CN116071359A - Battery aging degree detection method, electronic equipment and storage medium - Google Patents

Battery aging degree detection method, electronic equipment and storage medium Download PDF

Info

Publication number
CN116071359A
CN116071359A CN202310212696.3A CN202310212696A CN116071359A CN 116071359 A CN116071359 A CN 116071359A CN 202310212696 A CN202310212696 A CN 202310212696A CN 116071359 A CN116071359 A CN 116071359A
Authority
CN
China
Prior art keywords
convolution kernel
battery
convolution
image
size
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.)
Granted
Application number
CN202310212696.3A
Other languages
Chinese (zh)
Other versions
CN116071359B (en
Inventor
王芳
刘仕强
王文斌
王晓杰
杨亮
马天翼
姜成龙
韦振
Current Assignee (The listed assignees may be inaccurate. Google has not performed a legal analysis and makes no representation or warranty as to the accuracy of the list.)
China Automobile Information Technology Tianjin Co ltd
China Automotive Research New Energy Vehicle Inspection Center Tianjin Co ltd
Original Assignee
China Automobile Information Technology Tianjin Co ltd
China Automotive Research New Energy Vehicle Inspection Center Tianjin Co ltd
Priority date (The priority date 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 date listed.)
Filing date
Publication date
Application filed by China Automobile Information Technology Tianjin Co ltd, China Automotive Research New Energy Vehicle Inspection Center Tianjin Co ltd filed Critical China Automobile Information Technology Tianjin Co ltd
Priority to CN202310212696.3A priority Critical patent/CN116071359B/en
Publication of CN116071359A publication Critical patent/CN116071359A/en
Application granted granted Critical
Publication of CN116071359B publication Critical patent/CN116071359B/en
Active legal-status Critical Current
Anticipated expiration legal-status Critical

Links

Images

Classifications

    • GPHYSICS
    • G06COMPUTING; CALCULATING OR COUNTING
    • G06TIMAGE DATA PROCESSING OR GENERATION, IN GENERAL
    • G06T7/00Image analysis
    • G06T7/0002Inspection of images, e.g. flaw detection
    • G06T7/0004Industrial image inspection
    • GPHYSICS
    • G06COMPUTING; CALCULATING OR COUNTING
    • G06NCOMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
    • G06N3/00Computing arrangements based on biological models
    • G06N3/02Neural networks
    • G06N3/08Learning methods
    • GPHYSICS
    • G06COMPUTING; CALCULATING OR COUNTING
    • G06TIMAGE DATA PROCESSING OR GENERATION, IN GENERAL
    • G06T2207/00Indexing scheme for image analysis or image enhancement
    • G06T2207/20Special algorithmic details
    • G06T2207/20081Training; Learning
    • GPHYSICS
    • G06COMPUTING; CALCULATING OR COUNTING
    • G06TIMAGE DATA PROCESSING OR GENERATION, IN GENERAL
    • G06T2207/00Indexing scheme for image analysis or image enhancement
    • G06T2207/20Special algorithmic details
    • G06T2207/20084Artificial neural networks [ANN]
    • YGENERAL 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
    • Y02TECHNOLOGIES OR APPLICATIONS FOR MITIGATION OR ADAPTATION AGAINST CLIMATE CHANGE
    • Y02EREDUCTION OF GREENHOUSE GAS [GHG] EMISSIONS, RELATED TO ENERGY GENERATION, TRANSMISSION OR DISTRIBUTION
    • Y02E60/00Enabling technologies; Technologies with a potential or indirect contribution to GHG emissions mitigation
    • Y02E60/10Energy storage using batteries

Landscapes

  • Engineering & Computer Science (AREA)
  • Theoretical Computer Science (AREA)
  • Physics & Mathematics (AREA)
  • General Physics & Mathematics (AREA)
  • Biophysics (AREA)
  • Evolutionary Computation (AREA)
  • Health & Medical Sciences (AREA)
  • Life Sciences & Earth Sciences (AREA)
  • Artificial Intelligence (AREA)
  • Biomedical Technology (AREA)
  • Quality & Reliability (AREA)
  • Computational Linguistics (AREA)
  • Data Mining & Analysis (AREA)
  • Computer Vision & Pattern Recognition (AREA)
  • General Health & Medical Sciences (AREA)
  • Molecular Biology (AREA)
  • Computing Systems (AREA)
  • General Engineering & Computer Science (AREA)
  • Mathematical Physics (AREA)
  • Software Systems (AREA)
  • Image Analysis (AREA)

Abstract

The embodiment of the invention discloses a battery aging degree detection method, electronic equipment and a storage medium. The method comprises the following steps: acquiring an ultrasonic image of a battery to be detected, wherein the ultrasonic image reflects the distribution condition of electrolyte active substances in the battery; identifying the ultrasonic image by using a trained convolutional neural network to obtain the aging degree of the battery to be detected; the following operations are executed in each convolution layer of the network: performing channel transformation on the input image of the convolution layer by using the first convolution kernel to ensure that the quantity of the transformed channels is consistent with the total quantity of parameters of the second convolution kernel; weighting each parameter of the second convolution kernel according to the whole pixel information of each transformed channel image, wherein the whole pixel information can reflect the total content of active substances in the battery; and convolving the input image by using the weighted convolution check to obtain the output of the convolution layer. The embodiment can improve the detection accuracy and reduce the detection cost.

Description

Battery aging degree detection method, electronic equipment and storage medium
Technical Field
The embodiment of the invention relates to the field of artificial intelligence, in particular to a battery aging degree detection method, electronic equipment and a storage medium.
Background
With the development of new energy automobiles, the energy density and the output efficiency of the power battery are continuously improved, but the temperature monitoring and health evaluation technologies of the battery are obviously lagged. The detection of the aging degree of the battery is always a key problem of echelon utilization of the power battery of the new energy automobile and regeneration and recycling of the scrapped power battery material.
In the prior art, the aging degree of the battery is usually detected by using the charging current and the voltage, and the defects are low accuracy, a large amount of circulating charging and discharging data are required to be accumulated for training and learning, and the detection cost is high.
Disclosure of Invention
The embodiment of the invention provides a battery aging degree detection method, electronic equipment and a storage medium, which are used for detecting the aging degree of a battery through identifying an ultrasonic image of the battery, so that the detection accuracy is improved, and the detection cost is reduced.
In a first aspect, an embodiment of the present invention provides a method for detecting a battery aging degree, including:
acquiring an ultrasonic image of a battery to be detected, wherein the ultrasonic image reflects the distribution condition of electrolyte active substances in the battery;
identifying the ultrasonic image by using a trained convolutional neural network to obtain the aging degree of the battery to be detected;
wherein, each convolution layer in the trained convolution neural network comprises a first convolution kernel and a second convolution kernel, and the following operations are executed in each convolution layer:
performing channel transformation on the input image of the convolution layer by using the first convolution kernel to ensure that the quantity of the transformed channels is consistent with the total quantity of parameters of the second convolution kernel;
weighting each parameter of the second convolution kernel according to the whole pixel information of each transformed channel image, wherein the whole pixel information can reflect the total content of active substances in the battery;
and convolving the input image by using the weighted convolution check to obtain the output of the convolution layer.
In a second aspect, an embodiment of the present invention further provides an electronic device, including:
one or more processors;
a memory for storing one or more programs,
the one or more programs, when executed by the one or more processors, cause the one or more processors to implement the battery degradation level detection method of any one of the embodiments.
In a third aspect, an embodiment of the present invention further provides a computer readable storage medium having stored thereon a computer program, which when executed by a processor implements the method for detecting a battery aging degree according to any of the embodiments.
The embodiment of the invention provides a battery aging degree detection method, which detects the aging degree of a battery by identifying an ultrasonic image of the battery through an improved convolutional neural network. Meanwhile, the embodiment of the invention combines the characteristics of the battery ultrasonic image, replaces the convolution kernel in the traditional convolution neural network with a convolution layer, firstly verifies the channel transformation of the ultrasonic image and the input image of each layer in the convolution layer by utilizing a first convolution layer, and fits the nonlinear relation between the ultrasonic image and the aging degree to a certain degree through a simple structure; then, according to the characteristic that the aging degree of the battery is only related to the total content of the electrolyte active substances, extracting the whole pixel information of the ultrasonic image to reflect the total content of the active substances in the battery, ignoring the image details, omitting a complex detail feature extraction network (such as a attention mechanism and the like), and simultaneously not affecting the study of the relation between the ultrasonic image and the aging degree; finally, the convolution kernel parameters are weighted according to the whole pixel information of the image, so that the weighted convolution kernel can inherit the nonlinear relation fitted by the first convolution kernel and realize the dynamic adjustment of the parameters along with the total content of active substances in the battery. The whole network is simple in structure, the fitting of complex relations can be realized only through the first convolution kernel and the second convolution kernel, and the parameters of the convolution kernels can be flexibly adjusted for different ultrasonic images, so that the convolution kernels can better adapt to active material information reflected in the images, the degree of distinction between different images is enhanced, and the detection accuracy is improved; meanwhile, the key parameters to be trained only comprise parameters in the first convolution kernel and the second convolution kernel, so that the number of parameters is greatly reduced, and the calculation speed can be effectively improved in both the training stage and the using stage.
Drawings
In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the drawings that are needed in the description of the embodiments or the prior art will be briefly described, and it is obvious that the drawings in the description below are some embodiments of the present invention, and other drawings can be obtained according to the drawings without inventive effort for a person skilled in the art.
