Flying robot facing glass curtain wall and automated detection method

Abstract

The application discloses a kind of glass curtain wall-oriented flying robots and automated detection method. Wherein, flying robot includes robot body and multifunctional detection equipment installed on robot body;Robot body is equipped with wireless transmission device, vacuum adsorption device, support bracket device and flight system;Multifunctional detection equipment includes excitation source component, time-domain voltage signal acquisition sensor component, vibration data acquisition sensor component and image data acquisition sensor component;Excitation source component is installed on robot body by automatic telescopic arm;Time-domain voltage signal acquisition sensor component is arranged on the side of robot body;Vibration data acquisition sensor component is arranged on the top of robot body;Image data acquisition sensor component is arranged on the bottom of robot body;The application can realize the comprehensive automated detection of target glass curtain wall, realize the efficient and accurate detection of target glass curtain wall defect, high automation level, short detection cycle.

Description

Technical Field

[0001] This disclosure relates to the field of structural engineering damage detection, and in particular to a flying robot and automated detection method for glass curtain walls. Background Technology

[0002] Glass curtain walls are multifunctional building envelope structures that integrate lighting, rain protection, wind protection, and thermal insulation. With the increasing aesthetic demands of urban construction and the growing need for high-rise buildings, glass curtain walls, as a synthesis of architectural technology, function, structure, and art, have been widely used in the field of building structures, especially in high-rise buildings. Due to the diverse range of architectural designs and the structural characteristics required by high-rise buildings, the forms and structural support methods of glass curtain walls have become increasingly complex. Under the long-term effects of wind and temperature loads under normal conditions, high-rise glass curtain walls will inevitably develop some defects over time, such as scratches and spontaneous breakage of glass panels, aging and corrosion of structural adhesives, and stress relaxation and deformation of the supporting structure.

[0003] Currently, most glass curtain wall inspections rely on human operators working at heights to inspect the quality of each pane of glass. This manual inspection is inefficient, costly, and dangerous. Existing automated inspection methods, such as photoelasticity, are only for detecting spontaneous breakage defects in safety glass, while natural vibration methods can only detect defects in the bottom curtain wall structure and cannot meet the comprehensive damage inspection needs of high-rise glass curtain walls.

[0004] In summary, the existing publicly available solutions cannot detect hidden damage to glass panels, structural adhesives, and supporting structures. The results obtained are rather one-sided, resulting in limited testing areas, long testing cycles, and low testing accuracy. Summary of the Invention

[0005] In view of this, the present disclosure provides a flying robot and an automated inspection method for glass curtain walls, which at least partially solves the problems of poor detection accuracy, one-sided measurement results, and long inspection cycle in the prior art.

[0006] In a first aspect, the present disclosure provides a flying robot for glass curtain walls, including a robot body and a multi-functional detection device installed on the robot body;

[0007] The robot body is equipped with a wireless transmission device, a vacuum adsorption device, and a flight system;

[0008] The multifunctional detection device includes an excitation source component, a time-domain voltage signal acquisition sensor component, a vibration data acquisition sensor component, and an image data acquisition sensor component; the excitation source component is mounted on the robot body via an automatic telescopic arm; the time-domain voltage signal acquisition sensor component is disposed on the periphery of the robot body; the vibration data acquisition sensor component is disposed on the top of the robot body; and the image data acquisition sensor component is disposed on the bottom of the robot body.

[0009] Optionally, the excitation source assembly includes an automatic force hammer with an air damping unit;

[0010] The time-domain voltage signal acquisition sensor assembly includes an accelerometer and a microphone sensor;

[0011] The vibration data acquisition sensor assembly includes a Doppler laser and a microwave radar;

[0012] The image data acquisition sensor assembly includes an industrial camera and a hyperspectral camera.

[0013] Secondly, this application discloses an automated inspection method for glass curtain walls, which includes: constructing a three-dimensional model based on the actual information of the target glass curtain wall collected;

[0014] Finite element analysis is performed on the three-dimensional model to obtain simulated time-frequency response information; wherein, the simulated time-frequency response information includes displacement-time time-domain data, velocity-time time-domain data, and acceleration-time time-domain data;

[0015] Based on the simulated time-frequency response information, a detection strategy is obtained; wherein, the detection strategy includes a time-frequency measured sensor selection strategy, a time-frequency measured detection point setting position, and a time-frequency measured robot walking path, and the time-frequency measured detection point setting position is a measurement point position that can excite all target modes;

[0016] Based on the detection strategy, an excitation test is performed to obtain actual time-frequency response information; wherein, the actual time-frequency response information includes the first wave arrival time, vibration attenuation amplitude, energy spectrum, and root mean square deviation of energy;

[0017] Based on the actual time-frequency response information, the defect type of the target glass curtain wall is obtained.

[0018] Optionally, performing finite element analysis on the three-dimensional model to obtain simulated time-frequency response information includes: establishing a finite element model based on the three-dimensional model;

[0019] The finite element model is analyzed based on preset parameters to obtain the simulated time-frequency response information;

[0020] The preset parameters include the excitation point position and the data output point position; wherein, the data output point position is equipped with a displacement detection sensor, a velocity detection sensor and an acceleration detection sensor.

[0021] Optionally, the step of performing excitation testing based on the detection strategy to obtain actual time-frequency response information includes: based on the time-frequency measurement sensor selection strategy, deploying time-frequency measurement sensors at the time-frequency measurement detection point setting location, using a high-precision fully automatic force hammer to excite at the time-frequency measurement detection point setting location, and obtaining the actual time-frequency response information.

