An audio-visual fusion non-destructive testing robot for bridge cables or suspender

Abstract

The application discloses an audio-visual fusion nondestructive testing robot for bridge cable or suspender, and belongs to the technical field of nondestructive testing. The robot comprises a robot carrier, a machine vision detection module, an acoustic detection module and a central control module. The robot carrier comprises a caterpillar type clamping structure, which is used for supporting the movement operation of the robot on the bridge cable or suspender. The machine vision detection module comprises a high-definition camera and an image recognition unit, which are used for detecting the surface diseases on the bridge cable or suspender and outputting the detection result to the central control module. The acoustic detection module comprises an electromagnetic excitation assembly, a microphone array and a spectrum analysis unit, which are used for detecting the internal diseases on the bridge cable or suspender and outputting the detection result to the central control module. The central control module is used for controlling the movement of the robot, disease detection, fusing the surface disease result and the internal disease result, completing the depth analysis of the bridge cable or suspender and outputting the analysis result to a client.

Description

Technical Field

[0001] This invention belongs to the field of non-destructive testing technology, specifically relating to an audiovisual fusion non-destructive testing robot for bridge cables or hangers. Background Technology

[0002] Bridge cables / suspenders, as the "lifeline" of cable-stayed bridges and arch bridges, directly determine the structural safety and service life of the bridge. During long-term use, their outer PE sheaths are susceptible to various defects due to environmental factors such as wind and rain erosion, vehicle vibration, and sudden temperature changes. These defects include micro-cracks caused by material aging, localized damage from vehicle scrapes or construction collisions, surface mold due to prolonged high humidity, and pinpoint rust caused by the seepage of small amounts of corrosion products. While these external defects manifest as cosmetic damage, they compromise the first line of defense for the cables / suspenders, making it easier for subsequent corrosive media such as rainwater, salt, and dust to adhere or penetrate, accelerating the aging process of the sheath. If not inspected and repaired in time, the damage may expand, the protective function may completely fail, and ultimately indirectly threaten the overall structural safety of the cables / suspenders.

[0003] Traditional methods for inspecting the exterior of bridge cables / suspenders rely on manual inspection, requiring inspectors to climb the cables or use suspended platforms or aerial work platforms. This not only results in high operating costs and a significant risk of falls from heights, but also extremely low inspection efficiency. For example, inspecting a single 100-meter-long cable often requires 2-3 people working together for several hours. More importantly, manual inspection relies on visual observation, which is easily affected by lighting conditions (such as strong light reflection or dim lighting in rainy weather), perspective (such as difficulty in seeing the bottom of the sheath from a high vantage point), and the experience of the personnel. It is difficult to identify millimeter-level microcracks or hidden damage covered by dust, leading to large subjective judgment errors and a high rate of missed or misjudged defects.

[0004] To address the problems of traditional methods, existing technologies, combined with mature robotic processes, have led to the development of a cable-climbing robot capable of detecting surface defects. This robot utilizes machine vision to inspect bridge cables and hangers for defects. However, machine vision alone can only identify surface defects and cannot detect internal wire damage. Furthermore, it cannot combine internal and external damage to conduct in-depth analysis of bridge cable / hanger defects, thus failing to provide accurate data support for maintenance decisions. Summary of the Invention

[0005] In view of this, the purpose of the present invention is to provide an audiovisual fusion nondestructive testing robot for bridge cables or suspenders, in order to solve the above-mentioned technical problems.

[0006] To achieve the above objectives, the present invention provides the following technical solution: An audiovisual fusion non-destructive testing robot for bridge cables or suspenders includes a robot carrier, a machine vision inspection module, an acoustic inspection module, and a central control module. The robot carrier includes a tracked clamping structure to support the robot's movement on bridge cables or suspenders; the machine vision inspection module includes a high-definition camera and an image recognition unit to detect surface defects on bridge cables or suspenders and output the results to the central control module; the acoustic inspection module includes an electromagnetic excitation assembly, a microphone array, and a spectrum analysis unit to detect internal defects on bridge cables or suspenders and output the results to the central control module; the central control module controls the robot's movement, defect detection, and integrates surface and internal defect results to perform in-depth analysis of the bridge cables or suspenders and outputs the analysis results to the client.

[0007] Furthermore, the tracked clamping structure includes a track assembly, an elastic buffer, and a drive motor; The elastic buffer is used to adaptively adjust the clamping force according to the diameter of the bridge cable or gantry; the drive motor is used to drive the track assembly so that the robot moves axially along the cable or gantry.

[0008] Furthermore, the machine vision inspection module also includes a supplementary lighting component; a high-definition camera is arranged around the circumference of the robot carrier, and the supplementary lighting component is turned on and off synchronously with the high-definition camera. The image recognition unit is used to receive the cable body photos taken by the high-definition camera to complete the detection of surface defects on the bridge cables or suspenders. Once the robot reaches the target detection area, the central control module sends a first detection command to the machine vision detection module. Upon receiving the first detection command, the machine vision detection module performs the following operations: Control the high-definition camera to start real-time recording, and at the same time control the fill light component to perform light compensation operation on the recording area; The system acquires real-time camera images and sends them to the image recognition unit. The image recognition unit uses a pre-trained surface defect detection model to identify surface defects in the real-time camera images and sends the surface defect identification results to the central control module.

