Patent Application
20240221150
This application relates to the field of battery technologies, and in particular, to an electrode plate wrinkling detection method and system, a terminal, and a storage medium.
Existing batteries are subject to various defects during the winding production process, such as electrode plate wrinkling defects, electrode plate folding defects, and tab folding defects. Cells with wrinkled electrode plates have risks such as lithium precipitation induced short circuit and low capacity, so they receive attention during the production process. Generally, the cells produced by winding machines are gathered on a logistics line, and the appearance of the cells is observed manually and the cells are disassembled in a detection method to determine whether wrinkled electrode plates exist. After the electrode plate wrinkling defects of the cells have been detected, a scheduling personnel is assigned to adjust the machine. This detection method has the problems of lagged detection results and low accuracy of the detection results.
Therefore, there is an urgent need for a method capable of accurately detecting wrinkling of electrode plates in a timely manner.
An objective of this application is to propose an electrode plate wrinkling detection method and system, a terminal, and a storage medium, so as to solve the technical problems of lagged detection results obtained and low accuracy of the detection results of a detection method in the prior art.
According to a first aspect, an embodiment of this application provides an electrode plate wrinkling detection method. The method includes: acquiring images of cells with wrinkled electrode plates and images of cells with non-wrinkled electrode plates; performing type labeling processing on the images of the cells with wrinkled electrode plates and the images of the cells with non-wrinkled electrode plates, and constructing a database by using the labeled images; based on the database, training via a convolutional neural network to generate an electrode plate wrinkling detection model; testing and calibrating the electrode plate wrinkling detection model by using images outside the database; and performing electrode plate wrinkling detection on cell images obtained in a real-time manner during a battery winding process by using the tested and calibrated electrode plate wrinkling detection model.
In the technical solution of this embodiment of this application, in the electrode plate wrinkling detection method, the images of the cells with wrinkled electrode plates and the images of the cells with non-wrinkled electrode plates are first acquired, where the images of the cells with wrinkled electrode plates are images of defective cells with multiple types of wrinkled electrode plates having different wrinkling positions and different wrinkling textures, and the images of the cells with non-wrinkled electrode plates include images of intact cells and images of normal cells with wrinkling, stripe patterns, stains, and the like on separators caused by a production process. Then the acquired images are labeled with types, and the database is constructed using the labeled images. Based on the database, an artificial intelligence (AI) deep learning electrode plate wrinkling detection model is trained using the convolutional neural network. The generated electrode plate wrinkling detection model is capable of checking cell images and categorizing them into corresponding types. In addition, in this method, the images outside the database are further used to test the electrode plate wrinkling detection model, so as to improve the detection accuracy. To detect whether the cells have wrinkled electrode plates in a real-time manner and determine the defective cells, the cell images acquired during the battery winding process are taken as detection targets, which can not only discover the defective cells in a timely manner, but also remind a scheduling personnel to adjust the machine in a timely manner, reducing the electrode plate wrinkling probability in the whole winding process. In addition, this also avoids false negatives and false positives caused by observation on the appearance of the cells and the problem of loss of manpower and material resources caused by disassembly of the normal cells due to disassembly of the cells for determining whether the cells have wrinkled electrode plates.
In some embodiments, after the acquiring images of cells with wrinkled electrode plates and images of cells with non-wrinkled electrode plates, the electrode plate wrinkling detection method further includes: pre-processing the acquired images of the cells with wrinkled electrode plates and images of the cells with non-wrinkled electrode plates, where the pre-processing includes: performing cleansing on the images of the cells with wrinkled electrode plates and the images of the cells with non-wrinkled electrode plates, removing interference parts, and broadening, enhancing, and brightening the images of the cells with wrinkled electrode plates and the images of the cells with non-wrinkled electrode plates.
Pre-processing the images can avoid misjudgment caused by image factors, thus improving the accuracy of the subsequent type labeling processing and ultimately ensuring the accuracy of data in the database. Removing the interference parts of the images may refer to removing mechanical mechanisms in the images.
In some embodiments, the performing type labeling processing on the images of the cells with wrinkled electrode plates and the images of the cells with non-wrinkled electrode plates, and constructing a database by using the labeled images include: labeling the images of the cells with wrinkled electrode plates and the images of the cells with non-wrinkled electrode plates with types including wrinkling of electrode plate, wrinkling of separator, stripe pattern on separator, stain on separator, and full qualification, and constructing a database by using the images labeled with the types including wrinkling of electrode plate, wrinkling of separator, stripe pattern on separator, stain on separator, and full qualification.
Cells with wrinkled electrode plates have risks such as lithium precipitation induced short circuit and low capacity, so only the cells corresponding to the wrinkling of electrode plate type are labeled as defective cells. Wrinkling, stripe patterns, stains, and the like on separators do not affect the cell performance, and the problems such as stripe patterns and strains on the separators are unavoidable during the production process, so other cases except wrinkling of electrode plate are all labeled as normal cells.
In some embodiments, the testing and calibrating the electrode plate wrinkling detection model by using images outside the database includes: testing the electrode plate wrinkling detection model by using the images outside the database; under a condition that a test result meets a target requirement, performing electrode plate wrinkling detection on cell images obtained in a real-time manner during a battery winding process by using the tested and calibrated electrode plate wrinkling detection model; or under a condition that the test result does not meet the target requirement, retraining the electrode plate wrinkling detection model by modifying parameters of the convolutional neural network until the test result meets the target requirement, thereby improving the detection accuracy of the electrode plate wrinkling detection model.
In some embodiments, the performing electrode plate wrinkling detection on cell images obtained in a real-time manner during a battery winding process by using the tested and calibrated electrode plate wrinkling detection model includes: shooting to-be-inspected cells during a winding process to obtain images of the cells, inputting the images of the cells into the electrode plate wrinkling detection model for detection, when determining that defective cells with wrinkled electrode plates exist, discharging the defective cells into a defective product tank, or when determining that no defective cell with wrinkled electrode plates exists, discharging normal cells into a next production procedure.
This method has high accuracy and fast response speed. To expand sources of the cell images, an electrode plate wrinkling detection apparatus may be disposed at a position of a cell winding machine, specifically at a position directly facing the cell or at a position of an electrode plate tangent to the cell. The corresponding obtained cell images include images of a to-be-inspected cell and/or images of the electrode plate tangent to the to-be-inspected cell. The images of the electrode plate tangent to the to-be-inspected cell are typically images of an anode plate tangent to the to-be-inspected cell. This is because an anode plate and a cathode plate are made of different materials, and the cathode plate is hard and less prone to wrinkling.
In some embodiments, when it is determined that the defective cells with wrinkled electrode plates exist, an electrode plate wrinkling alarm prompt is further given so as to remind the scheduling personnel to adjust the machine in a timely manner, reducing the electrode plate wrinkling probability in the subsequent winding process.
According to a second aspect, an embodiment of this application provides an electrode plate wrinkling detection system. The system includes: an electrode plate wrinkling detection apparatus configured to acquire images of cells with wrinkled electrode plates and images of cells with non-wrinkled electrode plates; a processing module configured to: perform type labeling processing on the images of the cells with wrinkled electrode plates and the images of the cells with non-wrinkled electrode plates; and construct a database by using the labeled images; a training module configured to: based on the database, train via a convolutional neural network to generate an electrode plate wrinkling detection model; a testing module configured to test and calibrate the electrode plate wrinkling detection model by using images outside the database; and a detection module configured to perform electrode plate wrinkling detection on cell images obtained in a real-time manner during a battery winding process by using the tested and calibrated electrode plate wrinkling detection model.
In some embodiments, the electrode plate wrinkling detection system further includes: a pre-processing module configured to pre-process the acquired images of the cells with wrinkled electrode plates and images of the cells with non-wrinkled electrode plates, where the pre-processing includes: performing cleansing on the images of the cells with wrinkled electrode plates and the images of the cells with non-wrinkled electrode plates, removing interference parts, and broadening, enhancing, and brightening the images of the cells with wrinkled electrode plates and the images of the cells with non-wrinkled electrode plates.
In some embodiments, the processing module is further configured to: label the images of the cells with wrinkled electrode plates and the images of the cells with non-wrinkled electrode plates with types including wrinkling of electrode plate, wrinkling of separator, stripe pattern on separator, stain on separator, and full qualification; and construct a database by using the images labeled with the types including wrinkling of electrode plate, wrinkling of separator, stripe pattern on separator, stain on separator, and full qualification.
In some embodiments, the testing module is further configured to: test the electrode plate wrinkling detection model by using the images outside the database; under a condition that a test result meets a target requirement, communicatively transmit a signal indicating that the requirement is met to the detection module; or under a condition that the test result does not meet the target requirement, retrain the electrode plate wrinkling detection model by modifying parameters of the convolutional neural network until the test result meets the target requirement, and communicatively transmit the signal indicating that the requirement is met to the detection module; and the detection module is further configured to: in response to the received signal indicating that the requirement is met, perform the electrode plate wrinkling detection.
In some embodiments, the electrode plate wrinkling detection apparatus is further configured to: shoot to-be-inspected cells during a winding process to obtain images of the cells; and transmit the images of the cells to the detection module; and the detection module is further configured to: receive the images of the cells; check the images of the cells by using the electrode plate wrinkling detection model; when determining that defective cells with wrinkled electrode plates exist, output a signal indicating discharging the defective cells into a defective product tank; or when determining that no defective cell with wrinkled electrode plates exists, output a signal indicating discharging normal cells into a next production procedure.
In some embodiments, when determining that the defective cells with wrinkled electrode plates exist, the detection module further gives an electrode plate wrinkling alarm prompt.
According to a third aspect, an embodiment of this application provides a terminal. The terminal includes: at least one processor; and a memory communicatively connected to the at least one processor, where the memory stores instructions executable by the at least one processor, and the instructions are executed by the at least one processor to implement the electrode plate wrinkling detection method according to any one of the foregoing embodiments.
According to a fourth aspect, an embodiment of this application provides a storage medium storing a computer program, where the computer program is executed by a processor to implement the electrode plate wrinkling detection method according to any one of the foregoing embodiments.
The foregoing description is merely an overview of the technical solution of this application. For a better understanding of the technical means in this application such that they can be implemented according to the content of the specification, and to make the above and other objectives, features and advantages of this application more obvious and easier to understand, the following describes specific embodiments of this application.
In the accompanying drawings, unless otherwise specified, same reference signs throughout a plurality of accompanying drawings indicate the same or similar components or elements. These accompanying drawings are not necessarily drawn to scale. It should be noted that these accompanying drawings are merely intended to illustrate some embodiments disclosed in this application and should not be construed as any limitations on the scope of this application.
The following describes in detail the embodiments of technical solutions of this application with reference to the accompanying drawings. The following embodiments and accompanying drawing are merely intended to describe the technical solutions of this application more clearly and therefore are used as just examples which do not constitute any limitations on the protection scope of this application. The drawings schematically illustrate only parts related to the technical solutions of this application and do not represent their actual structure as products.
Unless otherwise defined, all technical and scientific terms used herein shall have the same meanings as commonly understood by those skilled in the art to which this application relates. The terms used herein are intended to merely describe the specific embodiments rather than to limit this application. The terms “include”, “comprise”, and “have” and any other variations thereof in the specification, claims and brief description of drawings of this application are intended to cover non-exclusive inclusions.
In the description of the embodiments of this application, the terms “first”, “second”, and the like are merely intended to distinguish between different objects, and shall not be understood as any indication or implication of relative importance or any implicit indication of the number, sequence or primary-secondary relationship of the technical features indicated. Unless otherwise stated, any method step disclosed herein are not necessarily performed in the definite sequence disclosed.
In this specification, reference to “embodiment” means that specific features, structures or characteristics described with reference to the embodiment may be incorporated in at least one embodiment of this application. The word “embodiment” appearing in various places in the specification does not necessarily refer to the same embodiment or an independent or alternative embodiment that is exclusive of other embodiments. It is explicitly or implicitly understood by persons skilled in the art that the embodiments described herein may be combined with other embodiments.
In the description of the embodiments of this application, the term “and/or” is only an associative relationship for describing associated objects, indicating that three relationships may be present. For example, A and/or B may indicate the following three cases: presence of only A, presence of both A and B, and presence of only B. In addition, the character “/” in this specification generally indicates an “or” relationship between contextually associated objects.
In the description of the embodiments of this application, the term “a plurality of” means more than two (inclusive). Similarly, “a plurality of groups” means more than two (inclusive) groups, and “a plurality of pieces” means more than two (inclusive) pieces.
In the description of the embodiments of this application, the orientations or positional relationships indicated by the technical terms “center”, “longitudinal” “transverse”, “length”, “width”, “thickness”, “upper”, “lower”, “front”, “rear”, “left”, “right”, “vertical”, “horizontal”, “top”, “bottom”, “inside”, “outside”, “clockwise”, “counterclockwise”, “axial”, “radial”, “circumferential”, and the like are based on the orientations or positional relationships as shown in the accompanying drawings. These terms are merely for ease and brevity of the description of the embodiments of this application rather than indicating or implying that the means or components mentioned must have specific orientations, or must be constructed or manipulated according to specific orientations, and therefore shall not be construed as any limitations on embodiments of this application.
In the description of the embodiments of this application, unless otherwise specified and defined explicitly, the terms “mount”, “connect”, “join”, and “fasten” should be understood in their general senses. For example, they may refer to a fixed connection, a detachable connection, or an integral connection, may refer to a mechanical connection or an electrical connection, and may refer to a direct connection, an indirect connection via an intermediate medium, an internal communication between two elements, or an interaction between two elements. Persons of ordinary skill in the art can understand specific meanings of these terms in this application as appropriate to specific situations.
Currently, from a perspective of the market development, application of traction batteries is being more extensive. Traction batteries have been not only used in energy storage power supply systems such as hydroelectric power plants, thermal power plants, wind power plants, and solar power plants, but also widely used in many other fields including electric transportation tools such as electric bicycles, electric motorcycles, and electric vehicles, military equipment, and aerospace. With the continuous expansion of application fields of traction batteries, market demands for traction batteries are also increasing.
A battery production process involves production of cell, and the quality of the cell directly affects the quality of the battery. A manufacturing procedure of a wound cell specifically includes: arranging cut electrode plates and separators in sequence, and then rotating a winding mandrel to wind the electrode plates and the separators to form a jelly roll wrapped layer by layer and served as a cell. Various defects occur during the winding production process, such as electrode plate wrinkling defects, electrode plate folding defects, and tab folding defects. Cells with wrinkled electrode plates have risks such as lithium precipitation induced short circuit and low capacity, so they require special attention.
The applicant has noted that in the prior art, after a batch of cells are wound, cells produced by winding machines are generally gathered on a logistics line, and wrinkling of electrode plate is detected in manners of manually observing the appearance of the cells and disassembling the cells. The manner of manually observing the appearance of the cells is not only low in accuracy, but also prone to false positives and false negatives that lead to outflow of defective cells, while the manner of disassembling the cells is prone to loss of normal cells due to disassembly of the normal cells, and a lot of manpower and material resources are wasted in a process of distinguishing between the normal cells and the defective cells. In addition, a long time is consumed in a process from production to discovery of wrinkling of electrode plate, and during this process, winding machines producing the defective cells continuously produce cells with wrinkled electrode plates, which increases production costs.
To discover the defective cells and the corresponding winding machines producing the defective cells in a timely manner, the applicant has found through research that conditions of electrode plates can be detected during winding production of the cells so as to discover wrinkling of electrode plate at an early stage, thus discovering the defective cells and the corresponding winding machines producing the defective cells at an early stage. Based on the foregoing considerations, the applicant, through in-depth research, has designed an electrode plate wrinkling detection method and system, a terminal, and a storage medium.
An electrode plate wrinkling detection method provided in some embodiments of this application includes:
In the electrode plate wrinkling detection method according to some embodiments of this application, the images of the cells with wrinkled electrode plates and the images of the cells with non-wrinkled electrode plates are first acquired, where the images of the cells with wrinkled electrode plates are images of defective cells with multiple types of wrinkled electrode plates having different wrinkling positions and different wrinkling textures, and the images of the cells with non-wrinkled electrode plates include images of intact cells and images of normal cells with wrinkling, stripe patterns, stains, and the like on separators caused by a production process. Then the acquired images are labeled with types, and the database is constructed using the labeled images. Based on the database, an artificial intelligence (AI) deep learning electrode plate wrinkling detection model is trained using the convolutional neural network. The generated electrode plate wrinkling detection model is capable of checking cell images and categorizing them into corresponding types. In addition, in this method, the images outside the database are further used to test the electrode plate wrinkling detection model, so as to improve the detection accuracy. To detect whether the cells have wrinkled electrode plates in a real-time manner and determine the defective cells, the cell images acquired during the battery winding process are taken as detection targets, which can not only discover the defective cells in a timely manner, but also remind a scheduling personnel to adjust the machine in a timely manner, reducing the electrode plate wrinkling probability in the whole winding process. In addition, this also avoids false negatives and false positives caused by observation on the appearance of the cells and the problem of loss of manpower and material resources caused by disassembly of the normal cells due to disassembly of the cells for determining whether the cells have wrinkled electrode plates.
It should be understood that the electrode plate may include an anode plate and a cathode plate. In terms of a lithium-ion battery, one or more of slurries such as graphite, silicon, silicon oxide, and graphite-mixed silicon are applied on the anode plate to form an anode active substance coating layer, and a slurry containing lithium ions is applied on the cathode plate to form a cathode active substance coating layer. A separator is sandwiched between the anode plate and the cathode plate.
The separator is a film used to separate positive and negative electrodes during electrolysis reaction so as to prevent energy loss caused by direct reaction in an electrolytic cell. In a structure of a lithium battery, the separator is one of important inner components. The performance of the separator determines an interface structure and internal resistance of the battery and directly affects the capacity, cycling performance, and safety performance of the battery. A separator with excellent performance plays an important role in improving the overall performance of the battery.
A cell is a single electrochemical cell including positive and negative electrodes formed by stacking or winding electrode plates and separators in a specified order, and is typically not used directly.
The electrode plate wrinkling detection method disclosed in some embodiments of this application may be used in, but is not limited to, a cell production and manufacturing process including a winding procedure of lithium-ion batteries, sodium-ion batteries, and the like. The electrode plate wrinkling detection method provided in some embodiments of this application helps to discover the wrinkled electrode plates in a timely manner during production and manufacturing of cells, thus discovering the defective cells and the winding machines producing the defective cells in a timely manner. In this way, the defective cells are accurately picked out and the winding machines producing the defective cells are adjusted so as to avoid continuous production of the defective cells, achieving the purpose of reducing production loss while avoiding outflow of the defective cells.
The following will describe the electrode plate wrinkling detection method and system, the terminal, and the storage medium provided in this application in detail with reference to specific embodiments.
Referring to
Step S1: Acquire images of cells with wrinkled electrode plates and images of cells with non-wrinkled electrode plates.
In the image acquisition stage, no less than 200 images including different types of images of cells with wrinkled electrode plates and images of cells with non-wrinkled electrode plates are shot by an electrode plate wrinkling detection apparatus. In some embodiments, more than 1000 images can be shot so as to cover as much as possible various types of defects occurring during the winding process. The images of the cells with wrinkled electrode plates are images of defective cells with multiple types of wrinkled electrode plates having different wrinkling positions and different wrinkling textures, and the images of the cells with non-wrinkled electrode plates include images of intact cells and images of normal cells with wrinkling, stripe patterns, stains, and the like on separators caused by a production process.
Referring to
Step S11′: Perform cleansing on the images of the cells with wrinkled electrode plates and the images of the cells with non-wrinkled electrode plates.
Step S12′: Remove interference parts.
Step S13′: Broaden, enhance, and brighten the images of the cells with wrinkled electrode plates and the images of the cells with non-wrinkled electrode plates.
The broadening treatment refers to generating similar but different to-be-trained samples by making a series of random changes to to-be-trained images, thereby enlarging the size of a training dataset. Randomly changing training samples can reduce the dependence of the model on some attributes, thereby improving generalization capability of the model. For example, an image is cut in different ways so that objects of interest appear at different positions, thereby reducing the dependence of the model on the positions where the objects appear. A fundamental objective is to increase the amount of information so that the model obtains higher-layer semantic information and gains a deeper understanding of the problem, thereby improving the accuracy (objectively, because of the enriched amount of information, it also has an effect of preventing overfitting).
The enhancing treatment refers to enhancing useful information in an image, which is a distortion process that aims to improve a visual effect of an image for an application scenario of a given image. Purposefully emphasizing overall or local characteristics of an image makes an original unclear image become clear or emphasizes some features of interest, expands difference between features of different objects in the image, and suppresses features of disinterest, thereby improving the quality of the image, enriching the amount of information, and enhancing the image interpretation and recognition effect to meet the needs of some special analyses.
The brightening treatment refers to adjusting factors such as brightness and color to reduce sensitivity of the model to color.
Pre-processing the acquired images of the cells with wrinkled electrode plates and images of the cells with non-wrinkled electrode plates can avoid misjudgment caused by image factors, thus improving the accuracy of the subsequent type labeling processing and ultimately ensuring the accuracy of data in the database. Removing the interference parts of the image may refer to removing mechanical mechanisms in the image and may also refer to removing other unnecessary interference parts in the image other than the cell. The broadening treatment on the image refers to generating similar but different training samples by making a series of random changes, thereby enlarging the size of the training dataset. The enhancing treatment on the image can improve the quality of the image. The brightening treatment on the image makes it easier to see some detailed information of the image.
In some embodiments, the electrode plate wrinkling detection method further includes the following step. Step S2: Perform type labeling processing on the images of the cells with wrinkled electrode plates and the images of the cells with non-wrinkled electrode plates, and construct a database by using the labeled images. Step S2 may include: labeling the images of the cells with wrinkled electrode plates and the images of the cells with non-wrinkled electrode plates with types including wrinkling of electrode plate, wrinkling of separator, stripe pattern on separator, stain on separator, and full qualification; and constructing a database by using the images labeled with the types including wrinkling of electrode plate, wrinkling of separator, stripe pattern on separator, stain on separator, and full qualification. Depending on actual needs, more types of labeling may be provided. Cells with wrinkled electrode plates have risks such as lithium precipitation induced short circuit and low capacity, so in some embodiments, only the cells corresponding to the wrinkling of electrode plate type are labeled as defective cells. Wrinkling, stripe patterns, stains, and the like on separators do not affect the cell performance, and the problems such as stripe pattern and strain on the separators are unavoidable during the production process, so in some embodiments, other cases except wrinkling of electrode plate are all labeled as normal cells. It should be understood that wrinkling of electrode plate typically refers to wrinkling of an anode plate. This is because the material of the anode plate is relatively soft, and wrinkling of electrode plate occurs during winding due to the quality of the material of the electrode plate, the winding process, or the equipment problem of the winding machine.
The image recognition technology is a technology for processing and analyzing the image via a computer to recognize different modes of targets and objects. In the conventional industry, an industrial camera is typically used to shoot images of to-be-inspected targets, and a computer is used to recognize simple image features such as width, length, and shape formed by gray values. However, the recognition accuracy is low for complex and variable image features. In some embodiments of this application, the electrode plate wrinkling detection method further includes the following step. Step S3: Based on the database, train via a convolutional neural network to generate an electrode plate wrinkling detection model. Features of images of the cells with wrinkled electrode plates and the images of the cells with non-wrinkled electrode plates, such as wrinkle width and wrinkle texture of a wrinkled region, and wrinkling of separator, stripe pattern on separator, and stain on separator of a non-wrinkled region are extracted, and the convolutional neural network is used for repeated training to generate an AI deep learning electrode plate wrinkling detection model. The AI deep learning electrode plate wrinkling detection model is a categorizer model. For example, during the detection process, it can categorize images into five types: wrinkling of electrode plate, wrinkling of separator, stripe pattern on separator, stain on separator, and full qualification. The cells with wrinkled electrode plates are recognized and distinguished by the trained AI deep learning electrode plate wrinkling detection model, overcoming the problem of low image feature recognition accuracy in the prior art caused by recognition based on gray values. The electrode plate wrinkling detection model is generated based on the convolutional neural network. The convolutional neural network is capable of automatically extracting, learning, and pre-processing features such as texture, width, and length of images to generate a categorizer. After training and iteration with a large number of images with different types and features, an electrode plate wrinkling detection model with high image recognition accuracy can be obtained.
In some embodiments, to improve the detection accuracy of the electrode plate wrinkling detection model, the electrode plate wrinkling detection method further includes the following step. Step S4: Test and calibrate the electrode plate wrinkling detection model by using images outside the database.
Step S4 may include: testing the electrode plate wrinkling detection model by using the images outside the database; determining whether a test result meets a target requirement; under a condition that the test result meets the target requirement, performing the step of performing electrode plate wrinkling detection by using the electrode plate wrinkling detection model, that is, putting the electrode plate wrinkling detection model into use; or under a condition that the test result does not meet the target requirement, retraining the AI deep learning electrode plate wrinkling detection model by modifying parameters of the neural network, such as weights and feature values until the test result meets the target requirement.
Testing the electrode plate wrinkling detection model by using the images outside the database can determine the detection accuracy of the electrode plate wrinkling detection model according to the test result. In some embodiments, the target requirement may mean that the accuracy reaches more than 90%. Detection results include types such as wrinkling of electrode plate, wrinkling of separator, stripe pattern on separator, stain on separator, and full qualification, which can further improve the detection accuracy.
In some embodiments, the electrode plate wrinkling detection method further includes the following step. Step S5: Perform electrode plate wrinkling detection on cell images obtained in a real-time manner during a battery winding process by using the tested and calibrated electrode plate wrinkling detection model. Step S5 may include the following steps. S51: Start an electrode plate wrinkling detection apparatus. S52: Shoot to-be-inspected cells during a winding process to obtain images of the cells. S53: Pre-process the images of the cells. S54: Input the pre-processed images of the cells into the electrode plate wrinkling detection model for detection, and determine whether wrinkled electrode plates exist.
When it is determined that wrinkled electrode plates exist, a determining result is output as defective cells, and the defective cells are discharged into a defective product (NG) tank and distinguished from normal cells; or when it is determined that no wrinkled electrode plate exists, the determining result is output as normal cells, and the normal cells are discharged into a next production procedure. The electrode plate wrinkling detection method provided in the embodiments of this application has high accuracy and fast response speed. In this embodiment, the electrode plate wrinkling detection apparatus may be disposed at a position of a cell winding machine, specifically at a position directly facing the cell. The obtained cell images are images of the to-be-inspected cell.
In some embodiments, when it is determined that wrinkled electrode plates exist, an electrode plate wrinkling alarm prompt is first given and the defective cells are then discharged into the NG tank so as to remind the scheduling personnel to adjust the machine in a timely manner, reducing the electrode plate wrinkling probability in the subsequent winding process.
In the cell 400, the lower separator 101, the cathode plate 201, the upper separator 102, and the anode plate 202 are tightly wound together. The lower separator 101 and the upper separator 102 are made of thin and soft materials, and wrinkling of the cathode plate 201 and the anode plate 202 can be shown through the separator. Images of the cell 400 are shot using the CCD detection mechanism 900, and wrinkling of electrode plate is analyzed and determined using an AI algorithm, so that wrinkling of the cathode plate can be detected online in a real-time manner, with a detection process as shown in
In another embodiment of this application, under a condition that the position of the CCD detection mechanism 900 is not changed, the lower separator 101, the anode plate 202, the upper separator 102, and the cathode plate 201 (that is, positions of the cathode plate 201 and the anode plate 202 are exchanged) are wound in sequence to form the cell 400. A front anode plate 202 of a to-be-inspected cell 400 is shot using the CCD detection mechanism 900 to obtain various types of cell images as shown in
In other embodiments of this application, to expand sources of the cell images, the electrode plate wrinkling detection apparatus may be disposed at a position of a cell winding machine, specifically at a position of an electrode plate tangent to the cell. The corresponding obtained cell images are images of the electrode plate tangent to the to-be-inspected cell. The images of the electrode plate tangent to the to-be-inspected cell are typically images of an anode plate tangent to the to-be-inspected cell. This is because an anode plate and a cathode plate are made of different materials, and the cathode plate is hard and less prone to wrinkling.
In the electrode plate wrinkling detection method provided in some embodiments of this application, the images of the cells with wrinkled electrode plates and the images of the cells with non-wrinkled electrode plates are first acquired, then the acquired images are labeled with types, and a database is constructed using the labeled images. Based on the database, an AI deep learning electrode plate wrinkling detection model is trained using the convolutional neural network. The generated electrode plate wrinkling detection model is capable of checking cell images and categorizing them into corresponding types. In addition, in this method, the images outside the database are further used to test the electrode plate wrinkling detection model, so as to improve the detection accuracy. To detect whether the cells have wrinkled electrode plates in a real-time manner and determine the defective cells, the cell images acquired during the battery winding process are taken as detection targets, which can not only discover the defective cells in a timely manner, but also remind a scheduling personnel to adjust the machine in a timely manner, reducing the electrode plate wrinkling probability in the whole winding process. In addition, this also avoids false negatives and false positives caused by observation on the appearance of the cells and the problem of loss of manpower and material resources caused by disassembly of the normal cells due to disassembly of the cells for determining whether the cells have wrinkled electrode plates.
A second aspect of this application provides an electrode plate wrinkling detection system. The system includes: an electrode plate wrinkling detection apparatus configured to acquire images of cells with wrinkled electrode plates and images of cells with non-wrinkled electrode plates; a processing module configured to: perform type labeling processing on the images of the cells with wrinkled electrode plates and the images of the cells with non-wrinkled electrode plates; and construct a database by using the labeled images; a training module configured to: based on the database, train via a convolutional neural network to generate an electrode plate wrinkling detection model; a testing module configured to test and calibrate the electrode plate wrinkling detection model by using images outside the database; and a detection module configured to perform electrode plate wrinkling detection on cell images obtained in a real-time manner during a battery winding process by using the tested and calibrated electrode plate wrinkling detection model.
In some embodiments, the electrode plate wrinkling detection system further includes: a pre-processing module configured to pre-process the acquired images of the cells with wrinkled electrode plates and images of the cells with non-wrinkled electrode plates, where the pre-processing includes: performing cleansing on the images of the cells with wrinkled electrode plates and the images of the cells with non-wrinkled electrode plates, removing interference parts, and broadening, enhancing, and brightening the images of the cells with wrinkled electrode plates and the images of the cells with non-wrinkled electrode plates.
In some embodiments, the processing module is further configured to: label the images of the cells with wrinkled electrode plates and the images of the cells with non-wrinkled electrode plates with types including wrinkling of electrode plate, wrinkling of separator, stripe pattern on separator, stain on separator, and full qualification; and construct a database by using the images labeled with the types including wrinkling of electrode plate, wrinkling of separator, stripe pattern on separator, stain on separator, and full qualification.
In some embodiments, the testing module is further configured to: test the electrode plate wrinkling detection model by using the images outside the database; under a condition that a test result meets a target requirement, communicatively transmit a signal indicating that the requirement is met to the detection module; or under a condition that the test result does not meet the target requirement, retrain the electrode plate wrinkling detection model by modifying parameters of the convolutional neural network until the test result meets the target requirement, and communicatively transmit the signal indicating that the requirement is met to the detection module; and the detection module is further configured to: in response to the received signal indicating that the requirement is met, perform the electrode plate wrinkling detection.
In some embodiments, the electrode plate wrinkling detection apparatus is further configured to: shoot to-be-inspected cells during a winding process to obtain images of the cells; and transmit the images of the cells to the detection module; and the detection module is further configured to: receive the images of the cells; check the images of the cells by using the electrode plate wrinkling detection model; when determining that defective cells with wrinkled electrode plates exist, output a signal indicating discharging the defective cells into a defective product tank; or when determining that no defective cell with wrinkled electrode plates exists, output a signal indicating discharging normal cells into a next production procedure.
In some embodiments, when determining that the defective cells with wrinkled electrode plates exist, the detection module further gives an electrode plate wrinkling alarm prompt.
Referring to
The memory 62 and the processor 61 are connected via a bus. The bus may include any quantity of interconnected buses and bridges, and the bus connects various circuits of one or more processors 61 and memories 62 together. The bus can also connect various other circuits such as a peripheral device, a voltage regulator, and a power management circuit together, which is not further described herein. A bus interface provides an interface between the bus and a transceiver. The transceiver may be a single element or a plurality of elements, such as a plurality of receivers and transmitters, providing units for communicating with various other apparatuses on a transmission medium. Data processed by the processor is transmitted over the wireless medium via an antenna. The antenna can also receive the data and transmit the data to the processor 61.
The processor 61 is responsible for managing the bus and general processing, and can also provide various functions including timing, peripheral interfacing, voltage regulation, power management, and other control functions. The memory 62 may be configured to store data used by processor 61 when performing operations.
A fourth aspect of this application provides a storage medium storing a computer program, where the computer program is executed by a processor to implement the electrode plate wrinkling detection method according to any one of the embodiments of this application.
Persons skilled in the art should understand that all or some of the steps of the method in the foregoing embodiments may be accomplished by hardware associated with program instructions. The program instructions are stored in the storage medium and include a plurality of instructions used to enable a device (which may be a single-chip microcomputer, a chip, or the like) or a processor to perform all or some of the steps of the method in the various embodiments of this application. The foregoing storage medium includes various media capable of storing program code, such as a USB flash drive, a portable hard drive, a read-only memory (ROM), a random access memory (RAM), a magnetic disk, or an optical disc.
In conclusion, it should be noted that the foregoing embodiments are for description of the technical solutions of this application only rather than for limiting this application. Although this application has been described in detail with reference to the foregoing embodiments, persons of ordinary skill in the art should appreciate that they can still make modifications to the technical solutions described in the embodiments or make equivalent replacements to some or all technical features thereof without departing from the scope of the technical solutions of the embodiments of this application. All such modifications and equivalent replacements shall fall within the scope of claims and specification of this application. In particular, as long as there is no structural conflict, the various technical features mentioned in the embodiments can be combined in any manner. This application is not limited to the specific embodiments disclosed in this specification but includes all technical solutions falling within the scope of the claims.
| Number | Date | Country | Kind |
|---|---|---|---|
| 202111249085.3 | Oct 2021 | CN | national |
This application is a continuation of International Application PCT/CN2022/124389, filed on Oct. 10, 2022, which claims priority to Chinese Patent Application No. 202111249085.3, filed on Oct. 26, 2021 and entitled “ELECTRODE PLATE WRINKLING DETECTION METHOD AND SYSTEM, TERMINAL, AND STORAGE MEDIUM”, which is incorporated herein by reference in its entirety.
| Number | Date | Country | |
|---|---|---|---|
| Parent | PCT/CN2022/124389 | Oct 2022 | WO |
| Child | 18610287 | US |
