A method and apparatus for encapsulation testing of high energy fuel cells

Abstract

The application discloses a kind of high-energy fuel cell packaging detection method and device, high-energy fuel cell includes jar body, spiral electrode tube and gas cylinder being arranged in jar body interior and the ceramic column of jar body bottom, method includes: obtaining and checking the quality inspection information of high-energy fuel cell;According to the quality inspection information of high-energy fuel cell, the image to be detected of high-energy fuel cell is extracted;Combining the image to be detected, the positional relationship of spiral electrode tube and gas cylinder, the quantity of ceramic column is analyzed, and first detection result is given;According to first detection result, the position of spiral electrode tube is optimized, the packaging of high-energy fuel cell is controlled, and packaging detection data is given;Combining the three-dimensional model of fuel cell constructed in advance, packaging detection data is analyzed, packaging detection result is given, and the packaging detection of high-energy fuel cell is completed.The quality of fuel cell before and after packaging is monitored and analyzed, and the use reliability and stability of fuel cell are guaranteed.

Description

Technical Field

[0001] This invention belongs to the technical field of battery packaging and testing, specifically relating to a packaging and testing method and apparatus for high-energy fuel cells. Background Technology

[0002] High-energy fuel cells have attracted widespread attention due to their advantages such as high energy density and environmental friendliness, which suggest broad application prospects in many fields. However, the packaging quality of fuel cells directly affects their performance, safety, and lifespan.

[0003] Currently, existing fuel cell packaging and testing technologies have many shortcomings. On the one hand, the testing accuracy is low, and the monitoring data is not comprehensive or accurate enough, making it difficult to achieve all-round and refined control of the entire packaging process. On the other hand, the lack of effective coordination and feedback mechanisms between the various modules of the testing system makes it difficult to achieve real-time and precise control of the packaging process, thus affecting the overall quality and production efficiency of the fuel cell. This can lead to problems such as poor sealing and damage to internal structural components in the packaged fuel cell, making it impossible to effectively control the quality of the fuel cell and thus affecting its reliability and stability in use.

[0004] Patent CN116899885A discloses a testing method for an unmanned full-inspection equipment in a lithium battery packaging production line, comprising: Step 1, the feeding mechanism transfers the lithium battery packs from the feeding conveyor belt to the positioning component on the testing platform; Step 2, the turntable component drives the positioning component to the first phase, where a first vision monitoring module and a second vision monitoring module acquire image information data of the first phase and upload it to a data analysis unit; Step 3, the turntable component drives the positioning component to the second phase, where the first and second vision monitoring modules acquire image information data of the second phase and upload it to the data analysis unit; Step 4, the data analysis unit analyzes the image information of the first and second phases and calculates the packaging structure data; Step 5, the turntable component drives the positioning component to the bottom of the unloading mechanism, where the unloading mechanism classifies and transports the products according to the test results, transferring qualified products to the unloading conveyor belt and unqualified products to the defective product conveyor belt. By using the first and second vision monitoring modules to collect image information and compare it before and after, accurate test data is obtained, improving the accuracy of the analysis and judgment of the test results. The analysis of image information requires complex algorithms. When the algorithm cannot support the complexity of the image information, the accuracy of the detection results cannot be guaranteed.

[0005] Monitoring and analyzing the quality of fuel cells to ensure their reliability and stability is a problem that needs to be solved. Summary of the Invention

[0006] To address the shortcomings of existing technologies, this invention provides a packaging and testing method and apparatus for high-energy fuel cells. The high-energy fuel cell includes a tank, a spiral electrode tube and a venting cylinder disposed inside the tank, and ceramic pillars at the bottom of the tank. The method includes: acquiring and verifying the quality inspection information of the high-energy fuel cell; extracting an image of the high-energy fuel cell to be tested based on the quality inspection information; analyzing the positional relationship between the spiral electrode tube and the venting cylinder, and the number of ceramic pillars, based on the image to be tested, and providing a first test result; optimizing the position of the spiral electrode tube based on the first test result, controlling the packaging of the high-energy fuel cell, and providing packaging test data; and analyzing the packaging test data based on a pre-constructed three-dimensional model of the fuel cell, providing a packaging test result, and completing the packaging test of the high-energy fuel cell. By monitoring and analyzing the quality of the fuel cell before and after packaging, the reliability and stability of the fuel cell are ensured.

[0007] In a first aspect, the present invention provides a packaging and testing method for a high-energy fuel cell. The high-energy fuel cell includes a tank, a spiral electrode tube disposed inside the tank, a venting cylinder, and a ceramic column at the bottom of the tank. Specifically, the method includes the following steps:

[0008] Obtain and verify the quality inspection information of high-energy fuel cells;

[0009] Based on the quality inspection information of high-energy fuel cells, extract the images of high-energy fuel cells to be inspected;

[0010] Based on the image to be tested, the positional relationship between the spiral electrode tube and the vent cylinder, as well as the number of ceramic columns, are analyzed to give the first test result;

[0011] Based on the first test results, the position of the spiral electrode tube is optimized, the packaging of the high-energy fuel cell is controlled, and the packaging test data is provided.

[0012] By combining a pre-constructed 3D model of the fuel cell, the packaging and testing data are analyzed, the packaging and testing results are given, and the packaging and testing of the high-energy fuel cell is completed.

[0013] Furthermore, by combining the image to be tested, the positional relationship between the spiral electrode tube and the venting cylinder, and the number of ceramic columns are analyzed to provide the first test results, specifically including:

[0014] The image to be detected is filtered, and the initial contour edges of the spiral electrode tube, the vent cylinder and the ceramic column in the image to be detected are extracted respectively.

[0015] The initial contour edge is traced, and the broken parts of the initial contour edge are repaired to obtain the target contour edge;

[0016] The target contour edge is compared and analyzed with a pre-defined standard contour to provide contour similarity data;

[0017] Based on a pre-set similarity threshold, the contour similarity data is judged, and the first detection result is given.

[0018] Furthermore, the image to be detected is filtered to extract the initial contour edges of the spiral electrode tube, the vent cylinder, and the ceramic column, respectively. Specifically, this includes:

[0019] The image to be detected is processed into grayscale to obtain a grayscale detection image;

[0020] The grayscale detection image is filtered to remove noise and obtain a standard detection image.

[0021] Analyze the gradients of each pixel in the standard detection image in the horizontal and vertical directions to obtain the horizontal gradient and vertical gradient;

[0022] The horizontal and vertical gradients of each pixel are fused to obtain the edge intensity and edge direction of each pixel;

[0023] Based on the edge intensity and edge direction of each pixel, edge pixels are selected from the standard detection image to obtain the initial contour edges of the spiral electrode tube, the vent cylinder, and the ceramic column.

[0024] Furthermore, the initial contour edge includes a first terminal portion, a second terminal portion, a spiral coil portion, and a breather cylinder portion;

[0025] The initial contour edge is traced, and broken parts of the initial contour edge are repaired to obtain the target contour edge, specifically including:

[0026] The edge pixel at the end of the first terminal portion that is far from the spiral coil portion is taken as the starting point of the first terminal tracking, and the first terminal tracking direction is set to vertically downward.

[0027] The edge pixel at the end of the second terminal portion that is far from the spiral coil portion is taken as the starting point of the second terminal tracking, and the tracking direction of the second terminal is set to vertically downward.

[0028] Based on the first terminal tracking start point of the first terminal portion and the second terminal tracking start point of the second terminal portion, and combined with the preset first terminal tracking direction and second terminal tracking direction, the edge pixels in the first terminal portion and the second terminal portion are traversed to give the broken parts of the first terminal portion and the second terminal portion.

[0029] The connection point between the first terminal and the spiral coil or the connection point between the second terminal and the spiral coil is taken as the starting point of spiral tracking.

[0030] Based on the spiral tracking starting point of the spiral coil section, and combined with the preset spiral tracking direction, the edge pixels in the spiral coil section are traversed and the spiral tracking direction is updated to give the broken part of the spiral coil section. The spiral tracking direction is determined based on the shape of the spiral coil section and the current spiral tracking starting point.

[0031] Based on the starting point of the ventilator cylinder tracking, combined with the preset ventilator cylinder tracking direction, the edge pixels in the ventilator cylinder are traversed to give the broken part of the ventilator cylinder.

[0032] The traversal process ends by combining the closed-loop analysis of the tracking directions of the first terminal, the second terminal, the spiral tracking direction, and the ventilator tracking direction.

[0033] Based on the length of the fractured portion, a corresponding repair strategy is matched to repair the fractured portion and obtain the target contour edge.

[0034] Furthermore, based on the length of the fractured portion, a corresponding repair strategy is matched to repair the fractured portion, resulting in the target contour edge, specifically including:

[0035] If the length of the broken part is less than the preset breaking threshold, the pixel values ​​of the corresponding pixels are filled according to the breaking start point and breaking end point of the broken part to obtain the target contour edge.

[0036] If the length of the fractured portion is greater than or equal to the fracture threshold, the fractured portion is filled based on the fracture line direction to obtain the target contour edge.

[0037] Furthermore, based on the fracture line direction of the fractured portion, the fractured portion is filled to obtain the target contour edge, specifically including:

[0038] Obtain the fracture start point, fracture end point, and corresponding start and end adjacent points of the fractured part;

[0039] Based on the initial adjacent points, analyze the direction of the initial straight line from the fracture initiation point to the fracture termination point;

[0040] Based on the adjacent points, analyze the direction of the end line line near the fracture initiation point of the fracture termination point.

[0041] By merging the starting and ending straight line directions, the direction of the break line is obtained;

[0042] Based on the direction of the fracture line, calculate the positions of adjacent points sequentially from the fracture start point or fracture end point, and fill the adjacent points to obtain the target contour edge.

[0043] Furthermore, the target contour edge is compared and analyzed with a pre-defined standard contour to provide contour similarity data, specifically including:

[0044] Based on the pixel values ​​of each edge pixel on the target contour edge, the mean value of the target edge is given;

[0045] The average value of the standard edge is given based on the pixel values ​​of each edge pixel on the standard contour.

[0046] By combining the target edge mean and the standard edge mean, the edge covariance of the target contour edge and the standard contour is given;

[0047] Based on the standard deviations of the target contour edge and the standard contour, the edge covariance is fused to provide contour similarity data.

[0048] Furthermore, based on the first test results, the position of the spiral electrode tube is optimized, the packaging of the high-energy fuel cell is controlled, and packaging test data is provided, specifically including:

[0049] Based on the first detection result, the position of the spiral electrode tube is determined, and a control command is output to position and limit the spiral electrode tube.

[0050] Based on the positional relationship between the sealing holes and the spiral electrode tubes of the high-energy fuel cell, the high-energy fuel cell is encapsulated.

[0051] The packaging and testing data of the high-energy fuel cell after packaging are collected. The packaging and testing data includes packaging and testing images and packaging performance data.

[0052] Furthermore, by combining the pre-constructed 3D model of the fuel cell, the packaging and testing data are analyzed, and the packaging and testing results are given, thus completing the packaging and testing of the high-energy fuel cell. Specifically, this includes:

[0053] Based on the package inspection image, perform stereo matching and depth calculation to construct a 3D model of the package;

[0054] Based on the encapsulated 3D model, the surface deformation of each part of the fuel cell in the encapsulated 3D model is analyzed, and the surface curvature of each part is given.

[0055] Analyze the differences between the surface curvature of each part and the surface curvature of the corresponding part of the fuel cell 3D model, and provide curvature difference data;

[0056] Determine the relationship between the curvature difference data and the pre-acquired standard curvature difference range, provide the packaging test results, and complete the packaging test of the high-energy fuel cell.

[0057] Secondly, the present invention also provides a packaging and testing apparatus for a high-energy fuel cell, employing the packaging and testing method for a high-energy fuel cell as described in any one of the above claims, comprising:

[0058] The data acquisition module is used to acquire and verify the quality inspection information of the high-energy fuel cell, which includes a tank, a spiral electrode tube and a venting cylinder installed inside the tank, and a ceramic column at the bottom of the tank.

[0059] The image extraction module is used to extract the image of the high-energy fuel cell to be inspected based on the quality inspection information of the high-energy fuel cell;

[0060] The image detection module is used to combine the image to be detected, analyze the positional relationship between the spiral electrode tube and the vent cylinder, the number of ceramic columns, and give the first detection result.

[0061] The packaging and testing module is used to optimize the position of the spiral electrode tube, control the packaging of the high-energy fuel cell, and provide packaging and testing data based on the first test result.

[0062] The encapsulation analysis module is used to analyze encapsulation test data by combining a pre-built 3D model of the fuel cell, provide encapsulation test results, and complete the encapsulation test of the high-energy fuel cell.

[0063] The present invention provides a packaging and testing method and apparatus for high-energy fuel cells, which has at least the following beneficial effects:

[0064] (1) By extracting the image of the high-energy fuel cell to be tested, analyzing the positional relationship between the spiral electrode tube and the venting cylinder and the number of ceramic columns, the first test result is given and the positional relationship of the spiral electrode tube is adjusted to control the packaging of the high-energy fuel cell. Then, the packaging test data is given and analyzed in combination with the pre-constructed three-dimensional model of the fuel cell. The packaging test of the high-energy fuel cell is completed, and the quality before and after the fuel cell is packaged is monitored and analyzed, ensuring the reliability and stability of the fuel cell.

[0065] (2) By automating the detection and analysis of the fuel cell packaging process, the packaging quality and efficiency of high-energy fuel cells are improved, the safety of the production environment is ensured, and the production cost is reduced.

[0066] (3) By performing grayscale processing on the image to be detected, the RGB image is converted into a grayscale image. This reduces the amount of data in the grayscale detection image while retaining the main structural information of the image to be detected, providing a data basis for subsequent analysis of the grayscale detection image and improving the efficiency of image analysis. Attached Figure Description

[0067] Figure 1 A flowchart illustrating the packaging and testing method for a high-energy fuel cell provided in an embodiment of the present invention;

[0068] Figure 2 This is a front view schematic diagram of a high-energy fuel cell provided in an embodiment of the present invention;

[0069] Figure 3 A flowchart providing the first detection result is provided for an embodiment of the present invention;

[0070] Figure 4 A flowchart for extracting the initial contour edges provided in an embodiment of the present invention;

[0071] Figure 5 A flowchart for obtaining the target contour edge is provided in an embodiment of the present invention;

[0072] Figure 6 This is a flowchart of a comparative analysis of the target contour edges provided in an embodiment of the present invention;

[0073] Figure 7 A flowchart providing packaging and testing data is provided for embodiments of the present invention;

[0074] Figure 8 A flowchart for determining the packaging test results provided in an embodiment of the present invention;

[0075] Figure 9 This is a structural block diagram of a high-energy fuel cell packaging and testing device provided in an embodiment of the present invention.

