CN117154256A - Electrochemical repair method for lithium battery - Google Patents
Electrochemical repair method for lithium battery Download PDFInfo
- Publication number
- CN117154256A CN117154256A CN202311342922.6A CN202311342922A CN117154256A CN 117154256 A CN117154256 A CN 117154256A CN 202311342922 A CN202311342922 A CN 202311342922A CN 117154256 A CN117154256 A CN 117154256A
- Authority
- CN
- China
- Prior art keywords
- temperature
- charging process
- time sequence
- feature
- local
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Pending
Links
- 238000000034 method Methods 0.000 title claims abstract description 220
- WHXSMMKQMYFTQS-UHFFFAOYSA-N Lithium Chemical compound [Li] WHXSMMKQMYFTQS-UHFFFAOYSA-N 0.000 title claims abstract description 35
- 229910052744 lithium Inorganic materials 0.000 title claims abstract description 35
- 230000008439 repair process Effects 0.000 title claims abstract description 32
- 230000008569 process Effects 0.000 claims abstract description 183
- 239000013598 vector Substances 0.000 claims description 149
- 239000011159 matrix material Substances 0.000 claims description 75
- 238000005070 sampling Methods 0.000 claims description 25
- 238000013527 convolutional neural network Methods 0.000 claims description 19
- 238000009826 distribution Methods 0.000 claims description 19
- 230000001965 increasing effect Effects 0.000 claims description 13
- 238000000605 extraction Methods 0.000 claims description 12
- 238000003062 neural network model Methods 0.000 claims description 12
- 230000003247 decreasing effect Effects 0.000 claims description 11
- 230000011218 segmentation Effects 0.000 claims description 4
- 238000005457 optimization Methods 0.000 claims 2
- 230000008859 change Effects 0.000 abstract description 31
- 238000012545 processing Methods 0.000 abstract description 13
- 238000011176 pooling Methods 0.000 description 15
- 230000006870 function Effects 0.000 description 14
- 230000010287 polarization Effects 0.000 description 9
- 230000004913 activation Effects 0.000 description 8
- 238000013528 artificial neural network Methods 0.000 description 8
- 239000000284 extract Substances 0.000 description 6
- 238000013507 mapping Methods 0.000 description 6
- 210000002569 neuron Anatomy 0.000 description 6
- 230000002829 reductive effect Effects 0.000 description 6
- QSNQXZYQEIKDPU-UHFFFAOYSA-N [Li].[Fe] Chemical compound [Li].[Fe] QSNQXZYQEIKDPU-UHFFFAOYSA-N 0.000 description 5
- 230000002776 aggregation Effects 0.000 description 4
- 238000004220 aggregation Methods 0.000 description 4
- 238000004458 analytical method Methods 0.000 description 4
- 238000010586 diagram Methods 0.000 description 4
- 238000005516 engineering process Methods 0.000 description 4
- 238000000547 structure data Methods 0.000 description 4
- 238000012549 training Methods 0.000 description 4
- HBBGRARXTFLTSG-UHFFFAOYSA-N Lithium ion Chemical compound [Li+] HBBGRARXTFLTSG-UHFFFAOYSA-N 0.000 description 3
- 230000008901 benefit Effects 0.000 description 3
- 238000004364 calculation method Methods 0.000 description 3
- 239000000306 component Substances 0.000 description 3
- 239000008358 core component Substances 0.000 description 3
- 229910001416 lithium ion Inorganic materials 0.000 description 3
- 230000007246 mechanism Effects 0.000 description 3
- 230000002441 reversible effect Effects 0.000 description 3
- 239000011149 active material Substances 0.000 description 2
- 239000010405 anode material Substances 0.000 description 2
- 238000013136 deep learning model Methods 0.000 description 2
- 238000007599 discharging Methods 0.000 description 2
- 230000000694 effects Effects 0.000 description 2
- 238000004146 energy storage Methods 0.000 description 2
- 230000002427 irreversible effect Effects 0.000 description 2
- GELKBWJHTRAYNV-UHFFFAOYSA-K lithium iron phosphate Chemical compound [Li+].[Fe+2].[O-]P([O-])([O-])=O GELKBWJHTRAYNV-UHFFFAOYSA-K 0.000 description 2
- 238000010801 machine learning Methods 0.000 description 2
- 239000000463 material Substances 0.000 description 2
- 230000036961 partial effect Effects 0.000 description 2
- 238000012300 Sequence Analysis Methods 0.000 description 1
- SOXUFMZTHZXOGC-UHFFFAOYSA-N [Li].[Mn].[Co].[Ni] Chemical compound [Li].[Mn].[Co].[Ni] SOXUFMZTHZXOGC-UHFFFAOYSA-N 0.000 description 1
- 230000004931 aggregating effect Effects 0.000 description 1
- 238000012512 characterization method Methods 0.000 description 1
- 238000004891 communication Methods 0.000 description 1
- 238000010276 construction Methods 0.000 description 1
- 230000007423 decrease Effects 0.000 description 1
- 238000013135 deep learning Methods 0.000 description 1
- 238000013461 design Methods 0.000 description 1
- 238000001514 detection method Methods 0.000 description 1
- 230000002708 enhancing effect Effects 0.000 description 1
- 230000002349 favourable effect Effects 0.000 description 1
- 230000006872 improvement Effects 0.000 description 1
- 230000001788 irregular Effects 0.000 description 1
- 238000003475 lamination Methods 0.000 description 1
- 238000005065 mining Methods 0.000 description 1
- 238000012986 modification Methods 0.000 description 1
- 230000004048 modification Effects 0.000 description 1
- 238000012544 monitoring process Methods 0.000 description 1
- 238000003058 natural language processing Methods 0.000 description 1
- 238000003012 network analysis Methods 0.000 description 1
- 238000010606 normalization Methods 0.000 description 1
- 238000003672 processing method Methods 0.000 description 1
- 230000001902 propagating effect Effects 0.000 description 1
- 238000011524 similarity measure Methods 0.000 description 1
- 230000005236 sound signal Effects 0.000 description 1
- 230000002123 temporal effect Effects 0.000 description 1
- 238000012546 transfer Methods 0.000 description 1
- 230000009466 transformation Effects 0.000 description 1
- 238000000844 transformation Methods 0.000 description 1
Classifications
-
- H—ELECTRICITY
- H01—ELECTRIC ELEMENTS
- H01M—PROCESSES OR MEANS, e.g. BATTERIES, FOR THE DIRECT CONVERSION OF CHEMICAL ENERGY INTO ELECTRICAL ENERGY
- H01M10/00—Secondary cells; Manufacture thereof
- H01M10/42—Methods or arrangements for servicing or maintenance of secondary cells or secondary half-cells
- H01M10/425—Structural combination with electronic components, e.g. electronic circuits integrated to the outside of the casing
-
- H—ELECTRICITY
- H01—ELECTRIC ELEMENTS
- H01M—PROCESSES OR MEANS, e.g. BATTERIES, FOR THE DIRECT CONVERSION OF CHEMICAL ENERGY INTO ELECTRICAL ENERGY
- H01M10/00—Secondary cells; Manufacture thereof
- H01M10/42—Methods or arrangements for servicing or maintenance of secondary cells or secondary half-cells
- H01M10/44—Methods for charging or discharging
-
- H—ELECTRICITY
- H01—ELECTRIC ELEMENTS
- H01M—PROCESSES OR MEANS, e.g. BATTERIES, FOR THE DIRECT CONVERSION OF CHEMICAL ENERGY INTO ELECTRICAL ENERGY
- H01M10/00—Secondary cells; Manufacture thereof
- H01M10/42—Methods or arrangements for servicing or maintenance of secondary cells or secondary half-cells
- H01M10/48—Accumulators combined with arrangements for measuring, testing or indicating the condition of cells, e.g. the level or density of the electrolyte
- H01M10/486—Accumulators combined with arrangements for measuring, testing or indicating the condition of cells, e.g. the level or density of the electrolyte for measuring temperature
-
- H—ELECTRICITY
- H01—ELECTRIC ELEMENTS
- H01M—PROCESSES OR MEANS, e.g. BATTERIES, FOR THE DIRECT CONVERSION OF CHEMICAL ENERGY INTO ELECTRICAL ENERGY
- H01M10/00—Secondary cells; Manufacture thereof
- H01M10/42—Methods or arrangements for servicing or maintenance of secondary cells or secondary half-cells
- H01M10/425—Structural combination with electronic components, e.g. electronic circuits integrated to the outside of the casing
- H01M2010/4271—Battery management systems including electronic circuits, e.g. control of current or voltage to keep battery in healthy state, cell balancing
-
- Y—GENERAL TAGGING OF NEW TECHNOLOGICAL DEVELOPMENTS; GENERAL TAGGING OF CROSS-SECTIONAL TECHNOLOGIES SPANNING OVER SEVERAL SECTIONS OF THE IPC; TECHNICAL SUBJECTS COVERED BY FORMER USPC CROSS-REFERENCE ART COLLECTIONS [XRACs] AND DIGESTS
- Y02—TECHNOLOGIES OR APPLICATIONS FOR MITIGATION OR ADAPTATION AGAINST CLIMATE CHANGE
- Y02E—REDUCTION OF GREENHOUSE GAS [GHG] EMISSIONS, RELATED TO ENERGY GENERATION, TRANSMISSION OR DISTRIBUTION
- Y02E60/00—Enabling technologies; Technologies with a potential or indirect contribution to GHG emissions mitigation
- Y02E60/10—Energy storage using batteries
Landscapes
- Engineering & Computer Science (AREA)
- Manufacturing & Machinery (AREA)
- Chemical & Material Sciences (AREA)
- Chemical Kinetics & Catalysis (AREA)
- Electrochemistry (AREA)
- General Chemical & Material Sciences (AREA)
- Microelectronics & Electronic Packaging (AREA)
- Secondary Cells (AREA)
Abstract
The application discloses an electrochemical repair method of a lithium battery, which is characterized in that the change of the temperature of the battery is monitored in real time in the charging process of the lithium battery, and a data processing and analyzing algorithm is introduced at the rear end to analyze the time sequence change condition of the temperature in the charging process, so that corresponding measures are taken to control the temperature based on the temperature change, such as reducing the charging current or suspending the charging.
Description
Technical Field
The present application relates to the field of lithium batteries, and more particularly, to an electrochemical repair method for lithium batteries.
Background
According to the difference of the anode materials, the power battery mainly comprises two major types of ternary power batteries and iron lithium power batteries, and nickel-cobalt-manganese-lithium (NCM) and lithium iron phosphate (LFP) are respectively used as the anode materials. From the aspect of performance analysis, the lithium iron power battery has low capacity and short endurance mileage, but has good safety and long cycle life. The market of the retired lithium iron power battery tends to enter the cascade utilization industry, the residual value of the lithium iron power battery can be fully exerted, the maximization of circular economy is realized, and the construction cost of an energy storage system is reduced. The ternary power battery has high specific capacity and long endurance mileage, but has shorter service life, and the safety performance is inferior to that of an iron lithium power battery, so that the ternary power battery is not suitable for the gradient utilization fields with complex application environments, such as an energy storage power station, a communication base station backup power supply and the like.
As the service life of lithium ion batteries increases, the capacity of lithium ion batteries gradually decreases, and this capacity loss limits the service life and performance of the batteries. However, retired lithium ion batteries are repairable. The capacity loss caused by dynamic polarization reasons belongs to reversible capacity loss, and can be compensated and repaired through a charge-discharge system; capacity loss caused by material structural fatigue and damage to the internal structure of the battery belongs to irreversible capacity loss, but can be partially recovered through proper physical repair. The battery repairing technology can greatly improve the step utilization efficiency and the product quality of the retired power battery.
The charge-discharge system is a common lithium battery repair method. By periodically charging and discharging the battery, the active material inside the battery can be promoted to be redistributed, thereby reducing polarization and improving the capacity of the battery. During charging, temperature is one of the important factors affecting battery performance and safety, and too high a temperature may cause further damage to the battery, while too low a temperature may affect battery performance.
Accordingly, an optimized electrochemical repair scheme for lithium batteries is desired.
Disclosure of Invention
The present application has been made to solve the above-mentioned technical problems. The embodiment of the application provides an electrochemical repair method of a lithium battery, which is characterized in that the change of the temperature of the battery is monitored in real time in the charging process of the lithium battery, and a data processing and analyzing algorithm is introduced at the rear end to analyze the time sequence change condition of the temperature in the charging process, so that corresponding measures are taken to control the temperature based on the change of the temperature, such as reducing the charging current or suspending the charging, and in such a way, the charging current value can be adaptively adjusted based on the actual condition of the temperature, thereby reducing the polarization phenomenon in the battery and improving the performance and the safety of the battery.
According to one aspect of the present application, there is provided an electrochemical repair method of a lithium battery, comprising:
acquiring temperature values of a plurality of preset time points in a preset time period in a charging process;
arranging the temperature values of the plurality of preset time points into a charging process temperature time sequence input vector according to a time dimension;
upsampling the charging process temperature time sequence input vector based on linear interpolation to obtain an upsampled charging process temperature time sequence input vector;
Carrying out local time sequence feature extraction on the up-sampling charging process temperature time sequence input vector to obtain a sequence of charging process temperature local time sequence feature vectors;
carrying out local temperature time sequence feature consistency association coding on each charging process temperature local time sequence feature vector in the sequence of the charging process temperature local time sequence feature vectors to obtain a local temperature time sequence feature consistency topology feature matrix;
performing association coding based on a graph structure on the sequence of the local time sequence feature vector of the charging process temperature and the local temperature time sequence feature consistency topology feature matrix to obtain a local time sequence feature of the consistency topology charging process temperature;
based on the consistent topology charging process temperature local timing characteristics, it is determined that the charging current value at the present point in time should be increased, decreased, or maintained.
