Battery remaining service life prediction method based on similarity feature enhancement, electronic equipment and readable storage medium

Abstract

The invention provides a battery remaining service life prediction method based on similarity feature enhancement. The method comprises the following steps of S1, extracting current data XS and voltage data XV of battery cyclic charging and discharging in real time; s2, determining a current map G1 (XS1, A1) according to the current data XS, and determining a voltage map G2 (XV1, A2) according to the voltage data XV; s3, designing a similarity feature enhancement module, and inputting the current diagram G1 and the voltage diagram G2 into the similarity feature enhancement module to obtain a comprehensive feature Hout; s4, inputting the comprehensive feature Hout into a prediction module in real time to obtain the health state of the battery; by adopting the comprehensive characteristics obtained by the method to predict the remaining service life of the battery, the prediction accuracy can be improved.

Description

Technical Field

[0001] The present invention relates to the field of battery life prediction, and in particular to a battery remaining service life prediction method based on similarity feature enhancement, an electronic device and a readable storage medium. Background Art

[0002] Rechargeable lithium-ion batteries play a vital role in many modern applications, especially in electric vehicles, where they are widely used, mainly due to their high energy density, long cycle life, low self-discharge rate, and low environmental pollution. As a product of industrialization, the performance of lithium-ion batteries shows degradation. Generally, the remaining useful life prediction (RUL) of lithium-ion batteries can be characterized by capacity. When the effective capacity of a lithium-ion battery is lower than 80% of the initial value, the battery is in an unstable working state; if the current capacity drops from 80% to 70%, the battery performance will drop sharply and be regarded as a battery failure; if the battery continues to operate, it may cause safety accidents, equipment may be damaged or cause battery explosion, etc. Therefore, accurately judging the RUL of lithium-ion batteries is a prerequisite for the full utilization and safe use of lithium-ion batteries, which can maximize the utilization of lithium-ion battery performance and extend its life, and provide subsequent predictive maintenance strategies and energy system management.

[0003] So far, typical methods for RUL prediction of lithium-ion batteries are generally divided into two categories: model-based methods and data-driven methods. Model-based methods describe the degradation behavior of batteries by simplifying the internal mechanism of the battery and then modeling the battery capacity decay condition. However, such methods rely on a large amount of prior knowledge of the system degradation mechanism. At the same time, in practice, it is difficult to accurately evaluate the RUL of lithium-ion batteries for batteries working in certain noisy and uncertain environments. Compared with model-based method research, data-driven methods can reduce the time cost of constructing models and reduce the interference of improper model establishment on subsequent life prediction. Current data-driven research methods include artificial neural networks, support vector regression, relevance vector machines, and recurrent neural networks. However, such methods can only process regular data and cannot well extract the hidden information and degradation laws of performance decay in battery historical data, resulting in deficiencies in model structure optimization and improvement of prediction model accuracy.

[0004] Therefore, in order to solve the above technical problems, it is urgent to propose a new technical means. Summary of the invention

[0005] In view of this, in order to improve the prediction accuracy of the remaining service life of a battery, the present invention proposes a method for predicting the remaining service life of a battery based on similarity feature enhancement, an electronic device and a readable storage medium.

[0006] The present invention provides a method for predicting the remaining service life of a battery based on similarity feature enhancement, comprising the following steps:

[0007] S1. Real-time extraction of current data during battery charge and discharge cycle X S and voltage data X V ;

[0008] S2. According to the current data X S Determine the current diagram G1(X S1 , A1), and according to the voltage data X V Determine the voltage diagram G2(X V1 ,A2);

[0009] Among them, X S1 and X V1 Respectively represent the current data X S and the voltage data X V The extracted features, A1 and A2 represent the current adjacency matrix and voltage adjacency matrix respectively;

[0010] S3. Design a similarity feature enhancement module, and input the current graph G1 and the voltage graph G2 into the similarity feature enhancement module to obtain a comprehensive feature H out ;

[0011] S4. Real-time processing of the comprehensive feature H out The data is input into the prediction module to obtain the remaining service life of the battery.

