Deep learning hydraulic metal structure corrosion detection method

Abstract

The invention relates to the technical field of hydraulic protection, in particular to a deep learning hydraulic metal structure corrosion detection method which comprises the following steps: S1, pretreatment: preparing an anti-rust coating, a fluorescent agent and a comprehensive treatment device, and preparing a mixing agent; a comprehensive treatment device is used for moving on the hydraulic metal surface, and meanwhile a mixing agent is smeared; a comprehensive processing device is used for collecting the hydraulic metal surface image smeared with the mixing agent, and an initial image is obtained; s2, fluorescence reflection detection and corrosion analysis, wherein in the using process of the hydraulic metal, a comprehensive treatment device is used for moving on the surface of the hydraulic metal; meanwhile, a comprehensive processing device is used for collecting a current hydraulic metal surface image, and a current image is obtained; and S3, rusting treatment: after the comprehensive treatment device judges the rusted area, the comprehensive treatment device scrapes rusted objects on the surface of the area and stores the rusted objects. According to the invention, the corrosion detection efficiency can be effectively improved.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of hydraulic protection, and in particular to a deep learning method for detecting corrosion of a hydraulic metal structure. BACKGROUND

[0002] Hydraulic metal structures (such as gates, pipelines, dam support structures, etc.) are exposed to a humid and corrosive environment for a long time, and are prone to corrosion problems. Corrosion not only reduces the strength of the structure, but also can cause safety hazards (such as gate failure, pipeline leakage), threatening the safety of water conservancy projects. In the traditional technology, the surface changes of the hydraulic metal structure are observed by manual visual observation to determine the corrosion condition of the hydraulic metal structure, which is low in efficiency.

[0003] The existing method for detecting corrosion of a hydraulic metal structure mainly collects the surface image of the hydraulic metal by an image collection device, and then constructs a detection model using deep learning technology to determine whether the pipeline surface is corroded. However, due to the complexity of the surface corrosion image of the hydraulic metal, the construction of the detection model is difficult, and the recognition process is complex, resulting in low detection efficiency and detection accuracy.

[0004] In summary, how to solve the problem of the existing method for detecting corrosion of a hydraulic metal structure, which has a large construction difficulty of the detection model and a complex recognition process, resulting in low detection efficiency and detection accuracy, has become a difficult problem to be solved in the field. Therefore, it is necessary to propose a more reasonable deep learning method for detecting corrosion of a hydraulic metal structure. SUMMARY

[0005] To solve the above problems, the present application provides a deep learning method for detecting corrosion of a hydraulic metal structure. By designing a mixed agent and improving the detection method, the corrosion area on the surface of the hydraulic metal can be quickly and accurately determined according to the fluorescence reflection while protecting the hydraulic metal, greatly reducing the detection difficulty and improving the detection efficiency and accuracy.

[0006] In order to achieve the above purpose, the technical scheme of the present application is as follows: a deep learning method for detecting corrosion of a hydraulic metal structure, comprising the following steps:

[0007] S1, pretreatment: preparing an anti-rust coating, a fluorescent agent and a comprehensive treatment device, mixing the anti-rust coating and the fluorescent agent by using the comprehensive treatment device to obtain a mixed agent; moving the comprehensive treatment device on the surface of the hydraulic metal while applying the mixed agent; and collecting the image of the surface of the hydraulic metal after applying the mixed agent by using the comprehensive treatment device to obtain an initial image.

[0008] S2, fluorescence reflection detection and corrosion analysis: during the use of the hydraulic metal, the comprehensive treatment device is used to move on the surface of the hydraulic metal; at the same time, the comprehensive treatment device is used to collect the current image of the surface of the hydraulic metal to obtain a current image, and the comprehensive treatment device is used to compare the current image with the initial image; the comprehensive treatment device identifies the corrosion area and the non-corrosion area on the surface of the hydraulic metal based on the reflected light of the fluorescent agent; if the reflected light of the fluorescent agent exists on the surface of the hydraulic metal, it indicates that the mixed agent in the area has not fallen off, and the comprehensive treatment device determines that the area has not been corroded; if the reflected light of the fluorescent agent is not identified on the surface of the hydraulic metal, it indicates that the mixed agent in the area has fallen off, and the comprehensive treatment device determines that the area has been corroded.

[0009] S3, corrosion treatment: when the comprehensive treatment device determines the corrosion area, the comprehensive treatment device scrapes the corrosion on the surface of the area and stores it; at the same time, new mixed agent is sprayed to the area; the comprehensive treatment device determines the corrosion severity according to the current image, and adjusts the spraying amount of the mixed agent according to the corrosion severity; the spraying amount of the mixed agent is proportional to the corrosion severity.

[0010] The above scheme has the following beneficial effects:

[0011] 1. In the prior art, although a detection model can be constructed based on deep learning technology, due to the complexity of the corrosion image of the surface of the hydraulic metal, a large number of corrosion images are needed to train the detection model to obtain a qualified detection model; resulting in difficulty in model construction, low detection efficiency, and poor detection accuracy if the training is insufficient; in the present application, the fluorescent agent is mixed in the anti-rust coating and then applied to the surface of the hydraulic metal; if the surface of the hydraulic metal is corroded, the anti-rust coating and the fluorescent agent in the area will fall off together, and the area will not produce fluorescent reflection; compared with the prior art, the present method is simpler, and only needs to determine whether there is fluorescent reflection to analyze whether the surface of the hydraulic metal is corroded, avoiding the interference of complex environment; thus, the detection difficulty is greatly reduced, thereby improving the detection efficiency and accuracy.

[0012] 2. In the prior art, the hydraulic metal needs to be detected and analyzed first, and then protected by other equipment and subsequent operation after obtaining the detection analysis result, resulting in untimely protection; in the present application, the protection and detection of the hydraulic metal are carried out synchronously, compared with the prior art, the hydraulic metal with corrosion can be protected in time, avoiding the corrosion of the corrosion area due to untimely treatment, resulting in more serious corrosion.

[0013] 3、The method compares the initial image of the water conservancy metal with the current image, quantifies the severity of the corrosion according to the change amount of the fluorescent reflection missing area in the initial image and the current image, and dynamically adjusts the spraying amount of the mixed agent according to the severity of the corrosion, thereby effectively improving the pertinence of protection and the utilization rate of the mixed agent.

[0014] Further, in S1, the comprehensive treatment device comprises a corrosion treatment system for detecting and treating corrosion and a wall-climbing robot, and the wall-climbing robot is provided with a mixing assembly for preparing a mixed agent, a conveying assembly for conveying the mixed agent, a spraying assembly for spraying the mixed agent, and a smearing assembly for smearing the mixed agent.

[0015] The side wall of the wall-climbing robot is provided with a mounting groove, and a mounting seat is fixedly connected to the inner side wall of the mounting groove. The mounting seat is provided with a driving assembly for driving the mixed assembly, the conveying assembly, and the smearing assembly to operate together.

[0016] The smearing assembly comprises a center rod rotationally fitted to the outer side wall of the mounting seat, a center plate fixedly sleeved on the center rod, and limit grooves symmetrically formed in the center plate. The side wall of the mounting seat is symmetrically rotationally fitted with rotating rods, and the rotating rods are fixedly connected with connecting rods. The ends of the connecting rods away from the rotating rods are fixedly connected with limit rods, and the limit rods are rotationally fitted with the limit grooves adjacent thereto. The ends of the rotating rods away from the mounting seat are fixedly connected with smearing discs, and the smearing discs are located outside the mounting groove. The center rod is provided with a scraping assembly for scraping corrosion. The limit rods are provided with a collecting assembly for collecting the surface image of the water conservancy metal.

[0017] Beneficial effects: The collecting assembly collects the surface image of the water conservancy metal as the wall-climbing robot moves vertically along the surface of the water conservancy metal. During the sliding process, the mixed assembly mixes the fluorescent agent with the anti-rust coating to prepare the mixed agent. The conveying assembly conveys the mixed agent to the spraying assembly. The spraying assembly sprays the mixed agent to the surface of the water conservancy metal.

[0018] During the movement of the wall-climbing robot, the rotating rods rotate under the driving of the driving assembly, thereby driving the smearing discs to rotate. Thus, the mixed agent is evenly smeared on the surface of the water conservancy metal, and the scraping assembly also scrapes and collects the corrosion on the surface of the water conservancy metal.

[0019] Further, the scraping assembly comprises an extension rod fixedly connected to the end of the center rod away from the mounting seat, and a scraper coaxially fixedly connected to the output shaft of the extension rod. The corrosion treatment system is used to control the operation of the extension rod, thereby adjusting the position of the scraper. The outer side wall of the wall-climbing robot is fixedly connected with a collecting frame. The collecting frame is located below the scraper, and a magnet piece is fixedly connected in the collecting frame.

[0020] Beneficial effects: when the rust treatment system identifies that there is rust on the water conservancy metal surface, the rust treatment system will control the telescopic rod to start, and move the scraper to the water conservancy metal surface. When the rotating rod rotates, the rotating rod will drive the connecting rod and the limiting rod to rotate. The limiting rod will repeatedly enter the limiting groove during rotation, thereby driving the center plate and the center rod to reciprocate. The center rod will drive the telescopic rod and the scraper to rotate, thereby scraping the rust to the collecting box. The rust falling into the collecting box will be attracted by the magnet piece for further detection by the operator.

[0021] Further, the mixing assembly comprises a plurality of mixing cavities formed in the mounting seat; each of the mixing cavities is communicated with an input pipe for inputting the rust-proof coating and the fluorescent agent; each of the rotating rods extends into the mixing cavity adjacent to the rotating rod and is in rotational cooperation with the inner side wall of the mixing cavity; each of the rotating rods in the mixing cavity is fixedly connected with a plurality of stirring rods; the inner side wall of one of the mixing cavities is fixedly connected with a driving member, and the output shaft of the driving member is coaxially fixedly connected with the rotating rod adjacent to the driving member; the rust treatment system is used for controlling the driving member to operate, thereby driving the rotating rod to rotate.

[0022] Beneficial effects: the rust-proof coating and the fluorescent agent are input into the mixing cavity through the input pipe; the rust treatment system controls the driving member to operate, thereby driving the output shaft of the driving member to rotate the rotating rod, and driving the stirring rod to rotate, thereby stirring and mixing the rust-proof coating and the fluorescent agent in the mixing cavity, so as to prepare the mixed agent for subsequent detection.

[0023] Further, the driving assembly comprises a driving cavity formed in the mounting seat; the driving cavity is located on the side of the mixing cavity away from the driving member; the inner side wall of the driving cavity is symmetrically rotatably connected with a first gear and a second gear, the first gear is in meshing cooperation with the second gear adjacent to the first gear; the adjacent second gears are in meshing cooperation with each other; the rotating rods all penetrate through the driving cavity and are in rotational cooperation with the side wall of the driving cavity; the rotating rods are all coaxially fixedly connected with the first gears adjacent to the rotating rods.

[0024] Beneficial effects: when the driving member drives one of the rotating rods to rotate, the rotating shaft drives the first gear connected with the rotating shaft to rotate, the first gear drives one of the second gears to rotate, the second gear drives the other second gear to rotate, and the other second gear drives the other first gear meshing with the other second gear to rotate, thereby driving the rotating shafts on both sides to synchronously rotate.

[0025] Further, the spraying assembly comprises a storage box fixedly connected to the outer side wall of the wall-climbing robot, the inner side wall of the storage box is vertically slidably connected with a piston plate, the top of the piston plate is hingedly connected with a hinge rod, and the top of the hinge rod is hingedly connected with a center plate; the top of the storage box is provided with a moving groove for the movement of the hinge rod; the side wall of the storage box is provided with a plurality of spraying openings; the spraying openings are all communicated with control valves; the rust treatment system is used for controlling the control valves to operate, thereby controlling the spraying of the mixed agent.

[0026] Beneficial effects: when the center plate reciprocating rotation under the push of the limiting rod, the center plate will drive the articulated rod in the moving groove; if you need to spray the mixture, the corrosion treatment system will open the control valve, when the center plate is downward, it will push the articulated rod and the piston plate to move downward, so as to spray the mixture stored inward through the spray port.

[0027] Further, the conveying assembly comprises a conveying cavity opened in the mounting seat, the conveying cavity is located between the mixing cavities; the mixing cavities are in communication with the conveying cavity; the inner side wall of the conveying cavity is symmetrically rotationally fitted with conveying teeth, adjacent conveying teeth are meshed with each other; the conveying cavity is in communication with the storage tank; the second gear is fixedly connected with a transmission rod near one side of the conveying cavity, the transmission rod extends into the conveying cavity and is fixedly connected with the conveying teeth adjacent thereto coaxially at the far end from the second gear.

[0028] Beneficial effects: when the second gear rotates, the second gear will drive the transmission rod to rotate, and in turn drive the conveying teeth to rotate together; by the principle of gear pump, when the gear slot above the conveying teeth changes from the meshing state to the separated state, a local vacuum is formed, which sucks the mixture in the mixing cavity; when the gear slot below the conveying teeth changes from the separated state to the meshing state, the volume of the area below the conveying teeth decreases, and the mixture in the conveying cavity is extruded into the storage tank by the conveying teeth; and since the conveying cavity is located above the storage tank, the conveying of the mixture will be more smooth due to the influence of gravity.

[0029] Further, the collecting assembly comprises a plurality of cameras, the limiting rods extend out of the mounting groove at the ends away from the connecting rods and are fixedly connected with the cameras adjacent thereto, and the corrosion treatment system is used for controlling the operation of the cameras.

[0030] Beneficial effects: when the limiting rod rotates, it will also drive the cameras to rotate, so that the cameras can collect more comprehensive images.

[0031] Further, a protective frame for protecting the rotating rods and the center rod is fixedly connected outside the mounting groove; the protective frame is provided with a telescopic slot for the telescopic rod to extend and retract and a plurality of application slots for the application discs to rotate; and a application plate for further applying the mixture is fixedly connected to the outer side wall of the protective frame.

[0032] Beneficial effects: the protective frame can protect each component in the mounting groove, improving the stability of the device; when the wall climbing robot moves, the application plate can further apply the mixture applied by the application discs, so that the mixture on the surface of the hydraulic metal is more uniform.

[0033] Further, the corrosion treatment system comprises a data acquisition module, a data analysis module and an intelligent control module.

[0034] The data acquisition module is used for collecting the initial image and the current image of the hydraulic metal through the cameras and transmitting them to the data analysis module.

[0035] The data analysis module is used for identifying the rust area and the non-rust area according to the reflected light of the fluorescent agent in the current image, generating a rust area report and transmitting the rust area report to the intelligent control module, and the rust analysis model is also used for comparing the initial image and the current image of the water conservancy metal, judging the rust expansion rate according to the rust area range change and the shooting interval time in the initial image and the current image, generating a rust change report and transmitting the rust change report to the intelligent control module.

[0036] The intelligent control module is used for extracting the rust area of the water conservancy metal according to the rust area report, and extracting the rust and smearing the mixed agent on the rust area, and adjusting the smearing amount of the mixed agent according to the rust change report.

[0037] Beneficial effects: unlike the existing analysis technology, the system identifies the rust area and the non-rust area according to the reflected light of the fluorescent agent in the current image of the water conservancy metal, in this way, the rust analysis model only needs to identify the area with fluorescent reflection and the area without fluorescent reflection, that is, the non-rust area and the rust area can be distinguished, and the area of the rust area can be judged according to the size of the area without fluorescent reflection, and the current image is compared with the initial image to judge the image change of the area, so as to analyze the severity of the rust and the rust rate; this way can effectively improve the detection efficiency and the detection accuracy.

[0038] Additional aspects and advantages of the application will be described in part below, some will become apparent from the following description, or will be learned by practice of the application. BRIEF DESCRIPTION OF DRAWINGS

[0039] Figure 1 It is a step schematic diagram of the deep learning water conservancy metal structure rust detection method of the application.

[0040] Figure 2 It is an isometric view of the comprehensive treatment device in the deep learning water conservancy metal structure rust detection method of the application.

[0041] Figure 3 It is an isometric view of the smearing assembly in the comprehensive treatment device of the application.

[0042] Figure 4 It is a sectional view of the mounting seat in the comprehensive treatment device of the application.

[0043] Figure 5 It is a sectional view of the spraying assembly in the comprehensive treatment device of the application.

[0044] The reference numerals in the accompanying drawings of the instruction manual include: 1. Wall-climbing robot; 2. Mounting base; 3. Center rod; 4. Center plate; 5. Rotating rod; 6. Connecting rod; 7. Limiting rod; 8. Spreading disc; 9. Telescopic rod; 10. Scraper; 11. Collection box; 12. Mixing chamber; 13. Stirring rod; 14. Gear motor; 15. Drive chamber; 16. First gear; 17. Second gear; 18. Storage box; 19. Piston plate; 20. Hinge rod; 21. Control valve; 22. Conveying gear; 23. Transmission rod; 24. Camera; 25. Protective frame; 26. Spreading disc. Detailed Implementation

[0045] The following detailed description illustrates the specific implementation method:

[0046] Example 1:

[0047] As attached Figures 1-5 As shown: A deep learning-based method for detecting corrosion in hydraulic metal structures, comprising the following steps:

[0048] S1, Pre-treatment: Prepare the anti-rust coating, fluorescent agent, and integrated treatment device. Use the integrated treatment device to stir and mix the anti-rust coating and fluorescent agent to obtain a mixture. Use the integrated treatment device to move on the surface of the hydraulic metal while applying the mixture. Use the integrated treatment device to collect an image of the surface of the hydraulic metal after applying the mixture to obtain an initial image.

[0049] S2, Fluorescence Reflection Detection and Corrosion Analysis: During the use of hydraulic metal components, an integrated processing device is moved across the surface of the hydraulic metal component. Simultaneously, the integrated processing device acquires an image of the current hydraulic metal surface, obtaining the current image. This current image is then compared with the initial image. The integrated processing device identifies corroded and non-corroded areas on the hydraulic metal surface based on the reflected light from the fluorescent agent. If the reflected light from the fluorescent agent is present on the hydraulic metal surface, it indicates that the agent in that area has not detached, and the integrated processing device will determine that no corrosion has occurred in that area. If no reflected light from the fluorescent agent is detected on the hydraulic metal surface, it indicates that the agent in that area has detached, and the integrated processing device will determine that corrosion has occurred in that area.

[0050] S3, Rust Treatment: After the integrated treatment device identifies the rusted area, it scrapes off the rust from the surface of the area and stores it; at the same time, it sprays a new mixture onto the area; the integrated treatment device judges the severity of the rust based on the current image and adjusts the amount of mixture sprayed according to the severity of the rust; the amount of mixture sprayed is directly proportional to the severity of the rust.

[0051] In S1, the comprehensive treatment device comprises a rust treatment system for detecting and treating rust and a wall-climbing robot 1 (for example, a magnetic adsorption type mobile robot or a vacuum adsorption type mobile robot in the prior art, and a magnetic adsorption type mobile robot is selected in the embodiment), and the wall-climbing robot 1 is provided with a mixing assembly for preparing a mixed agent, a conveying assembly for conveying the mixed agent, a spraying assembly for spraying the mixed agent, and an applying assembly for applying the mixed agent.

[0052] A mounting groove is formed in the side wall of the wall-climbing robot 1, and a mounting seat 2 is welded to the inner side wall of the mounting groove. The mounting seat 2 is provided with a driving assembly for driving the mixing assembly, the conveying assembly, and the applying assembly to operate together.

[0053] The applying assembly comprises a central rod 3 rotationally fitted to the outer side wall of the mounting seat 2, and a central plate 4 is fixedly sleeved on the central rod 3. The central plate 4 is symmetrically provided with limiting grooves. The side wall of the mounting seat 2 is symmetrically rotationally fitted with rotating rods 5, and the rotating rods 5 are each welded with a connecting rod 6. The connecting rod 6 is welded with a limiting rod 7 at the end away from the rotating rod 5, and the limiting rod 7 is rotationally fitted with the limiting groove adjacent thereto. The rotating rod 5 is welded with an applying disc 8 at the end away from the mounting seat 2, and the applying disc 8 is located outside the mounting groove. The central rod 3 is provided with a scraping assembly for scraping rust. The limiting rod 7 is provided with a collecting assembly for collecting the image of the waterwork metal surface.

[0054] The scraping assembly comprises an extension rod 9 bolted and connected to the end of the central rod 3 away from the mounting seat 2, and the output shaft of the extension rod 9 is coaxially bolted with a scraper 10. The rust treatment system is used to control the operation of the extension rod 9, so as to adjust the position of the scraper 10. The outer side wall of the wall-climbing robot 1 is welded with a collecting frame 11. The collecting frame 11 is located below the scraper 10, and the collecting frame 11 is fixedly bonded with a magnet sheet in the collecting frame 11.

[0055] The mixing assembly comprises a plurality of mixing cavities 12 formed in the mounting seat 2. The mixing cavities 12 are each communicated with an input pipe (not shown in the figure) for inputting a rust-proof coating and a fluorescent agent. The end of the rotating rod 5 away from the applying disc 8 extends into the mixing cavity 12 adjacent thereto, and the rotating rod 5 is rotationally fitted with the inner side wall of the mixing cavity 12 adjacent thereto. The portion of the rotating rod 5 located in the mixing cavity 12 is welded with a plurality of stirring rods 13. Figure 4 As shown in the figure, the inner side wall of the left mixing cavity 12 is bolted with a driving member, and the output shaft of the driving member is coaxially bolted with the left rotating rod 5. The rust treatment system is used to control the operation of the driving member, so as to drive the rotating rod 5 to rotate.

[0056] As shown in the figure, the inner side wall of the left mixing cavity 12 is bolted with a driving member, and the output shaft of the driving member is coaxially bolted with the left rotating rod 5. The rust treatment system is used to control the operation of the driving member, so as to drive the rotating rod 5 to rotate. Figure 4As shown, the driving assembly comprises a driving cavity 15 formed in the mounting seat 2; the driving cavity 15 is located on the side of the mixing cavity 12 away from the driving member; the inner side wall of the driving cavity 15 is symmetrically and rotationally fitted with a first gear 16 and a second gear 17, the first gear 16 is engaged with the second gear 17 adjacent thereto; the adjacent second gears 17 are engaged with each other; the rotating rods 5 all penetrate through the driving cavity 15 and are rotationally fitted with the side wall thereof; the rotating rods 5 are all coaxially and fixedly connected with the first gears 16 adjacent thereto by bolts.

[0057] The spraying assembly comprises a storage box 18 welded to the outer side wall of the wall-climbing robot 1, the inner side wall of the storage box 18 is vertically and slidingly fitted with a piston plate 19, the top of the piston plate 19 is hingedly connected with a hinge rod 20, the top of the hinge rod 20 is hingedly connected with the center plate 4; the top of the storage box 18 is provided with a moving groove for the movement of the hinge rod 20; the side wall of the storage box 18 is provided with a plurality of spraying openings; the spraying openings are all communicated with control valves 21; the rust treatment system is used for controlling the operation of the control valves 21, thereby controlling the spraying of the mixed agent.

[0058] The conveying assembly comprises a conveying cavity formed in the mounting seat 2, the conveying cavity is located between the mixing cavities 12; the mixing cavities 12 are all communicated with the conveying cavity; the inner side wall of the conveying cavity is symmetrically and rotationally fitted with conveying gears 22, the adjacent conveying gears 22 are engaged with each other; the conveying cavity is communicated with the storage box 18, the conveying cavity is located above the storage box 18; the second gears 17 all are fixedly connected with transmission rods 23 on the side close to the conveying cavity by bolts, the transmission rods 23 all extend into the conveying cavity on the side away from the second gears 17 and are all fixedly connected with the conveying gears 22 adjacent thereto by bolts.

[0059] The collecting assembly comprises a plurality of cameras 24, the limiting rods 7 all extend out of the mounting groove on the side away from the connecting rods 6 and are all fixedly connected with the cameras 24 adjacent thereto by bolts, the rust treatment system is used for controlling the operation of the cameras 24.

[0060] The rust treatment system comprises a data acquisition module, a data analysis module and an intelligent control module; the modules are signal-connected with each other; the modules are all signal-connected with the remote control end (such as a mobile phone, a computer or a tablet, etc., a mobile phone is selected as the remote control end in the embodiment) of the operator.

[0061] The data acquisition module is used for collecting the initial image and the current image of the hydraulic metal by the cameras 24 and transmitting the images to the data analysis module.

[0062] The data analysis module constructs a corrosion analysis model based on deep learning algorithms. The corrosion analysis model is used to identify rusted and non-rusted areas based on the reflected light of the fluorescent agent in the current image, generate a corrosion area report, and transmit it to the intelligent control module. The corrosion analysis model is also used to compare the initial image and the current image of the hydraulic metal, determine the corrosion expansion rate based on the changes in the range of rusted areas and the shooting interval in the initial image and the current image, generate a corrosion change report, and transmit it to the intelligent control module.

[0063] The intelligent control module is used to extract the rusted areas of hydraulic metals based on the rusted area report, and to extract the rust and apply a mixture to the rusted areas; it also adjusts the amount of mixture applied based on the rust change report.

[0064] In this embodiment, the driving component is a geared motor 14.

[0065] The specific implementation process is as follows:

[0066] by Figure 1 For example, before the hydraulic metal is put into use, the operator can deliver the anti-rust coating and fluorescent agent into the mixing chamber 12 through the input pipe; then control the wall-climbing robot 1 to move vertically along the surface of the hydraulic metal through the mobile phone, and start the reduction motor 14 at the same time.

[0067] by Figure 3 For example, during this process, the reduction motor 14 will drive the right rotating rod 5 and the right first gear 16 to rotate counterclockwise around their own axis, which in turn will drive the right second gear 17 to rotate clockwise around its own axis. The right second gear 17 will then drive the left second gear 17 to rotate counterclockwise around its own axis, and drive the left first gear 16 and the left rotating rod 5 to rotate clockwise around their own axis. At this time, the rotating rod 5 will drive the stirring rod 13 to rotate, thereby mixing the anti-rust coating and fluorescent agent in the mixing chamber 12 to prepare a mixture.

[0068] by Figure 3 and Figure 4 For example, when the second gear 17 rotates clockwise, it drives the transmission rod 23 to rotate, which in turn drives the conveying gear 22 to rotate together. Utilizing the principle of a gear pump, as the tooth grooves above the conveying gear 22 change from an engaged state to a disengaged state, a partial vacuum is formed, drawing the mixture from the mixing chamber 12. As the tooth grooves below the conveying gear 22 change from a disengaged state to an engaged state, the volume of the area below the conveying gear 22 decreases, and the mixture in the conveying chamber is squeezed into the storage tank 18 by the conveying gear 22. Since the conveying chamber is located above the storage tank 18, the mixture is affected by gravity, making the conveying process smoother. During the rotation of the conveying gear 22, it also further improves the mixing effect of the anti-rust coating and the fluorescent agent.

[0069] by Figure 3 For example, when the rotating rod 5 rotates, it will also drive the connecting rod 6 and the limiting rod 7 to rotate. The limiting rod 7 will drive the camera 24 to rotate, thereby collecting images of various areas of the hydraulic metal surface. The data acquisition module will obtain preliminary images of the hydraulic metal surface through the camera 24 and transmit them to the data analysis module for subsequent judgment on whether the hydraulic metal surface is corroded.

[0070] Meanwhile, when the right limit rod 7 enters the right limit groove, it causes the center plate 4 and the center rod 3 to rotate 90° clockwise, at which point the left limit groove rotates upward; when the left limit rod 7 enters the left limit groove, it causes the center plate 4 and the center rod 3 to rotate 90° counterclockwise, at which point the right limit groove rotates upward; during the rotation of the center plate 4, the center plate 4 will drive the hinge rod 20 to move up and down, and the hinge rod 20 will drive the piston plate 19 to move up and down in the storage tank 18. At this time, the operator can open the control valve 21 through the mobile phone, and the mixture in the storage tank 18 will be squeezed by the piston plate 19 through the control valve 21 and sprayed onto the hydraulic metal surface.

[0071] At the same time, the rotation of the rotating rod 5 will drive the coating disk 8 to rotate. In conjunction with the movement of the wall-climbing robot 1, the mixture will be evenly coated onto the surface of the hydraulic metal by the coating disk 8, thereby protecting the hydraulic metal and facilitating subsequent inspection.

[0072] When inspecting the hydraulic metal after use, the operator can control the wall-climbing robot 1 to move vertically along the surface of the hydraulic metal via a mobile phone, while simultaneously starting the reduction motor 14.

[0073] During the movement of the wall-climbing robot 1, the connecting rod 6, the limiting rod 7, and the camera 24 will rotate under the drive of the rotating rod 5. The data acquisition module will acquire the current image of the hydraulic metal surface through the camera 24 and transmit it to the data analysis module. The data analysis module will compare the current image with the initial image and determine whether there is corrosion at the current position.

[0074] During the rotation of the central rod 3, the central rod 3 will also drive the telescopic rod 9 and the scraper 10 to rotate. In the initial state, the telescopic rod 9 is in a retracted state. If the data analysis module determines that there is rust in the current area, the intelligent control module will control the telescopic rod 9 to extend, so that the scraper 10 is in contact with the surface of the hydraulic metal, thereby scraping the rust on its surface into the collection frame 11, and using the magnetic sheet to adsorb the rust, which is convenient for subsequent operators to conduct further inspection of the rust. At the same time, the intelligent control module will open the control valve 21 to spray the mixture into the area, and use the application plate 8 to spread the mixture evenly.

[0075] The corrosion analysis process is as follows: the data analysis module uses the corrosion analysis model to detect corrosion on the surface of the hydraulic metal; the corrosion analysis model identifies the corrosion area and the non-corrosion area of the surface of the hydraulic metal based on the reflected light of the fluorescent agent; if the surface of the hydraulic metal has fluorescent reflection, it indicates that the mixed agent in the area has not fallen off, and the corrosion analysis model determines that the area has not been corroded; if the surface of the hydraulic metal does not identify fluorescent reflection, it indicates that the mixed agent in the area has fallen off, and the corrosion analysis model determines that the area has been corroded; and the corrosion analysis model compares the initial image and the current image, and determines the change rate of the corrosion area according to the area difference of the corrosion area and the time interval between the two image acquisitions.

[0076] The calculation formula of the change rate is as follows:

[0077] V=S / T (1).

[0078] Wherein, S is the area difference of the corrosion area in the initial image and the current image; T is the acquisition time interval of the initial image and the current image, and V is the change rate; the greater the change rate V, the more serious the corrosion.

[0079] The embodiment can quickly and accurately determine the corrosion area on the surface of the hydraulic metal according to the fluorescent reflection while protecting the hydraulic metal, greatly reducing the detection difficulty and improving the detection efficiency and accuracy through the design of the mixed agent and the improvement of the detection idea.

[0080] Embodiment 2:

[0081] As shown in the accompanying drawings, Figure 2 different from the above-mentioned embodiments, a protective frame 25 for protecting the rotating rod 5 and the center rod 3 is welded outside the mounting groove; the protective frame 25 is provided with a telescopic groove for the telescopic rod 9 to extend and retract, and a plurality of smearing grooves for the smearing disc 8 to rotate; and a smearing plate 26 for further smearing the mixed agent is welded on the outer side wall of the protective frame 25.

[0082] The specific implementation process is as follows: during the movement of the wall-climbing robot 1, the protective frame 25 can protect each component in the mounting groove and improve the stability of the device; when the wall-climbing robot 1 moves, the smearing plate 26 can further smear the mixed agent smeared on the smearing disc 8, so that the mixed agent on the surface of the hydraulic metal is more uniform.

[0083] Obviously, the above-mentioned embodiments are only examples for clear illustration, and are not limitations on the embodiments. For ordinary skilled persons in the art, other different forms of changes or variations can be made on the basis of the above description. Here, it is not necessary and impossible to exhaust all the embodiments. The obvious changes or variations derived therefrom are still within the protection scope of the present application.

Claims

1. A deep learning-based method for detecting corrosion in hydraulic metal structures, characterized in that, Includes the following steps: S1, Pretreatment: Prepare anti-rust coating, fluorescent agent and comprehensive treatment device. Use the comprehensive treatment device to stir and mix the anti-rust coating and fluorescent agent to obtain a mixture. Use the comprehensive treatment device to move on the surface of the hydraulic metal while applying the mixture. An initial image was obtained by using a comprehensive processing device to acquire images of the hydraulic metal surface after the mixture was applied. S2, Fluorescence Reflection Detection and Corrosion Analysis: During the use of hydraulic metal components, a comprehensive processing device is moved across the surface of the hydraulic metal component. Simultaneously, the device acquires an image of the current surface of the hydraulic metal component, obtaining the current image. The device then compares the current image with the initial image. Based on the reflected light from the fluorescent agent, the device identifies corroded and non-corroded areas on the hydraulic metal surface. If the surface of the hydraulic metal component shows reflected light from the fluorescent agent, it indicates that the agent in that area has not detached, and the device determines that no corrosion has occurred in that area. If no reflected light from the fluorescent agent is detected on the surface of the hydraulic metal component, it indicates that the agent in that area has detached, and the device determines that corrosion has occurred in that area. S3, Rust Treatment: After the integrated treatment device identifies the rusted area, it scrapes off the rust from the surface of the area and stores it; at the same time, it sprays a new mixture onto the area; the integrated treatment device judges the severity of the rust based on the current image and adjusts the amount of mixture sprayed according to the severity of the rust; the amount of mixture sprayed is directly proportional to the severity of the rust.

2. The deep learning-based corrosion detection method for hydraulic metal structures according to claim 1, characterized in that, In S1, the integrated processing device includes a rust treatment system for detecting and treating rust and a wall-climbing robot (1). The wall-climbing robot (1) is equipped with a mixing component for preparing a mixture, a conveying component for conveying the mixture, a spraying component for spraying the mixture, and a coating component for applying the mixture. The wall-climbing robot (1) has a mounting slot on its side wall, and a mounting base (2) is fixedly connected to the inner side wall of the mounting slot. The mounting base (2) is equipped with a drive component for driving the mixing component, the conveying component and the coating component to operate together. The coating assembly includes a central rod (3) rotatably fitted to the outer wall of the mounting base (2), a central plate (4) fixedly fitted on the central rod (3), and symmetrically opened limit grooves on the central plate (4); a rotating rod (5) rotatably fitted to the side wall of the mounting base (2), a connecting rod (6) fixedly connected to each rotating rod (5), and a limit rod (7) fixedly connected to the end of each connecting rod (6) away from the rotating rod (5), and the limit rod (7) rotatably fitted with the limit groove adjacent to it; a coating disc (8) fixedly connected to the end of each rotating rod (5) away from the mounting base (2), and the coating disc (8) is located outside the mounting groove; a scraping assembly for scraping rust is provided on the central rod (3); and an acquisition assembly for acquiring images of the surface of hydraulic metal is provided on the limit rod (7).

3. The deep learning-based corrosion detection method for hydraulic metal structures according to claim 2, characterized in that, The scraping assembly includes a telescopic rod (9) fixedly connected to the end of the central rod (3) away from the mounting base (2), and a scraper (10) is fixedly connected to the output shaft of the telescopic rod (9) coaxially; the rust treatment system is used to control the operation of the telescopic rod (9) and thereby adjust the position of the scraper (10); a collection frame (11) is fixedly connected to the outer wall of the wall-climbing robot (1); the collection frame (11) is located below the scraper (10), and a magnet is fixedly connected inside the collection frame (11).

4. The deep learning-based corrosion detection method for hydraulic metal structures according to claim 3, characterized in that, The mixing assembly includes several mixing chambers (12) opened in the mounting base (2); each mixing chamber (12) is connected to an input pipe for inputting the anti-rust coating and fluorescent agent; the end of the rotating rod (5) away from the coating plate (8) extends into the mixing chamber (12) adjacent to it, and the rotating rod (5) rotates and engages with the inner wall of the mixing chamber (12) adjacent to it; the part of the rotating rod (5) located in the mixing chamber (12) is fixedly connected to several stirring rods (13); a driving component is fixedly connected to the inner wall of one of the mixing chambers (12), and the output shaft of the driving component is coaxially fixedly connected to the rotating rod (5) adjacent to it; the rust treatment system is used to control the operation of the driving component, thereby driving the rotating rod (5) to rotate.

5. The deep learning-based corrosion detection method for hydraulic metal structures according to claim 4, characterized in that, The drive assembly includes a drive cavity (15) inside the mounting base (2); the drive cavity (15) is located on the side of the mixing cavity (12) away from the drive component; the inner sidewall of the drive cavity (15) is symmetrically fitted with a first gear (16) and a second gear (17), the first gear (16) meshes with the second gear (17) adjacent to it; the adjacent second gears (17) mesh with each other; the rotating rods (5) all pass through the drive cavity (15) and are rotatably fitted with its sidewall; the rotating rods (5) are all coaxially fixedly connected to the first gear (16) adjacent to them.

6. The deep learning-based corrosion detection method for hydraulic metal structures according to claim 5, characterized in that, The spraying assembly includes a storage box (18) fixedly connected to the outer wall of the wall-climbing robot (1). A piston plate (19) is vertically slidably fitted on the inner wall of the storage box (18). A hinge rod (20) is hinged to the top of the piston plate (19). The top of the hinge rod (20) is hinged to the center plate (4). A moving groove for the hinge rod (20) to move is opened on the top of the storage box (18). Several spraying ports are opened on the side wall of the storage box (18). A control valve (21) is connected to each spraying port. The rust treatment system is used to control the operation of the control valve (21) and thus control the spraying of the mixture.

7. The deep learning-based corrosion detection method for hydraulic metal structures according to claim 6, characterized in that, The conveying assembly includes a conveying cavity opened in the mounting base (2), which is located between the mixing cavities (12); the mixing cavities (12) are all connected to the conveying cavity; the inner sidewall of the conveying cavity is symmetrically fitted with conveying teeth (22), and adjacent conveying teeth (22) mesh with each other; the conveying cavity is connected to the storage box (18); the second gear (17) is fixedly connected to a transmission rod (23) on the side near the conveying cavity, and the end of the transmission rod (23) away from the second gear (17) extends into the conveying cavity and is fixedly connected to the adjacent conveying tooth (22) on the same axis.

8. The deep learning-based corrosion detection method for hydraulic metal structures according to claim 7, characterized in that, The acquisition component includes several cameras (24), and the end of the limiting rod (7) away from the connecting rod (6) extends out of the mounting groove and is fixedly connected to the adjacent camera (24). The corrosion treatment system is used to control the operation of the camera (24).

9. The deep learning-based corrosion detection method for hydraulic metal structures according to claim 8, characterized in that, A protective frame (25) for protecting the rotating rod (5) and the center rod (3) is fixedly connected to the outside of the mounting slot; the protective frame (25) has a telescopic groove for the telescopic rod (9) to extend and retract and several coating grooves for the coating disc (8) to rotate; a coating plate (26) for further coating of the mixture is fixedly connected to the outer wall of the protective frame (25).

10. The deep learning-based corrosion detection method for hydraulic metal structures according to claim 9, characterized in that, The corrosion treatment system includes a data acquisition module, a data analysis module, and an intelligent control module; The data acquisition module is used to acquire initial and current images of the hydraulic metal via a camera (24) and transmit them to the data analysis module; The data analysis module constructs a corrosion analysis model based on deep learning algorithms. The corrosion analysis model is used to identify corrosion and non-corrosion areas based on the reflected light of the fluorescent agent in the current image, generate a corrosion area report and transmit it to the intelligent control module. The corrosion analysis model is also used to compare the initial image and the current image of the hydraulic metal, determine the corrosion expansion rate based on the changes in the range of corrosion areas in the initial image and the shooting interval, generate a corrosion change report and transmit it to the intelligent control module. The intelligent control module is used to extract the rusted areas of hydraulic metals based on the rusted area report, and to extract the rust and apply a mixture to the rusted areas; it also adjusts the amount of mixture applied based on the rust change report.

Images