Lithium battery operating data feature map extraction method, and apparatus, medium and product

Abstract

A lithium battery operating data feature map extraction method, and an apparatus, a medium and a product. The method comprises: acquiring operating data of a lithium battery of an energy storage system; applying a sliding window to select target operating data from among the operating data, so as to obtain a plurality of target data groups; ranking a plurality of pieces of operating data in each target data group according to the sequence of preset index moments, so as to obtain an initial matrix; performing mean-variance normalization processing on each column of the initial matrix, so as to obtain a normalized matrix; calculating a covariance matrix of the normalized matrix; performing singular value decomposition on the covariance matrix, so as to obtain a feature vector matrix; on the basis of the feature vector matrix, calculating a dimension reduction matrix corresponding to each target data group; scaling the dimension reduction matrix to construct a three-dimensional feature map matrix; and converting the three-dimensional feature map matrix into an image, so as to obtain a feature image of the lithium battery. Thus, the comprehensiveness, universality and robustness of feature extraction can be improved.

Description

A lithium battery operation data feature map extraction method, device, medium and product

[0001] The present application claims priority from the Chinese patent application No. 202410657865.9 filed on May 27, 2024, and entitled "A lithium battery operation data feature map extraction method, device, medium and product", the whole content of which is incorporated herein by reference. TECHNICAL FIELD

[0002] The present application relates to the field of energy storage of new energy power system, in particular to a lithium battery operation data feature map extraction method, device, medium and product. BACKGROUND

[0003] With the continuous expansion of new energy power generation capacity, the proportion of new energy power generation in the power grid is higher and higher. However, due to the small single machine capacity, large number, scattered distribution of new energy power generation, and the characteristics of significant intermittency, volatility, randomness, etc., high proportion of new energy grid connection will inevitably bring unprecedented challenges to power system supply and demand balance, safety and stability control, etc. Energy storage system is a key link to adjust the imbalance between new energy power generation and power system supply and demand, energy management and optimization. The core of the energy storage system is a large number of lithium ion batteries connected by series and parallel connection structure. Lithium batteries realize charging and discharging through internal electrochemical reaction. Due to the complexity of internal electrochemical reaction of lithium battery and the instability of environmental state, it is very difficult to directly describe the internal state change of the battery. In some cases, the key data features are extracted by collecting lithium battery operation data, and then the lithium battery operation state is evaluated by analyzing the feature data.

[0004] In summary, the lithium battery operation data feature extraction method has the following shortcomings: 1. The feature description is not comprehensive enough. The feature extraction method can only partially describe the data features in time domain and index dimension, and cannot fully represent the battery operation state. 2. Poor universality. Most of the feature extraction methods are designed for specific battery types or application scenarios, and lack of universality; 3. Poor robustness. The feature extraction method is often sensitive to noise and abnormal data, and has poor robustness; 4. Low efficiency of downstream algorithm adaptation. Due to the large scale of lithium battery operation data and the complexity of downstream algorithm, the feature extraction method often cannot meet the efficiency requirements of downstream data analysis algorithm. SUMMARY

[0005] The purpose of the present application is to provide a lithium battery operation data feature map extraction method, device, medium and product, which can improve the comprehensiveness, universality and robustness of feature extraction.

[0006] To achieve the above purpose, the present application provides the following solutions:

[0007] In a first aspect, a feature map extraction method for lithium battery operation data is provided. The method comprises: obtaining operation data of a lithium battery of an energy storage system; the operation data comprises voltage, current, temperature and charge and discharge capacity corresponding to each preset index moment; applying a sliding window, selecting target operation data from the operation data according to a preset size and a preset data interval of the sliding window, and obtaining a plurality of target data groups; sorting a plurality of operation data in each target data group according to the order of the preset index moment, and obtaining an initial matrix corresponding to each target data group; the row vector of the initial matrix is the operation data; the column vector of the initial matrix comprises a voltage column vector corresponding to the voltage, a current column vector corresponding to the current, a temperature column vector corresponding to the temperature, and a charge and discharge column vector corresponding to the charge and discharge capacity; performing mean-variance normalization processing on each column of the initial matrix to obtain a normalized matrix corresponding to each target data group; calculating a covariance matrix of the normalized matrix; performing singular value decomposition on the covariance matrix to obtain a feature vector matrix corresponding to each target data group; calculating a dimension reduction matrix corresponding to each target data group according to the feature vector matrix; scaling the dimension reduction matrix to construct a three-dimensional feature map matrix corresponding to each target data group; and converting the three-dimensional feature map matrix into an image to obtain a feature image of the lithium battery.

[0008] In a second aspect, a computer device is provided. The computer device comprises a memory, a processor, a computer program stored in the memory and executable on the processor, and the processor executes the computer program to implement the feature map extraction method for lithium battery operation data.

[0009] In a third aspect, a computer readable storage medium is provided. The computer readable storage medium stores a computer program, and the computer program is executed by a processor to implement the feature map extraction method for lithium battery operation data.

[0010] In a fourth aspect, a computer program product is provided. The computer program product comprises a computer program, and the computer program is executed by a processor to implement the feature map extraction method for lithium battery operation data.

[0011] According to the embodiments provided in the present application, the following technical effects are disclosed.

[0012] The feature map extraction method improves the comprehensiveness of lithium battery operation data feature extraction, improves the universality and robustness of lithium battery operation data feature extraction through data processing and data standardization method, and improves the execution efficiency of downstream algorithm through feature dimension reduction. BRIEF DESCRIPTION OF DRAWINGS

[0013] To more clearly illustrate the technical solutions in the embodiments of this application or the prior art, the drawings used in the embodiments will be briefly introduced below. Obviously, the drawings described below are only some embodiments of this application. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0014] Figure 1 is a schematic diagram of the lithium battery operation data feature map extraction method provided in the embodiment of this application.

[0015] Figure 2 is a visualization curve of the battery operation index data in the time direction provided in the embodiment of this application.

[0016] Figure 3 is a diagram of the internal structure of a computer device. Detailed Implementation

[0017] The technical solutions of the embodiments of this application will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of this application, and not all embodiments. Based on the embodiments of this application, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of this application.

[0018] The purpose of this application is to provide a method, apparatus, medium, and product for extracting feature maps from lithium battery operating data, aiming to improve the comprehensiveness, versatility, and robustness of feature extraction.

[0019] To make the above-mentioned objectives, features and advantages of this application more apparent and understandable, the application will be further described in detail below with reference to the accompanying drawings and specific embodiments.

[0020] Example 1

[0021] As shown in Figure 1, the lithium battery operation data feature map extraction method in this embodiment includes:

[0022] Step S1: Acquire the operating data of the lithium battery in the energy storage system; the operating data includes the voltage, current, temperature, and charge / discharge amount corresponding to each preset indicator time. In this embodiment, the voltage data is obtained by the PXI7065 battery data acquisition card, the temperature is obtained by the FBG fiber optic sensor, and the charge / discharge amount data is estimated by a coulomb meter + EIS analyzer.

[0023] Step S2: Apply a sliding window, and select target running data from the running data according to the preset size and preset data interval of the sliding window to obtain multiple target data groups.

[0024] In practical applications, as shown in FIG. 2, the input includes a feature map length m, a sliding window data interval p, and battery operation data containing index time, voltage, current, temperature, and charge and discharge amount, with a data length of n. The sliding window data interval refers to the step length, and the window size is the feature map length m.

[0025] Step S3: Sort the multiple operation data in each target data group according to the order before and after the preset index time, to obtain an initial matrix corresponding to each target data group; the row vector of the initial matrix is operation data; and the column vector of the initial matrix includes a voltage column vector corresponding to voltage, a current column vector corresponding to current, a temperature column vector corresponding to temperature, and a charge and discharge column vector corresponding to charge and discharge amount.

[0026] As a specific implementation, the method further includes:

[0027] When there is a null value in the initial matrix, a linear difference algorithm is applied to determine the numerical value of the null value.

[0028] In practical applications, the input data is arranged in ascending order according to the time column, and linear difference is performed on the null value position, as shown below:

[0029] wherein D k , V k , I k , T k , and Q k are the time, voltage, current, temperature, and charge and discharge amount of the kth row after sorting, and nan represents a null value. After data processing, the voltage, current, temperature, and charge and discharge amount are extracted to form an n x 4 matrix A = (V, I, T, Q), wherein V, I, T, and Q are n x 1 vectors after processing.

[0030] Step S4: Perform mean-variance normalization processing on each column of the initial matrix to obtain a normalized matrix corresponding to each target data group.

[0031] In practical applications, mean-variance normalization is performed on each column of the matrix A, and the formula is as follows:

[0032] wherein A kj is the element of the kth row and jth column of the matrix A.

[0033] Step S5: Calculate the covariance matrix of the normalized matrix.

[0034] In practical applications, the covariance matrix of the matrix S is calculated as follows:

[0035] wherein s k is a 1 x 4 row vector, representing the kth row of the matrix S, and the matrix is a 1x4 row vector, representing the average value vector of the matrix S in the column direction, represents the transpose matrix of

[0036] Step S6: Singular value decomposition is performed on the covariance matrix to obtain a feature vector matrix corresponding to each target data group.

[0037] In practical applications, singular value decomposition is performed on the covariance matrix W, W=UΣV, where U is a 4x4 left singular matrix; Σ is a singular value matrix, the main diagonal is a singular value, and the other elements are all 0; V is an n*n right singular matrix; U, Σ, and V can be obtained by solving the following equation group:

[0038] where h=1, 2, 3, 4, g=1, 2, …n, λ h , λ g is a singular value, that is, the main diagonal element value of Σ; u h , v g are feature vectors, respectively, and are column vectors of matrices U and V; A T is the transpose matrix of A.

[0039] Step S7: According to the feature vector matrix, the dimension reduction matrix corresponding to each target data group is calculated.

[0040] Step S7 includes:

[0041] Step S71: Arrange the feature vectors of the feature vector matrix in descending order according to the corresponding eigenvalues by column to obtain an updated feature vector matrix.

[0042] Step S72: According to the updated feature vector matrix, the dimension reduction matrix is calculated.

[0043] In practical applications, the feature vectors are arranged in descending order according to the corresponding eigenvalues from left to right by column, and U, Σ, and V are updated to calculate the dimension reduction matrix: B=U τ Σ τ V τ .

[0044] where U τ , Σ τ , V τ are matrices composed of the first τ rows of U, Σ, and V, respectively. When τ=3, B is an n*3 matrix.

[0045] The larger the eigenvalue, the more information the corresponding feature vector contains about the original matrix. The feature vectors are arranged in descending order from left to right to facilitate subsequent selection of feature vectors according to the dimension reduction dimension. The feature vectors and eigenvalues are in a one-to-one correspondence, and the feature vectors are reordered according to the eigenvalue order.

[0046] Step S8: scaling the dimensionality reduction matrix to construct a three-dimensional feature map matrix corresponding to each target data set.

[0047] Step S8 includes:

[0048] Step S81: scaling the dimensionality reduction matrix by column to obtain the scaled values of each element in each column.

[0049] Step S82: obtaining the scaled matrix according to the scaled values of each element in each column.

[0050] In practical applications, the matrix B is scaled by column to the range [0, 255], since the intensity value of the RGB color system is in the range [0, 255], which is shown as follows:

[0051] wherein are the maximum and minimum values of the i-th row and j-th column of the matrix B, respectively.

[0052] Step S83: determining a three-dimensional feature map matrix according to a preset feature map length from the scaled matrix.

[0053] Selecting a starting position i, i≤n-2(m-1)p, constructing a three-dimensional feature map matrix of m×m×3 according to the feature map length m: i =(X i Y i …… Z i ).

[0054] wherein:

[0055] wherein B' i,j is the i-th row and j-th column element of the matrix B'.

[0056] wherein, i refers to the i-th row of the matrix B', which is not related to the sliding window. According to the construction method of (X i , Y i , Z i ), the subscript i+2(m-1)p in the last element B'i+2(m-1)p,3 of the matrix should be less than or equal to n, so i≤n-2(m-1)p.

[0057] The reason for constructing a three-dimensional feature map matrix of m×m×n is that according to the RGB image system rule, m×m is the length and width of the picture, which also represents the number of pixel points, and each pixel point is represented by a three-tuple (R, G, B) respectively representing the (red, green, blue) color value intensity.

[0058] Step S9: converting the three-dimensional feature map matrix into an image to obtain a feature image of the lithium battery.

