Method and system for cleaning workpiece before flaw detection

By using the stain identification model to identify and allocate cleaning gears before workpiece flaw detection, combined with the use of dry ice cleaning devices, the problem of waste of dry ice resources in the existing dry ice cleaning devices is solved, achieving more efficient cleaning effects and resource conservation.

CN119870055AActive Publication Date: 2025-04-25SHENSHUO RAILWAY BRANCH CHINA SHENHUA ENERGY
View PDF 5 Cites 0 Cited by

Patent Information

Application Number
CN202411830816.7
Authority / Receiving Office
CN · China
Patent Type
Applications(China)
Current Assignee / Owner
Filing Date
2024-12-12
Publication Date
2025-04-25
Estimated Expiration
2044-12-12

AI Technical Summary

Technical Problem

Existing dry ice cleaning devices will cause a lot of waste of dry ice resources when ensuring adequate cleaning.

Method used

Stain recognition is performed through the stain recognition model, and the corresponding cleaning gear is allocated. Use a dry ice cleaning device to clean it. After each cleaning, the image acquisition and identification model will be used to identify whether there are still stains. If so, increase the cleaning gear and repeat cleaning until there are no stains.

Benefits of technology

A more complete cleaning effect is achieved, while avoiding the waste of dry ice resources and saving the use of dry ice resources.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN119870055A_ABST
    Figure CN119870055A_ABST
Patent Text Reader

Abstract

The invention provides a method for cleaning a workpiece before flaw detection, which comprises the following steps of: during cleaning, performing stain identification through a stain identification model, distributing a corresponding cleaning gear, cleaning by using a dry ice cleaning device, acquiring an image after each time of cleaning, identifying the stain by using the stain identification model, and cleaning an area still having the stain again. And after the surface of the workpiece is free of stains, the purpose of cleaning is achieved. The invention further provides a cleaning system before flaw detection of the workpiece, the cleaning system before flaw detection of the workpiece is applied to the cleaning method before flaw detection of the workpiece, the position of the image acquisition device is adjusted through the eight-axis linkage device, and the position of the dry ice spray gun is also adjusted through the eight-axis linkage device. The dry ice cleaning agent has the beneficial effects that the purpose of removing stains can be achieved, and meanwhile waste of dry ice resources is reduced.
Need to check novelty before this filing date? Find Prior Art

Description

Technical Field

[0001] The invention relates to the technical field of workpiece flaw detection, and in particular to a method and system for cleaning a workpiece before flaw detection. Background Art

[0002] Workpiece flaw detection is to inspect the workpiece through flaw detection equipment to detect cracks or damage in the workpiece. Before flaw detection, the stains on the surface of the workpiece need to be cleaned and removed. Common cleaning methods include boiling in a washing pool, manual high-pressure water gun (spray cleaning agent) flushing, manual steel wool wiping, etc. to remove surface dust, oil, rust and other stains. Among them, high-pressure water gun spraying and boiling in a washing pool both consume a lot of water resources, so some companies have begun to use dry ice cleaning. Through dry ice generating equipment and corresponding spray guns, high-pressure air is used to entrain dry ice particles and spray them onto the surface to be cleaned. The dry ice particles quickly sublimate when they come into contact with the surface of the object, absorbing a large amount of heat, and quickly freezing, brittle, and peeling off the dirt, thereby achieving a cleaning effect.

[0003] However, in order to ensure adequate cleaning, existing dry ice cleaning devices generally use a higher gear for flushing, but this will cause a large amount of dry ice resources to be wasted.

[0004] Therefore, it is necessary to study a cleaning method and system for workpieces before flaw detection to solve the above problems or alleviate the impact of the above problems. Summary of the invention

[0005] On the one hand, the present invention provides a method for cleaning a workpiece before flaw detection, which identifies stains through a stain recognition model, allocates corresponding cleaning gears for cleaning, and learns to improve the compatibility of the cleaning gear with the degree of dirtiness of the stains during the operation, so as to effectively solve the above problems or alleviate the impact of the above problems.

[0006] The cleaning method of the workpiece before flaw detection of the present invention comprises:

[0007] S100. The workpiece is transported to a cleaning station, where the surface of the workpiece is divided into several areas;

[0008] S200. Acquire images of each of the regions on the workpiece by an image acquisition device;

[0009] S300. Identify whether the image has stains and the degree of stains through a stain recognition model, mark the area corresponding to the image with stains as a stain area, and output a corresponding cleaning gear according to the degree of stains;

[0010] S400. Controlling the dry ice cleaning device to perform a cleaning operation, spraying dry ice particles on the stained area according to the cleaning gear;

[0011] S500. After cleaning is completed, the image acquisition device is used to obtain the cleaned images of each of the stained areas on the workpiece;

[0012] S600. Identify whether the cleaned image has stains by using the stain recognition model. If stains are recognized in the cleaned image, the stain area mark of the corresponding area is retained. Otherwise, the stain area mark of the corresponding area is cancelled.

[0013] S700. When all the images after cleaning are identified, if there is no stain area, the next step is executed; if there is a stain area, the cleaning gear is increased by one level based on the last cleaning, and the process returns to step S400;

[0014] S800. Transport the workpiece away from the cleaning station.

[0015] In one embodiment, the stain recognition model outputs the cleaning gear during cleaning by identifying the degree of dirtiness of the stains, the degree of dirtiness is divided into levels 1 to n from light to heavy, and the cleaning gear is divided into levels 1 to m from low to high according to the cleaning intensity, where m is greater than n; when the stain recognition model is created, the levels 1 to n of the degree of dirtiness correspond one to one with the levels 1 to n of the cleaning gears.

[0016] In one embodiment, between step S700 and step S800, the following steps are further included:

[0017] S701. Recording the initial dirtiness of each stained area in step S300;

[0018] S702. Record the cleaning position of each of the stained areas when they were last cleaned;

[0019] S703. The initial dirtiness level corresponds to the cleaning gear position during the last cleaning;

[0020] S704. Record the cleaning gears corresponding to the different levels of dirtiness into the dirt recognition model, overwriting the original corresponding relationship.

[0021] In one embodiment, step S200 specifically includes the following steps:

[0022] S201. Control the eight-axis linkage device to use the image acquisition device;

[0023] S202. Adjust the position of the image acquisition device through the eight-axis linkage device, so that the image acquisition device can capture images of each of the areas.

[0024] In one embodiment, step S400 specifically includes the following steps:

[0025] S401. Control the eight-axis linkage device to use the dry ice spray gun of the dry ice cleaning device;

[0026] S402. Adjust the position of the dry ice spray gun through the eight-axis linkage device so that the dry ice spray gun is directed toward any stain area;

[0027] S403. Controlling the dry ice cleaning device to start according to the cleaning gear corresponding to the stain area, and spraying dry ice particles toward the stain area;

[0028] S404. Repeat steps S402 and S403 to clean all stained areas.

[0029] Another aspect of the present invention further provides a workpiece cleaning system before flaw detection, the cleaning system is applied to the workpiece cleaning method before flaw detection as described above, the cleaning system comprises:

[0030] The processing silent room has a cleaning station in the center;

[0031] Two eight-axis linkage devices, the two eight-axis linkage devices are respectively arranged on both sides of the cleaning station; the eight-axis linkage device comprises a ground rail arranged on one side of the cleaning station, a lifting device is arranged on the slide of the ground rail, a six-axis mechanical arm is arranged on the slide of the lifting device, and a main joint is arranged at the end of the six-axis mechanical arm;

[0032] A dry ice cleaning device, comprising a dry ice generating device and a dry ice spray gun;

[0033] A dustproof device, comprising a dustproof box arranged at one side of the floor rail, a first storage rack being arranged on the outer side of the dustproof box, and a second storage rack being arranged inside the dustproof box; a second quick-change joint is placed on the second storage rack, a second clamp is arranged below the second quick-change joint, and the second clamp clamps an image acquisition device; a first quick-change joint is placed on the first storage rack, a first clamp is arranged at one end of the first quick-change joint, and the first clamp is used to clamp the dry ice spray gun; the second quick-change joint and the first quick-change joint can both be detachably connected to the main joint;

[0034] A central control device is used to control the eight-axis linkage device, the dry ice cleaning device and the image acquisition device.

[0035] In one embodiment, a plug-in plate groove is provided on one side wall of the dustproof box, a sealing plate is inserted in the plug-in plate groove, the extension direction of the sealing plate is a first direction, a reset spring is provided at one end of the sealing plate, a limiting guide groove is provided in the middle of the sealing plate and passes through the second direction, and two limiting guide columns arranged along the first direction are provided at the top of the inside of the dustproof box, and the two limiting guide columns both extend along the second direction and pass through the limiting guide groove.

[0036] In one embodiment, a avoidance groove is provided at the top of the dustproof box, a guiding rubber block is provided at one end of the sealing plate away from the return spring, and the end of the guiding rubber block away from the return spring has two guiding inclined surfaces.

[0037] In one embodiment, a first sealing rubber ring is sleeved on one end of the sealing plate close to the guide rubber block, and the first sealing rubber ring is used to seal the plug-in plate groove; a second sealing rubber ring is provided on the top of the sealing plate, and the second sealing rubber ring is used to seal the avoidance groove.

[0038] In one embodiment, the interior of the dustproof box is connected to the air outlet of a compressed air device, a micro air pump is provided on one side of the dustproof box, the air outlet of the micro air pump is connected to the interior of the dustproof box, and the air inlet of the micro air pump is connected to an external air source.

[0039] The cleaning method for a workpiece before flaw detection provided by the present invention has at least the following beneficial effects compared with the prior art:

[0040] The cleaning method of the workpiece before flaw detection of the present invention uses a stain recognition model to identify stains during cleaning, and allocates corresponding cleaning gears, uses a dry ice cleaning device for cleaning, and collects images after each cleaning and uses the stain recognition model to identify stains, and cleans the area where there are still stains again until there are no stains on the surface of the workpiece to complete the purpose of cleaning. In this way, the cleaning is more complete and no manual re-cleaning is required. In addition, the cleaning method of the present invention consumes more dry ice only during the initial cleaning. Each time the workpiece is cleaned, the cleaning gear corresponding to the current degree of dirtiness will be recorded and the original corresponding relationship in the stain recognition model will be overwritten. When dry ice is used for subsequent cleaning, a more suitable cleaning gear will be provided through the stain recognition model, which can not only achieve the purpose of removing stains, but also will not cause waste of dry ice resources, and can save a large amount of dry ice resources.

[0041] The present invention provides a cleaning system for workpieces before flaw detection. Since it is applied to the above-mentioned cleaning method for workpieces before flaw detection, it also has the above-mentioned beneficial effects. In addition, the cleaning system of the present invention adjusts the position of the image acquisition device through an eight-axis linkage device, which is convenient for taking images of various areas on the surface of the workpiece, and adjusts the position of the dry ice spray gun through the eight-axis linkage device, which is convenient for cleaning stains in various areas; in addition, a second storage rack, a first storage rack, a second quick-change joint, a first quick-change joint and a main joint are provided, so that the end of the eight-axis linkage device can be switched between the image acquisition device and the dry ice spray gun, which is also convenient for the storage of the image acquisition device and the dry ice spray gun when not in use. BRIEF DESCRIPTION OF THE DRAWINGS

[0042] Hereinafter, the present invention will be described in more detail based on embodiments and with reference to the accompanying drawings.

[0043] Figure 1 It is a flow chart of steps of a method for cleaning a workpiece before flaw detection according to an embodiment of the present invention;

[0044] Figure 2 It is a structural schematic diagram of a cleaning system for a workpiece before flaw detection according to an embodiment of the present invention;

[0045] Figure 3 It is a structural schematic diagram of an eight-axis linkage device according to an embodiment of the present invention;

[0046] Figure 4 is a schematic cross-sectional structure diagram of a dustproof box according to an embodiment of the present invention;

[0047] Figure 5 A schematic diagram of a connection structure between a dustproof box and a first storage rack according to an embodiment of the present invention;

[0048] Figure 6 The figure is a schematic structural diagram of a sealing plate according to an embodiment of the present invention.

[0049] In the drawings, the same reference numerals are used for the same components. The drawings are not drawn to scale.

[0050] Reference numerals:

[0051] 1- Processing silent room,

[0052] 2-Cleaning station,

[0053] 3-Dry ice cleaning device, 31-Dry ice generating equipment, 32-Dry ice spray gun,

[0054] 4-eight-axis linkage device, 41-ground rail, 42-lifting device, 43-six-axis robot arm, 44-main joint, 45-first quick-change joint, 46-first fixture, 47-second quick-change joint, 48-second fixture,

[0055] 5-dustproof device, 51-column, 52-dustproof box, 53-insertion plate slot, 54-sealing plate, 55-limiting guide slot,

[0056] 56-limit guide column, 57-reset spring, 58-avoidance groove, 59-second storage rack, 510-first storage rack,

[0057] 511-guiding rubber block, 512-first sealing rubber ring, 513-second sealing rubber ring, 514-micro air pump,

[0058] 6- Central control device,

[0059] 7- Image acquisition device. DETAILED DESCRIPTION

[0060] The present invention will be further described below in conjunction with the accompanying drawings.

[0061] Example 1

[0062] Please refer to Figure 1 to Figure 6 The cleaning method of the workpiece before flaw detection of the present invention comprises:

[0063] S100. The workpiece is transported to the cleaning station 2, and the surface of the workpiece is divided into several areas.

[0064] Among them, the cleaning station 2 is located in the processing silent room 1, which can reduce the noise pollution caused to the outside world. In addition, through the AGV transfer vehicle and the matching workpiece tooling, the workpiece is transported to the designated location, that is, the cleaning station 2, according to a fixed route.

[0065] S200. Acquire images of various regions on the workpiece through the image acquisition device 7 respectively.

[0066] In some examples, step S200 may specifically include the following steps:

[0067] S201 controls the eight-axis linkage device 4, and uses the image acquisition device 7;

[0068] S202. Adjust the position of the image acquisition device 7 through the eight-axis linkage device 4, so that the image acquisition device 7 can capture images of various areas.

[0069] Among them, the image acquisition device 7 uses a 2D camera, which can directly take pictures of various areas as images for subsequent analysis. The eight-axis linkage device 4 is a mechanism that can adjust the position in multiple directions, including a six-axis robotic arm 43, at the end of which a main joint 44 is provided. The image acquisition device 7 is connected to the second quick-change joint 47, and can be connected through the main joint 44 and the second quick-change joint 47 to achieve the purpose of taking the image acquisition device 7.

[0070] S300. Identify whether there are stains in the image and the degree of dirtiness of the stains through a stain recognition model, mark the area corresponding to the image with stains as a stain area, and output a corresponding cleaning gear according to the degree of dirtiness of the stains.

[0071] Among them, the stain recognition model is created based on the AIVI (Artificial Intelligence Vision Inspection) visual algorithm, and its creation method specifically includes the following process:

[0072] W1. Enter photos of various areas of the workpiece; optionally, the photos are taken manually;

[0073] W2. Select a photo and manually select the stain; the stain is used as the recognition feature, and the rest is used as the background;

[0074] W3. The algorithm automatically selects stains from other photos;

[0075] W4. Manually check each photo and eliminate the wrongly selected parts; that is, some parts of the background are close to stains, which will interfere with the recognition of the algorithm, so manual training is required;

[0076] W5. Classify the stains according to their degree of dirtiness, and store the correspondence between each degree of dirtiness and a cleaning gear;

[0077] W6. Manually select a number of photos and mark the stains in the photos according to the degree of dirtiness;

[0078] W7. The algorithm automatically matches the dirtiness level of stains in other photos;

[0079] W8. Manually check the dirt level of each photo and modify the obvious wrong marks;

[0080] After completing the above operations, the stain recognition model is created. The accuracy of stain recognition is improved through the artificially trained model. Furthermore, new photos can be added to the stain recognition model and multiple trainings can be performed to ensure the accuracy of recognition.

[0081] In some examples, the stain recognition model outputs the cleaning gear during cleaning by identifying the degree of dirtiness of the stains. The degree of dirtiness is divided into levels 1 to n from light to heavy, and the cleaning gear is divided into levels 1 to m from low to high according to the cleaning intensity, where m is greater than n. When the stain recognition model is created, the levels 1 to n of dirtiness correspond one to one with the levels 1 to n of cleaning gears.

[0082] Specifically, the degree of dirtiness is set from light to heavy as level 1, level 2, ..., level n, and the cleaning gear is set from low to high as level 1, level 2, ..., level n, ..., level m. When the stain recognition model is created, the level 1, level 2, ..., level n of the degree of dirtiness correspond to level 1, level 2, ..., level n, respectively. When the stain recognition model recognizes that there are stains in the image and matches the corresponding degree of dirtiness, it will further output the corresponding cleaning gear. The dry ice cleaning device 3 will clean the area corresponding to the image according to the cleaning gear. In addition, the cleaning gear includes the amount of dry ice, the injection pressure and the dry ice flow rate. The higher the gear, the larger the value of the above parameters, and the better the cleaning effect.

[0083] S400. Control the dry ice cleaning device 3 to perform a cleaning operation, and spray dry ice particles onto the stained area according to the cleaning gear.

[0084] In some examples, step S400 specifically includes the following steps:

[0085] S401 controls the eight-axis linkage device 4, and uses the dry ice spray gun 32 of the dry ice cleaning device 3;

[0086] S402. Adjust the position of the dry ice spray gun 32 through the eight-axis linkage device 4, so that the dry ice spray gun 32 is directed toward any stain area;

[0087] S403. Control the dry ice cleaning device 3 to start according to the cleaning gear corresponding to the stain area, and spray dry ice particles to the stain area;

[0088] S404. Repeat steps S402 and S403 to clean all stained areas.

[0089] Before S401, the ends of the eight-axis linkage device 4 drive the image acquisition device 7 to adjust the position through the main joint 44 and the second quick-change joint 47. When taking the dry ice spray gun 32, it is necessary to first separate the main joint 44 and the second quick-change joint 47, and then connect the main joint 44 and the first quick-change joint 45, and then connect the dry ice spray gun 32 to the first quick-change joint 45 to complete the use of the dry ice spray gun 32; and according to the dry ice usage parameters, injection pressure parameters and dry ice flow parameters provided by the cleaning gear, the purpose of cleaning stains is achieved.

[0090] S500. After cleaning is completed, the images of each stained area on the workpiece after cleaning are obtained through the image acquisition device 7. The acquisition method is the same as step S200, but before using the image acquisition device 7, it is necessary to first separate the main joint 44 and the first quick-change joint 45, and connect the main joint 44 and the second quick-change joint 47 to complete the storage of the dry ice spray gun 32 and the use of the image acquisition device 7.

[0091] S600. Use the stain recognition model to identify whether there are stains in the cleaned image. If stains are identified in the cleaned image, the corresponding area retains the stain area mark, otherwise the stain area mark of the corresponding area is cancelled. The principle is the same as step S300, that is, the cleaned image is recognized again by the stain recognition model, and the cleaned image with stains is identified, and the area corresponding to the cleaned image is output. These areas still retain the stain area mark, and continue to be cleaned.

[0092] S700. After all the images have been cleaned and recognized, if there is no stain area, the next step is executed; if there is a stain area, the cleaning level is increased by one level based on the last cleaning, and the process returns to step S400. If there is no stain area, it means that all stains have been cleaned, that is, the cleaning of the workpiece is completed, and if there is a stain area, it needs to be cleaned again.

[0093] In some examples, between step S700 and step S800, the following steps are also included:

[0094] S701. Recording the initial dirtiness of each stained area in step S300;

[0095] S702. Record the cleaning position of each stain area when it was last cleaned;

[0096] S703. The initial dirtiness level corresponds to the cleaning gear position during the last cleaning;

[0097] S704. Record the cleaning gears corresponding to the different levels of dirtiness into the stain recognition model, overwriting the original corresponding relationship.

[0098] In short, if a stain is located at level 1, the first cleaning will be carried out according to the cleaning gear output by the stain recognition model, that is, gear 1. However, because the cleaning is not clean, the stain area will be retained and the cleaning gear will be increased, that is, assigned to gear 2. Here, step S400 will be returned for cleaning. The dry ice cleaning device 3 will clean the stain area again according to gear 2. After step S600, it will be tested again. If it is not clean, the cleaning gear will be increased again, and step S400 will be returned until it is clean. Therefore, the cleaning gear of the last cleaning is matched with the level of the initial degree of dirtiness, covering the corresponding relationship in the stain recognition model. In this way, learning will be carried out and the stain recognition model will be updated every time cleaning is carried out. Except for the initial cleaning of the workpiece, which will consume more dry ice, the cleaning gear will be optimized in this way in the future to ensure that each level of dirtiness corresponds to the optimal cleaning gear, which can achieve the purpose of removing stains without wasting dry ice resources.

[0099] S800. Transport the workpiece away from the cleaning station 2; also transport the workpiece out of the processing silent room 1 by the AGV transfer vehicle.

[0100] In summary, the beneficial effects of the workpiece cleaning method before flaw detection of the present invention include:

[0101] When cleaning, stains are identified through the stain recognition model, and the corresponding cleaning gear is assigned. The dry ice cleaning device is used for cleaning, and images are collected after each cleaning and the stains are identified using the stain recognition model. The area with stains is cleaned again until there are no stains on the surface of the workpiece to complete the cleaning purpose. In this way, the cleaning is more complete and no manual cleaning is required. In addition, the cleaning method of the present invention consumes more dry ice only during the initial cleaning. Each time the workpiece is cleaned, the cleaning gear corresponding to the current degree of dirtiness will be recorded and the original corresponding relationship in the stain recognition model will be overwritten. When dry ice is used for subsequent cleaning, a more suitable cleaning gear will be provided through the stain recognition model, which can not only achieve the purpose of removing stains, but also will not cause waste of dry ice resources, and can save a large amount of dry ice resources.

[0102] Example 2

[0103] Based on the cleaning method steps of the above embodiment 1, Figure 2-Figure 6 As shown, the present invention provides a cleaning system for workpieces before flaw detection to which the above cleaning method can be applied, comprising:

[0104] Processing silent room 1, with a cleaning station 2 at the center;

[0105] Two eight-axis linkage devices 4 are respectively arranged on both sides of the cleaning station 2; the eight-axis linkage device 4 includes a ground rail 41 arranged on one side of the cleaning station 2, a lifting device 42 is arranged on the slide of the ground rail 41, a six-axis mechanical arm 43 is arranged on the slide of the lifting device 42, and a main joint 44 is arranged at the end of the six-axis mechanical arm 43;

[0106] A dry ice cleaning device 3, comprising a dry ice generating device 31 and a dry ice spray gun 32;

[0107] The dustproof device 5 comprises a dustproof box 52 arranged on one side of the ground rail 41, a first storage rack 510 is arranged on the outer side of the dustproof box 52, and a second storage rack 59 is arranged inside the dustproof box 52; the second storage rack 59 can be placed with a second quick-change joint 47, a second clamp 48 is arranged below the second quick-change joint 47, and the second clamp 48 clamps the image acquisition device 7; the first storage rack 510 can be placed with a first quick-change joint 45, a first clamp 46 is arranged at one end of the first quick-change joint 45, and the first clamp 46 is used to clamp the dry ice spray gun 32; the second quick-change joint 47 and the first quick-change joint 45 can be detachably connected to the main joint 44;

[0108] The central control device 6 is used to control the eight-axis linkage device 4, the dry ice cleaning device 3 and the image acquisition device 7.

[0109] like Figure 2 and Figure 3 As shown, the central control device 6 is a combination of an existing common computer and a programmable logic controller, which is a common control device and will not be described in detail here. It can control the movement of the slide of the ground rail 41, control the movement of the slide of the lifting device 42, control the position adjustment of the main joint 44 of the six-axis robot 43, control the dry ice cleaning device 3 to start according to the corresponding cleaning gear, control the image acquisition device 7 to take pictures of the surface of the workpiece, and control the main joint 44 to be connected with the second quick-change joint 47 or the first quick-change joint 45. The automatic connection of the quick-change joint is an existing technology, mostly electromagnetic adsorption, which will not be described in detail here.

[0110] In addition, if Figure 2As shown, the dry ice generating device 31 of the dry ice cleaning device 3 is arranged outside the processing silent room 1, which is convenient for the independent replacement and maintenance of the dry ice generating device 31. Furthermore, a dust collector is arranged inside the processing silent room 1 to discharge the dust and carbon dioxide formed after the cleaning inside the processing silent room 1 to the outside, so that workers can enter the processing silent room 1 for maintenance or other operations. A column 51 is arranged at the bottom end of the dustproof box 52, and the bottom end of the column 51 is connected to the ground of the processing silent room 1 to realize the installation of the dustproof box 52, so as to avoid placing it directly on the ground too far away from the main joint 44, which is inconvenient for subsequent operations.

[0111] In some examples, such as Figure 4 and Figure 5 As shown, a plug-in plate groove 53 is provided on one side wall of the dustproof box 52, a sealing plate 54 is inserted in the plug-in plate groove 53, the extension direction of the sealing plate 54 is the first direction, a return spring 57 is provided at one end of the sealing plate 54, a limiting guide groove 55 is provided in the middle of the sealing plate 54 and passes through the second direction, and two limiting guide columns 56 arranged along the first direction are provided at the top end of the dustproof box 52, and the two limiting guide columns 56 both extend along the second direction and pass through the limiting guide groove 55.

[0112] Specifically, Figure 4 and Figure 5 As shown, when the image acquisition device 7 needs to be taken during the cleaning operation, the main joint 44 is adjusted to the first storage rack 510 by the six-axis robot 43, and then the main joint 44 is controlled to separate from the first quick-change joint 45. Then the six-axis robot 43 adjusts the main joint 44 to move to the plug-in plate slot 53, and the sealing plate 54 is pushed by the main joint 44. The sealing plate 54 is restricted by two limiting guide pillars 56. The sealing plate 54 can move to the left toward the inside of the dustproof box 52 and compress the reset spring 57. When the main joint 44 moves to the second storage rack 59, the main joint 44 is controlled to be connected to the second quick-change joint 47 to complete the use of the image acquisition device 7.

[0113] In some examples, such as Figure 4 and Figure 5 As shown, a avoidance groove 58 is provided at the top of the dustproof box 52, a guiding rubber block 511 is provided at one end of the sealing plate 54 away from the return spring 57, and the end of the guiding rubber block 511 away from the return spring 57 has two guiding inclined surfaces.

[0114] Specifically, Figure 4 and Figure 5As shown, one guiding slope of the guiding rubber block 511 extends from the lower end to the middle, and the other guiding slope extends from the upper end to the middle toward the side away from the sealing plate 54. In this way, the six-axis robot 43 connected to the main joint 44 can be avoided through the avoidance groove 58 to avoid interfering with the movement of the main joint 44. When taking, after the main joint 44 is connected to the second quick-change joint 47, the main joint 44, the second quick-change joint 47 and the image acquisition device 7 are pulled upward. Under the action of the lower half of the guiding slope, the second quick-change joint 47 or the image acquisition device 7 will not be stuck, ensuring the normal use of the second quick-change joint 47 and the image acquisition device 7. When storing, the sealing plate 54 is pushed to move by the image acquisition device 7, and moves downward when it moves to the position of the second storage rack 59, and it is also affected by the upper half of the guiding slope to ensure the stability of storage.

[0115] In some examples, such as Figures 4 to 6 As shown, a first sealing rubber ring 512 is sleeved on one end of the sealing plate 54 close to the guiding rubber block 511 , and the first sealing rubber ring 512 is used to seal the plug-in plate groove 53 ; a second sealing rubber ring 513 is provided on the top of the sealing plate 54 , and the second sealing rubber ring 513 is used to seal the avoidance groove 58 .

[0116] Specifically, Figures 4 to 6 As shown, after the main joint 44 is away from the dustproof box 52, the sealing plate 54 is reset by the reset spring 57, and then the first sealing rubber ring 512 abuts against the inner wall of the plug-in plate groove 53, and the second sealing rubber ring 513 abuts against the inner top surface of the dustproof box 52 and surrounds the avoidance groove 58, so as to reduce the external dust from entering the dustproof box 52, thereby achieving a good dustproof effect.

[0117] In some examples, the interior of the dustproof box 52 is connected to the air outlet of the compressed air device, a micro air pump 514 is provided on one side of the dustproof box 52, the air outlet of the micro air pump 514 is connected to the interior of the dustproof box 52, and the air inlet of the micro air pump 514 is connected to an external air source.

[0118] Among them, the compressed air device is a commonly used air source supply equipment group in the factory. It provides pressurized gas to each workshop through a compressor and other machines. Each workshop uses the pressurized gas as the air source for pneumatic devices such as cylinders or pneumatic claws. The cleaning system of the present invention introduces the pressurized gas supplied by the compressed air device into the dustproof box 52 to ensure that the dustproof box 52 maintains a slightly positive pressure state, and external dust does not enter the dustproof box 52, thereby improving the dustproof effect. Further, if Figure 3 As shown, when the compressed air device is under maintenance or the air supply is temporarily stopped, the micro air pump 514 can be activated to fill the dustproof box 52 with a proper amount of gas so that the inside of the dustproof box 52 remains in a slightly positive pressure state to avoid a reduction in the dustproof effect.

[0119] In summary, the beneficial effects of the workpiece cleaning system before flaw detection of the present invention include:

[0120] The cleaning system of the present invention adjusts the position of the image acquisition device through an eight-axis linkage device, which is convenient for capturing images of various areas on the surface of the workpiece, and adjusts the position of the dry ice spray gun through the eight-axis linkage device, which is convenient for cleaning stains in various areas; in addition, a second storage rack, a first storage rack, a second quick-change joint, a first quick-change joint and a main joint are provided, so that the end of the eight-axis linkage device can be switched between the image acquisition device and the dry ice spray gun, which is also convenient for storing the image acquisition device and the dry ice spray gun when not in use. Furthermore, the cleaning system of the present invention seals the opening of the dustproof box through a sealing plate, a first sealing rubber ring and a second sealing rubber ring to prevent external dust from entering the dustproof box, and to prevent dust from adhering to the surface of the image acquisition device, resulting in unclear images being captured. Furthermore, the cleaning system of the present invention is provided with and by filling an appropriate amount of air into the dustproof box, further preventing external dust from entering the interior of the dustproof box, thereby achieving a higher dustproof effect.

[0121] Although the present invention has been described with reference to preferred embodiments, various modifications may be made thereto and parts thereof may be replaced by equivalents without departing from the scope of the present invention. In particular, the various technical features mentioned in the various embodiments may be combined in any manner as long as there are no structural conflicts. The present invention is not limited to the specific embodiments disclosed herein, but includes all technical solutions falling within the scope of the claims.

Claims

1. A method for cleaning a workpiece before flaw detection, characterized in that: The cleaning method comprises: S100. The workpiece is transported to a cleaning station, where the surface of the workpiece is divided into several areas; S200. Acquire images of each of the regions on the workpiece by an image acquisition device; S300. Identify whether the image has stains and the degree of stains through a stain recognition model, mark the area corresponding to the image with stains as a stain area, and output a corresponding cleaning gear according to the degree of stains; S400. Controlling the dry ice cleaning device to perform a cleaning operation, spraying dry ice particles on the stained area according to the cleaning gear; S500. After cleaning is completed, the image acquisition device is used to obtain the cleaned images of each of the stained areas on the workpiece; S600. Identify whether the cleaned image has stains by using the stain recognition model. If stains are recognized in the cleaned image, the stain area mark of the corresponding area is retained. Otherwise, the stain area mark of the corresponding area is cancelled. S700. When all the images after cleaning are identified, if there is no stain area, the next step is executed; if there is a stain area, the cleaning gear is increased by one level based on the last cleaning, and the process returns to step S400; S800. Transport the workpiece away from the cleaning station.

2. The method for cleaning a workpiece before flaw detection according to claim 1, characterized in that: The stain recognition model outputs the cleaning gear during cleaning by identifying the degree of dirtiness of the stains, the degree of dirtiness is divided into levels 1 to n from light to heavy, and the cleaning gear is divided into levels 1 to m from low to high according to the cleaning intensity, where m is greater than n; when the stain recognition model is created, the levels 1 to n of the degree of dirtiness correspond one to one with the levels 1 to n of the cleaning gear.

3. The method for cleaning a workpiece before flaw detection according to claim 1, characterized in that: The following steps are also included between step S700 and step S800: S701. Recording the initial dirtiness of each stained area in step S300; S702. Record the cleaning position of each of the stained areas when they were last cleaned; S703. The initial dirtiness level corresponds to the cleaning gear position during the last cleaning; S704. Record the cleaning gears corresponding to the different levels of dirtiness into the dirt recognition model, overwriting the original corresponding relationship.

4. The method for cleaning a workpiece before flaw detection according to claim 1, characterized in that: The step S200 specifically includes the following steps: S201. Control the eight-axis linkage device to use the image acquisition device; S202. Adjust the position of the image acquisition device through the eight-axis linkage device, so that the image acquisition device can capture images of each of the areas.

5. The method for cleaning a workpiece before flaw detection according to claim 4, characterized in that: The step S400 specifically includes the following steps: S401. Control the eight-axis linkage device to use the dry ice spray gun of the dry ice cleaning device; S402. Adjust the position of the dry ice spray gun through the eight-axis linkage device so that the dry ice spray gun is directed toward any stain area; S403. Control the dry ice cleaning device to start according to the cleaning gear corresponding to the stain area, and spray dry ice particles toward the stain area; S404. Repeat steps S402 and S403 to clean all stained areas.

6. A cleaning system for workpieces before flaw detection, characterized in that: The cleaning system is applied to the cleaning method of a workpiece before flaw detection according to any one of claims 1 to 5, and the cleaning system comprises: The processing silent room has a cleaning station in the center; Two eight-axis linkage devices, the two eight-axis linkage devices are respectively arranged on both sides of the cleaning station; the eight-axis linkage device comprises a ground rail arranged on one side of the cleaning station, a lifting device is arranged on the slide of the ground rail, a six-axis mechanical arm is arranged on the slide of the lifting device, and a main joint is arranged at the end of the six-axis mechanical arm; A dry ice cleaning device, comprising a dry ice generating device and a dry ice spray gun; A dustproof device, comprising a dustproof box arranged at one side of the floor rail, a first storage rack being arranged on the outer side of the dustproof box, and a second storage rack being arranged inside the dustproof box; a second quick-change joint is placed on the second storage rack, a second clamp is arranged below the second quick-change joint, and the second clamp clamps an image acquisition device; a first quick-change joint is placed on the first storage rack, a first clamp is arranged at one end of the first quick-change joint, and the first clamp is used to clamp the dry ice spray gun; the second quick-change joint and the first quick-change joint can both be detachably connected to the main joint; A central control device is used to control the eight-axis linkage device, the dry ice cleaning device and the image acquisition device.

7. The cleaning system for workpieces before flaw detection according to claim 6, characterized in that: A plug-in plate groove is provided on one side wall of the dustproof box, a sealing plate is inserted in the plug-in plate groove, the extending direction of the sealing plate is the first direction, a reset spring is provided at one end of the sealing plate, a limiting guide groove is provided in the middle of the sealing plate and passes through the second direction, and two limiting guide columns arranged along the first direction are provided at the top end of the dustproof box, and the two limiting guide columns both extend along the second direction and pass through the limiting guide groove.

8. The cleaning system for workpieces before flaw detection according to claim 7, characterized in that: The top of the dustproof box is provided with an avoidance groove, the end of the sealing plate away from the return spring is provided with a guide rubber block, and the end of the guide rubber block away from the return spring has two guide inclined surfaces.

9. The cleaning system for workpieces before flaw detection according to claim 8, characterized in that: A first sealing rubber ring is sleeved on one end of the sealing plate close to the guiding rubber block, and the first sealing rubber ring is used to seal the plug-in plate groove; a second sealing rubber ring is provided on the top of the sealing plate, and the second sealing rubber ring is used to seal the avoidance groove.

10. The cleaning system for workpieces before flaw detection according to claim 6, characterized in that: The interior of the dustproof box is connected to the air outlet of the compressed air device. A micro air pump is provided on one side of the dustproof box. The air outlet of the micro air pump is connected to the interior of the dustproof box, and the air inlet of the micro air pump is connected to an external air source.

Citation Information

Patent Citations

  • Dry ice cleaning control method for power transformation post insulator dry ice cleaning robot

    CN112354946A

  • Surface cleaning method and device

    CN112452998A

  • Laser device for laser cleaning

    CN114749429A

  • Motor train unit surface cleaning method based on machine learning and data driving digital twin platform

    CN116842830A

  • Laser cleaning apparatus and laser cleaning method therefor

    WO2021196421A1