health values ​​at each temperature; multiple smoothed DC internal resistance curves of the battery are determined based on the historical state data, and a base resistance value corresponding to each smoothed DC internal resistance curve is determined; a functional relationship model is established based on the multiple smoothed DC internal resistance curves and the base resistance value of each smoothed DC internal resistance curve to characterize the base resistance value of the battery and the temperature and health values ​​of the battery; a target base resistance value of the battery at a target temperature and a target health value, and a morphological function describing the change of DC internal resistance value with SOC, are determined based on the functional relationship model and the morphological function, and the prediction model is obtained based on the functional relationship model and the morphological function.

[0007] In one possible implementation, the morphological function includes a first morphological function for characterizing the left ascending wing and a second morphological function for characterizing the right ascending wing, wherein the first morphological function and the second morphological function are determined by: determining the first plateau region boundary and the second plateau region boundary of the target smooth DC internal resistance curve according to the functional relationship model; constructing the first morphological function based on the first plateau region boundary and the target substrate internal resistance value, and constructing the second morphological function based on the second plateau region boundary and the target substrate internal resistance value. In one possible implementation, the substrate internal resistance value corresponding to each smooth DC internal resistance curve is determined as follows: Based on the historical state dataset, multiple initial DC internal resistance sequences of the battery are determined; for each initial DC internal resistance sequence, a smoothing filter is applied to obtain a smooth DC internal resistance curve corresponding to that initial DC internal resistance sequence, and the bottom central region of the smooth DC internal resistance curve is determined; within the bottom central region, statistical outlier removal processing is performed on historical DC internal resistance measurements in the initial DC internal resistance sequence to obtain a purified data subset; local statistical values ​​of the purified data subset are calculated, and local minimum values ​​of the smooth DC internal resistance curve within the bottom central region are extracted; for each smooth DC internal resistance curve, the substrate internal resistance value corresponding to that smooth DC internal resistance curve is obtained based on the local statistical value and the local minimum value.

[0008] In one possible implementation, the boundaries of the first and second platform regions are determined as follows: Starting from the base resistance value of the smoothed DC resistance curve, the local first-order differential gradient value of each data point on the smoothed DC resistance curve is determined in a first direction, along with the relative increase of the DC resistance value corresponding to that data point relative to the base resistance value. The state of charge value corresponding to the data point that first simultaneously satisfies the condition that the local first-order differential gradient value continuously exceeds a preset gradient threshold and the relative increase first exceeds a preset amplitude threshold is determined as the boundary of the first platform region. Starting from the base resistance value of the smoothed DC resistance curve, the local first-order differential gradient value of each data point on the smoothed DC resistance curve is determined in a second direction, along with the relative increase of the DC resistance value corresponding to that data point relative to the base resistance value. The state of charge value corresponding to the data point that first simultaneously satisfies the condition that the local first-order differential gradient value continuously exceeds a preset gradient threshold and the relative increase first exceeds a preset amplitude threshold is determined as the boundary of the second platform region.

[0009] In one possible implementation, the first morphological function is represented by the following formula:

[0010] in, This is the predicted DC internal resistance value of the left ascending wing section. This is the substrate internal resistance value. For the left wing's ascent altitude parameters, For the left wing attenuation width parameter, This is the boundary of the first platform area. This represents the state of charge (SOC). The second morphological function is expressed by the following formula:

[0011] in, This is the predicted DC internal resistance value of the right ascending wing section. For the right wing's ascent altitude parameters, For the attenuation width parameter of the right wing, This is the boundary of the second platform area.

[0012] In one possible implementation, when the state of charge value is higher than a preset threshold, the predicted DC internal resistance value is calculated using the following formula:

[0013] in, The predicted DC internal resistance value corresponding to the i-th state of charge value that is higher than the preset threshold. For smoothing coefficients, The predicted DC internal resistance value corresponding to the (i-1)th state of charge value that is higher than the preset threshold. Let be the target stability constant.

[0014] Secondly, this application provides a battery DC internal resistance prediction device, the device comprising: an acquisition module for acquiring battery state data, the state data including the battery's state of charge value, temperature value, and health state value at a target time; and a prediction module for inputting the state data into a prediction model of the battery to obtain a predicted value of the battery's DC internal resistance, wherein the prediction model constructs a correspondence between the battery's DC internal resistance value and the battery's state of charge value, temperature value, and health state value based on the battery's historical state dataset.

[0015] Thirdly, this application also provides an electronic device, including: a processor, a memory, and a bus, wherein the memory stores machine-readable instructions executable by the processor, and when the electronic device is running, the processor communicates with the memory via the bus, and when the machine-readable instructions are executed by the processor, the steps of the method described above are performed.

[0016] Fourthly, this application also provides a computer-readable storage medium storing a computer program that, when executed by a processor, performs the steps of the method described above.

[0017] This application provides a method for predicting the DC internal resistance of a battery. The method includes: acquiring battery state data, including the battery's state of charge (SOC), temperature, and health status at a target time; and inputting the SOC data into a prediction model for the battery to obtain a predicted DC internal resistance value. The prediction model, based on a historical SOC dataset, constructs a correspondence between the battery's DC internal resistance value and its SOC, temperature, and health status values. This application achieves accurate prediction of the battery's DC internal resistance.

[0018] To make the above-mentioned objectives, features and advantages of this application more apparent and understandable, preferred embodiments are described below in detail with reference to the accompanying drawings. Attached Figure Description

[0019] To more clearly illustrate the technical solutions of the embodiments of this application, the accompanying drawings used in the embodiments will be briefly introduced below. It should be understood that the following drawings only show some embodiments of this application and should not be regarded as a limitation of the scope. For those skilled in the art, other related drawings can be obtained based on these drawings without creative effort.

[0020] Figure 1 A flowchart illustrating a method for predicting the DC internal resistance of a battery, provided in an embodiment of this application; Figure 2 This is a flowchart illustrating the construction of a prediction model provided in an embodiment of this application; Figure 3 A flowchart for determining the substrate internal resistance value corresponding to each smooth DC internal resistance curve provided in this application embodiment; Figure 4 A comparison chart of the modeling curves of this application and traditional methods provided for embodiments of this application; Figure 5 This is a schematic diagram of the battery DC internal resistance prediction device provided in the embodiments of this application; Figure 6 This is a schematic diagram of the structure of an electronic device provided in an embodiment of this application. Detailed Implementation

[0021] To make the objectives, technical solutions, and advantages of the embodiments of this application clearer, the technical solutions of the embodiments of this application will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of this application, and not all embodiments. The components of the embodiments of this application described and shown in the accompanying drawings can generally be arranged and designed in various different configurations. Therefore, the following detailed description of the embodiments of this application provided in the accompanying drawings is not intended to limit the scope of the claimed application, but merely represents selected embodiments of this application. Based on the embodiments of this application, every other embodiment obtained by those skilled in the art without inventive effort falls within the scope of protection of this application.

[0022] First, the applicable scenarios for this application will be introduced. This application can be applied to the field of power batteries.

[0023] Research has shown that lithium-ion batteries, as the mainstream electrochemical energy storage device, are crucial to the success or failure of applications such as electric vehicles and energy storage power stations due to their performance, safety, and lifespan. One of the core functions of a Battery Management System (BMS) is to accurately assess the battery's state, with the prediction of peak power being particularly critical, directly impacting the system's power performance, energy recovery efficiency, and operational safety.

[0024] Direct current resistance (DCR) is typically defined as the ratio of the total voltage drop across the battery terminals over a specified time under a specific pulse current to that current. It is a key parameter characterizing the combined effects of ohmic losses, electrochemical polarization, and concentration polarization within the battery. Accurate DCR values ​​are the physical basis for a battery management system (BMS) to precisely predict the battery's maximum input / output power.

[0025] However, the characterization of DCR is extremely complex, as it is strongly influenced by nonlinear factors such as State of Charge (SOC), temperature, current rate, and health status. Especially in low-temperature environments, the DCR value increases exponentially, and the measurement data is often accompanied by strong noise. Furthermore, its relationship curve with SOC often exhibits an asymmetric "J" or "L" shape, posing a significant challenge to accurate modeling.

[0026] Existing technologies have the following inherent limitations in DCR modeling: Traditional lookup table method: relies on massive offline calibration experiments, is costly, and consumes a large amount of storage resources in the BMS embedded system. This method is a static model, which cannot adapt to the performance degradation caused by battery aging, and the interpolation accuracy for uncalibrated operating conditions is insufficient.

[0027] Empirical formula fitting: When using methods such as global high-order polynomials to fit DCR data with complex shapes, overfitting or Runge phenomenon is very likely to occur, resulting in poor model generalization ability and prediction results deviating from physical reality.

[0028] State observers based on equivalent circuit models, such as the Extended Kalman Filter (EKF), rely on first- or second-order equivalent circuit models that contain structural errors in describing the complex electrochemical processes of batteries. Furthermore, these algorithms typically assume that noise follows a Gaussian distribution, while actual DCR measurement noise, especially at low temperatures, often exhibits non-Gaussian and nonlinear characteristics, leading to performance degradation or even divergence in the filtering algorithm.

[0029] Purely data-driven "black box" models, such as conventional neural networks, while possessing powerful nonlinear fitting capabilities, lack clear physical meaning in their model parameters, making prediction results difficult to interpret and reliability difficult to verify, thus failing to meet the requirements of high-safety-level applications such as automotive functional safety (ISO 26262).

[0030] Based on this, embodiments of this application provide a method, apparatus, electronic device, and medium for predicting the DC internal resistance of a battery, aiming to overcome at least one of the above-mentioned defects.

[0031] Please see Figure 1 , Figure 1 This is a flowchart illustrating a battery DC internal resistance prediction method provided in an embodiment of this application. Figure 1 As shown in the embodiments of this application, the battery DC internal resistance prediction method includes: S101, Obtain battery status data.

[0032] Specifically, the status data includes the battery's state of charge, temperature, and health status at the target time.

[0033] S102. Input the state data into the battery's prediction model to obtain the predicted value of the battery's DC internal resistance.

[0034] Here, the prediction model constructs a correspondence between the battery's DC internal resistance and its state of charge, temperature, and health status based on the battery's historical state data.

[0035] The following is through Figure 2 This section describes the specific process of building a prediction model.

[0036] Figure 2 This is a flowchart illustrating the construction of a prediction model provided in an embodiment of this application.

[0037] S201. Obtain the historical state dataset of the battery.

[0038] Here, the historical state dataset includes the battery at multiple temperature values ​​and the corresponding health state value for each temperature value, as well as the historical DC internal resistance measurement values ​​corresponding to each of the multiple state of charge values.

[0039] S202. Determine multiple smooth DC internal resistance curves of the battery based on historical state data, and determine the substrate internal resistance value corresponding to each smooth DC internal resistance curve.

[0040] The following is through Figure 3 This section describes the specific process for determining the substrate internal resistance value corresponding to each smooth DC internal resistance curve.

[0041] Figure 3 This is a flowchart for determining the substrate internal resistance value corresponding to each smooth DC internal resistance curve, provided in an embodiment of this application.

[0042] S301. Based on the historical state dataset, determine multiple initial DC internal resistance sequences for the battery.

[0043] Specifically, to uniformly characterize the substrate internal resistance values ​​obtained under multiple temperature and health status values... The sequence requires establishing a model that accurately describes its strong exponential dependence on temperature while also reflecting battery aging. For each temperature value and its corresponding state of health, there is a corresponding substrate internal resistance value. Internal resistance of each substrate It is an initial DC internal resistance sequence composed of the historical DC internal resistance measurements corresponding to multiple states of charge. It was determined.

[0044] S302. For each initial DC internal resistance sequence, perform smoothing filtering on the initial DC internal resistance sequence to obtain the smoothed DC internal resistance curve corresponding to the initial DC internal resistance sequence, and determine the bottom center region of the smoothed DC internal resistance curve. Within the bottom center region, perform statistical outlier removal processing on the historical DC internal resistance measurement values ​​in the initial DC internal resistance sequence to obtain a purified data subset. Calculate the local statistical values ​​of the purified data subset, and extract the local minimum value of the smoothed DC internal resistance curve within the bottom center region.

[0045] Specifically, for each initial DC internal resistance sequence Preprocessing is performed to obtain the smoothed DC resistance curve corresponding to the initial DC resistance sequence. This application employs an adaptive parameter Savitzky-Golay (SG) filtering strategy, establishing a SOC partitioning mapping rule to divide the entire SOC range into at least two regions, such as a low SOC steep region, a medium flat region, and a high SOC steep region. Based on the data points... In the region in question, the filter window length w and polynomial order p are dynamically selected to match the region, so as to effectively suppress noise while preserving the inherent shape characteristics of the DCR curve to the greatest extent.

[0046] Next, to robustly locate the bottom center index of the DCR curve. This paper proposes a bidirectional asymmetric window centroid method. Starting from the initial guess point (such as the minimum point of the smoothed curve), different widths are used in the low SOC and high SOC directions respectively. and Using a sliding window to calculate the local mean sequence and By analyzing the characteristics of the mean sequences on both sides, a stable bottom center region that can resist the effects of noise and curve asymmetry is identified.

[0047] Next, the initial DC internal resistance data for this region will be used. Statistical outlier removal is performed. An improved Z-score is calculated using robust statistics based on the median and median absolute deviation (MAD).

[0048] Where k is a proportionality constant, chosen such that k*MAD is approximately equal to the standard deviation when the data follows a Gaussian distribution. An optimal value for k is 1.4826 (i.e., 1 / 0.6745). [Remove] Outliers that are greater than a preset statistical threshold (e.g., 3.5).

[0049] Finally, local minima of the smoothed curves are extracted from the purified subset of data. and local statistics of the original data (such as minimum or median).

[0050] S303. For each smooth DC internal resistance curve, obtain the base internal resistance value corresponding to the smooth DC internal resistance curve based on the local statistical value and local minimum value corresponding to the smooth DC internal resistance curve.

[0051] Here, the corresponding base resistance value is obtained by weighted fusion of the corresponding local statistical values ​​and local minimum values. :

[0052] Among them, weighting factors It is an adjustable parameter in the range of [0, 1].

[0053] return Figure 2 S203. Based on multiple smooth DC internal resistance curves and the base internal resistance value of each smooth DC internal resistance curve, establish a functional relationship model to characterize the base internal resistance value of the battery and the battery temperature value and health status value.

[0054] Specifically, this application employs an empirical model inspired by the physical implications of the Arrhenius equation, but with parameters that are more engineering-oriented, to characterize the functional relationship model, in order to establish a model that can accurately describe its strong exponential dependence on temperature while also reflecting battery aging. Specifically, it is expressed as follows:

[0055] Where T is the Kelvin absolute temperature of the battery, and A(SOH) is the reference internal resistance factor, which mainly reflects the influence of SOH on the basic value of internal resistance. It is the equivalent temperature characteristic coefficient, with the unit Kelvin (K). It comprehensively characterizes the degree of drastic change in internal resistance with temperature (i.e., temperature sensitivity), and this sensitivity changes as the battery ages (SOH decreases).

[0056] As an example, to simplify the model, the reference internal resistance factor is used. and equivalent temperature characteristic coefficient Modeled as a linear function of SOH to capture the effects of aging:

[0057]

[0058] Among them, the model parameters are fitted using the nonlinear least squares method. , , and .

[0059] S204. Based on the functional relationship model, determine the target substrate internal resistance value of the battery at the target temperature value and the target health state value, and the shape function used to describe the change of DC internal resistance value with the state of charge, and obtain the prediction model based on the functional relationship model and the shape function.

[0060] Here, the hierarchical dynamic hybrid modeling strategy used in this application refers to the morphological function, which is a mathematical function used to describe the change of DC internal resistance value with the state of charge. The morphological function includes a first morphological function used to characterize the left ascending wing and a second morphological function used to characterize the right ascending wing. The functional relationship model, the first morphological function and the second morphological function are coupled to obtain the prediction model.

[0061] Specifically, the first morphological function and the second morphological function are determined in the following manner: Based on the functional relationship model, the boundaries of the first and second plateau regions of the target smooth DC internal resistance curve are determined; a first morphological function is constructed based on the first plateau region boundary and the target substrate internal resistance value, and a second morphological function is constructed based on the second plateau region boundary and the target substrate internal resistance value.

[0062] Among them, for the two rising wings on the left and right, i.e. (0, ]and[ Modeling is performed on the DCR increment within the interval [1]. Using a piecewise Gaussian decay function or a similar function, the first morphological function of the left ascending wing is expressed by the following formula:

[0063] in, This is the predicted DC internal resistance value for the left ascender wing. This is the substrate internal resistance value. For the left wing's ascent altitude parameters, For the left wing attenuation width parameter, This marks the boundary of the first platform area. This represents the state of charge (SOC).

[0064] The second morphological function is expressed by the following formula:

[0065] in, This is the predicted DC internal resistance value for the right ascender wing. For the right wing's ascent altitude parameters, For the attenuation width parameter of the right wing, This marks the boundary of the second platform area.

[0066] As an example, this application employs a multi-stage step-by-step optimization strategy from coarse to fine to solve for the model parameters. , , , ...), to avoid getting trapped in local optima. For example: in Phase 1, strong constraints are applied in the platform region. ≥ Phase 2 involves optimizing the wing parameters separately; Phase 3 involves global fine-tuning of all parameters and adding derivative continuity constraints at the boundaries.

[0067] It should be noted that for the preset high SOC region (e.g., SOC > 95%), a dynamic autoregressive clamping strategy is applied to enhance the engineering robustness of the model. When the SOC enters this region, the predicted DC internal resistance value is calculated using the following formula:

[0068] in, The predicted DC internal resistance value corresponding to the i-th state of charge value that is higher than the preset threshold. For smoothing coefficients, The predicted DC internal resistance value corresponding to the (i-1)th state of charge value that is higher than the preset threshold. Let be the target stability constant.