[0012] Furthermore, the feature enhancement module obtains the comprehensive feature H through the following steps: out :

[0013] S31. Using three layers of graph convolution layers to extract features of the current graph G1 and the voltage graph G2, respectively, to obtain current features and voltage features of each layer;

[0014] S32. Calculate the current characteristic enhancement coefficient and the voltage characteristic enhancement coefficient between each layer, including the current characteristic enhancement coefficient and the voltage characteristic enhancement coefficient between the first layer and the second layer, the first layer and the third layer, and the second layer and the third layer;

[0015] S33. Calculate the enhancement characteristic H between the first layer and the second layer according to the current characteristics and voltage characteristics of each layer, and the current characteristic enhancement coefficient and voltage characteristic enhancement coefficient between the layers. 12 , the enhanced feature H between the first and third layers 13 , and the enhanced feature H between the second and third layers 23 ;

[0016] S34. The enhancement feature H between the first layer and the second layer 12 , the enhanced feature H between the first layer and the third layer 13 and the enhancement feature H between the second and third layers 23 Splice and get the comprehensive feature H out ;

[0017] The comprehensive feature H out The expression is as follows:

[0018] H out =H 12 ||H 13 ||H 23

[0019] Among them, h 12 ||H 13 ||H 23 Represents the enhanced feature H 12 , H 13 and H 23 Make stitching.

[0020] Furthermore, the enhanced feature h 12 , H 13 and H 23 Calculated by the following formulas:

[0021]

[0022]

[0023]

[0024] Among them, a S12 、a S13 and a S23 represent the current characteristic enhancement coefficients between the first layer and the second layer, the first layer and the third layer, and the second layer and the third layer respectively; a V12 、a V13 and a V23 Respectively represent the voltage characteristic enhancement coefficients between the first layer and the second layer, the first layer and the third layer, and the second layer and the third layer; Indicates that the first layer current characteristics and the second layer current characteristics are spliced, Indicates that the first layer voltage feature and the second layer voltage feature are spliced; Indicates that the first layer current feature and the third layer current feature are spliced, Indicates that the first layer voltage feature and the third layer voltage feature are spliced; Indicates that the second layer current feature is spliced ​​with the third layer current feature, It indicates that the second layer voltage feature is spliced ​​with the third layer voltage feature.

[0025] Furthermore, the current characteristic enhancement coefficient a S12 、a S13 and a S23 Calculated by the following formula:

[0026]

[0027]

[0028]

[0029] Among them, sigmoid represents the sigmoid function, A S12 , A S13 and A S23 are the current cosine similarity matrices, R S1 , R S2 and R S3 Respectively represent the cosine feature similarity of the current graph in the first convolutional layer, the second convolutional layer, and the third convolutional layer, L1, L2, and L3 represent the first layer, the second layer, and the third layer, L1=1, L2=2, and L3=3;

[0030] The voltage characteristic enhancement factor a V12 、a V13 and a V23 The calculation principle and the current characteristic enhancement coefficient a S12 、a S13 and a S23 The calculation principle is the same as that of . During the calculation, the current feature similarity and current cosine similarity matrix in the current feature enhancement coefficient calculation formula are replaced with the corresponding voltage feature similarity and voltage cosine similarity matrix.

[0031] Furthermore, the current cosine similarity matrix A S12 , A S13 and A S23 Calculated by the following formula:

[0032]

[0033]

[0034]

[0035] in, and Respectively represent the norms of the first layer current characteristics, the second layer current characteristics and the third layer current characteristics;

[0036] The calculation principle of the voltage cosine similarity matrix is ​​the same as the calculation principle of the current cosine similarity matrix. During calculation, the current characteristics and the norm of the current characteristics in the formula are replaced with the corresponding voltage characteristics and the norm of the voltage characteristics.

[0037] Furthermore, the current characteristic and the voltage characteristic are calculated by the following formula:

[0038]

[0039]

[0040] in, represents the current characteristics output after the current graph undergoes the lth graph convolution, It represents the voltage features output after the voltage graph undergoes the lth graph convolution. l=1,2,3 correspond to three layers of graph convolution respectively. When l=1, X S1 and X V1 Respectively represent the current data X S and the voltage data X V The extracted features, A1 and A2 represent the current adjacency matrix and voltage adjacency matrix respectively. and They represent the normalization operation of the adjacency matrix, and They represent the weight matrices of current and voltage in the l-1 layer respectively, and ReLU represents the activation function.

