Guided wave-magnetic flux leakage composite intelligent detection equipment and method for bridge cable damage detection

Abstract

The invention relates to the field of bridge detection and structure health monitoring, and provides guided wave-magnetic flux leakage composite intelligent detection equipment and method for bridge cable damage detection. The equipment adopts a split type modular design and is composed of a climbing module, a data acquisition module and a calculation control module. The climbing module realizes stable climbing of the whole equipment on the surface of the cable; the data acquisition module integrates a magnetic flux leakage detection unit and a magnetostrictive guided wave detection unit to form a composite detection device; a data acquisition card, an embedded processor and the like are embedded in the calculation control module, and the calculation control module is used for signal analysis and processing. The method solves the problems of anchoring area detection blind areas, insufficient detection precision, low data processing efficiency and the like, and can be widely applied to daily detection and maintenance of cable bearing bridges such as cable-stayed bridges, suspension bridges and the like.

Description

Technical Field

[0001] This invention relates to the field of bridge engineering and structural health monitoring technology, specifically to a damage detection device suitable for cable-stayed bridges, suspension bridges and other cable-bearing structures, and in particular to a composite detection device combining guided wave and magnetic flux leakage nondestructive testing technologies and its intelligent evaluation method. Background Technology

[0002] With the rapid development of my country's transportation infrastructure, the number of long-span bridges is increasing daily. As a core load-bearing component of bridges, the health of the cables directly affects the overall safety of the bridge. Cables are constantly under high stress and subjected to multiple factors such as wind vibration, vehicle loads, and environmental corrosion, making them prone to corrosion, wire breakage, and other damage. If these damages are not detected and addressed in a timely manner, they may lead to sudden breakage accidents, causing significant casualties and economic losses.

[0003] Currently, bridge cable inspection mainly employs methods such as manual visual inspection, ultrasonic testing, magnetostrictive guided wave testing, and magnetic flux leakage testing. However, existing technologies have the following significant shortcomings: (1) Single detection method and limited coverage: Although magnetic flux leakage testing technology has high detection accuracy and can accurately locate local defects, it cannot effectively detect hidden parts such as anchorage areas due to its working principle limitations; although magnetostrictive guided wave testing technology can detect anchorage areas and is sensitive to corrosion damage, its detection accuracy and defect location capabilities need to be improved. Currently, both technologies are used in the field of cable inspection, but there are few studies on composite detection that organically combine the two and give full play to their respective advantages. (2) Signal quality is greatly affected by the field environment: During cable inspection, the sensor is lifted to a large height, resulting in weak magnetic flux leakage and guided wave signals, which are easily submerged in field noise and clutter signals, making it difficult to collect high-quality data. (3) Low data processing efficiency: Existing detection technologies acquire a large amount of raw data, and manual interpretation is time-consuming and laborious. There is a lack of effective algorithms to convert the raw signals into intuitive damage images and assessment results, and the detection efficiency is difficult to meet the needs of large-scale bridge maintenance.

[0004] Therefore, there is an urgent need for a new cable inspection technology that can overcome the above-mentioned defects. Summary of the Invention

[0005] This invention provides a guided wave-magnetic flux leakage composite intelligent detection device and method for bridge cable damage detection. The device combines the technical advantages of guided wave detection and magnetic flux leakage detection, realizes automatic climbing detection through a modular robotic detection platform, and introduces machine learning algorithms to intelligently process the detection signals. It effectively solves the problems of limited applicability, poor signal quality, and low data processing efficiency of single detection methods, and has the advantages of high detection accuracy, wide coverage, and high degree of intelligence.

[0006] In a first aspect, the present invention provides a guided wave-magnetic flux leakage composite intelligent detection device for bridge cable damage detection. The device includes a climbing module, a data acquisition module, and a computing control module. The computing control module embeds a signal processing device and a signal generator. The signal processing device includes a data acquisition card and an embedded processor, electrically connected to the data acquisition module via a connecting cable. The data acquisition module includes a magnetic flux leakage detection unit, a magnetostrictive guided wave detection unit, and a preamplifier circuit fixed to the magnetic flux leakage detection unit. The magnetic flux leakage detection unit includes a permanent magnet, an armature, and permalloy forming an excitation circuit, and a Hall element disposed between the excitation circuit and the target cable. Multiple magnetic flux leakage detection units are connected in a ring array around the target cable via a module connection device. The magnetostrictive guided wave detection unit includes a detection coil disposed below the Hall element, surrounding the outer edge of the target cable. The coil is connected to the signal generator in the computing control module and the preamplifier circuit in the data acquisition module via a transmit / receive switching circuit. The climbing module includes a pulley and a motor fixedly connected to the data acquisition module, used to drive the modular cable detection device to climb axially on the surface of the target cable.

[0007] As an optional solution, the armature is long and strip-shaped, axially arranged on the target cable, with permalloy fixedly connected to both sides below the armature, and a permanent magnet is provided below the armature between the two permalloy pieces.

[0008] As an optional solution, the climbing module is equipped with a shock absorption system; the shock absorption system includes a shock absorption connecting rod, a shock absorption belt, a shock absorption rubber-coated pulley, and a shock absorption spring; a pulley is provided at each end of the armature, and the pulley is a shock absorption rubber-coated pulley; one pulley is fixedly connected to the armature through the shock absorption spring; the other pulley is fixedly connected to the armature through the shock absorption connecting rod and connected to the motor through the shock absorption belt.

[0009] As an optional solution, the guided wave-leakage magnetic flux composite intelligent detection device also includes a battery to provide power to each module.

[0010] As an optional solution, the embedded processor incorporates a signal denoising and filtering algorithm and a trained machine learning model to process the input signal in real time and output cable damage assessment results.

[0011] As an optional solution, the number of magnetic flux leakage detection units is determined by the diameter of the target cable; the larger the diameter, the more units are required.

[0012] As an optional solution, the embedded processor is an embedded development board based on the Linux system; the embedded development board adopts a Raspberry Pi, which integrates a CPU, memory, GPIO pins and an Ethernet port, and is used to run the signal denoising and filtering algorithm and the machine learning model; the data acquisition card is a PXI-4496 integrated data acquisition card, which is used to convert the amplified analog signal from the digital acquisition module into a digital signal.

[0013] As an optional solution, the signal denoising and filtering algorithm adopts the wavelet transform algorithm to denoise the non-stationary abrupt signal generated by cable damage; the machine learning model is a deep neural network model, which is used to automatically extract cable damage features and identify the damage type, size and location; the input end of the calculation control module is connected to the output channel of the data acquisition card, and the output end outputs the damage assessment results.

[0014] A second aspect of the present invention provides a method for intelligent detection of bridge cable damage based on the above-mentioned guided wave-leakage magnetic flux composite intelligent detection device, comprising the following steps: S1: Select the corresponding number of magnetic flux leakage detection units and preamplifier circuits according to the diameter of the target cable, and combine them into a ring composite detection probe array through the module connection device, which is installed around the outer periphery of the target cable. S2: Start the climbing module. The guided wave-leakage magnetic flux composite intelligent detection equipment will climb steadily along the axis of the target cable, while the data acquisition module will work synchronously. S3: The excitation circuit of the leakage flux detection unit uniformly saturates the target cable with magnetization, and the Hall element collects the leakage flux signal caused by the defect; the magnetostrictive guided wave detection unit excites and receives the guided wave signal and collects the guided wave reflection signal. S4: The original leakage flux signal and guided wave signal are amplified by the preamplifier circuit and transmitted to the data acquisition card through the shielded cable, where they are converted into digital signals. The preamplifier circuit is located inside the data acquisition module and amplifies the original signals acquired by the Hall element and the detection coil to improve the signal-to-noise ratio and signal amplitude. The amplified signal is transmitted to the data acquisition card through the shielded cable. S5: The embedded processor incorporates a signal denoising and filtering algorithm and a trained machine learning model; the computational control module performs wavelet transform denoising and filtering on the digital signal to extract damage features; S6: Input the denoised and filtered signal into the trained machine learning model, and the model automatically identifies the damage type, size and location; S7: The output of the calculation and control module displays the cable damage assessment results in real time, completing the detection.

[0015] As an optional solution, the signal denoising and filtering algorithm adopts the wavelet transform algorithm, which is implemented through the wavelet transform library in LabVIEW. Compared with the traditional Fourier transform, it retains the time resolution and frequency resolution, while denoising the non-stationary abrupt signal caused by cable damage. The training method of the machine learning model includes: constructing a cable sample database containing different types and degrees of damage, extracting the leakage magnetic field and guided wave signal features of the samples, and performing supervised learning training so that the model learns the mapping relationship between damage features and damage type, size, and location.

[0016] The beneficial effects of this invention are as follows: This invention leverages the complementary advantages of waveguide and magnetic flux leakage technologies to achieve comprehensive detection of the cable anchorage zone and its entire length. The introduction of wavelet denoising and machine learning algorithms significantly improves signal processing efficiency and damage identification accuracy. The modular design of the detection equipment adapts to different cable diameters and solves the challenges of high-altitude detection. Compared to existing technologies, this invention improves detection accuracy by over 20% and data processing efficiency by more than double, providing a novel technical means for health monitoring of bridge cable structures. This invention can be widely applied to the daily inspection and maintenance of cable-stayed bridges, suspension bridges, and other cable-stayed structures, providing a new technical solution for the full life-cycle health monitoring of bridges. The intelligent detection method reduces reliance on the experience of inspection personnel, lowers labor costs, and improves detection efficiency, which is of great significance for ensuring the safe operation of bridges. Attached Figure Description

[0017] Figure 1 This is a schematic diagram of the overall three-dimensional structure of the guided wave-leakage magnetic flux composite detection device of the present invention.

[0018] Figure 2 This is a schematic diagram of the climbing module structure of the guided wave-leakage magnetic flux composite detection device of the present invention.

[0019] Figure 3 This is a schematic diagram of the data acquisition module structure of the guided wave-leakage magnetic flux composite detection device of the present invention.

[0020] Figure 4 This is a schematic diagram of the computational control module structure of the guided wave-leakage magnetic flux composite detection device of the present invention.

[0021] Explanation of the labels in the diagram: 1-Module connection device, 2-Signal processing equipment, 3-Connecting line, 4-Permanent magnet, 5-Permalloy, 6-Coil, 7-Target cable, 8-Shock-absorbing connecting rod, 9-Shock-absorbing belt, 10-Shock-absorbing rubber-coated pulley, 11-Armature, 12-Shock-absorbing spring, 13-Hall element, 14-Battery, 15-Servo motor. Detailed Implementation

[0022] The technical solution of the present invention will now be described in detail with reference to the accompanying drawings and specific embodiments.

[0023] like Figures 1-4 As shown, the present invention provides a guided wave-leakage magnetic flux composite detection device for bridge cable damage detection; including a data acquisition module, a calculation and control module, and a climbing module.

[0024] The data acquisition module includes a composite detection probe consisting of a magnetic flux leakage detection unit and a magnetostrictive guided wave detection unit, as well as a preamplifier circuit, all integrated within the module. The data acquisition module employs a ring array structure, with multiple magnetic flux leakage detection units forming a ring array around the target cable via module connection device 1. The number of magnetic flux leakage detection units is selected based on the diameter of the target cable 7; the larger the cable diameter, the more magnetic flux leakage detection units are required. The module connection device 1 uses a standardized interface to achieve mechanical connection and electrical conduction between the modules.

[0025] Specifically, the magnetic flux leakage detection unit consists of a permanent magnet 4, an armature 11, and a permalloy 5 forming an excitation circuit, and a Hall element 13 is arranged at the air gap of the excitation circuit to collect the magnetic flux leakage signal caused by the defect. The magnetostrictive guided wave detection unit uses a self-excited and self-receiving coil to excite and receive guided wave signals. During excitation, the signal generator in the calculation and control module generates an alternating electrical signal to drive the coil 6 to excite the guided wave. During reception, the guided wave reflection signal picked up by the coil 6 is amplified by the preamplifier circuit and transmitted to the data acquisition card of the calculation and control module through a shielded cable to achieve rapid screening of corrosion damage along the entire length of the cable and in the anchorage area. The permanent magnet 4, the armature 11, the target cable 7, and the air gap together form a closed magnetic circuit, establishing an excitation circuit along the cable axis. The coercivity and remanence of the permanent magnet 4 are used to achieve uniform saturation magnetization of the cable. The armature 11 is made of pure iron material with high relative permeability to establish a uniform saturation magnetization field along the cable axis to optimize the magnetic circuit efficiency.

[0026] Structurally, more specifically, the magnetic flux leakage detection unit includes a permanent magnet 4, an armature 11, and permalloy 5 forming an excitation circuit, as well as a Hall element 13 disposed between the excitation circuit and the target cable 7. The magnetostrictive guided wave detection unit includes a detection coil arranged below the Hall element 13, surrounding the outer edge of the target cable 7. The armature 11 is elongated and axially arranged on the target cable 7. Permalloy 5 is fixedly connected to both sides below the armature 11, and a permanent magnet 4 is disposed below the armature between the two pieces of permalloy 5. The signal amplification circuit is located inside the data acquisition module, adjacent to the Hall element and the detection coil, so that the original signal, especially the weak magnetic flux leakage signal and the guided wave reflection signal, is amplified nearby before transmission, which can effectively improve the signal-to-noise ratio and reduce the noise introduced by long-distance transmission. The coil is a self-excited and self-receiving coil, which can be connected to a signal generator for signal excitation and an amplification circuit for signal reception through a transmit / receive switching circuit, such as an analog switch or a relay. During the excitation phase, the coil is connected to the signal generator; during the reception phase, the coil is connected to the amplifier circuit. Through self-excitation and self-reception, the coil performs both the excitation function (generating an alternating magnetic field under the signal generator's drive to excite the guided wave) and the reception function (picking up the echo signal).

[0027] The climbing module is used to drive the guided wave-leakage magnetic flux composite intelligent detection device to climb axially on the target cable surface. It includes pulleys and a motor fixedly connected to the data acquisition module. Optionally, the number of climbing modules can be the same as the number of leakage magnetic flux detection units, i.e., each leakage magnetic flux detection unit is equipped with a climbing module. The climbing module carries a motor, a shock-absorbing connecting rod 8, a shock-absorbing belt 9, a shock-absorbing rubber-coated pulley 10, and a shock-absorbing spring 12. The motor is a servo motor 15, providing driving force for the robot, i.e., the entire detection device, to climb axially along the target cable 7. The shock-absorbing connecting rod 8, shock-absorbing belt 9, shock-absorbing rubber-coated pulley 10, and shock-absorbing spring 12 together constitute a multi-stage shock absorption structure, used to absorb vibrations caused by uneven cable surface during climbing, ensuring that the detection probe where the Hall element is located maintains a constant lift-off distance from the cable surface, thereby improving the stability of signal acquisition.

[0028] Structurally, more specifically, a shock-absorbing rubber-coated pulley 10 is provided at each end of the armature 11 of the data acquisition module; one pulley is fixedly connected to the armature 11 through a shock-absorbing spring 12; the other pulley is fixedly connected to the armature 11 through a shock-absorbing connecting rod 8 and connected to the servo motor 15 through a shock-absorbing belt 9.

[0029] The computing control module incorporates an embedded signal generator and signal processing device 2. The signal generation circuit generates the initial excitation signal. The signal processing device 2 includes a data acquisition card and an embedded processor, electrically connected to the data acquisition module via a connecting cable 3. It is embedded within the computing control module and uses a machine learning model as its core. The data acquisition card is a PXI-4496 integrated data acquisition card, used to convert the amplified analog signal from the data acquisition module into a digital signal. The embedded processor is a Raspberry Pi development board based on the Linux system, integrating a CPU, main control chip, memory, GPIO pins, and an Ethernet port, used to run signal processing algorithms and machine learning models. The power supply module uses battery 14 to provide operating power to each module.

[0030] Furthermore, the signal denoising and filtering algorithm employs a wavelet transform algorithm to denoise the non-stationary abrupt change signal caused by cable damage by preserving time and frequency resolution. The wavelet transform algorithm is implemented using the wavelet transform library in LabVIEW. The machine learning model is a deep neural network model, trained through supervised learning using a sample database containing different types and degrees of damage. The trained model is embedded in the computational control module to automatically extract cable damage features and identify the damage type, size, and location. The input of the computational control module is connected to the output channel of the data acquisition card, and the output directly provides the damage assessment result. During detection, the climbing module drives the entire detection device to climb upwards along the cable direction, while the data acquisition module simultaneously completes the detection, acquiring leakage magnetic signals and guided wave signals. The original signal is amplified by the amplification circuit and then converted into a digital signal by the data acquisition card. The embedded processor, with its embedded signal denoising and filtering algorithm and trained machine learning model, processes the digital signal in real time and outputs the cable damage assessment result, achieving real-time online assessment.

[0031] The composite intelligent detection device of the present invention has the following detection principle: When the magnetic flux leakage detection unit is working, the excitation circuit uniformly saturates the target cable, generating a uniform magnetic field inside the cable. When defects such as cracks, holes, or corrosion exist inside or on the surface of the cable, the magnetic permeability at the defect location changes, causing local distortion of the magnetic field and generating leakage flux. The Hall element collects the intensity and distribution characteristics of the leakage magnetic field signal, and by analyzing these signals, the type, size, and location of the defect can be determined.

[0032] Magnetostrictive guided wave detection units utilize the magnetostrictive effect of magnetic materials: under the action of an applied alternating magnetic field, the magnetic material undergoes expansion and contraction, thereby exciting guided waves to propagate along the cable; when the guided wave encounters a defect or boundary, a reflected wave is generated and picked up by the detection coil. By analyzing the propagation time and waveform characteristics of the guided wave, rapid screening of wire damage within the entire length of the cable (including the anchorage zone) can be achieved. In other words, the working principle of the magnetostrictive guided wave detection unit is as follows: magnetic materials undergo magnetostriction under the action of an applied magnetic field, manifested as changes in the material's length or shape with the magnetic field strength; by applying a certain magnetic field to monitor material deformation, the internal magnetic field distribution changes accordingly when the material is subjected to force or deformation, and the detection coil collects the magnetic field change signal to identify defects in the internal wires of the cable, which is particularly suitable for anchorage section detection.

[0033] The two detection signals are amplified by an amplifier circuit and transmitted through a shielded cable to a PXI-4496 data acquisition card for conversion into digital signals, which are then sent to a Raspberry Pi development board for processing. Due to electromagnetic interference in the field environment, the original signals are mixed with noise, and the cable damage signal is a non-stationary abrupt change signal, making it difficult for traditional Fourier transform to maintain both time and frequency resolution. Therefore, this invention runs a wavelet transform algorithm library in the LabVIEW environment on the Raspberry Pi development board to perform wavelet denoising filtering on the digital signal, effectively suppressing noise while preserving both time and frequency resolution, and extracting pure damage features.

[0034] Furthermore, this invention embeds the trained machine learning model into the computational control module. This model, trained through supervised learning using a sample database containing different types and degrees of damage, can automatically identify damage features and output the damage type, size, and location. During detection, the climbing module drives the robot to climb at a constant speed along the target cable axis, while the data acquisition module simultaneously collects leakage magnetic field and guided wave signals. After real-time processing, the damage assessment results are directly output, achieving automation and intelligence in the detection process.

[0035] The detection method described in this invention is implemented according to the following steps: First stage: Based on the diameter of the target cable 7, select the corresponding number of magnetic flux leakage detection units. Through the module connection device 1, combine multiple magnetic flux leakage detection units and their preamplifier circuits with the magnetostrictive guided wave detection units to form a ring composite detection probe, ensuring that the electrical connection of each module is reliable. The whole unit is installed around the outer periphery of the target cable 7.

[0036] Phase 2: Climbing Detection. The climbing module is activated, and the servo motor 15 drives the shock-absorbing rubber-coated pulley 10 to climb upwards along the surface of the target cable 7. The multi-stage shock absorption structure maintains a constant lifting distance between the Hall element probe and the cable. At the same time, the data acquisition module is activated, the excitation circuit establishes a stable magnetic field, the coil 6 excites the guided wave, and the Hall element 13 and the coil 6 synchronously acquire leakage magnetic signals and guided wave reflection signals.

[0037] The third stage: signal processing. The original signal is amplified by the signal amplification circuit and then transmitted to the PXI-4496 data acquisition card through a shielded cable, where it is converted into a digital signal. The Raspberry Pi development board uses the wavelet transform library to perform noise reduction and filtering on the digital signal and extract damage features.

[0038] Phase Four: Intelligent Assessment. The processed feature data is input into the trained machine learning model, which automatically identifies the damage type, size, and location, and displays the damage assessment results in real time at the output. The entire detection process is completed synchronously during the robot's climbing, achieving continuous detection along the entire length of the cable.

[0039] Thus, a guided wave-leakage magnetic flux composite detection device and intelligent evaluation method capable of efficiently and accurately detecting bridge cable damage has been completed.

[0040] Based on the above solution, it can be seen that the technical solution of the present invention has the following characteristics and advantages: 1) Significantly Improved Detection Performance: By employing a guided wave-magnetic leakage composite detection mechanism, the high-precision positioning advantage of magnetic leakage detection technology and the sensitivity of magnetostrictive guided wave detection technology to the anchorage zone are fully utilized. This optimizes the damage identification path to "raw signal acquisition → signal amplification → analog-to-digital conversion → wavelet denoising → feature extraction → machine learning identification → damage assessment," significantly improving the signal-to-noise ratio and defect identification accuracy. The composite detection scheme effectively overcomes the limitations of single detection methods in terms of applicability and accuracy, enabling comprehensive detection of internal wire damage along the entire length of the cable, including the anchorage zone.

[0041] 2) Advanced and Practical Structural Design: The modular detection unit design decouples climbing, data acquisition, and computing functions. Each module is independently packaged with standardized interfaces, facilitating maintenance and functional expansion. The climbing module employs permanent magnet adsorption and a multi-stage shock absorption structure to adapt to cables of different diameters and surface conditions. The data acquisition module uses a ring-shaped quick-release structure, allowing for the replacement of composite detection probes of different specifications according to the cable diameter. The computing and control module uses a general-purpose embedded platform, supporting the online deployment and updating of various machine learning models. This modular design significantly improves the adaptability and upgradeability of the detection equipment, filling the technological gap in high-altitude and complex area detection that traditional detection methods cannot cover.

[0042] 3) Significantly Enhanced Intelligence: By constructing a sample database containing different types and degrees of damage, a machine learning model is trained to automatically extract damage features. The trained algorithm is then embedded into the robot's computing control module, replacing traditional manual interpretation methods and significantly improving data processing efficiency and damage identification accuracy. The computing control module adopts an edge computing architecture to achieve real-time data processing and damage early warning at the inspection site, solving the technical problems of cumbersome data processing and low efficiency in traditional inspection methods.

[0043] 4) Reliable signal processing quality: Wavelet transform algorithm is used to denoise non-stationary damage signals, which is particularly suitable for processing abrupt signals caused by cable damage. The combination of amplification circuit and shielded transmission effectively suppresses on-site noise interference, ensuring high-quality raw data acquisition. The armature is made of pure iron material with high relative permeability, and the excitation circuit structure is optimized to improve magnetization uniformity.

[0044] The present invention has been described in detail above with reference to specific embodiments and exemplary examples. However, these descriptions should not be construed as limiting the present invention. Those skilled in the art will understand that various equivalent substitutions, modifications, or improvements can be made to the technical solutions and embodiments of the present invention without departing from the spirit and scope of the present invention, and all such modifications and improvements fall within the scope of the present invention.

Claims

1. A guided wave-leakage magnetic flux composite intelligent detection device for bridge cable damage detection, characterized in that, It includes a data acquisition module, a computing and control module, and a climbing module; The computing control module embeds a signal processing device and a signal generator. The signal processing device includes a data acquisition card and an embedded processor that are electrically connected to the data acquisition module. The data acquisition module includes a magnetic flux leakage detection unit, a magnetostrictive guided wave detection unit, and a preamplifier circuit fixed on the magnetic flux leakage detection unit; the magnetic flux leakage detection unit includes a permanent magnet constituting the excitation circuit, an armature and permalloy, and a Hall element disposed between the excitation circuit and the target cable. Multiple magnetic flux leakage detection units are connected by a module to form a ring array around the target cable; the magnetostrictive guided wave detection unit includes a coil arranged below the Hall element and surrounding the outer side of the target cable; the coil is connected to the signal generator and the preamplifier circuit respectively; The climbing module includes pulleys and a motor fixedly connected to the data acquisition module, which are used to drive the guided wave-leakage magnetic flux composite intelligent detection device to climb axially on the surface of the target cable.

2. The guided wave-leakage magnetic flux composite intelligent detection device according to claim 1, characterized in that, The armature is long and strip-shaped, axially arranged on the target cable. Permalloy is fixedly connected to both sides below the armature, and a permanent magnet is provided below the armature between the two pieces of permalloy.

3. The guided wave-leakage magnetic flux composite intelligent detection device according to claim 2, characterized in that, A pulley is provided at each end of the armature. The pulley is a shock-absorbing rubber-coated pulley. One pulley is fixedly connected to the armature through a shock-absorbing spring. The other pulley is fixedly connected to the armature through a shock-absorbing connecting rod and connected to the motor through a shock-absorbing belt.

4. The guided wave-leakage magnetic flux composite intelligent detection device according to claim 1, characterized in that, It also includes batteries to provide power to the various modules.

5. The guided wave-leakage magnetic flux composite intelligent detection device according to claim 1, characterized in that, The embedded processor is equipped with a signal denoising and filtering algorithm and a trained machine learning model to process the input signal in real time and output cable damage assessment results.

6. The guided wave-leakage magnetic flux composite intelligent detection device according to claim 1, characterized in that, The number of magnetic flux leakage detection units depends on the diameter of the target cable; the larger the diameter, the more units are required.

7. The guided wave-leakage magnetic flux composite intelligent detection device according to claim 5, characterized in that, The embedded processor is an embedded development board based on the Linux system; the embedded development board uses a Raspberry Pi, which integrates a CPU, memory, GPIO pins and an Ethernet port, and is used to run the signal denoising and filtering algorithm and the machine learning model; the data acquisition card is a PXI-4496 integrated data acquisition card, which is used to convert the amplified analog signal from the digital acquisition module into a digital signal.

8. The guided wave-leakage magnetic flux composite intelligent detection device according to claim 7, characterized in that, The signal denoising and filtering algorithm employs wavelet transform to denoise non-stationary abrupt change signals caused by cable damage; the machine learning model is a deep neural network model used to automatically extract cable damage features and identify damage type, size, and location; the input of the calculation control module is connected to the output channel of the data acquisition card, and the output output shows the damage assessment results.

9. A method for intelligent detection of bridge cable damage based on the guided wave-leakage magnetic flux composite intelligent detection device according to any one of claims 1-8, characterized in that, Includes the following steps: S1: Select the corresponding number of magnetic flux leakage detection units according to the diameter of the target cable, and combine them with the magnetic flux leakage detection units through the module connection device to form a ring composite detection probe array, which is installed around the outer periphery of the target cable; S2: Start the climbing module. The guided wave-leakage magnetic flux composite intelligent detection equipment climbs along the axis of the target cable, and the data acquisition module works synchronously. S3: The excitation circuit of the leakage flux detection unit uniformly saturates the target cable with magnetization, and the Hall element collects the leakage flux signal; the magnetostrictive guided wave detection unit excites and receives the guided wave signal. S4: The original leakage flux signal and guided wave signal are amplified by the preamplifier circuit and transmitted to the data acquisition card through the shielded cable, where they are converted into digital signals; S5: The embedded processor incorporates a signal denoising and filtering algorithm and a trained machine learning model; the computational control module performs wavelet transform denoising and filtering on the digital signal to extract damage features; S6: Input the denoised and filtered signal into the trained machine learning model to automatically identify the damage type, size and location; S7: The output of the calculation and control module displays the cable damage assessment results in real time, completing the detection.

10. The method according to claim 9, characterized in that, The signal denoising and filtering algorithm adopts the wavelet transform algorithm, which is implemented through the wavelet transform library in LabVIEW. By preserving the time resolution and frequency resolution, it performs denoising processing on the non-stationary abrupt signal caused by cable damage. The training method for the machine learning model includes: constructing a cable sample database containing different types and degrees of damage, extracting the leakage magnetic field and guided wave signal features of the samples, and performing supervised learning training so that the model learns the mapping relationship between damage features and damage type, size, and location.

Images