Fig. 1 is a flowchart of a method for detecting a battery aging degree according to an embodiment of the present invention.
Fig. 2 is a schematic diagram Of an ultrasonic image Of the same battery under different aging degrees according to an embodiment Of the present invention, where fig. 2 (a) is an ultrasonic image corresponding to SOH (State Of Health) =90%, fig. 2 (b) is an ultrasonic image corresponding to soh=80%, fig. 2 (c) is an ultrasonic image corresponding to soh=70%, fig. 2 (d) is an ultrasonic image corresponding to soh=60%, fig. 2 (e) is an ultrasonic image corresponding to soh=50%, and fig. 2 (f) is an ultrasonic image corresponding to soh=40%.
Fig. 3 is a schematic diagram of a conventional convolutional neural network.
Fig. 4 is a schematic diagram of an improved convolutional neural network provided by an embodiment of the present invention.
Fig. 5 is a schematic diagram of a side processing layer according to an embodiment of the present invention.
Fig. 6 is a schematic structural diagram of an electronic device according to an embodiment of the present invention.
Detailed Description
In order to make the objects, technical solutions and advantages of the present invention more apparent, the technical solutions of the present invention will be clearly and completely described below. It will be apparent that the described embodiments are only some, but not all, embodiments of the invention. All other embodiments, which can be made by one of ordinary skill in the art without undue burden from the invention, are within the scope of the invention.
In the description of the present invention, it should be noted that the directions or positional relationships indicated by the terms "center", "upper", "lower", "left", "right", "vertical", "horizontal", "inner", "outer", etc. are based on the directions or positional relationships shown in the drawings, are merely for convenience of describing the present invention and simplifying the description, and do not indicate or imply that the devices or elements referred to must have a specific orientation, be configured and operated in a specific orientation, and thus should not be construed as limiting the present invention. Furthermore, the terms "first," "second," and "third" are used for descriptive purposes only and are not to be construed as indicating or implying relative importance.
In the description of the present invention, it should also be noted that, unless explicitly specified and limited otherwise, the terms "mounted," "connected," and "connected" are to be construed broadly, and may be either fixedly connected, detachably connected, or integrally connected, for example; can be mechanically or electrically connected; can be directly connected or indirectly connected through an intermediate medium, and can be communication between two elements. The specific meaning of the above terms in the present invention will be understood in specific cases by those of ordinary skill in the art.
Fig. 1 is a flowchart of a method for detecting a battery aging degree according to an embodiment of the present invention. The method is suitable for detecting the aging degree of the same battery, and is executed by the electronic equipment. As shown in fig. 1, the method specifically includes:
s110, acquiring an ultrasonic image of the battery to be detected, wherein the ultrasonic image reflects the distribution condition of electrolyte active substances in the battery.
The ultrasonic image may be obtained by photographing the battery by ultrasonic waves. Since the reflection properties of the electrolyte in the battery against the ultrasonic waves are different from each other, the ultrasonic image can reflect the distribution of the electrolyte active material inside the battery, and the active material includes lithium ions and the like, for example. The active material content is an important index of the aging degree of the battery, so that the obtained ultrasonic images also show a large difference when the batteries with different aging degrees are photographed, as shown in fig. 2.
Specifically, the SOH values in percentage form are used in fig. 2 to characterize the aging degree of the battery, and the higher the SOH of the battery, the lower the aging degree, and it can be seen that the aging degree of the battery gradually decreases from fig. 2 (a) to fig. 2 (f). At the same time, the distribution of the electrolyte active material in the battery can be seen from each ultrasonic image, and if the pixel position in the middle region of the image (corresponding to the battery interior) is gray, it indicates that the active material is present in the position, and if it is black, it indicates that the active material is not present in the position. It can be seen that the active substance content is gradually reduced from fig. 2 (a) to fig. 2 (f). The following change law can be derived from this: the greater the gray area of the middle region of the ultrasound image, the higher the total active material content and the lower the battery aging.
And S120, identifying the ultrasonic image by using the trained convolutional neural network to obtain the aging degree of the battery to be detected.
Based on the change rule mentioned above, the present embodiment adopts the convolutional neural network to learn the relationship between the ultrasonic image and the aging degree of the battery, so that after the ultrasonic image of any one battery of the same type is input into the trained convolutional neural network, the network can automatically output the aging degree of the battery.
Conventional convolutional neural networks as shown in fig. 3, which is exemplary, include three layers of convolutional kernels 1, 2, 3, where black crosses represent the convolutional operations and layers 1 and 2 are the outputs of convolutional kernels 1 and 2, respectively. It can be seen that after the network training is completed, the parameters of each convolution kernel are fixed, and different ultrasonic images share the convolution kernel parameters. The method can not adjust convolution kernel parameters for different images, so that generalization capability is low, accuracy is low, and errors of different ultrasonic images of the same type tend to be consistent, and differences between the images can not be effectively distinguished. In view of the above drawbacks, the present embodiment improves a conventional convolutional neural network by replacing each convolution kernel in fig. 3 with a convolution layer including a plurality of convolution kernels, and dynamically adjusting parameters of the convolution kernels along with an input image in the convolution layer.
Fig. 4 is a schematic diagram of an improved convolutional neural network provided by an embodiment of the present invention. As shown in fig. 4, the network includes 3 convolutional layers, each of which includes a convolutional kernel and side processing layers. The side processing layer is used for extracting the whole pixel information of the input image of the convolution layer where the side processing layer is located, weighting another convolution kernel with fixed parameters according to the information, and the weighted convolution kernel is the convolution kernel. Since the input image of each convolution layer is either an ultrasonic image per se or a depth characteristic of an ultrasonic image, and can reflect the distribution condition of electrolyte active materials in the battery, the overall pixel information of the input image can reflect the total content of the active materials in the battery. In this embodiment, the convolution kernel parameters are weighted according to the total content of the active substances in the battery, so that the weighted convolution kernel can be dynamically adjusted according to different content of the active substances, which is more beneficial to improving generalization of the whole network and detecting precision of different aging degrees. It should be noted that, the present embodiment does not limit the number of convolution layers, and the 3 layers in fig. 4 are only examples.
Further, the structure of the side processing layer is shown in fig. 5, which includes two convolution kernels with fixed parameters, and for convenience of distinction and description, they are referred to as a first convolution kernel and a second convolution kernel, respectively, and in fact, the second convolution kernel is the above-mentioned "another convolution kernel with fixed parameters". Referring to fig. 5, in a trained convolutional neural network, the operations within each convolutional layer include the steps of:
step one, carrying out channel transformation on an input image of a convolution layer by utilizing a first convolution kernel, so that the quantity of the transformed channels is consistent with the total quantity of parameters of a second convolution kernel. Illustratively, the size of the input image is (batch size, w, h, y), where batch size represents the number of batches, w, h represent the size of the input image along the width and length directions, respectively, and y represents the number of channels of the input image. Correspondingly, the size of the first convolution kernel is (w 1, h1, y, x), wherein w1 and h1 respectively represent the size of the first convolution kernel along the width and length directions of the input image, and x represents the parameter total amount of the second convolution kernel; the second convolution kernel has dimensions (w 2, h 2), where w2 and h2 represent the dimensions of the second convolution kernel along the width and length directions of the input image, respectively. Since the number of channels after transformation coincides with the total amount of parameters of the second convolution kernel, x=w2×h2. The first convolution check with the size (w 1, h1, y, x) is used for convolving the input image with the size (batch size, w, h, y) to obtain an image with the size (batch size, w, h, x), and the conversion of the channel number can be realized. The nonlinear relation between the input and the output is fitted through the channel transformation of the first convolution kernel, and the nonlinear relation between the ultrasonic image and the aging degree is reflected to the extent.
And step two, weighting each parameter of the second convolution kernel according to the whole pixel information of each transformed channel image, wherein the whole pixel information can reflect the total content of active substances in the battery. For example, w×h pixels of each transformed channel image are added to obtain a pixel sum vector with a size of (1, x), where each element in the vector is a pixel sum, each pixel sum can reflect the total content of active substances in the battery, and these vectors are represented as (batch size, x) output layer 2 in fig. 5. After each vector is obtained, weighting is carried out on x pixels in each vector and x parameters of the second convolution kernel, so that a convolution kernel after the weighting is obtained, and each convolution kernel after the weighting corresponds to one image in batch processing. Specifically, taking fig. 5 as an example, mapping each pixel and each pixel to a weight between 0 and 1 by a Sigmoid function, to obtain an output layer 3 with a size (batch size, x); and then rearape function is adopted to rearrange the weights of x into a mask with the same size as the second convolution kernel (w 2, h 2), and the mask is used for dot multiplication with the second convolution kernel with the size of (w 2, h 2), so that weighted convolution kernels with the size of (w 2, h 2) are finally obtained. Because the aging degree of the battery is only related to the total content of the electrolyte active substances, the step focuses on the whole pixel information of the ultrasonic image through the summation of all pixels, and ignores the image details, thereby not only reducing the network complexity, but also not influencing the study of the relation between the ultrasonic image and the aging degree; and the convolution kernel parameters are weighted according to the whole pixel information of the image, so that the weighted convolution kernel can inherit the nonlinear relation fitted by the first convolution kernel and can be dynamically adjusted along with the total content of active substances in the battery.
And thirdly, convolving the input image by using the weighted convolution check to obtain the output of the convolution layer. Illustratively, each input image with the size (w, h) is subjected to two-dimensional convolution by using each weighted convolution kernel with the size (w 2, h 2), so as to obtain an output image with the size (batch size, w3, h 3), wherein w3 and h3 are the sizes of the output image along the width direction and the length direction respectively.
Optionally, the training process of the convolutional neural network includes: and acquiring ultrasonic images of the battery which is the same as the battery to be detected under different aging degrees. And training the convolutional neural network by utilizing each ultrasonic image, so that the output of the convolutional neural network is continuously approximate to the real aging degree corresponding to each ultrasonic image. After training, fixing parameters of each first convolution kernel and each second convolution kernel to obtain a trained convolution neural network.
The embodiment provides a battery aging degree detection method, which detects the aging degree of a battery through the identification of an improved convolutional neural network to an ultrasonic image of the battery, and compared with the detection method using charging current and voltage in the prior art, the detection method does not need to carry out a cyclic charging and discharging process, and effectively reduces the detection cost. Meanwhile, the embodiment combines the characteristics of the battery ultrasonic image, replaces the convolution kernel in the traditional convolution neural network with a convolution layer, firstly verifies the channel transformation of the ultrasonic image and the input image of each layer in the convolution layer by utilizing a first convolution layer, and fits the nonlinear relation between the ultrasonic image and the aging degree to a certain degree through a simple structure; then according to the characteristic that the aging degree of the battery is only related to the total content of the electrolyte active substances, carrying out full-pixel summation on the transformed image to pay attention to the whole pixel information of the ultrasonic image, neglecting the image details, omitting a complex detail feature extraction network (such as a attention mechanism and the like), and simultaneously not affecting the study on the relation between the ultrasonic image and the aging degree; finally, the convolution kernel parameters are weighted according to the whole pixel information of the image, so that the weighted convolution kernel can inherit the nonlinear relation fitted by the first convolution kernel and realize the dynamic adjustment of the parameters along with the total content of active substances in the battery. The whole network is simple in structure, fitting of complex relations can be achieved only through the first convolution kernel, the channel summation and the second convolution kernel, parameters of the convolution kernel can be flexibly adjusted according to different ultrasonic images, the convolution kernel can better adapt to active material information reflected in the images, distinguishing degree among different images is enhanced, and detection accuracy is improved; meanwhile, the key parameters to be trained only comprise parameters in the first convolution kernel and the second convolution kernel, so that the number of parameters is greatly reduced, and the calculation speed can be effectively improved in both the training stage and the using stage.
Fig. 6 is a schematic structural diagram of an electronic device according to an embodiment of the present invention, and as shown in fig. 6, the device includes a processor 60, a memory 61, an input device 62 and an output device 63; the number of processors 60 in the device may be one or more, one processor 60 being taken as an example in fig. 6; the processor 60, the memory 61, the input means 62 and the output means 63 in the device may be connected by a bus or other means, in fig. 6 by way of example.
The memory 61 is a computer readable storage medium, and may be used to store a software program, a computer executable program, and modules, such as program instructions/modules corresponding to the battery degradation detection method in the embodiment of the present invention. The processor 60 executes various functional applications of the device and data processing by running software programs, instructions and modules stored in the memory 61, i.e., implements the above-described battery degradation degree detection method.
The memory 61 may mainly include a storage program area and a storage data area, wherein the storage program area may store an operating system, at least one application program required for functions; the storage data area may store data created according to the use of the terminal, etc. In addition, the memory 61 may include high-speed random access memory, and may also include non-volatile memory, such as at least one magnetic disk storage device, flash memory device, or other non-volatile solid-state storage device. In some examples, memory 61 may further comprise memory remotely located relative to processor 60, which may be connected to the device via a network. Examples of such networks include, but are not limited to, the internet, intranets, local area networks, mobile communication networks, and combinations thereof.
The input means 62 may be used to receive entered numeric or character information and to generate key signal inputs related to user settings and function control of the device. The output 63 may comprise a display device such as a display screen.
The embodiment of the invention also provides a computer readable storage medium, on which a computer program is stored, which when executed by a processor implements the battery aging degree detection method of any of the embodiments.
The computer storage media of embodiments of the invention may take the form of any combination of one or more computer-readable media. The computer readable medium may be a computer readable signal medium or a computer readable storage medium. The computer readable storage medium can be, for example, but not limited to, an electronic, magnetic, optical, electromagnetic, infrared, or semiconductor system, apparatus, or device, or a combination of any of the foregoing. More specific examples (a non-exhaustive list) of the computer-readable storage medium would include the following: an electrical connection having one or more wires, a portable computer diskette, a hard disk, a Random Access Memory (RAM), a read-only memory (ROM), an erasable programmable read-only memory (EPROM or flash memory), an optical fiber, a portable compact disc read-only memory (CD-ROM), an optical storage device, a magnetic storage device, or any suitable combination of the foregoing. In this document, a computer readable storage medium may be any tangible medium that can contain, or store a program for use by or in connection with an instruction execution system, apparatus, or device.
The computer readable signal medium may include a propagated data signal with computer readable program code embodied therein, either in baseband or as part of a carrier wave. Such a propagated data signal may take any of a variety of forms, including, but not limited to, electro-magnetic, optical, or any suitable combination of the foregoing. A computer readable signal medium may also be any computer readable medium that is not a computer readable storage medium and that can communicate, propagate, or transport a program for use by or in connection with an instruction execution system, apparatus, or device.
Program code embodied on a computer readable medium may be transmitted using any appropriate medium, including but not limited to wireless, wireline, optical fiber cable, RF, etc., or any suitable combination of the foregoing.
Computer program code for carrying out operations of the present invention may be written in any combination of one or more programming languages, including an object oriented programming language such as Java, smalltalk, C ++ and conventional procedural programming languages, such as the C-programming language or similar programming languages. The program code may execute entirely on the user's computer, partly on the user's computer, as a stand-alone software package, partly on the user's computer and partly on a remote computer or entirely on the remote computer or server. In the case of a remote computer, the remote computer may be connected to the user's computer through any kind of network, including a Local Area Network (LAN) or a Wide Area Network (WAN), or may be connected to an external computer (for example, through the Internet using an Internet service provider).
Finally, it should be noted that: the above embodiments are only for illustrating the technical solution of the present invention, and not for limiting the same; although the invention has been described in detail with reference to the foregoing embodiments, it will be understood by those of ordinary skill in the art that: the technical scheme described in the foregoing embodiments can be modified or some or all of the technical features thereof can be replaced by equivalents; such modifications and substitutions do not depart from the essence of the corresponding technical solutions from the technical solutions of the embodiments of the present invention.

Claims (10)

1. A method for detecting the degree of battery aging, comprising:
acquiring an ultrasonic image of a battery to be detected, wherein the ultrasonic image reflects the distribution condition of electrolyte active substances in the battery;
identifying the ultrasonic image by using a trained convolutional neural network to obtain the aging degree of the battery to be detected;
wherein, each convolution layer in the trained convolution neural network comprises a first convolution kernel and a second convolution kernel, and the following operations are executed in each convolution layer:
performing channel transformation on the input image of the convolution layer by using the first convolution kernel to ensure that the quantity of the transformed channels is consistent with the total quantity of parameters of the second convolution kernel;
weighting each parameter of the second convolution kernel according to the whole pixel information of each transformed channel image, wherein the whole pixel information can reflect the total content of active substances in the battery;
and convolving the input image by using the weighted convolution check to obtain the output of the convolution layer.
2. The method of claim 1, wherein weighting parameters of the second convolution kernel based on the overall pixel information of each channel image comprises:
adding all pixels of each channel image;
and weighting each parameter of the second convolution kernel according to each pixel sum.
3. The method of claim 1, wherein the step of determining the position of the substrate comprises,
the first convolution kernel has a size of (w 1, h1, y, x), wherein w1 and h1 respectively represent the sizes of the first convolution kernel along the width and length directions of the input image, y represents the number of channels of the input image, and x represents the total parameter amount of the second convolution kernel;
the second convolution kernel has a size (w 2, h 2), where w2 and h2 represent the dimensions of the second convolution kernel along the width and length directions of the input image, x=w2×h2, respectively.
4. A method according to claim 3, wherein the input image has a size (batch size, w, h, y), wherein batch size represents the number of batches and w and h represent the dimensions of the input image in the width and length directions, respectively;
the channel transformation is performed on the input image of the convolution layer by using the first convolution kernel, so that the quantity of the transformed channels is consistent with the total quantity of parameters of the second convolution kernel, and the method comprises the following steps:
convolving an input image with the size (batch size, w, h, y) by using a first convolution kernel with the size (w 1, h1, y, x) to obtain an image with the size (batch size, w, h, x), and realizing the conversion of the number of channels;
the weighting each parameter of the second convolution kernel according to the whole pixel information of each transformed channel image includes:
adding w×h pixels of each channel image after transformation to obtain pixels with the size of (1, x) and vectors;
and weighting the x parameters of the second convolution kernel according to x pixel sums in each pixel sum vector to obtain a convolution kernel after the weighting of the batch size.
5. The method of claim 1, wherein weighting parameters of the second convolution kernel according to pixel sums comprises:
mapping each pixel sum to a weight between 0 and 1;
rearranging the weights into masks with the same size as the second convolution kernel;
and performing point multiplication on the mask and the second convolution kernel to realize parameter weighting.
6. The method of claim 1, further comprising, prior to said identifying said ultrasound image using a trained convolutional neural network to obtain a degree of aging of said battery to be detected:
acquiring ultrasonic images of the battery which is the same as the battery to be detected under different aging degrees;
and training the convolutional neural network by utilizing each ultrasonic image, so that the output of the convolutional neural network is continuously approximate to the real aging degree corresponding to each ultrasonic image.
7. The method as recited in claim 6, further comprising:
after training, fixing parameters of each first convolution kernel and each second convolution kernel to obtain a trained convolution neural network.
8. The method of claim 1, wherein the degree of aging of the battery is expressed in terms of a battery health value in percent.
9. An electronic device, comprising:
one or more processors;
a memory for storing one or more programs,
when the one or more programs are executed by the one or more processors, the one or more processors are caused to implement the battery degradation detection method of any one of claims 1 to 8.
10. A computer-readable storage medium, having stored thereon a computer program which, when executed by a processor, implements the battery degradation detection method according to any one of claims 1 to 8.
CN202310212696.3A 2023-03-08 2023-03-08 Battery aging degree detection method, electronic equipment and storage medium Active CN116071359B (en)

