A method for estimating state of health of lithium battery of energy storage power station based on charging current curve

Abstract

The application relates to a kind of energy storage power station lithium battery SOH estimation methods based on charging current curve, feature is extracted from the constant voltage charging current curve of battery, the influence of the initial point of charging is avoided, the accuracy of the estimation result is increased, and there is no special restriction on the size of charging current, with the advantages of wide application range;By extracting the statistical characteristics of constant voltage charging current curve and current first-order differential data, the correlation between the current curve and the aging of the battery is more accurately described, the input dimension of the model is reduced by using statistical indicators instead of time series data, and the estimation accuracy is improved.

Description

Technical Field

[0001] This invention relates to the field of battery energy storage technology, and specifically to a method for estimating the state of charge (SOH) of lithium batteries in energy storage power stations based on charging current curves. Background Technology

[0002] The safe operation of lithium batteries is crucial for energy storage power stations, necessitating real-time state of health (SOH) estimation for all lithium batteries. Since energy storage power stations primarily function as backup power sources, often playing a role in peak shaving and valley filling, the probability of lithium batteries being completely discharged is very low. Therefore, it is impossible to directly estimate the state of health based on the amount of discharge.

[0003] Existing State of Health (SOH) estimation methods, such as CN111443293A "A Data-Driven Method for Estimating the State of Health (SOH) of a Lithium-ion Battery," disclose a method for estimating the state of health (SOH) of a lithium-ion battery based on data, characterized by the following steps: 1) Real-time recording of charging data of the lithium-ion battery in constant current charging mode; 2) Calculating the capacity increment curve of the constant current charging voltage curve using a simplified dQ / dV processing method through capacity increment analysis; subsequently, determining the peak intensity and peak position voltage of peak 2 of the capacity increment curve as the feature vector for estimating SOH through grey relational analysis; 3) Constructing a support vector regression model with the peak intensity and peak position voltage as the feature vector as input and SOH as the output; 4) Integrating the differential evolution strategy with the grey wolf optimization algorithm to form an improved grey wolf optimization algorithm I. 5) The hyperparameters in the support vector regression model are jointly optimized using IGWO. The advantages of this invention are that it overcomes the shortcomings of existing technologies and has a reasonable and novel structural design. However, the SOH estimation method based on the complete charge-discharge curve characteristics requires deep discharge and ideal discharge conditions, making it difficult to adapt to the actual operating conditions of energy storage power stations. The SOH estimation method based on the constant current charging voltage curve characteristics faces the problem of an unstable initial charging point, limiting its application. The SOH estimation method based on the incremental capacity curve characteristics requires additional data noise reduction methods and is limited by the magnitude of the charging current. Existing SOH estimation methods based on the constant voltage charging process characteristics only consider the changing patterns of current data and fail to further explore the changing patterns of current differential data, making it difficult to extract more representative features and restricting the estimation accuracy. Summary of the Invention

[0004] To address the aforementioned technical problems, this application provides a method for estimating the state of charge (SOH) of lithium batteries in energy storage power stations based on charging current curves, comprising:

[0005] The lithium battery is subjected to a cycle test, which specifically involves charging the lithium battery using a constant current and constant voltage method, and then completely discharging the fully charged lithium battery using a constant current method. The charging and complete discharging operations of the lithium battery are repeated until the ratio of the current maximum discharge capacity of the lithium battery to the nominal capacity at the time of manufacture drops to the termination threshold.

[0006] A battery test dataset is constructed by collecting test data of lithium batteries during cycle testing. The test data includes the charging time, current data sequence, first-order difference sequence of current data, and maximum discharge capacity of lithium batteries during constant-voltage charging.

[0007] Based on the battery test dataset, lithium battery feature data is calculated, and the battery feature data is subjected to Min-Max normalization. A battery feature dataset is constructed based on the battery feature data after Min-Max normalization. The feature data includes the skewness value and Shannon entropy of the current data during constant voltage charging of the lithium battery, the skewness value and Shannon entropy of the first-order difference sequence of the current data, and the charging time. The battery feature dataset uses the battery health status as the label value.

[0008] The battery feature dataset is divided into a training set and a test set according to a preset ratio. A support vector regression model is built. The support vector regression model is trained using the training set and the performance of the support vector regression model is verified using the test set to obtain the trained support vector regression model.

