Method, device and medium for estimating state of charge of lithium ion battery

Abstract

This application provides a method, apparatus, device, and medium for estimating the state of charge (SOC) of a lithium-ion battery. The method includes: upon receiving a SOC estimation request for a battery to be estimated, acquiring the battery terminal voltage measurement value and a platform interval lookup table collected at the current moment, and determining a priori SOC estimate; further determining the platform interval attribute result corresponding to the priori SOC estimate; adjusting the observation noise covariance matrix based on the platform interval attribute result, the real-time slope of the open-circuit voltage-SOC relationship curve at the priori SOC estimate, and the battery operating current change rate; updating the Kalman gain based on the adjusted observation noise covariance matrix, and using the updated Kalman gain to correct the priori SOC estimate to obtain the target SOC estimate at the current moment. Through platform interval identification and a multi-parameter adaptive adjustment mechanism, the accuracy and stability of lithium-ion battery SOC estimation are improved.

Description

Technical Field

[0001] This invention relates to the field of battery technology, and in particular to a method, apparatus, device, and medium for estimating the state of charge of a lithium-ion battery. Background Technology

[0002] State of charge (SOC), a key parameter characterizing the remaining capacity of a lithium-ion battery, is crucial for the accurate estimation of which is a core function of battery management systems. Among various lithium-ion batteries, lithium iron phosphate (LFP) batteries are widely used in electric vehicles and energy storage systems due to their high safety, long cycle life, and cost advantages. However, the inherent open-circuit voltage-SOC relationship curve of this battery system exhibits a voltage plateau characteristic for most of its operating range. This strong nonlinearity presents a unique challenge to voltage observation-based SOC estimation methods.

[0003] Currently, the Extended Kalman Filter (EKF) algorithm has become one of the mainstream methods for estimating the state of charge (SOC) of power batteries due to its ability to effectively handle the nonlinear characteristics of systems. This algorithm establishes a state-space model of the battery and uses the SOC as a system state variable for optimal estimation. In practical implementation, traditional methods typically use a fixed observation noise covariance matrix to characterize the uncertainty in the terminal voltage measurement process, and achieve real-time SOC estimation through recursive calculation.

[0004] Existing extended Kalman filtering methods based on fixed observation noise covariance face significant limitations in lithium iron phosphate battery applications: First, in the plateau region of the open-circuit voltage-state-of-charge curve, the open-circuit voltage is extremely insensitive to changes in state of charge, and even small voltage measurement errors can lead to significant deviations in state of charge estimation, resulting in an amplification effect of observation noise. Second, the fixed observation noise covariance cannot adapt to the dynamic characteristics of different operating regions. It is difficult to effectively suppress estimation fluctuations in the plateau region, and it cannot fully utilize the correction effect of voltage observation information in the non-plateau region. At the same time, it lacks the ability to adaptively adjust noise under dynamic operating conditions such as drastic current changes, making it difficult to guarantee the accuracy of state of charge estimation under all operating conditions. Summary of the Invention

[0005] This application provides a method, apparatus, device, and medium for estimating the state of charge (SOC) of lithium-ion batteries. By identifying the plateau region and using a multi-parameter adaptive adjustment mechanism, it effectively solves the problem of observation noise amplification in the plateau region of lithium iron phosphate batteries using traditional EKF (Electronic Keying Method), and achieves a balance between SOC estimation accuracy and stability under all operating conditions.

[0006] In a first aspect, embodiments of this application provide a method for estimating the state of charge of a lithium-ion battery. The method is based on an extended Kalman filter model and includes:

[0007] When a request for state of charge estimation of the battery to be estimated is received, the battery terminal voltage measurement value collected at the current time and the pre-established plateau interval lookup table are obtained, and the prior state of charge estimate is determined based on the extended Kalman filter model; wherein, the plateau interval lookup table is determined by offline identification and calibration of at least one plateau interval in the open circuit voltage-state of charge relationship curve of the battery to be estimated.

[0008] Based on the prior state of charge estimate, the plateau interval lookup table is queried to determine the plateau interval attribute result of the open circuit voltage-state of charge relationship curve at the prior state of charge estimate;

[0009] Based on the platform interval attribute results, the real-time slope of the open-circuit voltage-state of charge relationship curve at the prior state of charge estimate, and the rate of change of battery operating current, the observation noise covariance matrix of the extended Kalman filter model is adjusted.

[0010] The Kalman gain is updated based on the adjusted observation noise covariance matrix, and the prior state of charge estimate is corrected using the updated Kalman gain and the terminal voltage measurement to obtain the target state of charge estimate at the current time.

[0011] Secondly, embodiments of this application also provide a device for estimating the state of charge of a lithium-ion battery. The device is based on an extended Kalman filter model and includes:

[0012] The data acquisition module is used to acquire the battery terminal voltage measurement value collected at the current moment and the pre-established plateau interval lookup table when a request for estimating the state of charge of the battery to be estimated is received, and to determine the prior state of charge estimate based on the extended Kalman filter model; wherein, the plateau interval lookup table is determined by offline identification and calibration of at least one plateau interval in the open circuit voltage-state of charge relationship curve of the battery to be estimated.

[0013] The platform attribute determination module is used to query the platform interval lookup table based on the prior state of charge estimate and determine the platform interval attribute result of the open circuit voltage-state of charge relationship curve at the prior state of charge estimate.

[0014] The noise covariance adjustment module is used to adjust the observation noise covariance matrix of the extended Kalman filter model based on the platform interval attribute results, the real-time slope of the open-circuit voltage-state of charge relationship curve at the prior state of charge estimate, and the battery operating current change rate.

[0015] The state of charge estimation module is used to update the Kalman gain based on the adjusted observation noise covariance matrix, and to correct the prior state of charge estimate using the updated Kalman gain and the terminal voltage measurement value, so as to obtain the target state of charge estimate at the current time.

[0016] Thirdly, embodiments of this application also provide an electronic device, which includes:

[0017] One or more processors;

[0018] Storage device for storing one or more programs.

[0019] When one or more programs are executed by one or more processors, the one or more processors implement a method for estimating the state of charge of a lithium-ion battery as described in any of the embodiments of this application.

[0020] Fourthly, embodiments of this application also provide a storage medium containing computer-executable instructions, which, when executed by a computer processor, are used to perform a method for estimating the state of charge of a lithium-ion battery as described in any of the embodiments of this application.

[0021] This application provides a method, apparatus, device, and medium for estimating the state of charge (SOC) of a lithium-ion battery. The method includes: upon receiving a request to estimate the SOC of a battery to be estimated, acquiring the battery terminal voltage measurement value collected at the current moment and a pre-established plateau interval lookup table, and determining a priori SOC estimate based on an extended Kalman filter model. The plateau interval lookup table is determined by offline identification and calibration of at least one plateau interval in the open-circuit voltage-SOC relationship curve of the battery to be estimated. Then, based on the priori SOC estimate, the plateau interval lookup table is queried to determine the plateau interval attribute result of the open-circuit voltage-SOC relationship curve at the priori SOC estimate. Further, based on the plateau interval attribute result, the real-time slope of the open-circuit voltage-SOC relationship curve at the priori SOC estimate, and the battery operating current change rate, the observation noise covariance matrix of the extended Kalman filter model is adjusted. Then, the Kalman gain is updated based on the adjusted observation noise covariance matrix, and the priori SOC estimate is corrected using the updated Kalman gain and the terminal voltage measurement value to obtain the target SOC estimate at the current moment. The technical solution of this application identifies the operating point characteristics by constructing a lookup table for the platform interval, and builds a multi-dimensional adaptive mechanism by combining the real-time slope and the rate of change of current. In the platform interval, the excessive influence of the voltage observation residual on the state correction is suppressed by increasing the observation noise covariance, so as to avoid drastic fluctuations in the SOC estimate. In the non-platform area and under dynamic operating conditions, the noise configuration is optimized by adjusting the adjustment factor to give full play to the correction role of the voltage observation value, and finally achieves a synergistic improvement in the accuracy and stability of SOC estimation under all operating conditions. Attached Figure Description

[0022] To more clearly illustrate the technical solutions of the exemplary embodiments of this application, the accompanying drawings used in describing the embodiments are briefly introduced below. Obviously, the accompanying drawings described are only a portion of the embodiments to be described in this application, and not all of them. For those skilled in the art, other drawings can be obtained from these drawings without any creative effort.

[0023] Figure 1 A flowchart illustrating a method for estimating the state of charge of a lithium-ion battery provided in an embodiment of this application;

[0024] Figure 2 A flowchart illustrating another method for estimating the state of charge of a lithium-ion battery provided in an embodiment of this application;

[0025] Figure 3 As involved in the embodiments of this application Figure 3 This is the open-circuit voltage-state-of-charge curve of a certain lithium-ion battery.

[0026] Figure 4A flowchart illustrating another method for estimating the state of charge of a lithium-ion battery provided in an embodiment of this application;

[0027] Figure 5 This is a schematic diagram comparing the state-of-charge estimation results of the extended Kalman filter method with fixed observation noise covariance and adaptive observation noise covariance involved in the embodiments of this application.

[0028] Figure 6 This is a schematic diagram comparing the state-of-charge estimation errors of the extended Kalman filter method with fixed observation noise covariance and adaptive observation noise covariance involved in the embodiments of this application.

[0029] Figure 7 A flowchart illustrating another method for estimating the state of charge of a lithium-ion battery provided in an embodiment of this application;

[0030] Figure 8 This is a flowchart illustrating another method for estimating the state of charge of a lithium-ion battery provided in an embodiment of this application. Detailed Implementation

[0031] The present application will now be described in further detail with reference to the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are merely illustrative of the application and not intended to limit it. Furthermore, it should be noted that, for ease of description, the accompanying drawings show only the parts relevant to the present application, not the entire structure.

[0032] Before introducing the technical solution provided in this application, the application scenarios of the solution can be explained first. This embodiment is applicable to various scenarios that require accurate estimation of the state of charge (SOC) of lithium-ion batteries. Currently, although the extended Kalman filter algorithm is widely used in battery SOC estimation, the traditional method with a fixed observation noise covariance design has obvious limitations. In practical applications, due to the significant plateau characteristics of the open-circuit voltage-SOC curve of batteries such as lithium iron phosphate, and the frequent drastic changes in operating current, fixed noise parameters are difficult to adapt to changes in observation characteristics under different operating conditions, which can easily lead to large deviations in the SOC estimation value. Therefore, there is an urgent need for a method that can dynamically adjust the observation noise covariance according to the actual operating state of the battery to improve the accuracy and stability of SOC estimation under all operating conditions. This embodiment constructs a plateau interval lookup table and combines multi-dimensional parameters such as real-time slope and current change rate to achieve adaptive adjustment of the observation noise covariance, thereby ensuring accurate and reliable SOC estimation results and effectively improving the SOC estimation performance of the battery management system.

[0033] Example 1

[0034] Figure 1This is a flowchart illustrating a method for estimating the state of charge (SOC) of a lithium-ion battery according to an embodiment of this application. This embodiment is applicable to any situation requiring accurate estimation of the SOC of a lithium-ion battery. The method can be executed by a lithium-ion battery SOC estimation device, which can be implemented in the form of software and / or hardware. The hardware can be a controller, such as a mobile terminal, a PC, or a server.

[0035] like Figure 1 As shown, the method includes:

[0036] S110. When a request for state of charge estimation of the battery to be estimated is received, the battery terminal voltage measurement value collected at the current time and the pre-established platform interval lookup table are obtained, and the prior state of charge estimate is determined based on the extended Kalman filter model.

[0037] Here, the battery to be estimated refers to a specific lithium-ion battery cell, battery module, or battery pack that needs to have its state of charge (SOC) estimated using the extended Kalman filter method. A SOC estimation request is a command signal generated during the operation of the battery management system, triggered periodically by the system or by specific external events such as significant changes in current / voltage or user query commands, requesting the immediate initiation or execution of a SOC estimation process.

[0038] Among them, the battery terminal voltage measurement value refers to the actual working voltage reading collected directly between the positive and negative terminals of the battery to be estimated at the current moment through the voltage sensing device without being connected to an external charging or discharging device. This value is an externally measurable electrical performance under the combined effect of various factors such as the internal chemical reaction state of the battery, load current, and internal resistance voltage drop, and serves as a key observation in the extended Kalman filter algorithm for comparing and correcting model predictions.

[0039] The platform interval lookup table is a data structure or database pre-established through offline calibration to store the characteristics of the flat voltage change region in the open-circuit voltage-state-charge curve of a specific battery to be estimated. The open-circuit voltage-state-charge curve is a characteristic curve determined through offline experiments, characterizing the relationship between the open-circuit voltage and state of charge of the battery under static equilibrium conditions; this curve reveals the inherent electrochemical characteristics of the battery. The platform interval lookup table is determined by offline identification and calibration of at least one platform interval in the open-circuit voltage-state-charge curve of the battery to be estimated. This can be understood as follows: when pre-constructing the platform interval lookup table, experimental tests can be conducted on the battery to be estimated to obtain its complete open-circuit voltage and state of charge static relationship curve. By analyzing the curve, the smooth sections (i.e., platform intervals) where the open-circuit voltage value changes very little or even remains basically unchanged with the state of charge can be identified. Then, the key attributes such as the boundary values, center points, and slope characteristics of these intervals are digitally extracted and stored, and finally organized into a structured data table that can be quickly retrieved by the online system. This provides a pre-prepared standard reference for dynamically judging the location characteristics of the working point during subsequent online estimation.

[0040] The Extended Kalman Filter (EKF) model is an algorithmic framework for state estimation problems in nonlinear systems. In this battery state-of-charge (SOC) estimation scenario, the model treats the battery's SOC as an internal state variable, uses a mathematical model of the battery to predict the state, and uses actual measured values ​​of the terminal voltage to correct the predicted values. Through iterative "prediction-correction" steps, the optimal estimate of the battery SOC, a quantity that cannot be directly measured, is finally achieved.

[0041] The prior state of charge (PSC) estimate is a preliminary estimate of the state calculated in the prediction step of the extended Kalman filter algorithm based on the system's previous state estimate and the battery's mathematical model. This estimate is made before the latest terminal voltage measurement is incorporated into the current state. The PSC estimate represents the result of forward extrapolation based solely on the system's own dynamic characteristics and serves as the basis for subsequent corrections using actual observations.

[0042] In this embodiment, when the battery management system detects the command signal to start the state of charge estimation process, it can read the battery terminal voltage measurement value of the battery to be estimated at the moment the command is triggered in real time through the voltage sensor. It can call the plateau interval lookup table established by offline identification and calibration of the plateau interval in the open circuit voltage-state of charge relationship curve of the battery to be estimated from the storage unit. At the same time, it drives the extended Kalman filter model to complete its prediction step, that is, to deduce the prior state of charge estimate value that has not yet incorporated the latest observation value at the current moment based on the state optimal estimate of the previous moment and the system input, thereby providing the necessary input conditions for subsequent adaptive correction based on the plateau interval attribute results.

[0043] For example, in a running electric vehicle battery management system, when a timed request for state of charge estimation for a specific power battery pack is triggered, the control unit immediately performs the following operations: reads the actual operating voltage (e.g., 3.65V) across the battery terminals using a voltage sensor as the measured battery terminal voltage value; simultaneously retrieves a platform range lookup table obtained from experimental calibration of the battery in flash memory; then, the extended Kalman filter model substitutes the final determined state of charge value (e.g., 62.1%) from the previous cycle and the newly acquired load current value (e.g., -15A) into the state equation, and calculates the uncorrected state of charge estimate (e.g., 61.8%) at the current moment using ampere-hour integration. This estimate is the prior state of charge estimate.

[0044] Based on the above embodiments, optionally, the specific implementation method for determining the prior state of charge estimate based on the extended Kalman filter model may include: inputting the target state of charge estimate from the previous moment and the currently collected load current value into the extended Kalman filter model to predict the prior state of charge estimate for the current moment.

[0045] Among them, the target charge state estimate of the previous moment refers to the optimal state estimate obtained by step S140 in the previous complete iteration cycle of the extended Kalman filter algorithm, which has incorporated the actual observation information at that moment. As the most reliable representation of the system's historical state, it is directly used as the initial state input for the new round of prediction calculation at the current moment. It is the key data to ensure that the state estimate maintains continuity and recursion in the time series.

[0046] The load current value refers to the real-time current data directly collected from the main circuit of the battery being estimated by a current sensor at the current moment. The load current value accurately represents the current intensity of the battery at that instant when it is in the discharging (positive current value) or charging (negative current value) operating mode.

[0047] In this embodiment, in the prediction step of the extended Kalman filter algorithm, the optimal state result obtained after the observation correction of the previous moment (i.e., the target state of charge estimate of the previous moment) and the load current value reflecting the battery charging and discharging intensity measured in real time are substituted into the state equation describing the dynamic characteristics of the battery for calculation. Through mathematical derivation (usually combined with the ampere-hour integration method), the uncorrected state of charge estimate before incorporating the new observation value at the current moment is estimated. This estimate is the prior state of charge estimate, which forms the basis for subsequent feedback correction using the terminal voltage measurement value.

[0048] S120. Based on the prior state of charge estimate, query the platform interval lookup table to determine the platform interval attribute result of the open circuit voltage-state of charge relationship curve at the prior state of charge estimate.

[0049] The plateau interval attribute result refers to the definitive conclusion regarding the positional characteristics of the estimated value on the open-circuit voltage-state-of-charge curve, obtained after querying the plateau interval lookup table based on the prior state-of-charge estimate. Specifically, the plateau interval attribute result includes either a plateau interval or a non-plateau interval.

[0050] In this embodiment, the prior state of charge estimate predicted by the extended Kalman filter model can be used as an input condition. It can be retrieved and matched in a pre-established plateau interval lookup table. By comparing the numerical range or performing table lookup operations, a qualitative judgment on the position of the estimated value on the open-circuit voltage-state of charge curve is finally output. That is, it clearly indicates whether the operating point falls within the calibrated plateau interval, thereby obtaining the plateau interval attribute result for subsequent adaptive filtering adjustment.

[0051] More specifically, the specific steps for determining the plateau interval attribute of the open-circuit voltage-state-of-charge curve at the prior estimate of the state of charge may include:

[0052] (1) Compare the prior state of charge estimates with the state of charge boundary values ​​of each platform interval recorded in the platform interval lookup table.

[0053] Among them, the state of charge boundary value refers to the two key state of charge values ​​determined for each platform interval to define the start and end range of the interval.

[0054] In this embodiment, after obtaining the prior state of charge estimate predicted by the extended Kalman filter model, it is used as a retrieval key value and compared one by one with the start and end state of charge boundary values ​​corresponding to each platform interval pre-stored in the platform interval lookup table. The logical judgment of the numerical range determines whether the prior state of charge estimate falls within the numerical range of any calibrated platform interval, thereby providing a direct basis for the generation of platform interval attribute results.

[0055] (2) Determine whether the prior state of charge estimate is within a plateau interval based on whether the prior state of charge estimate falls within the boundary of any plateau interval, and output the corresponding plateau interval attribute result.

[0056] In this embodiment, logical judgment can be used to determine whether the value of the current prior state of charge estimate is within the range defined by the state of charge boundary value of any platform interval recorded in the platform interval lookup table. If it is within the range, it is determined to be in a "platform interval", otherwise it is determined to be a "non-platform interval". Finally, this binary judgment conclusion is output as the platform interval attribute result.

[0057] For example, suppose the platform interval lookup table records the state of charge boundary value of a certain platform interval as 30% to 40%. When the prior state of charge estimate predicted by the extended Kalman filter model is 35%, the system determines that the value falls within the above boundary by numerical comparison, and then outputs "platform interval" as the platform interval attribute result; if the prior state of charge estimate is 45%, then the system outputs the judgment result of "non-platform interval" because it does not fall within any recorded boundary range.

[0058] S130. Based on the platform interval attribute results, the real-time slope of the open-circuit voltage-state of charge relationship curve at the prior state of charge estimate and the rate of change of battery operating current, adjust the observation noise covariance matrix of the extended Kalman filter model.

[0059] The real-time slope of the open-circuit voltage-state-of-charge (OCC) curve at the prior OCC estimate refers to the instantaneous rate of change of the OCC curve at a specific operating point corresponding to the prior OCC estimate, calculated based on the actual characteristics of the battery to be estimated, within the current estimation period. This real-time slope parameter characterizes the sensitivity of the open-circuit voltage to changes in OCC at this point and is a key indicator dynamically reflecting the nonlinear characteristics of the curve region where the battery is currently operating.

[0060] The battery operating current change rate refers to the rate of change of current calculated by monitoring the change of load current in the battery circuit under estimation over time within the current estimation period. This parameter quantitatively characterizes the dynamic fluctuation intensity of the battery's operating load, and its magnitude reflects the severity of the system's external operating conditions.

[0061] Among them, the observation noise covariance matrix is ​​a key parameter matrix used in the extended Kalman filter model to quantify the uncertainty in the observation equation. It characterizes the statistical characteristics of the deviation between the measured terminal voltage value and the true value caused by factors such as sensor error, model mismatch and environmental interference. Its magnitude directly affects the update weight of the Kalman gain, and thus determines the degree of confidence allocation between the prior charge state estimate and the measured terminal voltage value in the state estimation process.

[0062] Specifically, based on the platform interval attribute determination results, the operating state is divided into two basic modes: "platform interval" and "non-platform interval." Different coefficient weight calculation rules are preset for different modes. Then, combined with two dynamic parameters—the real-time slope of the open-circuit voltage-state-of-charge curve at the current prior estimate of the state of charge and the rate of change of the battery operating current—the value of the observation noise covariance matrix is ​​corrected in real time through weighted calculation. For example, a coefficient configuration with reduced voltage observation weight is used in the platform interval mode, while a coefficient configuration with increased voltage observation weight is used in the non-platform interval mode. This allows the adjustment of the observation noise covariance matrix to take into account both static operating point characteristics and dynamic operating condition changes.

[0063] S140. Update the Kalman gain based on the adjusted observation noise covariance matrix, and use the updated Kalman gain and terminal voltage measurement to correct the prior state of charge estimate, so as to obtain the target state of charge estimate at the current time.

[0064] In this context, Kalman gain refers to a key weighting matrix used in the update step of the extended Kalman filter model to optimally fuse the prior state estimate with the current observation. The value of the Kalman gain is jointly determined by the adjusted observation noise covariance matrix and the state prediction covariance matrix, specifically characterizing the degree of confidence in the terminal voltage measurement when correcting the prior state of charge estimate. Optimal correction of the prior state of charge estimate is achieved by multiplying the Kalman gain by the difference between the terminal voltage measurement and the model prediction.

[0065] The target state of charge (SBC) estimate refers to the final output SBC estimate after optimal correction. The target SBC estimate is obtained by weighting the difference between the prior SBC estimate and the measured terminal voltage using the Kalman gain updated based on the adjusted observation noise covariance matrix. It represents the most reliable estimate of the true SBC of the battery at the current moment and serves as the state input for calculating the prior SBC estimate at the next moment.

[0066] In this embodiment, the Kalman gain can be recalculated based on the dynamically adjusted observation noise covariance matrix. This gain is then used as the optimal weighting coefficient and multiplied by the difference between the measured terminal voltage and the predicted terminal voltage based on the prior state of charge estimate. Finally, this weighted correction term is added to the prior state of charge estimate to obtain the target state of charge estimate at the current moment after optimal data fusion.

[0067] This application provides a method for estimating the state of charge (SOC) of a lithium-ion battery. When a SOC estimation request for a battery to be estimated is received, the method acquires the battery terminal voltage measurement value collected at the current moment and a pre-established plateau interval lookup table. A priori SOC estimate is determined based on an extended Kalman filter model. The plateau interval lookup table is determined by offline identification and calibration of at least one plateau interval in the open-circuit voltage-SOC relationship curve of the battery to be estimated. Then, the method queries the plateau interval lookup table based on the prior SOC estimate to determine the plateau interval attribute result of the open-circuit voltage-SOC relationship curve at the prior SOC estimate. Further, based on the plateau interval attribute result, the real-time slope of the open-circuit voltage-SOC relationship curve at the prior SOC estimate, and the battery operating current change rate, the observation noise covariance matrix of the extended Kalman filter model is adjusted. The Kalman gain is then updated based on the adjusted observation noise covariance matrix, and the prior SOC estimate is corrected using the updated Kalman gain and the terminal voltage measurement value to obtain the target SOC estimate at the current moment. The technical solution of this application identifies the operating point characteristics by constructing a lookup table for the platform interval, and builds a multi-dimensional adaptive mechanism by combining the real-time slope and the rate of change of current. In the platform interval, the excessive influence of the voltage observation residual on the state correction is suppressed by increasing the observation noise covariance, so as to avoid drastic fluctuations in the SOC estimate. In the non-platform area and under dynamic operating conditions, the noise configuration is optimized by adjusting the adjustment factor to give full play to the correction role of the voltage observation value, and finally achieves a synergistic improvement in the accuracy and stability of SOC estimation under all operating conditions.

[0068] Example 2

[0069] Figure 2This diagram illustrates a method for estimating the state of charge of a lithium-ion battery according to an embodiment of this application. Based on the foregoing embodiments, this embodiment will provide a detailed explanation of the construction method of the platform interval lookup table. For specific implementation details, please refer to the technical solution of this embodiment. Technical terms that are the same as or corresponding to those in the above embodiments will not be repeated here.

[0070] like Figure 2 As shown, the method specifically includes the following steps:

[0071] S210. Construct a platform interval query table.

[0072] Optionally, the specific steps for constructing the platform range query table may include:

[0073] S2101. A fixed-size charge state window is used, which slides within the entire range from zero to full charge state according to a preset step size.

[0074] The charge state window refers to a fixed-width numerical analysis interval set within the entire charge state range (0% to 100%) of the battery during the construction of the platform interval lookup table. This window slides according to a preset step size to cover all areas to be analyzed.

[0075] In this embodiment, when analyzing the battery open-circuit voltage-state-of-charge curve offline, an analysis interval with a constant width (e.g., 5% state of charge range) is set. This interval starts from the 0% state of charge starting position and moves sequentially towards the 100% state of charge at fixed intervals (e.g., 1% state of charge). After each movement, the window covers a new continuous state of charge interval. By scanning the entire curve in this traversal method, feature parameters of each local interval and plateau interval identification can be achieved.

[0076] S2102. For the state of charge window at each sliding position, determine at least one characteristic parameter of the open-circuit voltage-state of charge relationship curve within the current state of charge interval.

[0077] The state of charge interval refers to the continuous, specific range of the state of charge covered by the window at each specific sliding position during the analysis of the open-circuit voltage-state of charge relationship curve using a fixed-size state of charge window.

[0078] At least one characteristic parameter includes a radius of curvature parameter, a rate of change of voltage parameter, and a voltage fluctuation parameter. The radius of curvature parameter is a geometric characteristic obtained by mathematically quantifying the curvature of the open-circuit voltage-state-of-charge curve at each sampling point within the current state of charge interval. It characterizes the degree to which the curve deviates from a straight line in this interval; a larger value indicates a smoother curve, making it a key indicator for identifying voltage stability when determining plateau intervals. The rate of change of voltage parameter is a characteristic obtained by calculating the maximum absolute value of the first derivative of the open-circuit voltage-state-of-charge curve at each sampling point within the current state of charge interval. It quantifies the degree of voltage change with state of charge in this interval; a smaller value indicates a more stable voltage within this interval, making it a key indicator for judging voltage stability when identifying plateau intervals. Voltage fluctuation parameter refers to the static characteristic quantity obtained by calculating the absolute difference between the highest and lowest voltage values ​​among all sampling points in the current state of charge interval of the open-circuit voltage-state of charge curve. It is used to quantify the overall fluctuation range of the voltage value in that interval. The smaller the value, the more concentrated the voltage distribution and the more significant the voltage plateau characteristics in that interval. It is an important criterion for evaluating voltage stability when identifying plateau intervals.

[0079] Specifically, the processing procedure is consistent for each state of charge interval. To clearly illustrate this technical solution, we will use one state of charge interval as an example. At each discrete position of the sliding window, the specific state of charge interval covered by the current window is treated as an independent analysis unit. For the curve segment corresponding to this interval, its geometric and differential features are calculated and extracted. These include the radius of curvature parameter reflecting the curvature characteristics of the curve, the voltage change rate parameter characterizing the sensitivity to voltage changes, and the voltage fluctuation parameter describing the voltage value distribution range. This provides a quantitative feature data set for subsequent platform interval determination.

[0080] In this embodiment, optionally, the specific implementation of determining at least one characteristic parameter of the open-circuit voltage-state-of-charge relationship curve within the current state-of-charge interval may include:

[0081] S1. For each sampling point of the open-circuit voltage-state-of-charge curve within the current state-of-charge interval, determine the radius of curvature parameter based on the actual curvature of each sampling point.

[0082] The actual degree of curvature refers to the precise mathematical description of the local morphology of the open-circuit voltage-state of charge curve at a specific sampling point using differential geometry. Specifically, it is the quantitative calculation of the curvature value at that point. Its value directly reflects the curvature intensity of the curve at that point—the larger the curvature value, the more severe the curvature; the smaller the curvature value, the closer the curve is to a straight line. This parameter is the basic geometric property for calculating the radius of curvature parameter.

[0083] In this embodiment, for each discrete voltage-state-of-charge data point within the charge state interval covered by the current analysis window, the curvature value corresponding to that point is calculated using differential geometry to quantify its local bending characteristics. Then, the reciprocal of the curvature value is converted into the radius of curvature. Finally, a representative value is extracted from the radius of curvature data of all sampling points in the interval using statistical methods, such as taking the mean, minimum, or a specific quantile value, as the radius of curvature parameter characterizing the straightness of the curve throughout the interval. Specifically, the formula for determining the actual degree of bending at each sampling point within the current charge state interval can be:

[0084]

[0085] In the formula, This indicates the actual curvature of the current sampling point. This represents the open-circuit voltage value at the current sampling point on the open-circuit voltage-state-of-charge curve. This represents the state of charge value at the current sampling point of the open-circuit voltage-state-of-charge curve.

[0086] S2. Determine the voltage change rate parameter based on the maximum absolute value of the first derivative corresponding to each sampling point.

[0087] In this embodiment, for all voltage-state-of-charge sampling points within the current state of charge interval, the first derivative of the open-circuit voltage with respect to the state of charge is calculated and its absolute value is taken. Then, the largest absolute value is selected from the derivative values ​​of all sampling points in the interval, and this maximum value is used as a quantitative index characterizing the fastest rate of change of voltage with the state of charge in the interval, i.e., the voltage change rate parameter.

[0088] S3. Determine the voltage fluctuation parameters based on the difference between the highest and lowest voltages at each sampling point.

[0089] In this embodiment, within the currently analyzed state of charge range, the values ​​of all open-circuit voltage sampling points are traversed to identify the maximum and minimum voltage values. The absolute difference between these two extreme values ​​is calculated, and this difference is used as an indicator to quantify the voltage value distribution range of the range, i.e., the voltage fluctuation parameter.

[0090] S2103. When at least one characteristic parameter data in the current state of charge interval meets the preset platform interval determination condition, the target platform interval is determined according to the current state of charge interval.

[0091] Among them, the plateau interval determination condition refers to a set of pre-defined logical rules for determining whether a state of charge interval has plateau interval characteristics. Its core is to compare the characteristic data such as the radius of curvature parameter, voltage change rate parameter, and voltage fluctuation parameter calculated for the current state of charge interval with their respective preset thresholds. When all characteristic parameters simultaneously meet their threshold requirements, the interval is determined to have plateau characteristics. This is the decision basis from quantitative characteristics to qualitative determination.

[0092] The target platform range refers to the continuous range of charged states that meets specific geometric and electrical characteristics and is finally determined after screening through the platform range judgment conditions.

[0093] In this embodiment, during the sliding window analysis process, if the characteristic data such as the radius of curvature parameter, voltage change rate parameter, and voltage fluctuation parameter calculated for a specific state of charge interval reach the preset threshold standard, then the interval is officially recognized as a valid interval that meets the platform characteristics, and it is used as the basic data unit for constructing the platform interval lookup table.

[0094] Based on the above embodiments, optionally, the specific implementation steps for determining the target platform range according to the current state of charge range may include:

[0095] S1. For each state of charge interval, when at least one characteristic parameter data corresponding to the current state of charge interval simultaneously satisfies its respective preset threshold condition, the current state of charge interval is marked as a potential plateau region.

[0096] Among them, potential plateau regions refer to those regions in the preliminary screening stage where the characteristic data such as the radius of curvature parameter, voltage change rate parameter, and voltage fluctuation parameter corresponding to the state of charge simultaneously meet their respective preset threshold conditions, and are therefore marked as candidate regions with the basic characteristics of plateau regions. These regions still need to undergo subsequent merging processing and width verification before they can be finally determined as target plateau regions.

[0097] In this embodiment, the preset threshold conditions include at least one of the following: the radius of curvature parameter corresponding to the current state of charge interval is greater than a first threshold; the voltage change rate parameter corresponding to the current state of charge interval is less than a second threshold; and the voltage fluctuation parameter corresponding to the current state of charge interval is less than a third threshold.

[0098] Specifically, when traversing and analyzing each state of charge interval, it is necessary to jointly judge multiple feature parameters extracted from it. If and only if the curvature radius parameter, voltage change rate parameter, and voltage fluctuation parameter of the interval all meet the preset standards, that is, the curvature radius parameter exceeds the first threshold to show flat characteristics, the voltage change rate parameter is lower than the second threshold to show slow change, and the voltage fluctuation parameter is less than the third threshold to indicate that the value is stable, it can be marked as a potential plateau region. This is a quantitative criterion for strictly screening plateau regions through multi-dimensional geometric and electrical characteristics.

[0099] For example, suppose that the calculated radius of curvature parameter of a certain state of charge interval is 15 (first threshold is 10), the voltage change rate parameter is 0.002 (second threshold is 0.005), and the voltage fluctuation parameter is 0.008 (third threshold is 0.01). Since the three feature parameters simultaneously satisfy the threshold conditions of "radius of curvature parameter > 10", "voltage change rate parameter < 0.005", and "voltage fluctuation parameter < 0.01", this state of charge interval can be marked as a potential plateau region and enter the subsequent merging verification process.

[0100] S2. Merge all adjacent or overlapping potential platform regions to obtain at least one candidate platform interval, and determine the charge state span corresponding to each candidate platform interval.

[0101] In this context, a candidate platform interval refers to a continuous charged state segment formed by merging all adjacent or overlapping potential platform intervals. The charged state span is the absolute range covered by the candidate platform interval on the charged state axis. Specifically, the charged state span is obtained by calculating the difference between the charged state values ​​of the termination boundary and the starting boundary of the interval. It is a quantitative indicator measuring the physical width of the platform interval and is used to determine whether the candidate platform interval reaches a preset minimum width threshold, thereby determining whether it can ultimately become the target platform interval.

[0102] In this embodiment, the intervals that are continuous or intersect on the charge state axis among the multiple initially marked potential platform regions are integrated and connected to form a larger continuous segment as a candidate platform interval. Then, the charge state difference between the start and end boundaries of each candidate platform interval is calculated to obtain the charge state span value that characterizes its coverage.

[0103] S3. For each candidate platform interval, if the charge state span of the current candidate platform interval is greater than or equal to the preset minimum width threshold, then the current candidate platform interval is determined as the target platform interval.

[0104] Among them, the preset minimum width threshold refers to a critical value of the charge state span that is set in advance to achieve effective platform interval identification.

[0105] Specifically, in the final screening stage, the width compliance of each candidate platform interval formed by merging is checked. When the value of its state of charge span reaches or exceeds the preset minimum width threshold, the interval is deemed to have sufficient platform characteristic continuity, and thus it is determined as the target platform interval.

[0106] S2104. Based on the charge state boundary information corresponding to each target platform interval, construct a platform interval lookup table.

[0107] Among them, the charge state boundary information refers to the specific start and end position values ​​of each target platform interval on the charge state axis, that is, the minimum charge state value and the maximum charge state value corresponding to the platform interval. These values ​​together define the effective range of the platform interval.

[0108] In this embodiment, the initial and final boundary values ​​of all the target platform intervals on the charge state axis can be used as core data and organized and stored according to a specific data structure to form a database that can be quickly queried during online estimation. The core function of this query table is to establish a mapping relationship between the charge state values ​​and the platform interval attributes.

[0109] For example, Figure 3 The open-circuit voltage-state-of-charge (SOC) curve of a certain lithium-ion battery is divided into three platforms—Platform 1, Platform 2, and Platform 3—based on the aforementioned platform interval identification strategy. For example, the first threshold is set to 150, the second threshold to 0.004, the third threshold to 0.015, and the preset minimum width threshold to 0.1. The platform interval identification results are as follows: Platform 1 has an SOC range of 39.47–48.78% and a voltage range of 3.2817–3.2838V; Platform 2 has an SOC range of 71.39–81.86% and a voltage range of 3.3209–3.3207V; Platform 3 has an SOC range of 83.32–92.82% and a voltage range of 3.3213–3.3236V.

