CN115808425B - Defect identification and application method in concrete member rebound detection process - Google Patents
Defect identification and application method in concrete member rebound detection process Download PDFInfo
- Publication number
- CN115808425B CN115808425B CN202310044327.8A CN202310044327A CN115808425B CN 115808425 B CN115808425 B CN 115808425B CN 202310044327 A CN202310044327 A CN 202310044327A CN 115808425 B CN115808425 B CN 115808425B
- Authority
- CN
- China
- Prior art keywords
- defect
- area
- wall surface
- wall
- rebound
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Active
Links
Images
Landscapes
- Investigating Materials By The Use Of Optical Means Adapted For Particular Applications (AREA)
Abstract
The invention relates to a defect identification and handling method in a concrete member rebound detection process. The defect identification method comprises the steps of firstly collecting wall images of a currently located area in real time, inputting the wall images into a trained wall defect identification model, identifying defect positions and defect characteristic categories in the wall images, and obtaining the defect pixel duty ratio. And then the number of defect characteristic categories and the defect pixel proportion in the wall image are weighted and aggregated to obtain a defect weighting result reflecting the defect degree of the wall. And finally, analyzing the relation between the defect weighted result and a preset judging threshold value. And when the defect weighting result is larger than a preset judging threshold value, judging that the wall surface of the currently positioned area is the defect wall surface. Otherwise, the wall surface is qualified. The defect identification method can effectively identify the defect wall surface area on the concrete member, thereby providing guidance for the rebound detection process and guaranteeing the smooth proceeding of the rebound detection process and the accuracy of the subsequent detection result.
Description
Technical Field
The invention relates to the technical field of building construction, in particular to a defect identification method in a concrete rebound detection process, a defect coping method, a computer terminal and a computer readable storage medium applying the defect identification method, and a rebound detection method of a concrete member applying the defect identification method and the defect coping method.
Background
The existing common concrete nondestructive testing technology generally has the advantages of rebound method intensity measurement, the rebound detection of concrete components generally needs manual operation of a rebound instrument, and for certain building components with higher operation danger coefficient and more areas to be detected, such as piers, building outer walls, dams and the like, the problems of low detection efficiency and limited detection precision exist, and the defects of low safety coefficient and inconvenient operation exist in the mode.
In recent years, with the increasing development of wall climbing robots, for concrete rebound detection, research and development personnel have proposed a scheme of using wall climbing robots or other technical means to replace manual concrete member rebound detection. However, in the practical application process, on the area which is originally arranged on the concrete member by the machine, as the blank wall surface may have defects which obviously cause deviation of rebound values, some defects even cause that the rebound operation of single or multiple measuring points cannot be completed by the rebound instrument, so that the rebound detection result is inaccurate, and further smooth progress of the rebound detection process of the concrete member is limited.
Disclosure of Invention
Based on the above, it is necessary to provide a defect identification and response method in the rebound detection process of a concrete member, aiming at the technical problem that the rebound detection result is not accurate enough because the defect of the wall surface on the concrete member area is difficult to determine in the prior art.
The invention discloses a defect identification method in a concrete rebound detection process, which is used for carrying out defect identification detection on the wall surface of one area to be detected of a concrete member after a rebound instrument is positioned to the wall surface of the area to be detected, so as to judge whether the wall surface of the area is a defect wall surface or not. The defect identification method comprises the following steps:
and acquiring wall images of the currently positioned area in real time.
Inputting the wall image into a pre-trained wall defect recognition model, recognizing defect positions and defect feature types in the wall image, and obtaining pixel duty ratios of all defects in the image according to the defect positions. Wherein the defect characteristic category comprises cracks, pits, honeycombs, pre-buried pipe orifices, floating slurry, oil dirt and coatings.
And carrying out weighted aggregation on the number of defect characteristic categories and the defect pixel duty ratio in the wall image to obtain a defect weighted result for reflecting the defect degree of the wall.
And analyzing the relation between the defect weighted result and a preset judging threshold value. And when the defect weighting result is larger than a preset judging threshold value, judging that the wall surface of the currently positioned area is the defect wall surface. Otherwise, judging the wall surface of the currently positioned area as a qualified wall surface.
As a further improvement of the scheme, before the wall image is input into the wall defect recognition model, gray processing and median filtering processing are further carried out on the acquired wall image, so that the wall image after preprocessing and reinforcing is further obtained.
As a further improvement of the scheme, the training method of the wall defect recognition model comprises the following steps of:
a plurality of wall defect image samples are acquired to form an original data set for model training. Wherein, each wall defect image sample is marked with a defect characteristic type true value and a defect position true value.
Initializing the constructed model to be trained, and setting a loss function and training parameters.
And carrying out defect identification on the wall defect image sample by using the model to be trained to obtain a defect feature type detection value and a defect position detection value in the wall defect image sample.
A first recognition loss between the defect feature class true value and the defect feature class detection value and a second recognition loss between the defect location true value and the defect location detection value are calculated based on the loss function.
And adjusting the neural network parameters of the model to be trained according to the first recognition loss and the second recognition loss, and further obtaining the wall defect recognition model after the iterative training is completed.
As a further improvement of the above-described scheme, the average light intensity value in a predetermined area around the region is acquired in real time before the wall image of each region is acquired. When the average light intensity value is lower than a preset light intensity threshold value, the wall surface of the current area is supplemented with light by a light source.
The invention also discloses a defect handling method in the concrete rebound detection process, which is used for deploying a qualified standby area for the rebound instrument when the rebound instrument is positioned to one of the areas on the wall surface to be detected of a concrete member and the wall surface of the area is identified as the defect wall surface by using any defect identification method. The defect coping method comprises the following steps:
and taking the currently positioned area as a reference area, acquiring the integral coordinates of the wall surface to be measured and the local coordinates of the reference area on the wall surface to be measured according to a component parameter database, and further dividing a deployable area corresponding to the reference area on the wall surface to be measured.
