A charging method and system of a new energy vehicle, an electronic device and a storage medium

Abstract

The application discloses a charging method and system of a new energy vehicle, an electronic device and a storage medium. The method comprises the following steps: acquiring end voltage and charging capacity time sequence data in a historical charging process of a battery; constructing a multi-order differential matrix of the end voltage and the charging capacity; performing curvature feature decomposition on the matrix; extracting a geometric invariant vector comprising a first principal curvature value, a second principal curvature value and a direction coupling factor; mapping all the vectors to a degenerate manifold space; calculating a manifold distance of adjacent projection coordinates in the manifold space; detecting a mutation point as a degenerate transition point; dividing the entire charging process into multiple degenerate stages and determining a current stage according to the degenerate transition point; calculating a modulus length deviation rate and a direction deviation rate of a current geometric invariant vector; obtaining a degenerate degree score through weighted harmonic averaging; matching basic waveform parameters from a template library according to the degenerate stage and the score and adjusting the basic waveform parameters; generating a positive-negative alternating rest charging current waveform; and charging the battery after applying a safety boundary constraint.

Description

Technical Field

[0001] This invention belongs to the field of new energy vehicle battery charging technology, specifically relating to a charging method, system, electronic device and storage medium for new energy vehicles. Background Technology

[0002] During long-term use, the internal electrochemical characteristics of the power batteries in new energy vehicles gradually degrade, manifested as a decrease in usable capacity and an increase in internal resistance. The degradation process of the battery is not uniform; with the accumulation of charge-discharge cycles, the battery will experience different degradation stages, each with different electrochemical states and tolerance.

[0003] Currently, conventional charging methods typically employ fixed charging strategies, such as constant current constant voltage charging or multi-stage constant current charging, using a uniform charging current waveform or one adjusted based on simple rules throughout the battery's lifespan. These methods fail to effectively utilize the rich degradation information inherent in the changes in terminal voltage with charging capacity during the charging process, nor do they fully extract the multi-order differential characteristics of terminal voltage with respect to charging capacity to finely characterize the evolution of the battery's internal electrochemical state.

[0004] Specifically, the curve of terminal voltage versus charging capacity reflects the phase transition process of the active materials inside the battery and the kinetics of lithium-ion transport. The curvature characteristics of the curve, such as the degree and direction of curvature, are closely related to the battery's degradation mode. However, existing methods generally lack the means to systematically extract and quantify these curvature characteristics, making it difficult to accurately identify the battery degradation stage and finely assess the degree of degradation during charging. Consequently, it is impossible to dynamically generate a charging current waveform that matches the actual degradation state of the battery.

[0005] The aforementioned problems cause the battery to continue to experience charging stresses similar to or incompatible with those in the early stages even after entering the mid-to-late stages of degradation. This accelerates the loss of active materials inside the battery and the occurrence of interfacial side reactions, ultimately shortening the overall lifespan of the battery. Summary of the Invention

[0006] In view of this, the present invention provides a charging method, system, electronic device and storage medium for new energy vehicles, in order to solve the problem that the existing technology generally lacks the means to systematically extract and quantify these curvature features, making it difficult to accurately identify the battery degradation stage and finely assess the degree of degradation during the charging process, and thus unable to dynamically generate a charging current waveform that matches the actual degradation state of the battery.

[0007] The technical solution adopted in this invention is as follows: A charging method for new energy vehicles, comprising: Step S1: Obtain the terminal voltage time-series data and charging capacity time-series data of the target vehicle battery during multiple historical charging processes. For each charging process, construct a multi-order differential matrix based on the terminal voltage time-series data and charging capacity time-series data. In step S1, the construction of the multi-order differential matrix based on the terminal voltage time-series data and the charging capacity time-series data specifically includes: Using the charging capacity time series data of a single charging process as the independent variable and the terminal voltage time series data as the dependent variable, the first, second, and third derivative sequences of the terminal voltage with respect to the charging capacity are calculated. The first, second, and third derivative sequences are concatenated column by column to form a multi-order differential matrix with three columns. Each row of the matrix corresponds to a sampling point, and the three columns correspond to the first, second, and third derivative values ​​at that sampling point, respectively.

[0008] Step S2: Perform curvature feature decomposition on the multi-order differential matrix to extract the geometric invariant vectors representing the degradation mode of the charging process; map the geometric invariant vectors of all historical charging processes to the degenerate manifold space to obtain the projected coordinates of each charging process. Step S2 specifically includes: Each row of the multi-order differential matrix is ​​regarded as a point in three-dimensional space. The covariance matrix is ​​calculated for the point set formed by all rows. The covariance matrix is ​​decomposed into eigenvalues. The eigenvector corresponding to the largest eigenvalue is taken as the first principal curvature direction, and the eigenvector corresponding to the second largest eigenvalue is taken as the second principal curvature direction. The ratio of the largest eigenvalue to the mean of the variance of the three-dimensional point set in each coordinate direction is taken as the first principal curvature value, the ratio of the second largest eigenvalue to the mean is taken as the second principal curvature value, and the absolute value of the cosine of the angle between the first principal curvature direction and the coordinate axis corresponding to the first column of the first derivative in the multi-order differential matrix is ​​taken as the directional coupling factor. The first principal curvature value, the second principal curvature value, and the directional coupling factor are combined into a geometric invariant vector; An isometric mapping algorithm is used to reduce all geometric invariant vectors to a two-dimensional plane, thereby obtaining the projected coordinates of each charging process in the degenerate manifold space.

[0009] Step S3: Calculate the manifold distance between time-adjacent projected coordinates in the degenerate manifold space, and detect abrupt changes in the manifold distance as degradation transition points; divide the entire charging process into multiple degradation stages with the degradation transition points as boundaries, and determine the current degradation stage of the battery. Step S3 specifically includes: In the degenerate manifold space, the manifold distance between each pair of adjacent projection points in the charging order is calculated to obtain a manifold distance sequence. The manifold distance is approximated by accumulating the Euclidean distance between sampling points on the connection line. A sliding window is used to scan the manifold distance sequence, and the median value and the median absolute deviation of the distance values ​​within each window are calculated. Points that exceed the median value plus three times the median absolute deviation of the window are marked as degenerate transition points. The entire charging process is divided into multiple degradation stages, with the degradation transition point as the boundary. Each degradation stage contains several consecutive charging processes. The mean of the projected coordinates within each degradation stage is calculated as the center coordinate of the stage. The projected coordinates of the current charging process are compared with the center coordinates of each stage, and the stage with the smallest Euclidean distance is taken as the current degradation stage of the battery.

[0010] Step S4: Calculate the magnitude and direction deviation of the geometric invariant vector of the current charging process, and obtain the degradation degree score by combining it with the statistical baseline of the current degradation stage. Step S4 specifically includes: Calculate the sum of squares of each component of the geometrically invariant vector, and then take the arithmetic square root of the sum of squares as the magnitude of the geometrically invariant vector; Collect the geometric invariant vectors of all historical charging processes within the current degradation stage, calculate the mean vector of all vectors, and normalize the mean vector to use it as the reference direction vector. Normalize the geometric invariant vector of the current charging process, calculate the dot product of the normalized vector and the reference direction vector, and obtain the cosine value of the direction deviation. The median value of the magnitude of all geometric invariant vectors within the degradation stage is taken as the magnitude baseline, and the median value of the direction deviation of all geometric invariant vectors from the cosine value is taken as the cosine baseline. Divide the absolute value of the difference between the current module length and the module length baseline by the module length baseline to obtain the module length deviation rate; divide the absolute value of the difference between the current direction deviation cosine value and the cosine baseline by the cosine baseline to obtain the direction deviation rate. The degradation score is obtained by taking a weighted harmonic average of the modulus deviation rate and the orientation deviation rate, multiplying it by 100 and limiting it to the range of 0 to 100.

[0011] Step S5: Based on the degradation stage and degree of degradation, match and adjust the basic waveform parameters from the preset waveform template library to generate a charging current waveform parameter set; Step S5 specifically includes: Based on the current degradation stage, a set of basic waveform parameters matching the current degradation stage is selected from the preset waveform template library. The set of basic waveform parameters includes positive current amplitude, positive duration, negative current amplitude, negative duration, and rest interval duration. The amplitude attenuation factor and the rest extension factor are calculated based on the degradation degree score. The higher the degradation degree score, the smaller the amplitude attenuation factor and the larger the rest extension factor. Multiply the positive and negative current amplitudes in the basic waveform parameter set by the amplitude attenuation factor, and multiply the rest interval duration by the rest extension factor to obtain the charging current waveform parameter set.

[0012] Step S6: Generate a charging current waveform based on the charging current waveform parameter set, apply safety boundary constraints to the charging current waveform, and charge the battery until the state of charge reaches the target value.

[0013] Step S6 specifically includes: A positive charging pulse segment is generated based on the positive current amplitude and positive duration; a negative discharging pulse segment is generated based on the negative current amplitude and negative duration; a no-current rest segment is generated based on the rest interval duration; and positive and negative alternating rest charging current waveforms are generated in sequence and cycle. During the charging process, the terminal voltage and surface temperature are collected in real time. When the terminal voltage exceeds the preset voltage safety limit, the amplitude of the subsequent positive current is compressed proportionally. When the surface temperature exceeds the preset temperature safety limit, the duration of the subsequent pause interval is extended proportionally. When the terminal voltage or surface temperature exceeds the preset limit threshold, the charging output is immediately cut off and locked. Once both drop below the safe recovery threshold, the lock is automatically released and charging resumes.

[0014] A charging system for a new energy vehicle, comprising a charging method for a new energy vehicle, including: Matrix construction module: Obtains terminal voltage time-series data and charging capacity time-series data of the target vehicle battery in multiple historical charging processes, and constructs a multi-order differential matrix based on the terminal voltage time-series data and charging capacity time-series data for each charging process; Eigenvalue decomposition module: Performs curvature eigenvalue decomposition on multi-order differential matrices to extract geometric invariant vectors representing the degradation modes of the charging process; maps the geometric invariant vectors of all historical charging processes to the degradation manifold space; Detection module: Calculates the manifold distance between temporally adjacent projected coordinates in the degenerate manifold space, detects abrupt changes in manifold distance as degradation transition points; divides the entire charging process into multiple degradation stages based on degradation transition points, and determines the current degradation stage of the battery. Scoring module: Calculates the magnitude and direction deviation of the geometric invariant vector of the current charging process, and obtains a degradation degree score by combining it with the statistical baseline of the current degradation stage; Parameter set acquisition module: Based on the degradation stage and degradation degree score, it matches and adjusts the basic waveform parameters from the preset waveform template library to generate a charging current waveform parameter set; Execution module: Generates a charging current waveform based on the charging current waveform parameter set, applies safety boundary constraints to the charging current waveform, and charges the battery until the state of charge reaches the target value.

[0015] An electronic device includes a processor, a memory controller, and a memory, wherein the memory stores computer-readable instructions that, when executed by the processor, implement a charging method for a new energy vehicle.

[0016] A computer storage medium storing computer-readable instructions, which, when executed by a processor, implement a charging method for a new energy vehicle.

[0017] In summary, due to the adoption of the above technical solution, the beneficial effects of the present invention are: 1. This invention constructs a multi-order differential matrix of terminal voltage versus charging capacity, enabling the extraction of curvature characteristics of the changes in the internal electrochemical state of the battery during charging at the higher-order derivative level. These characteristics are directly related to the phase transition process of the battery's active materials and reflect the essence of degradation more effectively than simply using terminal voltage or capacity data. After mapping the extracted geometric invariant vectors to the degradation manifold space, the degradation pattern of each charging process is transformed into a point in the space, making the degradation trajectory of the entire battery lifecycle visible. Different degradation stages naturally appear as clusters of points in different regions of the space.

[0018] 2. This invention uses manifold distance to automatically detect degradation transition points and divide the battery into stages, eliminating the need for manually setting cycle number thresholds. The resulting stage division more closely reflects the actual physical degradation process of the battery. It quantifies the degradation score by combining the deviations in both vector magnitude and direction, considering both changes in degradation amplitude and structural shifts in degradation patterns, resulting in a more comprehensive score. Finally, based on the degradation stage and score, the charging waveform parameters are dynamically adjusted to match the magnitude and interval of the charging current with the battery's current tolerance, thus delaying aging while maintaining charging efficiency. Attached Figure Description

[0019] The present invention will be described by way of example and with reference to the accompanying drawings, wherein: Figure 1 This is a schematic diagram of the process structure of the present invention; Figure 2 This is a schematic diagram of the process structure for step 3 of the present invention; Figure 3 This is a schematic diagram of the process structure for step 4 of the present invention. Detailed Implementation

[0020] To make the objectives, technical solutions, and advantages of the embodiments of the present invention clearer, the technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. The components of the embodiments of the present invention described and shown in the accompanying drawings can generally be arranged and designed in various different configurations.

[0021] Therefore, the following detailed description of the embodiments of the invention provided in the accompanying drawings is not intended to limit the scope of the claimed invention, but merely to illustrate selected embodiments of the invention. All other embodiments obtained by those skilled in the art based on the embodiments of the invention without inventive effort are within the scope of protection of the invention.

[0022] It should be noted that, unless otherwise specified, the embodiments and features described in this invention can be combined with each other.

[0023] It should be noted that similar labels and letters in the following figures indicate similar items. Therefore, once an item is defined in one figure, it does not need to be further defined and explained in subsequent figures.

[0024] In this invention, unless otherwise explicitly specified and limited, "above" or "below" the second feature can include direct contact between the first and second features, or contact between the first and second features through another feature between them. Furthermore, "above," "over," and "on top" of the second feature includes the first feature directly above or diagonally above the second feature, or simply indicates that the first feature is at a higher horizontal level than the second feature. "Below," "below," and "under" the second feature includes the first feature directly below or diagonally below the second feature, or simply indicates that the first feature is at a lower horizontal level than the second feature.

[0025] It should be noted that, unless otherwise specified, the embodiments and features described in this invention can be combined with each other.

[0026] Example 1

[0027] like Figures 1-3 As shown in the figure, an embodiment of the present invention discloses a charging method for new energy vehicles, including: Step S1: Obtain the terminal voltage time-series data and charging capacity time-series data of the target vehicle battery during multiple historical charging processes. For each charging process, construct a multi-order differential matrix based on the terminal voltage time-series data and charging capacity time-series data. In practical applications, the battery management system synchronously records the battery's terminal voltage at a fixed sampling period (e.g., 1 second) each time the vehicle is charged. and charging capacity Charging capacity Generally determined by charging current Obtained by integrating over time, i.e.

[0028] Alternatively, it can be obtained directly from the accumulated capacity value provided by the battery management system. This yields a sequence of terminal voltage changes over time. and the sequence of charging capacity changes over time At the same time One-to-one correspondence. To uncover subtle characteristics of the changes in the battery's internal electrochemical state during charging, this step does not directly use the original... - Instead of focusing on the curve itself, we construct a multi-order differential matrix to capture the higher-order curvature information of the curve.

[0029] In step S1, the construction of the multi-order differential matrix based on the terminal voltage time-series data and the charging capacity time-series data specifically includes: Using the charging capacity time series data of a single charging process as the independent variable and the terminal voltage time series data as the dependent variable, the first, second, and third derivative sequences of the terminal voltage with respect to the charging capacity are calculated. The first, second, and third derivative sequences are concatenated column by column to form a multi-order differential matrix with three columns. Each row of the matrix corresponds to a sampling point, and the three columns correspond to the first, second, and third derivative values ​​at that sampling point, respectively.

[0030] Before calculating the derivative, we can first examine the original... and The sequence is smoothed using a Gaussian filter to suppress measurement noise. Let the smoothed sequence be... and ,in The sampling point number, This represents the total number of sampling points. This is used to calculate... right The derivative is obtained using numerical differentiation. For points inside the sequence... ( Using the five-point central difference formula:

[0031]

[0032] in, , where is the capacity interval between adjacent sampling points. In actual calculations, the average of several intervals near this point can be taken to improve stability. Third derivative Then, based on the already obtained second derivative sequence, the central difference approximation is applied again, or higher-order central difference coefficients are directly used. For the sequence boundaries ( and ), and instead use the forward or backward difference formula, for example for The first derivative is obtained using the two-point forward difference:

[0033] Second-order and higher derivatives are derived recursively in a similar manner or using a one-sided difference scheme.

[0034] This results in three lengths. ( The derivative sequence of ) is denoted as . , , Concatenate them column-wise into a matrix:

[0035] Where the matrix The The row corresponds to the first There are 3 valid sampling points, and the elements in the three columns are the first derivative values ​​for that point. Second derivative value and third derivative value To eliminate the impact of the order-of-magnitude difference between different derivatives on subsequent geometric analysis, the following steps are taken: Perform maximum-minimum normalization column by column:

[0036] in, and They are the first The minimum and maximum values ​​of each column. After normalization, the values ​​in each column of the matrix are all within the range of... Within the range.

[0037] Step S2: Perform curvature feature decomposition on the multi-order differential matrix to extract the geometric invariant vectors representing the degradation mode of the charging process; map the geometric invariant vectors of all historical charging processes to the degenerate manifold space to obtain the projected coordinates of each charging process. Step S2 specifically includes: Each row of the multi-order differential matrix is ​​regarded as a point in three-dimensional space. The covariance matrix is ​​calculated for the point set formed by all rows. The covariance matrix is ​​decomposed into eigenvalues. The eigenvector corresponding to the largest eigenvalue is taken as the first principal curvature direction, and the eigenvector corresponding to the second largest eigenvalue is taken as the second principal curvature direction. The ratio of the largest eigenvalue to the mean of the variance of the three-dimensional point set in each coordinate direction is taken as the first principal curvature value, the ratio of the second largest eigenvalue to the mean is taken as the second principal curvature value, and the absolute value of the cosine of the angle between the first principal curvature direction and the coordinate axis corresponding to the first column of the first derivative in the multi-order differential matrix is ​​taken as the directional coupling factor. The first principal curvature value, the second principal curvature value, and the directional coupling factor are combined into a geometric invariant vector; An isometric mapping algorithm is used to reduce all geometric invariant vectors to a two-dimensional plane, thereby obtaining the projected coordinates of each charging process in the degenerate manifold space.

[0038] For the normalized multi-order differential matrix obtained during a single charging process Each row can be considered as a point in three-dimensional space. ,in These correspond to the normalized first, second, and third derivative values, respectively. (All) These points constitute a three-dimensional point set. To describe the curvature characteristics of the curve exhibited by this point set, its covariance matrix is ​​calculated:

[0039] in, Let be the mean vector of the point set. Eigenvalue decomposition yields three eigenvalues. and the corresponding unit eigenvector In differential geometry, the principal direction of the data point distribution reflects the local principal direction of the curve, and the largest eigenvalue... Corresponding feature vector It can be regarded as the first principal curvature direction. corresponding This is the direction of the second principal curvature.

[0040] To quantify the magnitude of the principal curvature, the mean variance of the point set in each coordinate direction is calculated:

[0041] in For the first The variance of the normalized derivative sequence is given. Therefore, the first principal curvature value is defined. Second principal curvature value .

[0042] and It reflects the degree of concentration of the point set along the main direction, and is closely related to the curvature of the charging voltage curve.

[0043] Simultaneously, to demonstrate the relationship between the principal curvature direction and the voltage change direction (i.e., the first derivative axis), the vector of the first principal curvature direction is calculated. unit vectors of the first derivative coordinate axes The absolute value of the cosine of the included angle is used as the directional coupling factor:

[0044] in for The first component. The closer to 1, the closer the principal curvature direction is to the first derivative axis; the closer to 0, the further away from the axis, which may correspond to different degradation modes.

[0045] The above three eigenvalues ​​are combined into a three-dimensional geometric invariant vector:

[0046] vector It is a compact representation of the shape of the differential curve in this charging process, and has rotation invariance (because the principal direction is based on the data distribution itself) and scaling invariance (the data range is fixed after normalization).

[0047] For all The above operation is repeated in each historical charging process to obtain a set of geometrically invariant vectors. To reveal the intrinsic manifold structure among these high-dimensional vectors, an isomap algorithm is used for dimensionality reduction. The specific process is as follows: calculate all... Construct a distance matrix using the Euclidean distance between them; determine the number of nearest neighbors. (usually taken) (5%~10%), for each Choose the closest one We construct a nearest neighbor graph by using each vector as its nearest neighbor, and treat non-nearest neighbor distances as infinite. We then use Dijkstra's algorithm to calculate the shortest path distance between all pairs of points in the graph, forming a geodesic distance matrix. ;right By applying multidimensional scaling (MDS), i.e., solving for the eigenvectors corresponding to the first two largest eigenvalues ​​after doubly centering, we obtain the two-dimensional projected coordinates:

[0048] In this way, each charging process is mapped to a point on the degenerate manifold plane, and the degenerate trajectory can be visualized on the plane.

[0049] Step S3: Calculate the manifold distance between time-adjacent projected coordinates in the degenerate manifold space, and detect abrupt changes in the manifold distance as degradation transition points; divide the entire charging process into multiple degradation stages with the degradation transition points as boundaries, and determine the current degradation stage of the battery. Step S3 specifically includes: In the degenerate manifold space, the manifold distance between each pair of adjacent projection points in the charging order is calculated to obtain a manifold distance sequence. The manifold distance is approximated by accumulating the Euclidean distance between sampling points on the connection line. A sliding window is used to scan the manifold distance sequence, and the median value and the median absolute deviation of the distance values ​​within each window are calculated. Points that exceed the median value plus three times the median absolute deviation of the window are marked as degenerate transition points. The entire charging process is divided into multiple degradation stages, with the degradation transition point as the boundary. Each degradation stage contains several consecutive charging processes. The mean of the projected coordinates within each degradation stage is calculated as the center coordinate of the stage. The projected coordinates of the current charging process are compared with the center coordinates of each stage, and the stage with the smallest Euclidean distance is taken as the current degradation stage of the battery.

[0050] After obtaining the two-dimensional projected coordinate sequence arranged in order of charging time. Then, the magnitude of the degradation mode change between adjacent charging processes is quantified. For the first... Second and third The projection points of the second charge are respectively and Because the manifold may be curved, the direct Euclidean distance may underestimate the true manifold distance. Here, a uniform insertion is made along the line connecting the two projection points. intermediate points (e.g.) ), calculate the Euclidean distance between adjacent points and sum them to obtain an approximate manifold distance:

[0051] when When sufficiently large, this summation approaches the Euclidean distance. In fact, for a two-dimensional plane, the manifold distance can be simplified to the Euclidean distance, but retaining this description covers a more general case. All neighboring points form a sequence of manifold distances: .

[0052] To automatically detect degradation transition points, the median absolute deviation (MAD) method using a sliding window is employed. The window width is set. (For example ), for the first in the sequence Distance values ​​( from arrive ), take Center, width Calculate the median value based on the neighborhood distance values. and median of absolute deviation :

[0053] If the conditions are met Then the first The point (i.e., the th point) Next to The transition point during the second charge is marked as a candidate degradation transition point. This criterion is based on robust statistics and is well-suited to non-normally distributed data.

[0054] To avoid multiple neighboring candidate points due to local fluctuations, adjacent candidate points with an interval of less than 3 charging cycles are merged into a group. Within each group, only the point with the largest manifold distance is retained as the final degradation transition point. These transition points divide the entire historical charging process into... A series of degradation stages .

[0055] For the latest charging process, its projected coordinates are: Calculate each stage The mean of all projected coordinates within the stage is used as the stage center:

[0056] Compare With each The Euclidean distance is used to find the stage with the smallest distance. This indicates the current stage of degradation of the battery.

[0057] Step S4: Calculate the magnitude and direction deviation of the geometric invariant vector of the current charging process, and obtain the degradation degree score by combining it with the statistical baseline of the current degradation stage. Step S4 specifically includes: Calculate the sum of squares of each component of the geometrically invariant vector, and then take the arithmetic square root of the sum of squares as the magnitude of the geometrically invariant vector; Collect the geometric invariant vectors of all historical charging processes within the current degradation stage, calculate the mean vector of all vectors, and normalize the mean vector to use it as the reference direction vector. Normalize the geometric invariant vector of the current charging process, calculate the dot product of the normalized vector and the reference direction vector, and obtain the cosine value of the direction deviation. The median value of the magnitude of all geometric invariant vectors within the degradation stage is taken as the magnitude baseline, and the median value of the direction deviation of all geometric invariant vectors from the cosine value is taken as the cosine baseline. Divide the absolute value of the difference between the current module length and the module length baseline by the module length baseline to obtain the module length deviation rate; divide the absolute value of the difference between the current direction deviation cosine value and the cosine baseline by the cosine baseline to obtain the direction deviation rate. The degradation score is obtained by taking a weighted harmonic average of the modulus deviation rate and the orientation deviation rate, multiplying it by 100 and limiting it to the range of 0 to 100.

[0058] Determine the current stage of the charging process. Subsequently, it is necessary to further quantify the degree of battery degradation during this stage. Let the current geometric invariant vector be... First, calculate the modulus:

[0059] Length This value comprehensively reflects the higher-order curvature and tilt of the charging curve. Typically, as the battery ages and electrochemical inhomogeneity intensifies, this value tends to increase.

[0060] Collection phase Geometric invariant vector of all historical charging processes Calculate their mean vector:

[0061] Will Normalization yields the reference direction vector. .Will After normalization, calculate the cosine similarity with the reference direction:

[0062] The closer a value is to 1, the more consistent the current degradation pattern is with the typical pattern within the stage; a significant deviation from 1 may indicate a new degradation trend.

[0063] The median value of the magnitude of all vectors within the stage As the baseline for the modulus, the median value of the cosine values ​​in all directions. As the cosine baseline ( For each historical vector and (cosine of the mean). The median value is used instead of the mean value to resist the effects of abnormal charging processes.

[0064] Calculate the modulus deviation rate:

[0065] Directional deviation rate:

[0066] These two deviation rates reflect the difference between the current charging process and the normal state of the stage, respectively, from the dimensions of magnitude and direction. To synthesize a single score, a weighted harmonic mean is used, as the harmonic mean is more sensitive to larger deviations and avoids overlooking any single indicator. Weights and satisfy This can be determined through offline aging experiments: the Pearson correlation coefficient between the modulus deviation rate and the battery capacity degradation rate is calculated separately. And the correlation coefficient between directional deviation rate and decay rate. Take the normalized absolute value as the weight, for example:

[0067] The weighted harmonic mean is calculated as follows (if a certain deviation rate is 0, then that term is ignored):

[0068] The final degradation score is linearly mapped to Interval:

[0069] The higher the score, the more severe the battery degradation.

[0070] Step S5: Based on the degradation stage and degree of degradation, match and adjust the basic waveform parameters from the preset waveform template library to generate a charging current waveform parameter set; Step S5 specifically includes: Based on the current degradation stage, a set of basic waveform parameters matching the current degradation stage is selected from the preset waveform template library. The set of basic waveform parameters includes positive current amplitude, positive duration, negative current amplitude, negative duration, and rest interval duration. The amplitude attenuation factor and the rest extension factor are calculated based on the degradation degree score. The higher the degradation degree score, the smaller the amplitude attenuation factor and the larger the rest extension factor. Multiply the positive and negative current amplitudes in the basic waveform parameter set by the amplitude attenuation factor, and multiply the rest interval duration by the rest extension factor to obtain the charging current waveform parameter set.

[0071] The preset waveform template library was built through extensive offline battery cycle aging tests. During the tests, for batteries at different degradation stages (corresponding to the stages defined by this method), different alternating positive and negative pulse charging waveform parameters were tested. Performance indicators such as charging time, temperature rise, and capacity degradation rate were recorded. The optimal basic waveform parameters for each stage were obtained through multi-objective optimization (e.g., minimizing charging time and minimizing capacity degradation rate). Each basic waveform parameter set includes: forward charging current amplitude. Positive duration negative discharge current amplitude Negative duration Rest interval duration These parameters are categorized by degradation stage. Store them separately.

[0072] During online charging, the current degradation stage is determined in step S3. The corresponding basic waveform parameter set is retrieved from the template library.

[0073] Then, the degree of degradation is used for scoring. Calculate two adjustment factors: amplitude attenuation factor and amplitude decay factor. Used to reduce the current intensity of charging pulses to avoid excessive stress on aging batteries; rest extension factor These are used to increase rest time, thus slowing temperature rise and polarization accumulation. They are derived from score mappings, such as linear mappings:

[0074] in, This is the amplitude attenuation intensity coefficient (e.g., 0.5), indicating that the amplitude can be attenuated by up to 50% when the score reaches 100; The rest duration extension factor (e.g., 1.0) indicates that the rest duration can be extended by up to 100% when the score is 100. These factors are calibrated through battery durability testing to ensure that the battery can still be safely charged in the later stages of degradation.

[0075] The adjusted charging current waveform parameter set is as follows: Forward current amplitude ; negative current amplitude ; The positive and negative durations remain unchanged from their base values ​​(or can be finely adjusted). rest interval duration ; These parameters constitute the set of parameters for the final output charging current waveform.

[0076] Step S6: Generate a charging current waveform based on the charging current waveform parameter set, apply safety boundary constraints to the charging current waveform, and charge the battery until the state of charge reaches the target value.

[0077] Step S6 specifically includes: A positive charging pulse segment is generated based on the positive current amplitude and positive duration; a negative discharging pulse segment is generated based on the negative current amplitude and negative duration; a no-current rest segment is generated based on the rest interval duration; and positive and negative alternating rest charging current waveforms are generated in sequence and cycle. During the charging process, the terminal voltage and surface temperature are collected in real time. When the terminal voltage exceeds the preset voltage safety limit, the amplitude of the subsequent positive current is compressed proportionally. When the surface temperature exceeds the preset temperature safety limit, the duration of the subsequent pause interval is extended proportionally. When the terminal voltage or surface temperature exceeds the preset limit threshold, the charging output is immediately cut off and locked. Once both drop below the safe recovery threshold, the lock is automatically released and charging resumes.

[0078] During charging, the current is output cyclically in the sequence of "positive pulse - negative pulse - rest". The positive pulse is characterized by its amplitude... Duration Output charging current; followed by a negative pulse with amplitude Duration Output discharge current (current direction opposite); then enter a zero-current rest state, which continues. Time. A complete cycle. This cycle repeats until the battery state of charge (SOC) reaches the target value (e.g., 90% or 100%).

[0079] During charging, the battery management system collects the battery terminal voltage in real time. and surface temperature Set a safe voltage limit. (e.g., 4.2V single-unit voltage multiplied by the number of series), upper limit of temperature safety. (For example, 45℃). If detected Then, the current amplitude of all subsequent positive pulses is compressed proportionally: ,in The voltage reduction factor (e.g., 0.9) is used until... Falling back to The following. If Then the duration of subsequent rest intervals will be extended proportionally: , This is a temperature extension factor (e.g., 1.2) to enhance heat dissipation. When Exceeding the limit threshold (e.g., 4.3V) or Exceeding the limit threshold (e.g., at 55℃), immediately disconnect the charging circuit and lock it. The system will continuously monitor until both return to the safe recovery threshold (e.g., and Below this value, the lock will automatically unlock, and charging will resume at a reduced amplitude. This ensures the safety of the charging process without completely stopping charging due to a brief over-limit.

[0080] Example 2

[0081] This embodiment proposes a charging system for new energy vehicles based on Embodiment 1, to implement a charging method for new energy vehicles, including: Matrix construction module: Obtains terminal voltage time-series data and charging capacity time-series data of the target vehicle battery in multiple historical charging processes, and constructs a multi-order differential matrix based on the terminal voltage time-series data and charging capacity time-series data for each charging process; Eigenvalue decomposition module: Performs curvature eigenvalue decomposition on multi-order differential matrices to extract geometric invariant vectors representing the degradation modes of the charging process; maps the geometric invariant vectors of all historical charging processes to the degradation manifold space; Detection module: Calculates the manifold distance between temporally adjacent projected coordinates in the degenerate manifold space, detects abrupt changes in manifold distance as degradation transition points; divides the entire charging process into multiple degradation stages based on degradation transition points, and determines the current degradation stage of the battery. Scoring module: Calculates the magnitude and direction deviation of the geometric invariant vector of the current charging process, and obtains a degradation degree score by combining it with the statistical baseline of the current degradation stage; Parameter set acquisition module: Based on the degradation stage and degradation degree score, it matches and adjusts the basic waveform parameters from the preset waveform template library to generate a charging current waveform parameter set; Execution module: Generates a charging current waveform based on the charging current waveform parameter set, applies safety boundary constraints to the charging current waveform, and charges the battery until the state of charge reaches the target value.

[0082] Example 3

[0083] This embodiment proposes an electronic device based on embodiment 1, including a processor, a memory controller, and a memory. The memory stores computer-readable instructions, which, when executed by the processor, implement a charging method for a new energy vehicle.

[0084] Example 4

[0085] This embodiment proposes a computer storage medium based on embodiment 1, which stores computer-readable instructions. When the instructions are executed by a processor, they implement a charging method for a new energy vehicle.

[0086] The circuits, electronic components, and modules involved are all existing technologies, which can be fully implemented by those skilled in the art, and need not be elaborated upon. The scope of protection of this invention does not involve any improvement to the software and methods.

[0087] The various embodiments in this specification are described in a progressive manner, with each embodiment focusing on the differences from other embodiments. The same or similar parts between the various embodiments can be referred to each other.

[0088] The above description of the disclosed embodiments enables those skilled in the art to make or use the invention. Various modifications to these embodiments will be readily apparent to those skilled in the art, and the general principles defined herein may be implemented in other embodiments without departing from the spirit or scope of the invention. Therefore, the invention is not to be limited to the embodiments shown herein, but is to be accorded the widest scope consistent with the principles and novel features disclosed herein.

Claims

1. A charging method for a new energy vehicle, characterized in that, include: Step S1: Obtain the terminal voltage time-series data and charging capacity time-series data of the target vehicle battery during multiple historical charging processes. For each charging process, construct a multi-order differential matrix based on the terminal voltage time-series data and charging capacity time-series data. Step S2: Perform curvature feature decomposition on the multi-order differential matrix to extract the geometric invariant vector representing the degradation mode of the charging process; Map the geometric invariant vectors of all historical charging processes to the degenerate manifold space to obtain the projected coordinates of each charging process; Step S3: Calculate the manifold distance between temporally adjacent projected coordinates in the degenerate manifold space, and detect abrupt changes in manifold distance as degenerate transition points; The entire charging process is divided into multiple degradation stages, with the degradation transition point as the boundary, and the current degradation stage of the battery is determined. Step S4: Calculate the magnitude and direction deviation of the geometric invariant vector of the current charging process, and obtain the degradation degree score by combining it with the statistical baseline of the current degradation stage. Step S5: Based on the degradation stage and degree of degradation, match and adjust the basic waveform parameters from the preset waveform template library to generate a charging current waveform parameter set; Step S6: Generate a charging current waveform based on the charging current waveform parameter set, apply safety boundary constraints to the charging current waveform, and charge the battery until the state of charge reaches the target value.

2. The charging method for a new energy vehicle according to claim 1, characterized in that, In step S1, the construction of the multi-order differential matrix based on the terminal voltage time-series data and the charging capacity time-series data specifically includes: Using the charging capacity time series data of a single charging process as the independent variable and the terminal voltage time series data as the dependent variable, the first, second, and third derivative sequences of the terminal voltage with respect to the charging capacity are calculated. The first, second, and third derivative sequences are concatenated column by column to form a multi-order differential matrix with three columns. Each row of the matrix corresponds to a sampling point, and the three columns correspond to the first, second, and third derivative values ​​at that sampling point, respectively.

3. The charging method for a new energy vehicle according to claim 1, characterized in that, Step S2 specifically includes: Each row of the multi-order differential matrix is ​​regarded as a point in three-dimensional space. The covariance matrix is ​​calculated for the point set formed by all rows. The covariance matrix is ​​decomposed into eigenvalues. The eigenvector corresponding to the largest eigenvalue is taken as the first principal curvature direction, and the eigenvector corresponding to the second largest eigenvalue is taken as the second principal curvature direction. The ratio of the largest eigenvalue to the mean of the variance of the three-dimensional point set in each coordinate direction is taken as the first principal curvature value, the ratio of the second largest eigenvalue to the mean is taken as the second principal curvature value, and the absolute value of the cosine of the angle between the first principal curvature direction and the coordinate axis corresponding to the first column of the first derivative in the multi-order differential matrix is ​​taken as the directional coupling factor. The first principal curvature value, the second principal curvature value, and the directional coupling factor are combined into a geometric invariant vector; An isometric mapping algorithm is used to reduce all geometric invariant vectors to a two-dimensional plane, thereby obtaining the projected coordinates of each charging process in the degenerate manifold space.

4. The charging method for a new energy vehicle according to claim 1, characterized in that, Step S3 specifically includes: In the degenerate manifold space, the manifold distance between each pair of adjacent projection points in the charging order is calculated to obtain a manifold distance sequence. The manifold distance is approximated by accumulating the Euclidean distance between sampling points on the connection line. A sliding window is used to scan the manifold distance sequence, and the median value and the median absolute deviation of the distance values ​​within each window are calculated. Points that exceed the median value plus three times the median absolute deviation of the window are marked as degenerate transition points. The entire charging process is divided into multiple degradation stages, with the degradation transition point as the boundary. Each degradation stage contains several consecutive charging processes. The mean of the projected coordinates within each degradation stage is calculated as the center coordinate of the stage. The projected coordinates of the current charging process are compared with the center coordinates of each stage, and the stage with the smallest Euclidean distance is taken as the current degradation stage of the battery.

5. A charging method for a new energy vehicle according to claim 3, characterized in that, Step S4 specifically includes: Calculate the sum of squares of each component of the geometrically invariant vector, and then take the arithmetic square root of the sum of squares as the magnitude of the geometrically invariant vector; Collect the geometric invariant vectors of all historical charging processes within the current degradation stage, calculate the mean vector of all vectors, and normalize the mean vector to use it as the reference direction vector. Normalize the geometric invariant vector of the current charging process, calculate the dot product of the normalized vector and the reference direction vector, and obtain the cosine value of the direction deviation. The median value of the magnitude of all geometric invariant vectors within the degradation stage is taken as the magnitude baseline, and the median value of the direction deviation of all geometric invariant vectors from the cosine value is taken as the cosine baseline. Divide the absolute value of the difference between the current module length and the module length baseline by the module length baseline to obtain the module length deviation rate; divide the absolute value of the difference between the current direction deviation cosine value and the cosine baseline by the cosine baseline to obtain the direction deviation rate. The degradation score is obtained by taking a weighted harmonic average of the modulus deviation rate and the orientation deviation rate, multiplying it by 100 and limiting it to the range of 0 to 100.

6. A charging method for a new energy vehicle according to claim 1, characterized in that, Step S5 specifically includes: Based on the current degradation stage, a set of basic waveform parameters matching the current degradation stage is selected from the preset waveform template library. The set of basic waveform parameters includes positive current amplitude, positive duration, negative current amplitude, negative duration, and rest interval duration. The amplitude attenuation factor and the rest extension factor are calculated based on the degradation degree score. The higher the degradation degree score, the smaller the amplitude attenuation factor and the larger the rest extension factor. Multiply the positive and negative current amplitudes in the basic waveform parameter set by the amplitude attenuation factor, and multiply the rest interval duration by the rest extension factor to obtain the charging current waveform parameter set.

7. A charging method for a new energy vehicle according to claim 1, characterized in that, Step S6 specifically includes: A positive charging pulse segment is generated based on the positive current amplitude and positive duration; a negative discharging pulse segment is generated based on the negative current amplitude and negative duration; a no-current rest segment is generated based on the rest interval duration; and positive and negative alternating rest charging current waveforms are generated in sequence and cycle. During the charging process, the terminal voltage and surface temperature are collected in real time. When the terminal voltage exceeds the preset voltage safety limit, the amplitude of the subsequent positive current is compressed proportionally. When the surface temperature exceeds the preset temperature safety limit, the duration of the subsequent pause interval is extended proportionally. When the terminal voltage or surface temperature exceeds the preset limit threshold, the charging output is immediately cut off and locked. Once both drop below the safe recovery threshold, the lock is automatically released and charging resumes.

8. A charging system for a new energy vehicle, characterized in that, A charging method for a new energy vehicle according to any one of claims 1-7, comprising: Matrix construction module: Obtains terminal voltage time-series data and charging capacity time-series data of the target vehicle battery in multiple historical charging processes, and constructs a multi-order differential matrix based on the terminal voltage time-series data and charging capacity time-series data for each charging process; Eigenvalue decomposition module: Performs curvature eigenvalue decomposition on multi-order differential matrices to extract geometric invariant vectors representing the degradation modes of the charging process; maps the geometric invariant vectors of all historical charging processes to the degradation manifold space; Detection module: Calculates the manifold distance between temporally adjacent projected coordinates in the degenerate manifold space, detects abrupt changes in manifold distance as degradation transition points; divides the entire charging process into multiple degradation stages based on degradation transition points, and determines the current degradation stage of the battery. Scoring module: Calculates the magnitude and direction deviation of the geometric invariant vector of the current charging process, and obtains a degradation degree score by combining it with the statistical baseline of the current degradation stage; Parameter set acquisition module: Based on the degradation stage and degradation degree score, it matches and adjusts the basic waveform parameters from the preset waveform template library to generate a charging current waveform parameter set; Execution module: Generates a charging current waveform based on the charging current waveform parameter set, applies safety boundary constraints to the charging current waveform, and charges the battery until the state of charge reaches the target value.

9. An electronic device, characterized in that, It includes a processor, a memory controller, and a memory, wherein the memory stores computer-readable instructions that, when executed by the processor, implement a charging method for a new energy vehicle according to any one of claims 1 to 7.

10. A computer storage medium storing computer-readable instructions thereon, characterized in that, When the instructions are executed by the processor, they implement a charging method for a new energy vehicle as described in any one of claims 1 to 7.

Images