A method for estimating the state of health of lithium-ion batteries based on open-circuit voltage curve reconstruction and a computer-readable storage medium.

Abstract

The present application relates to the technical field of battery, especially to a lithium ion battery health state estimation method based on open circuit voltage curve reconstruction and computer readable storage medium, comprising the following steps: extracting the first feature point related to the template lithium battery aging from the incremental capacity curve, and finding the second feature point corresponding to the first feature point from the open circuit voltage curve; obtaining the charging curve of the template lithium battery, dividing the charging curve into data segments of fixed length, and then constructing the open circuit voltage reconstruction model of the template lithium battery; optimizing the parameters of the open circuit voltage reconstruction model; obtaining the charging data segment of the target lithium battery, inputting the charging data segment into the open circuit voltage reconstruction model to obtain the aging feature point corresponding to the target lithium battery, and reconstructing the open circuit voltage capacity curve combined with the aging feature point to complete the health state estimation of the target lithium battery. The present application realizes accurate estimation of the health state of lithium battery.

Description

Technical Field

[0001] This invention relates to the field of battery technology, and in particular to a method for estimating the state of health of lithium-ion batteries based on open-circuit voltage curve reconstruction and a computer-readable storage medium. Background Technology

[0002] Lithium-ion batteries, as a key component of modern energy technology, offer strong support for applications in many fields, such as green transportation (electric vehicles), energy storage (electric ships and smart homes), communication equipment (wireless devices), and the aerospace industry, thanks to their fast charging and significant lifespan. However, with frequent charging and discharging, their electrochemical performance gradually degrades, potentially leading to increased internal resistance, reduced range, and even safety risks. Therefore, accurate capacity prediction of lithium-ion batteries is crucial. This technology not only helps optimize battery management systems but also identifies severe capacity degradation in advance through real-time monitoring. The early warning mechanism of the battery management system can effectively prevent emergency operations before critical performance is lost, thus ensuring the stable operation of the energy storage system and user safety. Through regular capacity maintenance or timely replacement, we can extend battery life, reduce operating costs, and minimize safety hazards caused by battery failure.

[0003] Battery capacity prediction, as one of the core tasks of battery management systems (BMS), has always been extensively studied. Currently, model-based methods, such as electrochemical models, use the internal chemical reaction principles of the battery to calculate and predict capacity. This method has a good theoretical foundation, but in practical applications, its effectiveness is significantly affected by external factors such as current characteristics, ambient temperature, and material properties, thus limiting the accuracy and robustness of the model. Summary of the Invention

[0004] This invention aims to at least solve one of the technical problems existing in related technologies. To this end, this invention provides a method for estimating the state of health of lithium-ion batteries based on open-circuit voltage curve reconstruction. By reconstructing the open-circuit voltage curve using partial charging segment data of the lithium battery, a rapid estimation of the lithium battery's state of health is achieved, improving the accuracy and robustness of the prediction.

[0005] This invention provides a method for estimating the state of health of lithium-ion batteries based on open-circuit voltage curve reconstruction, comprising the following steps:

[0006] S1, perform open-circuit battery testing on the template lithium battery to obtain the open-circuit voltage curve, and then calculate the incremental capacity curve of the template lithium battery through the open-circuit voltage curve.

[0007] S2, extract the first feature point related to the aging of the template lithium battery from the incremental capacity curve, and find the second feature point corresponding to the first feature point from the open circuit voltage curve;

[0008] S3, obtain the first feature points and the second feature points of a set amount of template lithium battery as a training dataset, obtain the charging curve of the template lithium battery, divide the charging curve into data segments of fixed length, and then construct the open circuit voltage reconstruction model of the template lithium battery.

[0009] S4, optimize the parameters of the open-circuit voltage reconstruction model, and determine the mapping relationship between the data segment and the aging feature points reflecting the health status of the template lithium battery;

[0010] S5. Obtain a charging data segment of the target lithium battery, input the charging data segment into the open circuit voltage reconstruction model to obtain aging characteristic points corresponding to the target lithium battery, and reconstruct the open circuit voltage capacity curve by combining the aging characteristic points to complete the health status estimation of the target lithium battery.