Priority Applications (1)

Application Number Priority Date Filing Date Title
CN202310212696.3A CN116071359B (en) 2023-03-08 2023-03-08 Battery aging degree detection method, electronic equipment and storage medium

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
CN202310212696.3A CN116071359B (en) 2023-03-08 2023-03-08 Battery aging degree detection method, electronic equipment and storage medium

Publications (2)

Publication Number Publication Date
CN116071359A true CN116071359A (en) 2023-05-05
CN116071359B CN116071359B (en) 2023-06-23

Family

ID=86173373

Family Applications (1)

Application Number Title Priority Date Filing Date
CN202310212696.3A Active CN116071359B (en) 2023-03-08 2023-03-08 Battery aging degree detection method, electronic equipment and storage medium

Country Status (1)

Country Link
CN (1) CN116071359B (en)

Cited By (2)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN117046864A (en) * 2023-08-28 2023-11-14 国能龙源环保有限公司 Photovoltaic module recycling method and device
CN117237349A (en) * 2023-11-14 2023-12-15 珠海市嘉德电能科技有限公司 Thermal runaway protection method, device, equipment and storage medium of battery management system

Citations (11)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN111505519A (en) * 2020-04-04 2020-08-07 江苏集萃华科智能装备科技有限公司 Lithium ion battery cycle life prediction method based on acoustic measurement
CN111581414A (en) * 2019-02-18 2020-08-25 北京京东尚科信息技术有限公司 Clothing identification, classification and retrieval method, device, equipment and storage medium
CN112634241A (en) * 2020-12-25 2021-04-09 周婷婷 Photovoltaic cell power attenuation analysis method and system based on artificial intelligence
CN113466706A (en) * 2021-07-26 2021-10-01 上海伟翔众翼新能源科技有限公司 Lithium battery echelon utilization residual life prediction method based on convolutional neural network
CN113516224A (en) * 2020-04-09 2021-10-19 深圳云天励飞技术有限公司 Neural network optimization method and device, electronic device and storage medium
CN113987926A (en) * 2021-10-20 2022-01-28 山东联安信息科技有限责任公司 Flywheel energy storage power generation vehicle health condition early warning system and method based on artificial intelligence
CN114239814A (en) * 2022-02-25 2022-03-25 杭州研极微电子有限公司 Training method of convolution neural network model for image processing
CN114384149A (en) * 2021-11-25 2022-04-22 西安交通大学 Energy storage device state detection method based on ultrasonic detection technology
US20220150143A1 (en) * 2020-11-12 2022-05-12 At&T Intellectual Property I, L.P. Classification of encrypted internet traffic with binary traffic vectors
CN115273050A (en) * 2022-07-01 2022-11-01 中国人民解放军空军工程大学 Limited sample trajectory target identification method capable of deeply separating and fusing convolutional neural network
CN115291122A (en) * 2022-08-24 2022-11-04 华中科技大学 Method for acquiring internal information of lithium ion battery based on ultrasonic reflection image

Patent Citations (11)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN111581414A (en) * 2019-02-18 2020-08-25 北京京东尚科信息技术有限公司 Clothing identification, classification and retrieval method, device, equipment and storage medium
CN111505519A (en) * 2020-04-04 2020-08-07 江苏集萃华科智能装备科技有限公司 Lithium ion battery cycle life prediction method based on acoustic measurement
CN113516224A (en) * 2020-04-09 2021-10-19 深圳云天励飞技术有限公司 Neural network optimization method and device, electronic device and storage medium
US20220150143A1 (en) * 2020-11-12 2022-05-12 At&T Intellectual Property I, L.P. Classification of encrypted internet traffic with binary traffic vectors
CN112634241A (en) * 2020-12-25 2021-04-09 周婷婷 Photovoltaic cell power attenuation analysis method and system based on artificial intelligence
CN113466706A (en) * 2021-07-26 2021-10-01 上海伟翔众翼新能源科技有限公司 Lithium battery echelon utilization residual life prediction method based on convolutional neural network
CN113987926A (en) * 2021-10-20 2022-01-28 山东联安信息科技有限责任公司 Flywheel energy storage power generation vehicle health condition early warning system and method based on artificial intelligence
CN114384149A (en) * 2021-11-25 2022-04-22 西安交通大学 Energy storage device state detection method based on ultrasonic detection technology
CN114239814A (en) * 2022-02-25 2022-03-25 杭州研极微电子有限公司 Training method of convolution neural network model for image processing
CN115273050A (en) * 2022-07-01 2022-11-01 中国人民解放军空军工程大学 Limited sample trajectory target identification method capable of deeply separating and fusing convolutional neural network
CN115291122A (en) * 2022-08-24 2022-11-04 华中科技大学 Method for acquiring internal information of lithium ion battery based on ultrasonic reflection image

Cited By (3)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN117046864A (en) * 2023-08-28 2023-11-14 国能龙源环保有限公司 Photovoltaic module recycling method and device
CN117237349A (en) * 2023-11-14 2023-12-15 珠海市嘉德电能科技有限公司 Thermal runaway protection method, device, equipment and storage medium of battery management system
CN117237349B (en) * 2023-11-14 2024-03-05 珠海市嘉德电能科技有限公司 Thermal runaway protection method, device, equipment and storage medium of battery management system

Also Published As

Publication number Publication date
CN116071359B (en) 2023-06-23

Similar Documents

Publication Publication Date Title
CN116071359B (en) Battery aging degree detection method, electronic equipment and storage medium
CN111028237A (en) Image segmentation method and device and terminal equipment
WO2024051655A1 (en) Method and apparatus for processing histopathological whole-slide image, and medium and electronic device
CN115546630A (en) Construction site extraction method and system based on remote sensing image characteristic target detection
CN117475158A (en) Error prediction method, device, equipment and medium for capacitive voltage transformer
CN116452810A (en) Multi-level semantic segmentation method and device, electronic equipment and storage medium
CN112270259A (en) SAR image ship target rapid detection method based on lightweight convolutional neural network
CN112200772A (en) Pox check out test set
CN116310582A (en) Classification model training method, image classification method, device, medium and equipment
CN117589792A (en) Ending position detection method, ending position detection device, computer equipment and storage medium
CN117014507A (en) Training method of task unloading model, task unloading method and device
CN113139617B (en) Power transmission line autonomous positioning method and device and terminal equipment
CN116090543A (en) Model compression method and device, computer readable medium and electronic equipment
CN115272682A (en) Target object detection method, target detection model training method and electronic equipment
CN116778534B (en) Image processing method, device, equipment and medium
CN118033452A (en) Method and terminal for predicting service life of feature extraction lithium ion battery
EP4328857A1 (en) Disparity estimation method and apparatus, and image processing device and storage medium
CN116894163B (en) Charging and discharging facility load prediction information generation method and device based on information security
CN116854111B (en) Purification method and system of electronic grade lithium hexafluorophosphate
CN118505687B (en) Photovoltaic panel defect detection method, storage medium and electronic device
WO2024051364A1 (en) Living body detection model training method and apparatus, and storage medium and terminal
CN117152783A (en) Formula labeling method and device, storage medium and electronic equipment
CN117496163A (en) Data processing method, device, electronic equipment and computer readable storage medium
CN118799339A (en) Image segmentation method, device, terminal equipment and computer readable storage medium
CN116264019A (en) Training method and device for silent image reproduction detection model and related equipment

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
GR01 Patent grant
GR01 Patent grant