Compared with the prior art, the electrochemical repair method for the lithium battery provided by the application has the advantages that the change of the temperature of the battery is monitored in real time in the charging process of the lithium battery, and the temperature time sequence change condition in the charging process is analyzed by introducing a data processing and analyzing algorithm at the rear end, so that corresponding measures are taken to control the temperature based on the temperature change, such as reducing the charging current or suspending the charging, and the charging current value can be adaptively adjusted based on the actual condition of the temperature in such a way, so that the polarization phenomenon in the battery is reduced, and the performance and the safety of the battery are improved.
Drawings
The above and other objects, features and advantages of the present application will become more apparent by describing embodiments of the present application in more detail with reference to the attached drawings. The accompanying drawings are included to provide a further understanding of embodiments of the application and are incorporated in and constitute a part of this specification, illustrate the application and together with the embodiments of the application, and not constitute a limitation to the application. In the drawings, like reference numerals generally refer to like parts or steps.
Fig. 1 is a flowchart of an electrochemical repair method of a lithium battery according to an embodiment of the present application;
fig. 2 is a system architecture diagram of an electrochemical repair method of a lithium battery according to an embodiment of the present application;
fig. 3 is a flowchart of substep S4 of the electrochemical repair method of a lithium battery according to an embodiment of the present application;
fig. 4 is a flowchart of substep S5 of the electrochemical repair method of a lithium battery according to an embodiment of the present application;
fig. 5 is a flowchart of substep S7 of the electrochemical repair method of a lithium battery according to an embodiment of the present application.
Fig. 6 is an effect diagram of electrochemical repair according to an embodiment of the present application.
Detailed Description
Hereinafter, exemplary embodiments according to the present application will be described in detail with reference to the accompanying drawings. It should be apparent that the described embodiments are only some embodiments of the present application and not all embodiments of the present application, and it should be understood that the present application is not limited by the example embodiments described herein.
As used in the specification and in the claims, the terms "a," "an," "the," and/or "the" are not specific to a singular, but may include a plurality, unless the context clearly dictates otherwise. In general, the terms "comprises" and "comprising" merely indicate that the steps and elements are explicitly identified, and they do not constitute an exclusive list, as other steps or elements may be included in a method or apparatus.
Although the present application makes various references to certain modules in a system according to embodiments of the present application, any number of different modules may be used and run on a user terminal and/or server. The modules are merely illustrative, and different aspects of the systems and methods may use different modules.
A flowchart is used in the present application to describe the operations performed by a system according to embodiments of the present application. It should be understood that the preceding or following operations are not necessarily performed in order precisely. Rather, the various steps may be processed in reverse order or simultaneously, as desired. Also, other operations may be added to or removed from these processes.
Hereinafter, exemplary embodiments according to the present application will be described in detail with reference to the accompanying drawings. It should be apparent that the described embodiments are only some embodiments of the present application and not all embodiments of the present application, and it should be understood that the present application is not limited by the example embodiments described herein.
The charge-discharge system is a common lithium battery repair method. By periodically charging and discharging the battery, the active material inside the battery can be promoted to be redistributed, thereby reducing polarization and improving the capacity of the battery. During charging, temperature is one of the important factors affecting battery performance and safety, and too high a temperature may cause further damage to the battery, while too low a temperature may affect battery performance. Accordingly, an optimized electrochemical repair scheme for lithium batteries is desired.
In the technical scheme of the application, an electrochemical repair method of a lithium battery is provided. Fig. 1 is a flowchart of an electrochemical repair method of a lithium battery according to an embodiment of the present application. Fig. 2 is a system architecture diagram of an electrochemical repair method of a lithium battery according to an embodiment of the present application. As shown in fig. 1 and 2, the electrochemical repair method of a lithium battery according to an embodiment of the present application includes the steps of: s1, acquiring temperature values of a plurality of preset time points in a preset time period in a charging process; s2, arranging the temperature values of the plurality of preset time points into a charging process temperature time sequence input vector according to a time dimension; s3, up-sampling based on linear interpolation is carried out on the temperature time sequence input vector of the charging process so as to obtain an up-sampling temperature time sequence input vector of the charging process; s4, carrying out local time sequence feature extraction on the up-sampling charging process temperature time sequence input vector to obtain a sequence of charging process temperature local time sequence feature vectors; s5, carrying out local temperature time sequence feature consistency association coding on each charging process temperature local time sequence feature vector in the sequence of the charging process temperature local time sequence feature vectors to obtain a local temperature time sequence feature consistency topology feature matrix; s6, performing association coding based on a graph structure on the sequence of the local time sequence feature vector of the temperature in the charging process and the local temperature time sequence feature consistency topology feature matrix to obtain the local time sequence feature of the temperature in the consistency topology charging process; s7, based on the local time sequence characteristics of the temperature of the consistent topological charging process, determining that the charging current value at the current time point should be increased, decreased or maintained.
In particular, the S1 obtains temperature values at a plurality of predetermined time points within a predetermined period of time during charging. In one example, temperature values at a plurality of predetermined time points within a predetermined period of time during the charging process may be acquired by a temperature sensor. It is worth mentioning that a temperature sensor is a device for measuring the temperature of an environment or an object. It is capable of converting temperature into electrical or digital signals for monitoring, control or recording.
Specifically, the step S2 is to arrange the temperature values at the predetermined time points into a charging process temperature time sequence input vector according to a time dimension. Considering that the temperature value in the charging process has a time sequence dynamic change rule in the time dimension, that is, the temperature value is continuously changed in the charging process, and the temperature values at the plurality of preset time points have a time sequence association relation. Therefore, in the technical scheme of the application, in order to capture and draw the time sequence dynamic change characteristics of the temperature values in the time dimension, the temperature values of the plurality of preset time points are required to be arranged into the charging process temperature time sequence input vector according to the time dimension, so that the temperature time sequence distribution information in the charging process is integrated.
Accordingly, in one possible implementation, the temperature values at the plurality of predetermined time points may be arranged as a charging process temperature timing input vector according to a time dimension by: collecting temperature value data of preset time points, and ensuring that each time point has a corresponding temperature value; determining the sequence of time points, and arranging the time points according to the time sequence in the charging process; creating a null vector for storing the temperature timing input; traversing the ordered time point list one by one from the earliest time point; for each time point, finding a corresponding temperature value; adding the temperature value to a temperature timing input vector; continuing to traverse the next time point, and repeating the step 5 and the step 6 until all the time points are traversed; after the traversal is completed, the obtained vector is the temperature time sequence input vector of the charging process arranged according to the time dimension.