[0076] Among them, 1. Tank body; 2. Spiral electrode tube; 21. First terminal part; 22. Spiral coil part; 23. Second terminal part; 3. Vent cylinder; 201. Data acquisition module; 202. Image extraction module; 203. Image detection module; 204. Packaging detection module; 205. Packaging analysis module. Detailed Implementation

[0077] To better understand the above technical solutions, a detailed description of the solutions will be provided below in conjunction with the accompanying drawings and specific embodiments. Obviously, the described embodiments are merely some, not all, of the embodiments of the present invention. All other embodiments obtained by those skilled in the art based on the embodiments of the present invention without creative effort are within the scope of protection of the present invention.

[0078] The terminology used in the embodiments of this invention is for the purpose of describing particular embodiments only and is not intended to limit the invention. The singular forms “a,” “the,” and “the” as used in the embodiments of this invention and the appended claims are also intended to include the plural forms, and “multiple” generally includes at least two unless the context clearly indicates otherwise.

[0079] It should also be noted that the terms "comprising," "including," or any other variations thereof are intended to cover non-exclusive inclusion, such that an article or device that comprises a list of elements includes not only those elements but also other elements not expressly listed, or elements inherent to such an article or device. Without further limitation, an element defined by the phrase "comprising one..." does not exclude the presence of other identical elements in the article or device that includes said element.

[0080] A fuel cell is a chemical device that directly converts the chemical energy of fuel into electrical energy. Because fuel cells convert the Gibbs free energy portion of the fuel's chemical energy into electrical energy through an electrochemical reaction, they are not limited by the Carnot cycle effect and are therefore highly efficient. Furthermore, fuel cells use fuel and oxygen as feedstock and have no mechanical transmission components, resulting in minimal emissions of harmful gases and a long service life.

[0081] High-energy fuel cells refer to fuel cells that use fuels with high energy density and high specific energy.

[0082] Energy density refers to the energy contained in a unit volume or unit mass of fuel. High-energy fuels have a much higher energy density than ordinary fuels. For example, liquid hydrogen is a common high-energy fuel. Its energy density is very high; the energy released when a unit mass of liquid hydrogen is burned is far greater than that of traditional fuels such as gasoline. Specific energy refers to the energy released when a unit mass of fuel is completely burned. A high-energy fuel has a high specific energy, meaning it can release more energy for the same mass. Examples include aviation kerosene and boron hydride. High-energy fuel cells can meet the high energy requirements of some special equipment or systems, helping to achieve more efficient and powerful power output or longer driving range.

[0083] This invention provides a packaging and testing method and apparatus for high-energy fuel cells. The high-energy fuel cell includes a tank 1, a spiral electrode tube 2 and a venting cylinder 3 disposed inside the tank 1, and ceramic pillars at the bottom of the tank 1. The method includes: acquiring and verifying the quality inspection information of the high-energy fuel cell; extracting an image of the high-energy fuel cell to be tested based on the quality inspection information; analyzing the positional relationship between the spiral electrode tube 2 and the venting cylinder 3, and the number of ceramic pillars, based on the image to be tested, and providing a first test result; optimizing the position of the spiral electrode tube 2 based on the first test result, controlling the packaging of the high-energy fuel cell, and providing packaging test data; and analyzing the packaging test data based on a pre-constructed three-dimensional model of the fuel cell, providing a packaging test result, and completing the packaging test of the high-energy fuel cell. By monitoring and analyzing the quality of the fuel cell before and after packaging, the reliability and stability of the fuel cell are ensured.

[0084] The encapsulation and testing method enables comprehensive monitoring and efficient testing of the fuel cell encapsulation process, thereby improving the quality of fuel cell encapsulation and testing, and ensuring the performance and safety of the fuel cell. Simultaneously, it enables the collection and uploading of data from the encapsulation and testing process, ensuring the traceability of process data.

[0085] like Figure 1 As shown in the figure, this embodiment of the invention provides a packaging and testing method for high-energy fuel cells, the specific steps of which are as follows:

[0086] S101: Obtain and verify the quality inspection information of high-energy fuel cells.

[0087] It is important to understand that, as a reference Figure 2 The high-energy fuel cell includes a tank 1, a spiral electrode tube 2 and a venting cylinder 3 disposed inside the tank 1, and a ceramic column at the bottom of the tank 1. In addition, the high-energy fuel cell includes an auxiliary mounting fixture, which is used to fix and position the tank 1 and other components, ensuring the stability of each component during assembly and operation. The auxiliary mounting fixture typically includes a base plate and multiple positioning components, which can be detachably fixed to the base plate to accommodate tanks 1 of different sizes. The tank 1 is the core of the entire high-energy fuel cell and is usually located at the center of the auxiliary mounting fixture, fixed by the positioning components of the auxiliary mounting fixture.

[0088] The spiral electrode tube 2 is typically installed inside the tank 1 to guide current flow and participate in electrochemical reactions. The position of the spiral electrode tube 2 needs precise adjustment to ensure proper alignment with other components within the tank 1 (such as the ceramic column). The tank 1 provides support and protection for the spiral electrode tube 2, ensuring its stable operation within the tank 1. The spiral electrode tube 2 includes a first terminal portion 21, a second terminal portion 23, and a spiral coil portion 22. The end of the first terminal portion 21 furthest from the vent cylinder 3 is fixed at the opening of the tank 1 using an auxiliary mounting clamp. The first terminal portion 21 passes through the cavity inside the vent cylinder 3 and connects to one end of the spiral coil portion 22. The spiral coil portion 22 is spirally wound around the outside of the vent cylinder 3. The free end of the spiral coil portion 22 connects to one end of the second terminal portion 23. The other end of the second terminal portion 23 is fixed at the opening of the tank 1 using an auxiliary mounting clamp. The ceramic column is located at the bottom of the tank 1 to fix and support the vent cylinder 3 and the spiral electrode tube 2, preventing short circuits and heat conduction. The ceramic column needs to be precisely aligned with the spiral electrode tube 2 to ensure good electrical isolation and thermal management. The spiral electrode tube 2 fits tightly with the ceramic column, and the size and position design of the ceramic column must ensure that the spiral electrode tube 2 remains stable during operation.

[0089] The vent cylinder 3 is used to discharge gases (such as water vapor) generated during the reaction process. The vent cylinder 3 is connected to the tank 1 via a pipe or channel, ensuring that gas can flow from inside the tank 1 to the vent cylinder 3. The location of the vent cylinder 3 needs to be optimized based on the structure and operational requirements of the tank 1 to ensure smooth gas discharge while avoiding interference with other components.

[0090] In addition to securing the tank body 1, the auxiliary installation fixture may also be used to secure the venting cylinder 3, ensuring its stability during operation. The fixture design needs to consider the size and position of the venting cylinder 3 to achieve optimal securing effect. The venting cylinder 3 is secured to the fuel cell tank body 1 together with the fixture, ensuring that it will not shift during operation.

[0091] Before verifying the quality inspection information of high-energy fuel cells, the packaging and testing methods also include:

[0092] Collect and test environment status, where the test environment status is data reflecting whether packaging and testing can be carried out.

[0093] Understandably, before encapsulating a fuel cell, the encapsulation environment and the components used in the encapsulation process must be inspected to ensure there are no problems before proceeding with the fuel cell encapsulation process.

[0094] Various types of sensors are deployed in the packaging environment to collect various data from the packaging environment. Based on this, it is determined whether the operating status of the detection environment and various devices is normal, and whether the operating conditions of the devices and the packaging detection environment meet the operating requirements.

[0095] When the environmental conditions are normal, it means that the fuel cell can be packaged. At this time, the quality inspection information of the high-energy fuel cell is collected by barcode scanner or vision sensor. The quality inspection information includes various battery parameters, including the fuel cell model, and the corresponding testing standards.

[0096] S102: Extract the image of the high-energy fuel cell to be inspected based on the quality inspection information of the high-energy fuel cell.

[0097] After the quality inspection information confirms that there are no problems, the high-energy fuel cell, including tank 1 and various components, is placed in the visual inspection area. The camera equipment deployed in the visual inspection area takes pictures and obtains the image to be inspected.

[0098] S103: Based on the image to be tested, analyze the positional relationship between the spiral electrode tube 2 and the ventilator 3, as well as the number of ceramic columns, and give the first test result.

[0099] Reference Figure 3 Specifically, it includes:

[0100] The image to be detected is filtered, and the initial contour edges of the spiral electrode tube 2, the air cylinder 3 and the ceramic column are extracted from the image to be detected.

[0101] The initial contour edge is traced, and the broken parts of the initial contour edge are repaired to obtain the target contour edge;

[0102] The target contour edge is compared and analyzed with a pre-defined standard contour to provide contour similarity data;

[0103] Based on a pre-set similarity threshold, the contour similarity data is judged, and the first detection result is given.

[0104] Furthermore, referring to Figure 4 The image to be detected is filtered to extract the initial contour edges of the spiral electrode tube 2, the vent cylinder 3, and the ceramic column, specifically including:

[0105] The image to be detected is processed into grayscale to obtain a grayscale detection image;

[0106] The grayscale detection image is filtered to remove noise and obtain a standard detection image.

[0107] Analyze the gradients of each pixel in the standard detection image in the horizontal and vertical directions to obtain the horizontal gradient and vertical gradient;

[0108] The horizontal and vertical gradients of each pixel are fused to obtain the edge intensity and edge direction of each pixel;

[0109] Based on the edge intensity and edge direction of each pixel, edge pixels are selected from the standard detection image to obtain the initial contour edges of the spiral electrode tube 2, the vent cylinder 3, and the ceramic column.

[0110] In the embodiments provided by this invention, a weighted average is performed on the pixel values ​​of each pixel in the image to be detected to obtain the grayscale pixel value corresponding to each pixel, thus providing a grayscale detection image. Specifically, this is expressed as follows:

[0111] Gray i =0.299*R i +0.587*G i +0.114*B i

[0112] Among them, Gray i Let R be the value of the i-th grayscale pixel in the grayscale detection image. i Let G be the red intensity of the i-th pixel in the image to be detected. i Let B be the green intensity of the i-th pixel in the image to be detected. iLet be the blue intensity of the i-th pixel in the image to be detected.

[0113] By performing grayscale processing on the image to be detected, converting the RGB image into a grayscale image, the amount of data in the grayscale detection image is reduced while retaining the main structural information of the image to be detected. This provides a data foundation for subsequent analysis of the grayscale detection image and improves the efficiency of image analysis.

[0114] In the embodiments provided by this invention, Gaussian filtering is used for filtering the grayscale detection image. Gaussian filtering is a linear smoothing filter used for image denoising, particularly for removing Gaussian noise. A Gaussian kernel is used to convolve the image, replacing each pixel value with the weighted average of its neighboring pixel values. First, based on the size of the grayscale detection image, a Gaussian kernel and filtering region are defined. Based on the filtered pixels in the grayscale detection image, and combined with a Gaussian function, a weight matrix is ​​calculated for each pixel within the corresponding filtering region. The Gaussian kernel and weight matrix are then convolved to update the pixel values ​​of the filtered pixels, completing the filtering process for the grayscale detection image and obtaining a standard detection image. Specifically, the Gaussian function is expressed as:

[0115]

[0116] Where σ is the Gaussian kernel, (x,y) is the coordinate of the pixel in the filtering region, and G(x,y) is the weight value of the pixel (x,y) in the weight matrix within the filtering region.

[0117] Gaussian filtering smooths an image by convolving a Gaussian kernel with each pixel and its neighborhood (the filtered region) to calculate a new pixel value, while preserving the main structure of the grayscale detection image and effectively removing high-frequency noise. The weights of the Gaussian kernel are determined by a Gaussian function, and the weights decrease as the distance from the center pixel increases.

[0118] Analyze the gradients of each pixel in the standard detection image in the horizontal and vertical directions to obtain the horizontal gradient and vertical gradient;

[0119] The horizontal and vertical gradients of each pixel are fused to obtain the edge intensity and edge direction of each pixel;

[0120] Based on the edge intensity and edge direction of each pixel, edge pixels are selected from the standard detection image to obtain the initial contour edges of the spiral electrode tube 2, the vent cylinder 3, and the ceramic column.

[0121] In the embodiments provided by this invention, the gradients of each pixel in the horizontal and vertical directions are calculated using the Sobel operator. The Sobel operator includes a horizontal operator and a vertical operator. The horizontal gradient and vertical gradient of each pixel are obtained by summing the neighborhood corresponding to each pixel with the horizontal and vertical operators, respectively. Taking a specific pixel as an example, if it is considered the center pixel, its neighborhood is a 3×3 region consisting of the center pixel and its eight surrounding pixels. Each value in this region is the pixel value of the corresponding pixel. The horizontal gradient is obtained by weighted summing the neighborhood matrix with the horizontal operator. The vertical gradient is obtained by weighted summing the neighborhood matrix with the vertical operator. The edge strength of the pixel is obtained by taking the square root of the sum of the squares of the horizontal and vertical gradients. The edge direction is obtained by substituting the ratio of the vertical and horizontal gradients into the inverse function of the tangent function.

[0122] Edge strength, specifically expressed as:

[0123]

[0124] Edge direction, specifically represented as:

[0125]

[0126] Where M is the edge intensity of the pixel, θ is the edge direction of the pixel, and G... x G represents the horizontal gradient of a pixel. y This represents the vertical gradient of a pixel.

[0127] If the edge intensity of each pixel reaches the preset edge intensity threshold, then the pixel is taken as an edge pixel to obtain the initial contour edge.

[0128] Edges in a standard detection image can be determined based on edge intensity and direction. A higher edge intensity value indicates a stronger edge. The edge intensity threshold is set according to the specific circumstances and is not specified here.

[0129] Furthermore, referring to Figure 5 The initial contour edge is traced, and broken parts of the initial contour edge are repaired to obtain the target contour edge, specifically including:

[0130] The initial contour edge is traced to obtain the broken part of the initial contour edge, wherein the initial contour edge includes a first terminal part 21, a second terminal part 23, a spiral coil part 22 and a breather cylinder part;

[0131] Based on the length of the fractured portion, different repair strategies are matched to repair the fractured portion and obtain the target contour edge.

[0132] Specifically, the edge pixel point at the end of the first terminal portion 21 that is away from the spiral coil portion 22 is taken as the starting point of the first terminal tracking, and the first terminal tracking direction is set to vertically downward;

[0133] The edge pixel at the end of the second terminal portion 23 that is away from the spiral coil portion 22 is taken as the starting point of the second terminal tracking, and the second terminal tracking direction is set to vertically downward.

[0134] Based on the first terminal tracking start point of the first terminal portion 21 and the second terminal tracking start point of the second terminal portion 23, and combined with the preset first terminal tracking direction and second terminal tracking direction, the edge pixels in the first terminal portion 21 and the second terminal portion 23 are traversed to give the broken parts of the first terminal portion 21 and the second terminal portion 23.

[0135] The connection point between the first terminal 21 and the spiral coil 22 or the connection point between the second terminal 23 and the spiral coil 22 is used as the starting point for spiral tracking.

[0136] Based on the spiral tracking starting point of the spiral coil section 22, and combined with the preset spiral tracking direction, the edge pixels in the spiral coil section 22 are traversed and the spiral tracking direction is updated to give the broken part of the spiral coil section 22. The spiral tracking direction is determined based on the shape of the spiral coil section 22 and the current spiral tracking starting point.

[0137] Based on the starting point of the ventilator cylinder 3 in the ventilator section, and combined with the preset ventilator cylinder 3 tracking direction, the edge pixels in the ventilator section are traversed to give the broken part of the ventilator section.

[0138] The traversal process ends by combining the closed-loop analysis of the tracking directions of the first terminal, the second terminal, the spiral tracking direction, and the tracking direction of the ventilator 3.

[0139] Based on the length of the fractured portion, a corresponding repair strategy is matched to repair the fractured portion and obtain the target contour edge.

[0140] In the embodiments provided by the present invention, the first terminal tracking start point is first determined from the first terminal portion 21 in the initial contour edge. For example, the edge pixel point at the end of the first terminal portion 21 away from the spiral coil portion 22 is taken as the first terminal tracking start point, and the first terminal tracking direction is set to vertically downward.

[0141] After determining the first terminal tracking start point, the system detects whether edge pixels exist along the terminal tracking direction, using the first terminal tracking start point as the starting point. If edge pixels exist, the first terminal tracking start point is updated, and the detected edge pixels are used as the new tracking start point, while the tracking direction is adjusted. For the first terminal section 21, which is in the vertical direction, there is no need to update the terminal tracking direction. The same applies to the second terminal section 23. However, for the spiral coil section 22, the spiral tracking direction needs to be updated multiple times according to the actual situation.

[0142] First, the spiral tracking starting point of the spiral coil section 22 is determined. For example, the connection point between the first terminal section 21 and the spiral coil section 22 can be used as the spiral tracking starting point, or the connection point between the second terminal section 23 and the spiral coil section 22 can be used as the spiral tracking starting point; there is no limitation on this. After determining the spiral tracking starting point, edge pixels are detected along the spiral tracking direction, using the spiral tracking starting point as the starting point. If edge pixels are found, the spiral tracking starting point is updated, and the detected edge pixels are used as the new spiral tracking starting point. The spiral tracking direction is adjusted as needed. The spiral tracking direction is determined based on the shape of the spiral coil section 22 and the current spiral tracking starting point. For example, if the spiral direction of the spiral coil section 22 makes a 45-degree angle with the horizontal line, the spiral tracking direction is set to the lower left corner at 45 degrees. Each time the spiral tracking direction is updated, the current spiral tracking direction is rotated 90 degrees clockwise and 90 degrees counterclockwise alternately to obtain the new spiral tracking direction. The tracking continues based on the adjusted spiral tracking direction. If no edge pixels are found in the adjusted spiral tracking direction, the spiral tracking direction is adjusted again until an edge pixel is found. This process is repeated until no new edge pixels are found in any of the spiral tracking directions. The adjustment of the spiral tracking direction is set according to the actual situation and is not limited thereto.

[0143] It is understandable that, since both the spiral electrode tube 2 and the ventilator 3 are inside the tank 1, the initial contour edges of the spiral electrode tube 2 and the ventilator 3 are together in the standard inspection image. Therefore, when performing contour tracking of each part of the spiral electrode tube 2, contour tracking of the ventilator 3 (ventilator section) is also performed.

[0144] Similar to the aforementioned contour tracking, the starting point for tracking the ventilator cylinder 3 is first determined. This can be the intersection of the ventilator cylinder and the first terminal 21, the upper left corner of the ventilator cylinder 3, or any other point; there are no limitations on this. Using the upper left corner as the starting point, the system detects the presence of edge pixels along the tracking direction (horizontal to the right) of the ventilator cylinder 3. If an edge pixel is found, the starting point is updated, and the detected edge pixel is used as the new starting point, while the tracking direction is adjusted. For the ventilator cylinder, if no new edge pixels are detected in the current tracking direction of the ventilator cylinder 3, the tracking direction is updated by rotating the rectangle vertically in the ventilator cylinder section until contour tracking is complete.

[0145] After each tracking iteration, it is determined whether the tracked line segment is a closed segment. If not, the tracking continues in the opposite direction from the starting point to another ending point; if so, the tracking of the current line segment ends. If multiple edge pixels exist during the tracking process, the nearest edge pixel is selected as the new tracking starting point, and the other edge pixels are used as tracking starting points for other lines. After tracking is completed, the various parts of the initial contour edge are determined, resulting in the broken portion of the initial contour edge.

[0146] Furthermore, based on the length of the fractured portion, a corresponding repair strategy is matched to repair the fractured portion, resulting in the target contour edge, specifically including:

[0147] If the length of the broken part is less than the preset breaking threshold, the pixel values ​​of the corresponding pixels are filled according to the breaking start point and breaking end point of the broken part to obtain the target contour edge.

[0148] If the length of the fractured portion is greater than or equal to the fracture threshold, the fractured portion is filled based on the fracture line direction to obtain the target contour edge.

[0149] In the embodiments provided by this invention, if the length of the fractured portion is less than a preset fracture threshold, it indicates that the number of pixels in the fractured portion is relatively small, and the pixel value of the fractured portion can be given by an interpolation algorithm. For example, the fracture start point (x1, y1) and fracture end point (x2, y2) can be connected by a straight line in linear interpolation, with the midpoint (x1, y1) being the midpoint. m ,y m The grayscale value g(x) m ,y m Specifically, it is expressed as:

[0150]

[0151] Where g(x1,y1) is the gray value at the fracture initiation point, and g(x2,y2) is the gray value at the fracture termination point.

[0152] Furthermore, based on the fracture line direction of the fractured portion, the fractured portion is filled to obtain the target contour edge, specifically including:

[0153] Obtain the fracture start point, fracture end point, and corresponding start and end adjacent points of the fractured part;

[0154] Based on the initial adjacent points, analyze the direction of the initial straight line from the fracture initiation point to the fracture termination point;

[0155] Based on the adjacent points, analyze the direction of the end line line near the fracture initiation point of the fracture termination point.

[0156] By merging the starting and ending straight line directions, the direction of the break line is obtained;

[0157] Based on the direction of the fracture line, calculate the positions of adjacent points sequentially from the fracture start point or fracture end point, and fill the adjacent points to obtain the target contour edge.

[0158] In the embodiments provided by this invention, the starting adjacent points include adjacent pixels in the direction of the break start point towards the break end point and adjacent pixels in the direction of the break start point away from the break end point. The ending adjacent points include adjacent pixels in the direction of the break end point towards the break start point and adjacent pixels in the direction of the break end point away from the break start point. Two points determine a line; therefore, the starting line direction can be determined based on the two starting adjacent points, and the ending line direction can be determined based on the two ending adjacent points. The average of the starting line direction and the ending line direction is used to obtain the break line direction. The positions of adjacent points are calculated sequentially from the break start point, and the adjacent points are filled to obtain the target contour edge. Alternatively, the positions of adjacent points can be calculated sequentially from the break end point, and the adjacent points are filled to obtain the target contour edge.

[0159] In a specific example, there exists a fracture start point P1(x1,y1), a fracture end point P2(x2,y2), and a starting adjacent point P. 1a (x 1a ,y 1a ), P 1b (x 1b ,y 1b ), end the adjacent point P 2a (x 2a ,y 2a ), P 2b (x 2b ,y 2b ), where P 1a For the adjacent point far from the fracture termination point, P1b For the adjacent point near the fracture termination point, P 2a For adjacent points near the fracture initiation point, P 2b The points adjacent to the fracture initiation point are far away from it.

[0160] The initial straight line direction θ1 is specifically represented as follows:

[0161]

[0162] The ending line direction θ2 is specifically represented as:

[0163]

[0164] Fracture line direction θ obj Specifically, it is expressed as:

[0165]

[0166] Based on the direction of the fracture line, the positions of adjacent points are calculated sequentially from the fracture initiation point, and these adjacent points are then filled in. Specifically, this is represented as follows:

[0167]

[0168] Among them, (x j ,y j (x1, y1) represents the coordinates of the adjacent point of the j-th filling, n represents the number of pixels that need to be filled in the broken part, L represents the distance between the starting point and the ending point of the break, and (x1, y1) represents the coordinates of the starting point of the break.

[0169] Furthermore, referring to Figure 6 The target contour edge is compared and analyzed with a pre-defined standard contour to provide contour similarity data, specifically including:

[0170] Based on the pixel values ​​of each edge pixel on the target contour edge, the mean value of the target edge is given;

[0171] The average value of the standard edge is given based on the pixel values ​​of each edge pixel on the standard contour.

[0172] By combining the target edge mean and the standard edge mean, the edge covariance of the target contour edge and the standard contour is given;

[0173] Based on the standard deviations of the target contour edge and the standard contour, the edge covariance is fused to provide contour similarity data.

[0174] It is understandable that after processing the image to be detected, the resulting target contour edges include the contours of the spiral electrode tube 2, the vent cylinder 3, and the ceramic column. For different target contour edges, there are corresponding standard contours.

[0175] Taking a target contour edge as an example, the pixel values ​​of the edge pixels on the target standard contour are [a1, a2, ..., a m The corresponding standard contour pixel values ​​are [b1, b2, ..., b]. m Then the marginal covariance Cov a,b Specifically, it is expressed as:

[0176]

[0177] in, The mean of the target margin. This is the standard marginal mean.

[0178] Contour Similarity Data NCC a,b Specifically, it is expressed as:

[0179]

[0180] Among them, NCC a,b ∈[-1,1], when the contour similarity data is 1, it means that the edge of the target contour and the corresponding standard contour are exactly the same or highly correlated. When the contour similarity data is -1, it means that the edge of the target contour and the corresponding standard contour are completely opposite (or negatively correlated). When the contour similarity data is 0, it means that there is no correlation between the edge of the target contour and the corresponding standard contour.

[0181] The results of contour similarity data do not depend on the size of the target standard contour, but only on the shape and relevance of the target standard contour.

[0182] S104: Based on the first test result, optimize the position of the spiral electrode tube 2, control the packaging of the high-energy fuel cell, and provide packaging test data.

[0183] Reference Figure 7 Specifically, it includes:

[0184] Based on the first detection result, the position of the spiral electrode tube 2 is determined, and a control command is output to position and restrict the spiral electrode tube 2.

[0185] Based on the positional relationship between the sealing hole of the high-energy fuel cell and the spiral electrode tube 2, the high-energy fuel cell is encapsulated.

[0186] The packaging and testing data of the high-energy fuel cell after packaging are collected. The packaging and testing data includes packaging and testing images and packaging performance data.

[0187] When the first test result is normal, it indicates that the high-energy fuel cell can be packaged. Before filling the high-energy fuel cell, the design specifications of the fuel cell are determined according to the fuel cell model, and the filling volume and filling speed are set. Different control methods are used for different filling weight stages. According to the weight of the impact of the feeding screw speed on filling accuracy, the first 90% of the target weight is fed at full speed using open-loop control. After rapidly approaching 90% of the target weight, it switches to low-speed precision feeding PID closed-loop control, setting an early shut-off threshold (such as 99% of the target value). The lag of the falling powder is used to complete the final filling, achieving a filling error control of ±1 gram. The control logic of fast first and then slow down improves the filling speed and ensures filling accuracy.

[0188] The filled fuel cell is transferred into the pressing and sealing mold. The spiral electrode tube 2 is positioned and restricted by the electrode tube positioning clamp, and the mold sensor detects that the spiral electrode tube 2 is in position. Based on control commands, a rodless magnetic drive cylinder is controlled to press the fuel cell cap positioning mold. The installation position of the constraint components is controlled by the tooling mold to ensure that the cap hole of the high-energy fuel cell is aligned with the spiral electrode tube 2. Furthermore, the pressure adjustment proportional valve is controlled by analog data feedback from the pressing mold pressure sensor to adjust the pressure of the rodless magnetic drive cylinder, ensuring a consistent size of the cap pressing gap. Welding is then performed. The laser welding machine control parameter formula is selected according to the specifications of the fuel cell, including scanning speed, scanning width, peak power, output duty cycle, pulse frequency, and welding speed. The servo turntable controls the pressing mold according to the set welding speed to perform cap positioning spot welding and circumferential welding.

[0189] After encapsulation, various sensors are deployed to test the design performance indicators of the encapsulated high-energy fuel cell. These include weighing, resistance, pressure holding, heating, temperature, and visual inspection. Encapsulation performance data includes data obtained from weighing, resistance, pressure holding, heating, and temperature measurements, while encapsulation inspection images include images obtained from visual inspection.

[0190] Among them, the weighing detection is to use a high-precision weighing sensor to detect the overall weight of the packaged high-energy fuel cell to determine whether there is any empty or missing parts, so as to ensure that the overall weight of the packaged product is controlled within the allowable error range of the design specifications.

[0191] The resistance detection is used to measure whether the spiral electrode tube 2 is displaced during the fuel filling and press-fit encapsulation process. If displacement occurs, it will cause a short circuit between the bodies of the spiral electrode tube 2 or a short circuit between the spiral electrode tube 2 and the housing. Specifically, the two ends of the tested electrode tube and the detection points on the housing are connected to the quick-connect test clips of the resistance measuring instrument according to the interface direction. The resistance measuring instrument gradually detects the resistance value of the electrode tube and the insulation resistance of the electrode tube to the fuel cell housing by switching the voltage level of the electromotive force of the detection output power supply. When the detected resistance value of the electrode tube falls within the resistance range required by the fuel cell specification design, it is regarded as qualified.

[0192] The pressure holding detection is to inject argon into the fuel cell through the gas-filled solenoid valve. When the detected pressure of the added gas is ≥3bar, the gas-filled solenoid valve will be automatically closed. After standing for 10 seconds to stabilize the backflushing pressure of the gas, the pressure holding detection timer is started, and the pressure holding values at the start and end are recorded. When the detection time reaches the set time, the system calculates the pressure holding leakage rate by calculating the pressure difference change within the set detection time, so as to determine whether the encapsulation of the fuel cell meets the designed sealing requirements.

[0193] The heating detection is to supply a high-frequency power of 20KHz-30KHz through the electrode tube heating power interface. When the high-frequency current passes through the spiral tube of the electrode tube, an alternating magnetic field will be generated around it. Due to the electromagnetic induction effect on the metal gas-permeable cylinder 3 inside the electrode tube, eddy currents will be generated inside the gas-permeable cylinder 3, so that the gas-permeable cylinder 3 heats up rapidly.

[0194] The temperature detection is to read the temperature rise change data of the electrode tube heating test through the K-type temperature sensor built in the fuel cell, calculate the temperature change rate within the detection time, judge the heating performance of the electrode tube and the temperature measurement performance of the temperature sensor, so as to ensure that the heating and temperature measurement functions of the fuel cell are normal.

[0195] The visual detection adopts a multi-angle trinocular CCD camera, which is distributed in three directions: left, middle and right, to eliminate the blind area of a single viewing angle. It synchronously collects images of the encapsulation appearance and accessories (plugs, bolts, pipeline interfaces, etc.) of the fuel cell at 30fps, realizing full coverage of the viewing angle in the detection area, with a resolution of up to 4096×3000 pixels.

[0196] S105: Combined with the pre-constructed three-dimensional model of the fuel cell, the encapsulation detection data is analyzed, the encapsulation detection result is given, and the encapsulation detection of the high-energy fuel cell is completed.

[0197] Reference Figure 8 , specifically including:

[0198] Perform stereo matching and depth calculation based on the encapsulation detection image to construct a three-dimensional encapsulation model;

[0199] Based on the three-dimensional encapsulation model, analyze the surface deformation of each part of the fuel cell in the three-dimensional encapsulation model, and give the corresponding surface curvature of each part;

[0200] Analyze the differences between the surface curvature of each part and the surface curvature of the corresponding part of the fuel cell 3D model, and provide curvature difference data;

[0201] Determine the relationship between the curvature difference data and the pre-acquired standard curvature difference range, provide the packaging test results, and complete the packaging test of the high-energy fuel cell.

[0202] In the embodiments provided by this invention, three-dimensional point cloud data is generated by combining the encapsulation detection images obtained from a tri-lens CCD camera. Model reconstruction is then performed on the three-dimensional point cloud data to complete the construction of the encapsulated three-dimensional model. The fuel cell in the encapsulated three-dimensional model is divided into different parts. This division can be done by average division or based on the correlation between different regions; there is no limitation on this. Next, the surface deformation of different parts is analyzed, and the surface curvature of each part is given. The difference between the surface curvature of different parts and the surface curvature of the corresponding part of the fuel cell three-dimensional model is calculated to obtain curvature difference data. The relationship between the curvature difference data and the pre-acquired standard curvature difference range is determined, and the encapsulation detection result is given. If the curvature difference data is within the standard curvature difference range, the encapsulation detection result is normal; if the curvature difference data is not within the standard curvature difference range, the encapsulation detection result is abnormal. The encapsulation detection result is output to prompt relevant personnel to make adjustments to the fuel cell encapsulation, thus completing the encapsulation detection of the high-energy fuel cell.

[0203] The curvature difference data corresponding to the encapsulated 3D model is the average of the curvature difference data corresponding to different parts. For judging the curvature difference data, a single judgment can be made on the curvature difference data corresponding to the encapsulated 3D model, or the curvature difference data of different parts can be judged separately, providing the encapsulation detection result. If the curvature difference data of different parts is judged separately, the encapsulation detection result is abnormal if the curvature difference data of one part is outside the standard curvature difference range.

[0204] The obtained 3D model of the packaged fuel cell is compared with the pre-stored 3D model of the fuel cell using template matching and feature extraction methods to detect subtle features such as the appearance integrity, surface flatness, presence of small cracks, and abnormal installation of external accessories.

[0205] Reference Figure 9 This invention provides a packaging and testing device for high-energy fuel cells, comprising:

[0206] The data acquisition module 201 is used to acquire and verify the quality inspection information of the high-energy fuel cell, wherein the high-energy fuel cell includes a tank 1, a spiral electrode tube 2 and a venting cylinder 3 disposed inside the tank 1, and a ceramic column at the bottom of the tank 1.

[0207] The image extraction module 202 is used to extract the image to be inspected of the high-energy fuel cell based on the quality inspection information of the high-energy fuel cell;

[0208] Image detection module 203 is used to combine the image to be detected, analyze the positional relationship between the spiral electrode tube 2 and the ventilator 3 and the number of ceramic columns, and give the first detection result;

[0209] The packaging and testing module 204 is used to optimize the position of the spiral electrode tube 2, control the packaging of the high-energy fuel cell, and provide packaging and testing data based on the first testing result.

[0210] The encapsulation analysis module 205 is used to analyze the encapsulation test data by combining the pre-built three-dimensional model of the fuel cell, give the encapsulation test results, and complete the encapsulation test of the high-energy fuel cell.

[0211] Those skilled in the art will understand that, for the sake of convenience and brevity, the specific working process of the described module can be referred to the corresponding process in the foregoing method embodiments, and will not be repeated here.

[0212] Although preferred embodiments of the invention have been described, those skilled in the art, upon learning the basic inventive concept, can make other changes and modifications to these embodiments. Therefore, the appended claims are intended to be interpreted as including both the preferred embodiments and all changes and modifications falling within the scope of the invention. Clearly, those skilled in the art can make various alterations and modifications to the invention without departing from its spirit and scope. Thus, if these modifications and variations of the invention fall within the scope of the claims and their equivalents, the invention is also intended to include these modifications and variations.

Claims

1. A packaging and testing method for a high-energy fuel cell, characterized in that, High-energy fuel cells include a tank, a spiral electrode tube and a venting cylinder disposed inside the tank, and a ceramic pillar at the bottom of the tank. The packaging and testing methods include: Obtain and verify the quality inspection information of high-energy fuel cells; Based on the quality inspection information of high-energy fuel cells, extract the images of high-energy fuel cells to be inspected; Based on the image to be tested, the positional relationship between the spiral electrode tube and the vent cylinder, as well as the number of ceramic columns, are analyzed to give the first test result; Based on the first test result, the position of the spiral electrode tube is optimized, the packaging of the high-energy fuel cell is controlled, and packaging test data is provided. Specifically, this includes: determining the position of the spiral electrode tube based on the first test result and outputting control commands to position and limit the spiral electrode tube; packaging the high-energy fuel cell based on the relationship between the sealing hole position of the high-energy fuel cell and the position of the spiral electrode tube; and collecting packaging test data of the high-energy fuel cell after packaging, wherein the packaging test data includes packaging test images and packaging performance data. By combining a pre-constructed 3D model of the fuel cell, analyzing the encapsulation inspection data, and providing the encapsulation inspection results, the encapsulation inspection of the high-energy fuel cell is completed. Specifically, this includes: performing stereo matching and depth calculation based on the encapsulation inspection images to construct a 3D encapsulation model; analyzing the surface deformation of each part of the fuel cell in the 3D encapsulation model and providing the surface curvature of each part; analyzing the difference between the surface curvature of each part and the surface curvature of the corresponding part in the 3D model of the fuel cell, and providing curvature difference data; determining the relationship between the curvature difference data and the pre-obtained standard curvature difference range, providing the encapsulation inspection results, and completing the encapsulation inspection of the high-energy fuel cell.

2. The packaging and testing method for high-energy fuel cells as described in claim 1, characterized in that, Based on the image to be tested, the positional relationship between the spiral electrode tube and the venting cylinder, and the number of ceramic columns are analyzed to provide the first test results, which include: The image to be detected is filtered, and the initial contour edges of the spiral electrode tube, the vent cylinder and the ceramic column in the image to be detected are extracted respectively. The initial contour edge is traced, and the broken parts of the initial contour edge are repaired to obtain the target contour edge; The target contour edge is compared and analyzed with a pre-defined standard contour to provide contour similarity data; Based on a pre-set similarity threshold, the contour similarity data is judged, and the first detection result is given.

3. The packaging and testing method for high-energy fuel cells as described in claim 2, characterized in that, The image to be detected is filtered, and the initial contour edges of the spiral electrode tube, the vent cylinder, and the ceramic column in the image are extracted respectively, specifically including: The image to be detected is processed into grayscale to obtain a grayscale detection image; The grayscale detection image is filtered to remove noise and obtain a standard detection image. Analyze the gradients of each pixel in the standard detection image in the horizontal and vertical directions to obtain the horizontal gradient and vertical gradient; The horizontal and vertical gradients of each pixel are fused to obtain the edge intensity and edge direction of each pixel; Based on the edge intensity and edge direction of each pixel, edge pixels are selected from the standard detection image to obtain the initial contour edges of the spiral electrode tube, the vent cylinder, and the ceramic column.

4. The packaging and testing method for high-energy fuel cells as described in claim 2, characterized in that, The initial contour edge includes a first terminal portion, a second terminal portion, a spiral coil portion, and a vent cylinder portion; The initial contour edge is traced, and broken parts of the initial contour edge are repaired to obtain the target contour edge, specifically including: The edge pixel at the end of the first terminal portion that is far from the spiral coil portion is taken as the starting point of the first terminal tracking, and the first terminal tracking direction is set to vertically downward. The edge pixel at the end of the second terminal portion that is far from the spiral coil portion is taken as the starting point of the second terminal tracking, and the tracking direction of the second terminal is set to vertically downward. Based on the first terminal tracking start point of the first terminal portion and the second terminal tracking start point of the second terminal portion, and combined with the preset first terminal tracking direction and second terminal tracking direction, the edge pixels in the first terminal portion and the second terminal portion are traversed to give the broken parts of the first terminal portion and the second terminal portion. The connection point between the first terminal and the spiral coil or the connection point between the second terminal and the spiral coil is taken as the starting point of spiral tracking. Based on the spiral tracking starting point of the spiral coil section, and combined with the preset spiral tracking direction, the edge pixels in the spiral coil section are traversed and the spiral tracking direction is updated to give the broken part of the spiral coil section. The spiral tracking direction is determined based on the shape of the spiral coil section and the current spiral tracking starting point. Based on the starting point of the ventilator cylinder tracking, combined with the preset ventilator cylinder tracking direction, the edge pixels in the ventilator cylinder are traversed to give the broken part of the ventilator cylinder. The traversal process ends by combining the closed-loop analysis of the tracking directions of the first terminal, the second terminal, the spiral tracking direction, and the ventilator tracking direction. Based on the length of the fractured portion, a corresponding repair strategy is matched to repair the fractured portion and obtain the target contour edge.

5. The packaging and testing method for high-energy fuel cells as described in claim 4, characterized in that, Based on the length of the fractured portion, a corresponding repair strategy is matched to repair the fractured portion, resulting in the target contour edge, specifically including: If the length of the broken part is less than the preset breaking threshold, the pixel values ​​of the corresponding pixels are filled according to the breaking start point and breaking end point of the broken part to obtain the target contour edge. If the length of the fractured portion is greater than or equal to the fracture threshold, the fractured portion is filled based on the fracture line direction to obtain the target contour edge.

6. The packaging and testing method for high-energy fuel cells as described in claim 5, characterized in that, Based on the fracture line direction of the fractured portion, the fractured portion is filled to obtain the target contour edge, specifically including: Obtain the fracture start point, fracture end point, and corresponding start and end adjacent points of the fractured part; Based on the initial adjacent points, analyze the direction of the initial straight line from the fracture initiation point to the fracture termination point; Based on the adjacent points, analyze the direction of the end line line near the fracture initiation point of the fracture termination point. By merging the starting and ending straight line directions, the direction of the break line is obtained; Based on the direction of the fracture line, calculate the positions of adjacent points sequentially from the fracture start point or fracture end point, and fill the adjacent points to obtain the target contour edge.

7. The packaging and testing method for high-energy fuel cells as described in claim 2, characterized in that, The target contour edge is compared and analyzed with a pre-defined standard contour to provide contour similarity data, specifically including: Based on the pixel values ​​of each edge pixel on the target contour edge, the mean value of the target edge is given; The average value of the standard edge is given based on the pixel values ​​of each edge pixel on the standard contour. By combining the target edge mean and the standard edge mean, the edge covariance of the target contour edge and the standard contour is given; Based on the standard deviations of the target contour edge and the standard contour, the edge covariance is fused to provide contour similarity data.

8. A packaging and testing device for a high-energy fuel cell, characterized in that, The packaging and testing method for high-energy fuel cells as described in any one of claims 1-7 includes: The data acquisition module is used to acquire and verify the quality inspection information of the high-energy fuel cell, which includes a tank, a spiral electrode tube and a venting cylinder installed inside the tank, and a ceramic column at the bottom of the tank. The image extraction module is used to extract the image of the high-energy fuel cell to be inspected based on the quality inspection information of the high-energy fuel cell; The image detection module is used to combine the image to be detected, analyze the positional relationship between the spiral electrode tube and the vent cylinder, the number of ceramic columns, and give the first detection result. The packaging and testing module is used to optimize the position of the spiral electrode tube, control the packaging of the high-energy fuel cell, and provide packaging and testing data based on the first testing result. Specifically, it includes: determining the position of the spiral electrode tube based on the first testing result and outputting control commands to position and limit the spiral electrode tube; packaging the high-energy fuel cell based on the relationship between the sealing hole position of the high-energy fuel cell and the position of the spiral electrode tube; and collecting packaging and testing data of the high-energy fuel cell after packaging, wherein the packaging and testing data includes packaging and testing images and packaging performance data. The encapsulation analysis module is used to analyze encapsulation inspection data by combining a pre-constructed 3D model of the fuel cell, provide encapsulation inspection results, and complete the encapsulation inspection of the high-energy fuel cell. Specifically, it includes: performing stereo matching and depth calculation based on the encapsulation inspection image to construct an encapsulation 3D model; analyzing the surface deformation of each part of the fuel cell in the encapsulation 3D model based on the encapsulation 3D model, and providing the surface curvature of each part; analyzing the difference between the surface curvature of each part and the surface curvature of the corresponding part in the fuel cell 3D model, and providing curvature difference data; determining the relationship between the curvature difference data and the pre-acquired standard curvature difference range, providing encapsulation inspection results, and completing the encapsulation inspection of the high-energy fuel cell.

Images