Internal defect detection system of prestressed concrete tower based on machine learning algorithm

Through a system based on machine learning algorithm, combined with ring body and ultrasonic probe, the internal defect detection and visual display of in-service concrete towers is realized, solving the problems of inefficiency and inaccurate information in the existing technology, and providing fast and efficient On-site inspection capability.

CN118777424BActive Publication Date: 2024-11-19HUBEI BRANCH OF CHINA TOWER CO LTD
View PDF 0 Cites 0 Cited by

Patent Information

Application Number
CN202411274791.7
Authority / Receiving Office
CN · China
Patent Type
Patents(China)
Current Assignee / Owner
Filing Date
2024-09-12
Publication Date
2024-11-19
Estimated Expiration
2044-09-12

AI Technical Summary

Technical Problem

The internal flaw detection scheme of existing concrete towers is inefficient, inaccurate information, and cannot conduct defect detection and visual display on site, especially in the maintenance process of in-service concrete towers.

Method used

Using a system based on machine learning algorithm, combining ring bodies, ultrasonic emission probes, multiple ultrasonic receiving probes, Imaging equipment and display devices, the pre-trained concrete crack identification model is imported to generate grayscale maps by extracting ultrasonic signal characteristic data. Conduct defect detection and visual display to realize automated and on-site inspection.

Benefits of technology

The internal defect detection results of concrete towers can be quickly and efficiently obtained without manual climbing, realizing on-site visual display, and improving detection efficiency and accuracy.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN118777424B_ABST
    Figure CN118777424B_ABST
Patent Text Reader

Abstract

The present application discloses a prestressed concrete tower internal defect detection system based on machine learning algorithm, which relates to the technical field of non-destructive flaw detection of concrete towers. The system includes a ring body, an ultrasonic transmitting probe, a plurality of ultrasonic receiving probes, an industrial intelligent machine device and a display device. The industrial intelligent machine device is used to first extract the ultrasonic signal characteristic data of each receiving probe, and then for each position in the cross section of the currently tested tower, the corresponding polar coordinate data and the polar coordinate data and characteristic data of each receiving probe are imported into a concrete crack recognition model based on a machine learning algorithm and pre-trained, and the confidence of cracks appearing at the corresponding position is output, and finally a grayscale image of the tower cross section is drawn based on the confidence to be transmitted to the display device as an internal defect detection result for visual display, so that the internal defect detection result can be obtained quickly, efficiently and accurately on site and displayed on site visually.
Need to check novelty before this filing date? Find Prior Art

Description

Technical Field

[0001] The present invention belongs to the technical field of nondestructive flaw detection of concrete pole towers, and in particular relates to a system for detecting internal defects of prestressed concrete pole towers based on a machine learning algorithm. Background Art

[0002] In recent years, prestressed reinforced concrete poles and towers have been widely used in the power industry due to their high strength and excellent durability, and have received widespread attention and recognition. Compared with traditional steel towers, prestressed reinforced concrete poles and towers have a series of advantages such as strong mounting capacity, low comprehensive cost and short production and construction period.

[0003] Concrete cracks are the damage caused by physical structural changes of concrete structure due to internal and external factors. They are also the main reason for the reduction of bearing capacity, durability and waterproofness of concrete structures. The number, type and size of concrete cracks are important indicators for stability analysis of concrete buildings.

[0004] At present, most of the existing concrete internal flaw detection solutions are handheld structures, and in-service concrete towers (i.e. prestressed reinforced concrete towers in service) are erected at the erection position, so maintenance personnel need to climb the poles and hold the instrument to obtain the internal information of the concrete towers, and the internal information of the concrete towers needs to be manually imported into the background server before the internal defect detection results of the concrete towers can be obtained through data analysis. This method will inevitably lead to various defects such as low efficiency, inaccurate information, inability to carry out work in special areas, and inability to obtain internal defect detection results on site and perform on-site visualization. Therefore, how to provide a new concrete internal flaw detection solution that can be applied to in-service concrete towers, so as to quickly, efficiently and accurately obtain internal defect detection results on site and perform on-site visualization, is a topic that technicians in this field urgently need to study. Summary of the invention

[0005] The purpose of the present invention is to provide a prestressed concrete tower internal defect detection system based on a machine learning algorithm to solve the problems of low efficiency, inaccurate information, inability to work in special areas, and inability to obtain internal defect detection results on site and perform on-site visual display when the existing concrete internal flaw detection scheme is used for in-service concrete towers.

[0006] In order to achieve the above object, the present invention adopts the following technical solutions:

[0007] In a first aspect, a prestressed concrete tower internal defect detection system based on a machine learning algorithm is provided, comprising an annular body, an ultrasonic transmitting probe, a plurality of ultrasonic receiving probes, an artificial intelligence device, and a display device, wherein the annular body is used to annularly surround an in-service concrete tower and vertically move up and down on the in-service concrete tower, the ultrasonic transmitting probe and the plurality of ultrasonic receiving probes are annularly spaced and arranged on the inner side of the annular body, and the artificial intelligence device is arranged on the outer side of the annular body;

[0008] The ultrasonic transmitting probe is communicatively connected to the IIoT device and is used to transmit ultrasonic signals to the in-service concrete tower under the control of the IIoT device;

[0009] The ultrasonic receiving probe is communicatively connected to the IIoT device, and is used to receive the ultrasonic signal reflected or transmitted by the in-service concrete tower, and transmit the received ultrasonic signal to the IIoT device in real time;

[0010] The IIoT device is communicatively connected to the display device and is used to process the ultrasonic signals from the multiple ultrasonic receiving probes in the following manner:

[0011] For each ultrasonic receiving probe among the plurality of ultrasonic receiving probes, performing feature data extraction processing on an ultrasonic signal from a corresponding probe to obtain corresponding ultrasonic signal feature data;

[0012] Determine the current measured height position of the in-service concrete tower according to the current climbing pole height of the annular body and the known arrangement heights of the ultrasonic transmitting probe and the plurality of ultrasonic receiving probes respectively relative to the annular body;

[0013] Determine, according to the known cross-section type of the in-service concrete tower, a current measured tower cross-section of the in-service concrete tower corresponding to the current measured height position, wherein the center point of the current measured tower cross-section is the annular center point of the annular body;

[0014] For each position in the cross section of the currently measured tower, the corresponding polar coordinate data and the polar coordinate data and ultrasonic signal characteristic data of each ultrasonic receiving probe are imported into a concrete crack recognition model based on a machine learning algorithm and has been pre-trained, and the confidence of the concrete crack occurring at the corresponding position is output, and the corresponding grayscale value negatively correlated with the confidence is calculated, wherein the polar coordinate data refers to the coordinate data in a polar coordinate system, and the polar coordinate system takes the center point of the tower cross section as the pole and the opposite direction of the ultrasonic emission direction as the polar axis direction;

[0015] According to the grayscale values ​​of the respective positions, a grayscale image of the cross section of the currently measured tower is obtained;

[0016] The current measured height position and the grayscale image of the current measured tower cross section are transmitted to the display device as the current internal defect detection result of the in-service concrete tower for visual display.

[0017] Based on the above invention content, a new solution for concrete internal flaw detection applicable to in-service concrete towers is provided, namely, it includes a ring body, an ultrasonic transmitting probe, multiple ultrasonic receiving probes, an artificial intelligence machine device and a display device, wherein the artificial intelligence machine device is used to first extract ultrasonic signal characteristic data of each receiving probe, and then, for each position in the cross section of the current measured tower, the corresponding polar coordinate data and the polar coordinate data and ultrasonic signal characteristic data of each receiving probe are imported into a concrete crack recognition model based on a machine learning algorithm and pre-trained, and the confidence of the occurrence of concrete cracks at the corresponding position is output, and finally, a grayscale image of the cross section of the current measured tower is drawn based on the confidence to be transmitted to the display device as the current internal defect detection result for visual display, so that there is no need to manually climb the pole to obtain the internal information of the concrete tower, and the internal defect detection results can be quickly, efficiently and accurately obtained on site and displayed on site visually, which is convenient for practical application and promotion.

[0018] In a possible design, a climbing pole driving mechanism is also included for driving the annular body to move vertically up and down on the in-service concrete pole tower, wherein a controlled end of the climbing pole driving mechanism is communicatively connected to the artificial intelligence machine device.

[0019] In a possible design, the climbing pole driving mechanism includes a vertical pole, an inclined pole, a driving wheel, an inclined spring and a motor, wherein the controlled end of the motor is used as the controlled end of the climbing pole driving mechanism;

[0020] The first end of the vertical rod is fixedly connected to the annular body, the first end of the oblique rod is movably hinged to the second end of the vertical rod, and the driving wheel is arranged at the second end of the oblique rod;

[0021] The inclined tension spring is always in a tensioned state, and the two ends of the inclined tension spring are respectively connected to the first end of the vertical rod and the second end of the inclined rod in a one-to-one correspondence, so as to ensure that the driving wheel always contacts the outer surface of the in-service concrete tower;

[0022] The motor is mounted on the inclined rod, and the output shaft of the motor is drivingly connected to the driving wheel.

[0023] In a possible design, a plurality of transverse connectors corresponding to the ultrasonic transmitting probe and the plurality of ultrasonic receiving probes are arranged annularly and spaced apart on the inner side of the annular body, wherein one end of the transverse connector is fixedly connected to the inner side of the annular body, and the other end of the transverse connector is fixedly connected to the vertical plate;

[0024] At least three universal wheels are arranged on the near-tower surface of the vertical plate, and the ultrasonic transmitting probe or the ultrasonic receiving probe corresponding to the vertical plate and the transverse connecting member is embedded in the near-tower surface, wherein the near-tower surface refers to a side surface close to the in-service concrete tower;

[0025] The transverse connecting member has automatic retractability to ensure that the corresponding at least three universal wheels always contact the outer surface of the in-service concrete tower.

[0026] In a possible design, the transverse connecting member includes a transverse rod, a transverse sleeve and an elastic member;

[0027] One end of the transverse rod is fixedly connected to the vertical plate, and the other end of the transverse rod extends into the interior of the transverse sleeve and is provided with an annular ear plate extending outward;

[0028] One end of the transverse sleeve is fixedly connected to the inner side of the annular body, and an annular baffle matching with the annular ear plate is arranged in the pipe opening of the other end of the transverse sleeve, and the inner cavity of the annular baffle is slidably matched with the transverse rod, and the inner cavity of the transverse sleeve is slidably matched with the annular ear plate;

[0029] The elastic member is located in the inner cavity of the transverse sleeve and is used to push or pull the transverse rod to move in a direction away from the annular body so as to make the transverse connecting member have automatic retractability.

[0030] In a possible design, the elastic member is a spring that is always in a tension state, wherein the spring is located between the annular ear plate and the annular baffle;

[0031] The spring is sleeved on the outer circumference of the transverse rod, and the two ends of the spring are fixedly connected to the annular ear plate and the annular baffle plate in a one-to-one correspondence, so as to pull the transverse rod to move in a direction away from the annular body.

[0032] In a possible design, the transverse rod and the annular body are both hollow structures, and the communication cable of the ultrasonic transmitting probe or the ultrasonic receiving probe passes through the corresponding transverse rod, the corresponding transverse sleeve and the annular body in sequence to communicate with the artificial intelligence device.

[0033] In a possible design, an extension plate is arranged on the far tower surface of the vertical plate, wherein the far tower surface refers to the other side surface away from the in-service concrete tower;

[0034] The extension plate is provided with a plug hole for inserting one end of the L-shaped rigid member, and the other end of the L-shaped rigid member is movably hinged to the annular body so as to lock the automatic retractability of the transverse connector through the plug-in cooperation between the L-shaped rigid member and the extension plate.

[0035] In a possible design, the far tower surfaces of any two circumferentially adjacent vertical plates are respectively connected to two ends of an elastic bandage in a one-to-one correspondence, wherein the far tower surface refers to the other side surface away from the in-service concrete tower.

[0036] In a possible design, the annular body is formed by joining two halves of a hoop body.

[0037] Beneficial effects of the above scheme:

[0038] (1) The present invention creatively provides a new solution for concrete internal flaw detection applicable to in-service concrete towers, namely, comprising a ring body, an ultrasonic transmitting probe, a plurality of ultrasonic receiving probes, an artificial intelligence device and a display device, wherein the artificial intelligence device is used to first extract ultrasonic signal characteristic data of each receiving probe, and then, for each position in the cross section of the currently measured tower, import the corresponding polar coordinate data and the polar coordinate data and ultrasonic signal characteristic data of each receiving probe into a concrete crack recognition model based on a machine learning algorithm and pre-trained, and output the confidence level of the concrete cracks appearing at the corresponding position. Finally, a grayscale image of the cross section of the currently measured tower is drawn based on the confidence level and transmitted to the display device as the current internal defect detection result for visual display. In this way, there is no need to manually climb the pole to obtain the internal information of the concrete tower, and the internal defect detection results can be quickly, efficiently and accurately obtained on site and displayed on site visually, which is convenient for practical application and promotion. BRIEF DESCRIPTION OF THE DRAWINGS

[0039] In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the drawings required for use in the embodiments or the description of the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying creative work.

[0040] Figure 1 A schematic diagram of the structure of a prestressed concrete tower internal defect detection system provided in an embodiment of the present application.

[0041] Figure 2A schematic diagram of the top-down position relationship of the annular body, probe, and AI device provided in an embodiment of the present application when used with an in-service concrete tower.

[0042] Figure 3 This is a schematic diagram of the top-down position relationship between the annular body, probe, and AI device provided in an embodiment of the present application and an in-service concrete tower before application.

[0043] Figure 4 Schematic diagram of the combined structure of the annular body, transverse connecting member and vertical plate provided in an embodiment of the present application.

[0044] Figure 5 A schematic diagram of the arrangement of the vertical panels near the tower surface provided in an embodiment of the present application.

[0045] Figure 6 This is a schematic diagram of the combined structure of two circumferentially adjacent vertical plates and an elastic bandage provided in an embodiment of the present application.

[0046] Figure 7 A schematic diagram of the side position relationship of the annular body, the probe and the climbing pole drive mechanism provided in the embodiment of the present application when used with an in-service concrete tower. DETAILED DESCRIPTION

[0047] In order to more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the present invention will be briefly introduced below in combination with the drawings and the description of the embodiments or the prior art. Obviously, the following description of the structures of the drawings is only some embodiments of the present invention. For ordinary technicians in this field, other embodiments can be obtained based on these embodiments without creative work. It should be noted that the description of these embodiments is used to help understand the present invention, but does not constitute a limitation of the present invention.

[0048] It should be understood that although the terms first and second, etc. may be used herein to describe various objects, these objects should not be limited by these terms. These terms are only used to distinguish one object from another object. For example, a first object can be referred to as a second object, and similarly, a second object can be referred to as a first object without departing from the scope of the exemplary embodiments of the present invention.

[0049] It should be understood that the term "and / or" that may appear in this document is merely a description of the association relationship between associated objects, indicating that there may be three relationships. For example, A and / or B can indicate three situations: A exists alone, B exists alone, or A and B exist at the same time. For another example, A, B and / or C can indicate the existence of any one of A, B and C or any combination of them. The term " / and" that may appear in this document describes another type of association object relationship, indicating that there may be two relationships. For example, A / and B can indicate two situations: A exists alone or A and B exist at the same time. In addition, the character " / " that may appear in this document generally indicates that the previous and next associated objects are in an "or" relationship.

[0050] Example

[0051] like Figures 1 to 7 As shown, the prestressed concrete tower internal defect detection system provided in this embodiment and based on the machine learning algorithm includes but is not limited to a ring body 1, an ultrasonic transmitting probe 2, a plurality of ultrasonic receiving probes 3, an artificial intelligence device 4 and a display device, wherein the ring body 1 is used to surround the in-service concrete tower 100 in an annular manner and move vertically up and down on the in-service concrete tower 100, the ultrasonic transmitting probe 2 and the plurality of ultrasonic receiving probes 3 are arranged in an annular manner at intervals on the inner side of the ring body 1, and the artificial intelligence device 4 is arranged on the outer side of the ring body 1. Figure 2 As shown, one ultrasonic transmitting probe 2 and seventeen ultrasonic receiving probes 3 (i.e., 18 probes in total) are arranged at intervals in the inner side of the annular body 1. In order to ensure that the in-service concrete tower 100 can be surrounded and released conveniently, as shown in FIG. Figure 3 As shown, the annular body 1 is preferably formed by joining two halves of the hoop body 10. In addition, the in-service concrete pole tower 100 refers to a concrete pole tower erected at an erection position, and its cross-sectional shape may be, but is not limited to, a circular cross-sectional shape, an annular cross-sectional shape or a square cross-sectional shape (the drawings of this embodiment show a circular cross-sectional shape as an example).

[0052] The ultrasonic transmitting probe 2 is communicatively connected to the IIoT device 4, and is used to transmit ultrasonic signals to the in-service concrete pole tower 100 under the control of the IIoT device 4; the ultrasonic receiving probe 3 is communicatively connected to the IIoT device 4, and is used to receive ultrasonic signals reflected or transmitted from the in-service concrete pole tower 100, and transmit the received ultrasonic signals to the IIoT device 4 in real time. The aforementioned ultrasonic transmitting probe 2 and ultrasonic receiving probe 3 can both be implemented using existing ultrasonic transceiver probes.

[0053] The AI ​​device 4 is communicatively connected to the display device, and is used to process the ultrasonic signals from the multiple ultrasonic receiving probes 3 according to the following steps S1 to S6.

[0054] S1. For each ultrasonic receiving probe 3 among the plurality of ultrasonic receiving probes 3, feature data extraction processing is performed on the ultrasonic signal from the corresponding probe to obtain corresponding ultrasonic signal feature data.

[0055] In the step S1, it is considered that when ultrasonic flaw detection of concrete is performed, the propagation of the ultrasonic signal can be regarded as a combined propagation in the concrete body, the light crack (defect) area and the wide crack (defect) (the composition ratio of each part can be regarded as determined by the ratio of the volume of each part to the total volume), and in good quality concrete, cracks (defects) are almost non-existent, and when cracks (defects) and other defects appear inside the concrete, the proportion of the concrete body to the total volume decreases sharply, and the cracks (defects) and other defects that appear are often filled with air or rainwater, resulting in different media in the propagation channel of the ultrasonic signal, a large amount of energy loss, and the ultrasonic propagation speed of the damaged material shows a sharp downward trend. Therefore, the ultrasonic wave velocity can be conventionally extracted and used as one of the characteristic data of the ultrasonic signal. It is also considered that cracks in concrete will cause reflection or refraction of ultrasonic waves during propagation, thereby causing changes in the amplitude and time of the signal, and the peak amplitude of the ultrasonic wave can be used to determine whether there are cracks in the concrete: if there are cracks (defects) in the concrete, the ultrasonic wave will be reflected or refracted when passing through the crack, causing the amplitude of the signal to decrease; if there is a waveform with a small amplitude in the ultrasonic signal, it may indicate that there are slight cracks (defects) in the concrete. Similarly, if the amplitude shows a sharp change, it indicates that there are wide cracks (defects) in the concrete. Therefore, the ultrasonic amplitude can also be conventionally extracted and used as one of the ultrasonic signal feature data. It is also considered that if there are cracks (defects) in the concrete, the propagation state of the ultrasonic wave will change significantly: generally speaking, the cracks are often filled with air or accumulated water. When the ultrasonic wave passes through such areas, it will slow down and be partially reflected, resulting in a longer time for the ultrasonic receiver to receive the complete ultrasonic signal. Therefore, the ultrasonic reflection time can also be conventionally extracted and used as one of the ultrasonic signal feature data. In addition to the ultrasonic wave velocity, the ultrasonic amplitude and the ultrasonic reflection time, other available features can also be used as the ultrasonic signal feature data, such as reflection waveform features. In addition, before performing the feature data extraction process, it is necessary to perform pre-processing such as denoising and / or filtering on the ultrasonic signal.

[0056] S2. Determine the current measured height position of the in-service concrete tower 100 according to the current climbing height of the ring body 1 and the known arrangement heights of the ultrasonic transmitting probe 2 and the plurality of ultrasonic receiving probes 3 respectively relative to the ring body 1.

[0057] In step S2, the current climbing pole height can be acquired in real time based on an existing conventional height measuring instrument (e.g., a height measuring sensor based on air pressure difference). Specifically, the current measured height position is equal to the current climbing pole height + the known arrangement height. In addition, the known arrangement height is generally equal to zero.

[0058] S3. According to the known cross-section type of the in-service concrete tower 100 , determine the current measured tower cross-section of the in-service concrete tower 100 corresponding to the current measured height position, wherein the center point of the current measured tower cross-section is the circumferential center point of the annular body 1 .

[0059] In step S3, the known section type can be obtained by input by the inspection personnel. When the in-service concrete pole tower 100 is a constant diameter pole, since the cross-sectional shape of the pole tower is always consistent from top to bottom, that is, the current measured pole tower section corresponding to the current measured height position is equal to the known pole tower section corresponding to the zero height position of the pole tower bottom (which can be input by the inspection personnel or obtained by on-site measurement); and when the in-service concrete pole tower 100 is a tapered pole, it is necessary to obtain the current measured pole tower section corresponding to the zero height position of the pole tower bottom, the known taper of the in-service concrete pole tower 100 (which can also be input by the inspection personnel) and the current measured height position, and finally obtain the current measured pole tower section corresponding to the current measured height position through conventional calculation.

[0060] S4. For each position in the current measured tower cross section, the corresponding polar coordinate data and the polar coordinate data and ultrasonic signal feature data of each ultrasonic receiving probe 3 are imported into a concrete crack recognition model based on a machine learning algorithm and has been pre-trained, and the confidence of the concrete crack occurring at the corresponding position is output, and the corresponding grayscale value negatively correlated with the confidence is calculated, wherein the polar coordinate data refers to the coordinate data in a polar coordinate system, and the polar coordinate system takes the center point of the current measured tower cross section as the pole, and the opposite direction of the ultrasonic emission direction as the polar axis direction.

[0061] In step S4, since the presence of concrete cracks at any position in the cross section of the currently measured tower will affect the ultrasonic signal characteristic data of each ultrasonic receiving probe 3, it is possible to reversely infer whether concrete cracks appear at any position based on the positional relationship between any position and each ultrasonic receiving probe 3 (i.e., reflected by the bipolar coordinate data) and the ultrasonic signal characteristic data of each ultrasonic receiving probe 3. The machine learning algorithm is a core artificial intelligence algorithm that specifically studies how computers simulate or implement human learning behavior to acquire new knowledge or skills and reorganize existing knowledge structures to continuously improve their own performance. It is the fundamental way to make computers intelligent; specifically, the machine learning algorithm preferably uses a linear regression algorithm based on the Python sklearn library to quickly and accurately find patterns in the data. Therefore, the concrete crack recognition model can be trained in advance based on a certain amount of positive sample data (i.e., the polar coordinate data of the position in the screenshot of the pole tower where concrete cracks are known to occur, as well as the polar coordinate data and ultrasonic signal characteristic data of each ultrasonic receiving probe 3) and negative sample data (i.e., the polar coordinate data of the position in the screenshot of the pole tower where concrete cracks are known to not occur, as well as the polar coordinate data and ultrasonic signal characteristic data of each ultrasonic receiving probe 3), through a conventional calibration verification modeling process (specifically including the calibration process and verification process of the model, that is, first by comparing the model simulation results with the measured data, and then adjusting the model parameters according to the comparison results, so that the simulation results are consistent with the actual process). Preferably, in the calibration verification modeling process of the concrete crack recognition model, the model parameters are tuned using a tree-structured Bayesian optimization algorithm. In addition, the machine learning algorithm can also be, but is not limited to, an artificial intelligence algorithm based on a support vector machine, a stochastic gradient descent method, a multivariate linear regression, a multilayer perceptron, a decision tree, a back propagation neural network, or a radial basis function network.

[0062] In step S4, the negative correlation between the grayscale value and the confidence level means that the higher the confidence level, the lower the grayscale value, and vice versa. Specifically, the grayscale value may be equal to 255-255×confidence level.

[0063] S5. According to the grayscale values ​​of the various positions, a grayscale image of the cross section of the currently measured tower is obtained.

[0064] In step S5, since the grayscale value is negatively correlated with the confidence level, the higher the confidence level, the lower the grayscale value, and the closer the corresponding pixel point is to white. Therefore, white can be used to directly present the area where concrete cracks may occur in the grayscale image of the current measured tower section, thereby obtaining an intuitive internal defect detection result of the in-service concrete tower.

[0065] S6. Transmitting the current measured height position and the grayscale image of the current measured tower cross section as the current internal defect detection result of the in-service concrete tower 100 to the display device for visual display.

[0066] In step S6, the display device may be specifically an electronic device held by the inspector, such as a mobile phone or a tablet computer. Since the current measured height position is bound to the grayscale image of the current measured tower section, during the inspection process, the three-dimensional inspection result of the internal defects of the in-service concrete tower 100 can be obtained by superimposing the grayscale images corresponding to each measured height position from bottom to top, and visualized on the display device, so that the inspector can quickly find the concrete cracks in the white area, and then determine whether the in-service concrete tower 100 needs to be scrapped.

[0067] Based on the detailed structure and working principle description of the internal defect detection system of the prestressed concrete tower, a new solution for concrete internal flaw detection applicable to in-service concrete towers is provided, which includes a ring body, an ultrasonic transmitting probe, multiple ultrasonic receiving probes, an artificial intelligence device and a display device, wherein the artificial intelligence device is used to first extract the ultrasonic signal characteristic data of each receiving probe, and then for each position in the cross section of the current measured tower, the corresponding polar coordinate data and the polar coordinate data and ultrasonic signal characteristic data of each receiving probe are imported into a concrete crack recognition model based on a machine learning algorithm and pre-trained, and the confidence of the concrete crack appearing at the corresponding position is output, and finally a grayscale image of the cross section of the current measured tower is drawn based on the confidence to be transmitted to the display device as the current internal defect detection result for visual display, so that there is no need to manually climb the pole to obtain the internal information of the concrete tower, and the internal defect detection results can be quickly, efficiently and accurately obtained on site and displayed on site visually, which is convenient for practical application and promotion.

[0068] Preferably, a pole climbing driving mechanism 6 is further included for driving the ring body 1 to move vertically up and down on the in-service concrete pole tower 100, wherein the controlled end of the pole climbing driving mechanism 6 is communicatively connected to the artificial intelligence device 4. In this way, the purpose of automatic pole climbing of the ring body 1 can be achieved through the pole climbing driving mechanism 6, which is further convenient and practical. Specifically, the climbing pole driving mechanism 6 includes but is not limited to a vertical rod 61, an inclined rod 62, a driving wheel 63, an inclined spring 64 and an electric motor, wherein the controlled end of the electric motor is used as the controlled end of the climbing pole driving mechanism 6; the first end of the vertical rod 61 is fixedly connected to the annular body 1, the first end of the inclined rod 62 is movably hinged to the second end of the vertical rod 61, and the driving wheel 63 is arranged at the second end of the inclined rod 62; the inclined spring 64 is always in a stretched state, and the two ends of the inclined spring 64 are respectively connected to the first end of the vertical rod 61 and the second end of the inclined rod 62 in a one-to-one correspondence, so as to ensure that the driving wheel 63 always contacts the outer surface of the in-service concrete pole tower 100; the motor is installed on the inclined rod 62, and the output shaft of the motor is transmission-connected to the driving wheel 63.

[0069] Preferably, a plurality of transverse connectors 11 corresponding to the ultrasonic transmitting probe 2 and the plurality of ultrasonic receiving probes 3 are arranged at intervals in annular direction on the inner side of the annular body 1, wherein one end of the transverse connector 11 is fixedly connected to the inner side of the annular body 1, and the other end of the transverse connector 11 is fixedly connected to the vertical plate 12; at least three universal wheels 121 are arranged on the near-tower surface of the vertical plate 12, and the ultrasonic transmitting probe 2 or the ultrasonic receiving probe 3 corresponding to the vertical plate 12 and the transverse connector 11 are embedded in the near-tower surface, wherein the near-tower surface refers to a side surface close to the in-service concrete tower 100; the transverse connector 11 has automatic retractability to ensure that the corresponding at least three universal wheels 121 always contact the outer surface of the in-service concrete tower 100. In this way, through the automatic retractability of the transverse connecting member 11 and the layout design on the near-tower surface, it can be ensured that the probe does not contact the outer surface of the in-service concrete tower 100 while approaching the outer surface of the in-service concrete tower 100, thereby ensuring that the received ultrasonic signal can accurately reflect the internal situation of the concrete tower while preventing the probe from being worn out during the process of climbing up and down the pole.

[0070] Further preferably, the transverse connecting member 11 includes but is not limited to a transverse rod 111, a transverse sleeve 112 and an elastic member 113; one end of the transverse rod 111 is fixedly connected to the vertical plate 12, the other end of the transverse rod 111 extends into the interior of the transverse sleeve 112 and is provided with an annular ear plate 114 extending outward; one end of the transverse sleeve 112 is fixedly connected to the inner side of the annular body 1, the other end of the transverse sleeve 112 is provided with an annular baffle 115 cooperating with the annular ear plate 114 in the pipe mouth, the internal cavity of the annular baffle 115 is slidably matched with the transverse rod 111, and the internal cavity of the transverse sleeve 112 is slidably matched with the annular ear plate 114; the elastic member 113 is located in the internal cavity of the transverse sleeve 112, and is used to push or pull the transverse rod 111 to move in a direction away from the annular body 1, so that the transverse connecting member 11 has automatic retractability.

[0071] Specifically, the elastic member 113 is a spring that is always in a stretched state, wherein the spring is located between the annular ear plate 114 and the annular baffle plate 115; the spring is sleeved on the outer circumference of the transverse rod 111, and the two ends of the spring are respectively fixedly connected to the annular ear plate 114 and the annular baffle plate 115 in a one-to-one correspondence, so as to pull the transverse rod 111 to move in a direction away from the annular body 1.

[0072] Specifically, the transverse rod 111 and the annular body 1 are both hollow structures, and the communication cable of the ultrasonic transmitting probe 2 or the ultrasonic receiving probe 3 passes through the corresponding transverse rod 111, the corresponding transverse sleeve 112 and the annular body 1 in sequence and then communicates with the artificial intelligence device 4.

[0073] Further preferably, an extension plate 13 is arranged on the far tower surface of the vertical plate 12, wherein the far tower surface refers to the other side surface away from the in-service concrete tower 100; the extension plate 13 is provided with a socket 131 for inserting one end of an L-shaped rigid member 14, and the other end of the L-shaped rigid member 14 is movably hinged to the annular body 1, so as to lock the automatic retractability of the transverse connecting member 11 through the plug-in cooperation between the L-shaped rigid member 14 and the extension plate 13.

[0074] Further preferably, the far pole tower surfaces of any two adjacent vertical plates 12 in the annular direction are respectively connected to the two ends of the elastic bandage 15 in a one-to-one correspondence, wherein the far pole tower surface refers to the other side surface away from the in-service concrete pole tower 100. Through the design of the elastic bandage 15, all the vertical plates 12 can be elastically connected together to improve their movement stability during the process of climbing up and down the pole.

[0075] In summary, the prestressed concrete tower internal defect detection system provided by this embodiment has the following technical effects:

[0076] (1) This embodiment provides a new solution for concrete internal flaw detection applicable to in-service concrete towers, namely, comprising a ring body, an ultrasonic transmitting probe, a plurality of ultrasonic receiving probes, an artificial intelligence device and a display device, wherein the artificial intelligence device is used to first extract ultrasonic signal characteristic data of each receiving probe, and then, for each position in the cross section of the currently measured tower, import the corresponding polar coordinate data and the polar coordinate data and ultrasonic signal characteristic data of each receiving probe into a concrete crack recognition model based on a machine learning algorithm and pre-trained, and output the confidence level of the concrete cracks appearing at the corresponding position. Finally, a grayscale image of the cross section of the currently measured tower is drawn based on the confidence level and transmitted to the display device as the current internal defect detection result for visual display. In this way, there is no need to manually climb the pole to obtain the internal information of the concrete tower, and the internal defect detection results can be quickly, efficiently and accurately obtained on site and displayed on site visually, which is convenient for practical application and promotion.

[0077] Finally, it should be noted that the above description is only a preferred embodiment of the present invention and is not intended to limit the protection scope of the present invention. Any modification, equivalent replacement, improvement, etc. made within the spirit and principle of the present invention shall be included in the protection scope of the present invention.

Claims

1. A prestressed concrete tower internal defect detection system based on machine learning algorithm, It is characterized in that The invention comprises an annular body (1), an ultrasonic transmitting probe (2), a plurality of ultrasonic receiving probes (3), an artificial intelligence device (4) and a display device, wherein the annular body (1) is used to surround an in-service concrete pole tower (100) in an annular direction and to move vertically up and down on the in-service concrete pole tower (100), the ultrasonic transmitting probe (2) and the plurality of ultrasonic receiving probes (3) are arranged in an annular manner at intervals on the inner side of the annular body (1), and the artificial intelligence device (4) is arranged on the outer side of the annular body (1); The ultrasonic transmitting probe (2) is communicatively connected to the artificial intelligence machine device (4) and is used to transmit an ultrasonic signal to the in-service concrete tower (100) under the control of the artificial intelligence machine device (4); The ultrasonic receiving probe (3) is communicatively connected to the artificial intelligence machine device (4) and is used to receive ultrasonic signals reflected or transmitted by the in-service concrete tower (100), and transmit the received ultrasonic signals to the artificial intelligence machine device (4) in real time; The AI ​​device (4) is communicatively connected to the display device and is used to process the ultrasonic signals from the multiple ultrasonic receiving probes (3) in the following manner: For each ultrasonic receiving probe (3) among the plurality of ultrasonic receiving probes (3), performing feature data extraction processing on an ultrasonic signal from a corresponding probe to obtain corresponding ultrasonic signal feature data; Determining the current measured height position of the in-service concrete pole tower (100) based on the current climbing pole height of the annular body (1) and the known arrangement heights of the ultrasonic transmitting probe (2) and the plurality of ultrasonic receiving probes (3) respectively relative to the annular body (1); Determining, based on the known cross-section type of the in-service concrete tower (100), a current measured tower cross-section of the in-service concrete tower (100) corresponding to the current measured height position, wherein the center point of the current measured tower cross-section is the annular center point of the annular body (1); For each position in the cross section of the currently measured tower, the corresponding polar coordinate data and the polar coordinate data and ultrasonic signal characteristic data of each ultrasonic receiving probe (3) are imported into a concrete crack recognition model based on a machine learning algorithm and has been pre-trained, and the confidence of the concrete crack occurring at the corresponding position is output, and the corresponding grayscale value negatively correlated with the confidence is calculated, wherein the polar coordinate data refers to the coordinate data in a polar coordinate system, and the polar coordinate system takes the center point of the cross section of the tower as the pole and the opposite direction of the ultrasonic emission direction as the polar axis direction; According to the grayscale values ​​of the respective positions, a grayscale image of the cross section of the currently measured tower is obtained; The current measured height position and the grayscale image of the current measured pole tower cross section are transmitted as the current internal defect detection result of the in-service concrete pole tower (100) to the display device for visual display.

2. The internal defect detection system of the prestressed concrete tower according to claim 1, It is characterized in that It also includes a pole climbing drive mechanism (6) for driving the annular body (1) to move vertically up and down on the in-service concrete pole tower (100), wherein a controlled end of the pole climbing drive mechanism (6) is communicatively connected to the artificial intelligence device (4).

3. The internal defect detection system of the prestressed concrete tower according to claim 2, It is characterized in that The climbing pole driving mechanism (6) comprises a vertical rod (61), an inclined rod (62), a driving wheel (63), an inclined spring (64) and an electric motor, wherein the controlled end of the electric motor is used as the controlled end of the climbing pole driving mechanism (6); The first end of the vertical rod (61) is fixedly connected to the annular body (1), the first end of the oblique rod (62) is movably hinged to the second end of the vertical rod (61), and the driving wheel (63) is arranged at the second end of the oblique rod (62); The inclined tension spring (64) is always in a tensioned state, and the two ends of the inclined tension spring (64) are respectively connected to the first end of the vertical rod (61) and the second end of the inclined rod (62) in a one-to-one correspondence, so as to ensure that the driving wheel (63) always contacts the outer surface of the in-service concrete pole tower (100); The motor is mounted on the inclined rod (62), and the output shaft of the motor is drivingly connected to the driving wheel (63).

4. The internal defect detection system of the prestressed concrete tower according to claim 1, It is characterized in that A plurality of transverse connecting members (11) corresponding to the ultrasonic transmitting probe (2) and the plurality of ultrasonic receiving probes (3) are arranged at intervals in the inner side of the annular body (1) in an annular direction, wherein one end of the transverse connecting member (11) is fixedly connected to the inner side of the annular body (1), and the other end of the transverse connecting member (11) is fixedly connected to the vertical plate (12); At least three universal wheels (121) are arranged on the near-tower surface of the vertical plate (12), and the ultrasonic transmitting probe (2) or the ultrasonic receiving probe (3) corresponding to the vertical plate (12) and the transverse connecting member (11) is embedded in the near-tower surface, wherein the near-tower surface refers to a side surface close to the in-service concrete tower (100); The transverse connecting member (11) has automatic retractability to ensure that the corresponding at least three universal wheels (121) always contact the outer surface of the in-service concrete pole tower (100).

5. The internal defect detection system of the prestressed concrete tower according to claim 4, It is characterized in that The transverse connecting member (11) comprises a transverse rod (111), a transverse sleeve (112) and an elastic member (113); One end of the transverse rod (111) is fixedly connected to the vertical plate (12), and the other end of the transverse rod (111) extends into the interior of the transverse sleeve (112) and is provided with an annular ear plate (114) extending outwardly; One end of the transverse sleeve (112) is fixedly connected to the inner side of the annular body (1); an annular baffle (115) cooperating with the annular ear plate (114) is provided in the pipe opening at the other end of the transverse sleeve (112); the inner cavity of the annular baffle (115) is slidably engaged with the transverse rod (111); and the inner cavity of the transverse sleeve (112) is slidably engaged with the annular ear plate (114); The elastic member (113) is located in the internal cavity of the transverse sleeve (112) and is used to push or pull the transverse rod (111) to move in a direction away from the annular body (1), so as to enable the transverse connecting member (11) to have automatic retractability.

6. The internal defect detection system of the prestressed concrete tower according to claim 5, It is characterized in that The elastic member (113) is a spring that is always in a stretched state, wherein the spring is located between the annular ear plate (114) and the annular baffle plate (115); The spring is sleeved on the outer circumference of the transverse rod (111), and the two ends of the spring are fixedly connected to the annular ear plate (114) and the annular baffle plate (115) in a one-to-one correspondence, so as to pull the transverse rod (111) to move in a direction away from the annular body (1).

7. The internal defect detection system of the prestressed concrete tower according to claim 5, It is characterized in that The transverse rod (111) and the annular body (1) are both hollow structures, and the communication cable of the ultrasonic transmitting probe (2) or the ultrasonic receiving probe (3) passes through the corresponding transverse rod (111), the corresponding transverse sleeve (112) and the annular body (1) in sequence, and then is communicatively connected to the artificial intelligence device (4).

8. The internal defect detection system of the prestressed concrete tower according to claim 4, It is characterized in that An extension plate (13) is arranged on the far tower surface of the vertical plate (12), wherein the far tower surface refers to the other side surface away from the in-service concrete tower (100); The extension plate (13) is provided with a plug hole (131) for inserting one end of an L-shaped rigid member (14), and the other end of the L-shaped rigid member (14) is movably hinged to the annular body (1), so that the automatic telescopic property of the transverse connecting member (11) is locked by plugging and matching the L-shaped rigid member (14) with the extension plate (13).

9. The internal defect detection system of the prestressed concrete tower according to claim 4, It is characterized in that The far tower surfaces of any two circumferentially adjacent vertical plates (12) are respectively connected to two ends of an elastic bandage (15) in a one-to-one correspondence, wherein the far tower surface refers to the other side surface away from the in-service concrete tower (100).

10. The internal defect detection system of the prestressed concrete tower according to claim 1, It is characterized in that The annular body (1) is formed by joining two halves of a hoop body (10) together.