In particular, the step S3 performs up-sampling based on linear interpolation on the charging process temperature timing input vector to obtain an up-sampled charging process temperature timing input vector. It should be understood that, in order to improve the capturing capability of the characteristic of the temperature fine change in the charging process, in the technical scheme of the application, the up-sampling based on the linear interpolation is further performed on the charging process temperature time sequence input vector to obtain an up-sampling charging process temperature time sequence input vector, so as to increase the density and smoothness of the temperature data, thereby facilitating the subsequent better representation of the time sequence characteristic of the temperature. It should be appreciated that by upsampling through linear interpolation, the data points in the original charging process temperature timing input vector may be interpolated, generating more data points. In this way, it is helpful to increase the resolution in the time dimension, making the change in temperature of the charging process more visible. Meanwhile, the linear interpolation can carry out smooth interpolation among sampling points, so that the influence of noise and abrupt change is reduced, and the continuity and stability of data are improved. That is, the upsampled charging process temperature timing input vector may provide more detailed and accurate charging process temperature timing change information, providing more abundant data for subsequent feature extraction and modeling processes. The method is favorable for enhancing the perceptibility of the model to temperature time sequence change in the charging process, improving the control precision of the charging current value and relieving the polarization phenomenon in the battery.
Notably, upsampling based on linear interpolation is a data processing method for increasing the number of data points in a time series or continuous data to obtain a higher time resolution or smoother data curve. The method uses linear interpolation to estimate the value of the newly added data point, and performs interpolation calculation based on the known original data point. The basic principle of linear interpolation is to assume that the numerical variation between two known data points is linear and then to scale between the two known data points based on the interpolation position. For example, if there are two known data points A and B, the time interval is t, the interpolation factor is m, then m data points are inserted between A and B. For each interpolation point, its ratio between A and B is determined based on its relative position, and then the ratio is applied to the difference between the values of A and B to calculate the value of the interpolation point. By upsampling based on linear interpolation, the density of the data can be increased, making the data smoother, and providing more detailed temporal resolution. This is often used in the fields of processing time-series data, signal processing, image processing, and the like.
Accordingly, in one possible implementation, the charging process temperature timing input vector may be up-sampled based on linear interpolation by: and acquiring an original charging process temperature time sequence input vector. This vector contains temperature data points during the charging process, typically acquired at regular intervals; the time interval between raw data points is calculated. By calculating the time difference between adjacent data points, the time interval between the original data points can be obtained; and determining an up-sampling multiple. The up-sampling multiple indicates how many times the new sampling rate is the original sampling rate. For example, if the upsampling multiple is 2, this means that the new sampling rate is twice the original sampling rate; a new time interval is calculated. Dividing the time interval between the original data points by the up-sampling multiple to obtain a new time interval; the number of new data points is calculated. The number of new data points is equal to the number of original data points multiplied by the upsampling multiple; a new time vector is created. Creating a new time vector for storing the time stamp of the interpolated data point according to the new time interval and the number of new data points; a new temperature vector is created. Creating a new temperature vector for storing the interpolated temperature data points according to the number of new data points; linear interpolation is performed. For each new data point, two adjacent data points are found in the original data point according to its time stamp. Calculating a temperature value of a new data point according to the time stamp and the temperature value of an adjacent data point by using a linear interpolation formula; and outputting the new time vector and the temperature vector as an up-sampled charging process temperature time sequence input vector.
Specifically, the step S4 is to perform local time sequence feature extraction on the up-sampling charging process temperature time sequence input vector to obtain a sequence of charging process temperature local time sequence feature vectors. In particular, in one specific example of the present application, as shown in fig. 3, the S4 includes: s41, vector segmentation is carried out on the up-sampling charging process temperature time sequence input vector so as to obtain a sequence of charging process temperature local time sequence input vectors; s42, performing feature extraction on the sequence of the local time sequence input vectors of the temperature of the charging process by a temperature time sequence feature extractor based on a deep neural network model to obtain the sequence of the local time sequence feature vectors of the temperature of the charging process.
Specifically, in S41, vector slicing is performed on the up-sampling charging process temperature timing input vector to obtain a sequence of charging process temperature local timing input vectors. It should be understood that, when extracting the time sequence variation feature of the charging process temperature, in order to better capture the time sequence variation feature information of the temperature value in the time dimension, vector segmentation is further required to be performed on the up-sampling charging process temperature time sequence input vector to obtain a sequence of charging process temperature local time sequence input vectors, so as to better extract the charging process temperature local time sequence detail variation feature information in different time periods.
It is noted that vector slicing is a process of splitting a vector into a plurality of sub-vectors according to a certain rule or condition. In vector slicing, an original vector is split into a plurality of smaller vectors, each sub-vector containing a portion of the elements in the original vector. Vector slicing can be used in a plurality of fields such as data processing, feature extraction, sequence analysis and the like. It can break up a larger vector into smaller parts, making the processing and analysis more flexible and efficient. The split sub-vectors may be operated independently or may be used as input for further calculations or model building.
Specifically, in S42, the sequence of the local time sequence input vectors of the temperature of the charging process is extracted by a temperature time sequence feature extractor based on a deep neural network model to obtain the sequence of the local time sequence feature vectors of the temperature of the charging process. In particular, in one specific example of the present application, the temperature timing feature extractor based on the deep neural network model is a temperature timing feature extractor based on a one-dimensional convolution layer. The sequence of the charging process temperature local time sequence input vector is extracted through feature extraction in a temperature time sequence feature extractor based on a one-dimensional convolution layer, so that local time sequence detail change feature information of the charging process temperature value in each local time period in a time dimension is extracted, and the sequence of the charging process temperature local time sequence feature vector is obtained. In this way, it is advantageous to perform time-series trend analysis of the charging process document value and control of the charging current value. Specifically, each layer of the one-dimensional convolution layer-based temperature time sequence feature extractor is used for respectively carrying out forward transfer on input data: carrying out convolution processing on input data to obtain a convolution characteristic diagram; pooling the convolution feature images based on a feature matrix to obtain pooled feature images; performing nonlinear activation on the pooled feature map to obtain an activated feature map; the output of the last layer of the one-dimensional convolution layer-based temperature time sequence feature extractor is a sequence of the charging process temperature local time sequence feature vectors, and the input of the first layer of the one-dimensional convolution layer-based temperature time sequence feature extractor is a sequence of the charging process temperature local time sequence input vectors.
It is noted that a one-dimensional convolutional layer is a neural network layer commonly used in deep learning for processing data having a time-series structure. It is one of the core components of Convolutional Neural Networks (CNNs). The one-dimensional convolution layer extracts features by performing one-dimensional convolution operations on the input data. It uses a learnable filter (also called a convolution kernel or kernel) to perform an element-by-element sliding window convolution operation on the input data. The weights of the filters are learned by back propagation during training to automatically capture features in the input data. The working principle of the one-dimensional convolution layer is as follows: input data: a one-dimensional convolutional layer receives one-dimensional input data, such as a coded representation of time-series data or text data; and (3) a filter: the one-dimensional convolution layer comprises a plurality of filters, each filter being a one-dimensional weight vector. The number of these filters can be set as desired; convolution operation: for each filter, a one-dimensional convolution layer performs an element-by-element convolution operation on the filter and the input data. This corresponds to sliding the filter over the input data and performing a dot product operation for each location; feature mapping: the result of the convolution operation generates a feature map, i.e. a representation comprising local features in the input data. The size of the feature map depends on the number of filters and the stride of the convolution operation; activation function: for each feature map, an activation function may be applied to introduce nonlinearities. Common activation functions include ReLU, sigmoid, and tanh, among others.