[0110] S220. When a request for state of charge estimation of the battery to be estimated is received, the battery terminal voltage measurement value collected at the current time and the pre-established platform interval lookup table are obtained, and the prior state of charge estimate value is determined based on the extended Kalman filter model.

[0111] The platform interval lookup table is determined by offline identification and calibration of at least one platform interval in the open-circuit voltage-state-of-charge curve of the battery to be estimated.

[0112] S230. Based on the prior state of charge estimate, query the platform interval lookup table to determine the platform interval attribute result of the open circuit voltage-state of charge relationship curve at the prior state of charge estimate.

[0113] S240. Based on the platform interval attribute results, the real-time slope of the open-circuit voltage-state of charge relationship curve at the prior state of charge estimate, and the rate of change of battery operating current, adjust the observation noise covariance matrix of the extended Kalman filter model.

[0114] S250. Update the Kalman gain based on the adjusted observation noise covariance matrix, and use the updated Kalman gain and terminal voltage measurement to correct the prior state of charge estimate, so as to obtain the target state of charge estimate at the current time.

[0115] The technical solution of this application embodiment uses a fixed-size state-of-charge window when constructing the platform interval lookup table. The window slides across the entire range from zero to full charge state with a preset step size. Then, for the state-of-charge interval corresponding to each sliding position of the state-of-charge window, at least one characteristic parameter of the open-circuit voltage-state-of-charge relationship curve within the current state-of-charge interval is determined. The at least one characteristic parameter includes a radius of curvature parameter, a voltage change rate parameter, and a voltage fluctuation parameter. Thus, when at least one characteristic parameter data within the current state-of-charge interval meets the preset platform interval determination condition, a target platform interval is determined based on the current state-of-charge interval. Finally, a platform interval lookup table is constructed based on the state-of-charge boundary information corresponding to each target platform interval. The technical solution of this application achieves accurate identification and extraction of plateau intervals in the open-circuit voltage-state-of-charge curve through a quantitative analysis method that combines sliding window and multi-feature parameter fusion. The sliding window mechanism ensures systematic scanning across the entire range. The curvature radius parameter, voltage change rate parameter, and voltage fluctuation parameter construct a complete plateau feature quantification system from three dimensions: geometric shape, change sensitivity, and numerical stability, respectively. Combined with preset threshold conditions and span verification, a strict plateau interval determination logic is formed. Finally, the plateau interval lookup table constructed through state-of-charge boundary information provides a highly reliable prior knowledge basis for adaptive adjustment of observation noise in the subsequent online estimation stage, thereby effectively improving the estimation stability of state of charge in the plateau region and the overall estimation accuracy under all operating conditions.

[0116] Example 3

[0117] Figure 4 This is a schematic diagram of a method for estimating the state of charge of a lithium-ion battery according to an embodiment of this application. Based on the foregoing embodiments, this embodiment provides a more detailed explanation of step S130, which describes how to determine the target code corresponding to the access terminal. For specific implementation details, please refer to the technical solution of this embodiment. Technical terms that are the same as or corresponding to those in the above embodiments will not be repeated here.

[0118] like Figure 4 As shown, the method specifically includes the following steps:

[0119] S310. When a request for state of charge estimation of the battery to be estimated is received, the battery terminal voltage measurement value collected at the current time and the pre-established platform interval lookup table are obtained, and the prior state of charge estimate value is determined based on the extended Kalman filter model.

[0120] The platform interval lookup table is determined by offline identification and calibration of at least one platform interval in the open-circuit voltage-state-of-charge curve of the battery to be estimated.

[0121] S320. Based on the prior state of charge estimate, query the platform interval lookup table to determine the platform interval attribute result of the open circuit voltage-state of charge relationship curve at the prior state of charge estimate.

[0122] S330. Based on the platform range attribute results, determine the platform range correction coefficient.

[0123] Among them, the platform area correction coefficient is a binary control parameter that is dynamically valued based on the platform area attribute results.

[0124] In this embodiment, the value of the platform range correction coefficient can be set according to the platform range attribute result obtained from the platform range lookup table, i.e., whether the current working point belongs to the "platform range result" or the "non-platform range result". If the platform range attribute result is a platform range result, the platform range correction coefficient is set to a preset positive value greater than zero; if the platform range attribute result is a non-platform range result, the platform range correction coefficient is set to zero.

[0125] For example, assuming that after querying the platform interval lookup table based on the prior state of charge estimate, the platform interval attribute result is "platform interval result", the platform area correction coefficient is set to the preset value of 0.8 according to the judgment rule; if the query result is "non-platform interval result", the system sets the platform area correction coefficient to 0.

[0126] S340. Determine the real-time slope of the open-circuit voltage-state-of-charge curve at the prior estimate of the state of charge.

[0127] In this embodiment, the instantaneous rate of change of the open-circuit voltage-state of charge curve corresponding to the specific operating point can be calculated based on the prior state of charge estimate predicted by the extended Kalman filter model at the current moment. This real-time slope characterizes the sensitivity of the open-circuit voltage to the change of the state of charge at this time, and is a key parameter that dynamically reflects the nonlinear characteristics of the curve region where the battery is currently operating.

[0128] S350. Determine the rate of change of the battery operating current based on the first current measurement value at the current moment and the second current measurement value at the previous moment.

[0129] The first and second current measurements refer to the real-time current data collected by the current sensor at two consecutive adjacent sampling times. The first current measurement represents the instantaneous current at the current moment, and the second current measurement represents the instantaneous current at the previous moment. The rate of change of battery operating current quantitatively characterizes the intensity of dynamic fluctuations in battery operating current per unit time.

[0130] In this embodiment, the battery operating current change rate parameter, which characterizes the dynamic change rate of the current, can be obtained by calculating the difference between the first current measurement value collected at the current moment and the second current measurement value collected at the previous moment, and then performing a ratio operation between the difference and the sampling time interval.

[0131] S360. Input the platform area correction coefficient, real-time slope, and battery operating current change rate into the observation noise covariance adjustment model to dynamically calculate and update the observation noise covariance matrix at the current moment.

[0132] The observation noise covariance adjustment model refers to using the observation noise covariance matrix of the previous time step as a benchmark, and introducing a plateau correction coefficient, the real-time slope of the open-circuit voltage-state-of-charge curve at the prior estimate of the state of charge, and the rate of change of the battery operating current as dynamic adjustment parameters. It then combines these parameters with a plateau adjustment factor and a current adjustment factor for weighted calculation, thereby outputting the mathematical expression of the observation noise covariance matrix at the current time step. In this embodiment, optionally, the expression formula of the observation noise covariance adjustment model includes:

[0133]

[0134] In the formula, This represents the observation noise covariance matrix at the current moment. This represents the observation noise covariance matrix at the previous time step. This represents the platform area correction factor. This represents the real-time slope of the open-circuit voltage-state-of-charge curve at the prior estimate of the state of charge. This indicates the rate of change of the battery's operating current. As a platform adjustment factor, For current regulation factor, and All are preset values ​​greater than zero.

[0135] In this embodiment, three real-time parameters—the platform correction coefficient set based on the platform interval attribute results, the real-time slope of the open-circuit voltage-state of charge relationship curve at the prior state of charge estimate, and the rate of change of battery operating current calculated based on the current measurement—can be substituted into the mathematical expression of the observation noise covariance adjustment model for calculation, thereby generating and updating the observation noise covariance matrix for state estimation at the current moment in real time.

[0136] S370. Based on the adjusted observation noise covariance matrix, the prior error covariance matrix at the current moment, and the real-time slope of the open-circuit voltage-state of charge relationship curve at the prior state of charge estimate, the Kalman gain at the current moment is calculated using the standard gain calculation formula of the extended Kalman filter.

[0137] The standard gain calculation formula refers to the standard equation for calculating Kalman gain using established mathematical relationships. Optionally, the standard gain calculation formula can be expressed as:

[0138]

[0139] in, This represents the Kalman gain at time k. Let the prior error covariance matrix at time k-1 be represented. This represents the real-time slope of the open-circuit voltage-state-of-charge curve at time k, relative to the prior estimate of the state of charge. The observation noise covariance matrix at time k.

[0140] In this embodiment, the observation noise covariance matrix dynamically updated by the observation noise covariance adjustment model, the prior error covariance matrix predicted by the extended Kalman filter model at the current moment, and the real-time slope of the open-circuit voltage-state of charge relationship curve at the prior state of charge estimate can be used as input parameters. These parameters are then substituted into the predetermined standard gain calculation formula in the extended Kalman filter algorithm to calculate the optimal Kalman gain value for state correction at the current moment.

[0141] S380. The prior state of charge estimate is corrected using the updated Kalman gain and terminal voltage measurement to obtain the target state of charge estimate at the current moment.

[0142] In this embodiment, the specific implementation method of correcting the prior state of charge estimate using the updated Kalman gain and terminal voltage measurement to obtain the target state of charge estimate at the current moment may include:

[0143] S3801. Compare the measured terminal voltage with the predicted terminal voltage based on the prior state of charge estimate to obtain the voltage observation residual.

[0144] In this embodiment, the voltage observation residual, which characterizes the deviation between the actual observation and the model prediction, is obtained by calculating the difference between the measured value of the battery terminal voltage at the current moment and the predicted value of the terminal voltage calculated by the extended Kalman filter model based on the prior state of charge estimate through the observation equation.

[0145] S3802. Multiply the updated Kalman gain by the voltage observation residual to obtain the state correction.

[0146] Specifically, the updated Kalman gain obtained by solving the standard gain calculation formula at the current moment is multiplied by a scalar multiplication with the voltage observation residual obtained by comparing the measured terminal voltage value and the predicted terminal voltage value, thereby calculating the state correction amount used to correct the prior state of charge estimate.

[0147] S3803. Based on the extended Kalman filter model, the prior charge state estimate is corrected using the state correction factor to obtain the optimal target charge state estimate at the current moment.

[0148] In this embodiment, the calculated state correction amount and the prior charge state estimate generated in the prediction step can be algebraically superimposed according to the state update mechanism of the extended Kalman filter algorithm. This data fusion method eliminates prediction bias and finally outputs the target charge state estimate at the current time that meets the optimal estimation criterion.

[0149] Under the same operating conditions, the state of charge estimation performance of the extended Kalman filter method with fixed observation noise covariance and the extended Kalman filter method with adaptive observation noise covariance (i.e., the technical solution provided in this embodiment) was compared and analyzed. The root mean square index of the state of charge estimation error of different methods during the test was evaluated. The comparison results are shown in Table 3. The root mean square index of the extended Kalman filter method with fixed observation noise covariance is 0.0124, and the root mean square index of the extended Kalman filter method with adaptive observation noise covariance is 0.0044. Figure 5 A schematic diagram comparing the state-of-charge estimation results of the extended Kalman filter method with fixed observation noise covariance and adaptive observation noise covariance. Figure 6 A schematic diagram comparing the state-of-charge estimation errors of the extended Kalman filter method with fixed observation noise covariance and adaptive observation noise covariance.

[0150] The technical solution of this application, when adjusting the observation noise covariance matrix of the extended Kalman filter model, specifically includes: determining the plateau region correction coefficient based on the plateau interval attribute results; determining the real-time slope of the open-circuit voltage-state-of-charge (SOC) relationship curve at the prior SOC estimate; determining the battery operating current change rate based on the first current measurement value at the current moment and the second current measurement value at the previous moment; inputting the plateau region correction coefficient, the real-time slope, and the battery operating current change rate into the observation noise covariance adjustment model, dynamically calculating and updating the observation noise covariance matrix at the current moment. The technical solution of this application, based on the plateau region correction coefficient set according to the plateau interval attribute results, actively enhances the observation noise within the plateau interval to suppress the excessive influence of voltage observation residuals on SOC correction, and accurately reflects the voltage sensitivity of the current operating point by combining the real-time slope parameter, ensuring that the noise adjustment matches the nonlinear characteristics of the curve; introducing the battery operating current change rate effectively senses the dynamics of external operating conditions, avoiding estimation bias caused by model mismatch when the current fluctuates drastically. This multi-parameter collaborative adjustment strategy significantly improves the estimation stability of the extended Kalman filter in the plateau range and its adaptive capability under all operating conditions, ultimately achieving overall optimization of the SOC estimation accuracy.

[0151] Example 4

[0152] Figure 7 This is a schematic diagram of a lithium-ion battery state-of-charge estimation device provided in an embodiment of this application. The device includes:

[0153] The data acquisition module 410 is used to acquire the battery terminal voltage measurement value collected at the current moment and the pre-established plateau interval lookup table when a request for estimating the state of charge of the battery to be estimated is received, and to determine the prior state of charge estimate based on the extended Kalman filter model; wherein, the plateau interval lookup table is determined by offline identification and calibration of at least one plateau interval in the open circuit voltage-state of charge relationship curve of the battery to be estimated.

[0154] The platform attribute determination module 420 is used to query the platform interval lookup table based on the prior state of charge estimate and determine the platform interval attribute result of the open circuit voltage-state of charge relationship curve at the prior state of charge estimate.

[0155] The noise covariance adjustment module 430 is used to adjust the observation noise covariance matrix of the extended Kalman filter model based on the platform interval attribute results, the real-time slope of the open-circuit voltage-state of charge relationship curve at the prior state of charge estimate, and the battery operating current change rate.

[0156] The state of charge estimation module 440 is used to update the Kalman gain based on the adjusted observation noise covariance matrix, and to correct the prior state of charge estimate using the updated Kalman gain and the terminal voltage measurement value, so as to obtain the target state of charge estimate at the current time.

[0157] This application provides a device for estimating the state of charge (SOC) of a lithium-ion battery. When the device receives a SOC estimation request for a battery to be estimated, it acquires the battery terminal voltage measurement value collected at the current moment and a pre-established plateau interval lookup table. Based on an extended Kalman filter (EPF) model, it determines a priori SOC estimate. The plateau interval lookup table is determined by offline identification and calibration of at least one plateau interval in the open-circuit voltage-SOC relationship curve of the battery to be estimated. Then, based on the priori SOC estimate, the device queries the plateau interval lookup table to determine the plateau interval attribute result of the open-circuit voltage-SOC relationship curve at the priori SOC estimate. Further, based on the plateau interval attribute result, the real-time slope of the open-circuit voltage-SOC relationship curve at the priori SOC estimate, and the battery operating current change rate, it adjusts the observation noise covariance matrix of the EPF model. Then, based on the adjusted observation noise covariance matrix, the Kalman gain is updated, and the updated Kalman gain and the terminal voltage measurement value are used to correct the priori SOC estimate, resulting in the target SOC estimate for the current moment. The technical solution of this application identifies the operating point characteristics by constructing a lookup table for the platform interval, and builds a multi-dimensional adaptive mechanism by combining the real-time slope and the rate of change of current. In the platform interval, the excessive influence of the voltage observation residual on the state correction is suppressed by increasing the observation noise covariance, so as to avoid drastic fluctuations in the SOC estimate. In the non-platform area and under dynamic operating conditions, the noise configuration is optimized by adjusting the adjustment factor to give full play to the correction role of the voltage observation value, and finally achieves a synergistic improvement in the accuracy and stability of SOC estimation under all operating conditions.

[0158] Based on the above-mentioned device, optionally, the lithium-ion battery state of charge estimation device also includes: a platform interval lookup table construction module, including:

[0159] A window sliding unit is used to slide a fixed-size charge state window in a preset step size across the entire range from zero to full charge state.

[0160] The window feature parameter determination unit is used to determine at least one feature parameter of the open-circuit voltage-state-of-charge relationship curve within the state of charge interval corresponding to each sliding position of the state of charge window; wherein, the at least one feature parameter includes a radius of curvature parameter, a rate of change of voltage parameter, and a voltage fluctuation parameter.

[0161] The platform identification unit is used to determine the target platform interval based on the current state of charge interval when the at least one feature parameter data within the current state of charge interval meets the preset platform interval determination condition.

[0162] The platform query table determination unit is used to construct the platform interval query table based on the charge state boundary information corresponding to each target platform interval.

[0163] Based on the above-mentioned device, optionally, a window feature parameter determination unit is used to determine the radius of curvature parameter for each sampling point of the open-circuit voltage-state of charge relationship curve within the current state of charge interval, based on the actual curvature of each sampling point; determine the voltage change rate parameter based on the maximum absolute value of the first derivative corresponding to each sampling point; and determine the voltage fluctuation parameter based on the difference between the highest and lowest voltages among the sampling points.

[0164] Based on the above-mentioned device, optionally, a platform identification unit is specifically used to mark the current state of charge interval as a potential platform region when the at least one feature parameter data corresponding to the current state of charge interval simultaneously meets its respective preset threshold conditions; merge all adjacent or overlapping potential platform regions to obtain at least one candidate platform interval, and determine the state of charge span corresponding to each candidate platform interval; for each candidate platform interval, if the state of charge span of the current candidate platform interval is greater than or equal to a preset minimum width threshold, then determine the current candidate platform interval as a target platform interval; wherein, the state of charge span is the absolute range covered by the candidate platform interval on the state of charge axis.

[0165] Based on the above-mentioned device, optionally, the preset threshold condition includes at least one of the following: the radius of curvature parameter corresponding to the current state of charge interval is greater than a first threshold; the voltage change rate parameter corresponding to the current state of charge interval is less than a second threshold; and the voltage fluctuation parameter corresponding to the current state of charge interval is less than a third threshold.

[0166] Based on the above-mentioned device, optionally, the data acquisition module 410 is also used to input the target state of charge estimation value of the previous moment and the currently acquired load current value into the extended Kalman filter model to predict the prior state of charge estimation value of the current moment.

[0167] Based on the above-mentioned device, optionally, the platform attribute determination module 420 includes:

[0168] The platform interval query unit is used to compare the prior state of charge estimate with the state of charge boundary values ​​of each platform interval recorded in the platform interval query table;

[0169] The platform attribute determination unit is used to determine whether the prior state of charge estimate is within a platform interval based on whether it falls within the boundary of any platform interval, and output the corresponding platform interval attribute result; wherein, the platform interval attribute result includes: platform interval result or non-platform interval result.

[0170] Based on the above-mentioned device, optionally, the noise covariance adjustment module 430 includes:

[0171] The correction coefficient determination unit is used to determine the platform area correction coefficient based on the platform area attribute results.

[0172] A real-time slope determination unit is used to determine the real-time slope of the open-circuit voltage-state-of-charge relationship curve at the prior state-of-charge estimate.

[0173] The current change rate determination unit is used to determine the battery operating current change rate based on the first current measurement value at the current moment and the second current measurement value at the previous moment;

[0174] The noise covariance adjustment unit is used to input the platform area correction coefficient, the real-time slope, and the battery operating current change rate into the observation noise covariance adjustment model, and dynamically calculate and update the observation noise covariance matrix at the current moment.

[0175] Based on the above-mentioned device, optionally, the expression formula of the observation noise covariance adjustment model includes:

[0176]

[0177] In the formula, This represents the observation noise covariance matrix at the current moment. This represents the observation noise covariance matrix at the previous time step. This represents the platform area correction factor. This represents the real-time slope of the open-circuit voltage-state-of-charge curve at the prior estimate of the state of charge. This represents the rate of change of the battery's operating current. As a platform adjustment factor, For current regulation factor, and All are preset values ​​greater than zero.

[0178] Based on the above-mentioned device, optionally, a correction coefficient determination unit is used to set the platform area correction coefficient to a preset positive value greater than zero if the platform area attribute result is a platform area result; and to set the platform area correction coefficient to zero if the platform area attribute result is a non-platform area result.

[0179] Based on the above device, optionally, the state of charge estimation module 440 is used to calculate the Kalman gain at the current time based on the adjusted observation noise covariance matrix, the prior error covariance matrix at the current time, and the real-time slope of the open-circuit voltage-state of charge relationship curve at the prior state of charge estimate, by using the standard gain calculation formula of extended Kalman filtering.

[0180] Based on the above-mentioned device, optionally, the state of charge estimation module 440 is further configured to compare the measured terminal voltage value with the predicted terminal voltage value based on the prior state of charge estimation value to obtain the voltage observation residual; multiply the updated Kalman gain with the voltage observation residual to obtain the state correction amount; and correct the prior state of charge estimation value using the state correction amount based on the extended Kalman filter model to obtain the optimal target state of charge estimation value at the current time.

[0181] The lithium-ion battery state-of-charge estimation device provided in this application embodiment can execute the lithium-ion battery state-of-charge estimation method provided in any embodiment of this application, and has the corresponding functional modules and beneficial effects of executing the method.

[0182] It is worth noting that the various units and modules included in the above system are only divided according to functional logic, but are not limited to the above division, as long as the corresponding functions can be achieved; in addition, the specific names of each functional unit are only for easy differentiation and are not used to limit the protection scope of the embodiments of this application.

[0183] Example 5

[0184] Figure 8 This is a schematic diagram of the structure of a controller provided in an embodiment of this application. Figure 8 A block diagram is shown of an exemplary controller 50 suitable for implementing embodiments of the present application. Figure 8 The controller 50 shown is merely an example and should not impose any limitations on the functionality and scope of use of the embodiments of this application.

[0185] like Figure 8 As shown, the controller 50 is presented in the form of a general-purpose computing device. The components of the controller 50 may include, but are not limited to: one or more processors or processing units 501, system memory 502, and bus 503 connecting different system components (including system memory 502 and processing unit 501).

[0186] Bus 503 represents one or more of several bus architectures, including a memory bus or memory controller, a peripheral bus, a graphics acceleration port, a processor, or a local bus using any of the various bus architectures. For example, these architectures include, but are not limited to, the Industry Standard Architecture (ISA) bus, the Micro Channel Architecture (MAC) bus, the Enhanced ISA bus, the Video Electronics Standards Association (VESA) local bus, and the Peripheral Component Interconnect (PCI) bus.

[0187] The controller 50 typically includes a variety of computer system readable media. These media can be any available media that can be accessed by the controller 50, including volatile and non-volatile media, and removable and non-removable media.

[0188] System memory 502 may include computer system readable media in the form of volatile memory, such as random access memory (RAM) 504 and / or cache memory 505. Controller 50 may further include other removable / non-removable, volatile / non-volatile computer system storage media. By way of example only, storage system 506 may be used to read and write non-removable, non-volatile magnetic media (…). Figure 8 Not shown; usually referred to as a "hard drive"). Although Figure 8 As not shown, a disk drive for reading and writing to a removable non-volatile disk (e.g., a "floppy disk") and an optical disk drive for reading and writing to a removable non-volatile optical disk (e.g., a CD-ROM, DVD-ROM, or other optical media) may be provided. In these cases, each drive may be connected to bus 503 via one or more data media interfaces. Memory 502 may include at least one program product having a set (e.g., at least one) of program modules configured to perform the functions of the embodiments of this application.

[0189] A program / utility 508 having a set (at least one) of program modules 507 may be stored, for example, in memory 502. Such program modules 507 include, but are not limited to, an operating system, one or more application programs, other program modules, and program data. Each or some combination of these examples may include an implementation of a network environment. Program modules 507 typically perform the functions and / or methods described in the embodiments of this application.

[0190] The controller 50 can also communicate with one or more external devices 509 (e.g., keyboard, pointing device, display 510, etc.), and with one or more devices that enable a user to interact with the controller 50, and / or with any device that enables the controller 50 to communicate with one or more other computing devices (e.g., network card, modem, etc.). This communication can be performed via input / output (I / O) interface 511. Furthermore, the controller 50 can also communicate with one or more networks (e.g., local area network (LAN), wide area network (WAN), and / or public networks, such as the Internet) via network adapter 512. As shown, network adapter 512 communicates with other modules of the controller 50 via bus 503. It should be understood that, although... Figure 8 As not shown, other hardware and / or software modules can be used in conjunction with controller 50, including but not limited to: microcode, device drivers, redundant processing units, external disk drive arrays, RAID systems, tape drives, and data backup storage systems.

[0191] The processing unit 501 executes various functional applications and page processing by running programs stored in the system memory 502, such as implementing the lithium-ion battery state of charge estimation method provided in the embodiments of this application.

[0192] Example 6

[0193] This application also provides a storage medium containing computer-executable instructions, which, when executed by a computer processor, are used to perform a method for estimating the state of charge of a lithium-ion battery. The method is based on an extended Kalman filter model and includes:

[0194] When a request for state of charge estimation of the battery to be estimated is received, the battery terminal voltage measurement value collected at the current time and the pre-established plateau interval lookup table are obtained, and the prior state of charge estimate is determined based on the extended Kalman filter model; wherein, the plateau interval lookup table is determined by offline identification and calibration of at least one plateau interval in the open circuit voltage-state of charge relationship curve of the battery to be estimated.

[0195] Based on the prior state of charge estimate, the plateau interval lookup table is queried to determine the plateau interval attribute result of the open circuit voltage-state of charge relationship curve at the prior state of charge estimate;

[0196] Based on the platform interval attribute results, the real-time slope of the open-circuit voltage-state of charge relationship curve at the prior state of charge estimate, and the rate of change of battery operating current, the observation noise covariance matrix of the extended Kalman filter model is adjusted.

[0197] The Kalman gain is updated based on the adjusted observation noise covariance matrix, and the prior state of charge estimate is corrected using the updated Kalman gain and the terminal voltage measurement to obtain the target state of charge estimate at the current time.

[0198] The computer storage medium in this application embodiment can be any combination of one or more computer-readable media. A computer-readable medium can be a computer-readable signal medium or a computer-readable storage medium. A computer-readable storage medium can be, for example,—but not limited to—an electrical, magnetic, optical, electromagnetic, infrared, or semiconductor system, apparatus, or device, or any combination thereof. More specific examples of computer-readable storage media (a non-exhaustive list) include: an electrical connection having one or more wires, a portable computer disk, a hard disk, random access memory (RAM), read-only memory (ROM), erasable programmable read-only memory (EPROM or flash memory), optical fiber, portable compact disk read-only memory (CD-ROM), optical storage device, magnetic storage device, or any suitable combination thereof. In this document, a computer-readable storage medium can be any tangible medium that contains or stores a program that can be used by or in conjunction with an instruction execution system, apparatus, or device.

[0199] Computer-readable signal media may include data signals propagated in baseband or as part of a carrier wave, carrying computer-readable program code. Such propagated data signals may take various forms, including but not limited to electromagnetic signals, optical signals, or any suitable combination thereof. Computer-readable signal media may also be any computer-readable medium other than computer-readable storage media, capable of sending, propagating, or transmitting programs for use by or in connection with an instruction execution system, apparatus, or device.

[0200] Program code contained on a computer-readable medium may be transmitted using any suitable medium, including—but not limited to—wireless, wire, optical fiber, RF, etc., or any suitable combination thereof.

[0201] Computer program code for performing the operations of the embodiments of this application can be written in one or more programming languages ​​or a combination thereof. Programming languages ​​include object-oriented programming languages—such as Java, Smalltalk, and C++—and conventional procedural programming languages—such as the "C" language or similar programming languages. The program code can be executed entirely on the user's computer, partially on the user's computer, as a standalone software package, partially on the user's computer and partially on a remote computer, or entirely on a remote computer or server. In cases involving remote computers, the remote computer can be connected to the user's computer via any type of network—including a local area network (LAN) or a wide area network (WAN)—or can be connected to an external computer (e.g., via the Internet using an Internet service provider).

[0202] Note that the above description is merely a preferred embodiment and the technical principles employed in this application. Those skilled in the art will understand that this application is not limited to the specific embodiments described herein, and various obvious changes, readjustments, and substitutions can be made without departing from the scope of protection of this application. Therefore, although this application has been described in detail through the above embodiments, this application is not limited to the above embodiments. Many other equivalent embodiments may be included without departing from the concept of this application, and the scope of this application is determined by the scope of the appended claims.

Claims

1. A method for estimating the state of charge (SOC) of a lithium-ion battery, the method being implemented based on an extended Kalman filter model, characterized in that... The method includes: When a request for state of charge estimation of the battery to be estimated is received, the battery terminal voltage measurement value collected at the current time and the pre-established plateau interval lookup table are obtained, and the prior state of charge estimate is determined based on the extended Kalman filter model; wherein, the plateau interval lookup table is determined by offline identification and calibration of at least one plateau interval in the open circuit voltage-state of charge relationship curve of the battery to be estimated. Based on the prior state of charge estimate, the plateau interval lookup table is queried to determine the plateau interval attribute result of the open circuit voltage-state of charge relationship curve at the prior state of charge estimate; Based on the platform interval attribute results, the real-time slope of the open-circuit voltage-state of charge relationship curve at the prior state of charge estimate, and the rate of change of battery operating current, the observation noise covariance matrix of the extended Kalman filter model is adjusted. The Kalman gain is updated based on the adjusted observation noise covariance matrix, and the prior state of charge estimate is corrected using the updated Kalman gain and the terminal voltage measurement to obtain the target state of charge estimate at the current time. The observation noise covariance matrix is ​​determined by an observation noise covariance adjustment model, the expression of which includes: In the formula, This represents the observation noise covariance matrix at the current moment. This represents the observation noise covariance matrix at the previous time step. This represents the platform area correction factor. This represents the real-time slope of the open-circuit voltage-state-of-charge curve at the prior estimate of the state of charge. This represents the rate of change of the battery's operating current. As a platform adjustment factor, For current regulation factor, and All are preset values ​​greater than zero; The step of determining the platform area correction coefficient based on the platform area attribute results includes: If the platform interval attribute result is a platform interval result, then the platform interval correction coefficient is set to a preset positive value greater than zero; If the platform interval attribute result is a non-platform interval result, then the platform interval correction coefficient is set to zero.

2. The method according to claim 1, characterized in that, The method further includes: constructing the platform interval query table, including: A fixed-size charge state window is used, which slides within the entire range from zero to full charge state according to a preset step size; For the state of charge window at each sliding position, determine at least one characteristic parameter of the open-circuit voltage-state of charge relationship curve within the current state of charge interval; wherein, the at least one characteristic parameter includes a radius of curvature parameter, a rate of change of voltage parameter, and a voltage fluctuation parameter; When at least one characteristic parameter data within the current state of charge interval meets the preset platform interval determination condition, the target platform interval is determined based on the current state of charge interval. Based on the charge state boundary information corresponding to each target platform interval, construct the platform interval lookup table.

3. The method according to claim 2, characterized in that, The determination of at least one characteristic parameter of the open-circuit voltage-state-of-charge relationship curve within the current state-of-charge interval includes: For each sampling point of the open-circuit voltage-state-of-charge curve within the current state-of-charge interval, the radius of curvature parameter is determined based on the actual curvature of each sampling point; The voltage change rate parameter is determined based on the maximum absolute value of the first derivative corresponding to each sampling point. The voltage fluctuation parameters are determined based on the difference between the highest and lowest voltages at each sampling point.

4. The method according to claim 2, characterized in that, When at least one characteristic parameter data within the current state of charge interval satisfies a preset platform interval determination condition, determining the target platform interval based on the current state of charge interval includes: For each of the stated state of charge intervals, when the at least one feature parameter data corresponding to the current state of charge interval simultaneously satisfies their respective preset threshold conditions, the current state of charge interval is marked as a potential plateau region. Merge all adjacent or overlapping potential platform regions to obtain at least one candidate platform interval, and determine the charge state span corresponding to each candidate platform interval; For each candidate platform interval, if the charge state span of the current candidate platform interval is greater than or equal to a preset minimum width threshold, then the current candidate platform interval is determined as the target platform interval; wherein, the charge state span is the absolute range covered by the candidate platform interval on the charge state axis.

5. The method according to claim 4, characterized in that, The preset threshold condition includes at least one of the following: The radius of curvature parameter corresponding to the current charge state interval is greater than the first threshold. The voltage change rate parameter corresponding to the current state of charge interval is less than the second threshold; and... The voltage fluctuation parameter corresponding to the current state of charge interval is less than the third threshold.

6. The method according to claim 1, characterized in that, The determination of the prior state of charge estimate based on the extended Kalman filter model includes: The target state of charge estimate from the previous moment and the currently acquired load current value are input into the extended Kalman filter model to predict the prior state of charge estimate for the current moment.

7. The method according to claim 1, characterized in that, The step of querying the plateau interval lookup table based on the prior state of charge estimate to determine the plateau interval attribute result of the open-circuit voltage-state of charge relationship curve at the prior state of charge estimate includes: The prior state of charge estimate is compared with the state of charge boundary values ​​of each platform interval recorded in the platform interval lookup table; Based on whether the prior state of charge estimate falls within the boundary of any plateau interval, determine whether the prior state of charge estimate is within a plateau interval and output the corresponding plateau interval attribute result; wherein, the plateau interval attribute result includes: plateau interval result or non-plateau interval result.

8. The method according to claim 1, characterized in that, The adjustment of the observation noise covariance matrix of the extended Kalman filter model based on the platform interval attribute results, the real-time slope of the open-circuit voltage-state-of-charge relationship curve at the prior state-of-charge estimate, and the battery operating current rate of change includes: Determine the real-time slope of the open-circuit voltage-state-of-charge curve at the prior estimate of the state of charge. The rate of change of the battery operating current is determined based on the first current measurement value at the current moment and the second current measurement value at the previous moment; The platform area correction coefficient, the real-time slope, and the battery operating current change rate are input into the observation noise covariance adjustment model to dynamically calculate and update the observation noise covariance matrix at the current moment.

9. The method according to claim 1, characterized in that, The update of the Kalman gain based on the adjusted observation noise covariance matrix includes: Based on the adjusted observation noise covariance matrix, the prior error covariance matrix at the current moment, and the real-time slope of the open-circuit voltage-state-of-charge curve at the prior state-of-charge estimate, the Kalman gain at the current moment is calculated using the standard gain calculation formula of the extended Kalman filter.

10. The method according to claim 1, characterized in that, The step of correcting the prior state of charge estimate using the updated Kalman gain and the measured terminal voltage to obtain the target state of charge estimate at the current moment includes: The measured terminal voltage value is compared with the predicted terminal voltage value based on the prior state of charge estimate to obtain the voltage observation residual; The updated Kalman gain is multiplied by the voltage observation residual to obtain the state correction. Based on the extended Kalman filter model, the prior state of charge estimate is corrected using the state correction amount to obtain the optimal target state of charge estimate at the current moment.

11. A device for estimating the state of charge of a lithium-ion battery, the device being implemented based on an extended Kalman filter model, characterized in that, The device includes: The data acquisition module is used to acquire the battery terminal voltage measurement value collected at the current moment and the pre-established plateau interval lookup table when a request for estimating the state of charge of the battery to be estimated is received, and to determine the prior state of charge estimate based on the extended Kalman filter model; wherein, the plateau interval lookup table is determined by offline identification and calibration of at least one plateau interval in the open circuit voltage-state of charge relationship curve of the battery to be estimated. The platform attribute determination module is used to query the platform interval lookup table based on the prior state of charge estimate and determine the platform interval attribute result of the open circuit voltage-state of charge relationship curve at the prior state of charge estimate. The noise covariance adjustment module is used to adjust the observation noise covariance matrix of the extended Kalman filter model based on the platform interval attribute results, the real-time slope of the open-circuit voltage-state of charge relationship curve at the prior state of charge estimate, and the battery operating current change rate. The state of charge estimation module is used to update the Kalman gain based on the adjusted observation noise covariance matrix, and to correct the prior state of charge estimate using the updated Kalman gain and the terminal voltage measurement value, so as to obtain the target state of charge estimate at the current time. The observation noise covariance matrix is ​​determined by an observation noise covariance adjustment model, the expression of which includes: In the formula, This represents the observation noise covariance matrix at the current moment. This represents the observation noise covariance matrix at the previous time step. This represents the platform area correction factor. This represents the real-time slope of the open-circuit voltage-state-of-charge curve at the prior estimate of the state of charge. This represents the rate of change of the battery's operating current. As a platform adjustment factor, For current regulation factor, and All are preset values ​​greater than zero; The step of determining the platform area correction coefficient based on the platform area attribute results includes: If the platform interval attribute result is a platform interval result, then the platform interval correction coefficient is set to a preset positive value greater than zero; If the platform interval attribute result is a non-platform interval result, then the platform interval correction coefficient is set to zero.

12. An electronic device, characterized in that, The electronic device includes: At least one processor; and A memory communicatively connected to the at least one processor; wherein, The memory stores a computer program that can be executed by the at least one processor to enable the at least one processor to perform the method for estimating the state of charge of a lithium-ion battery according to any one of claims 1-10.

13. A computer-readable storage medium, characterized in that, The computer-readable storage medium stores computer instructions that, when executed by a processor, implement the method for estimating the state of charge of a lithium-ion battery as described in any one of claims 1-10.

Images