[0009] The trained support vector retrospective model is imported into the battery management system of the energy storage power station to obtain the battery health status at the current moment.

[0010] Preferably, the current data sequence Ij of the j-th lithium battery during constant voltage charging is expressed by the formula:

[0011] Ij,n=[Ij,n,1,Ij,n,2,…,Ij,n,t,…,Ij,n,T];

[0012]

[0013] In the formula, I j，n，t This represents the current value of the j-th battery at the t-th sampling time during the n-th constant voltage charging process, where N is the number of cycle tests and T is the number of sampling times.

[0014] The first-order difference sequence ΔIj of the current data of the j-th lithium battery during constant voltage charging is expressed by the formula:

[0015] ΔI j,n =[ΔI j,n,1 ,ΔI j,n,2,…,ΔI j,n,t ,…,ΔI j,n,T-1 ];

[0016]

[0017] In the formula, △I j，n，t This represents the difference between the current value of the j-th battery at the (t-1)-th sampling time and the current value at the t-th sampling time during the n-th constant voltage charging process.

[0018] Preferably, the lithium battery characteristic data is calculated based on the battery test dataset as follows:

[0019] The skewness value of the current data sequence of the j-th lithium battery during the nth constant-voltage charging process is calculated using the following formula:

[0020]

[0021] I j, =[I j,1, I j,2, , ..., I j,, , ..., I j,, ];

[0022] In the formula, I j，n，skew Let I be the skewness value of the current data sequence of the j-th lithium battery during the n-th constant voltage charging process. j，n，m This represents the current data of the j-th lithium battery at the m-th sampling time during the n-th constant voltage charging process. Let be the average value of the current data sequence of the j-th lithium battery during the nth constant voltage charging process;

[0023] The Shannon entropy of the current data sequence of the j-th lithium battery during the nth constant-voltage charging process is calculated using the following formula:

[0024]

[0025] I j,shan =[I j,1,shan I j,2,shan , ..., I j,n,shan , ..., I j,N,shan ];

[0026] In the formula, p m I represents the current data sequence of the j-th lithium battery during the n-th constant voltage charging process. j Current data I appears in j,n,m The probability of the value of ;

[0027] The skewness value of the first-order difference sequence of the current data of the j-th lithium battery during the nth constant-voltage charging process is calculated by the following formula:

[0028]

[0029] △I j,skew =[△I j,1,skew , △I j,2,skew , ..., △I j,n,skew , ..., △I j,N,skew ];

[0030] In the formula, △I j，n，skew Let ΔI be the skewness value of the first-order difference sequence of the current data of the j-th lithium battery during the n-th constant-voltage charging process. j，n，m This represents the m-th data point in the first-order differential current data sequence of the j-th lithium battery during the n-th constant-voltage charging process. Let be the average value of the first-order difference data sequence of the current of the j-th lithium battery during the nth constant-voltage charging process;

[0031] The Shannon entropy of the first-order difference sequence of the current data of the j-th lithium battery during the nth constant-voltage charging process is calculated as follows:

[0032]

[0033] △I j,shan =[△I j,1,shan , △I j,2,shan , ..., △I j,n,shan , ..., △I j,N,shan ];

[0034] In the formula, p m ' represents the probability that the value of the m-th current data △Ij,m appears in the first-order difference sequence △Ij of the current data of the j-th lithium battery during the n-th constant voltage charging process.

[0035] Preferably, the battery health status is expressed by the formula:

[0036]

[0037] SOH j =[SOH j,1 SOH j,2 , ..., SOH j,n , ..., SOH j,N ];

[0038] In the formula, SOH j,n For the j-th lithium battery in the nth cycle test, SOH j Let Q be the set of battery health states of the j-th lithium battery during cycle testing. j , maxLet C be the maximum discharge capacity of the j-th lithium battery during the n-th constant current discharge process, and let C be the nominal capacity of the lithium battery.