It should be noted that, in other specific examples of the present application, the sequence of the charging process temperature local time sequence feature vectors may be obtained by performing feature extraction on the sequence of the charging process temperature local time sequence input vectors through a temperature time sequence feature extractor based on a deep neural network model in other manners, for example: a temperature timing feature extractor model based on a deep neural network is created. The model can be a neural network model formed by stacking a plurality of convolution layers, pooling layers and full connection layers, and is used for extracting features from time sequence input vectors; providing a sequence of charging process temperature local timing input vectors as inputs to a temperature timing feature extractor model; the temperature timing feature extractor model converts each local timing input vector into a corresponding local timing feature vector by forward propagation computation. These feature vectors can be extracted high-level abstract features with better characterization capabilities; the extracted sequence of charging process temperature local time series feature vectors may be used for further analysis, modeling, or other tasks. For example, they may be input into a subsequent fully connected layer for classification, regression, or other predictive tasks; according to specific requirements, proper regularization techniques such as batch normalization, dropout and the like can be added into the temperature time sequence feature extractor model so as to improve the generalization capability of the model and reduce overfitting.
It should be noted that, in other specific examples of the present application, the local time sequence feature extraction may be performed on the upsampled charging process temperature time sequence input vector in other manners to obtain a sequence of charging process temperature local time sequence feature vectors, for example: the window size is determined. The window size refers to the number of data points used to extract the local timing characteristics. Selecting a proper window size according to specific requirements and characteristics of a charging process; a sliding step is determined. The sliding step length refers to the moving distance of each sliding window. A portion of the window size may be selected as the sliding step size, or other suitable values may be selected; an empty local timing feature vector sequence is created. This sequence is used to store the extracted local timing feature vector; the upsampling charging process temperature timing input vector is traversed. Traversing the temperature time sequence input vector of the up-sampling charging process by taking the sliding step length as an interval from the beginning; for each window, local timing features are extracted. Within the current window, local timing features are extracted from the upsampled charging process temperature timing input vector. Various statistics, such as mean, variance, maximum, minimum, etc., may be used as local timing features; adding the extracted local time sequence features into a local time sequence feature vector sequence; continuing sliding the window until the temperature time sequence input vector of the whole up-sampling charging process is traversed; and taking the local time sequence characteristic vector sequence as the output of the temperature local time sequence characteristic vector in the charging process.
Particularly, the step S5 is to perform local temperature time sequence feature consistency association coding on each charging process temperature local time sequence feature vector in the sequence of charging process temperature local time sequence feature vectors to obtain a local temperature time sequence feature consistency topology feature matrix. In particular, in one specific example of the present application, as shown in fig. 4, the S5 includes: s51, calculating cosine similarity between any two charging process temperature local time sequence feature vectors in the sequence of the charging process temperature local time sequence feature vectors to obtain a local temperature time sequence feature consistency topology matrix; s52, the local temperature time sequence feature consistency topology matrix passes through a topology feature extractor based on a convolutional neural network model to obtain the local temperature time sequence feature consistency topology feature matrix.
Specifically, in S51, the cosine similarity between any two local time sequence feature vectors of the charging process temperature in the sequence of local time sequence feature vectors of the charging process temperature is calculated to obtain a topology matrix with consistency of local time sequence features. It is considered that there is a time-series-overall-based correlation between the temperature time series characteristics in the respective partial time periods due to the charging process of the lithium battery. In order to fully capture the change mode and rule of the temperature in the charging process so as to understand the dynamic change condition inside the battery, in the technical scheme of the application, the cosine similarity between any two charging process temperature local time sequence feature vectors in the sequence of the charging process temperature local time sequence feature vectors is further calculated so as to obtain a local temperature time sequence feature consistency topology matrix. It should be appreciated that by comparing the similarity between the charging process temperature local time series feature vectors for different time segments, the temperature correlation and trend of change in each local time segment can be captured and characterized. And by analyzing the topological matrix with the consistency of the local temperature time sequence characteristics, the mode and the rule of the temperature change in the battery charging process can be revealed, which is helpful for understanding the dynamic change in the battery and providing guidance for the subsequent charging strategy adjustment. For example, if the temperature change characteristics in a certain local time period are not consistent with those in other local time periods, the charging current can be adjusted according to the characteristic information so as to achieve better repairing effect.
Specifically, the step S52 is to pass the local temperature time sequence feature consistency topology matrix through a topology feature extractor based on a convolutional neural network model to obtain the local temperature time sequence feature consistency topology feature matrix. The local temperature time sequence feature consistency topology matrix is subjected to feature mining in a topology feature extractor based on a convolutional neural network model, so that consistency topology association feature information among local time sequence features related to temperature in each local time period in the charging process is extracted, and the local temperature time sequence feature consistency topology feature matrix is obtained. In particular, the method comprises the steps of,
notably, convolutional Neural Networks (CNNs) are a type of deep learning model that is primarily used to handle tasks with grid structure data, such as image and speech recognition. The core idea of the CNN is to realize feature extraction and classification of input data through components such as a convolution layer, a pooling layer, a full connection layer and the like. The following are the main components and features of CNN: convolution layer: the convolutional layer is a core component of the CNN that extracts local features of the input data by convolving the input data with a set of learnable filters (also called convolutional kernels). The convolution operation may capture spatial local relationships of the input data, such as edges, textures, etc. in the image; pooling layer: the pooling layer is used to reduce the spatial dimensions of the feature map while preserving important feature information. Common pooling operations include maximum pooling and average pooling, which extract the maximum or average value of the local area, respectively, as the pooling result; full tie layer: the fully connected layer connects the outputs of the convolution and pooling layers to one or more fully connected layers for classifying or regressing the extracted features. The neurons in the full-connection layer are connected with all neurons in the previous layer; activation function: in convolutional neural networks, an activation function is typically applied to the output of the convolutional layer and the fully-connected layer, introducing nonlinear transformations that increase the expressive power of the network. Common activation functions include ReLU, sigmoid, tanh, etc.; weight sharing: the convolutional layers in CNN use a strategy of weight sharing, i.e. apply the same convolutional kernel at different locations. Therefore, the number of parameters to be learned can be greatly reduced, and the efficiency and generalization capability of the model are improved. Through the stacking of multiple convolution layers and pooling layers, CNNs can extract abstract features of input data layer by layer, from low-level features (e.g., edges) to high-level features (e.g., object shapes). This makes CNNs excellent in computer vision tasks such as image classification, object detection, semantic segmentation, etc. Meanwhile, CNN can be applied in other fields, such as natural language processing and signal processing.
It should be noted that, in other specific examples of the present application, the local temperature time sequence feature vectors of the sequence of the local time sequence feature vectors of the charging process temperature may also be subjected to local temperature time sequence feature consistency association encoding in other manners to obtain a local temperature time sequence feature consistency topology feature matrix, for example: a window size (e.g., W) and stride (e.g., S) are defined. The window size represents the number of local feature vectors considered in the sequence, and the stride represents the sliding interval of the window in the sequence; sliding a window according to steps from a sequence of the local time sequence feature vectors of the temperature in the charging process, and taking a sub-sequence of the local feature vectors with the window size W each time; for the partial feature vector subsequence in each window, the similarity or consistency of its internal features is calculated. Various similarity measures such as euclidean distance, cosine similarity, etc. may be used; and constructing a local temperature time sequence characteristic consistency association coding matrix based on the result of similarity calculation. Each element of the matrix represents a degree of consistency between corresponding local feature vector subsequences; the local temperature time sequence characteristic consistency association coding matrix can be further processed, for example, a threshold value or weight is applied to screen or adjust a consistency relation; the finally obtained local temperature time sequence characteristic consistency topology characteristic matrix can be used for describing the consistency relation between the temperature local time sequence characteristic vectors in the charging process.
Specifically, the step S6 is to perform association encoding based on a graph structure on the sequence of the local time sequence feature vector of the charging process temperature and the local time sequence feature consistency topology feature matrix to obtain a local time sequence feature of the consistency topology charging process temperature. The method comprises the steps of taking each sound signal sampling window waveform characteristic vector in a sequence of the charging process temperature local time sequence characteristic vector as characteristic representation of a node, taking the local temperature time sequence characteristic consistency topological characteristic matrix as characteristic representation of an edge between nodes, and enabling the charging process temperature time sequence characteristic matrix and the local temperature time sequence characteristic consistency topological characteristic matrix obtained by two-dimensional arrangement of the sequence of the charging process temperature local time sequence characteristic vector to pass through a graph neural network model to obtain a consistency topological charging process temperature local time sequence characteristic matrix. Specifically, the graph neural network model performs graph structure data coding on the charging process temperature time sequence feature matrix and the local temperature time sequence feature consistency topology feature matrix through a learnable neural network parameter to obtain the consistency topology charging process temperature local time sequence feature matrix containing irregular local temperature time sequence consistency topology association features and temperature time sequence feature information in each local time sequence segment.
Notably, the Graph Neural Network (GNN) is a deep learning model for processing graph structure data. Unlike conventional Convolutional Neural Networks (CNNs) that are adapted to process grid structure data (e.g., images), GNNs are specifically designed to process data of non-grid structures, such as social networks, knowledge maps, molecular structures, and the like. The core idea of the graph neural network is to learn the representation of the nodes through local neighbor information propagation and aggregation between the nodes. The following are the main components and features of GNN: node represents: GNNs capture the characteristics of nodes by representing each node as a vector or matrix. Initially, each node has an initial feature vector. During the training process of the GNN, node representation is updated iteratively along with information propagation and aggregation; graph convolution layer: the graph convolution layer is the core component of the GNN for propagating and aggregating information of nodes in the graph structure. The graph convolution layer updates the representation of each node by considering the neighbor information of the node. The typical graph volume lamination adopts the methods of weighted average or splicing of neighbor nodes and the like to carry out information aggregation; and (5) a pooling layer: like the pooling layer in CNNs, the graph pooling layer serves to reduce the scale of the graph and preserve important node information. Common graph pooling methods include sub-sampling of graphs and graph structure aggregation of graphs; graph attention mechanism: to better handle node relationships in the graph, a graph annotation mechanism is introduced in GNN. The graph attention mechanism focuses more on important neighbor nodes by learning the attention weight between each node and its neighbor nodes; graph-level tasks and node-level tasks: GNNs can be used to solve graph-level tasks and node-level tasks. The graph-level tasks involve predicting or classifying the entire graph, such as graph classification and graph generation. Node-level tasks involve predicting or classifying each node, such as node classification and node embedding. The advantage of the graph neural network is the ability to process unstructured graph data and extract useful features from the global structure and local neighbor information of the graph. The method has wide application in the fields of social network analysis, recommendation systems, molecular design and the like.
In particular, the S7, based on the consistent topology charging process temperature local time sequence characteristics, determines that the charging current value at the present point in time should be increased, should be decreased, or should be maintained. In particular, in one specific example of the present application, as shown in fig. 5, the S7 includes: s71, performing characteristic scale on the local time sequence characteristic matrix of the temperature in the consistent topology charging process as a rank arrangement distribution soft matching of an imitation mask so as to obtain an optimized local time sequence characteristic matrix of the temperature in the consistent topology charging process; and S72, the temperature local time sequence feature matrix of the optimized consistency topology charging process passes through a classifier to obtain a classification result, wherein the classification result is used for indicating that the charging current value at the current time point should be increased, decreased or maintained.
Specifically, in S71, the feature scale of the local time sequence feature matrix of the consistent topology charging process temperature is used as the rank arrangement distribution soft matching of the imitation mask to obtain the local time sequence feature matrix of the optimized consistent topology charging process temperature. Particularly, in the technical scheme of the application, when the sequence of the local time sequence feature vectors of the charging process and the local temperature time sequence feature consistency topological feature matrix are obtained through a graph neural network model, each row feature vector of the local time sequence feature matrix of the consistent topology charging process expresses topological association expression of the temperature time sequence feature similarity topology between each local time sequence in the global time sequence based on the temperature time sequence feature similarity topology in the local time sequence, so if the temperature time sequence feature in each local time sequence is taken as a foreground object feature, the graph neural network model also introduces background distribution noise when carrying out feature topological association expression, and simultaneously carries out high rank distribution expression between vector-matrix, probability density mapping errors of the local time sequence feature matrix of the consistent topology charging process relative to associated feature time sequence distribution in each local time sequence domain are caused by time sequence space heterogeneous distribution of high-dimensional feature of the temperature value local time sequence association feature in the consistent topology charging process, thereby influencing the classification result obtained by a classifier. Based on this, the applicant of the present application performs a rank-permutation distribution soft-matching of a feature scale as an imitation mask on the consistent topology charging process temperature local time sequence feature matrix, for example denoted as M, specifically expressed as:
Wherein M is the local time sequence characteristic matrix of the temperature in the consistent topology charging process, and M i,j Is when the temperature of the consistent topology charging process is localThe eigenvalue of the (i, j) th position of the sequential eigenvalue matrix, S is the scale of the temperature local sequential eigenvalue matrix of the consistent topology charging process,the square of the Frobenius norm of the temperature local time sequence characteristic matrix of the consistent topological charging process is represented by M 2 Representing the two norms of the temperature local time sequence characteristic matrix of the consistent topology charging process, wherein alpha is a weighted hyper-parameter, exp (-) represents an exponential operation, m '' i,j Is the characteristic value of the (i, j) th position of the temperature local time sequence characteristic matrix in the optimized consistency topology charging process. Here, the feature scale as the rank arrangement distribution soft matching of the mimicking mask can focus the feature scale as the mimicking mask for mapping on the foreground object feature and ignore the background distribution noise when mapping the high-dimensional feature to be classified into the probability density space, and the distribution soft matching of the pyramid rank arrangement distribution by different norms of the temperature local time sequence feature matrix M in the consistent topology charging process effectively captures the correlation between the central area and the tail area of the probability density distribution, so that probability density mapping deviation caused by the spatial heterogeneous distribution of the high-dimensional feature of the temperature local time sequence feature matrix M in the consistent topology charging process is avoided, and the accuracy of the classification result obtained by the classifier of the temperature local time sequence feature matrix in the consistent topology charging process is improved. In this way, corresponding measures can be taken to control the temperature based on the change condition of the temperature, such as reducing the charging current or suspending the charging, and in this way, the charging current value can be adaptively adjusted based on the actual change condition of the temperature, so that the polarization phenomenon inside the battery is reduced, and the performance and the safety of the battery are improved.