[0009] Furthermore, once the robot reaches the target detection area, the central control module sends a second detection command to the acoustic detection module. Upon receiving the second detection command, the acoustic detection module performs the following operations: The target detection area is automatically tapped by an electromagnetic excitation component, and the sound signal generated by the tapping is collected in real time by a microphone array and transmitted to the spectrum analysis unit. The spectrum analysis unit analyzes the signal characteristics of the sound signal through Fourier transform and compares it with the pre-collected internal defect signal characteristics. Based on the comparison results, it determines whether there are internal defects inside the cable body and sends the internal defect identification results to the central control module.

[0010] Furthermore, the in-depth analysis of bridge cables or hangers, which integrates surface and internal defect results, includes: Acquire several surface disease results and internal disease results, along with their corresponding coordinate markers. Based on the coordinate markers, obtain the surface disease results and internal disease results for the same detection point, which are denoted as the first result and the second result. Using a pre-set damage association rule base, the causal relationship between the first and second results of the same detection point is determined to obtain the first analysis result; if only surface disease or only internal disease exists for the same detection point, the current detection point is determined to be an isolated disease and there is no causal relationship between the disease and the disease. The first analysis result is defined as disease prevention, and disease prevention suggestions related to the first analysis result are obtained through online retrieval. When any detection point has surface defects or internal defects, it is determined whether there are similar surface defects or internal defects at its axially continuous detection points. If so, it is determined that the corresponding axially continuous detection points and the current detection point have defects in the same area, and the defect results are extracted to generate a set of defects in the same area. The second analysis result is obtained based on the set of defects in the same area. The set of defects in the same area includes all the first and second results of the current detection point and the corresponding axially continuous detection points. The second analysis result is defined as disease treatment, and disease treatment suggestions related to the second analysis result are obtained through online retrieval. By integrating disease prevention and control recommendations, we can obtain disease control recommendations. The analysis results are obtained by integrating the first analysis results, the second analysis results, and the disease prevention and control recommendations. The results of surface disease and internal disease are integrated to complete the in-depth analysis of bridge cables or hangers.

[0011] Furthermore, the machine vision inspection module and the acoustic inspection module are started simultaneously; The acoustic detection module also includes a GPS component, which is used to locate the impact position of the electromagnetic vibration component. When generating internal defect identification results, the system generates marker information based on the impact position to mark the coordinates of the internal defect identification results. Obtain the relative coordinate information of the fixed positioning and detection area of ​​the electromagnetic excitation component and the high-definition camera; When generating surface defect identification results, a second marker is generated based on the corresponding marker information and relative coordinate information to mark the coordinates of the surface defect identification results.

[0012] Furthermore, using a pre-set damage association rule base, the causal relationship between the first and second results at the same detection point is determined, including: A damage association rule base is pre-deployed; the damage association rule base contains a mapping logic between external structural damage and internal functional failure, as well as corresponding association conditions covering different types of external damage and types of internal failure; external structural damage and external damage type are related to the first result, and internal functional failure and internal failure type are related to the second result. During the detection process, the central control module automatically extracts the first and second results from the same detection point, calls the damage association rule library to compare the first and second results in multiple dimensions, and if the two meet the mapping logic and corresponding association conditions in the damage association rule library, it is determined that there is a causal relationship between the two; where the causal relationship between the two is the mapping logic and corresponding association conditions that meet the requirements in the damage association rule library.

[0013] Furthermore, the second analysis results obtained from the disease cluster in the same region include: Extract the disease type, disease severity, and spatial distribution information of all detection points in the disease set of the same area corresponding to the first and second results; wherein, the spatial distribution information is determined based on the coordinate markings of the first and second results; The overall axial spread range of diseases in the same region is determined by all detection points in the disease set in the same region, and the circumferential coverage area of ​​diseases in the same region is determined by combining spatial distribution information. The dominant disease type and its proportion are determined based on the disease types of all first and second results in the disease set of the same region; Based on the severity of disease in all first and second results in the same disease set, and the axial distribution order of the detection points, the axial development law of disease severity is determined. The overall axial spread range, circumferential coverage area, dominant disease types and proportions, and axial development patterns of disease severity were obtained and integrated to obtain the second analysis results.

[0014] The beneficial effects of this invention are as follows: 1. This invention, through the synchronous collaboration of a machine vision module and an acoustic detection module, achieves for the first time the simultaneous detection of "surface defects and internal defects" in bridge cables / suspenders, solving the pain points of traditional single vision which can only identify surface defects and single acoustic which cannot record surface conditions. Furthermore, by using a pre-set damage association rule library, it matches surface defect results with internal defect results based on coordinate markers at the same location, accurately determining the causal relationship between the two defects. Simultaneously, isolated defects are individually marked to avoid "missed causal identification" or "false association," upgrading defect diagnosis from "single defect identification" to "root cause tracing diagnosis," significantly improving detection accuracy. 2. The tracked clamping structure of the robot carrier provided by this invention is equipped with an elastic buffer, which can adaptively adjust the clamping force according to different diameter bridge cables / rods. Combined with the drive motor, it can achieve stable movement along the cable axis, eliminating the need for manual climbing or using a suspended platform, thus completely avoiding the risk of falling from heights and reducing labor costs. In addition, the central control module can automatically trigger the detection process, and the GPS component and relative coordinate positioning technology ensure accurate marking of detection points. The synchronous start and stop of the supplementary lighting component and electromagnetic vibration further reduce manual intervention and further improve the level of intelligent detection. 3. The robot provided by this invention can not only perform disease detection, but also generate multi-dimensional second analysis results based on the disease set in the same area. At the same time, it combines the first analysis results to output "disease prevention suggestions" and "disease treatment suggestions", and finally integrates them into a complete prevention and control plan. This avoids the limitation of traditional detection that only outputs "disease exists" without operation and maintenance guidance. It directly provides bridge operation and maintenance personnel with full-chain support of "detection-analysis-decision", reducing the waste of resources caused by blind maintenance.

[0015] Other advantages, objectives, and features of the invention will be set forth in the following description and will be apparent to those skilled in the art in some respects, or may be learned by practice of the invention. The objects and other advantages of the invention may be realized and obtained by means of the structures particularly pointed out in the written description and the accompanying drawings.

[0016] The technical solution of the present invention will be further described in detail below with reference to the accompanying drawings and embodiments. Attached Figure Description

[0017] The accompanying drawings are provided to further illustrate the invention and form part of the specification. They are used in conjunction with embodiments of the invention to explain the invention and do not constitute a limitation thereof. In the drawings: Figure 1 This is a schematic diagram of a module of an audiovisual fusion nondestructive testing robot for bridge cables or suspenders, according to an embodiment of the present invention. Detailed Implementation

[0018] The preferred embodiments of the present invention will be described below with reference to the accompanying drawings. It should be understood that the preferred embodiments described herein are for illustration and explanation only and are not intended to limit the present invention.

[0019] like Figure 1 As shown, this invention proposes an audiovisual fusion non-destructive testing robot for bridge cables or suspenders, including a robot carrier, a machine vision inspection module, an acoustic inspection module, and a central control module; The robot carrier includes a tracked clamping structure to support the robot's movement on bridge cables or suspenders; the machine vision inspection module includes a high-definition camera and an image recognition unit to detect surface defects on bridge cables or suspenders and output the results to the central control module; the acoustic inspection module includes an electromagnetic excitation assembly, a microphone array, and a spectrum analysis unit to detect internal defects on bridge cables or suspenders and output the results to the central control module; the central control module controls the robot's movement, defect detection, and integrates surface and internal defect results to perform in-depth analysis of the bridge cables or suspenders, outputting the analysis results to the client. The working principle and beneficial effects of the above technical solution are as follows: This invention achieves the detection function through the coordinated operation of four core modules; the tracked clamping structure of the robot carrier provides overall support, the track assembly fits the surface of the cable / rod, and the elastic buffer can automatically adjust the clamping force according to the cable diameter to avoid slippage due to excessive looseness or damage to the cable due to excessive tightness; the drive motor drives the track assembly to rotate, moving the robot smoothly along the cable axis to the target detection area; in the machine vision detection module, a high-definition camera acquires images of the cable surface, and the image recognition unit identifies surface defects such as PE sheath cracks, damage, and rust seepage through a preset surface defect detection algorithm (such as feature extraction and classification model), and transmits the surface defect results (first result) to the central control module; in the acoustic detection module, the electromagnetic vibration assembly automatically strikes the cable, the microphone array collects the sound signal generated by the strike in real time, and the spectrum analysis unit uses Fourier transform to convert the sound signal into spectrum features, which are then compared with pre-stored... Internal defects (such as wire breakage and corrosion) signal characteristics are compared to determine the internal state and generate internal defect results (second results), which are also transmitted to the central control module. The central control module, as the core, controls the robot's movement speed and the start and stop sequence of the detection module. It also receives the two types of detection results, performs in-depth analysis through data fusion algorithms, and finally outputs the analysis results to the client. Through the above technical solution, traditional methods such as manual climbing and suspended platform operations are completely replaced, significantly reducing the risk of falls from heights for inspection personnel and reducing labor costs. The integration of machine vision and acoustic technology overcomes the technical limitations of single vision only being able to detect surfaces and single acoustic technology not being able to record surfaces, enabling simultaneous detection of internal and external defects of cables / suspenders. The collaborative control and in-depth analysis functions of the central control module upgrade the detection from "simple defect identification" to "full-dimensional diagnosis," meeting the core requirements of bridge structures for automated and comprehensive detection of cables / suspenders, and improving detection efficiency and accuracy.