[0039] Preferably, the battery feature data is processed by Min-Max normalization, expressed by the formula:

[0040]

[0041]

[0042]

[0043]

[0044]

[0045] In the formula, A j,d A j,skew A j,shan A j,△skew and A j,△shan Let I be the set of eigenvalues ​​of the characteristic data of the j-th lithium battery in the nth charge-discharge cycle after Min-Max normalization. j,n,d I j,n,skew I j,n,shan , △I j,n,skew and △I j,n,shan These are the feature values ​​of the j-th lithium battery in the nth charge-discharge cycle; max 0≤n≤N () represents the maximum value of the characteristic data of the j-th lithium battery during the cycle test, min 0≤n≤N () represents the minimum value of the characteristic data of the j-th lithium battery during the cycle test.

[0046] Preferably, a battery feature dataset Set2 is constructed based on the battery feature data after Min-Max normalization, expressed by the following formula:

[0047] Set 2,j ={A j,d A j,skew A j,shan A j,△skew A j,△shan SOH j};

[0048]

[0049]

[0050] In the formula, J represents the number of lithium batteries subjected to cycle testing.

[0051] Preferably, the constructed support vector regression model uses feature vectors as input and label values ​​as output. The mapping relationship between input and output is expressed by the formula:

[0052]

[0053] In the formula, x i and x j Let f(x) be the feature vector of the feature data, f(x) be the label value, and a be the feature vector of the feature data. i and a i * It is a Lagrange multiplier, b is the deviation, k(x) i ,x j The kernel function is expressed by the formula:

[0054]

[0055] In the formula, g represents the bandwidth of the Gaussian kernel function.

[0056] Preferably, the performance of the support vector regression model is verified using a test set to obtain the trained support vector regression model as follows:

[0057] The root mean square error coefficient is used as the evaluation index for validation performance. When the root mean square error coefficient of the support vector regression model on the test set is less than the preset threshold, the parameters of the support vector regression model are saved and the trained support vector regression model is obtained; otherwise, the steps of building and training the support vector regression model are repeated until the root mean square error coefficient of the support vector regression model is less than the preset threshold.

[0058] This application also provides an electronic device, including a memory, a processor, and a computer program stored in the memory and executable on the processor. When the processor executes the program, it implements a method for estimating the state of charge (SOH) of a lithium battery in an energy storage power station based on a charging current curve, as described in any embodiment of the present invention.

[0059] This application also provides a computer-readable storage medium having a computer program stored thereon, which, when executed by a processor, implements a method for estimating the state of harmonics (SOH) of a lithium battery in an energy storage power station based on a charging current curve, as described in any embodiment of the present invention.

[0060] Compared with the prior art, the beneficial effects of the present invention are:

[0061] This invention provides a method for estimating the state of harm (SOH) of lithium batteries in energy storage power stations based on charging current curves. It extracts features from the constant-voltage charging current curve of the battery, avoiding the influence of an unstable initial charging point and increasing the accuracy of the estimation results. Furthermore, it has no special limitations on the magnitude of the charging current, making it widely applicable. By extracting statistical features from the constant-voltage charging current curve and the first-order difference data of the current, it more accurately describes the correlation between the current curve and battery aging. By using statistical indicators instead of time-series data, it reduces the dimensionality of the model input and improves the estimation accuracy. Attached Figure Description

[0062] Figure 1 This is a flowchart of the method in an embodiment of the present invention. Detailed Implementation

[0063] 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. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0064] Example 1

[0065] Embodiment 1 of this invention discloses a method for estimating the state of harm (SOH) of lithium batteries in an energy storage power station based on the charging current curve. This invention processes the access control process of power equipment using a blockchain network consisting of a central server and multiple nodes. Figure 1 As shown, the specific steps include:

[0066] S1. Perform a cycle test on the lithium battery. Specifically, the cycle test involves charging the lithium battery using a constant current and constant voltage method, and then completely discharging the fully charged lithium battery using a constant current method. Repeat the charging and discharging operations of the lithium battery until the SOH of the lithium battery drops to the termination threshold. In this embodiment, SOH is defined as the ratio of the current maximum discharge capacity to the nominal capacity at the time of manufacture. The termination threshold is 70%.

[0067] S2. Collect test data of lithium battery in cycle test to construct battery test dataset, wherein the test data includes the charging time, current data sequence, first-order difference sequence of current data and maximum discharge capacity of lithium battery in constant current discharge process during constant voltage charging process.

