CN116539618A - Deep learning-based lithium ion battery active material damage degree testing method - Google Patents
Deep learning-based lithium ion battery active material damage degree testing method Download PDFInfo
- Publication number
- CN116539618A CN116539618A CN202310405813.8A CN202310405813A CN116539618A CN 116539618 A CN116539618 A CN 116539618A CN 202310405813 A CN202310405813 A CN 202310405813A CN 116539618 A CN116539618 A CN 116539618A
- Authority
- CN
- China
- Prior art keywords
- lithium ion
- ion battery
- active material
- pole piece
- damage degree
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Pending
Links
- 239000011149 active material Substances 0.000 title claims abstract description 76
- HBBGRARXTFLTSG-UHFFFAOYSA-N Lithium ion Chemical compound [Li+] HBBGRARXTFLTSG-UHFFFAOYSA-N 0.000 title claims abstract description 71
- 229910001416 lithium ion Inorganic materials 0.000 title claims abstract description 71
- 238000012360 testing method Methods 0.000 title claims abstract description 46
- 238000013135 deep learning Methods 0.000 title claims abstract description 15
- 230000003287 optical effect Effects 0.000 claims abstract description 36
- 238000000034 method Methods 0.000 claims abstract description 34
- 238000011065 in-situ storage Methods 0.000 claims abstract description 31
- 239000002245 particle Substances 0.000 claims abstract description 26
- 238000012549 training Methods 0.000 claims abstract description 22
- 239000000463 material Substances 0.000 claims abstract description 16
- 238000002372 labelling Methods 0.000 claims abstract description 7
- 238000013136 deep learning model Methods 0.000 claims abstract description 6
- 238000005498 polishing Methods 0.000 claims description 27
- 238000006243 chemical reaction Methods 0.000 claims description 19
- XKRFYHLGVUSROY-UHFFFAOYSA-N Argon Chemical compound [Ar] XKRFYHLGVUSROY-UHFFFAOYSA-N 0.000 claims description 18
- 238000007599 discharging Methods 0.000 claims description 13
- 238000013527 convolutional neural network Methods 0.000 claims description 11
- 125000004122 cyclic group Chemical group 0.000 claims description 10
- 238000010884 ion-beam technique Methods 0.000 claims description 10
- 229910052786 argon Inorganic materials 0.000 claims description 9
- 238000013528 artificial neural network Methods 0.000 claims description 7
- 238000000227 grinding Methods 0.000 claims description 7
- XUIMIQQOPSSXEZ-UHFFFAOYSA-N Silicon Chemical compound [Si] XUIMIQQOPSSXEZ-UHFFFAOYSA-N 0.000 claims description 6
- 238000000879 optical micrograph Methods 0.000 claims description 6
- 229910052710 silicon Inorganic materials 0.000 claims description 6
- 239000010703 silicon Substances 0.000 claims description 6
- OKTJSMMVPCPJKN-UHFFFAOYSA-N Carbon Chemical compound [C] OKTJSMMVPCPJKN-UHFFFAOYSA-N 0.000 claims description 5
- 229910002804 graphite Inorganic materials 0.000 claims description 5
- 239000010439 graphite Substances 0.000 claims description 5
- 238000012544 monitoring process Methods 0.000 claims description 5
- WHXSMMKQMYFTQS-UHFFFAOYSA-N Lithium Chemical compound [Li] WHXSMMKQMYFTQS-UHFFFAOYSA-N 0.000 claims description 4
- 229910052744 lithium Inorganic materials 0.000 claims description 4
- 229910012851 LiCoO 2 Inorganic materials 0.000 claims description 2
- 229910010707 LiFePO 4 Inorganic materials 0.000 claims description 2
- 229910013716 LiNi Inorganic materials 0.000 claims description 2
- RTAQQCXQSZGOHL-UHFFFAOYSA-N Titanium Chemical compound [Ti] RTAQQCXQSZGOHL-UHFFFAOYSA-N 0.000 claims description 2
- 238000003475 lamination Methods 0.000 claims description 2
- GELKBWJHTRAYNV-UHFFFAOYSA-K lithium iron phosphate Chemical compound [Li+].[Fe+2].[O-]P([O-])([O-])=O GELKBWJHTRAYNV-UHFFFAOYSA-K 0.000 claims description 2
- 238000007781 pre-processing Methods 0.000 claims description 2
- 238000007619 statistical method Methods 0.000 abstract description 3
- 238000002474 experimental method Methods 0.000 abstract description 2
- 238000011160 research Methods 0.000 abstract description 2
- 238000012512 characterization method Methods 0.000 description 7
- VYPSYNLAJGMNEJ-UHFFFAOYSA-N Silicium dioxide Chemical compound O=[Si]=O VYPSYNLAJGMNEJ-UHFFFAOYSA-N 0.000 description 6
- 239000004696 Poly ether ether ketone Substances 0.000 description 5
- JUPQTSLXMOCDHR-UHFFFAOYSA-N benzene-1,4-diol;bis(4-fluorophenyl)methanone Chemical compound OC1=CC=C(O)C=C1.C1=CC(F)=CC=C1C(=O)C1=CC=C(F)C=C1 JUPQTSLXMOCDHR-UHFFFAOYSA-N 0.000 description 5
- 229920002530 polyetherether ketone Polymers 0.000 description 5
- 238000007789 sealing Methods 0.000 description 5
- 235000012239 silicon dioxide Nutrition 0.000 description 4
- 235000012431 wafers Nutrition 0.000 description 4
- 238000005516 engineering process Methods 0.000 description 3
- 239000004744 fabric Substances 0.000 description 3
- 239000005357 flat glass Substances 0.000 description 3
- 239000010453 quartz Substances 0.000 description 3
- 229910001220 stainless steel Inorganic materials 0.000 description 3
- 239000010935 stainless steel Substances 0.000 description 3
- 238000005520 cutting process Methods 0.000 description 2
- 238000013461 design Methods 0.000 description 2
- 238000010586 diagram Methods 0.000 description 2
- 229910003460 diamond Inorganic materials 0.000 description 2
- 239000010432 diamond Substances 0.000 description 2
- 238000009792 diffusion process Methods 0.000 description 2
- 230000011218 segmentation Effects 0.000 description 2
- 239000002002 slurry Substances 0.000 description 2
- 238000010998 test method Methods 0.000 description 2
- HPTJABJPZMULFH-UHFFFAOYSA-N 12-[(Cyclohexylcarbamoyl)amino]dodecanoic acid Chemical compound OC(=O)CCCCCCCCCCCNC(=O)NC1CCCCC1 HPTJABJPZMULFH-UHFFFAOYSA-N 0.000 description 1
- 102100025471 Epiphycan Human genes 0.000 description 1
- 101001056751 Homo sapiens Epiphycan Proteins 0.000 description 1
- 239000002033 PVDF binder Substances 0.000 description 1
- 244000137852 Petrea volubilis Species 0.000 description 1
- 230000001133 acceleration Effects 0.000 description 1
- XAGFODPZIPBFFR-UHFFFAOYSA-N aluminium Chemical compound [Al] XAGFODPZIPBFFR-UHFFFAOYSA-N 0.000 description 1
- 229910052782 aluminium Inorganic materials 0.000 description 1
- 238000004458 analytical method Methods 0.000 description 1
- QVGXLLKOCUKJST-UHFFFAOYSA-N atomic oxygen Chemical compound [O] QVGXLLKOCUKJST-UHFFFAOYSA-N 0.000 description 1
- 230000009286 beneficial effect Effects 0.000 description 1
- 239000003795 chemical substances by application Substances 0.000 description 1
- 238000004146 energy storage Methods 0.000 description 1
- 238000011156 evaluation Methods 0.000 description 1
- 239000011888 foil Substances 0.000 description 1
- 239000003292 glue Substances 0.000 description 1
- 150000002500 ions Chemical class 0.000 description 1
- 230000007774 longterm Effects 0.000 description 1
- 238000002156 mixing Methods 0.000 description 1
- 239000000203 mixture Substances 0.000 description 1
- 238000000399 optical microscopy Methods 0.000 description 1
- 239000001301 oxygen Substances 0.000 description 1
- 229910052760 oxygen Inorganic materials 0.000 description 1
- 238000011192 particle characterization Methods 0.000 description 1
- 238000007517 polishing process Methods 0.000 description 1
- 229920002981 polyvinylidene fluoride Polymers 0.000 description 1
- 238000003825 pressing Methods 0.000 description 1
- 238000012545 processing Methods 0.000 description 1
- 238000012113 quantitative test Methods 0.000 description 1
- 239000011347 resin Substances 0.000 description 1
- 229920005989 resin Polymers 0.000 description 1
- 238000000926 separation method Methods 0.000 description 1
- 239000000377 silicon dioxide Substances 0.000 description 1
- 239000002904 solvent Substances 0.000 description 1
- 238000003756 stirring Methods 0.000 description 1
- 238000006467 substitution reaction Methods 0.000 description 1
- -1 superP Substances 0.000 description 1
- 230000002123 temporal effect Effects 0.000 description 1
- 238000001291 vacuum drying Methods 0.000 description 1
- 230000000007 visual effect Effects 0.000 description 1
- XLYOFNOQVPJJNP-UHFFFAOYSA-N water Substances O XLYOFNOQVPJJNP-UHFFFAOYSA-N 0.000 description 1
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N3/00—Computing arrangements based on biological models
- G06N3/02—Neural networks
- G06N3/04—Architecture, e.g. interconnection topology
- G06N3/0464—Convolutional networks [CNN, ConvNet]
-
- G—PHYSICS
- G01—MEASURING; TESTING
- G01N—INVESTIGATING OR ANALYSING MATERIALS BY DETERMINING THEIR CHEMICAL OR PHYSICAL PROPERTIES
- G01N21/00—Investigating or analysing materials by the use of optical means, i.e. using sub-millimetre waves, infrared, visible or ultraviolet light
- G01N21/84—Systems specially adapted for particular applications
- G01N21/88—Investigating the presence of flaws or contamination
- G01N21/8851—Scan or image signal processing specially adapted therefor, e.g. for scan signal adjustment, for detecting different kinds of defects, for compensating for structures, markings, edges
-
- G—PHYSICS
- G01—MEASURING; TESTING
- G01N—INVESTIGATING OR ANALYSING MATERIALS BY DETERMINING THEIR CHEMICAL OR PHYSICAL PROPERTIES
- G01N21/00—Investigating or analysing materials by the use of optical means, i.e. using sub-millimetre waves, infrared, visible or ultraviolet light
- G01N21/84—Systems specially adapted for particular applications
- G01N21/88—Investigating the presence of flaws or contamination
- G01N21/95—Investigating the presence of flaws or contamination characterised by the material or shape of the object to be examined
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N3/00—Computing arrangements based on biological models
- G06N3/02—Neural networks
- G06N3/08—Learning methods
-
- G—PHYSICS
- G01—MEASURING; TESTING
- G01N—INVESTIGATING OR ANALYSING MATERIALS BY DETERMINING THEIR CHEMICAL OR PHYSICAL PROPERTIES
