High-pressure hydrogen storage tank welding seam detection method and system based on image recognition

Abstract

The invention discloses a high-pressure hydrogen storage tank welding seam detection method and system based on image recognition. The method comprises the steps that high-pressure hydrogen storage tank welding seam image information is obtained; extracting welding seam features and corresponding welding seam feature values in the welding seam image of the high-pressure hydrogen storage tank; sending the weld joint characteristic value to a preset weld joint safety index evaluation model to obtain a safety index evaluation value of the corresponding weld joint; when the safety index evaluation value of the welding seam is larger than or equal to the safety index threshold value of the corresponding welding seam, the corresponding welding seam passes safety detection; and when the safety index evaluation value of the welding seam is smaller than the safety index threshold value of the corresponding welding seam, the position of the corresponding welding seam is extracted, and the position of the welding seam is sent to a preset management end to be displayed. The welding seam image is judged through the preset welding seam safety index evaluation model and the safety index threshold value of the welding seam, the welding seam detection accuracy is improved, and the quality guarantee period of the corresponding welding seam is prolonged.

Description

Technical Field

[0001] The present invention relates to the technical field of high-pressure hydrogen storage tank weld detection, and more specifically, to a high-pressure hydrogen storage tank weld detection method and system based on image recognition. Background Art

[0002] High-pressure hydrogen storage tanks are used to store hydrogen under high pressure to facilitate the transportation of hydrogen. The gaps in high-pressure hydrogen storage tanks will affect the transportation and loading safety of hydrogen. Currently, the welds of high-pressure hydrogen storage tanks are mainly detected through technologies such as laser scanning. This method can only determine the integrity of the high-pressure hydrogen storage tank at the time, and cannot evaluate the welding technology. The strength of the welding technology determines the duration for which the weld can maintain its integrity.

[0003] Therefore, the prior art has defects and needs to be improved urgently. Summary of the invention

[0004] In view of the above problems, the purpose of the present invention is to provide a high-pressure hydrogen storage tank weld detection method and system based on image recognition, which can improve the accuracy of weld detection and the shelf life of the corresponding weld.

[0005] A first aspect of the present invention provides a high-pressure hydrogen storage tank weld detection method based on image recognition, comprising:

[0006] Obtain image information of weld seams of high-pressure hydrogen storage tanks;

[0007] Extract weld features and corresponding weld feature values ​​from weld images of high-pressure hydrogen storage tanks;

[0008] The weld characteristic value is sent to a preset weld safety index evaluation model to obtain a safety index evaluation value of the corresponding weld;

[0009] When the safety index assessment value of a weld is greater than or equal to the safety index threshold of the corresponding weld, the corresponding weld passes the safety test;

[0010] When the safety index evaluation value of the weld is less than the safety index threshold of the corresponding weld, the position of the corresponding weld is extracted, and the position of the weld is sent to a preset management terminal for display;

[0011] The weld features include weld texture features, weld size features and weld outer contour features; the weld feature values ​​include weld texture feature values, weld size feature values ​​and weld outer contour feature values.

[0012] In this solution, the step of obtaining the weld texture feature value specifically includes:

[0013] Obtain the welding process of the corresponding weld and the shape and size of the weld;

[0014] According to the welding process of the corresponding weld, query in the preset weld shape and size table to obtain the standard shape and size of the corresponding weld;

[0015] Compare and analyze the shape and size of the weld with the standard shape and size of the corresponding weld to obtain similarity values;

[0016] The similarity value is multiplied by the weight coefficient of the shape and size of the corresponding weld in the weld texture feature, and then accumulated to obtain the corresponding weld texture feature value.

[0017] In this solution, the step of obtaining the weld size characteristic value specifically includes:

[0018] Extract the dimension elements in the weld dimension features and determine the values ​​of the corresponding dimension elements;

[0019] Calculate the difference between the value of the dimension element and the preset standard value of the corresponding dimension element to obtain the dimension element value difference;

[0020] Dividing the size element value difference by the preset standard value of the corresponding size element to obtain a first value of the corresponding size element;

[0021] Subtract the first value of the corresponding dimension element from the preset dimension element weight reference value to obtain the actual weight value of the corresponding dimension element;

[0022] The actual weight values ​​of different size elements are multiplied by the preset characteristic values ​​of the corresponding size elements, and then added up to obtain the weld size characteristic value.

[0023] In this solution, the step of obtaining the characteristic value of the weld outer contour specifically includes:

[0024] Extract pixel point cloud from the outer contour of the weld;

[0025] Based on the preset pixel point intervals, the pixel points in the pixel point cloud image in the outer contour line of the weld are divided, and the number of pixel points in each preset pixel point interval is calculated;

[0026] Extract the maximum number of pixels in the preset pixel interval and set it as the number of pixels on the standard contour line;

[0027] The number of pixel points on the standard contour line is divided by the total number of pixel points in the pixel point cloud image in the outer contour line of the weld to obtain a second value;

[0028] The second value is multiplied by a preset standard characteristic value of the weld outer contour to obtain a corresponding weld outer contour characteristic value.

[0029] In this solution, the step of sending the weld characteristic value to a preset weld safety index evaluation model to obtain a safety index evaluation value of the corresponding weld specifically includes:

[0030] Extracting weld texture feature values, weld size feature values ​​and weld outer contour feature values ​​from weld feature values;

[0031] The weld seam texture feature value is multiplied by a preset texture feature weight coefficient, the weld seam size feature value is multiplied by a preset size feature weight coefficient, and the weld seam outer contour feature value is multiplied by a preset outer contour feature weight coefficient, and then accumulated to obtain a weld seam safety index assessment value;

[0032] The sum of the preset texture feature weight coefficient, the preset size feature weight coefficient and the preset outer contour feature weight coefficient is equal to 1.

[0033] This plan includes:

[0034] Based on a preset time period, obtain historical fault data information of the high-pressure hydrogen storage tank;

[0035] Extract historical fault data information of high-pressure hydrogen storage tanks caused by gas leakage, and determine the historical gas leakage location and the number of times the same historical gas leakage location occurs;

[0036] According to the historical fault data information of the high-pressure hydrogen storage tank, determine the total number of corresponding historical faults;

[0037] According to the historical leakage data information of the high-pressure hydrogen storage tank, the historical leakage position and the number of times corresponding to the historical leakage position are obtained;

[0038] Divide the number of occurrences of the same historical leak location by the total number of historical faults to obtain the probability value of a fault occurring at the corresponding historical leak location;

[0039] If the probability value of a fault occurring at a historical air leakage location is greater than a preset probability threshold, the corresponding historical air leakage location is marked to obtain a marked historical air leakage location;

[0040] The marked historical air leakage position is sent to a preset management terminal for prompting.

[0041] In this solution, after sending the marked historical leakage position to the preset management terminal for prompting, it also includes:

[0042] The probability value of a failure at a corresponding historical gas leakage position is set as a safety index adjustment coefficient of the weld at the corresponding position;

[0043] The safety index adjustment coefficient of the corresponding weld is multiplied by the safety index threshold of the corresponding weld to obtain a safety index revision value of the corresponding weld;

[0044] The safety index threshold of the weld is added to the safety index revision value of the corresponding weld to obtain the revised safety index threshold of the corresponding weld.

[0045] A second aspect of the present invention provides a high-pressure hydrogen storage tank weld detection system based on image recognition, comprising: a parameter setting module, an image acquisition module, an image recognition module, a data storage module and a warning module;

[0046] The parameter setting module is used to set the parameters in the high-pressure hydrogen storage tank weld detection process;

[0047] The image acquisition module is used to acquire the weld image of the high-pressure hydrogen storage tank;

[0048] The image recognition module is used to recognize the weld image of the high-pressure hydrogen storage tank and determine whether the corresponding weld is safe;

[0049] The data storage module is used to store data for testing the weld images of the high-pressure hydrogen storage tank;

[0050] The warning module is used to prompt the preset management terminal.

[0051] In this solution, the image acquisition module is also used for:

[0052] Obtain the clarity of the weld image of the high-pressure hydrogen storage tank;

[0053] Determine whether the clarity of the high-pressure hydrogen storage tank weld image is greater than a preset clarity threshold, if so, the corresponding high-pressure hydrogen storage tank weld image is qualified; if not, the corresponding high-pressure hydrogen storage tank weld image is unqualified;

[0054] Based on the fact that the weld seam image of the high-pressure hydrogen storage tank is unqualified, the unqualified weld seam image of the high-pressure hydrogen storage tank is re-photographed.

[0055] In this solution, the image recognition module further includes: a weld feature and weld feature value extraction unit, a weld safety index evaluation unit, and a weld safety index determination unit;

[0056] The weld feature and weld characteristic value extraction unit is used to extract the weld feature and the corresponding weld characteristic value;

[0057] The weld safety index evaluation unit is used to calculate the weld safety index evaluation value according to the weld characteristic value;

[0058] The weld safety index determination unit is used to determine the safety index evaluation value of the weld and determine whether the corresponding weld passes the safety test.

[0059] The present invention discloses a high-pressure hydrogen storage tank weld detection method and system based on image recognition, which judges the weld image through a preset weld safety index evaluation model and a weld safety index threshold, thereby improving the accuracy of weld detection and the shelf life of the corresponding weld. BRIEF DESCRIPTION OF THE DRAWINGS

[0060] Figure 1 A flow chart of a high-pressure hydrogen storage tank weld detection method based on image recognition is shown in the present invention;

[0061] Figure 2 A block diagram of a high-pressure hydrogen storage tank weld detection system based on image recognition is shown in the present invention. DETAILED DESCRIPTION

[0062] In order to more clearly understand the above-mentioned purpose, features and advantages of the present invention, the present invention is further described in detail below in conjunction with the accompanying drawings and specific embodiments. It should be noted that the embodiments of the present application and the features in the embodiments can be combined with each other without conflict.

[0063] In the following description, many specific details are set forth to facilitate a full understanding of the present invention. However, the present invention may also be implemented in other ways different from those described herein. Therefore, the protection scope of the present invention is not limited to the specific embodiments disclosed below.

[0064] Figure 1 A flow chart of a high-pressure hydrogen storage tank weld detection method based on image recognition is shown in the present invention.

[0065] like Figure 1 As shown, the present invention discloses a high-pressure hydrogen storage tank weld detection method based on image recognition, comprising:

[0066] S101, obtaining image information of the weld of the high-pressure hydrogen storage tank;

[0067] S102, extracting weld features and corresponding weld feature values ​​in the weld image of the high-pressure hydrogen storage tank;

[0068] S103, sending the weld characteristic value to a preset weld safety index evaluation model to obtain a safety index evaluation value of the corresponding weld;

[0069] S104, when the safety index evaluation value of the weld is greater than or equal to the safety index threshold of the corresponding weld, the corresponding weld passes the safety test;

[0070] S105, when the safety index evaluation value of the weld is less than the safety index threshold of the corresponding weld, extract the position of the corresponding weld, and send the position of the weld to the preset management terminal for display.

[0071] According to an embodiment of the present invention, the weld features include weld texture features, weld size features and weld outer contour features; the weld feature values ​​include weld texture feature values, weld size feature values ​​and weld outer contour feature values; the weld features are extracted through the weld image, the weld features are digitized according to the weight parameters set in advance, and then the safety index evaluation value of the corresponding weld is calculated, and whether the process of the corresponding weld is qualified is determined by the safety index evaluation value of the weld, and if qualified, the corresponding weld passes the inspection.

[0072] According to an embodiment of the present invention, the step of obtaining the weld texture feature value specifically includes:

[0073] Obtain the welding process of the corresponding weld and the shape and size of the weld;

[0074] According to the welding process of the corresponding weld, query in the preset weld shape and size table to obtain the standard shape and size of the corresponding weld;

[0075] Compare and analyze the shape and size of the weld with the standard shape and size of the corresponding weld to obtain similarity values;

[0076] The similarity value is multiplied by the weight coefficient of the shape and size of the corresponding weld in the weld texture feature, and then accumulated to obtain the corresponding weld texture feature value.

[0077] It should be noted that when welding high-pressure hydrogen storage tanks, the corresponding technicians will record the welding process, such as laser welding, arc welding, etc. Each welding process has different characteristics, and the resulting weld texture characteristics are also different. The weld texture characteristics are composed of multiple shapes and sizes of the weld, such as weld width, weld excess height, penetration, weld thickness, etc. For example, the weld width is the straight-line distance between the two weld toes on the weld surface. The preset weld shape and size table stores various welding processes and the standard shapes and sizes of the corresponding welding processes. The similarity values ​​include the similarity values ​​of weld width, weld excess height, penetration, and thickness. The similarity value is calculated by comparing the shape and size of the weld with the standard shape and size of the corresponding weld. If the shape and size of the weld are within the standard shape and size of the corresponding weld, the corresponding similarity value is 100%. If the shape and size of the weld are larger than the standard shape and size of the corresponding weld, the shape and size of the weld is subtracted from the maximum value of the standard shape and size of the corresponding weld to obtain a first data difference; then the first data difference is divided by the maximum value of the standard shape and size of the corresponding weld to obtain a first data difference ratio; then the first data difference ratio is subtracted from 1 to obtain the value of the corresponding weld. and the standard shape and size of the corresponding weld; if the shape and size of the weld is smaller than the standard shape and size of the corresponding weld, the shape and size of the weld is subtracted from the minimum value in the standard shape and size of the corresponding weld to obtain a second data difference; the second data difference is divided by the minimum value in the standard shape and size of the corresponding weld to obtain a second data difference ratio, and then the second data difference ratio is subtracted from 1 to obtain a similarity between the shape and size of the corresponding weld and the standard shape and size of the corresponding weld; for example, the similarity value of the weld width, if the standard shape and size of the corresponding weld is 2 to 3 mm, and the current weld width is 1.9 mm, then the minimum value of 2 mm in the standard shape and size of the corresponding weld is extracted, and its weld width similarity value is (1-(2-1.9) / 2)*100%=95%; the weight coefficient occupied by the shape and size of each weld in the texture feature of the weld can be different, and the sum of the weight coefficients of the shapes and sizes of all welds in the weld texture feature is 1. For example, the weld width similarity value is multiplied by the corresponding weld width and the weight coefficient of the weld texture feature to obtain the characteristic value of the corresponding weld width, and then the characteristic value of the weld width and the characteristic values ​​of the shapes and sizes of other welds are accumulated to obtain the corresponding weld texture characteristic value.

[0078] According to an embodiment of the present invention, the step of obtaining the weld size characteristic value specifically includes:

[0079] Extract the dimension elements in the weld dimension features and determine the values ​​of the corresponding dimension elements;

[0080] Calculate the difference between the value of the dimension element and the preset standard value of the corresponding dimension element to obtain the dimension element value difference;

[0081] Dividing the size element value difference by the preset standard value of the corresponding size element to obtain a first value of the corresponding size element;

[0082] Subtract the first value of the corresponding dimension element from the preset dimension element weight reference value to obtain the actual weight value of the corresponding dimension element;

[0083] The actual weight values ​​of different size elements are multiplied by the preset characteristic values ​​of the corresponding size elements, and then added up to obtain the weld size characteristic value.

[0084] It should be noted that the dimension elements in the weld dimension feature include elements such as the weld width, weld length and weld thickness. The preset characteristic values ​​of all dimension elements are accumulated to obtain the preset standard characteristic value of the corresponding weld dimension.

[0085] According to an embodiment of the present invention, the step of obtaining the characteristic value of the weld outer contour specifically includes:

[0086] Extract pixel point cloud from the outer contour of the weld;

[0087] Based on the preset pixel point intervals, the pixel points in the pixel point cloud image in the outer contour line of the weld are divided, and the number of pixel points in each preset pixel point interval is calculated;

[0088] Extract the maximum number of pixels in the preset pixel interval and set it as the number of pixels on the standard contour line;

[0089] The number of pixel points on the standard contour line is divided by the total number of pixel points in the pixel point cloud image in the outer contour line of the weld to obtain a second value;

[0090] The second value is multiplied by a preset standard characteristic value of the weld outer contour to obtain a corresponding weld outer contour characteristic value.

[0091] It should be noted that the pixel points in the weld outer contour line are divided into intervals, an interval in which the corresponding pixel points are clustered is determined, and the interval in which the corresponding pixel points are clustered is used as a reference to determine the corresponding weld outer contour characteristic value; the preset standard characteristic value of the weld outer contour, the preset standard characteristic value of the weld size, and the preset standard characteristic value of the weld texture are equal, and the safety index threshold of the weld is less than or equal to the preset standard characteristic value of the weld outer contour.

[0092] According to an embodiment of the present invention, the step of sending the weld characteristic value to a preset weld safety index assessment model to obtain a safety index assessment value of the corresponding weld specifically includes:

[0093] Extracting weld texture feature values, weld size feature values ​​and weld outer contour feature values ​​from weld feature values;

[0094] The weld seam texture feature value is multiplied by a preset texture feature weight coefficient, the weld seam size feature value is multiplied by a preset size feature weight coefficient, and the weld seam outer contour feature value is multiplied by a preset outer contour feature weight coefficient, and then accumulated to obtain a weld seam safety index assessment value;

[0095] The sum of the preset texture feature weight coefficient, the preset size feature weight coefficient and the preset outer contour feature weight coefficient is equal to 1.

[0096] It should be noted that by extracting and setting the feature weight coefficient, the weight of each weld feature is calculated and then accumulated to obtain the weld safety index evaluation value.

[0097] According to an embodiment of the present invention, the invention comprises:

[0098] Based on a preset time period, obtain historical fault data information of the high-pressure hydrogen storage tank;

[0099] Extract historical fault data information of high-pressure hydrogen storage tanks caused by gas leakage, and determine the historical gas leakage location and the number of times the same historical gas leakage location occurs;

[0100] According to the historical fault data information of the high-pressure hydrogen storage tank, determine the total number of corresponding historical faults;

[0101] According to the historical leakage data information of the high-pressure hydrogen storage tank, the historical leakage position and the number of times corresponding to the historical leakage position are obtained;

[0102] Divide the number of occurrences of the same historical leak location by the total number of historical faults to obtain the probability value of a fault occurring at the corresponding historical leak location;

[0103] If the probability value of a fault occurring at a historical air leakage location is greater than a preset probability threshold, the corresponding historical air leakage location is marked to obtain a marked historical air leakage location;

[0104] The marked historical air leakage position is sent to a preset management terminal for prompting.

[0105] According to an embodiment of the present invention, for example, if the preset time period is 1 year, historical fault data information of the high-pressure hydrogen storage tank in the most recent year is extracted. There may be many reasons for the failure of the high-pressure hydrogen storage tank. The probability value of the failure at the historical leakage location is calculated based on the historical fault data of the high-pressure hydrogen storage tank caused by weld leakage, and the judgment is made based on the preset probability threshold.

[0106] According to an embodiment of the present invention, after sending the marked historical air leakage position to a preset management terminal for prompting, the method further includes:

[0107] The probability value of a failure at a corresponding historical gas leakage position is set as a safety index adjustment coefficient of the weld at the corresponding position;

[0108] The safety index adjustment coefficient of the corresponding weld is multiplied by the safety index threshold of the corresponding weld to obtain a safety index revision value of the corresponding weld;

[0109] The safety index threshold of the weld is added to the safety index revision value of the corresponding weld to obtain the revised safety index threshold of the corresponding weld.

[0110] It should be noted that the safety index threshold of the weld after revision is less than or equal to the preset standard characteristic value of the weld outer contour, the preset standard characteristic value of the weld size, or the preset standard characteristic value of the weld texture. When the safety index threshold of the weld after revision is greater than the preset standard characteristic value of the weld outer contour, the preset standard characteristic value of the weld size, or the preset standard characteristic value of the weld texture, it is replaced by the preset standard characteristic value of the weld outer contour, the preset standard characteristic value of the weld size, or the preset standard characteristic value of the weld texture.

[0111] Figure 2 A block diagram of a high-pressure hydrogen storage tank weld detection system based on image recognition is shown in the present invention.

[0112] like Figure 2 As shown, the second aspect of the present invention provides a high-pressure hydrogen storage tank weld detection system based on image recognition, including: a parameter setting module, an image acquisition module, an image recognition module, a data storage module and a warning module;

[0113] The parameter setting module is used to set the parameters in the high-pressure hydrogen storage tank weld detection process;

[0114] The image acquisition module is used to acquire the weld image of the high-pressure hydrogen storage tank;

[0115] The image recognition module is used to recognize the weld image of the high-pressure hydrogen storage tank and determine whether the corresponding weld is safe;

[0116] The data storage module is used to store data for testing the weld images of the high-pressure hydrogen storage tank;

[0117] The warning module is used to prompt the preset management terminal.

[0118] It should be noted that various weight parameters are set in advance through the parameter setting module, and then the weld image of the high-pressure hydrogen storage tank is extracted through the image acquisition module, and the weld image is sent to the image recognition module for image processing to determine whether the corresponding weld is safe. If it is safe, the corresponding high-pressure hydrogen storage tank passes the inspection; if it is unsafe, the corresponding high-pressure hydrogen storage tank cannot pass the inspection, and a warning message is generated, which is sent to the warning module to prompt the preset management end.

[0119] According to an embodiment of the present invention, the image acquisition module is further used for:

[0120] Obtain the clarity of the weld image of the high-pressure hydrogen storage tank;

[0121] Determine whether the clarity of the high-pressure hydrogen storage tank weld image is greater than a preset clarity threshold, if so, the corresponding high-pressure hydrogen storage tank weld image is qualified; if not, the corresponding high-pressure hydrogen storage tank weld image is unqualified;

[0122] Based on the fact that the weld seam image of the high-pressure hydrogen storage tank is unqualified, the unqualified weld seam image of the high-pressure hydrogen storage tank is re-photographed.

[0123] It should be noted that when acquiring the weld image, the image acquisition module also preprocesses the weld image, deletes the weld image whose clarity is less than or equal to the preset clarity threshold, and re-takes the weld image at the corresponding weld position.

[0124] According to an embodiment of the present invention, the image recognition module further includes: a unit for extracting weld features and weld feature values, a weld safety index evaluation unit, and a weld safety index determination unit;

[0125] The weld feature and weld characteristic value extraction unit is used to extract the weld feature and the corresponding weld characteristic value;

[0126] The weld safety index evaluation unit is used to calculate the weld safety index evaluation value according to the weld characteristic value;

[0127] The weld safety index determination unit is used to determine the safety index evaluation value of the weld and determine whether the corresponding weld passes the safety test.

[0128] It should be noted that the weld safety index assessment unit is set in the weld safety index assessment model. When the weld safety index assessment value is greater than or equal to the corresponding weld safety index threshold, the corresponding weld passes the safety inspection; when the weld safety index assessment value is less than the corresponding weld safety index threshold, the corresponding weld fails the safety inspection.

[0129] The present invention discloses a high-pressure hydrogen storage tank weld detection method and system based on image recognition, which judges the weld image through a preset weld safety index evaluation model and a weld safety index threshold, thereby improving the accuracy of weld detection and the shelf life of the corresponding weld.

[0130] In the several embodiments provided in the present application, it should be understood that the disclosed devices and methods can be implemented in other ways. The device embodiments described above are only schematic. For example, the division of the units is only a logical function division. There may be other division methods in actual implementation, such as: multiple units or components can be combined, or can be integrated into another system, or some features can be ignored or not executed. In addition, the coupling, direct coupling, or communication connection between the components shown or discussed can be through some interfaces, and the indirect coupling or communication connection of the devices or units can be electrical, mechanical or other forms.

[0131] The units described above as separate components may or may not be physically separated, and the components displayed as units may or may not be physical units; they may be located in one place or distributed on multiple network units; some or all of the units may be selected according to actual needs to achieve the purpose of the present embodiment.

[0132] In addition, all functional units in the embodiments of the present invention may be integrated into one processing unit, or each unit may be separately used as a unit, or two or more units may be integrated into one unit; the above-mentioned integrated units may be implemented in the form of hardware or in the form of hardware plus software functional units.

[0133] A person of ordinary skill in the art can understand that: all or part of the steps of implementing the above method embodiments can be completed by hardware related to program instructions, and the aforementioned program can be stored in a computer-readable storage medium. When the program is executed, it executes the steps of the above method embodiments; and the aforementioned storage medium includes: mobile storage devices, read-only memories (ROM, Read-Only Memory), random access memories (RAM, Random Access Memory), magnetic disks or optical disks, and other media that can store program codes.

[0134] Alternatively, if the above-mentioned integrated unit of the present invention is implemented in the form of a software function module and sold or used as an independent product, it can also be stored in a computer-readable storage medium. Based on this understanding, the technical solution of the embodiment of the present invention can be essentially or partly reflected in the form of a software product that contributes to the prior art. The computer software product is stored in a storage medium and includes several instructions for a computer device (which can be a personal computer, a server, or a network device, etc.) to execute all or part of the methods described in each embodiment of the present invention. The aforementioned storage medium includes: various media that can store program codes, such as mobile storage devices, ROM, RAM, magnetic disks or optical disks.

Claims

1. A high-pressure hydrogen storage tank weld detection method based on image recognition, characterized in that: include: Obtain image information of weld seams of high-pressure hydrogen storage tanks; Extract weld features and corresponding weld feature values ​​from weld images of high-pressure hydrogen storage tanks; The weld characteristic value is sent to a preset weld safety index evaluation model to obtain a safety index evaluation value of the corresponding weld; When the safety index assessment value of a weld is greater than or equal to the safety index threshold of the corresponding weld, the corresponding weld passes the safety test; When the safety index evaluation value of the weld is less than the safety index threshold of the corresponding weld, the position of the corresponding weld is extracted, and the position of the weld is sent to a preset management terminal for display; The weld features include weld texture features, weld size features and weld outer contour features; the weld feature values ​​include weld texture feature values, weld size feature values ​​and weld outer contour feature values.

2. A high-pressure hydrogen storage tank weld detection method based on image recognition according to claim 1, characterized in that: The step of obtaining the weld texture feature value specifically comprises: Obtain the welding process of the corresponding weld and the shape and size of the weld; According to the welding process of the corresponding weld, query in the preset weld shape and size table to obtain the standard shape and size of the corresponding weld; Compare and analyze the shape and size of the weld with the standard shape and size of the corresponding weld to obtain similarity values; The similarity value is multiplied by the weight coefficient of the shape and size of the corresponding weld in the weld texture feature, and then accumulated to obtain the corresponding weld texture feature value.

3. A high-pressure hydrogen storage tank weld detection method based on image recognition according to claim 1, characterized in that: The step of obtaining the weld size characteristic value specifically includes: Extract the dimension elements in the weld dimension features and determine the values ​​of the corresponding dimension elements; Calculate the difference between the value of the dimension element and the preset standard value of the corresponding dimension element to obtain the dimension element value difference; Dividing the size element value difference by the preset standard value of the corresponding size element to obtain a first value of the corresponding size element; Subtract the first value of the corresponding dimension element from the preset dimension element weight reference value to obtain the actual weight value of the corresponding dimension element; The actual weight values ​​of different size elements are multiplied by the preset characteristic values ​​of the corresponding size elements, and then added up to obtain the weld size characteristic value.

4. A high-pressure hydrogen storage tank weld detection method based on image recognition according to claim 1, characterized in that: The step of obtaining the weld outer contour characteristic value specifically comprises: Extract pixel point cloud from the outer contour of the weld; Based on the preset pixel point intervals, the pixel points in the pixel point cloud image in the outer contour line of the weld are divided, and the number of pixel points in each preset pixel point interval is calculated; Extract the maximum number of pixels in the preset pixel interval and set it as the number of pixels on the standard contour line; The number of pixel points on the standard contour line is divided by the total number of pixel points in the pixel point cloud image in the outer contour line of the weld to obtain a second value; The second value is multiplied by a preset standard characteristic value of the weld outer contour to obtain a corresponding weld outer contour characteristic value.

5. The method for detecting weld seams of high-pressure hydrogen storage tanks based on image recognition according to claim 1 is characterized in that: The step of sending the weld characteristic value to a preset weld safety index evaluation model to obtain a safety index evaluation value of the corresponding weld specifically includes: Extracting weld texture feature values, weld size feature values ​​and weld outer contour feature values ​​from weld feature values; The weld seam texture feature value is multiplied by a preset texture feature weight coefficient, the weld seam size feature value is multiplied by a preset size feature weight coefficient, and the weld seam outer contour feature value is multiplied by a preset outer contour feature weight coefficient, and then accumulated to obtain a weld seam safety index assessment value; The sum of the preset texture feature weight coefficient, the preset size feature weight coefficient and the preset outer contour feature weight coefficient is equal to 1.

6. The method for detecting weld seams of high-pressure hydrogen storage tanks based on image recognition according to claim 1 is characterized in that: include: Based on a preset time period, obtain historical fault data information of the high-pressure hydrogen storage tank; Extract historical fault data information of high-pressure hydrogen storage tanks caused by gas leakage, and determine the historical gas leakage location and the number of times the same historical gas leakage location occurs; According to the historical fault data information of the high-pressure hydrogen storage tank, determine the total number of corresponding historical faults; According to the historical leakage data information of the high-pressure hydrogen storage tank, the historical leakage position and the number of times corresponding to the historical leakage position are obtained; Divide the number of occurrences of the same historical leak location by the total number of historical faults to obtain the probability value of a fault occurring at the corresponding historical leak location; If the probability value of a fault occurring at a historical air leakage location is greater than a preset probability threshold, the corresponding historical air leakage location is marked to obtain a marked historical air leakage location; The marked historical air leakage position is sent to a preset management terminal for prompting.

7. A high-pressure hydrogen storage tank weld detection method based on image recognition according to claim 6, characterized in that: After sending the marked historical leakage position to the preset management terminal for prompting, the method further includes: The probability value of a failure at a corresponding historical gas leakage position is set as a safety index adjustment coefficient of the weld at the corresponding position; The safety index adjustment coefficient of the corresponding weld is multiplied by the safety index threshold of the corresponding weld to obtain a safety index revision value of the corresponding weld; The safety index threshold of the weld is added to the safety index revision value of the corresponding weld to obtain the revised safety index threshold of the corresponding weld.

8. A high-pressure hydrogen storage tank weld detection system based on image recognition, characterized in that: It includes: parameter setting module, image acquisition module, image recognition module, data storage module and warning module; The parameter setting module is used to set the parameters in the high-pressure hydrogen storage tank weld detection process; The image acquisition module is used to acquire the weld image of the high-pressure hydrogen storage tank; The image recognition module is used to recognize the weld image of the high-pressure hydrogen storage tank and determine whether the corresponding weld is safe; The data storage module is used to store data for testing the weld images of the high-pressure hydrogen storage tank; The warning module is used to prompt the preset management terminal.

9. A high-pressure hydrogen storage tank weld detection system based on image recognition according to claim 8, characterized in that: The image acquisition module is also used for: Obtain the clarity of the weld image of the high-pressure hydrogen storage tank; Determine whether the clarity of the high-pressure hydrogen storage tank weld image is greater than a preset clarity threshold, if so, the corresponding high-pressure hydrogen storage tank weld image is qualified; if not, the corresponding high-pressure hydrogen storage tank weld image is unqualified; Based on the fact that the weld seam image of the high-pressure hydrogen storage tank is unqualified, the unqualified weld seam image of the high-pressure hydrogen storage tank is re-photographed.

10. A high-pressure hydrogen storage tank weld detection system based on image recognition according to claim 8, characterized in that: The image recognition module further includes: a weld feature and weld feature value extraction unit, a weld safety index evaluation unit, and a weld safety index determination unit; The weld feature and weld characteristic value extraction unit is used to extract the weld feature and the corresponding weld characteristic value; The weld safety index evaluation unit is used to calculate the weld safety index evaluation value according to the weld characteristic value; The weld safety index determination unit is used to determine the safety index evaluation value of the weld and determine whether the corresponding weld passes the safety test.

Images