A method for controlling and treating the chiseling of an anchor hole

Abstract

The present invention provides a method for controlling and processing the chiseling of anchor holes. An image acquisition module is used to acquire images of the positions of anchor holes on the surface of a capping beam, and a deep convolutional neural network is used to identify the category of the current anchor hole at the current anchor hole position, so as to determine the type of the anchor hole at the current anchor hole position. Among them, the model processing module is used to construct a deep convolutional neural network. A lidar detection device is used to detect the current anchor hole position, determine the depth value at the center of the current anchor hole position and send it to the depth detection module. The image acquisition module determines the area to be polished at the current anchor hole position in real time according to the acquired image of the current anchor hole position. The model processing module calculates the polishing depth of the area to be polished at the current anchor hole position according to the type of the anchor hole at the current anchor hole position and the depth value at the center. The above processing method improves the efficiency and quality of chiseling and polishing of anchor holes.

Description

Technical Field

[0001] The present invention belongs to the field of process chiseling and grinding control data processing, and particularly relates to a method for controlling and processing the chiseling of an anchor hole. Background Art

[0002] A capping beam refers to a cross beam provided at the top of a row of pile piers to support, distribute, and transfer the loads of the upper structure. A reinforced concrete or under-reinforced concrete cross beam is provided on the pier or on the row of piles, and its main function is to support the upper structure of the bridge and transfer all the loads to the lower structure.

[0003] Currently, researchers mainly use the method of casting molds to produce capping beams modularly. After the capping beam is taken out of the mold, corresponding process treatments still need to be carried out further. For example: after the above-mentioned capping beam is formed, it needs to enter the chiseling station for chiseling treatment; the main operation of the above-mentioned chiseling station is to carry out grinding construction on the precast member of the capping beam. However, it is found that the position of the anchor hole (anchor cave) on the semi-finished capping beam is not easy to detect, and the grinding range of the grinding head of the grinding equipment is not easy to control for the anchor hole.

[0004] It is found through research that if the position of the anchor hole on the capping beam cannot be accurately detected and the grinding range cannot be reasonably controlled, the quality of chiseling and grinding cannot be improved, and at the same time, the product quality of the capping beam cannot be guaranteed. Summary of the Invention

[0005] In order to solve the technical problems existing in the above-mentioned prior art, the present application provides a method for controlling and processing the chiseling of an anchor hole.

[0006] In order to achieve the above object, the technical solution of the present invention is realized as follows:

[0007] The present invention provides a method for controlling and processing the chiseling of an anchor hole, which uses a controller to implement the processing operation of chiseling data: the controller includes an image acquisition module, a depth detection module, a sampled data input processing module, and a model processing module;

[0008] The image acquisition module is used to acquire images of the positions of the anchor holes on the surface of the capping beam, and a depth convolutional neural network is used to identify the category of the current anchor hole at the current anchor hole position, and determine the type of the anchor hole at the current anchor hole position; wherein, the model processing module is used to construct a depth convolutional neural network;

[0009] A lidar detection device is used to detect the current anchor hole position, determine the depth value at the center of the current anchor hole position (circular anchor hole), and send it to the depth detection module;

[0010] The sampled data input processing module inputs the type of the anchor hole at the current anchor hole position and the depth value at the center into the model processing module;

[0011] The image acquisition module determines the area to be polished at the current anchor hole position in real time based on the acquired image of the current anchor hole position; the model processing module calculates the polishing depth of the area to be polished at the current anchor hole position according to the anchor hole type at the current anchor hole position and the depth value at the center.

[0012] Preferably, as an implementable solution; before using the image acquisition module to acquire images of the anchor hole positions on the surface of the capping beam, the following operating steps are further included:

[0013] Use a lidar detection device to perform an initial height detection on the surface of the current capping beam. If it is found that the surface height detected currently is lower than or equal to the standard threshold, it is determined that the current surface of the current capping beam is the anchor hole position.

[0014] Preferably, as an implementable solution; use a lidar detection device to detect the current anchor hole position and determine the depth value at the center of the current anchor hole position (circular anchor hole), including the following operating steps:

[0015] Use a lidar detection device to acquire an image of the current anchor hole position to determine the edge trajectory image of the anchor hole position;

[0016] Determine the center of the current anchor hole position (circular anchor hole) as the target sampling point according to the edge trajectory image of the anchor hole position;

[0017] The lidar detection device performs laser detection on the target sampling point to determine the depth value at the center of the current anchor hole position.

[0018] Preferably, as an implementable solution; before using the image acquisition module to acquire images of the anchor hole positions on the surface of the capping beam, a depth convolutional neural network model is further established, including the following operating steps:

[0019] The model processing module acquires a sample image set by acquiring images of the anchor hole positions on the surface of the capping beam, and extracts the anchor hole shape features and the anchor hole edge texture image features on each sample image; a depth convolutional neural network is constructed based on the anchor hole shape features and the anchor hole edge texture image features.

[0020] Preferably, as an implementable solution; after calculating the polishing depth of the area to be polished at the current anchor hole position, the polishing actions on the current area to be polished are further adjusted in order.

[0021] Preferably, as an implementable solution; the adjustment of the order of the polishing actions on the current area to be polished includes the following operating steps:

[0022] Pool the areas with the same polishing depth in the current area to be polished into the same polishing set;

[0023] Repeat the above operation steps multiple times to obtain multiple sets of grinding areas with different grinding depths in the current area to be ground, and perform sorting on the grinding depths to obtain a list of sequences of grinding area groups;

[0024] When grinding, preferably select a group of grinding areas in the list of sequences of grinding area groups in the order from the larger grinding depth to the smaller grinding depth for grinding treatment.

[0025] Preferably, as an implementable solution; the adjustment of the order of the grinding actions on the current area to be ground includes the following operation steps:

[0026] Collect the areas with the same grinding depth in the current area to be ground into the same grinding set; at the same time, fuse and splice the adjacent areas in the current area to be ground to obtain the final multiple sets of grinding areas, and it is required that the difference value between the grinding depths of the current areas to be ground that can fuse and splice the adjacent areas is less than the standard difference value;

[0027] When grinding, preferably select multiple sets of grinding areas that can fuse and splice adjacent areas for grinding treatment.

[0028] Preferably, as an implementable solution; the image acquisition module determines the area to be ground at the current anchor hole position in real time according to the acquired image of the current anchor hole position, specifically including the following operation steps:

[0029] Use a lidar detection device to determine the current working position of the scarifying and grinding device in real time;

[0030] According to the current working position of the scarifying and grinding device, use an image acquisition device to follow the current working position of the scarifying and grinding device to perform image scanning and acquisition on the current bent cap;

[0031] After scanning, obtain the image of the current anchor hole position and finally determine the position where the current anchor hole position is located as the area to be ground at the current anchor hole position.

[0032] Preferably, as an implementable solution; the operation of preferably selecting multiple sets of grinding areas that can fuse and splice adjacent areas for grinding treatment when grinding includes the following operation steps:

[0033] When grinding, number the multiple sets of grinding areas that fuse and splice adjacent areas, and arrange them in descending order according to the number of adjacent areas fused and spliced within a set of grinding areas to obtain a list of grinding area groups;

[0034] Polish a set of polishing areas in descending order from the highest to the lowest in the polishing area group list. At the same time, when polishing each polishing area within a set of polishing areas in the current order, use the polishing area with the largest polishing depth among the multiple polishing areas within the group as the starting position, and determine the target polishing area as the last polishing area within the group;

[0035] The target polishing area is the polishing area with the smallest polishing depth among the multiple polishing areas within the group; and when polishing the next set of polishing areas, use the polishing area with the largest polishing depth among the multiple polishing areas within the next set of polishing areas as the starting position, and use the polishing area with the smallest polishing depth among the multiple polishing areas within the group as the last polishing area, and continuously repeat the above operations until all the polishing areas within all the groups in the polishing area group list are polished.

[0036] Preferably, as an implementable solution; after the lidar detection device performs laser detection on the target sampling point to determine the depth value at the center of the current anchor hole position, it further includes determining the depth distance of the target sampling point for detecting the next anchor hole position, specifically including the following operations:

[0037] The lidar detection device detects the current target sampling point, and waits for the image acquisition module to send an identification instruction for the next anchor hole position on the surface of the capping beam after the detection is completed;

[0038] After the lidar detection device receives the identification confirmation instruction, it tracks and determines the next anchor hole position on the surface of the capping beam; and performs laser detection on the target sampling point of the next anchor hole position.

[0039] An anchor hole roughening control processing method according to an embodiment of the present application uses two parameters as input reference quantities. When polishing the current anchor hole position, it considers both the depth value at the center of the current anchor hole position (circular anchor hole) and the anchor hole type at the current anchor hole position. In this way, different targeted roughening and polishing controls can be performed according to the anchor hole type of the current anchor hole position, and more precise polishing control can be further performed on the current anchor hole according to the depth value at the center; in this way, effective polishing for different shapes of anchor hole types and different depth anchor holes can be improved, avoiding the negative impacts of over-polishing and insufficient polishing depth. Description of the Drawings

[0040] The drawings described herein are used to provide a further understanding of the present application and constitute a part of the present application. The schematic embodiments of the present application and their descriptions are used to explain the present application and do not constitute an improper limitation to the present application. In the drawings:

[0041] Figure 1It is a main operation step flowchart of a method for controlling and processing the chiseling of an anchor hole in an embodiment of the present application;

[0042] Figure 2 It is a schematic diagram of a specific implementation process during the execution of step S100 of the method for controlling and processing the chiseling of an anchor hole in an embodiment of the present application;

[0043] Figure 3 It is a schematic diagram of another specific implementation process during the execution of step S200 of the method for controlling and processing the chiseling of an anchor hole in an embodiment of the present application;

[0044] Figure 4 It is a schematic diagram of a specific implementation process for adjusting the sequence of the grinding actions on the current area to be ground in the method for controlling and processing the chiseling of an anchor hole in an embodiment of the present application;

[0045] Figure 5 It is a schematic diagram of another specific implementation process for adjusting the sequence of the grinding actions on the current area to be ground in the method for controlling and processing the chiseling of an anchor hole in an embodiment of the present application;

[0046] Figure 6 It is a schematic diagram of another specific implementation process during the execution of step S400 of the method for controlling and processing the chiseling of an anchor hole in an embodiment of the present application;

[0047] Figure 7 It is a schematic diagram of a specific implementation process during the execution of step S200 of the method for controlling and processing the chiseling of an anchor hole in an embodiment of the present application. Detailed implementation manners

[0048] Next, the technical solutions of the present invention will be described clearly and completely with reference to the accompanying drawings. Obviously, the described embodiments are some but not all of the embodiments of the present invention. All other embodiments obtained by those of ordinary skill in the art without creative efforts based on the embodiments of the present invention belong to the scope of protection of the present invention.

[0049] In the description of the present invention, it should be noted that certain terms indicating orientation or positional relationships are only for the convenience of describing the present invention and simplifying the description, rather than indicating or implying that the device or element referred to must have a specific orientation, be constructed and operated in a specific orientation, and therefore should not be construed as a limitation of the present invention.

[0050] In the description of the present invention, it should be noted that the term "connection" should be understood in a broad sense. For example, it can be a fixed connection, a detachable connection, or an integral connection; it can be a mechanical connection or an electrical connection; it can be directly connected or indirectly connected through an intermediate medium, and it can be the communication inside two elements. For those of ordinary skill in the art, the specific meanings of the above terms in the present invention can be understood according to specific situations.

[0051] The present invention will be further described in detail below with specific implementation examples in conjunction with the accompanying drawings.

[0052] Embodiment 1

[0053] As Figure 1 shown, Embodiment 1 of the present invention provides a method for controlling and processing the chiseling of an anchor hole, which uses a controller to implement the processing operation of chiseling data: the controller includes an image acquisition module, a depth detection module, a sampled data input processing module, and a model processing module;

[0054] Step S100, use the image acquisition module to collect images of the anchor hole positions on the surface of the capping beam, and use a deep convolutional neural network to identify the category of the current anchor hole at the current anchor hole position, and determine the type of the anchor hole at the current anchor hole position; wherein, the model processing module is used to construct the deep convolutional neural network;

[0055] Step S200, use a lidar detection device to detect the current anchor hole position, determine the depth value at the center of the current anchor hole position (circular anchor hole) and send it to the depth detection module;

[0056] Step S300, the sampled data input processing module inputs the type of the anchor hole at the current anchor hole position and the depth value at the center into the model processing module;

[0057] Step S400, the image acquisition module determines the area to be polished at the current anchor hole position in real time according to the collected image of the current anchor hole position; the model processing module calculates the polishing depth of the area to be polished at the current anchor hole position according to the type of the anchor hole at the current anchor hole position and the depth value at the center.

[0058] Analyzing the main technical solution of the above anchor hole roughening control and treatment method, it can be seen that its main control unit, the main control unit, is the controller. It uses the controller to implement the processing operation of roughening data and uses the controller to implement the intelligent control operation of the output actions of the roughening equipment and the roughening head: The controller includes an image acquisition module, a depth detection module, a sampled data input processing module, a model processing module, a sample construction module, an acquisition module, and a model processing module; In the specific implementation process, the image acquisition module is used to acquire images of the anchor hole positions on the surface of the capping beam, and a deep convolutional neural network is used to perform category recognition on the current anchor hole at the current anchor hole position to determine the anchor hole type at the current anchor hole position; Among them, the model processing module is used to construct a deep convolutional neural network; A lidar detection device is used to detect the current anchor hole position, determine the depth value at the center of the current anchor hole position (circular anchor hole), and send it to the depth detection module; The sampled data input processing module inputs the anchor hole type at the current anchor hole position and the depth value at the center into the model processing module; The image acquisition module determines the area to be polished at the current anchor hole position in real time according to the acquired image of the current anchor hole position; The model processing module calculates the polishing depth of the area to be polished at the current anchor hole position according to the anchor hole type at the current anchor hole position and the depth value at the center.

[0059] The anchor hole roughening control and treatment method provided by the embodiment of the present invention uses double parameters as input reference quantities. When polishing the current anchor hole position, it not only considers the depth value at the center of the current anchor hole position (circular anchor hole), but also considers the anchor hole type at the current anchor hole position. In this way, different targeted roughening and polishing controls can be performed according to the anchor hole type at the current anchor hole position, and at the same time, more precise polishing control can be further performed on the current anchor hole according to the depth value at the center; In this way, effective polishing can be improved for different shapes of anchor hole types and different depth anchor holes, avoiding the negative impacts of over-polishing and insufficient polishing depth.

[0060] See Figure 2 , before using the image acquisition module to acquire images of the anchor hole positions on the surface of the capping beam during the execution of step S100, the following operating steps are also included:

[0061] Step S110, use a lidar detection device to perform an initial height detection on the current surface of the capping beam. If it is found that the currently detected surface height is lower than or equal to the standard threshold, it is determined that the current surface of the current capping beam is the anchor hole position.

[0062] See Figure 3 , during the execution of step S200, using a lidar detection device to detect the current anchor hole position and determine the depth value at the center of the current anchor hole position (circular anchor hole) includes the following operating steps:

[0063] Step S210: Use a lidar detection device to collect an image of the current anchor pocket position and determine the edge trajectory image of the anchor pocket position.

[0064] Step S220: Determine the center of the current anchor pocket position (circular anchor pocket) as the target sampling point according to the edge trajectory image of the anchor pocket position.

[0065] Step S230: The lidar detection device performs laser detection on the target sampling point to determine the depth value at the center of the current anchor pocket position.

[0066] It should be noted that in the technical solution of this embodiment, on the one hand, the above deep convolutional neural network can obtain an augmented image database by augmenting the database, and use the augmented database to train the deep convolutional neural network to obtain a grinding defect recognition model. By increasing the grinding defect images in the database, the training of the convolutional neural network is made more comprehensive, and the accuracy of the recognition model is improved; on the other hand, when real-time image analysis and recognition processing is required, the real-time image to be recognized is input into the grinding defect recognition model of the deep convolutional neural network, which improves the recognition speed of the real-time image, improves the recognition accuracy, ensures the grinding quality of the mold at the grinding and roughening station, quickly recognizes corresponding defects, and facilitates subsequent targeted processing for different grinding defects.

[0067] Preferably, as an implementable solution; before using the image acquisition module to collect an image of the anchor pocket position on the surface of the capping beam, it also includes establishing a deep convolutional neural network model, including the following operating steps:

[0068] The model processing module collects a sample image set by collecting an image of the anchor pocket position on the surface of the capping beam, and extracts the anchor pocket shape features and the anchor pocket edge texture image features on each sample image; a deep convolutional neural network is constructed based on the anchor pocket shape features and the anchor pocket edge texture image features.

[0069] It should be noted that in the technical solution of this embodiment, the model processing module collects a sample image set by collecting an image of the anchor pocket position on the surface of the capping beam, and extracts the anchor pocket shape features and the anchor pocket edge texture image features on each sample image; the above feature extraction involves anchor pocket shape features and anchor pocket edge texture image features (edge texture, edge pixels, etc.), and may even include RGB color features, etc. A deep convolutional neural network (the deep convolutional neural network is a ResNet network) is constructed through the above feature extraction, and a training set of the above deep convolutional neural network is formed for typical anchor pocket types, and the above deep convolutional neural network is continuously strengthened.

[0070] See Figure 4, after calculating the grinding depth of the area to be ground at the current anchor hole position, it further includes adjusting the order of the grinding actions for the current area to be ground.

[0071] The adjustment of the order of the grinding actions for the current area to be ground includes the following operating steps:

[0072] Step S510, gathering the areas with the same grinding depth in the current area to be ground into the same grinding set;

[0073] Step S520, repeatedly cycling the above operating steps to obtain multiple sets of grinding areas with different grinding depths in the current area to be ground and sorting them by grinding depth to obtain a sequence list of grinding area groups;

[0074] Step S530, when grinding, preferentially select a set of grinding areas in the sequence list of grinding area groups in the order from the larger grinding depth to the smaller grinding depth for grinding treatment.

[0075] It should be noted that in the technical solution of this embodiment, after calculating the grinding depth of the area to be ground at the current anchor hole position, it further includes adjusting the order of the grinding actions for the current area to be ground; analyzing the above technical solution, it can be seen that the grinding of the current area to be ground is not carried out disorderly; under the premise of meeting the technical requirements for ensuring the grinding quality, it is executed according to the following operating steps: repeatedly cycling the above operating steps to obtain multiple sets of grinding areas with different grinding depths in the current area to be ground and sorting them by grinding depth to obtain a sequence list of grinding area groups; when grinding, preferentially select a set of grinding areas in the sequence list of grinding area groups in the order from the larger grinding depth to the smaller grinding depth for grinding treatment; in this way, through the above operations, the order of the grinding actions for the current area to be ground can be adjusted, ensuring that the grinding of the area with a deeper depth starts first and gradually proceeds to the grinding area with a shallower grinding depth, thereby avoiding the influence of grinding from shallow to deep on the grinding quality and grinding accuracy (because when grinding from deep to shallow, the space for the grinding head to move horizontally and other movements will become larger).

[0076] It is found that by using the above grinding method, the grinding range can be reasonably controlled, the chiseling grinding quality can be improved, and at the same time, the product quality of the capping beam can be guaranteed.

[0077] See Figure 5 , the adjustment of the order of the grinding actions for the current area to be ground includes the following operating steps:

[0078] Step S540, gathering the areas with the same grinding depth in the current area to be ground into the same grinding set; at the same time, fusing and splicing the adjacent areas of the current area to be ground to obtain the final multiple sets of grinding areas, and it is required that the difference value between the grinding depths of the current areas to be ground that can fuse and splice the adjacent areas is less than the standard difference value;

[0079] Step S550: When grinding, preferably select multiple groups of grinding areas that can fuse and join adjacent areas for grinding.

[0080] It should be noted that in the technical solution of this embodiment, after calculating the grinding depth of the area to be ground at the current anchor hole position, it also includes adjusting the order of the grinding actions for the current area to be ground; analyzing the above technical solution, it can be seen that the grinding of the current area to be ground can be carried out in accordance with the principle of meeting the grinding efficiency: gathering the areas with the same grinding depth of the current area to be ground into the same grinding set; at the same time, fusing and joining the adjacent areas of the current area to be ground to obtain the final multiple groups of grinding areas, and it is required that the difference in the grinding depths of the current areas to be ground that can fuse and join adjacent areas is less than the standard deviation value; when grinding, preferably select multiple groups of grinding areas that can fuse and join adjacent areas for grinding.

[0081] It is found that using the above multiple groups of grinding areas that fuse and join adjacent areas for grinding can grind a large area at one time, thereby reducing the number of grinding repetitions and improving the grinding efficiency.

[0082] See Figure 6 , the image acquisition module determines the area to be ground at the current anchor hole position in real time according to the acquired image of the current anchor hole position, and specifically includes the following operation steps:

[0083] Step S410: Use the lidar detection device to determine and obtain the current working position of the chiseling and grinding device in real time.

[0084] Step S420: According to the current working position of the chiseling and grinding device, use the image acquisition device to follow the current working position of the chiseling and grinding device to perform image scanning and acquisition on the current bent cap.

[0085] Step S430: After scanning, obtain the image of the current anchor hole position and finally determine the position where the current anchor hole position is located as the area to be ground at the current anchor hole position.

[0086] It should be noted that in the technical solution of this embodiment, to collect real-time images, it is necessary to determine the current working position of the chiseling and grinding device (for example, regarding the grinding head as the working point and determining the three-dimensional space coordinates of the spatial working point, etc.), and then according to the current working position of the chiseling and grinding device, use the image acquisition device to follow the current working position of the chiseling and grinding device to perform image scanning and acquisition on the current anchor hole position, so as to obtain the real-time image data of the current anchor hole position.

[0087] Using the above-mentioned following grinding method, the operation position can be determined accurately in real time, and the current capping beam can be scanned and imaged. Obtaining the image of the current anchor hole position is the technical execution basis for subsequent operations.

[0088] Preferably, as an implementable solution; when grinding, it is preferable to select multiple groups of grinding areas that can fuse and stitch adjacent areas for grinding treatment, including the following operation steps:

[0089] When grinding, number the multiple groups of grinding areas that fuse and stitch adjacent areas, and arrange them in descending order according to the number of adjacent areas fused and stitched within a group of grinding areas to obtain a grinding area group list;

[0090] Grind a group of grinding areas in the descending order of the arrangement in the grinding area group list. At the same time, when grinding each grinding area within a group of grinding areas in the current order, use the grinding area with the largest grinding depth among the multiple grinding areas within the group as the starting position, and determine the target grinding area as the last grinding area within the group;

[0091] The target grinding area is the grinding area with the smallest grinding depth among the multiple grinding areas within the group; and when grinding the next group of grinding areas in the order, use the grinding area with the largest grinding depth among the multiple grinding areas within the next group of grinding areas as the starting position, and use the grinding area with the smallest grinding depth among the multiple grinding areas within the group as the last grinding area, and continuously repeat the above operations until all the grinding areas within all groups in the grinding area group list are ground.

[0092] Analyzing the above technical solution, it can be seen that the technical solution adopted in the embodiment of the present invention uses a real-time adjustment control logic during grinding; during grinding, multiple groups of grinding areas that are fused and spliced in adjacent areas are numbered (that is, each group of grinding areas has a number), and they are arranged in descending order according to the number of adjacent areas fused and spliced within a group of grinding areas to obtain a grinding area group list (that is, such a grinding area group list is a group list formed by multiple groups of grinding areas (or called an outer group list)); then the embodiment of the present application also designs a grinding order within the same grinding area group (that is, the inner group order, that is, the grinding area with the largest grinding depth in the multiple grinding areas within the group is used as the starting position, and the target grinding area is determined as the last grinding area within the group). In this way, through the above control logic, it is possible to perform preliminary grinding on a large area of continuously associated adjacent grinding areas outside the group, and at the same time start grinding from deep to shallow within the group. The above control method realizes dual-logic control of the outer group order and the inner group area order, and truly realizes intelligent, high-precision and reasonable grinding control; then it is also more convenient to switch from the grinding of the current order group to the next order group of grinding areas, because the last grinding action of the previous order is the grinding position with the shallowest grinding depth, so it is easier to reduce the upward movement amount of the grinding head and then move down to enter the grinding area of the next order.

[0093] See Figure 7 , after the lidar detection device performs laser detection on the target sampling point to determine the depth value at the center of the current anchor pocket position, it further includes determining the depth distance of the target sampling point for detecting the next anchor pocket position, specifically including the following operations:

[0094] Step S240, the lidar detection device detects the current target sampling point, and waits for the image acquisition module to send an identification instruction for the next anchor pocket position on the surface of the capping beam after the detection is completed;

[0095] Step S250, after the lidar detection device receives the identification confirmation instruction, it tracks and determines the next anchor pocket position on the surface of the capping beam; and performs laser detection on the target sampling point of the next anchor pocket position.

[0096] It should be noted that in the technical solution of this embodiment, the lidar detection device detects the current target sampling point, waits for the image acquisition module to send an identification instruction for the next anchor pocket position on the surface of the capping beam after the detection is completed; then tracks and determines the next anchor pocket position on the surface of the capping beam; through the above method, the anchor pocket position can be continuously determined, and the target sampling point of the next anchor pocket position can be continuously subjected to laser detection. Therefore, it can be said that the depth measurement of the target sampling point of the anchor pocket position and the image acquisition of the anchor pocket position in the embodiment of the present invention are carried out separately, and the operation execution instructions of the two do not interfere with each other, thus achieving a faster operation efficiency.

[0097] In summary, the anchor hole chiseling control method provided by the embodiments of the present invention can not only perform high-precision recognition of the dual-parameter model for the current anchor hole position, so as to perform targeted grinding operation (control the output of the grinding operation) on the current grinding area in real time, thereby significantly improving the quality and efficiency of anchor hole chiseling and grinding.

[0098] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention, rather than to limit them; although the present invention has been described in detail with reference to the foregoing embodiments, those of ordinary skill in the art should understand that they can still modify the technical solutions described in the foregoing embodiments, or perform equivalent replacements on some or all of the technical features; and these modifications or replacements do not make the essence of the corresponding technical solutions deviate from the scope of the technical solutions of the embodiments of the present invention.

Claims

1. A method for controlling and treating the chiseling of an anchor hole, characterized in that, It uses a controller to implement the processing operation of the roughening data: The controller includes an image acquisition module, a depth detection module, a sampled data input processing module, and a model processing module; The image acquisition module is used to acquire images of the anchor hole positions on the surface of the capping beam, and a depth convolutional neural network is used to identify the category of the current anchor hole at the current anchor hole position to determine the type of the anchor hole at the current anchor hole position; wherein, the model processing module is used to construct the depth convolutional neural network; A lidar detection device is used to detect the current anchor hole position to determine the depth value at the center of the current anchor hole position and send it to the depth detection module; The sampled data input processing module inputs the type of the anchor hole at the current anchor hole position and the depth value at the center into the model processing module; The image acquisition module determines the area to be polished at the current anchor hole position in real time according to the acquired image of the current anchor hole position; the model processing module calculates the polishing depth of the area to be polished at the current anchor hole position according to the type of the anchor hole at the current anchor hole position and the depth value at the center; After calculating the polishing depth of the area to be polished at the current anchor hole position, it further includes adjusting the order of the polishing actions for the current area to be polished; The adjustment of the order of the polishing actions for the current area to be polished includes the following operation steps: pooling the areas with the same polishing depth in the current area to be polished into the same polishing set; repeatedly looping through the above operation steps to obtain multiple groups of polishing areas with different polishing depths in the current area to be polished and sorting them by polishing depth to obtain a polishing area group sequence list; when polishing, preferentially select a group of polishing areas in the polishing area group sequence list from the larger polishing depth to the smaller polishing depth for polishing; or, The adjustment of the order of the polishing actions for the current area to be polished includes the following operation steps: Identifying the adjacent areas of the current area to be polished; at the same time, fusing and splicing the adjacent areas of the current area to be polished to obtain the final multiple groups of polishing areas, and requiring that the difference value between the polishing depths of the current areas to be polished that can fuse and splice the adjacent areas is less than the standard deviation value; when polishing, preferentially select multiple groups of polishing areas that can fuse and splice the adjacent areas for polishing.

2. The anchor hole roughening control processing method according to claim 1, characterized in that Before using the image acquisition module to acquire images of the anchor hole positions on the surface of the capping beam, the following operation steps are further included: Using a lidar detection device to perform an initial height detection on the current surface of the capping beam, and if it is found that the currently detected surface height is lower than or equal to the standard threshold, it is determined that the current surface of the current capping beam is the anchor hole position.

3. The anchor hole roughening control processing method according to claim 2, characterized in that Using a lidar detection device to detect the current anchor hole position to determine the depth value at the center of the current anchor hole position, including the following operation steps: Using a lidar detection device to acquire an image of the current anchor hole position to determine the edge trajectory image of the anchor hole position; Determining the center of the current anchor hole position as the target sampling point according to the edge trajectory image of the anchor hole position; The lidar detection device performs laser detection on the target sampling point to determine the depth value at the center of the current anchor hole position.

4. The anchor hole roughening control processing method according to claim 3, characterized in that Before using the image acquisition module to acquire images of the anchor hole positions on the surface of the pier cap, it also includes establishing a deep convolutional neural network model, which includes the following operation steps: The model processing module acquires a sample image set by acquiring images of the anchor hole positions on the surface of the pier cap, and extracts the anchor hole shape features and the anchor hole edge texture image features on each sample image; a deep convolutional neural network is constructed based on the anchor hole shape features and the anchor hole edge texture image features.

5. The method for controlling and processing the chiseling of the anchor hole according to claim 1, wherein The image acquisition module determines the area to be polished at the current anchor hole position in real time according to the acquired image of the current anchor hole position, which specifically includes the following operation steps: The current working position of the chiseling and grinding device is determined in real time through the lidar detection device; According to the current working position of the chiseling and grinding device, the image acquisition device follows the current working position of the chiseling and grinding device to perform image scanning and acquisition on the current pier cap; After scanning, the image of the current anchor hole position is obtained, and finally the position where the current anchor hole position is located is determined as the area to be polished at the current anchor hole position.

6. The anchor hole roughening control processing method according to claim 1, wherein, When grinding, it is preferred to select multiple groups of grinding areas that can fuse and stitch adjacent areas for grinding treatment, which includes the following operation steps: When grinding, the multiple groups of grinding areas that fuse and stitch adjacent areas are numbered, and they are arranged in descending order according to the number of adjacent areas fused and stitched within a group of grinding areas to obtain a grinding area group list; Grinding is performed on a group of grinding areas arranged from the highest to the lowest position in the grinding area group list. At the same time, when grinding each grinding area within a group of grinding areas at the current position, the grinding area with the largest grinding depth among the multiple grinding areas within the group is used as the starting position, and the target grinding area is determined to be the last grinding area within the group; The target grinding area is the grinding area with the smallest grinding depth among the multiple grinding areas within the group; and when grinding the next group of grinding areas, the grinding area with the largest grinding depth among the multiple grinding areas within the group of the next group of grinding areas is used as the starting position, and the grinding area with the smallest grinding depth among the multiple grinding areas within the group is used as the last grinding area, and the above operations are continuously repeated in a loop until all the grinding areas within all groups in the grinding area group list are ground.

7. The anchor hole roughening control processing method according to claim 1, characterized in that After the lidar detection device performs laser detection on the target sampling point to determine the depth value at the center of the current anchor hole position, it also includes determining the depth distance of the target sampling point for detecting the next anchor hole position, which specifically includes the following operations: The lidar detection device detects the current target sampling point, and waits for the image acquisition module to send an identification instruction for the next anchor hole position on the surface of the pier cap after the detection is completed; After the lidar detection device receives the identification confirmation instruction, it tracks and determines the next anchor hole position on the surface of the pier cap; laser detection is performed on the target sampling point of the next anchor hole position.

Images