[0011] A further improvement of the lithium-ion battery health state estimation method based on open-circuit voltage curve reconstruction in this invention is that step S2 specifically involves:

[0012] Extract the first coordinates corresponding to the two troughs and one peak in the incremental capacity curve, define the first coordinates as the first feature points, and then find the voltage value and charging capacity value corresponding to the open circuit voltage curve based on the first coordinates, and define the voltage value and the charging capacity value as the second feature points.

[0013] A further improvement of the lithium-ion battery health state estimation method based on open-circuit voltage curve reconstruction in this invention is that dividing the charging curve into fixed-length data segments in step S3 includes the following steps:

[0014] Define the voltage range corresponding to the fixed length. and sampling step size The sampling set is obtained. ,in, ,

[0015] The charge amount corresponding to the data in the sample set was calculated by combining the ampere-hour integration method.

[0016] ,

[0017] The input vector is obtained based on several charge quantities. Then, the open-circuit voltage reconstruction model is constructed based on the input vector;

[0018] in, Indicates the starting voltage. Indicates the termination voltage. Indicates the number of sampling points. This indicates the charging current of the template lithium battery. An expression representing the time corresponding to the charging voltage. Represents the sequence of sampling points. Indicates the first The amount of charge corresponding to each sampling point.

[0019] A further improvement of the lithium-ion battery health state estimation method based on open-circuit voltage curve reconstruction in this invention is that the optimization of the open-circuit voltage reconstruction model in step S4 includes:

[0020] S41, Calculate the gradient during the iteration process. First-order moment estimation Second-order moment estimation ,

[0021]

[0022] in, Indicates the first attenuation rate. Indicates the second attenuation rate. This represents the gradient of the objective function being optimized. Indicates the corresponding time step The parameter values ​​evaluated during the iteration at each time step. Indicates time step First-order moment estimation at time 10:00 Indicates time step The second moment estimate at time t;

[0023] S42, Calculate the first-order moment estimate after bias correction. Second-order moment estimation after bias correction ;

[0024]

[0025] S43, Update the parameters of the open-circuit voltage reconstruction model. ,

[0026]

[0027] in, This represents the initial learning rate. This indicates the setting parameters. Indicates the corresponding time step The parameter values ​​evaluated during the iteration at each time step.

[0028] A further improvement of the lithium-ion battery health state estimation method based on open-circuit voltage curve reconstruction in this invention is that step S5 combines the aging characteristic points to reconstruct the open-circuit voltage-capacity curve, thereby completing the health state estimation of the target lithium battery. The specific steps are as follows:

[0029] S51, the open-circuit voltage reconstruction model includes an initial reference dataset with aging feature points. The initial reference dataset includes L open-circuit voltages and L charging capacity values. The L open-circuit voltages and L charging capacity values ​​are arranged in chronological order to obtain the reference dataset. ;

[0030] S52, Match the aging feature points in the reference dataset with the aging feature points of the target lithium battery, and select the voltage offset coefficient. Capacity offset coefficient Curve scaling factor and construction Let be the loss coefficient. Combining the particle swarm optimization algorithm with the goal of minimizing the loss coefficient, we obtain the optimal matching coefficient.

[0031] S53, combining the optimal matching coefficient, transforms the aging feature points of the reference dataset to obtain the open-circuit voltage-capacity curve of the target lithium battery.

[0032]

[0033] in, This represents the charge amount at the aging characteristic point of the target lithium battery corresponding to the i-th sampling point. express The corresponding open-circuit voltage, This represents the open-circuit voltage of the aging characteristic point of the target lithium battery corresponding to the i-th sampling point in the reference dataset;

[0034] S54, the capacity of the open-circuit voltage-capacity curve corresponding to the upper limit cutoff voltage of the target lithium battery is the estimated capacity of the target lithium battery, combined with... The health status of the target lithium battery can then be obtained, among which, Indicates the estimated capacity. Indicates the rated capacity. This indicates the health status of the target lithium battery.

[0035] A further improvement of the lithium-ion battery health state estimation method based on open-circuit voltage curve reconstruction in this invention lies in the fact that the loss coefficient is expressed as... .

[0036] A further improvement of the lithium-ion battery health state estimation method based on open-circuit voltage curve reconstruction in this invention is that step S1 includes:

[0037] S11: Discharge the template lithium battery to the cutoff voltage at a rate of 0.5C;

[0038] S12: After the template lithium battery has been left to stand for 1 hour, it is charged in constant current and constant voltage mode at a rate of 0.5C. When the charging current drops to the set value, the complete charging capacity is recorded as the template lithium battery capacity.

[0039] S13: After the template lithium battery is left to stand for 15 minutes, it is discharged at a constant current rate of 0.05C until it is discharged to the cutoff voltage, thereby obtaining the open circuit voltage curve of the template lithium battery.

[0040] S14: The incremental capacity curve is calculated by combining the open-circuit voltage curve.

[0041]

[0042] in, This indicates the incremental capacity of the target lithium battery. This indicates the preset voltage interval. This represents the incremental change in capacity of the target lithium battery within a preset voltage interval. This represents the change in capacity of the target lithium battery within a preset voltage interval. This indicates the preset voltage change value.

[0043] The present invention also provides a computer-readable storage medium comprising a lithium-ion battery health state estimation method program based on open-circuit voltage curve reconstruction. When the lithium-ion battery health state estimation method program based on open-circuit voltage curve reconstruction is executed by a processor, it implements the lithium-ion battery health state estimation method based on open-circuit voltage curve reconstruction as described above.

[0044] The above-described one or more technical solutions in the embodiments of the present invention have broader applicability and can adaptively capture and predict battery performance changes; an open-circuit voltage reconstruction model is established by using the open-circuit voltage curves of several template lithium batteries, and the aging characteristic points of the target lithium battery can be predicted based on the open-circuit voltage reconstruction model. The open-circuit voltage capacity curve of the target lithium battery is reconstructed based on the aging characteristic points, and the health status of the target lithium battery can be intuitively obtained based on the changes in the open-circuit voltage capacity curve.

[0045] Additional aspects and advantages of the invention will be set forth in part in the description which follows, and in part will be obvious from the description, or may be learned by practice of the invention. Attached Figure Description

[0046] To more clearly illustrate the technical solutions in this invention or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are some embodiments of this invention. For those skilled in the art, other drawings can be obtained from these drawings without creative effort.

[0047] Figure 1 This is a flowchart of a lithium-ion battery health state estimation method based on open-circuit voltage curve reconstruction provided in an embodiment of the present invention.

[0048] Figure 2 These are the incremental capacity curves corresponding to template lithium batteries in different aging states in the embodiments of the present invention.

[0049] Figure 3 This is a structural diagram of the CNN-Transformer according to an embodiment of the present invention.

[0050] Figure 4 This is a fitted graph of the OCV curve of the target lithium battery in an embodiment of the present invention. Detailed Implementation

[0051] To make the objectives, technical solutions, and advantages of this invention clearer, the technical solutions of this invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some, not all, of the embodiments of this invention. All other embodiments obtained by those skilled in the art based on the embodiments of this invention without creative effort are within the scope of protection of this invention. The following embodiments are used to illustrate this invention but should not be used to limit the scope of this invention.

[0052] The following is combined with Figure 1 The present invention describes a method for estimating the state of health of a lithium-ion battery based on open-circuit voltage curve reconstruction, comprising the following steps:

[0053] S1. Perform open-circuit battery testing on the template lithium battery to obtain the open-circuit voltage curve, and then calculate the incremental capacity curve of the template lithium battery through the open-circuit voltage curve.

[0054] S2, extract the first feature point related to the aging of the template lithium battery from the incremental capacity curve, and find the second feature point corresponding to the first feature point from the open circuit voltage curve;

[0055] S3. Obtain the first and second feature points of a set amount of template lithium battery as a training dataset, obtain the charging curve of the template lithium battery, divide the charging curve into data segments of fixed length, and then construct the open circuit voltage reconstruction model of the template lithium battery.

[0056] S4, optimize the parameters of the open-circuit voltage reconstruction model, and determine the mapping relationship between data segments and aging characteristic points reflecting the health status of template lithium batteries;

[0057] S5. Obtain charging data segments of the target lithium battery, input the charging data segments into the open circuit voltage reconstruction model to obtain aging characteristic points corresponding to the target lithium battery, and combine the aging characteristic points to reconstruct the open circuit voltage capacity curve to complete the health status estimation of the target lithium battery.