[0068] S21. Preferably, the current data sequence Ij of the j-th lithium battery during constant voltage charging is expressed by the formula:

[0069] Ij,n=[Ij,n,1,Ij,n,2,…,Ij,n,t,…,Ij,n,T];

[0070]

[0071] In the formula, I j，n，t This represents the current value of the j-th battery at the t-th sampling time during the n-th constant voltage charging process, where N is the number of cycle tests and T is the number of sampling times.

[0072] The first-order difference sequence ΔIj of the current data of the j-th lithium battery during constant voltage charging is expressed by the formula:

[0073] ΔI j,n =[ΔI j,n,1 ,ΔI j,n,2 ,…,ΔI j,n,t ,…,ΔI j,n,T-1 ];

[0074]

[0075] In the formula, △I j，n，t This represents the difference between the current value of the j-th battery at the (t-1)-th sampling time and the current value at the t-th sampling time during the n-th constant voltage charging process.

[0076] S3. Based on the battery test dataset, calculate lithium battery feature data, perform Min-Max normalization on the battery feature data, and construct a battery feature dataset based on the battery feature data after Min-Max normalization. The feature data includes the skewness value and Shannon entropy of the current data during constant voltage charging of the lithium battery, the skewness value and Shannon entropy of the first-order difference sequence of the current data, and the charging time. The battery feature dataset uses the battery health status as the label value.

[0077] S31. Based on the battery test dataset, the lithium battery characteristic data is calculated as follows:

[0078] The skewness value of the current data sequence of the j-th lithium battery during the nth constant-voltage charging process is calculated using the following formula:

[0079]

[0080] I j,shew =[I j,1,shew I j,2,shew , ..., I j,n,shew , ..., I j,N,shew ];

[0081] In the formula, I j，n，skewLet I be the skewness value of the current data sequence of the j-th lithium battery during the n-th constant voltage charging process. j，n，m This represents the current data of the j-th lithium battery at the m-th sampling time during the n-th constant voltage charging process. Let be the average value of the current data sequence of the j-th lithium battery during the nth constant voltage charging process;

[0082] The Shannon entropy of the current data sequence of the j-th lithium battery during the nth constant-voltage charging process is calculated using the following formula:

[0083]

[0084] I j,shan =[I j,1,shan I j,2,shan , ..., I j,n,shan , ..., I j,N,shan ];

[0085] In the formula, p m I represents the current data sequence of the j-th lithium battery during the n-th constant voltage charging process. j Current data I appears in j,n,m The probability of the value of ;

[0086] The skewness value of the first-order difference sequence of the current data of the j-th lithium battery during the nth constant-voltage charging process is calculated by the following formula:

[0087]

[0088] △I j,skew =[△I j,1,skew , △I j,2,skew , ..., △I j,n,skew , ..., △I j,N,skew ];

[0089] In the formula, △I j，n，skew Let ΔI be the skewness value of the first-order difference sequence of the current data of the j-th lithium battery during the n-th constant-voltage charging process. j，n，m This represents the m-th data point in the first-order differential current data sequence of the j-th lithium battery during the n-th constant-voltage charging process. Let be the average value of the first-order difference data sequence of the current of the j-th lithium battery during the nth constant-voltage charging process;

[0090] The Shannon entropy of the first-order difference sequence of the current data of the j-th lithium battery during the nth constant-voltage charging process is calculated as follows:

[0091]

[0092] △I j,shan =[△I j,1,shan, △I j,2,shan , ..., △I j,n,shan , ..., △I j,N,shan ];

[0093] In the formula, p m ' represents the probability that the value of the m-th current data △Ij,m appears in the first-order difference sequence △Ij of the current data of the j-th lithium battery during the n-th constant voltage charging process;

[0094] S32. The battery health status is expressed by the formula:

[0095]

[0096] SOH j =[SOH j,1 SOH j,2 , ..., SOH j,n , ..., SOH j,N ];

[0097] In the formula, SOH j,n For the j-th lithium battery in the nth cycle test, SOH j Let Q be the set of battery health states of the j-th lithium battery during cycle testing. j , max Let C be the maximum discharge capacity of the j-th lithium battery during the n-th constant current discharge process, and let C be the nominal capacity of the lithium battery.

[0098] S33. Perform Min-Max normalization on the battery feature data, expressed by the formula:

[0099]

[0100]

[0101]

[0102]

[0103]

[0104] In the formula, A j,d A j,skew A j,shan A j,△skew and A j,△shan Let I be the set of eigenvalues ​​of the characteristic data of the j-th lithium battery in the nth charge-discharge cycle after Min-Max normalization. j,n,d I j,n,shew I j,n,shan , △I j,n,shew and △I j,n,shanThese are the feature values ​​of the j-th lithium battery in the nth charge-discharge cycle; max 0≤n≤N () represents the maximum value of the characteristic data of the j-th lithium battery during the cycle test, min 0≤n≤N () represents the minimum value of the characteristic data of the j-th lithium battery during the cycle test;

[0105] S34. Construct a battery feature dataset Set2 based on the battery feature data after Min-Max normalization, expressed by the following formula:

[0106] Set 2,j ={A j,d A j,skew A j,shan A j,△skew A j,△shan SOH j};

[0107]

[0108]

[0109] In the formula, J represents the number of lithium batteries subjected to cycle testing.

[0110] S4. Divide the battery feature dataset into a training set and a test set according to a preset ratio. In this embodiment, the training set and test set are divided in a ratio of 3:1.

[0111] Build a support vector regression model, train the support vector regression model using the training set, verify the performance of the support vector regression model using the test set, and obtain the trained support vector regression model.

[0112] S41. The constructed support vector regression model uses feature vectors as input and label values ​​as output. In this embodiment, the input, i.e., the feature data, is 5, and the output is 1. The mapping relationship between the input and output is expressed by the formula:

[0113] f(x) = w T x+b;

[0114] In the formula, x is the input, f(x) is the output, and w and b are the coefficient matrix and bias, respectively. To solve for w and b, the support vector regression model establishes the following calculation formula:

[0115]

[0116]

[0117] In the formula, C is the penalty term, and a i and ai * These are Lagrange multipliers. Solving them yields the final relationship between the input and output, expressed as a formula:

[0118]

[0119] In the formula, x i and x j Let f(x) be the feature vector of the feature data, f(x) be the label value, and a be the feature vector of the feature data. i and a i * It is a Lagrange multiplier, and b is the deviation;

[0120] Where, k(x) i ,x j The kernel function () is used to address the computational difficulties caused by high-dimensional inputs. Here, the Gaussian kernel function is used as the kernel function of the model, expressed by the formula:

[0121]

[0122] In the formula, g represents the bandwidth of the Gaussian kernel function;

[0123] S42. Validate the performance of the support vector regression model using the test set. The trained support vector regression model is as follows:

[0124] The root mean square error coefficient (RMSE) is used as the evaluation metric for validation performance. When the RMSE coefficient of the support vector regression model on the test set is less than a preset threshold, the parameters of the support vector regression model are saved, and the trained support vector regression model is obtained. Otherwise, the steps of building and training the support vector regression model are repeated until the RMSE coefficient of the support vector regression model is less than the preset threshold.

[0125] S5. Import the trained support vector retrospective model into the battery management system of the energy storage power station to obtain the battery health status at the current moment.

[0126] Example 2

[0127] The present invention also provides an electronic device, including a memory, a processor, and a computer program stored in the memory and executable on the processor. When the processor executes the program, it implements a method for estimating the state of charge (SOH) of a lithium battery in an energy storage power station based on a charging current curve as described in Embodiment 1.

[0128] Example 3

[0129] The present invention also provides a computer-readable storage medium having a computer program stored thereon, which, when executed by a processor, implements a method for estimating the state of charge (SOH) of a lithium battery in an energy storage power station based on a charging current curve as described in Embodiment 1.

[0130] It is worth noting that the system, electronic device and computer-readable storage medium described in this invention are all based on the same inventive concept as the method for estimating the SOH of a lithium battery in an energy storage power station based on the charging current curve in Embodiment 1, and the specific technical details will not be repeated here.

[0131] The above description is merely an embodiment of the present invention and does not limit the patent scope of the present invention. Any equivalent structural or procedural transformations made based on the content of the present invention's specification and drawings, or direct or indirect applications in other related technical fields, are similarly included within the patent protection scope of the present invention.

Claims

1. A method for estimating the state of charge (SOH) of a lithium battery in an energy storage power station based on the charging current curve, characterized in that, The specific steps are as follows: The lithium battery is subjected to a cycle test, which specifically involves charging the lithium battery using a constant current and constant voltage method, and then completely discharging the fully charged lithium battery using a constant current method. The charging and complete discharging operations of the lithium battery are repeated until the ratio of the current maximum discharge capacity of the lithium battery to the nominal capacity at the time of manufacture drops to the termination threshold. A battery test dataset is constructed by collecting test data of lithium batteries during cycle testing. The test data includes the charging time, current data sequence, first-order difference sequence of current data, and maximum discharge capacity of lithium batteries during constant-voltage charging. Based on the battery test dataset, lithium battery feature data is calculated, and the battery feature data is subjected to Min-Max normalization. A battery feature dataset is constructed based on the battery feature data after Min-Max normalization. The feature data includes the skewness value and Shannon entropy of the current data during constant voltage charging of the lithium battery, the skewness value and Shannon entropy of the first-order difference sequence of the current data, and the charging time. The battery feature dataset uses the battery health status as the label value. The battery feature dataset is divided into a training set and a test set according to a preset ratio. A support vector regression model is built. The support vector regression model is trained using the training set and the performance of the support vector regression model is verified using the test set to obtain the trained support vector regression model. The trained support vector retrospective model is imported into the battery management system of the energy storage power station to obtain the battery health status at the current moment.

2. The method for estimating the state of charge (SOH) of a lithium battery in an energy storage power station based on the charging current curve according to claim 1, characterized in that, The current data sequence Ij of the j-th lithium battery during constant voltage charging is expressed by the formula: I j，n =[I j，n，1 ,I j，n，2 ,...,I j，n，t ,...,I j，n，T ]; In the formula, I j，n，t This represents the current value of the j-th battery at the t-th sampling time during the n-th constant voltage charging process, where N is the number of cycle tests and T is the number of sampling times. The first-order difference sequence ΔI of the current data of the j-th lithium battery during constant voltage charging. j Expressed as a formula: ΔI j，n =[ΔI j，n，1 ,ΔI j，n，2 ,...,ΔI j，n，t ,...,ΔI j，n，T-1 ]; In the formula, ΔI j，n，t This represents the difference between the current value of the j-th battery at the (t-1)-th sampling time and the current value at the t-th sampling time during the n-th constant voltage charging process.

3. The method for estimating the state of charge (SOH) of a lithium battery in an energy storage power station based on the charging current curve according to claim 2, characterized in that, Based on the battery test dataset, the lithium battery characteristic data is calculated as follows: The skewness value of the current data sequence of the j-th lithium battery during the nth constant-voltage charging process is calculated using the following formula: I j，skew =I j，1，skew ,I j，2，skew ,...,I j，n，skew ,...,I j，N，skew ]; In the formula, I j，n，skew Let I be the skewness value of the current data sequence of the j-th lithium battery during the n-th constant voltage charging process. j，n，m This represents the current data of the j-th lithium battery at the m-th sampling time during the n-th constant voltage charging process. Let be the average value of the current data sequence of the j-th lithium battery during the nth constant voltage charging process; The Shannon entropy of the current data sequence of the j-th lithium battery during the nth constant-voltage charging process is calculated using the following formula: I j，shan =[I j，1，shan ,I j，2，shan ,...,I j，n，shan ,...,I j，N，shan ]; In the formula, p m I represents the current data sequence of the j-th lithium battery during the n-th constant voltage charging process. j Current data I appears in j，n，m The probability of the value of ; The skewness value of the first-order difference sequence of the current data of the j-th lithium battery during the nth constant-voltage charging process is calculated by the following formula: ΔI j，skew =[ΔI j，1，skew ,ΔI j，2，skew ,...,ΔI j，n，skew ,...,ΔI j，N，skew ]; In the formula, ΔI j，n，skew Let ΔI be the skewness value of the first-order difference sequence of the current data of the j-th lithium battery during the n-th constant-voltage charging process. j，n，m This represents the m-th data point in the first-order differential current data sequence of the j-th lithium battery during the n-th constant-voltage charging process. Let be the average value of the first-order difference data sequence of the current of the j-th lithium battery during the nth constant-voltage charging process; The Shannon entropy of the first-order difference sequence of the current data of the j-th lithium battery during the nth constant-voltage charging process is calculated as follows: ΔI j，shan =[ΔI j，1，shan ,ΔI j，2，shan ,...,ΔI j，n，shan ,...,ΔI j，N，shan ]; In the formula, p m ' represents the first-order difference sequence ΔI of the current data of the j-th lithium battery during the nth constant-voltage charging process. j The m-th current data ΔI appears in the data. j，m The probability of the value of .

4. The method for estimating the state of harmonics (SOH) of a lithium battery in an energy storage power station based on the charging current curve according to claim 1, characterized in that, The battery health status is expressed by the formula: SOH j =[SOH j，1 ,SOH j，2 ,...,SOH j，n ,...SOH j，N ]; In the formula, SOH j，n For the j-th lithium battery in the nth cycle test, SOH j Let Q be the set of battery health states of the j-th lithium battery during cycle testing. j，max Let C be the maximum discharge capacity of the j-th lithium battery during the n-th constant current discharge process, and let C be the nominal capacity of the lithium battery.

5. The method for estimating the state of charge (SOH) of a lithium battery in an energy storage power station based on the charging current curve according to claim 3, characterized in that, The battery feature data is processed using Min-Max normalization, expressed by the following formula: In the formula, A j，d A j，skew A j，shan A j，Δskew and A j，Δshan Let I be the set of eigenvalues ​​of the characteristic data of the j-th lithium battery in the nth charge-discharge cycle after Min-Max normalization. j，n，d I j，n，skew I j，n，shan ΔI j，n，skew and ΔI j，n，shan These are the feature values ​​of the j-th lithium battery in the nth charge-discharge cycle; max 0≤n≤N () represents the maximum value of the characteristic data of the j-th lithium battery during the cycle test, min 0≤n≤N () represents the minimum value of the characteristic data of the j-th lithium battery during the cycle test.

6. The method for estimating the state of charge (SOH) of a lithium battery in an energy storage power station based on the charging current curve according to claim 5, characterized in that, The battery feature dataset Set2 is constructed based on the battery feature data after Min-Max normalization, and is expressed by the following formula: Set 2，j ={A j，d ,A j，skew ,A j，shan ,A j，Δskew ,A j，Δshan ,SOH j }; In the formula, J represents the number of lithium batteries subjected to cycle testing.

7. The method for estimating the state of charge (SOH) of a lithium battery in an energy storage power station based on the charging current curve according to claim 1, characterized in that, The constructed support vector regression model uses feature vectors as input and label values ​​as output. The mapping relationship between input and output is expressed by the formula: In the formula, x i and x j Let f(x) be the feature vector of the feature data, f(x) be the label value, and α be the feature vector of the feature data. i and ɑ i * represents the Lagrange multiplier, b is the deviation, and k(x) is the Lagrange multiplier. i ,x j The kernel function is expressed by the formula: In the formula, g represents the bandwidth of the Gaussian kernel function.

8. The method for estimating the state of charge (SOH) of a lithium battery in an energy storage power station based on the charging current curve according to claim 1, characterized in that, The performance of the support vector regression model is validated using a test set. The trained support vector regression model is as follows: The root mean square error coefficient is used as the evaluation index for validation performance. When the root mean square error coefficient of the support vector regression model on the test set is less than the preset threshold, the parameters of the support vector regression model are saved, and the trained support vector regression model is obtained. Otherwise, repeat the steps of building and training the support vector regression model until the root mean square error coefficient of the support vector regression model is less than the preset threshold.

9. An electronic device comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, characterized in that, When the processor executes the program, it implements a method for estimating the state of harm (SOH) of a lithium battery in an energy storage power station based on a charging current curve, as described in any one of claims 1 to 8.

10. A computer-readable storage medium having a computer program stored thereon, characterized in that, When executed by the processor, the program implements a method for estimating the state of charge (SOH) of a lithium battery in an energy storage power station based on a charging current curve, as described in any one of claims 1 to 8.

Images