An automotive power battery collision test device and its system

Abstract

The present invention discloses an automobile power battery collision test device and its system, belonging to the technical field of battery testing. The present invention analyzes the position and depth of the impact depression of the automobile power battery after being impacted through a deformation analysis module, and then obtains the battery capacity of the automobile power battery after being impacted after a preset number of charge and discharge cycles through a life test module. According to the battery capacity, it is judged whether the position and depth of the impact depression affect battery safety. Finally, a pre-trained deep learning model is used to estimate the minimum depression depth affecting the safety of the automobile power battery at different positions, accurately providing a safety judgment after the battery is impacted and deformed. In the depression recognition module, through a pre-trained improved YOLOv8 model, the depression area in the image of the impacted surface is recognized, improving the accuracy of depression recognition, and thus improving the effectiveness and accuracy of the test.

Description

Technical Field

[0001] The present invention relates to the technical field of battery testing, in particular to a collision test device and system for automotive power batteries. Background Art

[0002] In the testing of automotive power batteries, the collision test of automotive power batteries is extremely important. After the battery is impacted, it will deform to varying degrees, that is, the battery will have dents. Sometimes the dents are very small, which makes people ignore that small dents may also affect the safety or lifespan of the battery, and further lead to driving hazards. Therefore, there is an urgent need for a test system that can obtain whether different dent degrees of the battery have an actual impact on the safe use of the battery. Summary of the Invention

[0003] In order to solve the technical problems existing in the prior art, the present invention provides a collision test system for automotive power batteries, which specifically includes:

[0004] An impact device for impacting an automotive power battery through an impact head;

[0005] A deformation analysis module, including an image acquisition module, a dent recognition module, an impact dent determination module, and a laser ranging device;

[0006] The image acquisition module is used to acquire the image of the impacted surface of the automotive power battery, and the image of the impacted surface includes an initial image and a post-impact image;

[0007] The dent recognition module is used to identify dent information in the image of the impacted surface, specifically: through a pre-trained improved YOLOv8 model, identify the dent area in the image of the impacted surface; use edge detection to identify the dent contour from the dent area;

[0008] The improved YOLOv8 model is improved based on the YOLOv8 model, specifically: replacing the ordinary convolution module in the backbone network of the YOLOv8 model with a Ghost convolution module; introducing a CA attention mechanism into the C2f module in the feature enhancement network of the YOLOv8 model;

[0009] The impact dent determination module is used to determine the impact dent obtained after the automotive power battery is impacted by the impact head in the post-impact image by comparing the dent information in the initial image and the post-impact image;

[0010] The laser ranging device is used to measure the dent depth of the impact dent;

[0011] A lifespan test module for obtaining the battery capacity after performing a preset number of charge and discharge cycles on the impacted automotive power battery;

[0012] A judgment module, configured to determine whether the dent depth at the position of the impact dent of the current automotive power battery affects the safety of the automotive power battery according to the battery capacity;

[0013] A prediction module, configured to estimate the minimum dent depth affecting the safety of the automotive power battery at different positions of the automotive power battery by using a pre-trained deep learning model.

[0014] Further, introducing the CA attention mechanism into the C2f module in the feature enhancement network of the YOLOv8 model specifically includes:

[0015] Input the feature map output by the splicing layer of the C2f module into the CA self-attention mechanism module. The CA self-attention mechanism module encodes the feature map in the horizontal and vertical directions to generate a horizontal direction attention map and a vertical direction attention map, and then fuses them to obtain a comprehensive attention map. The comprehensive attention map is used to weight the feature map to obtain an enhanced feature map, and the enhanced feature map is input into the second convolutional layer of the C2f module.

[0016] Further, measuring the dent depth of the impact dent specifically includes:

[0017] Obtain the distance h1 between the laser ranging device and the surface of the non-dented area of the impacted surface of the automotive power battery through the laser ranging device;

[0018] Obtain the position with the maximum gray value within the dent contour of the impact dent in the post-collision image and use its actual position on the impacted surface of the automotive power battery as the second ranging point;

[0019] After aligning the ranging beam of the laser ranging device with the second ranging point, obtain the distance h2 between the laser ranging device and the second ranging point;

[0020] The dent depth of the impact dent is the value of h2 minus h1.

[0021] Further, determining whether the dent depth at the position of the impact dent of the current automotive power battery affects the safety of the automotive power battery specifically includes:

[0022] If the battery capacity is less than or equal to the preset capacity threshold, the dent depth at the position of the impact dent of the current automotive power battery affects the safety of the automotive power battery.

[0023] Further, the system further includes a parameter acquisition module, configured to acquire the mass, initial velocity, and impact instantaneous velocity of the impact head.

[0024] Further, the judgment module is further configured to determine whether an object with the same impact head mass as the current one colliding with the automotive power battery at the same impact instantaneous speed affects the safety of the automotive power battery according to the battery capacity, specifically:

[0025] If the battery capacity is less than or equal to a preset capacity threshold, then an object with the same impact head mass as the current one colliding with the automotive power battery at the same impact instantaneous speed will affect the safety of the automotive power battery.

[0026] Further, the image acquisition module further includes an image preprocessing unit for preprocessing the initial image and the post-collision image of the automotive power battery.

[0027] The present invention also provides an automotive power battery collision test device, including:

[0028] An impact device for impacting the automotive power battery through an impact head;

[0029] An image acquisition device for acquiring the image of the impacted surface of the automotive power battery, and the image of the impacted surface includes an initial image and a post-collision image;

[0030] A life test device for obtaining the battery capacity after performing a preset number of charge and discharge cycles on the impacted automotive power battery;

[0031] A laser ranging device for measuring the depression depth of the impact depression;

[0032] An analysis platform provided with a depression recognition algorithm, an impact depression determination algorithm, a judgment algorithm, and a prediction algorithm;

[0033] The depression recognition algorithm is used to identify the depression area in the image of the impacted surface through a pre-trained improved YOLOv8 model, and use edge detection to identify the depression contour from the depression area;

[0034] The improved YOLOv8 model is improved based on the YOLOv8 model. Specifically, the ordinary convolution module in the backbone network of the YOLOv8 model is replaced with a Ghost convolution module; a CA attention mechanism is introduced into the C2f module in the feature enhancement network of the YOLOv8 model;

[0035] The impact depression is determined by the impact depression determination algorithm, specifically by comparing the depression information in the initial image and the post-collision image to determine the impact depression on the automotive power battery obtained after being impacted by the impact head in the post-collision image;

[0036] The judgment algorithm is used to determine whether the depression depth at the position of the impact depression of the current automotive power battery affects the safety of the automotive power battery according to the battery capacity;

[0037] The prediction algorithm uses a pre-trained deep learning model to estimate the minimum indentation depth affecting the safety of the automotive power battery at different positions of the automotive power battery.

[0038] Further, a CA attention mechanism is introduced into the C2f module in the feature enhancement network of the YOLOv8 model, specifically:

[0039] The feature map output by the splicing layer of the C2f module is input into the CA self-attention mechanism module. The CA self-attention mechanism module encodes the feature map in the horizontal and vertical directions to generate a horizontal direction attention map and a vertical direction attention map, and then fuses them to obtain a comprehensive attention map. The comprehensive attention map is used to weight the feature map to obtain an enhanced feature map, and the enhanced feature map is input into the second convolutional layer of the C2f module.

[0040] Further, the measurement of the indentation depth of the impact indentation is specifically:

[0041] The distance h1 between the laser ranging device and the surface of the non-indentation area of the impacted surface of the automotive power battery is obtained through the laser ranging device;

[0042] The position with the maximum gray value within the indentation contour of the impact indentation in the post-collision image is obtained, and its actual position on the impacted surface of the automotive power battery is used as the second ranging point;

[0043] After aligning the ranging beam of the laser ranging device with the second ranging point, the distance h2 between the laser ranging device and the second ranging point is obtained;

[0044] The indentation depth of the impact indentation is the value of h2 minus h1.

[0045] Compared with the prior art, the beneficial effects of the present invention are as follows:

[0046] The present invention analyzes the position and indentation depth of the impact indentation of the impacted automotive power battery through the deformation analysis module, and then obtains the battery capacity after the impacted automotive power battery undergoes a preset number of charge and discharge cycles through the life test module. It judges whether the position and indentation depth of the impact indentation affect the battery safety. Finally, a pre-trained deep learning model is used to estimate the minimum indentation depth affecting the safety of the automotive power battery at different positions of the automotive power battery, accurately providing a safety judgment after the battery is impacted and deformed. In the indentation recognition module, through the pre-trained improved YOLOv8 model, the indentation area in the impacted surface image is recognized, improving the accuracy of indentation recognition, and further improving the effectiveness and accuracy of the test. Description of the Drawings

[0047] The accompanying drawings here are incorporated into the specification and form a part of this specification, showing the embodiments in line with the present invention, and are used together with the specification to explain the principles of the present invention.

[0048] In order to more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the following will briefly introduce the accompanying drawings required for use in the description of the embodiments or the prior art. Obviously, for those of ordinary skill in the art, without creative efforts, other drawings can also be obtained based on these drawings.

[0049] Figure 1 It is a structural block diagram of a vehicle power battery collision test system of the present invention;

[0050] Figure 2 It is a structural block diagram of a deformation analysis module of a vehicle power battery collision test system of the invention. Specific embodiments

[0051] The following will clearly and completely describe the technical solutions in the embodiments of the present invention in conjunction with the accompanying drawings in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, rather than all of them. Based on the embodiments of the present invention, all other embodiments obtained by those of ordinary skill in the art without creative efforts fall within the scope of protection of the present invention.

[0052] It should be noted that all directional indications (such as up, down, left, right, front, back...) in the embodiments of the present invention are only used to explain the relative position relationship and movement conditions between components in a specific posture (as shown in the accompanying drawings). If this specific posture changes, the directional indications will also change accordingly.

[0053] In addition, the descriptions involving "first", "second", etc. in the present invention are only for descriptive purposes and cannot be understood as indicating or implying their relative importance or implicitly indicating the quantity of the indicated technical features. Thus, the features defined with "first" and "second" may explicitly or implicitly include at least one such feature. In addition, the technical solutions between various embodiments can be combined with each other, but it must be based on the fact that those of ordinary skill in the art can implement them. When the combination of technical solutions is contradictory or cannot be implemented, it should be considered that such a combination of technical solutions does not exist and is not within the scope of protection required by the present invention.

[0054] Embodiment 1

[0055] Refer to Figure 1 As shown, the present invention provides a vehicle power battery collision test system, specifically including:

[0056] An impact device, a deformation analysis module, a life test module, a judgment module, a prediction module, and a control center for communicating with the above modules.

[0057] An impact device for impacting an automotive power battery through an impact head.

[0058] The impact device can adopt impact heads of different shapes.

[0059] A deformation analysis module for analyzing the deformation state of the automotive power battery after being impacted.

[0060] Refer to Figure 2 As shown, the deformation analysis module includes: an image acquisition module, a dent recognition module, an impact dent determination module, and a laser ranging device.

[0061] An image acquisition module for acquiring an image of the impacted surface of the automotive power battery, where the image of the impacted surface includes an initial image and a post-impact image.

[0062] The initial image is an image of the impacted surface of the automotive power battery before being impacted, and the post-impact image is an image of the impacted surface of the automotive power battery after being impacted.

[0063] The image acquisition module includes a photographing device and a light source device.

[0064] It should be noted that when acquiring the initial image and the post-impact image, the shooting height and angle of the photographing device remain unchanged, and the lighting height, angle, and brightness of the light source device remain unchanged, and both the photographing device and the light source device are used to shoot and light the impacted surface of the automotive power battery directly.

[0065] The image acquisition module further includes an image preprocessing unit for preprocessing the initial image and the post-impact image of the automotive power battery, including but not limited to:

[0066] Grayscale processing, converting a color image into a grayscale image;

[0067] Filtering and denoising processing, removing noise points in the image to improve the image quality, and specifically, median filtering processing can be selected;

[0068] Histogram equalization processing, performing histogram equalization on the image after filtering and denoising processing to enhance the contrast of the image and make the dent more obvious.

[0069] A dent recognition module for identifying dent information in the image of the impacted surface.

[0070] Identifying the dent information in the image of the impacted surface specifically means:

[0071] Identify the sunken area in the collided surface image through a pre-trained improved YOLOv8 model;

[0072] Use edge detection to identify the sunken contour from the sunken area.

[0073] It should be noted that identifying the sunken information in the collided surface image is to identify the sunken information in the initial image and the post-collision image respectively.

[0074] The improved YOLOv8 model is improved based on the YOLOv8 model. Specifically:

[0075] Replace the ordinary convolution module in the backbone network of the YOLOv8 model with a Ghost convolution module;

[0076] Introduce the CA attention mechanism into the C2f module in the feature enhancement network of the YOLOv8 model.

[0077] The YOLOv8 model is mainly composed of three parts: the backbone network, the feature enhancement network, and the detection head.

[0078] Among them, the backbone network contains multiple Conv modules, that is, ordinary convolution modules. In the operation of ordinary convolution, each convolution kernel will independently generate a new feature map. There may be a large similarity, that is, redundancy, between these feature maps. And the parameter quantity and calculation amount of ordinary convolution are large, resulting in a high model complexity and insufficient running efficiency, which in turn leads to low detection efficiency. Therefore, in this solution, the ordinary convolution module in the backbone network of the YOLOv8 model is replaced with a Ghost convolution module. The Ghost convolution module first uses a small convolution kernel (such as 1×1 convolution) to perform convolution operation on the input image to generate a part of intermediate feature maps, and then uses depthwise separable convolution on the intermediate feature maps to generate Ghost feature maps. The intermediate feature maps and Ghost feature maps are concatenated to obtain the final output feature map. The Ghost convolution only needs a small number of ordinary convolution kernels and depthwise separable convolution kernels to generate a large number of feature maps, thereby reducing the parameter quantity and calculation amount.

[0079] The C2f module (CSP Bottleneck with 2Convo l ut ions) in the feature enhancement network is aimed at better aggregating multi-scale information while reducing the calculation amount and memory consumption. It includes a first convolution layer (Conv1), a feature map splitting layer (Sp l it), multiple bottleneck modules (Bott leneck), a concatenation layer (Concat), and a second convolution layer (Conv2)

[0080] The first convolution layer performs convolution transformation on the feature map output by the backbone network to generate a second intermediate feature map;

[0081] The feature map splitting layer is used to split the second intermediate feature map into two parts, one part is directly passed to the splicing layer, and the other part is passed to the bottleneck module;

[0082] The splicing layer splices the feature map processed by the bottleneck module with the intermediate feature map directly passed to the splicing layer to form a fused feature map;

[0083] The second convolutional layer performs convolutional transformation processing on the fused feature map.

[0084] Ordinary C2f modules mainly rely on local convolutional operations to extract features, which may cause the model to be unable to fully capture the global information in the feature map and the accuracy of distinguishing key features is insufficient. Therefore, this solution introduces the CA attention mechanism into the C2f module in the feature enhancement network of the YOLOv8 model, specifically as follows:

[0085] The feature map F output by the splicing layer of the C2f module is input into the CA self-attention mechanism module. The CA self-attention mechanism module encodes the feature map F in the horizontal and vertical directions to generate a horizontal direction attention map and a vertical direction attention map, and then fuses them to obtain a comprehensive attention map. The comprehensive attention map is used to weight the feature map F to obtain an enhanced feature map, and the enhanced feature map is input into the second convolutional layer of the C2f module.

[0086] The present invention introduces the CA attention mechanism into the C2f module, uses the comprehensive attention map to weight the feature map, and adjusts the weight of the feature map in this way, so as to better capture the global spatial relationship, thereby enhancing the feature extraction ability of the model and improving the accuracy of distinguishing key features.

[0087] The improved YOLOv8 model is trained with a large number of automotive power battery images with features of different concave regions.

[0088] The impact depression determination module is used to determine the impact depression in the post-collision image that belongs to the automotive power battery after being hit by the impact head by comparing the depression information in the initial image and the post-collision image. Specifically: compare the depression regions of the initial image and the post-collision image, and use the depression region that exists in the post-collision image but does not exist in the initial image as the impact depression.

[0089] The laser ranging device is used to measure the depression depth of the impact depression.

[0090] The measurement of the depression depth of the impact depression is specifically as follows:

[0091] Obtain the distance h1 between the laser ranging device and the surface of the non-depressed area of the hit surface of the automotive power battery through the laser ranging device;

[0092] Obtain the position with the maximum gray value within the depression contour of the impact depression in the post-impact image, and use its actual position on the impacted surface of the automotive power battery as the second ranging point;

[0093] After aligning the ranging beam of the laser ranging device with the second ranging point, obtain the distance h2 between the laser ranging device and the second ranging point;

[0094] The depression depth of the impact depression is the value obtained by subtracting h1 from h2.

[0095] The life test module is used to obtain the battery capacity after performing a preset number of charge and discharge cycles on the impacted automotive power battery.

[0096] The judgment module is used to judge whether the depression depth at the position of the impact depression of the current automotive power battery affects the safety of the automotive power battery according to the battery capacity, specifically:

[0097] If the battery capacity is less than or equal to the preset capacity threshold, the depression depth at the position of the impact depression of the current automotive power battery affects the safety of the automotive power battery.

[0098] The prediction module is used to estimate the minimum depression depth affecting the safety of the automotive power battery at different positions of the automotive power battery by using a pre-trained deep learning model

[0099] The deep learning model is trained with a large number of different positions, corresponding different depression depths, corresponding battery capacities obtained from the life test module, and corresponding judgment results obtained from the judgment module as training data.

[0100] Obtain different depression depths by adjusting the impact speed of the impact device.

[0101] In some embodiments, the system further includes a parameter acquisition module for acquiring the mass, initial velocity, and impact instantaneous velocity of the impact head.

[0102] The mass of the impact head is obtained in advance through a weight sensor; the initial velocity is set and adjusted manually; the impact instantaneous velocity is obtained through a velocity sensor.

[0103] The judgment module is further used to judge whether the collision of an object with the same mass as the current impact head with the automotive power battery at the same impact instantaneous velocity affects the safety of the automotive power battery according to the battery capacity, specifically:

[0104] If the battery capacity is less than or equal to the preset capacity threshold, the collision of an object with the same mass as the current impact head with the automotive power battery at the same impact instantaneous velocity will affect the safety of the automotive power battery.

[0105] Embodiment 2

[0106] The present invention also provides a collision test device for an automotive power battery, specifically including:

[0107] An impact device for impacting the automotive power battery through an impact head;

[0108] An image acquisition device for acquiring images of the impacted surface of the automotive power battery, where the impacted surface images include initial images and post-impact images;

[0109] A life test device for obtaining the battery capacity after performing a preset number of charge and discharge cycles on the impacted automotive power battery;

[0110] A laser ranging device for measuring the depth of the impact depression;

[0111] An analysis platform provided with a depression recognition algorithm, an impact depression determination algorithm, and a judgment algorithm;

[0112] The depression recognition algorithm is used to identify the depression area in the impacted surface image through a pre-trained improved YOLOv8 model, and use edge detection to identify the depression contour from the depression area;

[0113] The improved YOLOv8 model is improved based on the YOLOv8 model. Specifically, the ordinary convolution module in the backbone network of the YOLOv8 model is replaced with a Ghost convolution module; a CA attention mechanism is introduced into the C2f module in the feature enhancement network of the YOLOv8 model;

[0114] The impact depression is determined by the impact depression determination algorithm. Specifically, by comparing the depression information in the initial image and the post-impact image, the impact depression obtained after the automotive power battery is impacted by the impact head in the post-impact image is determined;

[0115] The judgment algorithm is used to judge whether the depth of the depression at the position of the impact depression of the current automotive power battery affects the safety of the automotive power battery according to the battery capacity.

[0116] The image acquisition device includes a shooting device and a light source device.

[0117] When acquiring the initial image and the post-impact image, the shooting height and angle of the shooting device remain unchanged, and the lighting height, angle, and brightness of the light source device remain unchanged, and both the shooting device and the light source device are shooting and lighting directly at the impacted surface of the automotive power battery.

[0118] The analysis platform is also equipped with an image pre - processing algorithm for pre - processing the initial image and the post - collision image of the automotive power battery, including but not limited to: grayscale processing, converting a color image into a grayscale image; filtering and denoising processing, removing noise points in the image to improve image quality, and specifically median filtering can be selected; histogram equalization processing, performing histogram equalization on the image after filtering and denoising to enhance the contrast of the image and make the depression more obvious.

[0119] Introduce the CA attention mechanism in the C2f module of the feature enhancement network of the YOLOv8 model, specifically:

[0120] Input the feature map F output by the splicing layer of the C2f module into the CA self - attention mechanism module. The CA self - attention mechanism module encodes the feature map F in the horizontal and vertical directions to generate a horizontal - direction attention map and a vertical - direction attention map, and then fuses them to obtain a comprehensive attention map. Use the comprehensive attention map to weight the feature map F to obtain an enhanced feature map, and input the enhanced feature map into the second convolutional layer of the C2f module.

[0121] In some embodiments, the automotive power battery collision test device further includes a weight sensor for obtaining the mass of the impact head and a speed sensor for obtaining the instantaneous impact speed of the impact head.

[0122] The judgment algorithm is also used to judge whether an object with the same mass as the current impact head colliding with the automotive power battery at the same instantaneous impact speed will affect the safety of the automotive power battery. Specifically:

[0123] If the battery capacity is less than or equal to the preset capacity threshold, then an object with the same mass as the current impact head colliding with the automotive power battery at the same instantaneous impact speed will affect the safety of the automotive power battery.

[0124] In some embodiments, the analysis platform is also equipped with a prediction algorithm for estimating the minimum estimated depression depth that affects the safety of the automotive power battery at different positions in the automotive power battery using a pre - trained deep - learning model.

[0125] Embodiment Three

[0126] The present invention also provides an electronic device, including: a processor, a sending device, an input device, an output device, and a memory. The processor can be implemented in ways such as a general-purpose CPU (Central Processing Unit), a microprocessor, an application-specific integrated circuit, or one or more integrated circuits, etc., and is used to execute relevant programs to implement the technical solutions provided in the embodiments of the present application. The memory can be implemented in forms such as a read-only memory (ROM), a static storage device, a dynamic storage device, or a random access memory (RAM), etc., and is used to store computer program code. The computer program code includes computer instructions. When the processor executes the computer instructions, the electronic device executes the method in any of the above possible implementation manners.

[0127] Embodiment 4

[0128] The present invention also provides a computer-readable storage medium. A computer program is stored in the computer-readable storage medium. The computer program includes program instructions. When the program instructions are executed by the processor of the electronic device, the processor is caused to execute the method in any of the above possible implementation manners.

[0129] The beneficial effects of the present invention are as follows:

[0130] The present invention analyzes the position and depth of the impact depression of the automotive power battery after being impacted through a deformation analysis module, and then obtains the battery capacity of the automotive power battery after being impacted after a preset number of charge and discharge cycles through a life test module. It determines whether the position and depth of the impact depression affect battery safety based on the battery capacity. Finally, it estimates the minimum depression depth affecting the safety of the automotive power battery at different positions of the automotive power battery using a pre-trained deep learning model, accurately providing a safety judgment for the deformed battery after being impacted. In the depression recognition module, through a pre-trained improved YOLOv8 model, it identifies the depression area in the image of the impacted surface, improving the accuracy of depression recognition, and further improving the effectiveness and accuracy of the test.

[0131] In the description of the specification, the description referring to terms such as "one embodiment", "example", "specific example", etc. means that the specific features, structures, materials, or characteristics described in connection with the embodiment or example are included in at least one embodiment or example of the present invention. In this specification, the schematic representations of the above terms do not necessarily refer to the same embodiment or example. Moreover, the specific features, structures, materials, or characteristics described can be combined in a suitable manner in any one or more embodiments or examples.

[0132] In addition, in each embodiment of the present application, each functional unit can be integrated into one processing unit, or each unit can exist physically alone, or two or more units can be integrated into one unit. The above-mentioned integrated unit can be implemented in the form of hardware or in the form of a software functional unit. If the integrated unit is implemented in the form of a software functional unit and sold or used as an independent product, it can be stored in a computer-readable storage medium. Based on such an understanding, the technical solution of the present application, in essence, or the part that contributes to the prior art, or all or part of the technical solution, can be embodied in the form of a software product. The computer software product is stored in a storage medium and includes multiple instructions for causing a computer device (which can be a personal computer, a server, or a network device, etc.) to execute all or part of the steps of the methods in each embodiment of the present application. The foregoing storage medium includes: various media that can store programs such as USB flash drives, mobile hard disks, read-only memories (ROM), random access memories (RAM), magnetic disks, or optical discs.

[0133] The above are only specific embodiments of the present invention, enabling those skilled in the art to understand or implement the present invention. Various modifications to these embodiments will be obvious to those skilled in the art. The general principles defined herein can be implemented in other embodiments without departing from the spirit or scope of the present invention. Therefore, the present invention will not be limited to these embodiments shown herein, but will conform to the widest scope consistent with the principles and novel features claimed herein.

Claims

1. An automotive power battery collision test system, characterized in that, Specifically, it includes: An impact device for impacting an automotive power battery through an impact head; A deformation analysis module, including an image acquisition module, a dent recognition module, an impact dent determination module, and a laser ranging device; The image acquisition module is used to acquire the image of the impacted surface of the automotive power battery, and the image of the impacted surface includes an initial image and a post-impact image; The dent recognition module is used to identify dent information in the image of the impacted surface. Specifically, it identifies the dent area in the image of the impacted surface through a pre-trained improved YOLOv8 model, and uses edge detection to identify the dent contour from the dent area; The improved YOLOv8 model is improved based on the YOLOv8 model. Specifically, the ordinary convolution module in the backbone network of the YOLOv8 model is replaced with a Ghost convolution module; a CA attention mechanism is introduced into the C2f module in the feature enhancement network of the YOLOv8 model. Specifically: The feature map output by the splicing layer of the C2f module is input into the CA self-attention mechanism module. The CA self-attention mechanism module encodes the feature map in the horizontal and vertical directions to generate a horizontal direction attention map and a vertical direction attention map, and then fuses them to obtain a comprehensive attention map. The comprehensive attention map is used to weight the feature map to obtain an enhanced feature map, and the enhanced feature map is input into the second convolutional layer of the C2f module; The impact dent determination module is used to determine the impact dent obtained after the automotive power battery is impacted by the impact head in the post-impact image by comparing the dent information in the initial image and the post-impact image; The laser ranging device is used to measure the dent depth of the impact dent; A life test module for obtaining the battery capacity after performing a preset number of charge and discharge cycles on the impacted automotive power battery; A judgment module for judging whether the dent depth at the position of the impact dent of the current automotive power battery affects the safety of the automotive power battery according to the battery capacity; A prediction module for estimating the minimum dent depth affecting the safety of the automotive power battery at different positions of the automotive power battery using a pre-trained deep learning model.

2. The automotive power battery collision test system according to claim 1, wherein The measurement of the dent depth of the impact dent is specifically: Obtaining the distance h1 between the laser ranging device and the surface of the non-dent area of the impacted surface of the automotive power battery through the laser ranging device; Obtaining the position with the maximum gray value within the dent contour of the impact dent in the post-impact image and taking its actual position on the impacted surface of the automotive power battery as the second ranging point; after aligning the ranging beam of the laser ranging device with the second ranging point, obtaining the distance h2 between the laser ranging device and the second ranging point; The dent depth of the impact dent is the value of h2 minus h1.

3. The automotive power battery collision test system according to claim 1, characterized in that, The judgment of whether the dent depth at the position of the impact dent of the current automotive power battery affects the safety of the automotive power battery is specifically: If the battery capacity is less than or equal to the preset capacity threshold, the dent depth at the position of the impact dent of the current automotive power battery affects the safety of the automotive power battery.

4. The automotive power battery collision test system according to claim 1, wherein The system further includes a parameter acquisition module for acquiring the mass, initial velocity, and impact instantaneous velocity of the impact head.

5. The automotive power battery collision test system according to claim 4, wherein, The judgment module is further configured to determine whether an object with the same impact head mass as the current one colliding with the automotive power battery at the same impact instantaneous speed will affect the safety of the automotive power battery according to the battery capacity, specifically: If the battery capacity is less than or equal to the preset capacity threshold, an object with the same impact head mass as the current one colliding with the automotive power battery at the same impact instantaneous speed will affect the safety of the automotive power battery.

6. The automotive power battery collision test system according to claim 1, characterized in that, The image acquisition module further includes an image preprocessing unit for preprocessing the initial image and the post-collision image of the automotive power battery.

7. An automobile power battery collision test device, which applies an automobile power battery collision test system according to any one of claims 1 to 6, and is characterized in that Including: An impact device for impacting the automotive power battery through an impact head; An image acquisition device for acquiring the image of the impacted surface of the automotive power battery, where the image of the impacted surface includes an initial image and a post-collision image; A life test device for obtaining the battery capacity after performing a preset number of charge and discharge cycles on the impacted automotive power battery; A laser ranging device for measuring the depression depth of the impact depression; An analysis platform equipped with a depression recognition algorithm, an impact depression determination algorithm, a judgment algorithm, and a prediction algorithm; The depression recognition algorithm is used to identify the depression area in the image of the impacted surface through a pre-trained improved YOLOv8 model, and use edge detection to identify the depression contour from the depression area; The improved YOLOv8 model is improved based on the YOLOv8 model. Specifically: replacing the ordinary convolution module in the backbone network of the YOLOv8 model with a Ghost convolution module; introducing a CA attention mechanism into the C2f module in the feature enhancement network of the YOLOv8 model. Specifically: Input the feature map output by the splicing layer of the C2f module into the CA self-attention mechanism module. The CA self-attention mechanism module encodes the feature map in the horizontal and vertical directions to generate a horizontal direction attention map and a vertical direction attention map, and then fuses them to obtain a comprehensive attention map. Use the comprehensive attention map to weight the feature map to obtain an enhanced feature map, and input the enhanced feature map into the second convolution layer of the C2f module; The impact depression is determined by the impact depression determination algorithm. Specifically, by comparing the depression information in the initial image and the post-collision image, the impact depression obtained after the automotive power battery is impacted by the impact head in the post-collision image is determined; The judgment algorithm is used to determine whether the depression depth at the position of the impact depression of the current automotive power battery affects the safety of the automotive power battery according to the battery capacity; The prediction algorithm uses a pre-trained deep learning model to estimate the minimum depression depth that affects the safety of the automotive power battery at different positions of the automotive power battery.

8. The automotive power battery collision test device according to claim 7, characterized in that, The measurement of the depression depth of the impact depression is specifically: Obtaining the distance h1 between the laser ranging device and the surface of the non-depression area of the impacted surface of the automotive power battery through the laser ranging device; Obtaining the position with the maximum gray value within the depression contour of the impact depression in the post-collision image and taking its actual position on the impacted surface of the automotive power battery as the second ranging point; After aligning the ranging beam of the laser ranging device with the second ranging point, obtaining the distance h2 between the laser ranging device and the second ranging point; The depth of the impact depression is the value of h2 minus h1.

Images