[0059] Step S9 comprises:

[0060] Step S91: convert the three-dimensional feature map matrix into an image according to the RGB rule to obtain a feature image corresponding to each target data group.

[0061] In actual application, the feature map matrix C i is converted into an image, that is, a feature image of the battery at the i moment.

[0062] Step S92: determine a feature image corresponding to the running data of the lithium battery according to the feature image corresponding to each target data group.

[0063] By extracting the battery running data feature image, the battery health state can be explicitly presented, which is used for intuitive visual diagnosis of the battery, earlier discovery of aging or thermal runaway signs, and improvement of fault detection (internal short circuit, thermal runaway) efficiency and accuracy. It can be used for battery life prediction under complex working conditions, optimization of battery management system, and assistance in intelligent early warning and precise maintenance of electric vehicles and energy storage power stations.

[0064] The beneficial effects of the present application are as follows:

[0065] 1. A lithium battery running data feature map extraction method is provided.

[0066] 2. The problem of one-sided feature extraction of the battery data feature extraction method is solved.

[0067] 3. The problem of poor universality of the battery data feature extraction method is solved.

[0068] 4. The robustness of the battery data feature extraction method is improved.

[0069] 5. The execution efficiency of the downstream algorithm is improved.

[0070] 6. The RGB image is used to represent the comprehensive features of the battery running state, improve the interpretability of the features, and better adapt the battery data to efficient deep learning algorithms.

[0071] Embodiment 2

[0072] A computer device comprises a memory, a processor to store a computer program on the memory and executable on the processor, and the processor executes the computer program to realize the lithium battery running data feature map extraction method in embodiment 1.

[0073] Embodiment 3

[0074] A computer readable storage medium having a computer program stored thereon, wherein the computer program is executed by a processor to realize the lithium battery running data feature map extraction method in embodiment 1.

[0075] Embodiment 4

[0076] A computer program product comprising a computer program which, when executed by a processor, implements the lithium battery operation data feature map extraction method in embodiment 1.

[0077] Embodiment 5

[0078] A computer device, which can be a database, can have an internal structure diagram as shown in FIG. 3. The computer device includes a processor, a memory, an input / output interface (I / O) and a communication interface. Among them, the processor, the memory and the input / output interface are connected through a system bus, and the communication interface is connected to the system bus through the input / output interface. Among them, the processor of the computer device is used to provide computing and control capabilities. The memory of the computer device includes a non-volatile storage medium and an internal memory. The non-volatile storage medium stores an operating system, a computer program and a database. The internal memory provides an environment for the operating system and the computer program in the non-volatile storage medium to run. The database of the computer device is used to store transactions to be processed. The input / output interface of the computer device is used to exchange information between the processor and external devices. The communication interface of the computer device is used to communicate with external terminals through network connection. The computer program is executed by the processor to implement the lithium battery operation data feature map extraction method in embodiment 1.

[0079] It should be noted that the object information (including but not limited to object device information, object personal information, etc.) and data (including but not limited to data for analysis, stored data, displayed data, etc.) involved in the present application are all information and data authorized by the object or fully authorized by all parties, and the collection, use and processing of related data need to comply with relevant laws, regulations and standards of relevant countries and regions.

[0080] Those skilled in the art can understand that all or part of the processes in the above-mentioned embodiment methods can be completed by instructing the relevant hardware through a computer program. The computer program can be stored in a non-volatile computer readable storage medium, and when the computer program is executed, the processes of the above-mentioned embodiments of the methods can be included. Any reference to memory, database or other medium used in the embodiments provided in the present application can include at least one of non-volatile and volatile memory. The non-volatile memory can include read-only memory (Read-Only Memory, ROM), magnetic tape, floppy disk, flash memory, optical storage, high-density embedded non-volatile memory, resistive memory (ReRAM), magnetoresistive random access memory (Magnetoresistive Random Access Memory, MRAM), ferroelectric memory (Ferroelectric Random Access Memory, FRAM), phase change memory (Phase Change Memory, PCM), graphene memory, etc. The volatile memory can include random access memory (Random Access Memory, RAM) or external cache memory, etc. As an illustration but not limitation, the RAM can be in various forms, such as static random access memory (Static Random Access Memory, SRAM) or dynamic random access memory (Dynamic Random Access Memory, DRAM), etc. The database involved in the embodiments provided in the present application can include at least one of a relational database and a non-relational database. The non-relational database can include a distributed database based on a block chain, etc., without being limited thereto. The processor involved in the embodiments provided in the present application can be a general-purpose processor, a central processing unit, a graphics processing unit, a digital signal processor, a programmable logic device, a data processing logic device based on quantum computing, etc., without being limited thereto.

[0081] Any combination of the technical features of the above embodiments can be made, and in order to make the description concise, all possible combinations of the technical features in the above embodiments are not described, however, as long as the combination of the technical features does not exist contradictory, it should be considered as the scope of the present application.

[0082] The principles and implementation modes of the present application are described by applying specific examples herein, and the above-mentioned embodiments are only used to help understand the method and its core idea of the present application; at the same time, for those skilled in the art, according to the idea of the present application, the specific implementation mode and application range will be changed. In conclusion, the content of the present application should not be understood as a limitation.

Claims

1. A method for extracting feature maps from lithium battery operating data, characterized in that, The method includes: Acquire the operating data of the lithium battery in the energy storage system; the operating data includes the voltage, current, temperature and charge / discharge amount at each preset index time. Using a sliding window, select target running data from the running data according to the preset size and preset data interval of the sliding window to obtain multiple target data groups; The multiple running data in each target data group are sorted according to the chronological order of the preset index time to obtain an initial matrix corresponding to each target data group; the row vector of the initial matrix is ​​the running data; the column vector of the initial matrix includes the voltage column vector corresponding to the voltage, the current column vector corresponding to the current, the temperature column vector corresponding to the temperature, and the charge / discharge quantity column vector corresponding to the charge / discharge quantity. The mean and variance are normalized for each column of the initial matrix to obtain the normalized matrix corresponding to each target data group; Calculate the covariance matrix of the normalized matrix; Singular value decomposition is performed on the covariance matrix to obtain the feature vector matrix corresponding to each target data group. Based on the feature vector matrix, calculate the dimensionality reduction matrix corresponding to each target data group; The dimensionality reduction matrix is ​​scaled to construct a three-dimensional feature map matrix corresponding to each target data group; The three-dimensional feature map matrix is ​​converted into an image to obtain the feature image of the lithium battery.

2. The lithium battery operation data feature map extraction method according to claim 1, characterized in that, The method further includes: When there are null values ​​in the initial matrix, the linear interpolation algorithm is used to determine the value of the null value.

3. The method for extracting feature maps of lithium battery operating data according to claim 1, characterized in that, Based on the feature vector matrix, calculate the dimensionality reduction matrix corresponding to each target data group, including: Arrange the eigenvectors of the eigenvector matrix in descending column order according to their corresponding eigenvalues ​​to obtain the updated eigenvector matrix; Calculate the dimensionality reduction matrix based on the updated feature vector matrix.

4. The method for extracting feature maps of lithium battery operating data according to claim 1, characterized in that, The dimensionality reduction matrix is ​​scaled to construct a three-dimensional feature map matrix corresponding to each target data group, including: The dimensionality reduction matrix is ​​scaled column-wise to obtain the scaled values ​​of each element in each column; Based on the scaled values ​​of each element in each column, a scaled matrix is ​​obtained; From the scaled matrix, a three-dimensional feature map matrix is ​​determined according to the preset feature map length.

5. The method for extracting feature maps of lithium battery operating data according to claim 1, characterized in that, The three-dimensional feature map matrix is ​​converted into an image to obtain the feature image of the lithium battery, including: The three-dimensional feature map matrix is ​​converted into an image according to the RGB rules to obtain the feature image corresponding to each target data group; Based on the feature image corresponding to each target data group, the feature image corresponding to the operating data of the lithium battery is determined.

6. A computer device, comprising: The memory and processor contain a computer program stored in the memory and executable on the processor, characterized in that the processor executes the computer program to implement the lithium battery operating data feature map extraction method according to any one of claims 1-5.

7. A computer-readable storage medium having a computer program stored thereon, characterized in that, When the computer program is executed by the processor, it implements the lithium battery operating data feature map extraction method according to any one of claims 1-5.

8. A computer program product, comprising a computer program, characterized in that, When the computer program is executed by the processor, it implements the lithium battery operating data feature map extraction method according to any one of claims 1-5.

Images