Specifically, in S72, the local time sequence feature matrix of the temperature of the charging process of the optimized consistency topology is passed through a classifier to obtain a classification result, where the classification result is used to indicate that the charging current value at the current time point should be increased, should be decreased or should be maintained. More specifically, the temperature local time sequence feature matrix of the optimized consistency topology charging process is unfolded into classification feature vectors based on row vectors or column vectors; performing full-connection coding on the classification feature vectors by using a plurality of full-connection layers of the classifier to obtain coded classification feature vectors; and passing the coding classification feature vector through a Softmax classification function of the classifier to obtain the classification result. The classification label of the classifier is a control label of a charging current value, and after the classification result is obtained, the charging current value is adaptively adjusted based on the change condition of temperature, so that the polarization phenomenon in the battery is reduced, and the performance and the safety of the battery are improved.
A Classifier (Classifier) refers to a machine learning model or algorithm that is used to classify input data into different categories or labels. The classifier is part of supervised learning, which performs classification tasks by learning mappings from input data to output categories.
The fully connected layer (Fully Connected Layer) is one type of layer commonly found in neural networks. In the fully connected layer, each neuron is connected to all neurons of the upper layer, and each connection has a weight. This means that each neuron in the fully connected layer receives inputs from all neurons in the upper layer, and weights these inputs together, and then passes the result to the next layer.
The Softmax classification function is a commonly used activation function for multi-classification problems. It converts each element of the input vector into a probability value between 0 and 1, and the sum of these probability values equals 1. The Softmax function is commonly used at the output layer of a neural network, and is particularly suited for multi-classification problems, because it can map the network output into probability distributions for individual classes. During the training process, the output of the Softmax function may be used to calculate the loss function and update the network parameters through a back propagation algorithm. Notably, the output of the Softmax function does not change the relative magnitude relationship between elements, but rather normalizes them. Thus, the Softmax function does not change the characteristics of the input vector, but simply converts it into a probability distribution form.
It should be noted that, in other specific examples of the present application, it may also be determined that the charging current value at the present time point should be increased, decreased or maintained based on the local time sequence feature of the consistent topology charging process temperature in other manners, for example: firstly, collecting temperature and charging current data of a consistent topology charging process; extracting temperature local time sequence characteristics from the collected data; and according to the extracted characteristics, a model is established to predict the charging current value of the current time point. A machine learning algorithm, such as a regression model or a time series model, may be used to construct the predictive model; the model is trained using historical data. Taking the known temperature local time sequence characteristics and corresponding charging current values as input and output, and learning the relation between the known temperature local time sequence characteristics and corresponding charging current values through model training; and predicting the local time sequence characteristic of the temperature at the current time point by using the trained model to determine whether the charging current value at the current time point should be increased, decreased or maintained. According to the output of the model, the difference between the current value and the predicted value can be judged, and then the operation to be adopted is determined.
In summary, the electrochemical repair method of the lithium battery according to the embodiment of the application is explained, which monitors the change of the battery temperature in real time during the charging process of the lithium battery, and introduces a data processing and analyzing algorithm at the rear end to analyze the temperature time sequence change condition during the charging process, so as to take corresponding measures based on the temperature change to control the temperature, such as reducing the charging current or suspending the charging.
In the embodiment of the application, as shown in fig. 6, the capacity loss caused by dynamic polarization reasons belongs to reversible capacity loss, and can be compensated and repaired by a charge-discharge system (see fig. 4); capacity loss caused by material structural fatigue and damage to the internal structure of the battery belongs to irreversible capacity loss, but can be partially recovered through proper physical repair. The battery repairing technology can greatly improve the utilization efficiency and the product quality of the retired power battery cascade, and belongs to a breakthrough core technology of the retired power battery cascade utilization industry.
The foregoing description of the embodiments of the present disclosure has been presented for purposes of illustration and description, and is not intended to be exhaustive or limited to the embodiments disclosed. Many modifications and variations will be apparent to those of ordinary skill in the art without departing from the scope and spirit of the various embodiments described. The terminology used herein was chosen in order to best explain the principles of the embodiments, the practical application, or the improvement of technology in the marketplace, or to enable others of ordinary skill in the art to understand the embodiments disclosed herein.
Claims (8)
1. A method for electrochemical repair of a lithium battery, comprising:
acquiring temperature values of a plurality of preset time points in a preset time period in a charging process;
arranging the temperature values of the plurality of preset time points into a charging process temperature time sequence input vector according to a time dimension;
upsampling the charging process temperature time sequence input vector based on linear interpolation to obtain an upsampled charging process temperature time sequence input vector;
carrying out local time sequence feature extraction on the up-sampling charging process temperature time sequence input vector to obtain a sequence of charging process temperature local time sequence feature vectors;
carrying out local temperature time sequence feature consistency association coding on each charging process temperature local time sequence feature vector in the sequence of the charging process temperature local time sequence feature vectors to obtain a local temperature time sequence feature consistency topology feature matrix;
performing association coding based on a graph structure on the sequence of the local time sequence feature vector of the charging process temperature and the local temperature time sequence feature consistency topology feature matrix to obtain a local time sequence feature of the consistency topology charging process temperature;
based on the consistent topology charging process temperature local timing characteristics, it is determined that the charging current value at the present point in time should be increased, decreased, or maintained.
2. The electrochemical repair method of a lithium battery according to claim 1, wherein performing local time series feature extraction on the upsampled charging process temperature time series input vector to obtain a sequence of charging process temperature local time series feature vectors, comprises:
vector segmentation is carried out on the up-sampling charging process temperature time sequence input vector so as to obtain a sequence of charging process temperature local time sequence input vectors;
and extracting features of the sequence of the charging process temperature local time sequence input vectors through a temperature time sequence feature extractor based on a deep neural network model so as to obtain the sequence of the charging process temperature local time sequence feature vectors.
3. The electrochemical repair method of a lithium battery according to claim 2, wherein the temperature time series feature extractor based on the deep neural network model is a temperature time series feature extractor based on a one-dimensional convolution layer.
4. The electrochemical repair method of a lithium battery according to claim 3, wherein performing local temperature timing feature consistency association encoding on each charging process temperature local timing feature vector in the sequence of charging process temperature local timing feature vectors to obtain a local temperature timing feature consistency topology feature matrix, comprises:
Calculating cosine similarity between any two charging process temperature local time sequence feature vectors in the sequence of the charging process temperature local time sequence feature vectors to obtain a local temperature time sequence feature consistency topology matrix;
and the local temperature time sequence feature consistency topology matrix passes through a topology feature extractor based on a convolutional neural network model to obtain the local temperature time sequence feature consistency topology feature matrix.
5. The electrochemical repair method of a lithium battery according to claim 4, wherein performing graph structure-based associative coding on the sequence of the charging process temperature local time sequence feature vectors and the local temperature time sequence feature consistency topology feature matrix to obtain a consistency topology charging process temperature local time sequence feature, comprises:
and the sequence of the local time sequence feature vector of the charging process temperature and the local temperature time sequence feature consistency topological feature matrix are processed through a graph neural network model to obtain a local time sequence feature matrix of the consistency topological charging process temperature as the local time sequence feature of the consistency topological charging process temperature.
6. The method of electrochemical repair of a lithium battery of claim 5, wherein determining that a charge current value at a present point in time should be increased, decreased, or maintained based on the consistent topological charging process temperature local time sequence feature comprises:
Performing soft matching on the characteristic scale of the temperature local time sequence characteristic matrix in the consistent topology charging process as the rank arrangement distribution of the imitation mask so as to obtain an optimized temperature local time sequence characteristic matrix in the consistent topology charging process;
and (3) passing the temperature local time sequence characteristic matrix of the optimized consistency topology charging process through a classifier to obtain a classification result, wherein the classification result is used for indicating that the charging current value at the current time point should be increased, decreased or maintained.
7. The electrochemical repair method of a lithium battery according to claim 6, wherein performing a rank-permutation distribution soft-matching of a feature scale as a simulated mask on the consistent topology charging process temperature local time series feature matrix to obtain an optimized consistent topology charging process temperature local time series feature matrix, comprises:
performing characteristic scale on the temperature local time sequence characteristic matrix in the consistent topology charging process by using the following optimization formula as a rank arrangement distribution soft match of an imitation mask so as to obtain the temperature local time sequence characteristic matrix in the optimized consistent topology charging process;
wherein, the optimization formula is:
wherein M is the local time sequence characteristic matrix of the temperature in the consistent topology charging process, and M i,j Is the characteristic value of the (i, j) th position of the temperature local time sequence characteristic matrix of the consistent topology charging process, S is the scale of the temperature local time sequence characteristic matrix of the consistent topology charging process,the square of the Frobenius norm of the temperature local time sequence characteristic matrix of the consistent topological charging process is represented by M 2 Representing the two norms of the temperature local time sequence characteristic matrix of the consistent topology charging process, wherein alpha is a weighted hyper-parameter, exp (-) represents an exponential operation, m '' i,j Is the characteristic value of the (i, j) th position of the temperature local time sequence characteristic matrix in the optimized consistency topology charging process.
8. The electrochemical repair method of a lithium battery according to claim 7, wherein passing the optimized consistent topology charging process temperature local time series characteristic matrix through a classifier to obtain a classification result, the classification result being used to indicate that a charging current value at a present point in time should be increased, decreased or maintained, comprises:
expanding the temperature local time sequence feature matrix of the optimized consistency topology charging process into classification feature vectors based on row vectors or column vectors;
performing full-connection coding on the classification feature vectors by using a plurality of full-connection layers of the classifier to obtain coded classification feature vectors; and
And the coding classification feature vector is passed through a Softmax classification function of the classifier to obtain the classification result.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202311342922.6A CN117154256A (en) | 2023-10-17 | 2023-10-17 | Electrochemical repair method for lithium battery |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202311342922.6A CN117154256A (en) | 2023-10-17 | 2023-10-17 | Electrochemical repair method for lithium battery |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN117154256A true CN117154256A (en) | 2023-12-01 |
Family
ID=88902827
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202311342922.6A Pending CN117154256A (en) | 2023-10-17 | 2023-10-17 | Electrochemical repair method for lithium battery |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN117154256A (en) |
Cited By (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN117419828A (en) * | 2023-12-18 | 2024-01-19 | 南京品傲光电科技有限公司 | New energy battery temperature monitoring method based on optical fiber sensor |
-
2023
- 2023-10-17 CN CN202311342922.6A patent/CN117154256A/en active Pending
Cited By (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN117419828A (en) * | 2023-12-18 | 2024-01-19 | 南京品傲光电科技有限公司 | New energy battery temperature monitoring method based on optical fiber sensor |
| CN117419828B (en) * | 2023-12-18 | 2024-05-03 | 南京品傲光电科技有限公司 | New energy battery temperature monitoring method based on optical fiber sensor |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| Khaireddin et al. | Facial emotion recognition: State of the art performance on FER2013 | |
| CN108805083B (en) | Single-stage video behavior detection method | |
| CN107506740B (en) | Human body behavior identification method based on three-dimensional convolutional neural network and transfer learning model | |
| CN109697434B (en) | Behavior recognition method and device and storage medium | |
| Springenberg et al. | Improving deep neural networks with probabilistic maxout units | |
| CN110929622A (en) | Video classification method, model training method, device, equipment and storage medium | |
| Ding et al. | Where to prune: Using LSTM to guide data-dependent soft pruning | |
| CN111008618B (en) | Self-attention deep learning end-to-end pedestrian re-identification method | |
| CN110941734A (en) | Depth unsupervised image retrieval method based on sparse graph structure | |
| CN112884742A (en) | Multi-algorithm fusion-based multi-target real-time detection, identification and tracking method | |
| CN114332578A (en) | Image anomaly detection model training method, image anomaly detection method and device | |
| Dai et al. | Hybrid deep model for human behavior understanding on industrial internet of video things | |
| CN117154256A (en) | Electrochemical repair method for lithium battery | |
| KR20220098991A (en) | Method and apparatus for recognizing emtions based on speech signal | |
| CN114118225A (en) | Method, system, electronic device and storage medium for predicting remaining life of generator | |
| CN115294563A (en) | 3D point cloud analysis method and device based on Transformer and capable of enhancing local semantic learning ability | |
| CN116110089A (en) | Facial expression recognition method based on depth self-adaptive metric learning | |
| CN110111365B (en) | Training method and device based on deep learning and target tracking method and device | |
| CN117148197A (en) | Lithium ion battery life prediction method based on integrated transducer model | |
| CN115983497A (en) | Time sequence data prediction method and device, computer equipment and storage medium | |
| Wang et al. | A novel multiface recognition method with short training time and lightweight based on ABASNet and H-softmax | |
| CN114780767A (en) | Large-scale image retrieval method and system based on deep convolutional neural network | |
| CN117725491A (en) | SCITET-based power system fault state detection and classification method | |
| CN113780584A (en) | Label prediction method, apparatus, storage medium and program product | |
| Li | Parallel two-class 3D-CNN classifiers for video classification |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination |