A method, system, equipment and medium for detecting the pre-lithiation depth of graphite
By using image recognition technology and model prediction, the accuracy and cost issues of graphite pre-lithiation depth detection have been solved, achieving rapid, accurate, and low-cost detection results.
Patent Information
- Application Number
- CN202310754957.4
- Authority / Receiving Office
- CN · China
- Patent Type
- Patents(China)
- Current Assignee / Owner
- Filing Date
- 2023-06-25
- Publication Date
- 2025-11-14
- Estimated Expiration
- 2043-06-25
AI Technical Summary
Existing methods for detecting the depth of graphite prelithiation suffer from problems such as long testing time, high cost, and inaccurate results, making it difficult to achieve rapid and accurate detection.
Image recognition technology is used to preprocess graphite images, extract features, and map lithiation depth. A lithiation depth quantification model is used for prediction, and a convolutional neural network and a linear regression model are combined to achieve rapid and accurate detection of graphite prelithiation depth.
It enables rapid and accurate detection of the pre-lithiation depth of graphite, reduces detection costs, and does not damage the graphite material, thus offering advantages of high efficiency and high precision.
Smart Images
Figure CN116797571B_ABST
Abstract
Description
Technical Field
[0001] This invention relates to the field of lithium-ion battery testing technology, and in particular to a method, system, equipment and medium for detecting the pre-lithiation depth of graphite. Background Technology
[0002] Lithium-ion batteries are widely used high-performance batteries, and graphite is one of the main electrode materials. In battery production, graphite pre-lithiation is a crucial process. During preparation, graphite reacts with lithium ions to pre-intercalate a certain amount of lithium ions, thus offsetting the irreversible capacity loss during the first cycle and improving the capacity and cycle life of the graphite electrode. The pre-lithiation depth of graphite is one of the key factors affecting lithium-ion battery performance. Insufficient pre-lithiation depth results in no improvement in battery capacity and energy density. Excessive pre-lithiation depth can lead to dendrite growth and metallic lithium deposition during charge and discharge, shortening battery life. Therefore, real-time monitoring and control of graphite pre-lithiation depth is of great significance for improving battery performance and lifespan.
[0003] Currently, the main methods for detecting the pre-lithiation depth of graphite include physical and chemical testing methods. Physical testing methods include electrochemical testing and X-ray diffraction testing, but these methods suffer from problems such as long testing time, high cost, and difficulty in accurately reflecting the pre-lithiation depth of graphite. Chemical testing methods include chemical analysis and mass spectrometry analysis. Although the test results are accurate, they require a large number of chemical reagents and instruments, and the cost is also relatively high.
[0004] The rise of image recognition technology, with its high precision, efficiency, and non-destructive testing advantages, perfectly meets the needs of rapid screening of electrode materials. Furthermore, the different visual characteristics of graphite anodes in different lithium-intercalation states satisfy the capture requirements of image recognition methods. Therefore, developing a rapid, accurate, and low-cost method and system for detecting the depth of graphite pre-lithiation by combining image recognition technology has significant practical implications. Summary of the Invention
[0005] The purpose of this invention is to provide a method, system, device, and medium for detecting the prelithiation depth of graphite, so as to achieve rapid, accurate, and low-cost detection of the prelithiation depth of graphite.
[0006] To achieve the above objectives, the present invention provides the following solution:
[0007] A method for detecting the pre-lithiation depth of graphite includes:
[0008] Acquire the target graphite image;
[0009] The target graphite image is preprocessed to obtain a target grayscale image;
[0010] Feature extraction is performed on the target grayscale image to obtain target grayscale feature values;
[0011] A lithiation depth quantization model is used to map the target grayscale feature value to lithiation depth to obtain the predicted lithiation depth value; the lithiation depth quantization model is a correlation model between grayscale feature value and lithiation depth value.
[0012] Optionally, the process of determining the lithiation depth quantification model includes:
[0013] Obtain a training dataset; the training dataset includes several sample graphite images and the measured lithiation depth values corresponding to each sample graphite image;
[0014] The sample graphite image is preprocessed to obtain a sample grayscale image;
[0015] Feature extraction is performed on the sample grayscale image to obtain sample grayscale feature values;
[0016] A linear regression model was used to linearly correlate the sample grayscale feature values and the measured lithiation depth values corresponding to the graphite images of each sample, thus obtaining a lithiation depth quantification model.
[0017] Optionally, the target graphite image is preprocessed to obtain a target grayscale image, specifically including:
[0018] The target graphite image is enhanced to obtain an enhanced target image;
[0019] The enhanced target image is filtered to obtain a denoised target image;
[0020] The target denoised image is converted to grayscale to obtain a target grayscale image.
[0021] Optionally, the graphite pre-lithiation depth detection method further includes:
[0022] Determine whether the predicted lithiation depth value is within the standard range of lithiation depth, and obtain the determination result;
[0023] If the predicted lithiation depth value is within the standard range of lithiation depth, then the judgment result is that the target graphite electrode is well pre-lithiated.
[0024] If the predicted lithiation depth value is less than the lower limit of the standard range of lithiation depth values, then the judgment result is that the target graphite electrode is under-lithiated.
[0025] If the predicted lithiation depth value is greater than the upper limit of the standard range of lithiation depth, then the judgment result is that the target graphite electrode is over-lithiated.
[0026] If the judgment result indicates that the target graphite electrode is under-lithiated or over-lithiated, a corresponding alarm signal is generated; if the judgment result indicates that the target graphite electrode is well-lithiated, a completion signal is generated.
[0027] Optionally, feature extraction is performed on the target grayscale image to obtain target grayscale feature values, specifically including:
[0028] Image recognition technology is used to determine the gray value of each pixel in the target grayscale image, and the average value of the gray values of all pixels is calculated to obtain the target grayscale feature value.
[0029] Optionally, feature extraction is performed on the target grayscale image to obtain target grayscale feature values, specifically including:
[0030] The target grayscale image is subjected to feature extraction using a feature extraction model to obtain target grayscale feature values; the feature extraction model is obtained based on training of a convolutional neural network.
[0031] Optionally, the convolutional neural network includes: an input layer, a convolutional layer, a pooling layer, and a fully connected layer connected in sequence; the convolutional layer has a kernel size of 3×3 and 32 channels; the pooling layer uses max pooling and has a pooling window size of 2×2; and the fully connected layer has 128 neurons.
[0032] A graphite pre-lithiation depth detection system, comprising:
[0033] The image acquisition module is used to acquire the target graphite image;
[0034] The image processing module is used to preprocess the target graphite image to obtain a target grayscale image;
[0035] The feature extraction module is used to extract features from the target grayscale image to obtain target grayscale feature values;
[0036] The lithiation depth prediction module is used to map the target grayscale feature value to lithiation depth using a lithiation depth quantization model to obtain the predicted lithiation depth value; the lithiation depth quantization model is a correlation model between grayscale feature value and lithiation depth value.
[0037] An electronic device includes a memory and a processor, the memory storing a computer program, and the processor running the computer program to enable the electronic device to perform the graphite pre-lithiation depth detection method described above.
[0038] A computer-readable storage medium storing a computer program that, when executed by a processor, implements the above-described method for detecting the depth of graphite pre-lithiation.
[0039] According to specific embodiments provided by the present invention, the present invention discloses the following technical effects:
[0040] The graphite pre-lithiation depth detection method provided by this invention preprocesses the target graphite image to obtain a target grayscale image, and further extracts the target grayscale feature values. Then, using a correlation model between the grayscale feature values and the lithiation depth value (i.e., a lithiation depth quantification model), the grayscale feature values are mapped to the lithiation depth to obtain a predicted lithiation depth value. Since the color of graphite changes with the lithium-ion concentration, this characteristic allows the calculation of the lithium-ion concentration of graphite by determining the grayscale feature values. This invention analyzes the lithiation depth of the graphite surface by calculating the grayscale feature values, thereby achieving real-time monitoring of the graphite pre-lithiation depth. It can quickly and accurately obtain the lithiation depth value of the target graphite electrode, and this process does not require destructive testing of the graphite, reducing detection costs. Attached Figure Description
[0041] To more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the drawings used in the embodiments will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.
[0042] Figure 1 A flowchart of the graphite pre-lithiation depth detection method provided by the present invention;
[0043] Figure 2 A diagram of the device for implementing the graphite pre-lithiation depth detection method provided by the present invention;
[0044] Figure 3 A detailed flowchart of the graphite pre-lithiation depth detection method provided by the present invention;
[0045] Figure 4 This is a schematic diagram of the lithiation depth quantification model provided by the present invention.
[0046] Symbol explanation:
[0047] Conveyor belt-1, camera-2, processor-3, remote terminal-4. Detailed Implementation
[0048] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.
[0049] The purpose of this invention is to provide a method, system, device, and medium for detecting the prelithiation depth of graphite, so as to achieve rapid, accurate, and low-cost detection of the prelithiation depth of graphite.
[0050] To make the above-mentioned objects, features and advantages of the present invention more apparent and understandable, the present invention will be further described in detail below with reference to the accompanying drawings and specific embodiments.
[0051] Example 1
[0052] This invention provides a method for detecting the pre-lithiation depth of graphite. For example... Figure 1 , Figure 2 and Figure 3 As shown, the method includes:
[0053] Step S1: Acquire the target graphite image. The target graphite image is an image of the electrode edge and central region of the target graphite electrode, acquired using an industrial high-definition camera.
[0054] like Figure 2 As shown, the graphite electrode is located on conveyor belt 1 and is transported via conveyor belt 1. Camera 2 is located directly above conveyor belt 1 and acquires images of the graphite electrode on conveyor belt 1 under fixed light intensity conditions. The acquisition area can cover the edge and center area of the electrode. Image sampling is performed at fixed time intervals, and the acquired images are transmitted to processor 3 for image processing. In addition, if the electrode is not uniformly lithiated, a stop signal is issued if a single pixel with a large difference in grayscale value appears in the same sampling area at a fixed time point.
[0055] Furthermore, to avoid the impact of environmental changes on grayscale values, camera 2 needs to be equipped with a supplementary light source, with a brightness adjustable from 20,000 to 110,000 LUX. Additionally, camera 2 can also be fitted with an adjustable focal length lens. The acquired electrode images or videos are in 4K resolution and used as input for analysis.
[0056] Step S2: Preprocess the target graphite image to obtain a target grayscale image.
[0057] This step specifically includes: enhancing the target graphite image to obtain an enhanced target image; filtering the enhanced target image to obtain a denoised target image; and converting the denoised target image to grayscale to obtain a grayscale target image.
[0058] Specifically, image processing techniques are used to preprocess digital images, including enhancing, denoising, and converting them to grayscale to remove background noise and enhance image contrast. Image processing techniques include, but are not limited to, image enhancement algorithms, median filtering algorithms, and grayscale conversion algorithms.
[0059] Step S3: Extract features from the target grayscale image to obtain target grayscale feature values.
[0060] As an optional implementation, image recognition technology is used to determine the gray value of each pixel in the target grayscale image, and the average value of the gray values of all pixels is calculated to obtain the target grayscale feature value.
[0061] Furthermore, when collecting a large number of pixels, it is necessary to simplify the pixels to reduce the amount of data to be processed.
[0062] As another optional implementation, a feature extraction model is used to extract features from the target grayscale image to obtain target grayscale feature values; the feature extraction model is based on a convolutional neural network trained on it. The convolutional neural network is trained using a grayscale-processed training dataset, and the model parameters are updated through backpropagation and optimization algorithms. During training, the input is a grayscale image, and the target is the average grayscale value. In this embodiment, the input image is 300×300 pixels in size and is used as a single-channel grayscale image input.
[0063] The convolutional neural network comprises: an input layer, a convolutional layer, a pooling layer, and a fully connected layer connected in sequence; the convolutional layer has a kernel size of 3×3 and 32 channels; the pooling layer uses max pooling with a pooling window size of 2×2; and the fully connected layer has 128 neurons.
[0064] Furthermore, the feature extraction model can be trained and optimized by importing a large number of graphite anode images with different lithiation depths to improve the model's accuracy and robustness.
[0065] Step S4: Using a lithiation depth quantization model, the target grayscale feature value is mapped to lithiation depth to obtain a predicted lithiation depth value; the lithiation depth quantization model is a correlation model between grayscale feature values and lithiation depth values.
[0066] The process of determining the lithiation depth quantization model includes: acquiring a training dataset; the training dataset includes several sample graphite images and the measured lithiation depth values corresponding to each sample graphite image; preprocessing the sample graphite images to obtain sample grayscale images; extracting features from the sample grayscale images to obtain sample grayscale feature values; and using a linear regression model to linearly correlate the sample grayscale feature values corresponding to each sample graphite image with the measured lithiation depth values to obtain the lithiation depth quantization model.
[0067] In addition, a linear function can be used to linearly fit the sample grayscale feature values and the measured lithiation depth values corresponding to each sample graphite image, thus obtaining the correlation model between the two, i.e., the lithiation depth quantification model, see [link to relevant documentation]. Figure 4.
[0068] Specifically, for the processed image, the extracted grayscale values are input into the lithiation depth quantization model for comparison. The training dataset in the lithiation depth quantization model includes graphite sample images with different lithiation depths and their corresponding lithiation depth values. The images and lithiation depth values used in the training dataset are obtained by disassembling graphite half-cells after constant current charging to different states of charge. The correlation between grayscale values and lithiation depth values is generated based on the provided training dataset. Image recognition technology is used to statistically analyze the grayscale value of each pixel in the image, and the average grayscale value of the image is calculated by summing and averaging. The average grayscale value is then fitted and correlated with the given image lithiation depth value to achieve a correspondence between the image grayscale value and the lithiation depth value.
[0069] Furthermore, the method further includes: determining whether the predicted lithiation depth value is within the standard range of lithiation depth, and obtaining a determination result; if the predicted lithiation depth value is within the standard range of lithiation depth, the determination result indicates that the target graphite electrode is well pre-lithiated; if the predicted lithiation depth value is less than the lower limit of the standard range of lithiation depth, the determination result indicates that the target graphite electrode is insufficiently pre-lithiated; if the predicted lithiation depth value is greater than the upper limit of the standard range of lithiation depth, the determination result indicates that the target graphite electrode is excessively pre-lithiated; if the determination result indicates that the target graphite electrode is insufficiently pre-lithiated or excessively pre-lithiated, a corresponding alarm signal is generated; if the determination result indicates that the target graphite electrode is well pre-lithiated, a completion signal is generated.
[0070] Specifically, if the lithiation depth value corresponding to the characteristic value of the tested sample meets the standard value range, a completion signal is issued, and the matched lithiation depth value is transmitted to the output module. If it exceeds or falls below the standard lithiation depth value range, an alarm signal is issued. The standard lithiation depth value is divided into three intervals, defined as insufficient pre-lithiation, good pre-lithiation, and excessive pre-lithiation, based on the lithiation depth value. The standard lithiation depth value range is calculated based on the initial irreversible capacity of the graphite material and the difference in capacity between the positive and negative electrodes. The alarm signal, the completion signal, and the matched lithiation depth value are all displayed through the remote terminal 4 to remind relevant personnel to handle the situation promptly.
[0071] Example 2
[0072] To implement the method corresponding to Embodiment 1 above and achieve the corresponding functions and technical effects, a graphite pre-lithiation depth detection system is provided below. This system includes:
[0073] The image acquisition module is used to acquire the target graphite image.
[0074] The image processing module is used to preprocess the target graphite image to obtain a target grayscale image.
[0075] The feature extraction module is used to extract features from the target grayscale image to obtain target grayscale feature values.
[0076] The lithiation depth prediction module is used to map the target grayscale feature value to lithiation depth using a lithiation depth quantization model to obtain the predicted lithiation depth value; the lithiation depth quantization model is a correlation model between grayscale feature value and lithiation depth value.
[0077] Example 3
[0078] This invention also provides an electronic device, including a memory and a processor. The memory stores a computer program, and the processor runs the computer program to enable the electronic device to perform the graphite pre-lithiation depth detection method of Embodiment 1. The electronic device may be a server.
[0079] Furthermore, such as Figure 2 As shown, the electronic device also includes an image acquisition module and a remote terminal 4. The image acquisition module, located directly above the conveyor belt 1, consists of a camera 2, a supplementary light source, and a communication interface. The image acquisition module acquires images of the target graphite and transmits them to the processor 3. The processor 3 includes an image acquisition module, an image processing module, a feature extraction module, and a lithiation depth prediction module. These modules work collaboratively through data transmission to obtain a predicted lithiation depth value and determine whether it falls within the standard lithiation depth range. The remote terminal 4 displays an alarm signal, a completion signal, and the matched lithiation depth value based on the judgment result.
[0080] In addition, the present invention also provides a computer-readable storage medium storing a computer program that, when executed by a processor, implements the graphite pre-lithiation depth detection method in Embodiment 1.
[0081] In summary, the graphite pre-lithiation depth detection method, system, equipment, and medium provided by this invention, based on image recognition technology, can achieve real-time monitoring and high-throughput screening of graphite pre-lithiation depth, while also offering advantages of high precision and high efficiency. Secondly, this technology does not require destructive testing of graphite, avoiding damage and waste. Thirdly, this technology can be applied to various types of lithium-ion battery production lines, demonstrating excellent versatility.
[0082] The various embodiments in this specification are described in a progressive manner, with each embodiment focusing on its differences from other embodiments. Similar or identical parts between embodiments can be referred to interchangeably. For the systems disclosed in the embodiments, since they correspond to the methods disclosed in the embodiments, the descriptions are relatively simple; relevant parts can be referred to the method section.
[0083] This document uses specific examples to illustrate the principles and implementation methods of the present invention. The descriptions of the above embodiments are only for the purpose of helping to understand the method and core ideas of the present invention. Furthermore, those skilled in the art will recognize that, based on the ideas of the present invention, there will be changes in the specific implementation methods and application scope. Therefore, the content of this specification should not be construed as a limitation of the present invention.
Claims
1. A method for detecting the pre-lithiation depth of graphite, characterized in that, include: Acquire the target graphite image; The target graphite image is preprocessed to obtain a target grayscale image; Feature extraction is performed on the target grayscale image to obtain target grayscale feature values; A lithiation depth quantization model is used to map the target grayscale feature value to lithiation depth to obtain the predicted lithiation depth value. The lithiation depth quantification model is a correlation model between grayscale feature values and lithiation depth values; The process of determining the lithiation depth quantification model includes: Obtain a training dataset; the training dataset includes several sample graphite images and the measured lithiation depth values corresponding to each sample graphite image; The sample graphite image is preprocessed to obtain a sample grayscale image; Feature extraction is performed on the sample grayscale image to obtain sample grayscale feature values; A linear regression model was used to linearly correlate the sample grayscale feature values and the measured lithiation depth values corresponding to the graphite images of each sample, thus obtaining a lithiation depth quantification model.
2. The method for detecting the pre-lithiation depth of graphite according to claim 1, characterized in that, The target graphite image is preprocessed to obtain a target grayscale image, specifically including: The target graphite image is enhanced to obtain an enhanced target image; The enhanced target image is filtered to obtain a denoised target image; The target denoised image is converted to grayscale to obtain a target grayscale image.
3. The method for detecting the pre-lithiation depth of graphite according to claim 1, characterized in that, Also includes: Determine whether the predicted lithiation depth value is within the standard range of lithiation depth, and obtain the determination result; If the predicted lithiation depth value is within the standard range of lithiation depth, then the judgment result is that the target graphite electrode is well pre-lithiated. If the predicted lithiation depth value is less than the lower limit of the standard range of lithiation depth values, then the judgment result is that the target graphite electrode is under-lithiated. If the predicted lithiation depth value is greater than the upper limit of the standard range of lithiation depth, then the judgment result is that the target graphite electrode is over-lithiated. If the judgment result indicates that the target graphite electrode is under-lithiated or over-lithiated, a corresponding alarm signal is generated; if the judgment result indicates that the target graphite electrode is well-lithiated, a completion signal is generated.
4. The method for detecting the pre-lithiation depth of graphite according to claim 1, characterized in that, Feature extraction is performed on the target grayscale image to obtain target grayscale feature values, specifically including: Image recognition technology is used to determine the gray value of each pixel in the target grayscale image, and the average value of the gray values of all pixels is calculated to obtain the target grayscale feature value.
5. The method for detecting the pre-lithiation depth of graphite according to claim 1, characterized in that, Feature extraction is performed on the target grayscale image to obtain target grayscale feature values, specifically including: The target grayscale image is subjected to feature extraction using a feature extraction model to obtain target grayscale feature values; the feature extraction model is obtained based on training of a convolutional neural network.
6. The method for detecting the pre-lithiation depth of graphite according to claim 5, characterized in that, The convolutional neural network includes: an input layer, a convolutional layer, a pooling layer, and a fully connected layer connected in sequence; the convolutional layer has a kernel size of 3×3 and 32 channels; the pooling layer uses max pooling and has a pooling window size of 2×2; the fully connected layer has 128 neurons.
7. A graphite pre-lithiation depth detection system, characterized in that, include: The image acquisition module is used to acquire the target graphite image; The image processing module is used to preprocess the target graphite image to obtain a target grayscale image; The feature extraction module is used to extract features from the target grayscale image to obtain target grayscale feature values; The lithiation depth prediction module is used to map the target grayscale feature value to lithiation depth using a lithiation depth quantization model to obtain the predicted lithiation depth value. The lithiation depth quantification model is a correlation model between grayscale feature values and lithiation depth values; The process of determining the lithiation depth quantification model includes: Obtain a training dataset; the training dataset includes several sample graphite images and the measured lithiation depth values corresponding to each sample graphite image; The sample graphite image is preprocessed to obtain a sample grayscale image; Feature extraction is performed on the sample grayscale image to obtain sample grayscale feature values; A linear regression model was used to linearly correlate the sample grayscale feature values and the measured lithiation depth values corresponding to the graphite images of each sample, thus obtaining a lithiation depth quantification model.
8. An electronic device, characterized in that, The device includes a memory and a processor, the memory being used to store a computer program, and the processor running the computer program to cause the electronic device to perform the graphite pre-lithiation depth detection method as described in any one of claims 1 to 6.
9. A computer-readable storage medium, characterized in that, It stores a computer program that, when executed by a processor, implements the graphite pre-lithiation depth detection method as described in any one of claims 1 to 6.
Citation Information
Patent Citations
Fuel cell graphite plate surface molding smoothness detection method and device
CN112345548A
Method and system for predicting remaining service life of battery
CN114636932A