Intelligent AI linkage-based spraying concrete wall surface maintenance system and method

Abstract

The application discloses a kind of based on intelligent AI linkage spraying concrete wall maintenance system and method, the scheme is by obtaining the environment data around the concrete wall to be maintained and the monitoring image of the concrete wall to be maintained, and accordingly identify the dryness and humidity of corresponding area on the concrete wall to be maintained, concrete color parameter;Again, according to the environment data obtained, monitoring image and the corresponding wall parameter data identified judge that the wall area needs to be sprayed and maintained, and control the spraying unit of corresponding position to start spraying.This maintenance scheme can automatically identify the area to be maintained on the wall based on the dryness and humidity of the wall of building, the characteristics such as concrete color, and linkage spraying maintenance, effectively solve the problem that concrete wall cannot be maintained in time, realize green construction.

Description

TECHNICAL FIELD

[0001] The present application relates to the construction engineering construction technology, specifically to the wall maintenance technology of concrete structure. BACKGROUND

[0002] With the continuous development of high-rise, super high-rise, higher requirements for the quality of the building and green construction, the wall maintenance of concrete structure becomes a hot spot in the field.

[0003] The conventional maintenance means of the wall is artificial watering and smearing maintenance liquid. This way has many problems in the actual operation process, such as water cannot stay on the wall for a long time, and if the outdoor temperature is low after pouring in the outdoor environment, the temperature of the wall column concrete will be low, which is not conducive to the maintenance of the concrete; at the same time, a large number of uneven artificial spraying not only affects the maintenance quality of the concrete, but also increases the labor cost and wastes water resources.

[0004] Therefore, how to efficiently spray and maintain the concrete wall surface is a problem to be solved in the field. SUMMARY

[0005] In view of the problems existing in the artificial watering based maintenance scheme of the existing concrete structure wall, the purpose of the present application is to provide an intelligent AI linkage spraying concrete wall surface maintenance system, and based on the system, a concrete wall surface maintenance method is also provided. The maintenance scheme can automatically identify the to-be-maintained area on the wall based on the humidity, concrete color and other characteristics of the building wall, and linkage spraying maintenance, effectively solve the problem that the concrete wall cannot be maintained in time, and realize green construction.

[0006] In order to achieve the above purpose, the intelligent AI linkage spraying concrete wall surface maintenance system provided by the present application comprises a frame body, a plurality of intelligent identification units, a data processing unit, a water supply pipeline, a plurality of spraying units and a background management unit.

[0007] The frame body is erected on the concrete wall surface to be maintained.

[0008] The water supply pipeline is erected on the frame body, and the plurality of spraying units are distributed on the frame body and connected to the water supply pipeline; the sprayable area formed between the plurality of spraying units can cover the entire concrete wall surface to be maintained.

[0009] The plurality of intelligent identification units are distributed on the frame body and connected to the data processing unit, and the identifiable area formed between the plurality of intelligent identification units can cover the entire concrete wall surface to be maintained; each intelligent identification unit can obtain the monitoring image of the concrete wall surface to be maintained in the identifiable area, and perform deep learning classification on the obtained monitoring image.

[0010] The data processing unit is respectively connected with a plurality of spraying units, a plurality of intelligent recognition units and a background management unit, and data interaction between the plurality of spraying units, the plurality of intelligent recognition units and the background management unit is completed;

[0011] The background management unit analyzes and processes the monitoring image classified by deep learning transmitted by each intelligent recognition unit, identifies and judges the wall region needing to be sprayed and maintained, and controls the spraying unit at the corresponding position to start spraying.

[0012] Further, the water supply pipeline comprises a main water supply pipe, an electromagnetic valve, a branch water supply pipe and a fire standpipe, the main water supply pipe is connected with the fire standpipe through the electromagnetic valve and the pressure reducing valve, the main water supply pipe is connected with the branch water supply pipe, and the electromagnetic valve is controlled by the background management unit.

[0013] Further, the plurality of spraying units are distributed in an array and connected through the branch water supply pipeline.

[0014] Further, the plurality of intelligent recognition units are arranged at the top of the frame body.

[0015] Further, the intelligent recognition unit adopts a deep learning model to extract low-level features of an image, further learns high-level features of the image based on the mentioned low-level features of the image, aggregates all the features, completes image classification, and distinguishes different concrete image states.

[0016] Further, the background management unit statistically analyzes the monitoring image classified by deep learning of the intelligent recognition unit, and determines the spraying and maintenance state of the corresponding concrete wall according to the statistical analysis result.

[0017] In order to achieve the above purpose, the intelligent AI linkage spraying concrete wall surface maintenance method provided by the present application comprises:

[0018] Obtaining a monitoring image of a concrete wall surface to be maintained, classifying the obtained monitoring image by deep learning, and identifying the spraying and maintenance state of the corresponding region of the concrete wall to be maintained;

[0019] According to the obtained monitoring image classified by deep learning, the wall region needing to be sprayed and maintained is identified and judged, and the spraying unit at the corresponding position is controlled to start spraying.

[0020] Further, the deep learning model is used to extract low-level features of an image in the maintenance method, further learn high-level features of the image based on the mentioned low-level features of the image, aggregate all the features, complete image classification, and distinguish different concrete image states.

[0021] Further, the maintenance method further comprises the step of adjusting the water pressure of the spraying.

[0022] Further, the maintenance method further comprises the step of adjusting the water pressure of the spraying.

[0023] The concrete wall maintenance scheme provided by the application can automatically identify the state of the concrete wall surface, automatically maintain the poured concrete, ensure the quality of the poured concrete, and realize green construction of the construction site by timely opening and closing the spraying device through the intelligent identification system and saving water. BRIEF DESCRIPTION OF DRAWINGS

[0024] The application will be further described below in combination with the drawings and specific embodiments.

[0025] Figure 1 is a sectional view of the intelligent AI linkage spraying concrete wall maintenance system in the embodiment of the application;

[0026] Figure 2 is a sectional view of the intelligent AI linkage spraying concrete wall maintenance system in the embodiment of the application;

[0027] Figure 3 is a sectional view of the intelligent AI linkage spraying concrete wall maintenance system in the embodiment of the application;

[0028] Figure 4 is a sectional view of the intelligent AI linkage spraying concrete wall maintenance system in the embodiment of the application;

[0029] Figure 5 is a sectional view of the intelligent AI linkage spraying concrete wall maintenance system in the embodiment of the application; DETAILED DESCRIPTION

[0030] In order to make the technical means, creative features, purposes and effects realized by the application easy to understand, the application will be further described below in combination with specific drawings.

[0031] In order to make the technical means, creative features, purposes and effects realized by the application easy to understand, the application will be further described below in combination with specific drawings.

[0032] The application innovatively uses image recognition technology, obtains the environment data around the concrete wall to be maintained and the monitoring image of the concrete wall surface to be maintained, and identifies the dryness and wetness and the concrete color parameters of the corresponding area on the concrete wall to be maintained.

[0033] On this basis, the wall body area needing to be sprayed for maintenance is determined according to the obtained environmental data, monitoring images and the identified corresponding wall body parameter data, and the spraying unit at the corresponding position is controlled to start spraying.

[0034] Specifically, in the present application, autonomous deep learning of images is completed through a deep learning model, that is, feature learning or representation learning of images is performed, so that the ability to distinguish images corresponding to different concrete states is achieved, and accordingly, classification of whether the concrete is in a maintenance state is completed.

[0035] Further, in the deep learning model in the present application, a 'low-level' feature of an image, such as the color, shape and contour of concrete, is extracted through a 'low-level' neural network (i.e., a shallow neural network), and the extracted 'low-level' feature information of the image is input to a 'high-level' neural network (i.e., a deep neural network);

[0036] Further, a 'high-level' feature of an image, such as water stains, concrete pores and stains, is learned by the 'high-level' neural network (i.e., the deep neural network);

[0037] Finally, all 'high-level' features of images obtained through deep learning are aggregated to a final fully connected layer by the 'high-level' neural network (i.e., the deep neural network), thereby completing the image classification task.

[0038] In the present application, statistical analysis is performed on the obtained environmental data, monitoring images and the identified corresponding wall body parameter data, and a decision voting method is used to identify the maintenance state, so that the wall body area needing to be sprayed for maintenance is determined, and the spraying unit at the corresponding position is controlled to start spraying.

[0039] In the present application, the collected monitoring images classified through deep learning are counted, and the accuracy of the actual recognition result is improved based on a large number of images.

[0040] Specifically, the collected monitoring images classified through deep learning are counted and analyzed, and a decision voting is performed on the maintenance state thereof according to the principle of'minority submits to majority', and whether the current concrete needs to be maintained is determined according to the result.

[0041] On this basis, the present application can further report the processed data to a cloud platform, and control the localization display of parameters, so as to realize the organic integration of the environment and related parameters with the corresponding monitoring pictures.

[0042] As an example, the parameters herein are generally the working records of spraying for maintenance, such as the spraying start time and the spraying duration.

[0043] Finally, the scheme can control the corresponding spraying unit of the position needing spraying maintenance and adjust the spraying water pressure size by adjusting the water pressure when it is determined that the poured concrete wall of the building needs maintenance, so as to realize the accurate grid spraying maintenance control, and not only realize the spraying maintenance only for the position needing spraying maintenance, but also match the corresponding spraying maintenance mode (such as spraying water pressure size, spraying time, etc.) according to the to-be-maintained state of the position needing spraying maintenance.

[0044] Referring to Figure 1 With Figure 2 It shows an example scheme of the concrete wall surface maintenance system based on intelligent AI linkage provided by the present application based on the foregoing scheme.

[0045] As can be seen from the figure, the concrete wall surface maintenance system based on intelligent AI linkage mainly comprises water pump 1, metal spraying device 2, sub-water supply pipe 3, main control board switch 4, intelligent camera 5, data processing unit 6, mobile device 7, main water supply pipe 8, frame body 9 and background management unit.

[0046] The frame body 9 in the system is attached to the concrete wall 10 to be maintained, as shown in the figure, the frame body 9 in the example is specifically attached to the two side surfaces of the concrete wall 10 to be maintained, i.e. attached to the inner side surface and the outer side surface of the concrete wall 10, for carrying other components in the system.

[0047] The water pump 1, sub-water supply pipe 3 and main water supply pipe 8 in the system cooperate with each other to form a corresponding water supply pipeline, and are integrally erected on the frame body 9, for supplying water to the corresponding metal spraying device 2.

[0048] The main water supply pipe 8 is integrally attached to the frame body 9 and climbs with the frame body 9.

[0049] Specifically, the main water supply pipe 8 in the system is set according to the number of fire risers of the frame body, and is set in priority to the fire riser.

[0050] The main water supply pipe 8 in the system is connected to the fire riser through a corresponding pressure reducing valve and an electromagnetic valve. As an example, the fire riser is connected to the pressure reducing valve for pressure reduction, the pressure reducing valve is connected to the electromagnetic valve, the on-off state is controlled by the electromagnetic valve, and the main water supply pipe is connected by the electromagnetic valve through a variable pipe (i.e. a corresponding water pipe variable pipe).

[0051] As an example, the main water supply pipe 8 is preferably composed of a corresponding PVC pipe.

[0052] The water supply sub-pipe 3 is attached to the frame body 9, for connecting the main water supply pipe 8 and the corresponding metal spraying device 2.

[0053] The water supply branch pipe 3 in the system is preferably communicated with the main water supply pipe 8 through the corresponding variable (i.e. the corresponding water pipe variable) when setting.

[0054] The main water supply pipe 8 is preferably composed of the corresponding PVC pipe as an example.

[0055] The water pump 1 here is used to pump the water obtained from the lower main water supply pipe 8 from the fire riser to the water supply branch pipe 3 of the upper concrete wall body outer frame, and then spray the wall body for maintenance through the metal spraying device 2 communicated with the water supply branch pipe 3.

[0056] Referring to Figure 3 , in this example, the water supply pipeline is formed by cooperating the water pump 1, the branch water supply pipe 3 and the main water supply pipe 8, and based on the fire riser, wherein the fire riser 11 is connected with the pressure reducing valve 12 to reduce the high pressure water in the fire riser 11, the pressure reducing valve 12 is connected with the electromagnetic valve 13, the on-off state of the whole water supply pipeline is controlled by the electromagnetic valve 13, the electromagnetic valve 13 is connected with the main water supply pipe 8 through the first variable (i.e. the corresponding water pipe variable) 14, and the main water supply pipe 8 is communicated to the corresponding branch water supply pipe 3 through the second variable (i.e. the corresponding water pipe variable) 15, thereby forming the whole water supply pipeline attached to the frame 9, so as to cover the corresponding concrete wall surface.

[0057] For Figure 1 the example shown, the corresponding frame 9 is arranged on the inner side and the outer side of the concrete wall 10, and two sets of water supply pipelines are arranged. Therefore, in the specific implementation, two fire risers are arranged corresponding to the inner side and the outer side of the concrete wall 10, and the water supply pipelines attached to the two frames 9 are formed based on the two fire risers, which are specifically as described above, and will not be described here.

[0058] The metal spraying device 2 in this example forms the corresponding spraying unit to spray and maintain the concrete. In this example, according to the actual maintenance requirements, multiple metal spraying devices 2 are adopted, which are distributed on the frame 9 and connected with the corresponding water supply branch pipe 3, thereby realizing the spraying and maintenance of the wall body.

[0059] As a preferred, the multiple metal spraying devices 2 in this example form a sprayable area between the multiple metal spraying devices 2 when being distributed and arranged, which can cover the whole concrete wall surface to be maintained.

[0060] Taking the example of the diagram scheme, in this example, the corresponding water supply pipelines are arranged on the inner side and the outer side of the concrete wall 10, and four rows of spraying devices are arranged along the height direction of the frame 9 distributed on the inner side and the outer side of the concrete wall 10 in this example, each row is divided into five columns, and the up-down and left-right spacing is 1m, thereby forming 20 arrayed spraying points for each side of the concrete wall 10 to be maintained.

[0061] The master control switch 4 in this example is a manual master control switch of the metal spraying device 2, which can be used to adjust the water pressure.

[0062] As an example, in this example, each set of water supply pipeline is controlled by the corresponding electromagnetic valve 13 to control the on-off of the entire water supply pipeline. Therefore, in the pipeline of the inlet of the electromagnetic valve in each set of water supply pipeline, a pressure regulating valve is connected in series to control the pressure of the medium air to the cylinder, so as to adjust the water pressure. The pressure regulating valve is controlled by the master control switch 4, thereby realizing manual control to adjust the spraying state of the corresponding metal spraying device 2.

[0063] The intelligent camera 5 in this example constitutes an intelligent recognition unit of the entire system front end, which is specifically distributed on the frame body 9 and is connected with the data processing unit 6. The intelligent camera 5 can obtain surrounding environment data and monitoring images of the concrete wall surface to be maintained in the identifiable area, and identify the dryness and humidity of the concrete wall body to be maintained and the color parameters of the concrete in the identifiable area.

[0064] The intelligent camera 5 in this example runs a corresponding deep learning model, which completes the learning of images through a deep learning algorithm, has the ability to distinguish images corresponding to different concrete states, and accordingly completes the classification of whether the concrete is in a maintenance state.

[0065] The deep learning model has a shallow neural network, a deep neural network, and a fully connected layer.

[0066] The deep learning model is aimed at the concrete wall surface images captured by the intelligent camera 5. First, the shallow neural network processes the captured concrete wall surface images to extract “low-level” features of the images, such as concrete color, shape, and contour, and inputs the extracted “low-level” features of the images to the deep neural network. The deep neural network further learns “high-level” features of the images based on the received “low-level” features of the images, such as water stains, concrete holes, and stains. Finally, the deep neural network aggregates all the “high-level” features obtained to the final fully connected layer to complete the image classification task.

[0067] After the intelligent camera 5 completes the image classification of the captured concrete wall surface images through the deep learning model, the classified images are transmitted to the background management unit, which discriminates and votes according to the classification task.

[0068] According to the actual maintenance needs, multiple intelligent cameras 5 are preferably used for each concrete wall surface to be maintained in the example. When the multiple intelligent cameras 5 are distributed on the frame body 9, the identifiable area formed between them can cover the entire concrete wall surface to be maintained.

[0069] As an example, when the corresponding intelligent camera 5 is deployed, the corresponding intelligent camera 5 is preferably arranged at the top and middle of the frame body; the intelligent camera 5 distributed at the top of the frame body is used to obtain the state image of the top to the middle region of the concrete wall surface to be maintained; the intelligent camera 5 distributed at the middle of the frame body is used to obtain the state image of the middle to the bottom region of the concrete wall surface to be maintained. At the same time, the intelligent camera 5 distributed at the top of the frame body and the intelligent camera 5 distributed at the top of the frame body are staggered in the height direction of the frame body, so as to realize the comprehensive coverage of the entire concrete wall surface to be maintained.

[0070] Thus, for each concrete wall surface to be maintained, the platform management unit can synchronously collect image information of different regions of the concrete wall surface to be maintained through the plurality of field-deployed intelligent cameras 5, and discriminates one by one to determine the maintenance state of the concrete. If maintenance is required, the spray maintenance system can be opened to spray and maintain by controlling the electromagnetic valve in the water supply pipeline of the spray maintenance system distributed along the concrete.

[0071] As an example, the intelligent camera 5 distributed and arranged in this way can further realize the timely capture and real-time identification of the environment and the wall state, so as to timely understand the state of the entire concrete wall surface to be maintained.

[0072] The data processing unit 6 in the system is the hub of the entire system, and is connected with the plurality of metal spray devices 2, the plurality of intelligent cameras 5 and the background management unit, respectively, to complete the data interaction between the plurality of metal spray devices 2, the plurality of intelligent cameras 5 and the background management unit.

[0073] As an example, the data processing unit 6 in the system is mainly composed of a data receiver and an LED screen. The data receiver is used to receive the data identified and sent by the intelligent camera 5, to forward to the background management unit, and to receive the control data sent by the background management unit, to realize local display through the LED screen and synchronous sending to the corresponding metal spray device 2, to realize the linkage control of the metal spray device 2.

[0074] As an example, when the data processing unit 6 displays the local control data, different colors can be used to represent different spray maintenance states. For example, if the LED screen in the data processing unit displays red, it means that the metal spray device 2 is linked to be opened and performs spray maintenance; if the LED screen in the data processing unit displays green, it means that the metal spray device 2 is linked to be closed and the spray maintenance is completed.

[0075] The mobile device 7 in the system, as a mobile client, can interact with the background management unit to view the maintenance situation.

[0076] As an example, the mobile device 7 in this instance consists of a smartphone and a smart APP running on the smartphone. The smart APP communicates with the back-end management unit via 4G and 5G networks to view the implementation and maintenance status.

[0077] The specific composition of the mobile device 7 is not limited here; it can be determined according to actual needs.

[0078] The back-end management unit in this system is the remote computing and control center of the entire system. It can determine the concrete wall area that needs to be sprayed for curing based on the environmental data, monitoring images and corresponding wall parameter data transmitted by each smart camera 5, and control the metal spraying device 2 at the corresponding location to start spraying.

[0079] The back-end management unit in this system performs statistical analysis on the monitoring images (i.e., classified by deep learning) uploaded by each smart camera 5 deployed on site, thereby identifying the corresponding concrete wall's condition for spray curing.

[0080] This backend management unit further improves the accuracy of actual identification by collecting a large number of surveillance images classified through deep learning and performing statistical analysis.

[0081] The following example illustrates the implementation process of intelligent judgment of the status of concrete walls awaiting spray curing in the background management unit of this system.

[0082] For each concrete wall to be cured, this backend management unit collects monitoring images uploaded by each smart camera 5 deployed on the surface of that concrete wall, which are classified through deep learning. It then performs statistical analysis on the large number of monitoring images collected through deep learning, prioritizing the "majority rule" statistical analysis principle to vote on the curing status of the corresponding concrete wall and determine whether the concrete needs curing.

[0083] Because each of the five smart cameras deployed on site directly performs deep learning classification processing on each image of the concrete wall area it captures, it determines whether the concrete wall area corresponding to the image is in a state that requires spray curing or does not require spray curing.

[0084] In this way, for each concrete wall surface to be maintained, the background management unit collects the monitoring images classified by deep learning uploaded by each intelligent camera 5 deployed corresponding to the concrete wall surface to be maintained at the same time. The concrete wall surface area corresponding to the plurality of monitoring images collected can cover the entire concrete wall surface. Further analysis of the classification information of each monitoring image, i.e., the corresponding concrete wall area is in a need for spraying maintenance state or a non-need for spraying maintenance state. If the number of images classified as needing spraying maintenance is greater than the number of images classified as not needing spraying maintenance in the plurality of monitoring images collected, it is determined that the concrete wall surface to be maintained at this time is in a need for spraying maintenance state. If the number of images classified as needing spraying maintenance is less than the number of images classified as not needing spraying maintenance in the plurality of monitoring images collected, it is determined that the concrete wall surface to be maintained at this time is in a non-need for spraying maintenance state.

[0085] On this basis, the background management unit further collects monitoring images classified by deep learning uploaded by each intelligent camera 5 deployed corresponding to the concrete wall surface to be maintained at adjacent different times, and further performs statistical analysis and processing accordingly to analyze and determine the spraying maintenance state of the concrete wall surface to be maintained at different times. The specific process is the same as above and will not be described here.

[0086] Finally, the background management unit performs statistical analysis on the spraying maintenance state of the concrete wall surface to be maintained at all times. Based on the statistical analysis principle of "minority obeying majority", if the number of times that the concrete wall surface to be maintained is determined to be in a need for spraying maintenance state is greater than the number of times that the concrete wall surface to be maintained is determined to be in a non-need for spraying maintenance state, it is finally determined that the concrete wall surface to be maintained is in a need for spraying maintenance state, and the spraying maintenance system corresponding to the concrete wall surface to be maintained is started to spray and maintain the concrete wall surface. If the number of times that the concrete wall surface to be maintained is determined to be in a need for spraying maintenance state is less than the number of times that the concrete wall surface to be maintained is determined to be in a non-need for spraying maintenance state, it is finally determined that the concrete wall surface to be maintained is in a non-need for spraying maintenance state, and the spraying maintenance system corresponding to the concrete wall surface to be maintained is not started, and the next determination period is entered.

[0087] Through the aforementioned two-layer statistical analysis of "minority obeying majority" in the background management unit of the system, the maintenance state of each concrete wall surface to be maintained can be accurately determined, and the efficiency and accuracy of the entire system for spraying maintenance can be improved.

[0088] The concrete wall maintenance system based on intelligent AI linkage spraying of the present example can effectively solve the problem of regular maintenance of concrete wall surface. The system can be installed through traditional frame, climbing formwork and other frames in actual application, and the poured concrete can be maintained in time to ensure the quality of concrete pouring. At the same time, the metal spraying device is turned on and off in time through the intelligent identification system, water is saved, and green construction of the construction site is realized.

[0089] Further, the system can realize manual control of spraying maintenance based on the main control board switch 4 to apply to various unexpected situations on site.

[0090] The concrete wall maintenance system based on intelligent AI linkage spraying of the present example can effectively solve the problem of regular maintenance of concrete wall surface. The system can be installed through traditional frame, climbing formwork and other frames in actual application, and the poured concrete can be maintained in time to ensure the quality of concrete pouring. At the same time, the metal spraying device is turned on and off in time through the intelligent identification system, water is saved, and green construction of the construction site is realized.

[0091] Referring to Figure 4 The concrete wall maintenance system based on intelligent AI linkage spraying of the present example realizes the basic process of automatic spraying maintenance for concrete wall surface.

[0092] On each concrete wall surface that needs to be maintained, the concrete wall maintenance system based on intelligent AI linkage spraying is erected and arranged through traditional frame, climbing formwork and other frames.

[0093] When the system is working, the multiple intelligent cameras 5 deployed towards the concrete wall surface to be maintained respectively identify and capture images in the corresponding area of the concrete wall surface to be maintained in real time or at regular intervals, and each image is subjected to deep learning classification processing to determine the concrete dryness, concrete color and spraying maintenance state of the corresponding area of the concrete wall surface to be maintained, and the corresponding parameter data is transmitted to the data processing unit 6.

[0094] The data processing unit 6 transmits the picture data parameters identified by the intelligent AI algorithm-based cameras to the background management unit.

[0095] The background management unit statistically analyzes the collected parameters, and votes according to the dryness, color and spraying maintenance state of the concrete wall surface to be maintained identified by the multiple intelligent cameras 5 respectively:

[0096] If it is determined that maintenance is needed, the processed data is reported to the cloud platform, and the control parameters are fed back to the data processing unit 6; the data processing unit 6 realizes local display, such as LED display red light, and linkage of the metal spraying device 2 corresponding to the corresponding area of the concrete wall surface to be maintained to spray the wall;

[0097] If it is determined that the maintenance is normal, the background will transmit the data to the data processing unit 6, and the localization display is realized through the data processing unit 6, such as the LED display green light, and the camera is re-captured at the same time, and the metal spraying device 2 is in the closed state.

[0098] Referring to Figure 5 The flowchart of the concrete wall maintenance system based on the intelligent AI linkage spraying is shown in the figure.

[0099] When the power supply is interrupted or other objective factors affect the spraying maintenance, the maintenance system can be manually controlled by the main control board switch 4 to be in the open state; at this time, the main control board switch 4 will control the corresponding water pump 1 switch to be opened. At this time, the water pump 1 will send the fire riser of the lower water supply main pipe to the water supply branch pipe 3 of the upper concrete wall body outer frame, and the metal spraying device 2 connected therewith will spray the wall body for maintenance.

[0100] In this process, the water pressure can be adjusted by manually controlling the main control board switch 4, and the specific implementation process is as above, and the data processing unit 6 links the main control board switch 4 to realize the local display.

[0101] The above shows and describes the basic principles, main features and advantages of the present application. Those skilled in the art should understand that the present application is not limited to the above examples, and the above examples and descriptions in the specification are only to illustrate the principles of the present application. Without departing from the spirit and scope of the present application, various changes and improvements can be made to the present application, and these changes and improvements all fall within the scope of the present application. The scope of protection of the present application is defined by the appended claims and their equivalents.

Claims

1. A smart AI-based linkage spraying concrete wall maintenance system, characterized in that, The system comprises a frame body, a plurality of intelligent recognition units, a data processing unit, a water supply pipeline, a plurality of spraying units and a background management unit. The frame body is arranged on the concrete wall surface to be maintained. The water supply pipeline is arranged on the frame body, and the plurality of spraying units are distributed on the frame body and connected to the water supply pipeline. The plurality of intelligent recognition units are distributed on the frame body and connected to the data processing unit. The data processing unit is connected to the plurality of spraying units, the plurality of intelligent recognition units and the background management unit. The background management unit identifies the maintenance state by decision voting based on the obtained environmental data, monitoring images and the corresponding wall parameters identified by each intelligent recognition unit, judges the wall area that needs to be sprayed and maintained, and controls the spraying unit at the corresponding position to start spraying.

2. The smart AI-based linked spraying concrete wall maintenance system according to claim 1, wherein The water supply pipeline comprises a main water supply pipeline, an electromagnetic valve, a branch water supply pipeline and a fire riser. 3.The smart AI-based linkage spraying concrete wall maintenance system according to claim 1, characterized in that, The plurality of spraying units are arranged in an array and connected by the branch water supply pipeline.

4. The smart AI-based linked spraying concrete wall maintenance system according to claim 1, wherein The plurality of intelligent recognition units are arranged at the top of the frame body. 5.The smart AI-based linkage spraying concrete wall maintenance system according to claim 1, wherein The intelligent recognition unit uses a deep learning model to extract low-level features of images, further learns high-level features of images based on the mentioned low-level features of images, aggregates all features, completes image classification and distinguishes different concrete image states.

6. The smart AI-based linked spraying concrete wall maintenance system according to claim 1, wherein The background management unit statistically analyzes the monitoring images classified by the intelligent recognition unit, and determines the corresponding concrete wall spraying maintenance state according to the statistical analysis result.

7. A method for maintaining a sprayed concrete wall surface based on intelligent AI linkage, characterized by, The method comprises: The method comprises: The method comprises:

8. The smart AI-based linked spraying concrete wall maintenance method according to claim 7, characterized in that, The method comprises: 9.The smart AI linkage-based spraying concrete wall maintenance method according to claim 7, characterized in that, The maintenance method comprises the following steps: collecting monitoring images of the concrete wall body; classifying the monitoring images by using deep learning; and determining a spraying maintenance state of the concrete wall body according to a result of statistical analysis on the classified monitoring images. 10.The smart AI linkage-based spraying concrete wall maintenance method according to claim 7, wherein, The maintenance method further comprises a step of adjusting the water pressure of the spraying.

Images