[0058] In a preferred embodiment of the lithium-ion battery health state estimation method based on open-circuit voltage curve reconstruction of the present invention, step S2 specifically involves:

[0059] Extract the first coordinates corresponding to the two troughs and one peak in the incremental capacity curve, define the first coordinates as the first feature points, and then find the voltage value and charging capacity value corresponding to the open circuit voltage curve based on the first coordinates, and define the voltage value and charging capacity value as the second feature points.

[0060] By extracting the first coordinates corresponding to two troughs and one peak, values ​​within a smaller period can be extracted, improving the accuracy of data extraction.

[0061] In a specific implementation case, such as Figure 3 As shown, the open-circuit voltage reconstruction model includes a convolutional layer, a pooling layer, a fully connected layer, a position encoding layer, a multi-head self-attention mechanism layer, a residual connection or normalization layer, a feedforward connection layer, a residual connection or normalization layer, a fully connected layer, and an output layer.

[0062] Further, step S3 involves constructing an open-circuit voltage reconstruction model for the template lithium battery. The specific steps are as follows:

[0063] S31, construct a dual-channel convolutional layer and a max pooling layer, and connect a pooling layer after each convolutional layer to reduce the risk of overfitting;

[0064] S32, First convolutional layer: For the input vector H, one-dimensional convolution and max pooling operations are applied.

[0065]

[0066] in The weights of the convolution kernel, This represents the bias term of the convolution kernel, where * indicates the convolution operation, and... It is an activation function used to add non-linearity. This layer extracts features from time series data by applying convolution operations. For pooling operations, This is the output of the first convolutional layer;

[0067] S33, the second convolutional layer follows immediately after the first convolutional layer, and the output of the first convolutional layer... As input to the second convolutional layer:

[0068]

[0069] in, The weights of the second convolutional kernel. This is the bias of the second convolutional kernel. This is the output of the second convolutional layer;

[0070] To construct the convolutional neural network (CNN) part of the open-circuit voltage reconstruction model;

[0071] S34, the Transformer part is constructed. The model uses four attention heads, which are composed of units such as encoder, multi-head self-attention mechanism, feedforward neural network, residual connection and layer normalization.

[0072] The self-attention mechanism can be represented as:

[0073] in, Matrix representing query, key, and value respectively. is the dimension of the key matrix, used to scale the dot product result to prevent gradient vanishing. This indicates a transpose calculation.

[0074] The core computational process of the self-attention mechanism is as follows: It is the attention head of each self-attention mechanism;

[0075]

[0076] in, It is a specific transformation weight that concatenates the results of the multi-head self-attention mechanism. , , These are the weights used in the self-attention mechanism.

[0077] To enhance the model's ability to fit complex states, two fully connected layers are added to improve its expressive power. The feedforward neural network is essentially a two-layer fully connected layer. The first layer uses ReLU as its activation function, while the second layer does not use an activation function. Its implementation process is shown in the following equation:

[0078]

[0079] For each sublayer (such as the self-attention layer and the feedforward network), residual connections and layer normalization are applied, as shown in the following equation: ;

[0080] This completes the CNN-Transformer structure of the open-circuit voltage reconstruction model.

[0081] Furthermore, step S3, which divides the charging curve into data segments of fixed length, includes the following steps:

[0082] Define the voltage range corresponding to a fixed length. and sampling step size The sampling set is obtained. ,in, ,

[0083] The charge amount corresponding to the data in the sample set was calculated by combining the ampere-hour integration method.

[0084] ,

[0085] The input vector is obtained based on several charge quantities. Then, an open-circuit voltage reconstruction model is constructed based on the input vector;

[0086] in, Indicates the starting voltage. Indicates the termination voltage. Indicates the number of sampling points. This indicates the charging current of the template lithium battery. An expression representing the time corresponding to the charging voltage. Represents the sequence of sampling points. Indicates the first The amount of charge corresponding to each sampling point.

[0087] Furthermore, the optimization of the open-circuit voltage reconstruction model in step S4 includes:

[0088] S41, Calculate the gradient during the iteration process. First-order moment estimation Second-order moment estimation ,

[0089]

[0090] in, Indicates the first attenuation rate. , Indicates the second attenuation rate. , This represents the gradient of the objective function being optimized. Indicates the corresponding time step The parameter values ​​evaluated during the iteration at each time step. Indicates time step First-order moment estimation at time 10:00 Indicates time step The second moment estimate at time t;

[0091] S42, Calculate the first-order moment estimate after bias correction. Second-order moment estimation after bias correction ;

[0092]

[0093] S43, Update the parameters of the open-circuit voltage reconstruction model. ,

[0094]

[0095] in, This represents the initial learning rate, in one specific embodiment, Set it to 0.001. This indicates the setting parameters; in one specific embodiment, Set as , Indicates the corresponding time step The parameter values ​​evaluated during the iteration at each time step.

[0096] Specifically, it also includes using the mean absolute percentage error as the basic loss function, and using the mean absolute percentage error to calculate the average percentage error between the predicted value and the true value, thereby ensuring that the model can effectively adjust the parameters during training to minimize the prediction error and prevent overfitting.

[0097] Furthermore, step S5 combines aging feature point reconstruction to obtain the open-circuit voltage-capacity curve, thereby completing the health status estimation of the target lithium battery. The specific steps are as follows:

[0098] S51, the open-circuit voltage reconstruction model includes an initial reference dataset with aging feature points. The initial reference dataset consists of L open-circuit voltages and L charging capacity values. The L open-circuit voltages and L charging capacity values ​​are arranged in chronological order to obtain the reference dataset. ,

[0099] S52, Match the aging feature points in the reference dataset with the aging feature points of the target lithium battery, and select the voltage offset coefficient. Capacity offset coefficient Curve scaling factor and construction Let be the loss coefficient. Combining the particle swarm optimization algorithm with the goal of minimizing the loss coefficient, we obtain the optimal matching coefficient.

[0100] S53, combining the optimal matching coefficient, transforms the aging feature points of the reference dataset to obtain the open-circuit voltage-capacity curve of the target lithium battery.

[0101]

[0102] in, This represents the charge amount at the aging characteristic point of the target lithium battery corresponding to the i-th sampling point. express The corresponding open-circuit voltage, This represents the open-circuit voltage of the aging characteristic point of the target lithium battery corresponding to the i-th sampling point in the reference dataset;

[0103] S54, the capacity of the open-circuit voltage-capacity curve corresponding to the upper limit cutoff voltage of the target lithium battery is the estimated capacity of the target lithium battery, combined with... The health status of the target lithium battery can then be obtained, among which, Indicates the estimated capacity. Indicates the rated capacity. This indicates the health status of the target lithium battery.

[0104] In one specific embodiment, the upper cutoff voltage of the target lithium battery is measured, and the corresponding estimated capacity can be obtained based on the upper cutoff voltage, i.e., the corresponding capacity can be obtained. This allows us to find the corresponding... Thus, the calculation is obtained .

[0105] Specifically, the loss coefficient is expressed as .

[0106] In a specific implementation case, such as Figure 4 As shown, Figure 4 To predict the health status of the target lithium battery, the reference OCV curve is the capacity-voltage curve generated by the reference dataset in the open-circuit voltage reconstruction model, the real OCV curve is the capacity-voltage curve of the target lithium battery, and the estimated OCV curve is the capacity-voltage curve of the target lithium battery obtained after processing by the open-circuit voltage reconstruction model. Based on the capacity-voltage curve and knowing the current voltage of the lithium battery, the current capacity of the target lithium battery can be estimated, and thus the health status of the target lithium battery can be estimated.

[0107] Further, step S1 includes:

[0108] S11: Discharge the template lithium battery to the cutoff voltage at a rate of 0.5C;

[0109] S12: After the template lithium battery is left to stand for 1 hour, it is charged in constant current and constant voltage mode at a rate of 0.5C. When the charging current drops to a set value, in one specific embodiment, the set value is 0.02mAh, the complete charging capacity is recorded as the template lithium battery capacity.

[0110] S13: After the template lithium battery is left to stand for 15 minutes, it is discharged at a constant current rate of 0.05C until it is discharged to the cutoff voltage, thereby obtaining the open circuit voltage curve of the template lithium battery.

[0111] S14: The incremental capacity curve is calculated by combining the open-circuit voltage curve.

[0112]

[0113] in, This indicates the incremental capacity of the target lithium battery. This indicates the preset voltage interval. This represents the incremental change in capacity of the target lithium battery within a preset voltage interval. This represents the change in capacity of the target lithium battery within a preset voltage interval. This indicates the preset voltage change value.

[0114] In a specific implementation case, such as Figure 2 As shown, by obtaining the incremental capacity curves of lithium batteries with different templates, the lithium batteries with different templates were cycled 60 times, 146 times, 232 times, 332 times and 420 times respectively, and the incremental capacity curves were calculated by the OCV curve.

[0115] The present invention also provides a computer-readable storage medium including a lithium-ion battery health state estimation method program based on open-circuit voltage curve reconstruction. When the lithium-ion battery health state estimation method program based on open-circuit voltage curve reconstruction is executed by a processor, the above-mentioned lithium-ion battery health state estimation method based on open-circuit voltage curve reconstruction is implemented.

[0116] The above-described one or more technical solutions in the embodiments of the present invention have broader applicability and can adaptively capture and predict battery performance changes; an open-circuit voltage reconstruction model is established by using the open-circuit voltage curves of several template lithium batteries, and the aging characteristic points of the target lithium battery can be predicted based on the open-circuit voltage reconstruction model. The open-circuit voltage capacity curve of the target lithium battery is reconstructed based on the aging characteristic points, and the health status of the target lithium battery can be intuitively obtained based on the changes in the open-circuit voltage capacity curve.

[0117] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention, and not to limit them; although the present invention has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that modifications can still be made to the technical solutions described in the foregoing embodiments, or equivalent substitutions can be made to some of the technical features; and these modifications or substitutions do not cause the essence of the corresponding technical solutions to deviate from the spirit and scope of the technical solutions of the embodiments of the present invention.

Claims

1. A lithium-ion battery state-of-health estimation method based on open-circuit voltage curve reconstruction, characterized in that, The method comprises the following steps: S1, open circuit battery test is performed on a template lithium battery to obtain an open circuit voltage curve, and then an incremental capacity curve of the template lithium battery is calculated through the open circuit voltage curve; S2, a first feature point related to the aging of the template lithium battery is extracted from the incremental capacity curve, and a second feature point corresponding to the first feature point is found on the open circuit voltage curve; S3, the first feature point and the second feature point of a set amount of template lithium batteries are obtained as a training data set, a charging curve of the template lithium battery is obtained, the charging curve is divided into data segments of a fixed length, and then an open circuit voltage reconstruction model of the template lithium battery is constructed; S4, parameters of the open circuit voltage reconstruction model are optimized, and a mapping relationship between the data segments and an aging feature point reflecting the health state of the template lithium battery is determined; S5, a charging data segment of a target lithium battery is obtained, the charging data segment is input into the open circuit voltage reconstruction model to obtain an aging feature point corresponding to the target lithium battery, and an open circuit voltage capacity curve is reconstructed by combining the aging feature point, so as to complete the health state estimation of the target lithium battery.

2. The open-circuit voltage curve reconstruction based lithium-ion battery state-of-health estimation method according to claim 1, characterized in that, Step S2 is specifically, The first coordinates corresponding to two troughs and one peak in the incremental capacity curve are extracted, the first coordinates are defined as first feature points, then the voltage value and the charging capacity value corresponding to the first coordinates on the open circuit voltage curve are found, and the voltage value and the charging capacity value are defined as second feature points.

3. The open-circuit voltage curve fitting based lithium-ion battery state-of-health estimation method of claim 1, wherein, In step S3, dividing the charging curve into data segments of a fixed length comprises the following steps: defining a voltage range corresponding to the fixed length and a sampling step yielding a sample set wherein , The charge amount corresponding to the data in the sampling set is calculated by combining the ampere-hour integral method, ， Input vector from several charge amounts and further constructing the open circuit voltage reconstruction model from the input vector; wherein, represents a start voltage, represents an end voltage, represents a number of sampling points, represents a charging current of the template lithium battery, represents an expression of a charging voltage corresponding time, represents a sequence of sampling points, represents a charge amount corresponding to the th sampling point.

4. The lithium-ion battery state-of-health estimation method based on open-circuit voltage curve reconstruction of claim 3, wherein, In step S4, optimizing the open circuit voltage reconstruction model comprises: S41, calculating gradient in iteration process , first moment estimation , second moment estimation , wherein, denotes a first decay rate, denotes a second decay rate, denotes a gradient of the objective function being optimized, denotes a parameter value evaluated in an iteration at a time step denotes a parameter value evaluated in an iteration at a time step denotes a first moment estimate at a time step denotes a first moment estimate at a time step denotes a second moment estimate at a time step denotes a second moment estimate at a time step S42, calculate the first moment estimate after bias correction , the second moment estimate after bias correction ; S43, updating the parameters of the open circuit voltage reconstruction model, wherein, denotes an initial learning rate, denotes a set parameter, denotes a parameter value evaluated in an iteration at a time step moment in time.

5. The lithium-ion battery state-of-health estimation method based on open-circuit voltage curve reconstruction of claim 4, wherein, In step S5, the open circuit voltage capacity curve is reconstructed by combining the aging feature point, so as to complete the health state estimation of the target lithium battery, and the specific steps are as follows: S51, the open circuit voltage reconstruction model is provided with an initial reference data set of aging feature points, and the initial reference data set includes L open circuit voltages and L charge capacity values, and the L open circuit voltages and the L charge capacity values are arranged in time sequence to obtain a reference data set ; S52, match the aging feature points in the reference data set with the aging feature points of the target lithium battery, select the voltage offset coefficient , the capacity offset coefficient , the curve stretching coefficient and the structure as the loss coefficient, combine the particle swarm algorithm to minimize the loss coefficient as the optimization target, and obtain the optimal matching coefficient; S53, combining the optimal matching coefficient, the data of the aging feature point of the reference data set is transformed to obtain the open circuit voltage capacity curve of the target lithium battery, wherein, represents the charge amount of the aging feature point corresponding to the target lithium battery at the i-th sampling point, represents the open circuit voltage corresponding to the i-th sampling point in the reference data set, represents the open circuit voltage of the aging feature point corresponding to the target lithium battery at the i-th sampling point in the reference data set;​ S54, the capacity of the open-circuit voltage capacity curve corresponding to the upper limit cut-off voltage of the target lithium battery is the estimated capacity of the target lithium battery, combined with The state of health of the target lithium battery can be obtained, wherein, represents the estimated capacity, represents the rated capacity, represents the state of health of the target lithium battery.

6. The open-circuit voltage curve reconstruction based lithium-ion battery state-of-health estimation method according to claim 5, characterized in that, The loss coefficient is expressed as .

7. The open-circuit voltage curve fitting based lithium-ion battery state-of-health estimation method of claim 1, wherein, Step S1 comprises: S11: discharge the template lithium battery to the cut-off voltage at 0.5C rate; S12: after the template lithium battery is left for 1 hour, charge the template lithium battery at 0.5C rate in constant current and constant voltage mode, when the charging current decreases to a set value, record the complete charging capacity as the capacity of the template lithium battery; S13: after the template lithium battery is left for 15 minutes, discharge the template lithium battery at 0.05C rate until the discharge voltage reaches the cut-off voltage, so as to obtain the open circuit voltage curve of the template lithium battery; S14: calculate the incremental capacity curve in combination with the open circuit voltage curve, wherein, represents an incremental capacity of the target lithium battery, represents a preset voltage interval, represents a capacity incremental change value of the target lithium battery within the preset voltage interval, represents a capacity change value of the target lithium battery within the preset voltage interval, represents a preset voltage change value.

8. A computer-readable storage medium, characterized in that, The computer readable storage medium comprises a lithium ion battery health state estimation method program based on open circuit voltage curve reconstruction, and the lithium ion battery health state estimation method program based on open circuit voltage curve reconstruction is executed by the processor to realize the lithium ion battery health state estimation method based on open circuit voltage curve reconstruction in any one of claims 1 to 7.

Images