[0022] Optionally, the locations of the time-frequency measurement detection points include the structural adhesive bonding area corresponding to the outer surface of the facade and the weather-resistant adhesive area between the two glass panels on the outer side of the facade.

[0023] Optionally, the time-frequency measurement sensor selection strategy includes ensuring that the sensor's frequency testing range covers the target detection frequency and that the sensor's sensitivity matches the structural characteristics of the glass curtain wall.

[0024] Optionally, obtaining the defect type of the target glass curtain wall based on the actual time-frequency response information includes: performing morphological analysis, threshold analysis, digital image analysis, and hyperspectral analysis on the actual time-frequency response information to determine whether there is apparent damage; if so, performing refined re-measurement on the defect location with apparent damage to obtain the defect type of the target glass curtain wall.

[0025] If not, perform main frequency analysis, amplitude analysis, impact-response analysis, and impact-acoustic vibration analysis on the actual time-frequency response information to determine whether there is any hidden damage; if so, perform detailed re-measurement of the defect location with hidden damage to obtain the defect type of the target glass curtain wall.

[0026] If not, the next measurement point will be automatically detected based on the robot's travel path as described in the time-frequency measurement.

[0027] Optionally, the target glass curtain wall is not prone to spontaneous breakage.

[0028] Optionally, before constructing the 3D model, the method further includes: determining whether the target glass curtain wall has a preset crack state based on the actual information of the target glass curtain wall;

[0029] If so, the target glass curtain wall is determined to be an abnormal area, and a new target area without the preset crack state is selected for detection.

[0030] Based on the target area to be detected, an updated 3D model is constructed;

[0031] The preset crack state is a multi-crack state; the abnormal region includes areas where spontaneous explosion is possible.

[0032] The first aspect of this disclosure discloses a flying robot for glass curtain walls, which is a multi-functional integrated flying inspection robot capable of flexible operation on glass curtain wall structures. It can be excited by an excitation source component and capture apparent damage through an image data acquisition sensor component. Simultaneously, under the action of vibration, a time-domain voltage signal acquisition sensor component and a vibration data acquisition sensor component can capture and analyze vibration signals. Overall, it achieves efficient fusion and collaborative testing of impact-acoustic vibration, impact-resonance, machine vision, hyperspectral, and laser vibration methods, realizing integrated, synchronous, and accurate testing of both apparent and hidden damage to glass curtain walls. This can compensate for the shortcomings of existing detection methods, overcome the technical bottleneck of high-altitude operations, improve the automation level of glass curtain wall structural damage detection, and significantly improve testing efficiency and identification accuracy.

[0033] The automated inspection method for glass curtain walls provided in this disclosure first collects actual information about the target glass curtain wall to ensure that the constructed model is consistent with the target glass curtain wall information; then, under ideal conditions, time-frequency simulation analysis is performed on the constructed three-dimensional model to obtain theoretical data information to guide the actual measurement of the target glass curtain wall; finally, the measured time-frequency data is compared and analyzed to accurately and comprehensively obtain the defect type of the target glass curtain wall, with a high degree of automation, high accuracy, high reliability, and high safety level.

[0034] The method disclosed in this application uses a flying robot as a carrier to conduct automated field tests, enabling damage detection of glass curtain walls at high altitudes and in areas inaccessible to inspection personnel. It also achieves non-destructive testing of glass curtain wall structures. Furthermore, this solution efficiently integrates impact-acoustic vibration, impact-response, microwave radar, laser Doppler vibration measurement, machine vision, and hyperspectral testing methods. During the testing process, it enables efficient detection and real-time analysis of the safety performance of glass curtain wall structures, effectively improving the accuracy and efficiency of damage detection testing.

[0035] The above description is merely an overview of the technical solution disclosed herein. In order to better understand the technical means of this disclosure and to implement it in accordance with the contents of the specification, and to make the above and other objects, features and advantages of this disclosure more apparent and understandable, preferred embodiments are described below in detail with reference to the accompanying drawings. Attached Figure Description

[0036] To more clearly illustrate the technical solutions of the embodiments of this disclosure, the drawings used in the embodiments will be briefly introduced below. Obviously, the drawings described below are only some embodiments of this disclosure. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0037] Figure 1 This is a flowchart illustrating a specific embodiment of the automated inspection method for glass curtain walls in this application.

[0038] Figure 2 for Figure 1 A flowchart illustrating the method for obtaining simulated time-frequency response information.

[0039] Figure 3 This is a structural diagram of a flying robot facing a glass curtain wall.

[0040] Figure 4 for Figure 3 Another perspective diagram.

[0041] Figure 5 for Figure 4 A magnified view of part A in the image.

[0042] Figure 6 This is a partial enlarged view of the support bracket device.

[0043] Figure 7 This is a schematic diagram showing the loosening of the supporting structure.

[0044] Figure 8 This is a schematic diagram of the point support structure on the glass panel.

[0045] Figure 9 This is a schematic diagram of the structural adhesive bonding area corresponding to the outer surface of the facade.

[0046] Figure 10 This is a flowchart illustrating another specific embodiment of the automated inspection method for glass curtain walls in this application.

[0047] Figure 11 This is a schematic diagram of the root mean square deviation under different operating conditions.

[0048] Figure 12 This is a schematic diagram of the Pearson correlation coefficient under different operating conditions.

[0049] Figure 13 This is a distribution diagram of the energy coefficients of each subband of the wavelet in the undamaged state.

[0050] Figure 14 This is a distribution diagram of the energy coefficients of each subband of the wavelet under unilateral damage conditions.

[0051] Figure 15 This is a distribution diagram of the energy coefficients of each subband of the wavelet under bilateral damage conditions.

[0052] Figure 16 This is a distribution diagram of the energy coefficients of each subband of the wavelet under the three-sided damage state.

[0053] Figure 17This is the wavelet time-frequency plot under the undamaged state.

[0054] Figure 18 This is a wavelet time-frequency plot under unilateral damage conditions.

[0055] Figure 19 This is a wavelet time-frequency plot under bilateral damage conditions.

[0056] Figure 20 This is the wavelet time-frequency plot under the three-sided damage state.

[0057] Figure 21 This is a schematic diagram of the automated detection system for defects in glass curtain wall structures as described in this application.

[0058] Explanation of reference numerals in the attached figures:

[0059] 1. Flying robot; 11. Wireless transmission device; 12. Flight wing; 13. Wing protection frame; 14. Vacuum suction cup; 15. Support bracket device; 16. Air vibration damping unit; 2. Industrial camera; 3. High-frequency speaker; 4. Hyperspectral camera; 5. Microwave radar; 6. Micro-vibration Doppler vibrometer; 7. Low-frequency microphone; 8. Automatic force hammer; 9. Low-frequency acceleration sensor; 10. Telescopic movable bracket; 20. Glass panel; 21. Point support structure; 31. Weather-resistant adhesive area; 41. Hidden damage; 42. Surface damage. Detailed Implementation

[0060] The present disclosure will now be described in further detail with reference to the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are for illustrative purposes only and are not intended to limit the scope of the disclosure. Furthermore, it should be noted that, for ease of description, only the parts relevant to the present disclosure are shown in the accompanying drawings.

[0061] It should be noted that, where there is no conflict, the embodiments and features described in this disclosure can be combined with each other. The technical solutions of this disclosure will now be described in detail with reference to the accompanying drawings and embodiments.

[0062] Unless otherwise stated, the exemplary implementations / embodiments shown are to be understood as providing exemplary features of various details that provide ways in which the technical concepts of this disclosure can be implemented in practice. Therefore, unless otherwise stated, the features of various implementations / embodiments may be additionally combined, separated, interchanged and / or rearranged without departing from the technical concepts of this disclosure.

[0063] The use of crosshairs and / or shading in the accompanying drawings is generally used to clarify the boundaries between adjacent components. Thus, unless otherwise stated, the presence or absence of crosshairs or shading does not convey or indicate any preference or requirement for the specific material, material properties, dimensions, proportions, commonalities between the illustrated components, or any other characteristics, properties, etc., of the components. Furthermore, in the accompanying drawings, the dimensions and relative dimensions of components may be exaggerated for clarity and / or descriptive purposes. When exemplary embodiments can be implemented differently, a specific process sequence may be performed in a different order than that described. For example, two consecutively described processes may be performed substantially simultaneously or in the reverse order of their description. Furthermore, the same reference numerals denote the same components.

[0064] When a component is referred to as being "on" or "above" another component, "connected to," or "joined to" another component, the component may be directly on, directly connected to, or directly joined to the other component, or there may be intermediate components. However, when a component is referred to as being "directly on" another component, "directly connected to," or "directly joined to" another component, there are no intermediate components. Therefore, the term "connection" can refer to a physical connection, an electrical connection, etc., and may or may not have intermediate components.

[0065] For descriptive purposes, this disclosure may use spatial relative terms such as “below,” “under,” “below,” “down,” “above,” “above,” “higher,” and “side (e.g., in a “sidewall”)” to describe the relationship between one component and another component as shown in the accompanying drawings. In addition to the orientations depicted in the drawings, the spatial relative terms are also intended to encompass different orientations of the device during use, operation, and / or manufacture. For example, if the device in the drawings is flipped, a component described as “below” or “under” another component or feature would subsequently be positioned “above” said other component or feature. Thus, the exemplary term “below” can encompass both “above” and “below” orientations. Furthermore, the device may be otherwise positioned (e.g., rotated 90 degrees or in other orientations), thus interpreting the spatial relative descriptive terms used herein accordingly.

[0066] The terminology used herein is for the purpose of describing particular embodiments and is not intended to be limiting. As used herein, unless the context clearly indicates otherwise, the singular forms “a” and “the” are intended to include the plural forms as well. Furthermore, when the terms “comprising” and / or “including” and variations thereof are used in this specification, it indicates the presence of the stated features, integrals, steps, operations, parts, components, and / or groups thereof, but does not exclude the presence or addition of one or more other features, integrals, steps, operations, parts, components, and / or groups thereof. It should also be noted that, as used herein, the terms “substantially,” “about,” and other similar terms are used as approximate terms rather than as terms of degree, thus explaining the inherent biases in measurements, calculated values, and / or provided values ​​that would be recognized by one of ordinary skill in the art.

[0067] In this application, since glass curtain walls have different specifications and types, a three-dimensional model of the target glass curtain wall is constructed before the actual measurement. The three-dimensional model includes all the details of the target glass curtain wall, especially the structural adhesive, weather-resistant adhesive, and support system of the curtain wall, which are set with different damping parameters and contact types to ensure that the constraints of each part of the constructed three-dimensional model are consistent with the actual situation.

[0068] Reference Figure 1 The first aspect of this application discloses an automated inspection method for glass curtain walls, the method comprising the following steps:

[0069] S10, Based on the actual information of the target glass curtain wall, construct a three-dimensional model (i.e., construct a three-dimensional model of the target glass curtain wall). The actual information includes the structure, dimensions, materials, and other information of the target glass curtain wall.

[0070] S20, Time-Frequency Simulation Analysis: Perform finite element analysis on the three-dimensional model to obtain simulated time-frequency response information (time-domain data curves, curves showing the change of vibration amplitude over time).

[0071] The simulated time-frequency response information includes displacement-time time-domain data, velocity-time time-domain data, and acceleration-time time-domain data.

[0072] In this embodiment, other finite element analysis software such as ANSYS and ABAQUS can be used to perform finite element analysis on the three-dimensional model.

[0073] S30: Based on the simulated time-frequency response information, a detection strategy is obtained.

[0074] The detection strategy includes the selection strategy for time-frequency measurement sensors, the setting position of time-frequency measurement detection points, and the walking path of time-frequency measurement robots. The setting position of time-frequency measurement detection points is the measurement point position that can excite all target modes.

[0075] S40, Time-Frequency Response Analysis: Based on the detection strategy, excitation tests are performed to obtain actual time-frequency response information.

[0076] The actual time-frequency response information includes the arrival time of the first wave, the vibration attenuation amplitude, the energy spectrum, and the root mean square deviation of the energy.

[0077] S50, Defect Analysis: Based on actual time-frequency response information, obtain the defect type of the target glass curtain wall.

[0078] The automated inspection method for glass curtain walls provided in this disclosure first collects actual information about the target glass curtain wall to ensure that the constructed model is consistent with the target glass curtain wall information; then, under ideal conditions, time-frequency simulation analysis is performed on the constructed three-dimensional model to obtain theoretical data information to guide the actual measurement of the target glass curtain wall; finally, the measured time-frequency data is compared and analyzed to accurately and comprehensively obtain the defect type of the target glass curtain wall, with a high degree of automation, high accuracy, high reliability, and high safety level.

[0079] Reference Figure 2 The method for obtaining simulated time-frequency response information specifically includes the following steps:

[0080] S21. Based on the three-dimensional model, a finite element model is established.

[0081] S22, based on preset parameters, analyzes the finite element model to obtain simulated time-frequency response information.

[0082] The preset parameters include the excitation point location (i.e., the load application point location) and the data output point location; the data output point location is equipped with a displacement detection sensor, a velocity detection sensor, and an acceleration detection sensor.

[0083] In this embodiment, the method for obtaining actual time-frequency response information specifically includes: based on the selection strategy of time-frequency measurement sensor, deploying time-frequency measurement sensors at the location of the time-frequency measurement detection point, and using a high-precision fully automatic force hammer to excite at the location of the time-frequency measurement detection point to obtain actual time-frequency response information.

[0084] The selection strategy for time-frequency measurement sensors includes ensuring that the sensor's frequency testing range covers the target detection frequency and that the sensor's sensitivity matches the structural characteristics of the glass curtain wall.

[0085] Reference Figures 3 to 5 In this embodiment, the target glass curtain wall does not spontaneously explode; the excitation test is conducted using a flying robot 1.

[0086] Specifically, the flying robot 1 includes a robot body and multifunctional testing equipment. The robot body is equipped with a wireless transmission device 11, a vacuum adsorption device, a support bracket device 15, and a flight system. Using the flying robot as a carrier, the flight system and vacuum adsorption system enable the carrier robot to move stably on the glass curtain wall and resist high-altitude wind resistance. The support bracket device 15 allows the flying robot to adhere to the adjacent sides of the glass unit under test, eliminating the influence of the robot's added mass and its own vibration on the glass curtain wall unit. Through the integrated multifunctional testing equipment, comprehensive and synchronous detection of defects in the glass curtain wall is achieved, improving the accuracy of damage testing and solving the technical bottlenecks faced in non-destructive testing of glass curtain wall structures.

[0087] In this embodiment, the vacuum adsorption device is preferably a vacuum suction cup 14; the flight system includes a flight wing 12 and a wing protection frame 13.

[0088] Furthermore, referring to Figure 6 The vacuum suction cup also has an air vibration damping unit 16 to eliminate the influence of the vibration of the curtain wall glass adsorbed by the robot, and to ensure that the vibration signal measured by the vibration measurement module is the true vibration information of the curtain wall unit being tested.

[0089] The multifunctional detection equipment includes an excitation source component, a time-domain voltage signal acquisition sensor component, a vibration data acquisition sensor component, and an image data acquisition sensor component. The excitation source component is mounted on the robot body via an automatic telescopic arm (i.e., a telescopic movable support 10). The time-domain voltage signal acquisition sensor component is located on the periphery of the robot body. The vibration data acquisition sensor component is located on the top of the robot body. The image data acquisition sensor component is located on the bottom of the robot body.

[0090] The time-domain voltage signal acquisition sensor assembly includes an accelerometer and a low-frequency microphone sensor to acquire time-frequency voltage signals; the vibration data acquisition sensor assembly acquires vibration data including vibration data of the glass panel 20 acquired by Doppler laser and microwave radar 5; the image data acquisition sensor assembly is used to acquire image data, including surface images and flatness data of the glass panel 20 (e.g., data on apparent damage 42), and infrared hyperspectral structural composition change data (e.g., data on adhesive strip aging).

[0091] Preferably, the excitation source component includes an automatic hammer 8 and a high-pitched horn 3, wherein the automatic hammer 8 is an automatic hammer with an air damping and vibration isolation unit to eliminate the vibration of the robot caused by the hammer during the striking process.

[0092] Specifically, a wirelessly controlled flying inspection robot equipped with a support bracket device 15 containing a vacuum suction cup 14 and an air vibration damping unit 16 is positioned on both sides of the glass curtain wall unit structure under test. It can be individually excited by impact and excited by the high-pitched speaker 3, or both simultaneously, thus enabling simultaneous testing of apparent damage 42 and hidden damage 41 of the curtain wall. In this application, the flying robot 1 is used as a carrier. The flight system and support bracket device 15 (vacuum suction cup 14 with air vibration damping unit 16) enable the carrier robot to move stably on the glass curtain wall and resist high-altitude wind resistance.

[0093] Excitation is achieved by the impact of an automatic hammer 8 or a high-pitched horn 3. The industrial camera 2 and the hyperspectral camera 4 capture the apparent damage 42. At the same time, under the action of vibration, the low-frequency microphone 7, the low-frequency acceleration sensor 9, the microwave radar 5, and the laser Doppler (i.e., the micro-vibration Doppler vibrometer 6) can capture and analyze the vibration signal. Overall, the efficient integration and collaborative testing of impact-acoustic vibration, impact-response, machine vision, hyperspectral and laser vibration methods are realized. This enables integrated, synchronous and accurate testing of the apparent damage 42 and the hidden damage 41 of the glass curtain wall. It can make up for the shortcomings of existing detection methods, break through the technical bottleneck of high-altitude operation, improve the automation level of glass curtain wall structural damage detection, and significantly improve testing efficiency and identification accuracy.

[0094] Reference Figures 7 to 9 The flying robot 1 disclosed in this application is capable of detecting apparent damage 42 on the glass panel 20, hidden damage 41 on the glass panel 20, hidden damage 41 on the point support structure 21, and point support structure on the glass panel 20.

[0095] Among them, apparent damage 42 includes surface cracks; hidden damage 41 includes loosening of point support structures, etc.

[0096] The locations of the time-frequency test points include the structural adhesive bonding area corresponding to the outer surface of the facade and the weather-resistant adhesive area between the two glass panels on the exterior of the facade.

[0097] When the time-frequency measurement test point is set at the structural adhesive bonding area corresponding to the outer surface of the facade, the test format is: single-emission-single-receiver. The time-frequency measurement sensor selection strategy includes one or two accelerometers. A force hammer is used to excite the sensor near the center of the curtain wall. The two accelerometers collect the time-domain data of the response within a specified time. Further post-processing of the time-domain curves can yield the arrival time of the first wave, the attenuation amplitude of the vibration during propagation, the energy spectrum, and the root mean square deviation of the energy.

[0098] When the actual frequency measurement test point is set at the weather-resistant sealant area 31 between the two glass panels on the exterior of the facade (i.e., the reference test points adjacent to the weather-resistant sealant between the two glass panels on the exterior of the facade), an accelerometer is placed at each of the reference test points adjacent to the weather-resistant sealant. A force hammer is applied to one side of one of the accelerometers. The accelerometer can collect time-domain data within a specified time range. Based on this time-domain data, inversion calculations are performed to obtain the thickness of the propagation medium and the elastic material parameters. The thickness (based on historical construction design data) can be used to determine whether the weather-resistant sealant has been improperly applied or whether this has caused leakage. The elastic material parameters (based on GB / T14683 "Silicone Building Sealants") can be used to determine whether the weather-resistant sealant has experienced local hardening. The remaining service life of the weather-resistant sealant can also be calculated.

[0099] Furthermore, the method for obtaining the defect type of the target glass curtain wall includes: performing morphological analysis, threshold analysis, digital image analysis, and hyperspectral analysis on the actual time-frequency response information to determine whether apparent damage 42 exists; if so, performing a refined re-measurement of the defect location where apparent damage 42 exists to obtain the defect type of the target glass curtain wall.

[0100] If not, perform main frequency analysis, amplitude analysis, impact-response analysis, and impact-acoustic vibration analysis on the actual time-frequency response information to determine whether there is hidden damage 41; if so, perform detailed re-measurement of the defect location where there is hidden damage 41 to obtain the defect type of the target glass curtain wall.

[0101] If not, the next measurement point will be automatically detected based on the robot's actual travel path measured in time and frequency.

[0102] Wavelet transform is performed on the actual time-frequency response information to obtain the working time-frequency energy value; it is determined whether the working time-frequency energy value belongs to the first interval range. If so, the defect type of the target glass curtain wall is directly obtained; the first interval range includes the non-damaged segment and the multi-damaged segment.

[0103] If not, process the working time-frequency energy value according to the preset error to obtain the corrected working time-frequency energy data; based on the second interval range, obtain the defect type of the target glass curtain wall.

[0104] Based on the preset energy value range, the damage interval of the working time-frequency energy value is initially determined; the preset energy value range includes a first interval range and a second interval range, wherein the first interval range includes a no-damage segment and a multi-damage segment, and the second interval range includes a no-single-damage segment, a single-double-damage segment, and a double-triple-damage segment.

[0105] In this application, since glass curtain walls have different specifications and types, a three-dimensional model of the target glass curtain wall is constructed before the actual measurement. The three-dimensional model includes all the details of the target glass curtain wall, especially the structural adhesive, weather-resistant adhesive, and support system of the curtain wall, which are set with different damping parameters and contact types to ensure that the constraints of each part of the constructed three-dimensional model are consistent with the actual situation.

[0106] For collecting actual information, the following technical solutions can be adopted: For glass curtain walls at high altitudes, the following feasible solutions can be used to collect their actual information: 1) Manual on-site survey: Dispatch professional surveyors to the site for on-site surveys, obtaining information such as the shape, area, and structure of the curtain wall through visual observation and measuring instruments; 2) Aerial drone (UAV) photography: Use high-definition cameras from aerial drones (UAVs) to photograph the shape and details of the curtain wall to obtain its three-dimensional information. If more detailed information is needed, multi-angle, multi-height fixed-point photography and photography of light reflection at different frequency bands can be used; 3) High-altitude robot: Use a high-altitude robot, suspending it on the surface of the curtain wall to conduct detection and surveying to obtain detailed information about the glass curtain wall; 4) 3D laser scanning: Use a 3D laser scanner to scan the glass curtain wall to obtain high-precision, large-area point cloud data, and then use point cloud reconstruction tools to construct a three-dimensional model of the curtain wall. Of course, different combinations of methods can also be used according to different needs to achieve the goal of accurately collecting actual information about the curtain wall.

[0107] In another embodiment, before constructing the three-dimensional model, the method further includes: determining whether the target glass curtain wall has a preset crack state based on the actual information of the target glass curtain wall;

[0108] If so, the target glass curtain wall is determined to be an abnormal area, and a new target area without the preset crack state is selected (i.e., the test is moved to the next test area).

[0109] An updated 3D model is constructed based on the target region to be detected.

[0110] The preset crack state is a multi-crack state; the abnormal area includes areas where spontaneous explosions are possible.

[0111] It should be noted that the simulation in this application is for the purpose of guiding actual testing. Therefore, during the initial data collection, a preliminary judgment can be made on the target glass curtain wall before modeling, that is, to determine whether the target glass curtain wall is prone to spontaneous breakage. If spontaneous breakage is present, forcibly conducting actual testing will cause greater damage to the target glass curtain wall. Therefore, there is no need to conduct subsequent actual testing, and it is meaningless to simulate a target glass curtain wall that is prone to spontaneous breakage.

[0112] Furthermore, the methods for determining whether a target glass curtain wall is prone to spontaneous breakage can include the following points: 1) If the cracked edge of the glass is neat, the number of cracks is small, and the crack line is a tortuous single line or double line, then spontaneous breakage can be determined; 2) If the crack line of the glass is perpendicular to the edge of the glass, then spontaneous breakage can be considered, and it should be further determined whether the phenomenon is caused by bending stress or by defects at the edge of the glass; 3) If the crack line in the middle area of ​​the glass is mostly arc-shaped, then spontaneous breakage can be determined.

[0113] Reference Figure 10 In another embodiment of this application, the automated inspection method for glass curtain walls specifically includes: starting a flying robot and flying to the target glass curtain wall; 1) visually inspecting for surface defects; 2) determining whether surface defects exist; if so, stopping the tapping, recording the position of the glass and the surface defect with the visual device, then changing the measuring point, and the flying robot flying to the next measuring point for testing; if not, lowering the support bracket and increasing the negative pressure of the vacuum suction cup to a stable and measurable state, then controlling the hammer to start tapping, and exciting through a high-pitched speaker, and the test begins; 3) obtaining the low-frequency microphone time history curve, the low-frequency acceleration time history curve, and the low-frequency laser scanning Doppler time history curve, and then calculating the root mean square deviation, similarity coefficient, and time domain amplitude; 4) determining whether the basic indicators obtained by different methods are consistent with the simulated benchmark values; if not, retesting; if so, performing non-contact measurement; 5) using a hammer-low-frequency microphone for impact-acoustic vibration acquisition, and using a hammer-low-frequency scanning laser Doppler for information acquisition, obtaining the stiffness error matrix, flexibility coefficient matrix, and power spectral density. 6) Initially determine whether hidden damage exists. If not, proceed to the next measurement point, statistically analyze the results to complete the on-site test, verify the finite element model, and output a data analysis test report. If damage exists, use contact measurement and a hammer-low-frequency triaxial velocities meter for impact-response acquisition to obtain wavelet transform energy values, energy proportions of each wavelet subband, and wavelet time-frequency diagrams to determine the damage location. Use a hammer-low-frequency uniaxial accelerometer for longitudinal measurement to obtain frequency shift amplitude, nonlinear phase, and loss factor to quantify hidden damage. Then proceed to the next measurement point. Finally, statistically analyze the results to complete the on-site test of the glass curtain wall, verify the finite element model, and output a data analysis test report.

[0114] Wavelet transform can provide high resolution in both time and frequency. In multi-sensor testing of curtain walls, especially in contact testing, it can be used to further improve resolution and provide more comprehensive structural status information. Its basic theoretical formula is:

[0115] Discrete wavelet transform formula: Where X[m, n] are the coefficients after wavelet transform, x[k] is the original signal, and ψ m，n [k] is the wavelet basis function.

[0116] The above formula can be further decomposed and reconstructed, where the decomposed formula is:

[0117] Where X[m, n] are detail coefficients, x[k] is the original signal, h[m] is the impulse response of the low-pass filter, and φ(t) is the scaling function.

[0118] Furthermore, the reconstruction formula is:

[0119] Where x[k] is the reconstructed signal and X[m, n] are detail coefficients.

[0120] In specific testing, optical cameras and hyperspectral cameras are first used to identify microcracks in the curtain wall glass and weather-resistant sealant. Since regulations require all glass curtain walls under construction to use tempered safety glass, which currently only exhibits spontaneous breakage, this testing system categorizes spontaneous breakage as a multi-crack condition. Thus, combined with the poor sealing problem caused by microcracks in the exposed weather-resistant sealant, both are classified as microcrack visual inspection issues. Based on this, optical cameras and deep learning algorithms are used to determine whether the glass exhibits a multi-crack condition to determine if spontaneous breakage has occurred, and whether it is close to a multi-crack condition to determine whether a force hammer test should be performed on the glass panel. If the glass is in a multi-crack condition or close to a multi-crack condition, the test is stopped, and the test is moved to the next test area. In addition, crack identification and shape integrity identification can be performed on exposed weather-resistant sealant on the exterior facade. The principle is the same as the glass multi-crack state test. The hyperspectral supplement is used in situations with large indoor and outdoor temperature differences and poor visibility, such as rainy days, cloudy days, and foggy days. By using thermal spectral sensitivity, the sealing defects of weather-resistant sealant can be quickly identified, thereby supplementing the limitations of optical camera application environments.

[0121] Secondly, after confirming that the panel is free of multiple cracks, the inspection robot will move to the designated test point to perform a force hammer excitation test. The force hammer uses a rubber / hard plastic hammer head, and different excitation forces are used according to the required modes at different excitation positions. The modal shape, modal damping, and modal frequency reflected by the working mode are compared with the simulated mode under different damage types (weatherproof adhesive damage, support structure defects). After modal testing, time-frequency analysis is performed. The robot returns to the designated test position, and the robotic arm places the low-frequency accelerometer and low-frequency microphone in the designated positions to prepare for testing. After the hammer strikes, the transmission characteristics of stress waves and acoustic-structure coupling characteristics at different positions are recorded. The detection point can be at the center of the glass panel to detect the overall working time-frequency state of the unit panel, or it can be detected between the weather-resistant sealant of two adjacent glass panels. The detection content is the waveguide characteristics generated by the plate as a waveguide. By comparing with actual glass curtain walls in service and specific damage tests and simulations, the precise identification of defect locations is obtained by considering factors such as dispersion characteristics, material property inversion, acoustic-structure characteristics, time-frequency amplitude changes, and energy attenuation. Microwave radar and miniature laser Doppler are used as supplementary testing methods to compare the test results of the low-frequency accelerometer and low-frequency microphone, achieving benchmark testing. That is, multiple methods are used to test the same object simultaneously to obtain the same results, ensuring the accuracy of the test results.

[0122] Finally, the data is transmitted via a 5G network. The obtained time-domain curves are then systematically analyzed on a high-performance workstation. The analysis results are used to identify defects in individual glass curtain walls and to qualitatively determine the health status of individual glass curtain walls. The identification results are then uploaded to a cloud platform to display the health status characteristics of the entire glass curtain wall envelope in the form of an image, thus providing relevant reference for government departments and other organizations.

[0123] Reference Figure 11 The root mean square deviation includes the root mean square deviation under no-damage condition, the root mean square deviation under unilateral damage condition, the root mean square deviation under bilateral damage condition, and the root mean square deviation under trilateral damage condition.

[0124] Reference Figure 12 The similarity coefficients include Pearson correlation coefficients under no-damage, unilateral-damage, bilateral-damage, and trilateral-damage conditions.

[0125] Reference Figures 13 to 16 The energy proportions of each wavelet subband include the energy coefficient distributions of each wavelet subband under the undamaged state, the energy coefficient distributions of each wavelet subband under the unilateral damage state, the energy coefficient distributions of each wavelet subband under the bilateral damage state, and the energy coefficient distributions of each wavelet subband under the trilateral damage state.

[0126] Reference Figures 17 to 20The wavelet time-frequency plots include wavelet time-frequency plots under no-damage conditions, wavelet time-frequency plots under unilateral damage conditions, wavelet time-frequency plots under bilateral damage conditions, and wavelet time-frequency plots under trilateral damage conditions.

[0127] Reference Figure 21 The second aspect of this application discloses an automated inspection system for glass curtain walls, the system comprising:

[0128] The module is configured to build a 3D model based on the actual information of the target glass curtain wall.

[0129] The first acquisition module is configured to perform finite element analysis on the three-dimensional model to obtain simulated time-frequency response information; wherein, the simulated time-frequency response information includes displacement-time time-domain data, velocity-time time-domain data, and acceleration-time time-domain data;

[0130] The second acquisition module is configured to obtain a detection strategy based on the simulated time-frequency response information; wherein, the detection strategy includes a time-frequency measured sensor selection strategy, a time-frequency measured detection point setting position, and a time-frequency measured robot walking path, and the time-frequency measured detection point setting position is a measurement point position that can excite all target modes;

[0131] The third acquisition module is configured to perform excitation testing based on the detection strategy to acquire actual time-frequency response information; wherein, the actual time-frequency response information includes the first wave arrival time, vibration attenuation amplitude, energy spectrum, and root mean square deviation of energy;

[0132] The analysis module is configured to obtain the defect type of the target glass curtain wall based on the actual time-frequency response information.

[0133] In the description of this specification, the references to terms such as "one embodiment / mode," "some embodiments / modes," "example," "specific example," or "some examples," etc., indicate that a specific feature, structure, material, or characteristic described in connection with that embodiment / mode or example is included in at least one embodiment / mode or example of this application. In this specification, the illustrative expressions of the above terms do not necessarily refer to the same embodiment / mode or example. Moreover, the specific features, structures, materials, or characteristics described may be combined in any suitable manner in one or more embodiments / modes or examples. Furthermore, without contradiction, those skilled in the art can combine and integrate the different embodiments / modes or examples described in this specification, as well as the features of different embodiments / modes or examples.

[0134] Furthermore, the terms "first" and "second" are used for descriptive purposes only and should not be construed as indicating or implying relative importance or implicitly specifying the number of technical features indicated. Thus, a feature defined as "first" or "second" may explicitly or implicitly include at least one of that feature. In the description of this application, "multiple" means at least two, such as two, three, etc., unless otherwise explicitly specified.

[0135] Those skilled in the art should understand that the above embodiments are merely for the purpose of clearly illustrating this disclosure, and are not intended to limit the scope of this disclosure.

Claims

1. An automated inspection method for glass curtain walls, characterized in that, The method includes: A three-dimensional model is constructed based on the actual information collected about the target glass curtain wall. Finite element analysis is performed on the three-dimensional model to obtain simulated time-frequency response information; wherein, the simulated time-frequency response information includes displacement-time time-domain data, velocity-time time-domain data, and acceleration-time time-domain data; Based on the simulated time-frequency response information, a detection strategy is obtained; wherein, the detection strategy includes a time-frequency measured sensor selection strategy, a time-frequency measured detection point setting position, and a time-frequency measured robot walking path, and the time-frequency measured detection point setting position is a measurement point position that can excite all target modes; Based on the detection strategy, an excitation test is performed to obtain actual time-frequency response information; wherein, the actual time-frequency response information includes the first wave arrival time, vibration attenuation amplitude, energy spectrum, and root mean square deviation of energy; Based on the actual time-frequency response information, the defect type of the target glass curtain wall is obtained.

2. The automated inspection method for glass curtain walls according to claim 1, characterized in that, The step of performing finite element analysis on the three-dimensional model to obtain simulated time-frequency response information includes: Based on the aforementioned three-dimensional model, a finite element model is established; The finite element model is analyzed based on preset parameters to obtain the simulated time-frequency response information; The preset parameters include the excitation point position and the data output point position; wherein, the data output point position is equipped with a displacement detection sensor, a velocity detection sensor and an acceleration detection sensor.

3. The automated inspection method for glass curtain walls according to claim 2, characterized in that, The step of performing stimulus testing based on the detection strategy to obtain actual time-frequency response information includes: Based on the time-frequency measurement sensor selection strategy, time-frequency measurement sensors are deployed at the time-frequency measurement detection point locations, and a high-precision fully automatic force hammer is used to excite the time-frequency measurement detection point locations to obtain the actual time-frequency response information.

4. The automated inspection method for glass curtain walls according to claim 3, characterized in that, The locations of the time-frequency measurement test points include the structural adhesive bonding area corresponding to the outer surface of the facade and the weather-resistant adhesive area between the two glass panels on the outer side of the facade.

5. The automated inspection method for glass curtain walls according to claim 3, characterized in that, The time-frequency measurement sensor selection strategy includes ensuring that the sensor's frequency testing range covers the target detection frequency and that the sensor's sensitivity matches the structural characteristics of the glass curtain wall.

6. The automated inspection method for glass curtain walls according to claim 5, characterized in that, The method of obtaining the defect type of the target glass curtain wall based on the actual time-frequency response information includes: The actual time-frequency response information is subjected to morphological analysis, threshold analysis, digital image analysis, and hyperspectral analysis to determine whether there is apparent damage; if so, the defect locations with apparent damage are re-measured in detail to obtain the defect type of the target glass curtain wall. If not, perform main frequency analysis, amplitude analysis, impact-response analysis, and impact-acoustic vibration analysis on the actual time-frequency response information to determine whether there is any hidden damage; if so, perform detailed re-measurement of the defect location with hidden damage to obtain the defect type of the target glass curtain wall. If not, the next measurement point will be automatically detected based on the robot's travel path as described in the time-frequency measurement.

7. The automated inspection method for glass curtain walls according to any one of claims 1-6, characterized in that, The target glass curtain wall is not subject to spontaneous explosion.

8. The automated inspection method for glass curtain walls according to any one of claims 1-6, characterized in that, The process before building the 3D model also includes: Based on the actual information of the target glass curtain wall, determine whether the target glass curtain wall has a preset crack state; If so, the target glass curtain wall is determined to be an abnormal area, and a new target area without the preset crack state is selected for detection. Based on the target area to be detected, an updated 3D model is constructed; The preset crack state is a multi-crack state; the abnormal region includes areas where spontaneous explosion is possible.

9. A flying robot facing a glass curtain wall, characterized in that, The flying robot for performing the excitation test in the automated inspection method for glass curtain walls according to any one of claims 1-8 includes a robot body and a multi-functional inspection device installed on the robot body. The robot body is equipped with a wireless transmission device, a vacuum adsorption device, a support bracket device, and a flight system. The multifunctional detection device includes an excitation source component, a time-domain voltage signal acquisition sensor component, a vibration data acquisition sensor component, and an image data acquisition sensor component; the excitation source component is mounted on the robot body via an automatic telescopic arm; the time-domain voltage signal acquisition sensor component is disposed on the periphery of the robot body; the vibration data acquisition sensor component is disposed on the top of the robot body; and the image data acquisition sensor component is disposed on the bottom of the robot body.

10. The flying robot facing a glass curtain wall according to claim 9, characterized in that, The excitation source assembly includes an automatic force hammer with an air damping unit; The time-domain voltage signal acquisition sensor assembly includes an accelerometer and a microphone sensor; The vibration data acquisition sensor assembly includes a Doppler laser and a microwave radar; The image data acquisition sensor assembly includes an industrial camera and a hyperspectral camera.

Images