A method and device for identifying stirring quality of lithium battery slurry

Abstract

The present application provides a method and device for identifying the stirring quality of lithium battery slurry, wherein the method comprises: obtaining a slurry image to be detected, wherein the slurry image to be detected is a current color image of the slurry to be detected collected during the stirring process of the lithium battery slurry; grayscale processing is performed on the slurry image to be detected to obtain a target grayscale image; the target grayscale image is input into a trained machine learning model to predict the stirring quality information of the slurry image to be detected, and the stirring quality information is used to evaluate the kneading solid content or viscosity of the slurry to be detected. The present application solves the technical problem of inaccuracy caused by manual identification of stirring quality in the prior art, and solves the technical problem of low detection efficiency caused by sampling detection, and achieves the technical effect of improving detection accuracy and efficiency.

Description

Technical Field

[0001] The present application relates to the technical field of lithium battery manufacturing, and in particular to a method and device for identifying the stirring quality of lithium battery slurry. Background Art

[0002] In the process of manufacturing lithium batteries, it is necessary to stir the battery cell slurry of the lithium battery and conduct quality inspection on the stirred slurry. In the prior art, the slurry stirring process is mainly dry stirring, and the basic process is dry powder mixing-wetting-dispersion-adjusting viscosity. In the wetting stage, it is necessary to judge the kneading solid content, which can be understood as the process of gradually adding solvents to mix. In the viscosity adjustment stage, the viscosity is judged, which can be understood as the process of gradually adding a small amount of NMP (N-methylpyrrolidone, a solvent commonly used in lithium battery production, used to dissolve lithium salts and polymer electrolytes) to dilute. Manual visual inspection is used to determine whether the kneading solid content of the slurry in the wetting stage meets the requirements and whether the viscosity in the viscosity adjustment stage meets the requirements. The evaluation method relies on manual visual inspection during the stirring process and manual sampling to detect the stirring and kneading solid content and viscosity. However, the manual visual inspection method is highly subjective, and its reliability is low due to the different abilities of visual inspectors; although the manual sampling detection method has high accuracy, it is not conducive to improving mass production efficiency and requires regular shutdown detection. Summary of the invention

[0003] In view of this, the purpose of the present application is to at least provide a method and device for identifying the stirring quality of lithium battery slurry, by collecting an image of the slurry to be detected during the stirring process of the lithium battery slurry, gray-processing the image of the slurry to be detected, and then inputting it into a pre-trained machine learning model, so as to predict the stirring quality information of the slurry to be detected through the machine learning model, thereby solving the technical problem of inaccuracy caused by relying on manual identification of stirring quality in the prior art, and solving the technical problem of low detection efficiency caused by relying on sampling detection, thereby achieving the technical effect of improving detection accuracy and efficiency.

[0004] This application mainly includes the following aspects:

[0005] In a first aspect, an embodiment of the present application provides a method for identifying the stirring quality of lithium battery slurry, the method comprising: obtaining an image of a slurry to be detected, wherein the image of the slurry to be detected is a current color image of the slurry to be detected collected during the stirring process of the lithium battery slurry; performing grayscale processing on the image of the slurry to be detected to obtain a target grayscale image; inputting the target grayscale image into a trained machine learning model to predict stirring quality information of the image of the slurry to be detected, and the stirring quality information is used to evaluate the kneading solid content or viscosity of the slurry to be detected.

[0006] Optionally, the grayscale processing of the slurry image to be detected to obtain a target grayscale image includes: converting the slurry image to be detected into an initial grayscale image based on a triplet of each pixel of the slurry image to be detected; and performing histogram equalization on the initial grayscale image to obtain a target grayscale image of the slurry image to be detected.

[0007] Optionally, the triplet includes a first component, a second component and a third component, and according to the triplet of each pixel of the slurry image to be detected, the slurry image to be detected is converted into an initial grayscale image, including: for each pixel of the slurry image to be detected, multiplying the first component in the triplet of the pixel by the corresponding first coefficient to obtain a first product, multiplying the second component in the triplet of the pixel by the corresponding second coefficient to obtain a second product, multiplying the third component in the triplet of the pixel by the corresponding third coefficient to obtain a third product, and taking the sum of the first product, the second product and the third product as the grayscale value of the pixel to obtain the initial grayscale image.

[0008] Optionally, the slurry image to be detected is obtained in the following manner: a lithium battery slurry picture is acquired by an image acquisition device according to a preset acquisition cycle, wherein the preset acquisition cycle is used to limit the position of a stirring blade in the lithium battery slurry picture; a preset area of ​​the lithium battery slurry picture is used as the slurry image to be detected to remove the portion other than the lithium battery slurry in the lithium battery slurry picture.

[0009] Optionally, the machine learning model is trained in the following manner: obtaining multiple historical slurry images collected during the stirring process of lithium battery slurry; performing grayscale processing on each historical slurry image to obtain a historical grayscale image corresponding to each historical slurry image; determining historical stirring quality information corresponding to each historical grayscale image, and the historical stirring quality information is used to describe the stirring quality of the lithium battery slurry in each historical grayscale image; using each historical grayscale image as sample data and the historical stirring quality information corresponding to each historical slurry image as a label to train the machine learning model.

[0010] Optionally, the method further comprises: sending stirring quality information of the image of the slurry to be detected to a display screen to prompt a user of the kneading solid content or viscosity of the slurry to be detected.

[0011] In the second aspect, an embodiment of the present application also provides a device for identifying the stirring quality of lithium battery slurry, the device comprising: an acquisition module, used to acquire a slurry image to be detected, the slurry image to be detected is a current color image of the slurry to be detected collected during the stirring process of the lithium battery slurry; a determination module, used to perform grayscale processing on the slurry image to be detected to obtain a target grayscale image; a prediction module, used to input the target grayscale image into a trained machine learning model to predict the stirring quality information of the slurry image to be detected, and the stirring quality information is used to evaluate the kneading solid content or viscosity of the slurry to be detected.

[0012] Optionally, the determination module is further used to convert the slurry image to be detected into an initial grayscale image according to the triplet of each pixel of the slurry image to be detected; and perform histogram equalization on the initial grayscale image to obtain a target grayscale image of the slurry image to be detected.

[0013] In a third aspect, an embodiment of the present application further provides an electronic device, comprising: a processor, a memory and a bus, wherein the memory stores machine-readable instructions executable by the processor, and when the electronic device is running, the processor and the memory communicate through the bus, and the machine-readable instructions are executed by the processor to execute the steps of the method for identifying the stirring quality of lithium battery slurry described in the first aspect or any possible implementation manner of the first aspect.

[0014] In a fourth aspect, an embodiment of the present application further provides a computer-readable storage medium, on which a computer program is stored. When the computer program is executed by a processor, the steps of the method for identifying the stirring quality of lithium battery slurry described in the first aspect or any possible embodiment of the first aspect are executed.

[0015] The embodiment of the present application provides a method and device for identifying the stirring quality of lithium battery slurry, the method comprising: obtaining a slurry image to be detected, the slurry image to be detected being a current color image of the slurry to be detected collected during the stirring process of the lithium battery slurry; grayscale processing the slurry image to be detected to obtain a target grayscale image; inputting the target grayscale image into a trained machine learning model to predict the stirring quality information of the slurry image to be detected, the stirring quality information being used to evaluate the kneading solid content or viscosity of the slurry to be detected. The present application collects the slurry image to be detected during the stirring process of the lithium battery slurry, grayscale processes the slurry image to be detected, and then inputs the pre-trained machine learning model to predict the stirring quality information of the slurry to be detected through the machine learning model, thereby solving the inaccurate technical problem caused by relying on manual identification of stirring quality in the prior art, and solving the technical problem of low detection efficiency caused by relying on sampling detection, thereby achieving the technical effect of improving detection accuracy and efficiency.

[0016] In order to make the above-mentioned objects, features and advantages of the present application more obvious and easy to understand, preferred embodiments are specifically cited below and described in detail with reference to the attached drawings. BRIEF DESCRIPTION OF THE DRAWINGS

[0017] In order to more clearly illustrate the technical solutions of the embodiments of the present application, the drawings required for use in the embodiments will be briefly introduced below. It should be understood that the following drawings only show certain embodiments of the present application and therefore should not be regarded as limiting the scope. For ordinary technicians in this field, other related drawings can be obtained based on these drawings without paying creative work.

[0018] Figure 1 A flow chart of a method for identifying stirring quality of lithium battery slurry provided in an embodiment of the present application is shown.

[0019] Figure 2 A flow chart showing the steps of performing grayscale processing on a slurry image to be detected to obtain a target grayscale image provided by an embodiment of the present application.

[0020] Figure 3 A functional module diagram of a lithium battery slurry stirring quality identification device provided in an embodiment of the present application is shown.

[0021] Figure 4 A schematic structural diagram of an electronic device provided in an embodiment of the present application is shown. DETAILED DESCRIPTION

[0022] To make the purpose, technical scheme and advantages of the embodiments of the present application clearer, the technical scheme in the embodiments of the present application will be clearly and completely described below in conjunction with the drawings in the embodiments of the present application. It should be understood that the drawings in the present application only serve the purpose of explanation and description and are not used to limit the scope of protection of the present application. In addition, it should be understood that the schematic drawings are not drawn in real proportion. The flowchart used in this application shows the operations implemented according to some embodiments of the present application. It should be understood that the operations of the flowchart can be implemented out of sequence, and the steps without logical context can be reversed in order or implemented simultaneously. In addition, those skilled in the art, under the guidance of the content of the present application, can add one or more other operations to the flowchart, or remove one or more operations from the flowchart.

[0023] In addition, the described embodiments are only a part of the embodiments of the present application, rather than all the embodiments. The components of the embodiments of the present application described and shown in the drawings here can be arranged and designed in various configurations. Therefore, the following detailed description of the embodiments of the present application provided in the drawings is not intended to limit the scope of the application claimed for protection, but merely represents the selected embodiments of the present application. Based on the embodiments of the present application, all other embodiments obtained by those skilled in the art without making creative work belong to the scope of protection of the present application.

[0024] In the prior art, the adhesion or viscosity of the lithium battery slurry during the stirring process is manually identified to determine the stirring state of the lithium battery slurry, but manual identification is prone to inaccurate identification; or, the adhesion or viscosity of the lithium battery slurry during the stirring process is detected by stopping the machine for sampling, but the detection is time-consuming, which affects the normal operation of the mixer and affects the efficiency.

[0025] Based on this, an embodiment of the present application provides a method and device for identifying the stirring quality of lithium battery slurry. The present application collects an image of the slurry to be detected during the stirring process of the lithium battery slurry, grayscale processes the image of the slurry to be detected, and then inputs it into a pre-trained machine learning model to predict the stirring quality information of the slurry to be detected through the machine learning model, thereby solving the technical problem of inaccuracy caused by relying on manual identification of stirring quality in the prior art, and solving the technical problem of low detection efficiency caused by relying on sampling detection, thereby achieving the technical effect of improving detection accuracy and efficiency.

[0026] The details are as follows:

[0027] See also Figure 1 , Figure 1 This is a flow chart of a method for identifying the stirring quality of lithium battery slurry provided in an embodiment of the present application. Figure 1As shown, the stirring quality identification method of lithium battery slurry provided in the embodiment of the present application includes the following steps:

[0028] S101: Acquire an image of a slurry to be detected.

[0029] The slurry image to be detected is a current color image of the slurry to be detected collected during the stirring process of the lithium battery slurry.

[0030] Among them, the raw materials constituting the lithium battery slurry are placed in a mixer for stirring. The mixer first stirs the raw materials to make the lithium battery slurry in a wetting stage. The wetting stage is used to fully mix the raw materials. The kneading solid content in the wetting stage is used to determine whether the raw materials are fully mixed. In other words, the kneading solid content is used to measure the mixing degree of the raw materials constituting the lithium battery slurry.

[0031] After the wetting stage is executed, dispersion is performed. After the dispersion stage is executed, solvent is continued to be added to enter the viscosity adjustment stage. The viscosity adjustment stage is used to adjust the viscosity of the lithium battery slurry. The viscosity of the lithium battery slurry during the stirring process is determined by adjusting the viscosity of the viscosity stage. In other words, viscosity is used to measure the viscosity of the lithium battery slurry. The wetting stage is a process in which the powder and liquid of the raw materials of the lithium battery slurry are mixed at the macro level so that the air or other impurities adsorbed on the surface of the powder are replaced by the liquid. This step requires the detection of the kneading solid content. The wetting of the powder in the liquid phase is an important prerequisite for whether the powder can be evenly dispersed. Poor wetting will cause agglomeration and agglomeration, which will affect the further dispersion and mixing later. There is no need for image confirmation in the dispersion stage, and there is no need to add slurry in the dispersion stage. It only needs to ensure that the stirring time reaches the preset dispersion stirring time, which means that the dispersion stage is over. The dispersion stage is only to make the solid particles agglomerate under the action of mechanical force at the micro level after the wetting stage, depolymerize and disperse, and ensure that the depolymerized slurry is stabilized and homogenized to prevent re-agglomeration.

[0032] The wetting stage, the dispersion stage and the viscosity adjustment stage are all performed in the mixer. Furthermore, the slurry image to be detected is a current color image obtained by photographing the slurry to be detected in the mixer during the stirring process of the lithium battery slurry; or, the slurry image to be detected is a preset area of ​​the current color image photographed by the image acquisition device during the stirring process of the lithium battery slurry.

[0033] The slurry image to be detected is obtained in the following manner: a lithium battery slurry image is acquired by an image acquisition device according to a preset acquisition cycle, wherein the preset acquisition cycle is used to limit the position of the stirring blade in the lithium battery slurry image; a preset area of ​​the lithium battery slurry image is used as the slurry image to be detected to remove the portion other than the lithium battery slurry in the lithium battery slurry image.

[0034] During the wetting stage and the viscosity adjustment stage, the stirring speed of the mixer is constant, and the preset acquisition cycle of the image acquisition device for capturing the lithium battery slurry image can be set according to the stirring speed and the installation position of the image acquisition device, thereby avoiding the lithium battery slurry image from including a large number of stirring blade parts.

[0035] Since the captured lithium battery slurry image may include parts other than the lithium battery slurry, such as the inner wall of the mixer and a small part of the blades, the lithium battery slurry part in the lithium battery slurry image can be retained by setting a preset area to remove the parts other than the lithium battery slurry to obtain the slurry image to be tested.

[0036] That is, the slurry image to be detected is the current color image captured by the image acquisition device during the stirring process of the lithium battery slurry, or the slurry image to be detected is a preset area of ​​the current color image captured by the image acquisition device during the stirring process of the lithium battery slurry.

[0037] That is, the preset area is the portion of the lithium battery slurry in the current color image captured by the image acquisition device during the stirring process of the lithium battery slurry. The preset area can be set by the user according to the position set by the image acquisition device.

[0038] S102: grayscale processing is performed on the slurry image to be detected to obtain a target grayscale image.

[0039] See also Figure 2 , Figure 2 A flowchart of the steps of grayscale processing the slurry image to be detected to obtain a target grayscale image provided by an embodiment of the present application. Figure 2 As shown, the grayscale processing of the slurry image to be detected to obtain a target grayscale image includes:

[0040] S1021: Convert the slurry image to be detected into an initial grayscale image according to the triplet of each pixel of the slurry image to be detected.

[0041] That is, each pixel of the slurry image to be detected corresponds to a triplet, which refers to the RGB components in a pixel, namely, red component, green component, and blue component. The first component is the red component, the second component is the green component, and the third component is the blue component.

[0042] The triplet includes a first component, a second component and a third component. The converting the slurry image to be detected into an initial grayscale image based on the triplet of each pixel of the slurry image to be detected includes: for each pixel of the slurry image to be detected, multiplying the first component in the triplet of the pixel by the corresponding first coefficient to obtain a first product, multiplying the second component in the triplet of the pixel by the corresponding second coefficient to obtain a second product, multiplying the third component in the triplet of the pixel by the corresponding third coefficient to obtain a third product, and taking the sum of the first product, the second product and the third product as the grayscale value of the pixel to obtain the initial grayscale image.

[0043] The grayscale value corresponding to each pixel is calculated by the following formula:

[0044] G i ′=a×R i +b×G i +c×B i (1)

[0045] As shown in formula (1), G i ′ represents the gray value of the i-th pixel, R i represents the first component of the i-th pixel, G i represents the second component of the i-th pixel, B i represents the third component of the i-th pixel, a represents the first component and the corresponding first coefficient, a can be set to 0.299, b represents the second component and the corresponding second coefficient, b can be set to 0.587, c represents the third component and the corresponding third coefficient, c can be set to 0.114.

[0046] Then, the slurry image to be detected is converted from a current color image to a grayscale image.

[0047] S1022: Performing histogram equalization on the initial grayscale image to obtain a target grayscale image of the slurry image to be detected.

[0048] Specifically, according to the grayscale value of each pixel of the initial grayscale image, the number of pixels of each grayscale value is counted; for each grayscale value, the number of pixels of the grayscale value is compared with the total number of pixels of the initial grayscale image to obtain the probability distribution of each grayscale value; according to the probability distribution of each grayscale value, the cumulative distribution function of each grayscale value is calculated; the cumulative distribution function of each grayscale value is multiplied by a preset grayscale value to obtain a target grayscale value of each grayscale value; the grayscale value of each pixel in the initial grayscale image is replaced with the corresponding target grayscale value to obtain a target grayscale image.

[0049] That is to say, for each grayscale value, the probability distribution of the grayscale value is summed with the probability distribution of each grayscale value smaller than the grayscale value to obtain the cumulative distribution function of the grayscale value.

[0050] The preset grayscale value is the upper limit of the grayscale value range of the preset grayscale level, or the preset grayscale value is the preset grayscale level minus 1. For example, if the preset grayscale level is 256, the grayscale value range is [0, 255], and the preset grayscale value is 255; if the preset grayscale level is 256, the preset grayscale value is 255.

[0051] S103: Input the target grayscale image into a trained machine learning model to predict the stirring quality information of the slurry image to be detected.

[0052] The stirring quality information is used to evaluate the kneading solid content or viscosity of the slurry to be detected. The stirring quality information includes: high kneading solid content, low kneading solid content, normal kneading solid content, high viscosity, low viscosity, and normal viscosity of the lithium battery slurry in the slurry image to be detected.

[0053] The machine learning model is trained in the following manner: a plurality of historical slurry images collected during the stirring process of the lithium battery slurry are obtained; each historical slurry image is gray-processed respectively to obtain a historical gray-scale image corresponding to each historical slurry image; historical stirring quality information corresponding to each historical gray-scale image is determined, and the historical stirring quality information is used to describe the stirring quality of the lithium battery slurry in each historical gray-scale image; each historical gray-scale image is used as sample data, and the historical stirring quality information corresponding to each historical slurry image is used as a label to train the machine learning model.

[0054] Exemplarily, multiple historical slurry images are collected during the stirring process of lithium battery slurry, and the historical slurry image is a historical color image captured by an image acquisition device during the stirring process of lithium battery slurry, or the historical slurry image is a preset area of ​​a historical color image captured by an image acquisition device during the stirring process of lithium battery slurry.

[0055] That is to say, the slurry image to be detected needs to correspond to the historical slurry image; when the historical slurry image is a historical color image taken by the image acquisition device during the stirring process of the lithium battery slurry, the slurry image to be detected is the current color image taken by the image acquisition device during the stirring process of the lithium battery slurry; when the historical slurry image is a preset area of ​​the historical color image taken by the image acquisition device during the stirring process of the lithium battery slurry, the slurry image to be detected is the preset area of ​​the current color image taken by the image acquisition device during the stirring process of the lithium battery slurry.

[0056] Each historical slurry image is gray-scaled in step S102 to obtain a historical gray-scale image corresponding to each historical slurry image. The historical gray-scale image is annotated by the user to annotate the historical mixing quality information corresponding to each historical gray-scale image.

[0057] The historical stirring quality information is used to evaluate the kneading solid content or viscosity of the lithium battery slurry in the historical grayscale image. The historical stirring quality information includes: high kneading solid content, low kneading solid content, normal kneading solid content, high viscosity, low viscosity, and normal viscosity of the lithium battery slurry in the historical grayscale image.

[0058] In the wetting stage, the kneading solid content changes from high to low. For example, when the kneading solid content is above 80%, the lithium battery slurry is too dry to form a mass; when the kneading solid content is between 80% and 70%, the lithium battery slurry is slightly dry when it forms a mass; when the kneading solid content is between 80% and 75%, the lithium battery slurry is slightly dry when it forms a mass; when the kneading solid content is between 75% and 67%, the lithium battery slurry forms a mass well, and there is no obvious moisture in the cross section; when the kneading solid content is between 67% and 65%, the lithium battery slurry forms a mass well, and there is appropriate moisture in the cross section; when the kneading solid content is between 65% and 62%, the lithium battery slurry can form a mass, but cannot be shaped, and the cross section has appropriate moisture; when the kneading solid content is between 62% and 60%, the lithium battery slurry can form a mass, but cannot be shaped, and the cross section is too wet; when the kneading solid content is below 60%, it is too wet and cannot form a mass.

[0059] That is to say, the kneading solid content is between 80% and 60% and is in the normal range. Furthermore, when the kneading solid content is above 80%, the lithium battery slurry is too dry to form a mass, and the kneading solid content is considered high; when the kneading solid content is between 80% and 60%, the lithium battery slurry can form a mass, and the kneading solid content is considered normal; when the kneading solid content is below 60%, the lithium battery slurry cannot form a mass, and the kneading solid content is considered low. As for the kneading solid content that is considered low, high or normal, it depends on the actual ratio of the raw materials of the lithium battery slurry.

[0060] During the viscosity adjustment stage, the viscosity changes from high to low. Viscosity also has different viscosities that can be detected. The normal viscosity range is determined according to the requirements, generally 4000-8000 mPa.s or 8000-12000 mPa.s. If it is greater than the upper limit of the normal viscosity range, the viscosity is considered high, and if it is less than the lower limit of the viscosity range, the viscosity is considered low. If it is within the normal viscosity range, the viscosity is considered normal. The normal viscosity range also needs to be set according to the actual ratio.

[0061] That is to say, the historical stirring quality information can be the percentage of kneaded solid content and specific viscosity data of the lithium battery slurry gradually detected during the historical stirring process of the lithium battery slurry, and the historical stirring quality information corresponding to each historical grayscale image is determined by the preset normal range of viscosity and normal range of kneaded solid content; or, the historical stirring quality information corresponding to each historical grayscale image is directly determined by the operator through manual visual inspection.

[0062] Since the lithium battery slurry is first in the wetting stage and then gradually enters the viscosity adjustment stage during the stirring process, therefore, each historical grayscale image can only correspond to one historical stirring quality information.

[0063] Therefore, each historical grayscale image is used as sample data, and the historical stirring quality information corresponding to each historical slurry image is used as a label to train the machine learning model, so that the trained machine learning model can identify the corresponding stirring quality information through the input grayscale image.

[0064] Furthermore, the stirring quality information includes high kneading solid content, low kneading solid content, normal kneading solid content, high viscosity, low viscosity, and normal viscosity for the lithium battery slurry in the slurry image to be detected. Among them, high kneading solid content, low kneading solid content, high viscosity, and low viscosity are abnormal quality, and normal kneading solid content and normal viscosity are normal quality.

[0065] Exemplarily, each historical grayscale image is used as sample data, and the historical mixing quality information of each historical grayscale image is used as a label to train a machine learning model.

[0066] Alternatively, for each historical grayscale image, the historical grayscale image is segmented into multiple cropped images of uniform size; each cropped image is data enhanced to obtain multiple enhanced images, wherein the data enhancement method includes: small-angle image rotation, image flipping, image random fusion, grid masking, and convex surface superposition; each enhanced image is used as sample data, and the historical mixing quality information of the historical grayscale image corresponding to each enhanced image is used as a label to train the machine learning model. Thus, by segmenting the historical grayscale image and performing data enhancement, the training set of the machine learning model is increased.

[0067] Alternatively, if the historical slurry image is a historical color image captured by an image acquisition device during the stirring process of the lithium battery slurry, the stirring detection area and historical stirring quality information of each historical slurry image marked by the user are determined, and the stirring detection area refers to the area in the historical slurry image where only the lithium battery slurry is present; each historical grayscale image is used as sample data, and the stirring detection area and historical stirring quality information of each historical grayscale image are used as labels to train the machine learning model. Furthermore, the slurry image to be detected is the current color image captured by the image acquisition device during the stirring process of the lithium battery slurry. After the slurry image to be detected is input into the trained machine learning model, the stirring detection area and stirring quality information of the slurry image to be detected can be obtained.

[0068] Specifically, all sample data need to be divided into a training sample set and a test sample set according to a preset ratio to train the machine learning model.

[0069] Exemplarily, the machine learning model in the present application can be a YOLOv5 model, which is trained and hyperparameter optimized by a training sample set to obtain a YOLOv5 model after the training is completed; the test sample set is input into the YOLOv5 model obtained after the training is completed, and the model evaluation index is calculated based on the obtained results, so as to determine whether the YOLOv5 model is trained based on the model evaluation index.

[0070] There are four main model evaluation indicators, namely precision, recall, mAP@0.5, and mAP@[.5:.95]. Among them, Recall reflects the ratio of the number of samples whose mixing quality information is correctly predicted to be abnormal quality to the number of samples whose actual historical mixing quality information is abnormal quality; Precision reflects the ratio of the number of samples whose mixing quality information is correctly predicted to be abnormal quality to the number of samples whose predicted mixing quality information is abnormal quality. mAP@0.5 (mean-average precision) refers to the average value of the area under the precision-recall curve under different historical mixing quality information conditions when the confidence level is 0.5; mAP@[0.5:0.95] refers to the average value of the area under the precision-recall curve under different confidence levels, starting from 0.5 and with a step size of 0.05 up to 0.95.

[0071] Furthermore, the loss function is gradually reduced during the model training process using the training sample set, and the model evaluation index of the model is steadily improved to convergence using the test sample set to complete the training of the model.

[0072] The method also includes: sending the stirring quality information of the slurry image to be detected to the display screen to prompt the user of the kneading solid content or viscosity of the slurry to be detected. Since the kneading solid content gradually decreases during the wetting process, it can be understood as a process in which the kneading solid content gradually decreases by gradually increasing the solvent, and the lithium battery slurry is gradually diluted. Furthermore, in the wetting process, it is first identified that the kneading solid content of the slurry image to be detected is high, and then when it is identified that the kneading solid content of the slurry image to be detected is normal, the wetting process can be terminated and the dispersion stage can be entered. In theory, it is rarely possible to identify a low kneading solid content during the wetting process. That is to say, each time the stirring quality information is identified through the image of the slurry to be detected during the wetting process, it can be sent to the display screen to prompt the user of the kneading solid content of the slurry to be detected. Then, when the kneading solid content of the slurry image to be detected is identified as normal, the stirring process of the lithium battery slurry is controlled to enter the dispersion stage, that is, when the kneading solid content is normal, the time of the dispersion stage is set to the preset dispersion stirring time. When the preset dispersion stirring time ends, it means that the dispersion stage ends, and then the slurry image to be detected is collected, and the stirring quality information is identified through the slurry image to be detected. In the viscosity adjustment stage, the viscosity changes from high to low, that is, due to the process of gradually adding a small amount of NMP to dilute, that is, gradually adding NMP to reduce the viscosity. Then, in the viscosity adjustment stage, the viscosity of the slurry image to be detected is first identified as high, and then when the viscosity of the slurry image to be detected is identified as normal, the viscosity adjustment stage can be ended, and then in the viscosity adjustment stage, it is theoretically rare to identify a low viscosity. That is to say, in the viscosity adjustment stage, each stirring quality information identified by the image of the slurry to be detected can be sent to the display screen to prompt the user of the viscosity of the slurry to be detected. Then, when the viscosity of the slurry image to be detected is identified as normal, the user determines that the stirring process of the lithium battery slurry is completed.

[0073] That is, after predicting the stirring quality information of the slurry image to be detected, the stirring quality information of the slurry image to be detected is directly sent to the display screen, so that the user can determine the stirring quality information of the slurry image to be detected through the display screen, thereby determining whether the stirrer needs to be stopped. Furthermore, after determining that the stirring process of the lithium battery slurry is completed, the user can stop the stirrer when the viscosity is low or the kneading solid content is low.

[0074] Based on the same application concept, the embodiments of the present application also provide a lithium battery slurry stirring quality identification device corresponding to the lithium battery slurry stirring quality identification method provided in the above embodiments. Since the principle of solving the problem by the device in the embodiments of the present application is similar to the lithium battery slurry stirring quality identification method in the above embodiments of the present application, the implementation of the device can refer to the implementation of the method, and the repeated parts will not be repeated.

[0075] like Figure 3 As shown, Figure 3 The functional module diagram of a device for identifying the stirring quality of lithium battery slurry provided in an embodiment of the present application is as follows: The device 10 for identifying the stirring quality of lithium battery slurry comprises: an acquisition module 101 , a determination module 102 and a prediction module 103 .

[0076] Among them, the acquisition module 101 is used to obtain the slurry image to be detected, and the slurry image to be detected is the current color image of the slurry to be detected collected during the stirring process of the lithium battery slurry; the determination module 102 is used to perform grayscale processing on the slurry image to be detected to obtain a target grayscale image; the prediction module 103 is used to input the target grayscale image into a trained machine learning model to predict the stirring quality information of the slurry image to be detected, and the stirring quality information is used to evaluate the kneading solid content or viscosity of the slurry to be detected.

[0077] The determination module is further used to convert the slurry image to be detected into an initial grayscale image according to the triplet of each pixel of the slurry image to be detected; and perform histogram equalization on the initial grayscale image to obtain a target grayscale image of the slurry image to be detected.

[0078] Based on the same application concept, see Figure 4 As shown, it is a schematic diagram of the structure of an electronic device provided in an embodiment of the present application, the electronic device 20 includes: a processor 201, a memory 202 and a bus 203, the memory 202 stores machine-readable instructions executable by the processor 201, when the electronic device 20 is running, the processor 201 and the memory 202 communicate through the bus 203, and the machine-readable instructions are executed by the processor 201 when running to perform the steps of the stirring quality identification method of lithium battery slurry as described in any of the above embodiments.

[0079] Specifically, when the machine-readable instructions are executed by the processor 201, the following processing can be performed: obtaining a slurry image to be detected, wherein the slurry image to be detected is a current color image of the slurry to be detected collected during the stirring process of the lithium battery slurry; performing grayscale processing on the slurry image to be detected to obtain a target grayscale image; inputting the target grayscale image into a trained machine learning model to predict the stirring quality information of the slurry image to be detected, and the stirring quality information is used to evaluate the kneading solid content or viscosity of the slurry to be detected.

[0080] Based on the same application concept, an embodiment of the present application further provides a computer-readable storage medium, on which a computer program is stored. When the computer program is executed by a processor, the steps of the stirring quality identification method for lithium battery slurry provided in the above embodiment are executed.

[0081] Specifically, the storage medium can be a general storage medium, such as a mobile disk, a hard disk, etc. When the computer program on the storage medium is run, the above-mentioned lithium battery slurry stirring quality identification method can be executed, by collecting the slurry image to be detected during the stirring process of the lithium battery slurry, gray-processing the slurry image to be detected, and then inputting a pre-trained machine learning model to predict the stirring quality information of the slurry to be detected through the machine learning model, thereby solving the technical problem of inaccuracy caused by relying on manual identification of stirring quality in the prior art, and solving the technical problem of low detection efficiency caused by relying on sampling detection, thereby achieving the technical effect of improving detection accuracy and efficiency.

[0082] Those skilled in the art can clearly understand that, for the convenience and simplicity of description, the specific working process of the system and device described above can refer to the corresponding process in the aforementioned method embodiment, and will not be repeated here. In the several embodiments provided in the present application, it should be understood that the disclosed system, device and method can be implemented in other ways. The device embodiments described above are merely schematic. For example, the division of the units is only a logical function division. There may be other division methods in actual implementation. For example, multiple units or components can be combined or integrated into another system, or some features can be ignored or not executed. Another point is that the mutual coupling or direct coupling or communication connection shown or discussed can be through some communication interfaces, indirect coupling or communication connection of devices or units, which can be electrical, mechanical or other forms.

[0083] The units described as separate components may or may not be physically separated, and the components shown as units may or may not be physical units, that is, they may be located in one place or distributed on multiple network units. Some or all of the units may be selected according to actual needs to achieve the purpose of the solution of this embodiment.

[0084] In addition, each functional unit in each embodiment of the present application may be integrated into one processing unit, or each unit may exist physically separately, or two or more units may be integrated into one unit.

[0085] If the function is implemented in the form of a software functional unit and sold or used as an independent product, it can be stored in a non-volatile computer-readable storage medium that is executable by a processor. Based on this understanding, the technical solution of the present application, or the part that contributes to the prior art or the part of the technical solution, can be embodied in the form of a software product, and the computer software product is stored in a storage medium, including a number of instructions to enable a computer device (which can be a personal computer, a server, or a network device, etc.) to perform all or part of the steps of the method described in each embodiment of the present application. The aforementioned storage medium includes: U disk, mobile hard disk, read-only memory (ROM), random access memory (RAM), disk or optical disk, etc., various media that can store program codes.

[0086] The above are only specific implementations of the present application, but the protection scope of the present application is not limited thereto. Any technician familiar with the technical field can easily think of changes or substitutions within the technical scope disclosed in the present application, which should be included in the protection scope of the present application. Therefore, the protection scope of the present application should be based on the protection scope of the claims.

Claims

1. A method for identifying stirring quality of lithium battery slurry, characterized in that: The method comprises: Acquire a slurry image to be detected, wherein the slurry image to be detected is a current color image of the slurry to be detected collected during the stirring process of the lithium battery slurry, wherein the current color image includes: a color image corresponding to a wetting stage for mixing powder and liquid of raw materials of the lithium battery slurry, or a color image corresponding to a viscosity adjustment stage for adjusting the viscosity of the lithium battery slurry, wherein the wetting stage is located before the viscosity adjustment stage during the stirring process; Performing grayscale processing on the slurry image to be detected to obtain a target grayscale image; The target grayscale image is input into the trained machine learning model to predict the stirring quality information of the slurry image to be tested, and the stirring quality information is used to evaluate the kneading solid content or viscosity of the slurry to be tested, wherein the stirring quality information obtained from the color image corresponding to the wetting stage is the kneading solid content of the slurry to be tested, and the stirring quality information obtained from the color image corresponding to the viscosity adjustment stage is the viscosity of the slurry to be tested.

2. The method according to claim 1, characterized in that The grayscale processing of the slurry image to be detected to obtain a target grayscale image includes: According to the triplet of each pixel of the slurry image to be detected, converting the slurry image to be detected into an initial grayscale image; The initial grayscale image is subjected to histogram equalization to obtain a target grayscale image of the slurry image to be detected.

3. The method according to claim 2, characterized in that The triplet includes a first component, a second component and a third component, According to the triplet of each pixel of the slurry image to be detected, the slurry image to be detected is converted into an initial grayscale image, comprising: For each pixel of the slurry image to be detected, a first component in the triplet of the pixel is multiplied by the corresponding first coefficient to obtain a first product, a second component in the triplet of the pixel is multiplied by the corresponding second coefficient to obtain a second product, and a third component in the triplet of the pixel is multiplied by the corresponding third coefficient to obtain a third product, and a sum of the first product, the second product and the third product is used as the gray value of the pixel to obtain the initial gray image.

4. The method according to claim 1, characterized in that: The image of the slurry to be tested is obtained by: Collecting a lithium battery slurry image by an image acquisition device according to a preset acquisition cycle, wherein the preset acquisition cycle is used to limit the position of a stirring blade in the lithium battery slurry image; The preset area of ​​the lithium battery slurry picture is used as the slurry image to be detected, so as to remove the part other than the lithium battery slurry in the lithium battery slurry picture.

5. The method according to claim 1, characterized in that The machine learning model is trained by: Acquire multiple historical slurry images collected during the stirring process of lithium battery slurry; Performing grayscale processing on each historical slurry image respectively to obtain a historical grayscale image corresponding to each historical slurry image; Determine historical stirring quality information corresponding to each historical grayscale image, where the historical stirring quality information is used to describe the stirring quality of the lithium battery slurry in each historical grayscale image; Each historical grayscale image is used as sample data, and the historical mixing quality information corresponding to each historical slurry image is used as a label to train the machine learning model.

6. The method according to claim 1, characterized in that The method further comprises: The stirring quality information of the slurry image to be detected is sent to a display screen to prompt a user of the kneading solid content or viscosity of the slurry to be detected.

7. A device for identifying stirring quality of lithium battery slurry, characterized in that: The device comprises: an acquisition module, for acquiring an image of a slurry to be detected, wherein the image of the slurry to be detected is a current color image of the slurry to be detected collected during the stirring process of the lithium battery slurry, wherein the current color image includes: a color image corresponding to a wetting stage for mixing powder and liquid of raw materials of the lithium battery slurry, or a color image corresponding to a viscosity adjustment stage for adjusting the viscosity of the lithium battery slurry, wherein the wetting stage is located before the viscosity adjustment stage during the stirring process; A determination module, used for performing grayscale processing on the slurry image to be detected to obtain a target grayscale image; A prediction module is used to input the target grayscale image into a trained machine learning model to predict the stirring quality information of the slurry image to be detected, and the stirring quality information is used to evaluate the kneading solid content or viscosity of the slurry to be detected, wherein the stirring quality information obtained from the color image corresponding to the wetting stage is the kneading solid content of the slurry to be detected, and the stirring quality information obtained from the color image corresponding to the viscosity adjustment stage is the viscosity of the slurry to be detected.

8. The device according to claim 7, characterized in that The determination module is further used to convert the slurry image to be detected into an initial grayscale image according to the triplet of each pixel of the slurry image to be detected; and perform histogram equalization on the initial grayscale image to obtain a target grayscale image of the slurry image to be detected.

9. An electronic device, characterized in that: include: A processor, a memory and a bus, wherein the memory stores machine-readable instructions executable by the processor, and when the electronic device is running, the processor and the memory communicate through the bus, and the machine-readable instructions are executed by the processor to execute the steps of the method for identifying the stirring quality of lithium battery slurry as described in any one of claims 1 to 6.

10. A computer-readable storage medium, characterized in that: The computer-readable storage medium stores a computer program, and when the computer program is executed by a processor, the steps of the method for identifying the stirring quality of lithium battery slurry according to any one of claims 1 to 6 are executed.

Images