[0041] Furthermore, the current adjacency matrix A1 and the voltage adjacency matrix A2 are determined by the following method:

[0042] Calculate feature X S1 With current data X S The cosine similarity of and based on feature X S1 With current data X S The cosine similarity constructs the current adjacency matrix A1;

[0043] Calculate feature X V1 With voltage data X V The cosine similarity of and based on feature X V1 With voltage data X V The voltage adjacency matrix A2 is constructed based on the cosine similarity.

[0044] Furthermore, the feature X S1 and X V1 Calculated by the following formula:

[0045] X S1 =MLP(X S)

[0046] X V1 =MLP(X V )

[0047] Among them, MLP(X S ) indicates that the current data X is extracted using a multi-layer perceptron S The features in MLP(X V ) indicates that the voltage data X is extracted using a multi-layer perceptron V Features in .

[0048] Accordingly, the present invention further provides an electronic device, comprising:

[0049] A memory and a processor, wherein the memory is used to store a computer program, and when the computer program is executed by the processor, the method for predicting the remaining service life of a battery based on similarity feature enhancement is implemented.

[0050] Correspondingly, the present invention also proposes a readable storage medium, in which computer instructions are stored. When the computer instructions are executed by a processor, the above-mentioned battery remaining service life prediction method based on similarity feature enhancement is implemented.

[0051] Beneficial effects of the present invention: The present invention obtains a comprehensive feature that integrates and enhances the current feature and the voltage feature by determining the current graph and the voltage graph, and inputting the current graph and the voltage graph into a similarity feature enhancement module. The comprehensive feature is used to predict the remaining service life of the battery, which can improve the accuracy of the prediction. BRIEF DESCRIPTION OF THE DRAWINGS

[0052] The present invention will be further described below in conjunction with the accompanying drawings and embodiments:

[0053] Figure 1 It is a flow chart of the present invention.

[0054] Figure 2 This is a flow chart for calculating current characteristics and voltage characteristics of the present invention.

[0055] Figure 3 The present invention provides a flow chart for calculating the current characteristic enhancement coefficient and the voltage characteristic enhancement coefficient.

[0056] Figure 4 The present invention is a flow chart for calculating the current cosine similarity matrix and the voltage cosine similarity matrix. DETAILED DESCRIPTION

[0057] The present invention is further described below in conjunction with the accompanying drawings:

[0058] The present invention provides a method for predicting the remaining service life of a battery based on similarity feature enhancement, comprising the following steps:

[0059] S1. Real-time extraction of current data during battery charge and discharge cycle X S and voltage data X V ;

[0060] S2. According to the current data X S Determine the current diagram G1(X S1 , A1), and according to the voltage data X V Determine the voltage diagram G2(X V1 ,A2);

[0061] Among them, X S1 and X V1 Respectively represent the current data X S and the voltage data X V The extracted features, A1 and A2 represent the current adjacency matrix and voltage adjacency matrix respectively;

[0062] S3. Design a similarity feature enhancement module, and input the current graph G1 and the voltage graph G2 into the similarity feature enhancement module to obtain a comprehensive feature H out ;

[0063] S4. Real-time processing of the comprehensive feature H out The remaining service life of the battery is obtained by inputting the information into the prediction module. The above method can improve the accuracy of the prediction of the remaining service life of the battery.

[0064] In this embodiment, in step S1, the current data X during the battery charge and discharge cycle is extracted in real time. S and voltage data X V ; The extracted current data X S and voltage data X V Perform preprocessing, such as data cleaning and denoising. Extracting real-time data helps predict the remaining service life of the battery in real time; preprocessing the data can eliminate errors, correct outliers in the data, and remove noise. Among them, preprocessing operations such as data cleaning and denoising are existing technologies and will not be described in detail here.

[0065] In this embodiment, in step S2, according to the current data X S Determine the current diagram G1(X S1 , A1), and according to the voltage data X V Determine the voltage diagram G2(X V1 ,A2);

[0066] Among them, X S1 and X V1 Respectively represent the current data X Sand the voltage data X V The extracted features, A1 and A2 represent the current adjacency matrix and voltage adjacency matrix respectively;

[0067] The feature X S1 and X V1 Calculated by the following formula:

[0068] X S1 =MLP(X S )

[0069] X V1 =MLP(X V )

[0070] Among them, MLP(X S ) indicates that the current data X is extracted using a multi-layer perceptron S The features in MLP(X V ) indicates that the voltage data X is extracted using a multi-layer perceptron V Features in;

[0071] Multilayer perceptrons are able to handle complex nonlinear relationships, thus better capturing complex patterns in data. In power systems, current and voltage signals often contain rich nonlinear features, which are crucial for fault diagnosis and prediction. It is very difficult to directly classify and identify the original current and voltage data, so feature extraction is needed to simplify the task and improve the accuracy and efficiency of classification. MLP can significantly improve the accuracy of classification results by extracting features layer by layer. Among them, the process of extracting features using multilayer perceptrons is an existing technology and will not be described here.

[0072] The current adjacency matrix A1 and the voltage adjacency matrix A2 are determined by the following method:

[0073] Calculate feature X S1 With current data X S The cosine similarity of and based on feature X S1 With current data X S The cosine similarity constructs the current adjacency matrix A1;

[0074] Calculate feature X V1 With voltage data X V The cosine similarity of and based on feature X V1 With voltage data X VThe voltage adjacency matrix A2 is constructed by the cosine similarity. Among them, calculating cosine similarity is a prior art, and its calculation process is not repeated here. Cosine similarity only cares about the direction of the vector and is not affected by the length of the vector, which makes it particularly effective when processing sparse vectors. For the data in the current (voltage), there may be current (voltage) changes of different amplitudes. The cosine similarity can ignore these amplitude differences and focus on the direction of the current (voltage) change. When using cosine similarity to construct the current (voltage) graph, it has the advantages of normalized measurement, no influence of vector length, low computational complexity and high directional sensitivity, which can effectively help analyze and predict current (voltage) behavior.

[0075] In this embodiment, in step S3, a similarity feature enhancement module is designed, and the current graph G1 and the voltage graph G2 are input into the similarity feature enhancement module to obtain a comprehensive feature H. out The feature enhancement module obtains the comprehensive feature H through the following steps: out :

[0076] S31. Using three layers of graph convolution layers to extract features from the current graph G1 and the voltage graph G2, respectively, to obtain current features and voltage features of each layer; Figure 2 As shown;

[0077] The current characteristic and the voltage characteristic are calculated by the following formula:

[0078]

[0079]

[0080] in, represents the current characteristics output after the current graph undergoes the lth graph convolution, It represents the voltage features output after the voltage graph undergoes the lth graph convolution. l=1,2,3 correspond to three layers of graph convolution respectively. When l=1, X S1 and X V1 Respectively represent the current data X S and the voltage data X V The extracted features, A1 and A2 represent the current adjacency matrix and voltage adjacency matrix respectively. and They represent the normalization operation of the adjacency matrix, and They represent the weight matrices of current and voltage in the l-1 layer respectively, and ReLU represents the activation function.

[0081] Graph convolution operates directly on the nodes of the graph, taking full account of the structural characteristics of the graph. This means that when processing current and voltage graphs, graph convolution can better capture the complex relationships and dependencies between nodes, thereby extracting more meaningful features; Graph Convolution Network (GCN) can automatically learn and extract low-level to high-level features in graph data, which is very beneficial for feature extraction of current and voltage graphs. With this ability, GCN can effectively extract key feature information from complex electrical systems; Graph Convolution Network has significant advantages in processing current and voltage graphs. This ability enables GCN to better adapt and extract features when processing complex electrical system data.

[0082] S32. Calculate the current characteristic enhancement coefficient and the voltage characteristic enhancement coefficient between each layer, including the current characteristic enhancement coefficient and the voltage characteristic enhancement coefficient between the first layer and the second layer, the first layer and the third layer, and the second layer and the third layer; Figure 3 As shown;

[0083] The current characteristic enhancement coefficient a S12 、a S13 and a S23 Calculated by the following formula:

[0084]

[0085]

[0086]

[0087] Among them, sigmoid represents the sigmoid function, A S12 , A S13 and A S23 are the current cosine similarity matrices, R S1 , R S2 and R S3 Respectively represent the cosine feature similarity of the current graph in the first convolutional layer, the second convolutional layer, and the third convolutional layer, L1, L2, and L3 represent the first layer, the second layer, and the third layer, L1=1, L2=2, and L3=3;

[0088] The voltage characteristic enhancement factor a V12 、a V13 and a V23 The calculation principle and the current characteristic enhancement coefficient a S12 、a S13 and a S23 The calculation principle is the same, and the calculation formula is as follows:

[0089]

[0090]

[0091]

[0092] Among them, A V12 , A V13 and A V23 are the voltage cosine similarity matrices, R V1 , R V2 and R V3 They represent the cosine feature similarity of the voltage graph in the first, second, and third convolutional layers, respectively.

[0093] Further, the calculation process of the current cosine similarity matrix and the voltage cosine similarity matrix is ​​as follows: Figure 4 As shown;

[0094] The current cosine similarity matrix A S12 , A S13 and A S23 Calculated by the following formula:

[0095]

[0096]

[0097]

[0098] in, and Respectively represent the norms of the first layer current characteristics, the second layer current characteristics and the third layer current characteristics;

[0099] The voltage cosine similarity matrix A V12 , A V13 and A V23 The calculation principle of is the same as that of the current cosine similarity matrix, and the calculation formula is as follows:

[0100]

[0101]

[0102]

[0103] in, and They represent the norms of the first layer voltage features, the second layer voltage features, and the third layer voltage features respectively.

[0104] If we directly calculate the cosine similarity between the current graph convolution and the voltage graph convolution layers, we can calculate the current cosine similarity matrix A S12, A S13 and A S23 , and the voltage cosine similarity matrix A V12 , A V13 and A V23 ; The greater the similarity between layer features, the better the corresponding node features and the features of the surrounding neighboring nodes, thus generating similarity. However, this situation ignores the fact that the high similarity may also be due to the fact that the feature is redundant and does not generate feature interactions with other nodes. In order to solve the aforementioned problem, the cosine feature similarity of each layer of graph convolution features is calculated by cosine similarity to obtain the node connection status of the features of this layer. The cosine feature similarity difference is multiplied by the above-mentioned corresponding cosine similarity matrix, which can effectively avoid the fact that the high similarity is caused by redundant features. Among them, by calculating the feature similarity difference between corresponding layers, the feature similarity with a smaller difference means that the corresponding node feature is redundant. S33. Calculate the enhanced feature H between the first layer and the second layer according to the current features and voltage features of each layer, as well as the current feature enhancement coefficient and voltage feature enhancement coefficient between each layer. 12 , the enhanced feature H between the first and third layers 13 , and the enhanced feature H between the second and third layers 23 ;

[0105] The enhanced feature H 12 , H 13 and H 23 Calculated by the following formulas:

[0106]

[0107]

[0108]

[0109] Among them, a S12 、a S13 and a S23 Respectively represent the current characteristic enhancement coefficients between the first layer and the second layer, the first layer and the third layer, and the second layer and the third layer; a V12 、a V13 and a V23 Respectively represent the voltage characteristic enhancement coefficients between the first layer and the second layer, the first layer and the third layer, and the second layer and the third layer; Indicates that the first layer current features are spliced ​​with the second layer current features. Indicates that the voltage features of the first layer are concatenated with the voltage features of the second layer; It means that the first layer current features are spliced ​​with the third layer current features. Indicates that the voltage characteristics of the first layer are concatenated with the voltage characteristics of the third layer; It means that the second layer current features are spliced ​​with the third layer current features. It indicates that the voltage features of the second layer are concatenated with the voltage features of the third layer. The concatenated layer and layer features are multiplied based on the feature enhancement coefficient to achieve feature enhancement.

[0110] S34. The enhanced feature H between the first layer and the second layer 12 , the enhanced feature H between the first and third layers 13 and the enhanced feature H between the second and third layers 23 Splice and get the comprehensive feature H out ;

[0111] The comprehensive feature H out The expression is as follows:

[0112] H out =H 12 ||H 13 ||H 23

[0113] Among them, H 12 ||H 13 ||H 23 Represents the enhanced feature H 12 , H 13 and H 23 Through the above method, the feature enhancement coefficient and the splicing features between layers are enhanced, and then the features of the current and voltage corresponding layers are added to realize the hierarchical fusion operation of current and voltage features.

[0114] In this embodiment, in step S4, the comprehensive feature H is out Input into the prediction module to obtain the remaining service life of the battery.

[0115] The prediction module may be a fully connected layer, as well as a neural network and a deep learning model capable of predicting various types, and no specific limitation is imposed on the prediction module herein.

[0116] Accordingly, the present invention further provides an electronic device, comprising:

[0117] A memory and a processor, wherein the memory is used to store a computer program, and when the computer program is executed by the processor, the method for predicting the remaining service life of a battery based on similarity feature enhancement is implemented.

[0118] Correspondingly, the present invention also proposes a readable storage medium, in which computer instructions are stored. When the computer instructions are executed by a processor, the above-mentioned battery remaining service life prediction method based on similarity feature enhancement is implemented.

[0119] In the above embodiments of the present application, the description of each embodiment has its own emphasis. For parts that are not described in detail in a certain embodiment, please refer to the relevant description of other embodiments.

[0120] Those skilled in the art can understand that all or part of the processes in the above method embodiments can be completed by instructing the relevant hardware through a computer program, and the computer program can be stored in a non-volatile computer-readable storage medium. When the computer program is executed, it can include the processes of the embodiments of the above methods. Among them, any reference to memory, storage, database or other media used in the embodiments provided in this application can include non-volatile and / or volatile memory. Non-volatile memory can include read-only memory (ROM), programmable ROM (PROM), electrically programmable ROM (EPROM), electrically erasable programmable ROM (EEPROM) or flash memory. Volatile memory can include random access memory (RAM) or external cache memory. As an illustration and not limitation, RAM is available in many forms, such as static RAM (SRAM), dynamic RAM (DRAM), synchronous DRAM (SDRAM), double data rate SDRAM (DDRSDRAM), enhanced SDRAM (ESDRAM), synchronous link (Synchlink), DRAM (SLDRAM), memory bus (Rambus), direct RAM (RDRAM), direct memory bus dynamic RAM (DRDRAM), and memory bus dynamic RAM (RDRAM).

[0121] Finally, it should be noted that the above embodiments are only used to illustrate the technical solution of the present invention rather than to limit it. Although the present invention has been described in detail with reference to the preferred embodiments, those skilled in the art should understand that the technical solution of the present invention can be modified or replaced by equivalents without departing from the purpose and scope of the technical solution of the present invention, which should be included in the scope of the claims of the present invention.

Claims

1. A method for predicting remaining battery life based on similarity feature enhancement, characterized in that: The following steps are involved: S1. Real-time extraction of current data during battery charge and discharge cycle X S and voltage data X V ; S2. According to the current data X S Determine the current diagram G1(X S1 , A1), and according to the voltage data X V Determine the voltage diagram G2(X V1 ,A2); Among them, X S1 and X V1 Respectively represent the current data X S and the voltage data X V The extracted features, A1 and A2 represent the current adjacency matrix and voltage adjacency matrix respectively; S3. Design a similarity feature enhancement module, and input the current graph G1 and the voltage graph G2 into the similarity feature enhancement module to obtain a comprehensive feature H out ; S4. Real-time processing of the comprehensive feature H out The data is input into the prediction module to obtain the remaining service life of the battery.

2. The method for predicting remaining battery life based on similarity feature enhancement according to claim 1, characterized in that: The feature enhancement module obtains the comprehensive feature H through the following steps out : S31. Using three layers of graph convolution layers to extract features from the current graph G1 and the voltage graph G2 respectively, to obtain current features and voltage features of each layer; S32. Calculate the current characteristic enhancement coefficient and the voltage characteristic enhancement coefficient between each layer, including the current characteristic enhancement coefficient and the voltage characteristic enhancement coefficient between the first layer and the second layer, the first layer and the third layer, and the second layer and the third layer; S33. Calculate the enhancement characteristic H between the first layer and the second layer according to the current characteristics and voltage characteristics of each layer, and the current characteristic enhancement coefficient and voltage characteristic enhancement coefficient between the layers. 12 , the enhanced feature H between the first and third layers 13 , and the enhanced feature H between the second and third layers 23 ; S34. The enhancement feature H between the first layer and the second layer 12 , the enhanced feature H between the first layer and the third layer 13 and the enhancement feature H between the second and third layers 23 Splice and get the comprehensive feature H out ; The comprehensive feature H out The expression is as follows: H out =H 12 ||H 13 ||H 23 Among them, H 12 ||H 13 ||H 23 Represents the enhanced feature H 12 , H 13 and H 23 Make splicing.

3. The method for predicting remaining battery life based on similarity feature enhancement according to claim 2, characterized in that: The enhanced feature H 12 , H 13 and H 23 Calculated by the following formulas: Among them, a S12 、a S13 and a S23 represent the current characteristic enhancement coefficients between the first layer and the second layer, the first layer and the third layer, and the second layer and the third layer respectively; a V12 、a V13 and a V23 Respectively represent the voltage characteristic enhancement coefficients between the first layer and the second layer, the first layer and the third layer, and the second layer and the third layer; Indicates that the first layer current characteristics and the second layer current characteristics are spliced, Indicates that the first layer voltage feature and the second layer voltage feature are spliced; Indicates that the first layer current feature and the third layer current feature are spliced, Indicates that the first layer voltage feature and the third layer voltage feature are spliced; Indicates that the second layer current feature is spliced ​​with the third layer current feature, It indicates that the second layer voltage feature is spliced ​​with the third layer voltage feature.

4. The method for predicting remaining battery life based on similarity feature enhancement according to claim 3 is characterized in that: The current characteristic enhancement coefficient a S12 、a S13 and a S23 Calculated by the following formula: Among them, sigmoid represents the sigmoid function, A S12 , A S13 and A S23 are the current cosine similarity matrices, R S1 , R S2 and R S3 Respectively represent the cosine feature similarity of the current graph in the first convolutional layer, the second convolutional layer, and the third convolutional layer, L1, L2, and L3 represent the first layer, the second layer, and the third layer, L1=1, L2=2, and L3=3; The voltage characteristic enhancement factor a V12 、a V13 and a V23 The calculation principle and the current characteristic enhancement coefficient a S12 、a S13 and a S23 The calculation principle is the same as that of . During the calculation, the current feature similarity and current cosine similarity matrix in the current feature enhancement coefficient calculation formula are replaced with the corresponding voltage feature similarity and voltage cosine similarity matrix.

5. The method for predicting remaining battery life based on similarity feature enhancement according to claim 4 is characterized in that: The current cosine similarity matrix A S12 , A S13 and A S23 Calculated by the following formula: in, and Respectively represent the norms of the first layer current characteristics, the second layer current characteristics and the third layer current characteristics; The calculation principle of the voltage cosine similarity matrix is ​​the same as the calculation principle of the current cosine similarity matrix. During calculation, the current characteristics and the norm of the current characteristics in the formula are replaced with the corresponding voltage characteristics and the norm of the voltage characteristics.

6. The method for predicting remaining battery life based on similarity feature enhancement according to claim 3 or claim 5, characterized in that: The current characteristic and the voltage characteristic are calculated by the following formula: in, It represents the current characteristics output by the current graph after the lth graph convolution. It represents the voltage features output after the voltage graph undergoes the lth graph convolution. l=1,2,3 correspond to three layers of graph convolution respectively. When l=1, X S1 and X V1 Respectively represent the current data X S and the voltage data X V The extracted features, A1 and A2 represent the current adjacency matrix and the voltage adjacency matrix respectively, and They represent the normalization operation of the adjacency matrix, and They represent the weight matrices of current and voltage in the l-1 layer respectively, and ReLU represents the activation function.

7. The method for predicting remaining battery life based on similarity feature enhancement according to claim 6, characterized in that: The current adjacency matrix A1 and the voltage adjacency matrix A2 are determined by the following method: Calculate feature X S1 With current data X S The cosine similarity of and based on feature X S1 With current data X S The cosine similarity of constructs the current adjacency matrix A1; Calculate feature X V1 With voltage data X V The cosine similarity of and based on feature X V1 With voltage data X V The voltage adjacency matrix A2 is constructed based on the cosine similarity.

8. The method for predicting remaining battery life based on similarity feature enhancement according to claim 7, characterized in that: The feature X S1 and X V1 Calculated by the following formula: X S1 =MLP(X S ) X V1 =MLP(X V ) Among them, MLP(X S ) indicates that the current data X is extracted using a multi-layer perceptron S The features in MLP(X V ) indicates that the voltage data X is extracted using a multi-layer perceptron V Features in .

9. An electronic device, characterized in that: include: A memory and a processor, wherein the memory is used to store a computer program, and when the computer program is executed by the processor, the method for predicting the remaining service life of a battery based on similarity feature enhancement described in any one of claims 1 to 8 is implemented.

10. A readable storage medium, characterized in that: The readable storage medium stores computer instructions, and when the computer instructions are executed by a processor, the method for predicting the remaining useful life of a battery based on similarity feature enhancement according to any one of claims 1 to 8 is implemented.

Images