[0020] In one embodiment, the tracked clamping structure includes a track assembly, an elastic buffer, and a drive motor; The elastic buffer is used to adaptively adjust the clamping force according to the diameter of the bridge cable or gantry; the drive motor is used to drive the track assembly so that the robot moves axially along the cable of the bridge cable or gantry. The working principle of the above technical solution is as follows: This solution is an optimized design for the tracked clamping structure of the robot carrier; the tracked clamping structure includes three core components: track assembly, elastic buffer, and drive motor; the elastic buffer is made of spring or elastic rubber, with one end connected to the track assembly and the other end fixed to the carrier body. When the robot is attached to cables / rods of different diameters, the elastic buffer adaptively adjusts the contact pressure between the track assembly and the cable through its own deformation to ensure appropriate clamping force; the drive motor is a servo motor, connected to the transmission mechanism of the track assembly. The central control module can send speed control signals to the drive motor, and the drive motor adjusts the output speed according to the signal, thereby driving the track assembly to rotate at a set speed, realizing the movement of the robot along the cable axis; the speed can be reduced for fine inspection and increased for rapid inspection; the entire structure does not require manual adjustment of clamping force or movement speed, and is fully adapted through mechanical self-adaptation and electronic control; The beneficial effects of the above technical solution are as follows: the adaptive design of the elastic buffer component allows the robot to be compatible with cables / rods of different diameters (such as the common 50-300mm diameter range), eliminating the need to customize special carriers for different cables and greatly improving the equipment's versatility; the controllable speed function of the drive motor allows for flexible switching of working modes according to testing needs, ensuring both high data acquisition density during fine testing and improved efficiency during rapid inspection, balancing testing accuracy and operational efficiency; compared to traditional fixed clamping structures, the adaptive clamping design avoids the tedious operation of manual adjustment, while preventing excessive tightness from damaging the PE sheath of the cable, extending the cable's service life and ensuring the safety of equipment testing.

[0021] In one embodiment, the machine vision inspection module further includes a supplementary lighting component; a high-definition camera is arranged around the robot carrier, the supplementary lighting component is started and stopped synchronously with the high-definition camera, and the image recognition unit is used to receive cable photos taken by the high-definition camera to complete the detection of surface defects on the bridge cables or suspenders; Once the robot reaches the target detection area, the central control module sends a first detection command to the machine vision detection module. Upon receiving the first detection command, the machine vision detection module performs the following operations: Control the high-definition camera to start real-time recording, and at the same time control the fill light component to perform light compensation operation on the recording area; The system acquires real-time camera images and sends them to the image recognition unit. The image recognition unit uses a pre-trained surface defect detection model to identify surface defects in the real-time camera images and sends the surface defect identification results to the central control module. The working principle of the above technical solution is as follows: In addition to the basic high-definition camera and image recognition unit, the machine vision inspection module adds a supplementary lighting component, and the high-definition cameras are evenly arranged around the circumference of the robot carrier (e.g., 4 cameras) to ensure that the entire circumference of the cable body is covered without blind spots; when the robot reaches the target detection area, the central control module sends the first detection command to the module: after the command is triggered, the high-definition camera immediately starts real-time imaging and synchronously controls the start and stop of the supplementary lighting component - the supplementary lighting component uses an LED light source, which can automatically adjust the brightness according to the ambient light intensity (e.g., increase the brightness when the light is insufficient and turn it off when the light is sufficient), to avoid strong light reflection or dim light causing image blurring; the high-definition camera transmits the collected surface image to the image recognition unit in real time, and the image recognition unit calls the pre-trained surface defect detection model (e.g., a classification model based on a convolutional neural network) to identify and locate features such as cracks, damage, and rust in the image, generate a surface defect result (first result) containing the defect type and location, and feed it back to the central control module; The beneficial effects of the above technical solution are as follows: the circumferential arrangement design of the high-definition camera completely solves the problem of blind spots caused by the cylindrical structure of the cable body in traditional single-view cameras, ensuring that no surface of the cable body is missed in detection; the synchronous start-stop and brightness adaptive function of the supplementary lighting component effectively offsets the impact of changes in ambient light on image quality, ensuring that clear images can still be obtained in complex environments such as cloudy days, strong light, and backlight, thus improving the accuracy of surface defect identification; the command-triggered process of the central control module realizes the automated control of visual inspection without the need for manual intervention to start, stop, or adjust parameters. Combined with the automatic recognition function of the image recognition unit, it greatly reduces the error of subjective judgment by humans and improves the standardization and efficiency of surface inspection.

[0022] In one embodiment, when the robot reaches the target detection area, the central control module sends a second detection command to the acoustic detection module. Upon receiving the second detection command, the acoustic detection module performs the following operations: The target detection area is automatically tapped by an electromagnetic excitation component, and the sound signal generated by the tapping is collected in real time by a microphone array and transmitted to the spectrum analysis unit. The spectrum analysis unit analyzes the signal characteristics of the sound signal through Fourier transform and compares it with the pre-collected internal defect signal characteristics. Based on the comparison results, it determines whether there are internal defects inside the cable body and sends the internal defect identification results to the central control module. The working principle of the above technical solution is as follows: When the robot arrives at the target detection area, the central control module sends a second detection command to the acoustic detection module. After the command is triggered, the electromagnetic vibration component of the acoustic detection module is activated first, and automatically taps the surface of the cable body at a preset frequency (e.g., 10-1000Hz), simulating manual tapping but with more stable force and frequency. The microphone array next to the electromagnetic vibration component (e.g., 6-8 microphones arranged in a ring) collects the sound signal generated by the tapping in real time. During the collection process, the microphone array reduces environmental noise interference through spatial filtering technology. The collected sound signal is transmitted to the spectrum analysis unit, which uses the Fourier transform algorithm to convert the time-domain sound signal into frequency-domain spectrum features and extracts key parameters such as frequency peak and attenuation rate. Then, these parameters are compared with the acoustic feature parameters of the "cable body without internal defects" collected in advance. If the parameter deviation exceeds the preset threshold (e.g., the frequency peak is lower than the normal range, or the attenuation rate is faster than the normal range), it is determined that there are defects such as fracture and corrosion inside the cable body. An internal defect result (second result) containing the defect type and severity is generated and transmitted to the central control module. The beneficial effects of the above technical solution are as follows: the automatic tapping of the electromagnetic excitation component replaces manual tapping, avoiding detection errors caused by uneven tapping force and unstable frequency, while significantly improving internal detection efficiency (single-point detection time is reduced to the second level); the noise suppression of the microphone array and the Fourier transform processing of the spectrum analysis unit enhance the anti-interference ability and feature extraction accuracy of the sound signal, solving the problems of strong subjective judgment and susceptibility to environmental noise in traditional manual listening; the entire acoustic detection process is triggered by instructions from the central control module, realizing automated operation without the need for manual handheld equipment detection, adapting to high-altitude and long-distance detection scenarios of cables / rods, and improving the reliability and applicability of internal defect detection.

[0023] In one embodiment, fusing surface defect results and internal defect results to complete an in-depth analysis of bridge cables or hangers includes: Acquire several surface disease results and internal disease results, along with their corresponding coordinate markers. Based on the coordinate markers, obtain the surface disease results and internal disease results for the same detection point, which are denoted as the first result and the second result. Using a pre-set damage association rule base, the causal relationship between the first and second results of the same detection point is determined to obtain the first analysis result; if only surface disease or only internal disease exists for the same detection point, the current detection point is determined to be an isolated disease and there is no causal relationship between the disease and the disease. The first analysis result is defined as disease prevention, and disease prevention suggestions related to the first analysis result are obtained through online retrieval. When any detection point has surface defects or internal defects, it is determined whether there are similar surface defects or internal defects at its axially continuous detection points. If so, it is determined that the corresponding axially continuous detection points and the current detection point have defects in the same area, and the defect results are extracted to generate a set of defects in the same area. The second analysis result is obtained based on the set of defects in the same area. The set of defects in the same area includes all the first and second results of the current detection point and the corresponding axially continuous detection points. The second analysis result is defined as disease treatment, and disease treatment suggestions related to the second analysis result are obtained through online retrieval. By integrating disease prevention and control recommendations, we can obtain disease control recommendations. The analysis results are obtained by integrating the first analysis results, the second analysis results, and the disease prevention and control recommendations. The analysis results are then integrated with the surface disease results and the internal disease results to complete the in-depth analysis of the bridge cables or hangers. The working principle of the above technical solution is as follows: The core of this solution is the deep analysis function of the central control module, which is divided into three major steps: First, coordinate matching and result association. The central control module obtains the surface disease results (first result) and internal disease results (second result) with coordinate markings. Based on the coordinate markings, it filters out the two types of results at the same detection point. It ensures that the two types of results accurately correspond to the same physical location, laying a spatial consistency foundation for subsequent causal analysis. For the first result or second result that does not match the corresponding result (such as only surface cracks but no internal abnormalities), it is temporarily stored as "isolated disease to be determined" and left for separate marking later. The second step is to generate the first analysis result and call the pre-set damage association rule base (built-in mapping logic between external damage and internal failure). If two types of results exist at the same location, compare whether they meet the association conditions in the rule base. If they do, determine the causal relationship of the disease (such as surface damage leading to internal corrosion), define it as "disease prevention", and retrieve prevention suggestions. If only one type of result exists, mark it as an isolated disease. Step 3: The second analysis results are generated. The detection points with diseases are screened, and it is determined whether there are similar diseases in the continuous axial points. If so, all the results of these points are extracted to generate a "set of diseases in the same area". Step 4: Final result integration. The central control module integrates the first analysis results (cause and effect relationship of the disease + prevention recommendations) and the second analysis results (spread range / dominant type / development pattern of the disease in the same area + treatment recommendations) in a "structured document" format (such as JSON or PDF). This ensures that each analysis conclusion is accompanied by corresponding detection data (such as original images and spectrum diagrams of cause and effect relationships, coordinate distribution maps of the same area analysis). Finally, the data is transmitted to the client (bridge operation and maintenance monitoring platform) via wireless communication for operation and maintenance personnel to directly access and make decisions. It is worth noting that when conducting network searches, the central control module preferentially accesses bridge operation and maintenance knowledge bases (such as the national bridge inspection database and industry maintenance standard document library) via 4G / 5G wireless communication interfaces, and searches using "cause and effect relationship of defects" as keywords (such as "sheath damage leads to steel wire corrosion") to automatically extract matching defect prevention suggestions (such as "regularly (every 3 months) check the integrity of the PE sheath and repair damage with an area ≥5mm² in a timely manner" and "apply weather-resistant anti-corrosion coating to areas with high incidence of sheath damage") and associates them with the first analysis results; Similarly, the disease treatment based on the second analysis results is retrieved using the above method, which will not be elaborated here; The beneficial effects of the above technical solution are as follows: by using coordinate matching and rule base judgment, the causal relationship between internal and external defects can be realized, solving the problem that traditional detection can only identify internal and external defects individually and cannot trace the root cause of the defects; the analysis function of the defect set in the same area clarifies the spatial distribution and development trend of the defects, avoiding the limitation of "single point detection" in being unable to determine the spread range of the defects; the output of integrated prevention and control suggestions allows the detection results to be directly transformed into implementable operation and maintenance solutions, breaking through the limitation of traditional detection "only reporting the disease, not treating the disease", providing bridge operation and maintenance personnel with full-chain support of "diagnosis + solution", and reducing the waste of resources caused by blind maintenance.

[0024] In one embodiment, the machine vision detection module and the acoustic detection module are started simultaneously. The acoustic detection module also includes a GPS component, which is used to locate the impact position of the electromagnetic vibration component. When generating internal defect identification results, the system generates marker information based on the impact position to mark the coordinates of the internal defect identification results. Obtain the relative coordinate information of the fixed positioning and detection area of ​​the electromagnetic excitation component and the high-definition camera; When generating surface defect identification results, second marker information is generated based on the corresponding marker information and relative coordinate information to mark the coordinates of the surface defect identification results. The working principle of the above technical solution is as follows: This solution achieves synchronous collaboration and precise coordinate marking between the machine vision and acoustic detection modules. First, when the central control module sends a detection command, it simultaneously triggers the machine vision and acoustic detection modules to start, ensuring that the two types of detection are carried out in the same time period and the same detection area, avoiding misalignment of points due to time sequence differences. Second, the GPS component of the acoustic detection module locates the striking position of the electromagnetic excitation component in real time, generates a coordinate mark at that position, and directly associates it with the internal defect results (second result). At the same time, the fixed relative coordinates between the electromagnetic excitation component and the high-definition camera are pre-measured and stored (e.g., the camera is at 0° and 90° circumferential positions of the excitation component, 5cm away from the excitation point). When the high-definition camera acquires a surface image, it calls the GPS coordinates of the excitation component, combines them with the relative coordinates to calculate the coordinates of the camera acquisition point, generates second mark information, and associates it with the surface defect results (first result). Through synchronous startup and relative coordinate conversion, it ensures that the coordinate marks of the two types of detection results correspond accurately. The beneficial effects of the above technical solution are as follows: synchronous start-up design ensures the spatiotemporal consistency of visual and acoustic detection, avoiding the problem of "non-overlapping detection areas" caused by asynchronous module start-up and shutdown, and providing a basis for subsequent correlation of results at the same location; the combination of GPS positioning and relative coordinate conversion solves the problem of independent visual and acoustic point markings and inaccurate matching in traditional detection, greatly improving the coordinate correspondence accuracy of the two types of results and reducing misjudgments of causal relationships caused by coordinate misalignment; accurate coordinate marking also provides a reliable basis for the subsequent delineation of the range of disease clusters in the same area, ensuring the accuracy of disease spread range analysis and improving the reliability of overall detection and analysis.

[0025] In one embodiment, using a pre-set damage association rule base, determining the causal relationship between the first and second results at the same detection point includes: A damage association rule base is pre-deployed; the damage association rule base contains a mapping logic between external structural damage and internal functional failure, as well as corresponding association conditions covering different types of external damage and types of internal failure; external structural damage and external damage type are related to the first result, and internal functional failure and internal failure type are related to the second result. During the detection process, the central control module automatically extracts the first and second results from the same detection point, calls the damage association rule library to compare the first and second results from multiple dimensions, and if the two meet the mapping logic and corresponding association conditions in the damage association rule library, it is determined that there is a causal relationship between the two; where the causal relationship between the two is the mapping logic and corresponding association conditions that meet the requirements in the damage association rule library. The working principle of the above technical solution is as follows: The damage association rule base is pre-built with two types of core content: one is the mapping logic between external structural damage (related to the first result, such as PE sheath damage and rust seepage) and internal functional failure (related to the second result, such as steel wire corrosion and breakage) (e.g., "sheath damage → steel wire corrosion"); the other is the specific association conditions for different damage types (e.g., "sheath damage area > 10mm² and the corresponding area steel wire corrosion characteristic signal deviation > 15%"). After the central control module extracts the first result and the second result of the same detection point, it calls the rule base for multi-dimensional comparison: first, it matches whether the mapping logic between external damage and internal failure corresponds, and then verifies whether the association conditions under the mapping logic are met. If both are met, it is determined that there is a causal relationship between the two types of results (the causal relationship is the matched mapping logic and association conditions). If either condition is not met, it is determined that there is no causal relationship between the two types of results. The beneficial effects of the above technical solution are as follows: the preset damage association rule base standardizes the "causal relationship between internal and external defects", avoids subjective differences when manually determining causality, ensures consistent causal determination results under different detection scenarios and different operators, and improves the standardization of analysis; the multi-dimensional comparison logic (mapping logic + association conditions) avoids misjudgment caused by single condition matching (such as judging internal corrosion only because of surface damage, ignoring the matching between the degree of damage and the internal signal deviation), and greatly improves the accuracy of causal relationship determination; the built-in design of the rule base facilitates subsequent updates and optimizations based on different bridge types and cable materials, improving the adaptability and scalability of the technical solution.

[0026] In one embodiment, obtaining a second analysis result based on a set of diseases in the same region includes: Extract the disease type, disease severity, and spatial distribution information of all detection points in the disease set of the same area corresponding to the first and second results; wherein, the spatial distribution information is determined based on the coordinate markings of the first and second results; The overall axial spread range of diseases in the same region is determined by all detection points in the disease set in the same region, and the circumferential coverage area of ​​diseases in the same region is determined by combining spatial distribution information. The dominant disease type and its proportion are determined based on the disease types of all first and second results in the disease set of the same region; Based on the severity of disease in all first and second results in the same disease set, and the axial distribution order of the detection points, the axial development law of disease severity is determined. The overall axial spread range, circumferential coverage area, dominant disease types and proportions, and axial development patterns of disease severity were obtained and integrated to obtain the second analysis results. The working principle of the above technical solution is as follows: This solution elaborates on the generation process of the second analysis result, based on the disease set in the same region: First step, information extraction: Extract the first and second results of all detection points from the disease set in the same region, obtain the disease type (such as surface cracks, internal fractures) and disease degree (such as crack length, corrosion area) of each result, and determine the spatial distribution information of each point according to the coordinate markings; Second step, range analysis: Combine the coordinate distribution of all points to determine the starting point and ending point of the spread of the disease in the cable body axis (overall axial spread range), and the coverage angle in the circumferential direction of the cable body (circumferential coverage area); Third step, type and trend analysis: Count the number of all disease types in the set, determine the dominant disease type with the highest proportion (such as 70% surface damage), and compare the disease degree according to the axial distribution order of the points to determine whether the disease is gradually aggravated, alleviated, or stabilized (the axial development law of disease degree); Fourth step, result integration: Integrate the overall axial spread range, circumferential coverage area, dominant disease type and proportion, and the axial development law of disease degree to form the second analysis result; The beneficial effects of the above technical solutions are as follows: Analyzing the spread of diseases through spatial distribution information solves the problem that traditional single-point detection cannot determine the boundary of disease impact, providing a basis for maintenance personnel to delineate key treatment areas; statistical analysis of dominant disease types helps maintenance personnel identify core diseases within the same area, avoiding "average effort" in treatment and improving the targeting of treatment; judging the axial development law of disease severity allows tracing the direction and speed of disease spread, providing data support for predicting disease development trends and formulating subsequent detection cycles; the overall second analysis results describe diseases in the same area from three dimensions: "spatial range + type + trend," making maintenance decisions more scientific and avoiding blind treatment or omission of key areas.

[0027] Finally, it should be noted that the above preferred embodiments are only used to illustrate the technical solutions of the present invention and are not intended to limit it. Although the present invention has been described in detail through the above preferred embodiments, those skilled in the art should understand that various changes can be made to it in form and detail without departing from the scope defined by the claims of the present invention.

Claims

1. An audio-visual fusion non-destructive testing robot for a bridge cable or a suspender, characterized in that, It includes a robot carrier, a machine vision inspection module, an acoustic inspection module, and a central control module; The robot carrier includes a tracked clamping structure to support the robot's movement on bridge cables or suspenders; the machine vision inspection module includes a high-definition camera and an image recognition unit to detect surface defects on bridge cables or suspenders and output the results to the central control module; the acoustic inspection module includes an electromagnetic excitation assembly, a microphone array, and a spectrum analysis unit to detect internal defects on bridge cables or suspenders and output the results to the central control module; the central control module controls the robot's movement, defect detection, and integrates surface and internal defect results to perform in-depth analysis of the bridge cables or suspenders and outputs the analysis results to the client.

2. The audiovisual fusion nondestructive testing robot for bridge cables or suspenders according to claim 1, characterized in that, The tracked clamping structure includes a track assembly, an elastic buffer, and a drive motor; The elastic buffer is used to adaptively adjust the clamping force according to the diameter of the bridge cable or gantry; the drive motor is used to drive the track assembly so that the robot moves axially along the cable or gantry.

3. The audiovisual fusion nondestructive testing robot for bridge cables or suspenders according to claim 1, characterized in that, The machine vision inspection module also includes a supplementary lighting component; a high-definition camera is arranged around the circumference of the robot carrier, and the supplementary lighting component starts and stops synchronously with the high-definition camera. The image recognition unit is used to receive the cable photos taken by the high-definition camera to complete the detection of surface defects on the bridge cables or suspenders. Once the robot reaches the target detection area, the central control module sends a first detection command to the machine vision detection module. Upon receiving the first detection command, the machine vision detection module performs the following operations: Control the high-definition camera to start real-time recording, and at the same time control the fill light component to perform light compensation operation on the recording area; The system acquires real-time camera images and sends them to the image recognition unit. The image recognition unit uses a pre-trained surface defect detection model to identify surface defects in the real-time camera images and sends the surface defect identification results to the central control module.

4. A non-destructive testing robot for bridge cables or suspenders according to claim 1 or 3, characterized in that, Once the robot reaches the target detection area, the central control module sends a second detection command to the acoustic detection module. Upon receiving the second detection command, the acoustic detection module performs the following operations: The target detection area is automatically tapped by an electromagnetic excitation component, and the sound signal generated by the tapping is collected in real time by a microphone array and transmitted to the spectrum analysis unit. The spectrum analysis unit analyzes the signal characteristics of the sound signal through Fourier transform and compares it with the pre-collected internal defect signal characteristics. Based on the comparison results, it determines whether there are internal defects inside the cable body and sends the internal defect identification results to the central control module.

5. The audiovisual fusion nondestructive testing robot for bridge cables or suspenders according to claim 1, characterized in that, A comprehensive analysis of bridge cables or hangers, integrating surface and internal defect findings, includes: Acquire several surface disease results and internal disease results, along with their corresponding coordinate markers. Based on the coordinate markers, obtain the surface disease results and internal disease results for the same detection point, which are denoted as the first result and the second result. Using a pre-set damage association rule base, the causal relationship between the first and second results of the same detection point is determined to obtain the first analysis result; if only surface disease or only internal disease exists for the same detection point, the current detection point is determined to be an isolated disease and there is no causal relationship between the disease and the disease. The first analysis result is defined as disease prevention, and disease prevention suggestions related to the first analysis result are obtained through online retrieval. When any detection point has surface defects or internal defects, it is determined whether there are similar surface defects or internal defects at its axially continuous detection points. If so, it is determined that the corresponding axially continuous detection points and the current detection point have defects in the same area, and the defect results are extracted to generate a set of defects in the same area. The second analysis result is obtained based on the set of defects in the same area. The set of defects in the same area includes all the first and second results of the current detection point and the corresponding axially continuous detection points. The second analysis result is defined as disease treatment, and disease treatment suggestions related to the second analysis result are obtained through online retrieval. By integrating disease prevention and control recommendations, we can obtain disease control recommendations. The analysis results are obtained by integrating the first analysis results, the second analysis results, and the disease prevention and control recommendations. The results of surface disease and internal disease are integrated to complete the in-depth analysis of bridge cables or hangers.

6. The audiovisual fusion nondestructive testing robot for bridge cables or suspenders according to claim 5, characterized in that, The machine vision inspection module and the acoustic inspection module start up simultaneously; The acoustic detection module also includes a GPS component, which is used to locate the impact position of the electromagnetic vibration component. When generating internal defect identification results, the system generates marker information based on the impact position to mark the coordinates of the internal defect identification results. Obtain the relative coordinate information of the fixed positioning and detection area of ​​the electromagnetic excitation component and the high-definition camera; When generating surface defect identification results, a second marker is generated based on the corresponding marker information and relative coordinate information to mark the coordinates of the surface defect identification results.

7. The audiovisual fusion nondestructive testing robot for bridge cables or suspenders according to claim 5, characterized in that, Using a pre-set damage association rule base, the causal relationship between the first and second results at the same detection point is determined, including: A damage association rule base is pre-deployed; the damage association rule base contains a mapping logic between external structural damage and internal functional failure, as well as corresponding association conditions covering different types of external damage and types of internal failure; external structural damage and external damage type are related to the first result, and internal functional failure and internal failure type are related to the second result. During the detection process, the central control module automatically extracts the first and second results from the same detection point, calls the damage association rule library to compare the first and second results in multiple dimensions, and if the two meet the mapping logic and corresponding association conditions in the damage association rule library, it is determined that there is a causal relationship between the two; where the causal relationship between the two is the mapping logic and corresponding association conditions that meet the requirements in the damage association rule library.

8. The audiovisual fusion nondestructive testing robot for bridge cables or suspenders according to claim 5, characterized in that, The second analysis results, based on the disease set in the same region, include: Extract the disease type, disease severity, and spatial distribution information of all detection points in the disease set of the same area corresponding to the first and second results; wherein, the spatial distribution information is determined based on the coordinate markings of the first and second results; The overall axial spread range of diseases in the same region is determined by all detection points in the disease set in the same region, and the circumferential coverage area of ​​diseases in the same region is determined by combining spatial distribution information. The dominant disease type and its proportion are determined based on the disease types of all first and second results in the disease set of the same region; Based on the severity of disease in all first and second results in the same disease set, and the axial distribution order of the detection points, the axial development law of disease severity is determined. The overall axial spread range, circumferential coverage area, dominant disease types and proportions, and axial development patterns of disease severity were obtained and integrated to obtain the second analysis results.

Images