[0069] Specifically, the boundaries of the first platform area and the second platform area are determined in the following manner: Starting from the base resistance value of the smooth DC internal resistance curve, the local first-order differential gradient value of each data point on the smooth DC internal resistance curve is determined in the first direction, and the relative increase of the DC internal resistance value of the data point relative to the base internal resistance value is determined. The state of charge value corresponding to the data point that first simultaneously satisfies that the local first-order differential gradient value continuously exceeds the preset gradient threshold and the relative increase exceeds the preset amplitude threshold for the first time is determined as the boundary of the first platform region.

[0070] Starting from the base resistance value of the smooth DC internal resistance curve, the local first-order differential gradient value of each data point on the smooth DC internal resistance curve is determined in the second direction, and the relative increase of the DC internal resistance value corresponding to the data point relative to the base internal resistance value is determined. The state of charge value corresponding to the data point that first simultaneously satisfies that the local first-order differential gradient value continuously exceeds the preset gradient threshold and the relative increase exceeds the preset amplitude threshold for the first time is determined as the boundary of the second plateau region.

[0071] Specifically, in order to accurately identify the plateau region boundary of the DCR curve and Starting with the base resistance value, the system proceeds sequentially towards both high and low SOC directions, employing a hybrid gradient and relative threshold driving strategy. A point is considered a platform boundary only if it simultaneously satisfies the following conditions: 1) Its local first-order difference gradient first sustains exceeding the gradient threshold. .

[0072] 2) Its value is relative to relative increase / First time exceeding the amplitude threshold ,Right now:

[0073] As an example, after model deployment, using newly acquired online DCR measurements, recursive least squares (RLS) with a forgetting factor λ or a similar online identification algorithm, the parameters in the model most relevant to aging and temperature characteristics (such as those described) are identified. and The coefficients of the evolution law are fine-tuned online.

[0074] Finally, the model is deployed to the vehicle's BMS to run in real time. Real-time operating conditions, namely the current temperature value, SOC, and the current vehicle SOC, are input into the prediction model to predict the current DCR resistance value.

[0075] As an example, this application can also compute the confidence score Conf of the model output in parallel, which is a weighted fusion of multiple dimensions:

[0076] in, Based on recent signal-to-noise ratio measurements; Residuals based on RLS online correction; Based on the location of the current operating point in the offline calibration data distribution; Based on whether the current SOH exceeds the calibration range.

[0077] Example: The following section uses the DCR modeling of a 100Ah lithium iron phosphate (LFP) battery under a 10-second discharge at -10°C as an example to elaborate on the technical solution of this application.

[0078] Step S100: Data acquisition and adaptive preprocessing.

[0079] An initial DC internal resistance sequence covering the 0-100% SOC range was obtained using a high-precision battery testing system. A SOC partition lookup table, Zone_Map, is established. In this embodiment, Zone_Map is defined as: {[0,0.15, 3, 1], [0.15, 0.90, 11, 2], [0.90, 1.0, 5, 1]}, where each row has the format [SOC_start, SOC_end, window_length, poly_order]. For each input data point, the SG filtering function is called according to the interval where its SOC_i is located, and the corresponding parameters are used for filtering to obtain high-fidelity smooth DC internal resistance curve data. .

[0080] Step S200: Robust bottom center positioning.

[0081] Take smooth curve The index of the global minimum point is used as the initial guess point. .from Starting from the low SOC direction, adopt a width A sliding window with a width of 5 is used towards the high SOC direction. Using a sliding window with a value of 9, calculate the local mean sequence. and Find the index of each minimum point. and The final robust bottom center area The weighted average is obtained as follows:

[0082] Step S300: Substrate internal resistance identification.

[0083] In Within a centered window (e.g., containing 9 data points), the raw data Outlier removal is performed. Calculate the number of outliers within this window. The median and median absolute deviation (MAD). For each point Calculate its improved Z-score:

[0084] Among them, if | If the value is greater than 3.5, mark the point as an outlier and remove it. On the cleaned subset of data, extract the local minimum of the smoothed curve. Local minimum of the original data With a weight α=0.7, the substrate internal resistance under the current operating condition is calculated. .

[0085] Step S400: Cross-condition physical modeling.

[0086] Repeat steps S100-S300 to obtain data at multiple temperature points. (e.g., -30°C, -20°C, 0°C, 25°C) and different aging conditions (e.g., 100%, 90%, 80%) value.

[0087] Substitute all data points into the cross-condition model based on the empirical temperature-dependent model described in step S400. Use the Levenberg-Marquardt nonlinear least squares algorithm to fit all four model parameters at once. , , and .

[0088] Steps S500-S800: Morphological feature extraction, modeling and optimization.

[0089] On the smooth curve at -10°C, from Scan to both sides and set gradient thresholds and amplitude threshold (e.g., 5%), identify platform boundaries and A piecewise Gaussian decay function is used to model the two rising segments. The model parameters are solved using a multi-stage step-by-step optimization strategy. When SOC > 95%, dynamic autoregressive clamping is enabled, with β = 0.9 and C = 0.9. The value at SOC=95%.

[0090] Steps S900-S1000: Online prediction and update, and assessment of confidence level.

[0091] After the model is deployed to BMS, the parameters are processed using the RLS algorithm with a forgetting factor λ=0.995. and It performs online updates. Simultaneously, it computes the model confidence score (Conf) in parallel, providing decision support for upper-layer applications. Figure 4 The diagram illustrates the difference between the proposed method (solid line) and traditional polynomial fitting (dashed line), showing that the DCR prediction value can be obtained for each SOC. The proposed method effectively suppresses noise, avoids oscillations, and achieves stable clamping in the high SOC region.

[0092] Compared with existing methods, this application has the following significant advantages: 1. Ultra-high precision and robustness: Through adaptive preprocessing, robust benchmark identification and fine morphological modeling, it can effectively resist strong noise and outlier interference, and accurately characterize DCR under various harsh working conditions.

[0093] 2. Deep physical interpretation: the core of the model Based on an empirical temperature-dependent model inspired by physical laws, its key parameters are strongly correlated with battery aging and temperature sensitivity. The model is transparent, reliable, easy to verify, and easy to apply in engineering.

[0094] 3. Full life cycle adaptability: It has online self-correction capability and can track the performance degradation of the battery caused by aging in real time, achieving high-precision evaluation throughout the entire life cycle.

[0095] 4. Decision-level reliability: It not only provides high-precision predicted values, but also quantifies the confidence level of these values, providing a key basis for BMS to achieve more intelligent and safer hierarchical power control and fault diagnosis strategies.

[0096] Based on the same inventive concept, this application also provides a battery DC internal resistance prediction device corresponding to the battery DC internal resistance prediction method. Since the principle of the device in this application is similar to the battery DC internal resistance prediction method described above in this application, the implementation of the device can refer to the implementation of the method, and the repeated parts will not be described again.

[0097] Please see Figure 5 , Figure 5 This is a schematic diagram of a battery DC internal resistance prediction device provided in an embodiment of this application. The battery DC internal resistance prediction device 500 includes: The acquisition module 501 is used to acquire the battery's state data, which includes the battery's state of charge, temperature, and health status at a target time.

[0098] The prediction module 502 is used to input the state data into the prediction model of the battery to obtain the predicted value of the DC internal resistance of the battery. The prediction model constructs the correspondence between the DC internal resistance value of the battery and the state of charge value, temperature value and health value of the battery based on the historical state dataset of the battery.

[0099] Please see Figure 6 , Figure 6 This is a schematic diagram of the structure of an electronic device provided in an embodiment of this application. Figure 6 As shown, the electronic device 600 includes a processor 610, a memory 620, and a bus 630.

[0100] The memory 620 stores machine-readable instructions that can be executed by the processor 610. When the electronic device 600 is running, the processor 610 and the memory 620 communicate via the bus 630. When the machine-readable instructions are executed by the processor 610, the steps of the method in the above method embodiment can be performed. For specific implementation methods, please refer to the method embodiment, which will not be repeated here.

[0101] This application also provides a computer-readable storage medium storing a computer program. When the computer program is run by a processor, it can execute the steps of the method in the above method embodiments. For specific implementation details, please refer to the method embodiments, which will not be repeated here.

[0102] Those skilled in the art will understand that, for the sake of convenience and brevity, the specific working processes of the systems, devices, and units described above can be referred to the corresponding processes in the foregoing method embodiments, and will not be repeated here.

[0103] In the several embodiments provided in this application, it should be understood that the disclosed systems, apparatuses, and methods can be implemented in other ways. The apparatus embodiments described above are merely illustrative. For example, the division of units is only a logical functional division, and in actual implementation, there may be other division methods. Furthermore, multiple units or components may be combined or integrated into another system, or some features may be ignored or not executed. Additionally, the shown or discussed mutual couplings, direct couplings, or communication connections may be through some communication interfaces; indirect couplings or communication connections between devices or units may be electrical, mechanical, or other forms.

[0104] The units described as separate components may or may not be physically separate. The components shown as units may or may not be physical units; that is, they may be located in one place or distributed across multiple network units. Some or all of the units can be selected to achieve the purpose of this embodiment according to actual needs.

[0105] In addition, the functional units in the various embodiments of this application can be integrated into one processing unit, or each unit can exist physically separately, or two or more units can be integrated into one unit.

[0106] If the aforementioned functions are implemented as software functional units and sold or used as independent products, they can be stored in a processor-executable, non-volatile, computer-readable storage medium. Based on this understanding, the technical solution of this application, in essence, or the part that contributes to the prior art, or a portion of the technical solution, can be embodied in the form of a software product. This computer software product is stored in a storage medium and includes several instructions to cause a computer device (which may be a personal computer, server, or network device, etc.) to execute all or part of the steps of the methods described in the various embodiments of this application. The aforementioned storage medium includes various media capable of storing program code, such as USB flash drives, portable hard drives, read-only memory (ROM), random access memory (RAM), magnetic disks, or optical disks.

[0107] Finally, it should be noted that the above-described embodiments are merely specific implementations of this application, used to illustrate the technical solutions of this application, and not to limit them. The scope of protection of this application is not limited thereto. Although this application has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that any person skilled in the art can still modify or easily conceive of changes to the technical solutions described in the foregoing embodiments, or make equivalent substitutions for some of the technical features, within the scope of the technology disclosed in this application. Such modifications, changes, or substitutions do not cause the essence of the corresponding technical solutions to deviate from the spirit and scope of the technical solutions of the embodiments of this application, and should all be covered within the scope of protection of this application. Therefore, the scope of protection of this application should be determined by the scope of the claims.

Claims

1. A method for predicting the DC internal resistance of a battery, characterized in that, The method includes: Acquire battery status data, which includes the battery's state of charge, temperature, and health status at a target time. The state data is input into the battery's prediction model to obtain the predicted DC internal resistance value of the battery. The prediction model, based on the battery's historical state dataset, constructs a correspondence between the battery's DC internal resistance value and its state of charge, temperature, and health status values. The prediction model is constructed in the following manner: Obtain the historical state dataset of the battery, which includes the historical DC internal resistance measurement values ​​corresponding to each of the multiple state of charge values ​​under multiple temperature values ​​and the health state value corresponding to each temperature value; Based on the historical state dataset, multiple smooth DC internal resistance curves of the battery are determined, and the substrate internal resistance value corresponding to each smooth DC internal resistance curve is determined. Based on the plurality of smoothed DC internal resistance curves and the base internal resistance value of each smoothed DC internal resistance curve, a functional relationship model is established to characterize the base internal resistance value of the battery and the temperature value and health status value of the battery. Based on the aforementioned functional relationship model, the target substrate internal resistance value of the battery at the target temperature and target health state values ​​is determined, along with a shape function describing the change of the battery's DC internal resistance with respect to its state of charge. The prediction model is then derived based on the functional relationship model and the shape function. The substrate internal resistance value corresponding to each smooth DC internal resistance curve is determined in the following way: Based on the historical state dataset, multiple initial DC internal resistance sequences of the battery are determined; For each initial DC internal resistance sequence, a smoothing filter is applied to the initial DC internal resistance sequence to obtain the smoothed DC internal resistance curve corresponding to the initial DC internal resistance sequence. The bottom center region of the smoothed DC internal resistance curve is determined. Within the bottom center region, statistical outlier removal is performed on the historical DC internal resistance measurements in the initial DC internal resistance sequence to obtain a purified data subset. The local statistical values ​​of the purified data subset are calculated, and the local minimum value of the smoothed DC internal resistance curve within the bottom center region is extracted. For each smooth DC internal resistance curve, the base resistance value corresponding to the smooth DC internal resistance curve is obtained based on the local statistical value and local minimum value corresponding to the smooth DC internal resistance curve.

2. The method according to claim 1, characterized in that, The morphological functions include a first morphological function for characterizing the left ascending wing and a second morphological function for characterizing the right ascending wing. The first morphological function and the second morphological function are determined in the following manner: Based on the aforementioned functional relationship model, determine the first and second plateau region boundaries of the target smooth DC internal resistance curve; The first morphological function is constructed based on the boundary of the first platform region and the internal resistance value of the target substrate, and the second morphological function is constructed based on the boundary of the second platform region and the internal resistance value of the target substrate.

3. The method according to claim 2, characterized in that, The boundaries of the first platform area and the second platform area are determined in the following manner: Starting from the base resistance value of the smooth DC internal resistance curve, the local first-order differential gradient value of each data point on the smooth DC internal resistance curve is determined in the first direction, and the relative increase of the DC internal resistance value corresponding to the data point relative to the base internal resistance value is determined. The state of charge value corresponding to the data point that first simultaneously satisfies that the local first-order differential gradient value continuously exceeds the preset gradient threshold and the relative increase exceeds the preset amplitude threshold for the first time is determined as the boundary of the first platform region. Starting from the base resistance value of the smooth DC internal resistance curve, the local first-order differential gradient value of each data point on the smooth DC internal resistance curve is determined in the second direction, and the relative increase of the DC internal resistance value corresponding to the data point relative to the base internal resistance value is determined. The state of charge value corresponding to the data point that first simultaneously satisfies that the local first-order differential gradient value continuously exceeds a preset gradient threshold and the relative increase first exceeds a preset amplitude threshold is determined as the boundary of the second platform region.

4. The method according to claim 2, characterized in that, The first morphological function is expressed by the following formula: in, This is the predicted DC internal resistance value of the left ascending wing section. This is the substrate internal resistance value. For the left wing's ascent altitude parameters, For the left wing attenuation width parameter, This is the boundary of the first platform area. This is the state of charge value. The second morphological function is expressed by the following formula: in, This is the predicted DC internal resistance value of the right ascending wing section. For the right wing's ascent altitude parameters, For the attenuation width parameter of the right wing, This is the boundary of the second platform area.

5. The method according to claim 4, characterized in that, When the state of charge value is higher than the preset threshold, the predicted DC internal resistance value is calculated using the following formula: in, The predicted DC internal resistance value corresponding to the i-th state of charge value that is higher than the preset threshold. For smoothing coefficients, The predicted DC internal resistance value corresponding to the (i-1)th state of charge value that is higher than the preset threshold. Let be the target stability constant.

6. A battery DC internal resistance prediction device, characterized in that, The device includes: The acquisition module is used to acquire the battery's status data, which includes the battery's state of charge, temperature, and health status at a target time. The prediction module is used to input the state data into the prediction model of the battery to obtain the predicted value of the DC internal resistance of the battery. The prediction model constructs the correspondence between the DC internal resistance value of the battery and the state of charge value, temperature value and health value of the battery based on the historical state dataset of the battery. The prediction module is further configured to acquire a historical state dataset of the battery, which includes historical DC internal resistance measurements corresponding to multiple states of charge (SOC) values ​​at multiple temperature values ​​and corresponding health values ​​at each temperature value; determine multiple smoothed DC internal resistance curves of the battery based on the historical state dataset, and determine the base resistance value corresponding to each smoothed DC internal resistance curve; establish a functional relationship model characterizing the base resistance value of the battery and the temperature and health values ​​of the battery based on the multiple smoothed DC internal resistance curves and the base resistance value of each smoothed DC internal resistance curve; determine the target base resistance value of the battery at the target temperature and target health values, and a shape function describing the change of the DC internal resistance value of the battery with the SOC, based on the functional relationship model and the shape function; and obtain the prediction model based on the functional relationship model and the shape function. The prediction module is further configured to determine multiple initial DC internal resistance sequences of the battery based on the historical state dataset; for each initial DC internal resistance sequence, perform smoothing filtering on the initial DC internal resistance sequence to obtain a smoothed DC internal resistance curve corresponding to the initial DC internal resistance sequence, and determine the bottom central region of the smoothed DC internal resistance curve; within the bottom central region, perform statistical outlier removal processing on the historical DC internal resistance measurements in the initial DC internal resistance sequence to obtain a purified data subset; calculate the local statistical values ​​of the purified data subset; and extract the local minimum value of the smoothed DC internal resistance curve within the bottom central region; for each smoothed DC internal resistance curve, obtain the base resistance value corresponding to the smoothed DC internal resistance curve based on the local statistical value and the local minimum value corresponding to the smoothed DC internal resistance curve.

7. An electronic device, characterized in that, include: The device includes a processor, a memory, and a bus, wherein the memory stores machine-readable instructions executable by the processor, and when the electronic device is in operation, the processor communicates with the memory via the bus, and the processor executes the machine-readable instructions to perform the steps of the method as described in any one of claims 1 to 5.

8. A computer-readable storage medium, characterized in that, The computer-readable storage medium stores a computer program that, when executed by a processor, performs the steps of the method as described in any one of claims 1 to 5.

Images