- G01N21/00—Investigating or analysing materials by the use of optical means, i.e. using sub-millimetre waves, infrared, visible or ultraviolet light
- G01N21/84—Systems specially adapted for particular applications
- G01N21/88—Investigating the presence of flaws or contamination
- G01N21/8851—Scan or image signal processing specially adapted therefor, e.g. for scan signal adjustment, for detecting different kinds of defects, for compensating for structures, markings, edges
- G01N2021/8854—Grading and classifying of flaws
- G01N2021/8861—Determining coordinates of flaws
-
- G—PHYSICS
- G01—MEASURING; TESTING
- G01N—INVESTIGATING OR ANALYSING MATERIALS BY DETERMINING THEIR CHEMICAL OR PHYSICAL PROPERTIES
- G01N21/00—Investigating or analysing materials by the use of optical means, i.e. using sub-millimetre waves, infrared, visible or ultraviolet light
- G01N21/84—Systems specially adapted for particular applications
- G01N21/88—Investigating the presence of flaws or contamination
- G01N21/8851—Scan or image signal processing specially adapted therefor, e.g. for scan signal adjustment, for detecting different kinds of defects, for compensating for structures, markings, edges
- G01N2021/8854—Grading and classifying of flaws
- G01N2021/8861—Determining coordinates of flaws
- G01N2021/8864—Mapping zones of defects
-
- G—PHYSICS
- G01—MEASURING; TESTING
- G01N—INVESTIGATING OR ANALYSING MATERIALS BY DETERMINING THEIR CHEMICAL OR PHYSICAL PROPERTIES
- G01N21/00—Investigating or analysing materials by the use of optical means, i.e. using sub-millimetre waves, infrared, visible or ultraviolet light
- G01N21/84—Systems specially adapted for particular applications
- G01N21/88—Investigating the presence of flaws or contamination
- G01N21/8851—Scan or image signal processing specially adapted therefor, e.g. for scan signal adjustment, for detecting different kinds of defects, for compensating for structures, markings, edges
- G01N2021/8887—Scan or image signal processing specially adapted therefor, e.g. for scan signal adjustment, for detecting different kinds of defects, for compensating for structures, markings, edges based on image processing techniques
-
- Y—GENERAL TAGGING OF NEW TECHNOLOGICAL DEVELOPMENTS; GENERAL TAGGING OF CROSS-SECTIONAL TECHNOLOGIES SPANNING OVER SEVERAL SECTIONS OF THE IPC; TECHNICAL SUBJECTS COVERED BY FORMER USPC CROSS-REFERENCE ART COLLECTIONS [XRACs] AND DIGESTS
- Y02—TECHNOLOGIES OR APPLICATIONS FOR MITIGATION OR ADAPTATION AGAINST CLIMATE CHANGE
- Y02E—REDUCTION OF GREENHOUSE GAS [GHG] EMISSIONS, RELATED TO ENERGY GENERATION, TRANSMISSION OR DISTRIBUTION
- Y02E60/00—Enabling technologies; Technologies with a potential or indirect contribution to GHG emissions mitigation
- Y02E60/10—Energy storage using batteries
Landscapes
- Engineering & Computer Science (AREA)
- Physics & Mathematics (AREA)
- General Health & Medical Sciences (AREA)
- Health & Medical Sciences (AREA)
- Life Sciences & Earth Sciences (AREA)
- Theoretical Computer Science (AREA)
- General Physics & Mathematics (AREA)
- Computing Systems (AREA)
- Software Systems (AREA)
- Evolutionary Computation (AREA)
- Computational Linguistics (AREA)
- Molecular Biology (AREA)
- Biophysics (AREA)
- General Engineering & Computer Science (AREA)
- Biomedical Technology (AREA)
- Mathematical Physics (AREA)
- Data Mining & Analysis (AREA)
- Artificial Intelligence (AREA)
- Pathology (AREA)
- Immunology (AREA)
- Chemical & Material Sciences (AREA)
- Analytical Chemistry (AREA)
- Biochemistry (AREA)
- Signal Processing (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Battery Electrode And Active Subsutance (AREA)
Abstract
The invention relates to a lithium ion battery active material damage degree testing method based on deep learning. Comprising the following steps: preparing a sample of the section of the pole piece of the lithium ion battery; acquiring a cross-sectional image by an optical microscope as an original data set; distinguishing and labeling the original data set image as a training data set of the deep learning model; building and training a deep learning model; constructing an in-situ optical microscope observation system of the lithium ion battery active material, and carrying out in-situ observation experiments; inputting an in-situ observation image result into the deep learning model to finish intelligent identification of particles of the lithium ion battery active material to be observed; and judging the gray value of the image by adopting a threshold setting method, thereby realizing quantitative statistical analysis of the damage degree of the lithium ion battery material. The invention can carry out automatic and intelligent quantitative statistical analysis on the damage degree of the active material of the same pole piece in different charge and discharge stages, and can be widely applied to the research on the evolution mechanism of the damage degree of the lithium ion battery material.
Description
Technical Field
The invention belongs to the technical field of lithium ion battery material characterization, and relates to a lithium ion battery active material damage degree testing method based on deep learning.
Background
Lithium ion batteries have been the technology of choice for a range of portable energy storage device solutions over the past few decades due to their high energy density and capacity. In order to meet the higher requirements of the green power grid, new energy automobiles and other emerging application scenes on cycle life, power and energy density, advanced lithium ion battery technology still needs to be continuously developed. Lithium ion battery active materials are important factors in determining the performance of lithium ion batteries. During charge and discharge, lithium ions are intercalated and deintercalated in the active material, and simultaneously the volume of the active material expands and contracts. Diffusion induced stresses are created during diffusion of lithium ions within the active material. Along with long-term cyclic charge and discharge, the lithium ion battery material can generate cracks with different degrees so as to influence the performance of the lithium ion battery. Therefore, the exploration of the crack evolution mechanism of the lithium ion battery material has important guiding significance for optimizing the design of the lithium ion battery material.
Typically, researchers observe cross-sectional cracks in particles using SEM to characterize damage to lithium ion battery materials. However, this has problems in terms of both spatial and temporal dimensions. On a spatial scale, researchers typically choose several particles as characterization results. In practical situations, the active material contained in the battery pole piece has non-uniformity, namely, a researcher can obtain a characterization result of good active material structure and a characterization result of serious active material cracks in the same sample by selecting a specific observation area. Therefore, the method for testing the damage degree of the lithium ion battery active material has statistical significance, the overall damage condition is tested in as many particle characterization results as possible, and the method for manually identifying and counting cannot be used for measuring the damage of hundreds of active materials. In terms of time scale, researchers usually adopt a method for characterizing the damage condition of an active material by disassembling the battery off-site characterization, can only obtain a characterization image at a single moment, and cannot test the evolution condition of the damage degree of the active material under different charge and discharge cycle numbers. Therefore, the test of the damage degree of the lithium ion battery active material has time continuity, and the real-time observation method is required to observe, record and analyze the fixed area of the same pole piece at different moments.
Disclosure of Invention
Aiming at the problems of the background technology, the invention provides a lithium ion battery active material damage degree statistical method based on deep learning.
A lithium ion battery active material damage degree testing method based on deep learning comprises the following steps:
(1) And (3) constructing a target test active material intelligent identification tool:
s1, grinding and polishing the section of a lithium ion battery pole piece by adopting a polishing means to obtain a flat and smooth sample of the section of the battery pole piece;
s2, acquiring a section image of the lithium ion battery pole piece by using an optical microscope as an original data set;
s3, distinguishing and labeling active materials of the lithium ion battery, which are subjected to target test, in the pole piece image of the original data set, drawing the active materials in the section of the pole piece of the lithium ion battery by using a polygon by using a professional labeling tool LabelMe, and storing coordinate data of each vertex on a pixel scale so as to obtain a training data set;
s4, constructing a deep learning model Mask R-CNN, selecting a training data set, preprocessing a training data set image, inputting the training data set image into a neural network for training, calculating a loss result after gradient back propagation, continuously inputting the training data into the neural network when a specified period is not reached, and storing neural network parameters as a target test active material intelligent recognition tool when the specified period is reached;
(2) Real-time monitoring of lithium ion battery active material damage based on in-situ optical microscope observation platform:
s5, constructing an in-situ optical microscope observation platform of the lithium ion battery active material, and monitoring the sectional image of the lithium ion battery pole piece in real time in the cyclic charge and discharge process;
s6, based on an intelligent identification tool for target test active materials, intelligent identification of target test lithium ion battery active materials in battery pole piece optical microscope images recorded in real time in the cyclic charge and discharge process is completed, and mask images and coordinate information of each active material in the images are obtained;
(3) And testing and evaluating the damage degree of the lithium ion battery active material:
s7, judging the image gray value of each active material obtained by the test in the step 2 by adopting a method of setting a threshold value, dividing the particle surface into a crack area and a block area, wherein the crack degree of each active material = crack area/particle area, and testing and analyzing the crack degree of each active material to obtain the overall damage degree of active particles distributed by the pole piece to be observed.
In some embodiments, where the lithium ion battery active material is an active material particle in a battery pole piece, it may be lithium iron phosphate LiFePO 4 Lithium cobaltate LiCoO 2 Ternary material LiNi of NCM x Co y Mn 1-x-y O 2 Lithium titanate Li 4 Ti 5 O 12 Graphite C and silicon Si.
In some embodiments, the polishing method for preparing the lithium-ion battery pole-piece cross-section sample in step S1 may use mechanical polishing, argon ion beam polishing, or focused ion beam polishing.
In some embodiments, the in-situ optical microscope observation system in step S5 is composed of an optical microscope, a lithium ion battery charge-discharge device, an optical in-situ reaction tank, and a computer; the pole piece to be observed, the counter electrode and the diaphragm in the optical in-situ reaction tank are clamped in a lamination mode, the cross sections of one sides of the pole piece to be observed, the counter electrode and the diaphragm facing the window are guaranteed to be level, and the sides are tightly attached to the window of the reaction tank; the electrode to be observed is polished by an argon ion beam to obtain a smooth section, and the section is used as a section which is clung to a window of the reaction tank.
The cyclic charge and discharge process in the step S5 is as follows: and setting a charging and discharging process step of multiplying power charging, standing, multiplying power discharging, standing and circulating for the optical in-situ reaction tank, wherein the voltage interval is 2.8-4.7V, and the charging and discharging multiplying power is set to be 0.5C.
The invention has the beneficial effects that the invention discloses a method for testing the damage degree of the lithium ion battery active material based on deep learning, which is improved compared with the prior art in that:
the invention designs a lithium ion battery active material damage degree testing method based on deep learning, which is based on a Mask R-CNN image example segmentation model, acquires a lithium ion battery material pole piece image through an in-situ optical microscope method in the cyclic charge and discharge process of a lithium ion battery, predicts active particles in the image by combining the Mask R-CNN model to realize rapid and intelligent separation, and simultaneously realizes the test analysis of the damage degree by using a method for judging gray values by setting a threshold value. The method can carry out automatic and intelligent quantitative tests on the damage degree of the active material of the same pole piece in different charge and discharge stages.
Drawings
FIG. 1 is a flow chart of a method for testing the damage degree of a lithium ion active battery material based on deep learning;
FIG. 2a is an original view of an optical microscope image of a cross section polished by NCM811 pole piece obtained by disassembly of a soft pack battery of an embodiment of the present application;
FIG. 2b is a plot of LabelMe labeling results based on optical microscopy images;
FIG. 2c is a mask image corresponding to a annotation image;
FIG. 3a is a schematic view of an in situ optical microscope observation platform used in the process of testing the damage degree of the active material of the lithium ion battery;
FIG. 3b is a schematic illustration of an optical in situ reaction cell;
fig. 4 is a diagram of a result of intelligent identification of an active material by using a target test active material intelligent identification tool based on a Mask R-CNN model for a pole piece to be observed in an embodiment of the present application.
Fig. 5a is a schematic diagram of an active particle damage condition obtained based on a gray value threshold after performing particle intelligent recognition on a pole piece to be observed in the embodiment of the application;
fig. 5b is a graph showing the results of the test of the damage degree of the fresh pole piece and the pole piece after the cycle at the cut-off voltage of 4.7V.
Detailed Description
The invention is further illustrated by the following drawings and examples of embodiments, which are not intended to limit the scope of the invention.
Implementation example: a lithium ion battery active material damage degree testing method based on deep learning specifically comprises the following steps:
(1) And (3) constructing a target test active material intelligent identification tool:
preparing a lithium ion battery pole piece sample: commercial NCM811/Graphite soft pack cells with a capacity of 4Ah were selected and disassembled in an argon filled glove box (water oxygen values were all below 0.1 PPM). The obtained NCM811 pole piece is soaked in DMC solvent for 1h, the NCM811 pole piece is taken out and dried in a vacuum drying oven at 80 ℃ for 12h, and then the NCM811 pole piece is taken out.
Pre-polishing a sample of the section of the lithium ion battery pole piece: the NCM811 pole piece was cut into 10mm x 10mm square pieces. Two pieces of 10 mm-10 mm silicon wafers with the thickness of 1mm are selected, and 10 pieces of NCM811 square piece and the other piece of silicon wafer are sequentially bonded on one piece of silicon wafer by using 502 glue. Placing the sample block in a metallographic specimen cold mosaic mold, adopting a Stahl 40200029Epofix series cold mosaic material, mixing and stirring resin and a curing agent in a mass ratio of 25:3 for 2min, pouring the mixture into the mold, and taking out the cold mosaic block from the mold after the cold mosaic block is completely cured after 12 h.
Grinding and polishing the cold mosaic blocks by adopting a Sitel Tegramin-20 automatic grinding and polishing machine, and respectively grinding the surfaces of the samples by adopting 400-mesh, 800-mesh, 1200-mesh, 2000-mesh and 4000-mesh sand paper for 1min, 1.5min, 2min, 2.5min and 3min; then respectively adopting a diamond polishing solution of 3 mu m and a diamond polishing solution of 1 mu m and a corresponding Dac polishing cloth and Nap polishing cloth to perform rough polishing on the surface of the sample for 5 minutes; finally, polishing the sample surface by using 0.05 mu m silicon dioxide polishing solution and Chem polishing cloth to obtain a flat sample surface; the pressure of the cold insert in all polishing processes is 5N, and the rotating speed of the cold insert and the rotating speed of the turntable are both 50r/min and rotate in the same direction;
and grinding the cold insert materials except for the clamping part of the silicon wafer after final polishing to obtain a sample block with the size of about 10 mm-3 mm.
Polishing a sample of the section of the lithium ion battery pole piece: and (3) polishing the pre-polished surface of the sample block by using a table ion grinding and polishing instrument SEMPrep2 by using an argon ion beam, so that a flat and smooth boundary section of a large number of NCM811 pole piece samples can be obtained by one-time argon ion beam cutting.
Acquisition of the original dataset: NCM811 cross-sectional images were recorded using an EC Epiplan 100x/0.85M27 objective lens of a Zeiss Axio Imager 1 metallographic microscope, as shown in FIG. 2 a.
Acquisition of training data sets: NCM811 particles in the original dataset image were labeled using the LabelMe labeling tool. Individual particles were drawn by visual recognition using polygons and named "particle_001", "particle_002", "particle_003" … … as shown in fig. 2 b. And (3) saving the marked image by json, and then generating a corresponding Mask image, as shown in fig. 2c, and entering a Mask R-CNN model for subsequent training and prediction.
Building and training an example segmentation deep learning model: the Mask R-CNN model was used to intelligently identify NCM811 active particles. The experimental environment for model building is computer configuration: CPU: AMD EPYC 7543 32-Core Processor@3395.407MHz, GPU: NVIDIA A40 (48G), memory: 90G. The depth requirement framework uses Tensorflow 1.13.1 and Keras 2.2.5. The parallel computing framework adopts CUDA 11.0, and the deep neural network acceleration library is cuDNN 8.0.5.
Inputting the marked training data set into a Mask R-CNN model to start training, setting the batch processing size as 1, setting the iteration cycle number as 100 and setting the learning rate as 0.001 in the training process of the Mask R-CNN model. And after training is completed, obtaining available Mask R-CNN model parameters as intelligent identification tools for target test active materials.
(2) Real-time monitoring of lithium ion battery active material damage based on in-situ optical microscope observation platform:
preparing a pole piece to be observed: the NCM811 active material, superP, PVDF and NMP solutions were mixed into NCM811 slurry, and the slurry was coated with aluminum foil current collector to make NCM811 pole pieces, which were subjected to argon ion beam cutting to obtain a flat, smooth cross section to be observed.
Assembling an optical in-situ reaction tank: the optical in-situ reaction tank comprises a PEEK shell, a PEEK upper cover, a quartz glass window, a sealing flat gasket, a sealing ring, a stainless steel conductive column, a battery material clamping block and a fastening screw. And the PEEK upper cover is fixedly connected with the PEEK shell by using a fastening screw, and the sealing is realized by pressing quartz window glass, a sealing ring and a sealing flat gasket which are arranged on the PEEK shell. Placing the prepared NCM811 pole piece to be observed, the PP diaphragm and the graphite pole piece of the counter electrode between two battery material clamping blocks, clamping and fixing, and attaching the clamping blocks to quartz window glass, wherein one side of the cross section of the pole piece to be observed polished by an argon ion beam is attached to the quartz window glass; the conductive path of the optical in-situ reaction cell was formed by conducting the NCM811 anode and the graphite cathode using stainless steel conductive posts, as shown in fig. 3 b.
Building an in-situ optical microscope observation system: placing the optical in-situ reaction tank on an objective table of a zeiss Axio Imager 1 metallographic microscope, and focusing on an NCM811 pole piece to be observed by using an EC Epiplan 100x/0.85M27 objective lens; connecting positive and negative wires of the new Wei charge-discharge equipment with positive and negative stainless steel conductive posts of the optical in-situ reaction tank respectively to prepare for carrying out charge-discharge experiments on the optical in-situ reaction tank; the computer is respectively connected with the optical microscope and the charging and discharging equipment to realize the control of the optical microscope, the control of the charging and discharging equipment, the recording of the image result of the optical microscope and the recording of the charging and discharging result of the optical in-situ reaction tank, as shown in figure 3 a; the light microscope is adjusted to dark field mode and a clear light microscope image is obtained using adjustment of the appropriate exposure time and cool-warm color values.
In situ optical microscopy observation test: setting a charging and discharging process step of multiplying power charging, standing, multiplying power discharging, standing and circulating for the optical in-situ reaction tank, wherein the voltage interval is 2.8-4.7V, and the charging and discharging multiplying power is set to be 0.5C; and (3) respectively recording optical microscope images of the section of the pole piece to be observed in an initial state and every 10 circles, and carrying out intelligent identification of the active materials in the pole piece on the pole piece images obtained by the in-situ observation test by combining with the intelligent identification tool of the target test active materials to obtain mask images and coordinate information of each active material, as shown in fig. 4.
(3) And testing and evaluating the damage degree of the lithium ion battery active material:
for each NCM811 particle, the crack is generally black, the block is generally light white, a threshold value is set to judge the gray value of each particle, the particle area is divided into a crack area and a block area, as shown in fig. 5a, the damage degree of a single active particle = crack area/particle area, and the damage degree of each NCM811 particle is tested and evaluated to obtain the overall damage degree condition of the active particles distributed by the pole piece to be observed; the damage degree evolution condition in the cyclic charge-discharge process can be obtained by comparing and analyzing the crack degree conditions of the active particles with different cycle numbers, as shown in fig. 5 b.
So far, a target test active material intelligent recognition tool is constructed by a Mask R-CNN model based on the image training of the cross section optical microscope of the lithium ion battery material, and the automatic intelligent recognition and test evaluation of the damage degree of a large number of lithium ion battery active material particles under a large spatial scale are realized; meanwhile, an in-situ optical microscope observation test method is combined, and a test means of the damage degree evolution condition of the lithium ion battery active material is further expanded from a time scale, so that the damage degree evolution of the lithium ion battery active material in the cyclic charge and discharge process is obtained. The method provides a reliable and advanced characterization test method for the research of the evolution mechanism of the damage degree of the lithium ion battery active material.
The foregoing is merely a preferred embodiment of the present application, but the scope of the present application is not limited thereto, and any changes or substitutions easily conceivable by those skilled in the art within the technical scope of the present application should be covered in the scope of the present application. Therefore, the protection scope of the present application shall be subject to the protection scope of the claims.
Claims (5)
1. A method for testing the damage degree of an active material of a lithium ion battery based on deep learning is characterized by comprising the following steps,
(1) And (3) constructing a target test active material intelligent identification tool:
s1, grinding and polishing the section of a lithium ion battery pole piece by adopting a polishing means to obtain a flat and smooth sample of the section of the battery pole piece;
s2, acquiring a section image of the lithium ion battery pole piece by using an optical microscope as an original data set;
s3, distinguishing and labeling active materials of the lithium ion battery, which are subjected to target test, in the pole piece image of the original data set, drawing the active materials in the section of the pole piece of the lithium ion battery by using a polygon by using a professional labeling tool LabelMe, and storing coordinate data of each vertex on a pixel scale so as to obtain a training data set;
s4, constructing a deep learning model Mask R-CNN, selecting a training data set, preprocessing a training data set image, inputting the training data set image into a neural network for training, calculating a loss result after gradient back propagation, continuously inputting the training data into the neural network when a specified period is not reached, and storing neural network parameters as a target test active material intelligent recognition tool when the specified period is reached;
(2) Real-time monitoring of lithium ion battery active material damage based on in-situ optical microscope observation platform:
s5, constructing an in-situ optical microscope observation platform of the lithium ion battery active material, and monitoring the sectional image of the lithium ion battery pole piece in real time in the cyclic charge and discharge process;
s6, based on an intelligent identification tool for target test active materials, intelligent identification of target test lithium ion battery active materials in battery pole piece optical microscope images recorded in real time in the cyclic charge and discharge process is completed, and mask images and coordinate information of each active material in the images are obtained;
(3) And testing and evaluating the damage degree of the lithium ion battery active material:
s7, judging the image gray value of each active material obtained by the test in the step 2 by adopting a method of setting a threshold value, dividing the particle surface into a crack area and a block area, wherein the crack degree of each active material = crack area/particle area, and testing and analyzing the crack degree of each active material to obtain the overall damage degree of active particles distributed by the pole piece to be observed.
2. The method for testing the damage degree of the lithium ion battery active material based on deep learning according to claim 1, wherein the lithium ion battery active material to be tested is active material particles in a battery pole piece, and can be lithium iron phosphate LiFePO 4 Lithium cobaltate LiCoO 2 Ternary material LiNi of NCM x Co y Mn 1-x-y O 2 Lithium titanate Li 4 Ti 5 O 12 Graphite C and silicon Si.
3. The method for testing the damage degree of the lithium ion battery active material based on deep learning according to claim 1, wherein the polishing method for preparing the lithium ion battery pole piece section sample is mechanical polishing, argon ion beam polishing or focused ion beam polishing.
4. The method for testing the damage degree of the lithium ion battery active material based on deep learning according to claim 1, wherein the in-situ optical microscope observation platform consists of an optical microscope, lithium ion battery charging and discharging equipment, an optical in-situ reaction tank and a computer; the pole piece to be observed, the counter electrode and the diaphragm in the optical in-situ reaction tank are clamped in a lamination mode, the cross sections of one sides of the pole piece to be observed, the counter electrode and the diaphragm facing the window are guaranteed to be level, and the sides are tightly attached to the window of the reaction tank; the electrode to be observed is polished by an argon ion beam to obtain a smooth section, and the section is used as a section which is clung to a window of the reaction tank.
5. The method for testing the damage degree of the lithium ion battery active material based on deep learning is characterized in that the cyclic charge and discharge process in the step S5 is as follows: and setting a charging and discharging process step of multiplying power charging, standing, multiplying power discharging, standing and circulating for the optical in-situ reaction tank, wherein the voltage interval is 2.8-4.7V, and the charging and discharging multiplying power is set to be 0.5C.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202310405813.8A CN116539618A (en) | 2023-04-17 | 2023-04-17 | Deep learning-based lithium ion battery active material damage degree testing method |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202310405813.8A CN116539618A (en) | 2023-04-17 | 2023-04-17 | Deep learning-based lithium ion battery active material damage degree testing method |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN116539618A true CN116539618A (en) | 2023-08-04 |
Family
ID=87446093
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202310405813.8A Pending CN116539618A (en) | 2023-04-17 | 2023-04-17 | Deep learning-based lithium ion battery active material damage degree testing method |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN116539618A (en) |
Cited By (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN117784388A (en) * | 2024-02-28 | 2024-03-29 | 宁波永新光学股份有限公司 | High dynamic range metallographic image generation method based on camera response curve |
-
2023
- 2023-04-17 CN CN202310405813.8A patent/CN116539618A/en active Pending
Cited By (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN117784388A (en) * | 2024-02-28 | 2024-03-29 | 宁波永新光学股份有限公司 | High dynamic range metallographic image generation method based on camera response curve |
| CN117784388B (en) * | 2024-02-28 | 2024-05-07 | 宁波永新光学股份有限公司 | High dynamic range metallographic image generation method based on camera response curve |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| Mei et al. | Understanding of Li‐plating on graphite electrode: detection, quantification and mechanism revelation | |
| CN111141755B (en) | Method for detecting internal defects of battery cell | |
| CN107991615B (en) | Method for representing performance of battery cathode material by lithium content | |
| CN110333463B (en) | Battery cell consistency screening method and system | |
| CN116539618A (en) | Deep learning-based lithium ion battery active material damage degree testing method | |
| CN112240986B (en) | Evaluation method for lithium separation and uniformity of large-size soft-package lithium ion battery | |
| CN109061478A (en) | A method of it is tested using EIS and carries out lithium ion battery service life qualitative forecasting | |
| CN114217238A (en) | Method for predicting cycle life of lithium ion battery | |
| CN112698208A (en) | System and method for in-situ measurement of Young modulus and partial molar volume of lithium battery material | |
| CN113189498A (en) | Nondestructive lithium analysis detection method | |
| Chen et al. | Li+ concentration waves in a liquid electrolyte of Li-ion batteries with porous graphite-based electrodes | |
| Wen et al. | The scaling of charging rate and cycle number of commercial batteries | |
| CN115548482A (en) | Lithium supplementing method, battery preparation method and battery | |
| Li et al. | Accelerated aging of lithium-ion batteries: Bridging battery aging analysis and operational lifetime prediction | |
| US11555860B2 (en) | Method for testing and evaluating mechanical performance of lithium ion battery electrode based on nano-indentation technology | |
| CN109655098A (en) | The failure analysis method of secondary cell battery core | |
| CN215219088U (en) | Device for predicting lithium battery capacity by formation data | |
| CN113607774A (en) | Electrode strain field in-situ monitoring device and method for marking fluorescent quantum dot speckles | |
| CN116165563A (en) | Ship power lithium battery health grade identification method based on neutron imaging | |
| Xu et al. | Understanding the process of lithium deposition on a graphite anode for better lithium-ion batteries | |
| CN114859233A (en) | Lithium ion battery capacity loss prediction method oriented to full life utilization | |
| CN111948554B (en) | Method for reducing mechanical degradation of lithium ion battery | |
| CN114778633A (en) | Single-layer particle electrode for electrochemical analysis and electrochemical analysis method | |
| CN113483686A (en) | Electrode material strain measurement method and device | |
| CN113921892A (en) | Lithium ion battery electrolyte infiltration judgment method |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination |