Battery lithium precipitation detection method and system
The battery discharge curve characteristics are extracted through one-dimensional convolutional network and logistic regression model, and the problems of single lithium-extraction detection information and low accuracy in the prior art are solved, and high-accuracy lithium-extraction detection are achieved.
Patent Information
- Application Number
- CN202411134964.5
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2024-08-19
- Publication Date
- 2024-11-12
AI Technical Summary
In the prior art, lithium-ion detection relies on small negative numbers in the initial stage of the voltage fluctuation curve, and the information is single, resulting in low detection accuracy and manual dismantling of the battery consumes manpower and material resources.
The one-dimensional convolutional network is used to extract all the features of the battery discharge curve, and combined with the logistic Stie's regression model, the battery lithium-ion detection model is trained to detect lithium-ion in real time, and the detection accuracy is improved through data preprocessing and feature extraction.
By extracting all the characteristics of the battery discharge curve, the accuracy of lithium-ion detection is significantly improved, and the problems of single feature extraction and low detection accuracy in the prior art are solved.
Smart Images

Figure CN118938020A_ABST
Abstract
Description
Technical Field
[0001] The invention relates to a battery lithium deposition detection method and system, belonging to the technical field of battery lithium deposition detection. Background Art
[0002] Lithium deposition often occurs during the production of lithium ions. In the application process, it is found that lithium deposition not only reduces battery performance and significantly shortens cycle life, but also limits the fast charging capacity of the battery and may cause catastrophic consequences such as combustion and explosion. Therefore, how to effectively detect lithium deposition is of great significance to battery quality.
[0003] At present, the commonly used lithium plating detection is to manually disassemble the battery for observation, which consumes a lot of manpower and material resources. In recent years, lithium plating detection has been widely studied and paid attention to. The core task of lithium plating detection is to detect based on the process data generated during the battery production process. There are related studies that use voltage fluctuation curves for non-disassembly lithium plating detection, but this method only judges whether a value less than zero appears in the initial stage of the curve. The information used is single and the detection effect is poor. Summary of the invention
[0004] The technical problem to be solved by the present invention is to provide a battery lithium deposition detection method and system, which utilizes a one-dimensional convolutional network to extract all information of a battery discharge curve to perform lithium deposition discrimination, thereby greatly improving the accuracy of lithium deposition detection.
[0005] The present invention adopts the following technical solutions to solve the above technical problems:
[0006] A battery lithium deposition detection method comprises the following steps:
[0007] Step 1, obtaining the battery cell discharge curve time series data and the corresponding battery lithium deposition situation;
[0008] Step 2, preprocessing the battery cell discharge curve time series data obtained in step 1;
[0009] Step 3, extracting features from the preprocessed battery cell discharge curve time series data;
[0010] Step 4, taking the features extracted in step 3 as input and the corresponding battery lithium deposition situation as output, training the logistic regression model to obtain a battery lithium deposition detection model;
[0011] Step 5, obtaining the time series data of the battery cell discharge curve in real time, and performing the same processing as steps 2 and 3, inputting the features extracted from the real-time time series data into the battery lithium plating detection model for detection, and obtaining the lithium plating detection result.
[0012] By extracting all the features of the battery cell discharge curve time series data, using the extracted features and the corresponding lithium plating conditions to train a logistic regression model, a battery lithium plating detection model is obtained. The battery lithium plating detection model is used to perform lithium plating detection, which solves the problems of single feature extraction and low detection accuracy in the prior art.
[0013] As a preferred embodiment of the method of the present invention, in step 2, the preprocessing includes removing outliers and filling missing values, wherein:
[0014] The 6σ principle is used to eliminate outliers, and all data outside the upper and lower bounds are eliminated. The calculation formula is as follows:
[0015]
[0016] Among them, U and L represent the upper and lower bounds of 6σ respectively, and d i represents the time series data of the battery discharge curve, n represents the total amount of data before outliers are removed, Indicates all d i The mean of
[0017] After removing outliers, the interval window mean filling method is used to fill the missing values. The calculation formula is as follows:
[0018]
[0019] Among them, p represents the filling value, w′ represents the window size, w′ is an even number, and i represents the missing data position.
[0020] As a preferred embodiment of the method of the present invention, in step 3, a one-dimensional convolutional network is used to extract the features of the preprocessed battery cell discharge curve time series data. The one-dimensional convolutional network includes three layers of convolution, the first layer of convolution uses a 1×5 convolution kernel, the second layer of convolution uses a 1×3 convolution kernel, and the third layer of convolution uses a 1×3 convolution kernel.
[0021] As a preferred embodiment of the method of the present invention, in step 4, the logistic regression model is trained using the gradient descent method, and the loss function is as follows:
[0022]
[0023] Among them, L(W) represents the loss function, Y represents the actual category, and X i Represents data features, y i Represents the predicted category, and n represents the amount of preprocessed data.
[0024] A battery lithium deposition detection system, the battery lithium deposition detection system comprises: a data acquisition module, a data preprocessing module, a feature extraction module, a model processing module and a detection output module; wherein,
[0025] The data acquisition module is used to obtain the battery cell discharge curve time series data and the corresponding battery lithium deposition situation, and is also used to obtain the battery cell discharge curve time series data in real time;
[0026] The data preprocessing module is used to preprocess the acquired battery cell discharge curve time series data, and is also used to preprocess the battery cell discharge curve time series data acquired in real time;
[0027] The feature extraction module is used to extract features from the preprocessed battery cell discharge curve time series data, and is also used to extract features from the preprocessed real-time battery cell discharge curve time series data;
[0028] The model processing module is used to take the features extracted from the pre-processed battery discharge curve time series data as input, and the corresponding battery lithium deposition situation as output, train the logistic regression model, and obtain the battery lithium deposition detection model;
[0029] The detection output module is used to input the features extracted from the pre-processed real-time battery cell discharge curve time series data into the battery lithium deposition detection model to obtain the battery lithium deposition detection result.
[0030] A computer device comprises a memory, a processor, and a computer program stored in the memory and capable of running on the processor, wherein the processor implements the steps of the battery lithium deposition detection method when executing the computer program.
[0031] A computer-readable storage medium stores a computer program, and when the computer program is executed by a processor, the steps of the battery lithium deposition detection method are implemented.
[0032] Compared with the prior art, the present invention adopts the above technical solution and has the following technical effects:
[0033] The present invention provides a battery lithium deposition detection method and system based on CNN+LR to solve the problem of using single information to judge lithium deposition in the prior art. The discharge curve features are extracted through a one-dimensional convolutional network, and then a logistic regression model is used to distinguish lithium deposition; all information of the discharge curve can be extracted, greatly improving the accuracy of lithium deposition detection. BRIEF DESCRIPTION OF THE DRAWINGS
[0034] Figure 1 It is a discharge curve convolution extraction feature map of the present invention;
[0035] Figure 2 It is a flow chart of the battery lithium deposition detection method of the present invention. DETAILED DESCRIPTION
[0036] The embodiments of the present invention are described in detail below, and examples of the embodiments are shown in the accompanying drawings. The embodiments described below with reference to the accompanying drawings are exemplary and are only used to explain the present invention, and cannot be interpreted as limiting the present invention.
[0037] The present invention proposes a battery lithium deposition detection method based on CNN+LR to solve the problem of using single information to judge lithium deposition in the prior art. The discharge curve features are extracted through a one-dimensional convolutional neural network (CNN), and then the logistic regression model (LR) is used to distinguish lithium deposition. Figure 2 As shown, the specific steps are as follows:
[0038] Step S1, collecting battery discharge data and lithium deposition conditions;
[0039] Step S2: The data collected from the production line contains a small amount of missing or outliers. In view of this, the collected data is processed as follows:
[0040] Outlier processing: Outlier elimination uses the 6σ principle, and the calculation formula is as follows:
[0041]
[0042] Among them, U represents the upper bound of 6σ, L represents the lower bound of 6σ, and d i Indicates the time series data of the battery discharge curve, n indicates the amount of data, Represents the data mean. All data outside the upper and lower bounds are removed.
[0043] Missing value processing: The interval window mean filling method is used, and the calculation formula is as follows:
[0044]
[0045] Among them, w′ represents the window size (even number), i represents the position of the missing data point, and p represents the filled value.
[0046] Step S3, after data preprocessing, use a one-dimensional convolutional network to extract the features of the battery cell discharge curve time series data, as follows:
[0047] conv=(conv1,conv2,conv3)
[0048] Among them, conv represents convolution and, the first layer conv1 uses a 1×5 convolution kernel, the second layer conv2 uses a 1×3 convolution kernel, and the third layer conv3 uses a 1×3 convolution kernel, such as Figure 1 shown.
[0049] Step S4, after the one-dimensional convolution extracts the features, the logistic regression model is input, as follows:
[0050]
[0051] Among them, X represents the feature set, 1 in the X set is to adapt the intercept of the logistic regression model, m represents the feature dimension, n represents the amount of data, x represents the feature value, Y represents the label set, and y is 0 or 1 (0 represents no lithium deposition, 1 represents lithium deposition).
[0052] The logistic regression model is as follows:
[0053]
[0054] Among them, w is the model parameter, b is the intercept, x is the feature, and f(x) represents the logistic regression model.
[0055] The loss function is as follows:
[0056]
[0057] Train the logistic regression model using gradient descent.
[0058] The convolution kernel parameters and logistic regression model are obtained. The logistic regression model is expressed as follows:
[0059]
[0060] Step S5, obtaining real-time online battery cell discharge timing data, performing outlier elimination and missing value filling through a data preprocessing module, performing feature extraction through a feature extraction module, and performing model prediction through a model processing module.
[0061] The present invention also proposes a battery lithium deposition detection system, which includes a data acquisition module, a data preprocessing module, a feature extraction module, a model processing module and a detection output module; wherein,
[0062] The data acquisition module is used to obtain the battery cell discharge curve time series data and the corresponding battery lithium deposition situation, and is also used to obtain the battery cell discharge curve time series data in real time;
[0063] The data preprocessing module is used to preprocess the acquired battery cell discharge curve time series data, and is also used to preprocess the battery cell discharge curve time series data acquired in real time;
[0064] The feature extraction module is used to extract features from the preprocessed battery cell discharge curve time series data, and is also used to extract features from the preprocessed real-time battery cell discharge curve time series data;
[0065] The model processing module is used to take the features extracted from the pre-processed battery discharge curve time series data as input, and the corresponding battery lithium deposition situation as output, train the logistic regression model, and obtain the battery lithium deposition detection model;
[0066] The detection output module is used to input the features extracted from the preprocessed real-time battery cell discharge curve time series data into the battery lithium deposition detection model to obtain the battery lithium deposition detection results.
[0067] Based on the same inventive concept, an embodiment of the present application provides a computer device, including a memory, a processor, and a computer program stored in the memory and executable on the processor. When the processor executes the computer program, the steps of the aforementioned battery lithium deposition detection method are implemented.
[0068] Based on the same inventive concept, an embodiment of the present application provides a computer-readable storage medium, which stores a computer program. When the computer program is executed by a processor, the steps of the aforementioned battery lithium deposition detection method are implemented.
[0069] Those skilled in the art will appreciate that embodiments of the present invention may be provided as methods, systems, or computer program products. Therefore, the present invention may take the form of a complete hardware embodiment, a complete software embodiment, or an embodiment combining software and hardware. Moreover, the present invention may take the form of a computer program product implemented on one or more computer-usable storage media (including but not limited to disk storage, CD-ROM, optical storage, etc.) containing computer-usable program code.
[0070] The present invention is described with reference to flowcharts and / or block diagrams of methods, devices (systems), and computer program products according to embodiments of the present invention. It should be understood that each process and / or block in the flowchart and / or block diagram, as well as the combination of processes and / or blocks in the flowchart and / or block diagram, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, a special-purpose computer, an embedded processor, or other programmable data processing device to produce a machine, so that the instructions executed by the processor of the computer or other programmable data processing device generate instructions for implementing the processes in the flowchart and / or block diagram. Figure 1 A process or multiple processes and / or boxes Figure 1 A device that provides the functions specified in a block or multiple blocks.
[0071] These computer program instructions may also be stored in a computer-readable memory capable of directing a computer or other programmable data processing device to operate in a specific manner, so that the instructions stored in the computer-readable memory produce an article of manufacture comprising an instruction device, which implements the process Figure 1 A process or multiple processes and / or boxes Figure 1A function specified in one or more boxes.
[0072] These computer program instructions can also be loaded onto a computer or other programmable data processing device so that a series of operating steps are executed on the computer or other programmable device to produce a computer-implemented process, thereby providing instructions for implementing the process. Figure 1 A process or multiple processes and / or boxes Figure 1 The steps for the functions specified in one or more boxes.
[0073] The above embodiments are only for illustrating the technical idea of the present invention, and cannot be used to limit the protection scope of the present invention. Any changes made on the basis of the technical solution in accordance with the technical idea proposed by the present invention shall fall within the protection scope of the present invention.
Claims
1. A battery lithium deposition detection method, characterized in that: The steps include: Step 1, obtaining the battery cell discharge curve time series data and the corresponding battery lithium deposition situation; Step 2, preprocessing the battery cell discharge curve time series data obtained in step 1; Step 3, extracting features from the preprocessed battery cell discharge curve time series data; Step 4, taking the features extracted in step 3 as input and the corresponding battery lithium deposition situation as output, training the logistic regression model to obtain a battery lithium deposition detection model; Step 5, obtaining the time series data of the battery cell discharge curve in real time, and performing the same processing as steps 2 and 3, inputting the features extracted from the real-time time series data into the battery lithium plating detection model for detection, and obtaining the lithium plating detection result.
2. The battery lithium deposition detection method according to claim 1, characterized in that: In step 2, the preprocessing includes removing outliers and filling missing values, wherein: The 6σ principle is used to eliminate outliers, and all data outside the upper and lower bounds are eliminated. The calculation formula is as follows: Among them, U and L represent the upper and lower bounds of 6σ respectively, and d i represents the time series data of the battery discharge curve, n represents the total amount of data before outliers are removed, Indicates all d i The mean of After removing outliers, the interval window mean filling method is used to fill the missing values. The calculation formula is as follows: Among them, p represents the filling value, w′ represents the window size, w′ is an even number, and i represents the missing data position.
3. The battery lithium deposition detection method according to claim 1, characterized in that: In the step 3, a one-dimensional convolutional network is used to extract the features of the preprocessed battery cell discharge curve time series data. The one-dimensional convolutional network includes three layers of convolution, the first layer of convolution uses a 1×5 convolution kernel, the second layer of convolution uses a 1×3 convolution kernel, and the third layer of convolution uses a 1×3 convolution kernel.
4. The battery lithium deposition detection method according to claim 1, characterized in that: In step 4, the logistic regression model is trained using the gradient descent method, and the loss function is as follows: Among them, L(W) represents the loss function, Y represents the actual category, and X i Represents data features, y i Represents the predicted category, and n represents the amount of preprocessed data.
5. A battery lithium deposition detection system, characterized in that: The battery lithium deposition detection system includes: a data acquisition module, a data preprocessing module, a feature extraction module, a model processing module and a detection output module; wherein, The data acquisition module is used to obtain the battery cell discharge curve time series data and the corresponding battery lithium deposition situation, and is also used to obtain the battery cell discharge curve time series data in real time; The data preprocessing module is used to preprocess the acquired battery cell discharge curve time series data, and is also used to preprocess the battery cell discharge curve time series data acquired in real time; The feature extraction module is used to extract features from the preprocessed battery cell discharge curve time series data, and is also used to extract features from the preprocessed real-time battery cell discharge curve time series data; The model processing module is used to take the features extracted from the pre-processed battery discharge curve time series data as input, and the corresponding battery lithium deposition situation as output, train the logistic regression model, and obtain the battery lithium deposition detection model; The detection output module is used to input the features extracted from the pre-processed real-time battery cell discharge curve time series data into the battery lithium deposition detection model to obtain the battery lithium deposition detection result.
6. A computer device comprising a memory, a processor, and a computer program stored in the memory and capable of running on the processor, characterized in that: When the processor executes the computer program, the steps of the battery lithium deposition detection method as described in any one of claims 1 to 4 are implemented.
7. A computer-readable storage medium storing a computer program, characterized in that: When the computer program is executed by a processor, the steps of the battery lithium deposition detection method as described in any one of claims 1 to 4 are implemented.
Citation Information
Cited By
Sodium ion battery online sodium precipitation detection method, system, equipment and medium
CN120539607A