The deployment spacing of adjacent regions, and the edge spacing of the regions from the ends of the members, are set to define at least one candidate region within the deployable area.
Positioning a resiliometer to each candidate area in turn, and marking the walls of each candidate area as defective walls or acceptable walls in turn according to the defect identification method as claimed in any one of claims 1 to 4.
Counting the number of all candidate areas marked as qualified walls in the deployable area, and making the following decision based on the number:
when the qualification number of the alternative area is more than or equal to 1, the alternative area closest to the reference area is used as the standby area.
And when the qualification number of the alternative areas is less than 1, counting defect weighting results of the reference area and all the alternative areas, selecting an area with the smallest defect weighting result from the defect weighting results as a standby area, marking the standby area as abnormal, and sending the standby area to an interactive end.
As a further improvement of the above scheme, the deployment interval of the adjacent areas is not higher than 2m. The edge spacing between the region and the end of the member is no more than 0.5m and no less than 0.2m.
The invention also discloses a rebound detection method of the concrete member, which is used for controlling a rebound detector to move along the wall surface to be detected of a standard concrete member and keeping the axis of the rebound detector vertical to the wall surface to be detected so as to carry out rebound detection. The thickness, length and height directions of the wall surface to be measured are defined as an X axis, a Y axis and a Z axis respectively. The rebound detection method comprises the following steps:
s1, respectively obtaining a preset track I and a preset track II which are matched with a wall surface to be detected according to a component parameter database.
The first preset track is sequentially connected with first positioning points corresponding to all the areas of the wall surface to be detected in series. The second preset track is sequentially connected with the second positioning points corresponding to all the measuring points in each measuring area in series. The planes of the first preset track and the second preset track are parallel to the Y-Z plane.
S2, after the resiliometer is positioned to an initial point on the first preset track, the resiliometer is controlled to move to a positioning point along the first preset track, and the positioning of the measuring area is completed.
S3, judging whether the wall surface of the currently positioned area is a defect wall surface by adopting any defect identification method.
S4, when the wall surface of the currently positioned area is a defect wall surface, adopting any defect coping method to deploy a qualified standby area for the resiliometer, and controlling the resiliometer to be positioned to the standby area.
S5, if the wall surface of the currently positioned area is a qualified wall surface or the resiliometer is deployed to a qualified standby area, the resiliometer is controlled to move to one of positioning points II along a preset track II so as to finish positioning of the measuring point.
S6, controlling the resiliometer to approach the wall surface to be measured from an initial axial position along the X-axis direction until the rebound test of the current measuring point is completed, collecting rebound value data of the current time, and then recovering the resiliometer to the initial axial position.
S7, executing S5-S6 circularly until the rebound value data acquisition of all the measuring points in the current area is completed.
S8, judging whether the current detection area for completing rebound data acquisition is the last detection area in the first preset track. And if so, ending the moving process along the first preset track. Otherwise, returning to the step S2 to control the resiliometer to continue to move to the next positioning point I until the rebound data acquisition is completed for all the preset number of areas of the wall surface to be detected.
S9, calculating the average rebound value of each measuring area according to the rebound data of the measuring points of each measuring area.
As a further improvement of the above scheme, in S4, on the wall surface to be measured, the standby area and all areas corresponding to the preset track one are independent from each other.
The invention also discloses a computer terminal which comprises a memory, a processor and a computer program stored on the memory and capable of running on the processor, wherein the processor realizes the steps of the defect identification method in the concrete rebound detection process when executing the program.
The invention also discloses a computer readable storage medium, on which a computer program is stored which, when executed by a processor, implements the steps of the defect identification method of any one of the above in the concrete rebound detection process.
Compared with the prior art, the technical scheme disclosed by the invention has the following beneficial effects:
1. the defect identification method is used for acquiring a wall image of a region to be detected on the wall surface of the concrete member after the resiliometer is positioned to the region, and identifying and obtaining the defect position, the defect characteristic category and the defect pixel ratio in the wall surface image by utilizing a pre-trained wall surface defect identification model. And finally, carrying out weighted aggregation on the number of all defect characteristic categories and the defect pixel proportion in the detection area to obtain a defect weighted result capable of reflecting the defect degree of the wall surface, and judging the detection area as the defect or the qualified wall surface according to the magnitude relation between the weighted result and a preset judgment threshold value. The technical scheme of the invention can identify the defect area on the wall surface in the rebound detection process of the machine, and provide guidance for the rebound detection process of each area, thereby ensuring the smooth proceeding of the rebound detection process and being beneficial to indirectly improving the accuracy of the subsequent rebound detection result.
2. According to the defect coping method, after the original detection area is identified as the defect detection area, a plurality of candidate detection areas with qualified wall quality are deployed for the rebound instrument preliminarily, and then the standby detection areas with qualified wall quality are determined, so that the replacement detection areas can be provided for the defect detection areas in the rebound detection process of the machine, the rebound detection process can be smoothly carried out, and the accuracy of the final rebound detection result can be directly improved.
3. The rebound detection method of the concrete member is realized by the following steps: the defect area in the originally planned virtual area can be effectively searched out without manually drawing actual area grid lines by a detector, manually controlling the rebound instrument, improving rebound detection precision and guaranteeing the relative safety of the detector, and the qualified standby area is redeployed for the rebound instrument aiming at the defect area, so that the rebound detection precision and efficiency are further improved.
Drawings
FIG. 1 is a flow chart of a method for detecting rebound of a concrete member in example 1 of the present invention;
fig. 2 is a schematic perspective view of a wall climbing robot according to embodiment 1 of the present invention disposed on a wall surface to be measured of a concrete member;
FIG. 3 is an enlarged view of FIG. 2 at A;
FIG. 4 is a schematic diagram illustrating the distribution of a plurality of measurement areas on a wall surface to be measured in embodiment 1 of the present invention;
FIG. 5 is a schematic diagram illustrating the correspondence between the plurality of measurement areas and the plurality of positioning points I in FIG. 4;
FIG. 6 is a schematic diagram of a plurality of first positioning points connected into a first predetermined track in FIG. 5;
fig. 7 is a schematic diagram of correspondence between a single measurement area and a plurality of second positioning points in embodiment 1 of the present invention;
FIG. 8 is a schematic diagram of a plurality of second positioning points connected to form a second preset track in FIG. 7;
FIG. 9 is a flowchart of a defect identification method in embodiment 1 of the present invention;
FIG. 10 is a flowchart of a defect coping method in embodiment 1 of the present invention;
FIG. 11 is a schematic diagram of the present invention in which a plurality of alternative regions are planned for the reference region in embodiment 1;
fig. 12 is a schematic perspective view of a wall surface inspection robot in embodiment 4 of the present invention;
fig. 13 is a schematic perspective view of the wall surface inspection robot in fig. 9.
Description of main reference numerals: 1. a wall surface to be measured; 2. a wall climbing device; 21. a mechanical arm; 22. a vacuum chuck; 3. a resiliometer; 4. a measuring area; 5. a driving device; 51. an X-axis linear motor; 52. a Y-axis linear motor; 53. a Z-axis linear motor; 6. wall climbing robot; 7. a carrier.
The foregoing general description of the invention will be described in further detail with reference to the drawings and detailed description.
Detailed Description
The following description of the embodiments of the present invention will be made clearly and completely with reference to the accompanying drawings, in which it is apparent that the embodiments described are only some embodiments of the present invention, but not all embodiments. All other embodiments, which can be made by those skilled in the art based on the embodiments of the invention without making any inventive effort, are intended to be within the scope of the invention.
It is noted that when an element is referred to as being "mounted to" another element, it can be directly on the other element or intervening elements may also be present. When an element is referred to as being "disposed on" another element, it can be directly on the other element or intervening elements may also be present. When an element is referred to as being "secured to" another element, it can be directly secured to the other element or intervening elements may also be present.
Unless defined otherwise, all technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the art to which this invention belongs. The terminology used herein in the description of the invention is for the purpose of describing particular embodiments only and is not intended to be limiting of the invention. The term "or/and" as used herein includes any and all combinations of one or more of the associated listed items.
Example 1
The embodiment provides a rebound detection method of a concrete member, which is used for controlling a rebound detector to move along a wall surface to be detected of a standard concrete member and keeping the axis of the rebound detector perpendicular to the wall surface to be detected for rebound detection.
Referring to fig. 2 and 3, the rebound detection method may be implemented by controlling a wall climbing robot 6 to climb a wall on a wall 1 to be measured of a concrete member, and the rebound apparatus 3 may be mounted on the wall climbing robot 6 with its axis maintained perpendicular to the wall 1 to be measured. In addition, a driving structure for driving the resiliometer 3 to approach or depart from the wall surface 1 to be detected can be arranged on the wall climbing robot 6, so that the rebound detection method can be realized. The wall climbing robot 6 can adopt a bionic robot in some embodiments, is similar to a gecko, can be a bionic adhesion material at the contact end with a wall, can also be a vacuum chuck and other structures, automatically performs wall climbing walking through a built-in controller, and can also be manually controlled by manpower. In other embodiments, the wall climbing robot may also adopt an existing automatic wall climbing structure, so long as the condition of interference such as rebound reaction force, impact force, self gravity and the like generated by the wall in the rebound detection process is satisfied, and the robot can also be stably positioned or moved. As for the specific construction and principle of the wall climbing robot, the specific construction and principle are not used as the invention point of the rebound detection method in the present embodiment, and will not be described herein.
The thickness, length and height directions of the wall surface to be detected are defined as an X axis, a Y axis and a Z axis respectively, and the rebound detection method of the embodiment comprises the following steps:
s1, respectively obtaining a preset track I and a preset track II which are matched with a wall surface to be detected according to a component parameter database.
The first preset track is sequentially connected with first positioning points corresponding to all the areas of the wall surface to be detected in series. The second preset track is sequentially connected with the second positioning points corresponding to all the measuring points in each measuring area in series. The planes of the first preset track and the second preset track are parallel to the Y-Z plane.
The component parameter database may include: the method comprises the steps of forming a plurality of standard concrete members, namely, the shape and size parameters of the standard concrete members, the number of the measuring areas on the wall surface of each standard concrete member, the arrangement mode parameters of the measuring areas, the unit area of the measuring areas, and the coordinate parameters of a first positioning point and a second positioning point corresponding to the wall surface of each standard concrete member. Since concrete elements are generally pre-designed or pre-fabricated in the construction process, by building such an element parameter database, applications in a variety of standardized buildings, such as concrete walls, beams or columns of different floors, may be satisfied, even if not belonging to the same floor, but having substantially common relevant dimensional parameters. In addition, the component parameter database may be updated and optimized in real time.
It should be understood that the wall surface to be measured of the concrete member is a plane, a corresponding X-Y-Z coordinate system can be established by acquiring the relevant parameters of each standard concrete member and taking a certain point on the wall surface to be measured as an origin, and the coordinate parameters of each positioning point one and the positioning point two are the coordinates in the coordinate system, and similarly, the preset track one and the preset track two can also be generated in the coordinate system. In addition, according to the coordinate parameters of the first positioning point and the second positioning point, the first preset track and the second preset track corresponding to the wall surface of each standard concrete member can be generated by Direct path networks (direct path algorithm) through the data processor in advance, direct path networks is connected with all the points of the destination to obtain the most direct path, namely, all the first positioning points are connected to obtain the first preset track, and all the second positioning points are connected to obtain the second preset track. Of course, in other embodiments, the preset trajectory may be solved by using Minimal path networks (shortest path network) algorithm, so as to obtain a shorter optimized trajectory.
Referring to fig. 4, in this embodiment, taking a building outer wall as a concrete member, the length of the building outer wall is 5m, the height of the building outer wall is 3m, and 10 areas (the areas are marked as 4 in the drawing) can be arranged on the building outer wall. Fig. 4 shows a schematic distribution of 10 areas on the wall 1 to be measured. The area of each area can be controlled to be 0.04m 2 I.e. square areas with a side length of 0.2m, and equally dividing a single area into 16 measuring points. The distance between two adjacent areas should not be greater than 2m, the distance between the areas and the end of the member or the edge of the construction joint should not be greater than 0.5m, and should not be less than 0.2m. Other arrangement methods and pretreatment of components before detection can be performed by referring to the guidance in JG3/T-2011 of the industry standard technical specification for detecting the compressive strength of concrete by a rebound method, and are not repeated here.
Referring to fig. 5 and fig. 6, in fig. 5, a plurality of first positioning points respectively correspond to a plurality of measuring areas; in fig. 6, a plurality of positioning points can be connected to form a local route of a preset track one by the above straight line path algorithm, and the route is parallel to the wall surface to be measured.
Referring to fig. 7 and 8, fig. 7 is a schematic diagram showing a plurality of measurement points and a plurality of second positioning points in a single area, and each area is divided into measurement points of 16 grids. In fig. 8, a plurality of second positioning points may be connected to form a second predetermined track, which is parallel to the wall surface to be measured.
It should be noted that, the above-mentioned division of the area, and the areas and the measuring points shown in the drawings do not draw actual lines on the surface of the concrete member, but write the relevant parameter information of the area, the measuring point, the first positioning point, the second positioning point, the first preset track and the second preset track into the relevant control element of the wall climbing robot 6 in the form of program codes, and can control the rebound instrument to move according to the program codes in the following actual rebound instrument moving and rebound detecting processes.
S2, after the resiliometer is positioned to an initial point on the first preset track, the resiliometer is controlled to move to a positioning point along the first preset track, and the preliminary area positioning is completed.
In this embodiment, the initial point of the preset trajectory one may be set at one specific corner of the concrete member. As shown in fig. 2, whenever the wall climbing robot 6 is positioned to the corner of the member in a uniform positioning manner, the resiliometer 3 is automatically aligned and positioned at an initial point of the first preset track, and then the wall climbing robot 6 can automatically control the resiliometer to move to the first positioning point corresponding to the first area along the first preset track according to the current position of the wall climbing robot 6.
S3, judging whether the wall surface of the currently positioned area is a defect wall surface or not. Specifically, the present embodiment further provides a defect identification method in the concrete rebound detection process, which is configured to perform defect identification detection on the wall surface of the currently located area, so as to determine whether the wall surface of the area is a defective wall surface.
Referring to fig. 9, the defect identifying method includes the following steps, S31 to S34.
S31, acquiring wall images of the currently located area in real time.
In this embodiment, an industrial camera may be used to capture a wall image, and before capturing, an average light intensity value in a preset area around the area may be acquired in real time, and when the average light intensity value is lower than a preset light intensity threshold, the wall of the current area is further supplemented by a light source. In addition, under the actual shooting scene, the wall image may be affected by factors such as dim light, backlight, noise and the like, so that the wall image needs to be preprocessed and enhanced, and the requirement of subsequent image recognition is met. Specifically, gray processing and median filtering processing can be performed on the acquired wall image, so that the wall image after pretreatment and enhancement is obtained.
S32, inputting the wall image into a pre-trained wall defect recognition model, recognizing defect positions and defect feature types in the wall image, and obtaining pixel duty ratios of all defects in the image according to the defect positions. Wherein the defect characteristic category comprises cracks, pits, honeycombs, pre-buried pipe orifices, floating slurry, oil dirt and coatings.
The wall defect recognition model can adopt the existing neural network model, such as YOLO and the like, and needs to be trained before the model is used. Specifically, the training method of the wall defect recognition model comprises the following steps of (1) - (5).
(1) A plurality of wall defect image samples are acquired to form an original data set for model training. Wherein, each wall defect image sample is marked with a defect characteristic type true value and a defect position true value.
(2) Initializing the constructed model to be trained, and setting a loss function and training parameters.
(3) And carrying out defect identification on the wall defect image sample by using the model to be trained to obtain a defect feature type detection value and a defect position detection value in the wall defect image sample.
(4) A first recognition loss between the defect feature class true value and the defect feature class detection value and a second recognition loss between the defect location true value and the defect location detection value are calculated based on the loss function.
(5) And adjusting the neural network parameters of the model to be trained according to the first recognition loss and the second recognition loss, and further obtaining the wall defect recognition model after the iterative training is completed.
S33, carrying out weighted aggregation on the defect feature class number and the defect pixel ratio in the wall image to obtain a defect weighted result for reflecting the wall defect degree.
In the invention, compared with the number of defect characteristic categories, the defect pixel ratio occupies higher weight, and specific weight distribution is also determined according to the building parameters (such as surface roughness, water-cement ratio, water consumption, sand ratio and the like) of the concrete member and combined with experience values. The weighting process is performed because the defect pixel ratio of the surface of the wall surface is high in consideration of some wall surfaces, but the defect characteristic types are few, for example, only one thicker embedded pipe orifice is positioned in the area, and the other areas are smooth, and in this case, if only the defect characteristic types are considered, the area may be considered to be a qualified wall surface. In addition, there are some wall surfaces with more defect characteristics, but the defect pixel ratio is low, and in this case, the rebound detection of the area is still affected to some extent.
S34, analyzing the relation between the defect weighted result and a preset judging threshold value. And when the defect weighting result is larger than a preset judging threshold value, judging that the wall surface of the currently positioned area is the defect wall surface. Otherwise, judging the wall surface of the currently positioned area as a qualified wall surface.
In this embodiment, the preset determination threshold may also be determined according to the building parameters of the concrete member and in combination with empirical values.
S4, when the wall surface of the currently positioned area is a defect wall surface, disposing a qualified standby area for the resiliometer, and controlling the resiliometer to be positioned to the standby area.
It should be noted that the present invention can be applied to concrete members (such as piers, building walls, dam bodies, etc.) with larger detection areas, so that the total area of all the areas on the wall surface to be detected only occupies a small part of the wall surface to be detected, and if the individual areas on the original preset track one are defective wall surfaces, the replacement qualified standby areas can be redeployed near the preset track one. Specifically, the embodiment also provides a defect handling method in the concrete rebound detection process, which is used for deploying a qualified standby area for a rebound instrument when the defect identification method identifies that the wall surface of the area is a defect wall surface, and the standby area is mutually independent from all areas corresponding to a preset track one on the wall surface to be detected.
Referring to FIG. 10, the defect coping method includes the following steps S41-S44.
S41, taking the currently positioned area as a reference area, acquiring the integral coordinates of the wall surface to be measured and the local coordinates of the reference area on the wall surface to be measured according to a component parameter database, and further dividing the deployable area corresponding to the reference area on the wall surface to be measured.
S42, setting the deployment interval of adjacent areas and the edge interval between the areas and the end parts of the components, so as to plan a plurality of alternative areas in the deployable areas. In this embodiment, the deployment interval between adjacent areas is not higher than 2m. The edge spacing between the region and the end of the member is no more than 0.5m and no less than 0.2m.
Referring to fig. 11, a is disposed around the left side region 1 on the original predetermined track one 1 、a 2 、a 3 、a 4 、a 5 Five candidate areas are provided, although other numbers of candidate areas may be provided for other areas to be deployed.
In addition, in this embodiment, the aforementioned component parameter database may further include an alternative area deployment mode of each standard concrete component, and the relevant drawings are drawn on the computer according to the relevant detection personnel in the early stage, and written into the component parameter database in the form of program codes. The alternative test may also be called directly from the component parameter database each time the alternative test is planned in S42.
S43, sequentially positioning the resiliometer to each candidate area, and simultaneously sequentially marking the wall surface of each candidate area as a defect wall surface or a qualified wall surface according to the defect identification method in the embodiment 1.
S44, counting the number of alternative areas marked as qualified wall surfaces in the deployable area, and making the following decision according to the number:
when the qualification number of the alternative area is more than or equal to 1, the alternative area closest to the reference area is used as the standby area. It should be noted that, if there are a plurality of candidate areas closest to the reference area, one of the candidate areas may be selected randomly as the candidate area.
When the qualified number of the alternative areas is less than 1, counting defect weighted results of the reference area and all the alternative areas, selecting an area with the smallest defect weighted result from the reference area and the alternative areas as a standby area, marking the standby area as abnormal, and sending the abnormal standby area to an interaction end, so as to remind relevant detection personnel for subsequent adjustment or improvement.
S5, if the wall surface of the currently positioned area is a qualified wall surface or the resiliometer is deployed to a qualified standby area, the resiliometer is controlled to move to one of positioning points II along a preset track II so as to finish positioning of the measuring point.
S6, controlling the resiliometer to approach the wall surface to be measured from an initial axial position along the X-axis direction until the rebound test of the current measuring point is completed, collecting rebound value data of the current time, and then recovering the resiliometer to the initial axial position.
In addition, by acquiring the real-time distance between the housing of the resiliometer 3 and the wall surface 1 to be measured, the following determination is made according to the real-time distance:
and when the real-time distance is a first preset distance value, judging that the rebound instrument is at the initial axial position.
And when the real-time distance is a second preset distance value, judging that the flicking rod of the resiliometer is just contacted with the wall surface to be measured.
And when the real-time distance is a third preset distance value, judging that the bouncing rod of the resiliometer triggers the rebound, and completing the rebound test of the current measuring point.
Wherein the first distance preset value is less than the second distance preset value and less than the third distance preset value. The first distance preset value, the second distance preset value and the third distance preset value can be specifically set according to the size parameters of the rebound detection system and the rebound instrument.
S7, executing S5-S6 circularly until the rebound value data acquisition of all the measuring points in the current area is completed.
In this embodiment, 16 positioning points are set in total, that is, 16 times of rebound value data are acquired, that is, the rebound value data acquisition of all the measuring points in the current area is completed.
S8, judging whether the current detection area for completing rebound data acquisition is the last detection area in the first preset track. And if so, ending the moving process along the first preset track. Otherwise, returning to the step S2 to control the resiliometer to continue to move to the next positioning point I until all the areas finish the rebound data acquisition. If the current area where the rebound data acquisition is completed is the 10 th area, then it is determined that the concrete member has completed rebound detection of all areas, and then a notification of completion of detection can be sent to a detection person or related personnel, so that the person can detach the rebound detection system from the concrete member or perform rebound detection of the next concrete member.
S9, calculating the average rebound value of each measuring area according to the rebound data of the measuring points of each measuring area.
In this embodiment, 16 measurement points are provided in each measurement area, and the average rebound value of the measurement area in each measurement area can be obtained by adopting an algorithm of clipping an average value, that is, 3 maximum rebound values and 3 minimum rebound values are removed, and then the remaining 10 rebound value data are used for calculating the average rebound value of the measurement area. The calculation formula of the average rebound value of the measuring area is as follows:
in the method, in the process of the invention,R m mean rebound values for the zones are indicated.R i Represent the firstiThe rebound values of the measuring points,i=1，2，…10。
in other embodiments, the strength of the concrete member may also be calculated according to the concrete strength calculation formula provided in the industry standard described above. Specifically, the concrete strength of each zone of the member requires the average rebound value obtained aboveR m And average carbonization depth value of each regiond m ) And calculated from the annex table in the procedure, and will not be described in detail herein.
Example 2
The present embodiment provides a computer terminal comprising a memory, a processor, and a computer program stored on the memory and executable on the processor.
The computer terminal may be a smart phone, a tablet computer, a notebook computer, etc. capable of executing a program. The processor may be a central processing unit (Central Processing Unit, CPU), controller, microcontroller, microprocessor, or other data processing chip in some embodiments. The processor is typically used to control the overall operation of the computer device. In this embodiment, the processor is configured to execute the program code stored in the memory or process the data. The steps of the defect identification method in the concrete springback detection process in embodiment 1 can be realized when the processor executes the program.
Example 3
The present embodiment provides a computer-readable storage medium having stored thereon a computer program which, when executed by a processor, implements the steps of the defect identification method in the concrete rebound detection process in embodiment 1.
The computer readable storage medium may include flash memory, hard disk, multimedia card, card memory (e.g., SD or DX memory, etc.), random Access Memory (RAM), static Random Access Memory (SRAM), read Only Memory (ROM), electrically Erasable Programmable Read Only Memory (EEPROM), programmable Read Only Memory (PROM), magnetic memory, magnetic disk, optical disk, etc. In some embodiments, the storage medium may be an internal storage unit of a computer device, such as a hard disk or memory of the computer device. In other embodiments, the storage medium may also be an external storage device of the computer device, such as a plug-in hard disk, a Smart Media Card (SMC), a Secure Digital (SD) Card, a Flash memory Card (Flash Card) or the like. Of course, the storage medium may also include both internal storage units of the computer device and external storage devices. In this embodiment, the memory is typically used to store an operating system and various application software installed on the computer device. In addition, the memory can be used to temporarily store various types of data that have been output or are to be output.
Example 4
Referring to fig. 12 and 13, the present embodiment provides a wall surface detection robot, which can use the rebound detection method of the concrete member in embodiment 1 to drive a rebound gauge to move along a wall surface to be detected of a standard concrete member, and keep the axis of the rebound gauge perpendicular to the wall surface to be detected for rebound detection.
The wall inspection robot includes a wall climbing device 2, a driving device 5, a carrying platform 7 and a controller, and may further include a clamping assembly for mounting the resiliometer 3 on the movable end of the driving device 5, and a power supply for supplying power to the robot.
The stage 7 may be formed in a rectangular parallelepiped shape, and a controller may be provided therein. The stage 7 is kept parallel to the wall surface 1 to be measured, i.e., parallel to the Y-Z plane in embodiment 1, during rebound detection.
The wall climbing device 2 comprises a mechanical arm 21 and a vacuum chuck 22. The robot arm 21 may be provided with four sets, the number of vacuum chucks 22 corresponding thereto. In other embodiments, the robotic arms 21 may be arranged in other numbers, such as six or eight groups. The mechanical arm 21 can adopt a multi-degree-of-freedom mechanical arm, and each joint of the mechanical arm 21 can be driven by a servo motor. One end of each group of mechanical arms 21 is rotatably arranged near the corner of the carrying platform 7, and the other end of each group of mechanical arms can be fixedly connected with a corresponding vacuum chuck 22.
Wherein, the carrier 7 can be driven to move along the surface of the wall 1 to be measured by the cooperative movement of the plurality of groups of mechanical arms 21. Wherein, during movement, a part of the mechanical arms 21 are in adsorption action with the wall surface 1 to be detected, and the other part of the mechanical arms 21 change the gesture, so that new adsorption points are selected on the wall surface 1 to be detected, and the cooperative movement of a plurality of groups of mechanical arms 21 is realized in sequence. In addition, when each group of mechanical arms 21 stays on the wall surface 1 to be measured, the corresponding vacuum chuck 22 is in an adsorption state, and when the posture of the mechanical arms 21 needs to be changed, the vacuum chuck 22 removes adsorption.
In addition, each vacuum cup 22 may be connected by its own nipple to a vacuum device (e.g., a vacuum generator) that may be removably placed beside the concrete element to adjust the suction status of the respective vacuum cup 22.
The driving device 5 comprises three sets of linear motors: an X-axis linear motor 51, a Y-axis linear motor 52, and a Z-axis linear motor 53. The Z-axis linear motor 53 may be provided with two sets parallel to each other, which are fixedly mounted on the side of the stage 7 near the wall surface 1 to be measured, and the extending direction of the Z-axis linear motor 53 is parallel to the Z-axis. The Y-axis linear motor 52 is provided with a group, and both ends thereof are fixedly mounted with the moving ends of the two groups of Z-axis linear motors 53, respectively. The X-axis linear motor 51 is also provided with a set of which the bottom can be fixedly mounted on the movable end of the Y-axis linear motor 52. So far, the resiliometer 3 can be fixedly installed on the movable end of the X-axis linear motor 51 through the clamping assembly, so that the driving device 5 can drive the resiliometer 3 to move along the X, Y, Z axial direction in a certain movable range near the carrying platform 7, and further, when the resiliometer 3 reaches the vicinity of a certain area 4, all positioning points two corresponding to all measuring points in the area respectively reach along the Y-Z plane, and when the resiliometer 3 reaches a certain positioning point two, the resiliometer reaches corresponding measuring points along the X-axis direction and completes one rebound detection.
The clamping assembly may include two arcuate clamping plates secured by bolts, one of which is secured to the movable end of the X-axis linear motor 51.
The controller and the power supply can be arranged in the inner cavity of the carrying platform 7, and the controller is electrically connected with each linear motor of the driving device 5, each servo motor of the wall climbing device 2 and the vacuum equipment, so that the running state of each component is controlled. The controller may store a related control program, and the processing unit thereof may implement the steps of the rebound detection method in embodiment 1 when executing the control program. In addition, the power supply may power the various motors on the robot. The power can adopt the battery, and the electric quantity of storage can satisfy at least one and await measuring the complete rebound detection of wall 1 can.
The technical features of the above-described embodiments may be arbitrarily combined, and all possible combinations of the technical features in the above-described embodiments are not described for brevity of description, however, as long as there is no contradiction between the combinations of the technical features, they should be considered as the scope of the description.
The above examples merely represent a few embodiments of the present invention, which are described in more detail and are not to be construed as limiting the scope of the invention. It should be noted that it will be apparent to those skilled in the art that several variations and modifications can be made without departing from the spirit of the invention, which are all within the scope of the invention. Accordingly, the scope of the invention should be assessed as that of the appended claims.
Claims (7)
1. The defect coping method in the concrete rebound detection process is characterized by being used for deploying a qualified standby area for a rebound instrument when the rebound instrument is positioned to one of areas on a wall surface to be detected of a concrete member and a defect recognition method is applied to recognize that the wall surface of the area is a defect wall surface; the defect identification method comprises the following steps:
acquiring wall images of a currently positioned area in real time;
inputting the wall image into a pre-trained wall defect recognition model, recognizing defect positions and defect characteristic categories in the wall image, and obtaining pixel duty ratios of all defects in the image according to the defect positions; wherein the defect characteristic categories comprise cracks, pits, honeycombs, pre-buried pipe orifices, floating slurry, oil dirt and coatings;
performing weighted aggregation on the defect feature class number and the defect pixel ratio in the wall image to obtain a defect weighted result for reflecting the wall defect degree;
analyzing the relation between the defect weighted result and a preset judging threshold value; when the defect weighting result is larger than the preset judging threshold value, judging that the wall surface of the currently positioned area is a defect wall surface; otherwise, judging the wall surface of the current positioning area as a qualified wall surface;
the defect coping method includes the steps of:
taking the currently positioned area as a reference area, acquiring the integral coordinates of the wall surface to be measured and the local coordinates of the reference area on the wall surface to be measured according to a component parameter database, and further dividing a deployable area corresponding to the reference area on the wall surface to be measured;
setting the deployment interval of adjacent areas and the edge interval between the areas and the end parts of the members, so as to plan at least one alternative area in the deployable area;
sequentially positioning the resiliometer to each alternative area, and simultaneously sequentially marking the wall surface of each alternative area as a defect wall surface or a qualified wall surface according to the defect identification method;
counting the number of all candidate areas marked as qualified wall surfaces in the deployable area, and making the following decision based on the number of candidate areas:
when the qualification number of the alternative areas is more than or equal to 1, the alternative area closest to the reference area is used as a standby area;
and when the qualification number of the alternative areas is less than 1, counting the defect weighting results of the reference area and all the alternative areas, selecting the area with the smallest defect weighting result from the defect weighting results as the alternative area, marking the alternative area as abnormal, and sending the abnormal alternative area to an interaction end.
2. The method for coping with defects in concrete rebound detection according to claim 1, wherein a deployment pitch of adjacent areas is not higher than 2m; the edge spacing between the region and the end of the member is no more than 0.5m and no less than 0.2m.
3. The method for coping with defects in concrete springback detection according to claim 1, wherein the collected wall image is further subjected to gray-scale processing and median filtering processing before the wall image is input into the wall defect recognition model, thereby obtaining a wall image after preprocessing reinforcement.
4. The method for coping with defects in concrete rebound detection according to claim 1, wherein the training method of the wall defect recognition model comprises the steps of:
obtaining a plurality of wall defect image samples to form an original data set for model training; wherein, each wall defect image sample is marked with a defect feature class true value and a defect position true value;
initializing a constructed model to be trained, and setting a loss function and training parameters;
performing defect identification on the wall defect image sample by using the model to be trained to obtain a defect characteristic type detection value and a defect position detection value in the wall defect image sample;
calculating a first recognition loss between the defect feature class true value and a defect feature class detection value, and a second recognition loss between the defect location true value and the defect location detection value, based on the loss function;
and adjusting the neural network parameters of the model to be trained according to the first recognition loss and the second recognition loss, and further obtaining the wall defect recognition model after iterative training is completed.
5. The method for coping with defects in concrete rebound detection according to claim 1, wherein an average light intensity value in a predetermined area around each of the areas is acquired in real time before acquiring a wall image of the each of the areas; and when the average light intensity value is lower than a preset light intensity threshold value, supplementing light to the wall surface of the current area through a light source.
6. The rebound detection method of the concrete member is characterized by comprising the steps of controlling a rebound gauge to move along a wall surface to be detected of a standard concrete member, and keeping the axis of the rebound gauge perpendicular to the wall surface to be detected so as to carry out rebound detection; defining the thickness, length and height directions of the wall surface to be measured as an X axis, a Y axis and a Z axis respectively; the rebound detection method comprises the following steps:
s1, respectively acquiring a preset track I and a preset track II which are matched with a wall surface to be detected according to a component parameter database;
the first preset track is sequentially connected with first positioning points corresponding to all the areas of the wall surface to be detected in series; the second preset track is sequentially connected with second positioning points corresponding to all the measuring points in each measuring area in series; the planes of the first preset track and the second preset track are parallel to the Y-Z plane;
s2, after the resiliometer is positioned to an initial point on the first preset track, controlling the resiliometer to move to one positioning point along the first preset track to complete positioning of a measuring area;
s3, judging whether the wall surface of the currently positioned area is a defect wall surface by adopting the defect identification method as set forth in any one of claims 1, 3, 4 and 5;
s4, when the wall surface of the currently positioned area is a defect wall surface, deploying a qualified standby area for the resiliometer by adopting the defect coping method according to any one of claims 1 to 5, and controlling the resiliometer to be positioned to the standby area;
s5, if the wall surface of the currently positioned area is a qualified wall surface or the resiliometer is deployed to a qualified standby area, controlling the resiliometer to move to one of the positioning points II along the preset track II so as to finish positioning of the measuring point;
s6, controlling the resiliometer to approach the wall surface to be measured from an initial axial position along the X-axis direction until the rebound test of the current measuring point is completed and the rebound value data are acquired, so that the resiliometer is restored to the initial axial position;
s7, circularly executing S5-S6 until the rebound value data acquisition of all the measuring points in the current area is completed;
s8, judging whether a current detection zone for completing rebound data acquisition is the last detection zone in the first preset track; if yes, ending the moving process along the first preset track; otherwise, returning to the step S2 to control the resiliometer to continue to move to the next positioning point I until the rebound data acquisition is completed for all the preset number of areas of the wall surface to be detected;
s9, calculating the average rebound value of each measuring area according to the rebound data of the measuring points of each measuring area.
7. The method according to claim 6, wherein in S4, the standby area and all areas corresponding to the first predetermined track are independent of each other on the wall surface to be measured.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202310044327.8A CN115808425B (en) | 2023-01-30 | 2023-01-30 | Defect identification and application method in concrete member rebound detection process |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202310044327.8A CN115808425B (en) | 2023-01-30 | 2023-01-30 | Defect identification and application method in concrete member rebound detection process |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN115808425A CN115808425A (en) | 2023-03-17 |
| CN115808425B true CN115808425B (en) | 2023-05-05 |
Family
ID=85487626
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202310044327.8A Active CN115808425B (en) | 2023-01-30 | 2023-01-30 | Defect identification and application method in concrete member rebound detection process |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN115808425B (en) |
Families Citing this family (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN116539463B (en) * | 2023-06-20 | 2023-10-13 | 广东惠和工程检测有限公司 | Concrete strength detection method |
Family Cites Families (10)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CH679953A5 (en) * | 1989-12-07 | 1992-05-15 | Proceq Sa | |
| EP1679710B1 (en) * | 2005-01-10 | 2013-08-14 | Nero Ag | Method for writing data blocks on a block addressable storage medium using defect management |
| EP2461324A1 (en) * | 2007-02-14 | 2012-06-06 | Panasonic Corporation | Optical disc device |
| CN104483330B (en) * | 2014-11-11 | 2017-03-29 | 浙江大学 | The dynamic assessment method of concrete surface crack real-time monitoring system and cracking risk |
| CN106872472A (en) * | 2017-01-17 | 2017-06-20 | 中交武汉港湾工程设计研究院有限公司 | Surface Quality of Concrete method of determination and evaluation |
| CN110672822A (en) * | 2019-09-23 | 2020-01-10 | 郑州航空工业管理学院 | Nondestructive testing system for structural defects of bridge concrete |
| CN111999195A (en) * | 2020-08-27 | 2020-11-27 | 福建省永富建设集团有限公司 | Nondestructive testing system for structural defects of concrete |
| CN112147228B (en) * | 2020-09-28 | 2023-03-10 | 廊坊市阳光建设工程质量检测有限公司 | Method for establishing strength measurement curve for detecting concrete strength by using rebound ultrasonic angle measurement comprehensive method |
| CN217766224U (en) * | 2022-05-28 | 2022-11-08 | 武汉科技大学 | Assembled reinforced concrete beam column concatenation steel fast overhaul device for node |
| CN115619778A (en) * | 2022-12-06 | 2023-01-17 | 南京迈能能源科技有限公司 | Power equipment defect identification method and system, readable storage medium and equipment |
-
2023
- 2023-01-30 CN CN202310044327.8A patent/CN115808425B/en active Active
Also Published As
| Publication number | Publication date |
|---|---|
| CN115808425A (en) | 2023-03-17 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN115808425B (en) | Defect identification and application method in concrete member rebound detection process | |
| CN109840900B (en) | Fault online detection system and detection method applied to intelligent manufacturing workshop | |
| CN110825088B (en) | Multi-view vision guiding ship body cleaning robot system and cleaning method | |
| CN113134683A (en) | Laser marking method and device based on machine learning | |
| CN113780111B (en) | Pipeline connector defect accurate identification method based on optimized YOLOv3 algorithm | |
| CN106813570B (en) | The identification of elongated cylindrical object dimensional and localization method based on line-structured light scanning | |
| Li et al. | Development of a robot system for applying putty on plastered walls | |
| CN108280820A (en) | A kind of automatic repair system of nuclear island factory building clear-water concrete metope and its implementation | |
| CN113280209B (en) | System for detecting pipeline excess, use method of system and detection method | |
| CN117790380A (en) | Accurate control operation method for wafer cleaning mechanical arm | |
| CN116362548B (en) | Standard knot availability management method, system, device and storage medium | |
| CN108789426A (en) | The mechanical people of surface treatment | |
| CN116765569A (en) | Robot post nail roller surfacing path planning method based on point cloud | |
| CN117218511A (en) | Visual identification method for crack of magnetic powder inspection and magnetic powder inspection method for crack | |
| CN109342439B (en) | Unmanned aerial vehicle-based cable structure appearance detection method | |
| CN116597329A (en) | Bridge crack detection system and method based on machine vision | |
| Wang et al. | Robot floor‐tiling control method based on finite‐state machine and visual measurement in limited FOV | |
| CN110146204A (en) | A kind of intelligent residual stress x-ray measuring device and its measurement method | |
| CN109764822A (en) | Large-sized gantry crane beam deformation measurement method based on image and geometric profile | |
| CN114408131A (en) | Bionic robot fish for daily inspection of underwater part of bridge | |
| Aoyama et al. | Autonomous measurement of physical model shape for reverse engineering | |
| CN112330670A (en) | Tunnel drilling robot drilling method and device, drilling robot and medium | |
| Hu et al. | The next-best-view for workpiece localization in robot workspace | |
| CN116572264B (en) | Soft mechanical arm free eye system target tracking method based on light weight model | |
| CN116713514B (en) | Accurate positioning system and method for machining holes of cast parts based on machine vision |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant |