Method and system for detecting defects of small-caliber non-metal composite pipe

Abstract

The invention belongs to the field of pipeline detection, and relates to a small-caliber nonmetal composite pipe defect detection method and system, and the method comprises the following steps: S1, collecting a magnetic flux leakage signal, video information and position information; s2, according to the magnetic flux leakage signal, obtaining a corrosion depth through a fast Fourier transform algorithm of a Hanning window; s3, based on the video information, the flattening deformation amount and the perforation diameter of the composite pipe are obtained through a defect recognition model; s4, a corrected corrosion position, a corrected flattening position and a corrected perforation position are obtained; and S5, determining an excavation position by an operator based on the defect signal received by the positioning receiver, the corrected corrosion position, the corrected flattening position and the corrected perforation position. The small-size inner detection device is used for detecting the defects of the small-caliber non-metal composite pipe, and the problems that an existing detector is not suitable for the small-caliber non-metal composite pipe, the detection result is low in accuracy, and the universality is poor are solved.

Description

Technical Field

[0001] This invention relates to the field of pipeline inspection, and specifically discloses a method and system for detecting defects in small-diameter non-metallic composite pipes. Background Technology

[0002] Composite pipes are composed of multiple layers of non-metallic materials, typically connected by metal joints. The combined length of these pipes is usually tens of kilometers. In practical applications, due to corrosion, external damage, and other factors, composite pipes may develop defects such as perforation and flattening. Corrosion and leakage are also common at metal joints, requiring repair by operators. Current detectors for detecting defects in composite pipes, such as acoustic or electromagnetic wave detectors, have the following problems: they cannot penetrate the non-metallic layer to identify metal joints, resulting in low accuracy; wireless transmission is susceptible to interference from the pipeline environment, and the transmission distance is less than 500 meters, further reducing accuracy; existing detectors are bulky and unsuitable for composite pipes with a diameter less than 150 mm, leading to poor versatility. Summary of the Invention

[0003] The purpose of this invention is to provide a method and system for detecting defects in small-diameter non-metallic composite pipes, solving the problems of existing detectors being unsuitable for small-diameter composite pipes, having low accuracy of detection results, and poor versatility.

[0004] The specific solution of the present invention is as follows: A method for detecting defects in small-diameter non-metallic composite pipes includes the following steps: S1. When the internal detection device moves inside the composite tube, it collects leakage magnetic field signals, video information and position information through fiber optic cable twisters and nylon cable twisters, and associates the position information with the leakage magnetic field signals and video signals respectively. S2. Detect metal joints based on magnetic flux leakage signals. When a metal joint is detected, obtain the corrosion depth using the fast Fourier transform algorithm of the Hanning window based on the magnetic flux leakage signal, and obtain the corrosion location based on the location information associated with the magnetic flux leakage signal. S3. Based on video information, obtain the flattening deformation amount and perforation diameter of the composite pipe through a defect identification model, and obtain the flattening position and perforation position of the composite pipe based on the location information associated with the video information. S4. Based on the mileage recorded by the fiber optic cable winch, the corrosion location, flattening location, and perforation location are corrected to obtain the corrected corrosion location, corrected flattening location, and corrected perforation location. S5. When the corrosion depth, flattening deformation amount, and perforation diameter are obtained, the trigger signal transmitter emits a defect signal. The operator determines the excavation location based on the defect signal received by the positioning receiver and the corrected corrosion location, flattening location, and perforation location.

[0005] In some embodiments, detecting a metal connector based on a magnetic flux leakage signal includes: The magnetic flux leakage signal includes signals from four magnetic flux leakage probes; the metal connector is identified based on the four magnetic flux leakage probe signals through a majority voting mechanism.

[0006] In some embodiments, identifying a metal connector based on a majority voting mechanism using signals from four magnetic flux leakage probes includes: The fluctuation amplitude of each magnetic flux leakage probe signal is determined to be greater than the threshold. If the result is that the fluctuation amplitude of at least three magnetic flux leakage probe signals is greater than the threshold, then a metal connector is detected; if the result is that the fluctuation amplitude of no at least three magnetic flux leakage probe signals is greater than the threshold, then no metal connector is detected.

[0007] In some embodiments, obtaining the corrosion depth based on the leakage magnetic field signal using the Hanning window's fast Fourier transform algorithm includes: The time-domain signal is obtained from the leakage magnetic field signal. The time-domain signal is truncated or padded with zeros to make the length of the time-domain signal the same as the length of the Hanning window. The time-domain signal is multiplied point by point by the Hanning window to obtain the windowed time-domain signal. The windowed time-domain signal is converted into a frequency domain spectrum by Fast Fourier Transform (FFT). The frequency domain spectrum is analyzed to extract the characteristic frequency offset. The corrosion depth is obtained based on the characteristic frequency offset.

[0008] In some embodiments, the formula for calculating the corrosion depth is: , Where d is the corrosion depth, Δf is the characteristic frequency offset, and k is the calibration coefficient.

[0009] In some embodiments, the defect identification model is a YOLOv8 model that is optimized and updated using a focus loss function.

[0010] In some embodiments, the formula for calculating the flattening deformation is: , Where T is the amount of flattening deformation, D1 is the standard diameter of the composite pipe, and D2 is the measured diameter of the flattened part of the composite pipe.

[0011] In some embodiments, a data preprocessing step is also included, wherein the data preprocessing includes: Data cleaning, filtering, and standardization processes.

[0012] This invention also relates to a defect detection system for small-diameter non-metallic composite pipes, comprising an internal detection device, a computing device, a fiber optic signal modulator, a fiber optic cable reel, a nylon cable reel, and a positioning receiver. The internal detection device sequentially includes a camera locator, a data video acquisition unit, a power supply, a magnetic flux leakage sensor, a data processor, a signal transmitter, and a fiber optic transceiver. The internal detection device is placed inside the composite pipe. The computing device, the fiber optic signal modulator, and the fiber optic cable reel are located outside one port of the composite pipe, and the nylon cable reel is located outside the other port of the composite pipe. The camera locator is connected to the nylon cable in the nylon cable reel. The fiber optic transceiver is connected to one end of the fiber optic cable in the fiber optic cable reel, and the other end of the fiber optic cable is connected to the input end of the fiber optic signal modulator. The output end of the fiber optic signal modulator is connected to the input end of the computing device. The fiber optic cable reel is connected to the computing device, and the positioning receiver is wirelessly connected to the signal transmitter.

[0013] In some embodiments, the magnetic flux leakage sensor is an array-type Hall sensor, which includes four magnetic flux leakage probes arranged in pairs overlapping each other at 30°.

[0014] Compared with the prior art, the present invention has the following advantages and beneficial effects: 1. This invention uses an internal detection device to detect metal joints, obtains corrosion depth using the Hanning window's fast Fourier transform algorithm, acquires the flattening deformation and perforation diameter of the composite pipe using a defect identification model, corrects the defect location based on the mileage recorded by the fiber optic cable winch, and further corrects the defect location to determine the excavation location using a positioning receiver. This improves the accuracy and versatility of defect detection results for small-diameter non-metallic composite pipes and increases the maintenance efficiency of operators. The internal detection device, composed of a camera locator, data video acquisition unit, power supply, magnetic flux leakage sensor, data processor, signal transmitter, and fiber optic transceiver, is small in size and suitable for small-diameter composite pipes. Attached Figure Description

[0015] Figure 1 This is a flowchart of a defect detection method for small-diameter non-metallic composite pipes according to an embodiment of the present invention.

[0016] Figure 2 This is a circuit diagram of a small-diameter non-metallic composite pipe defect detection system according to an embodiment of the present invention.

[0017] Reference numerals: 1-Computing device, 2-Fiber optic signal modulator, 3-Fiber optic cable winch, 4-Nylon cable winch, 5-Positioning receiver, 6-Camera locator, 7-Data video acquisition device, 8-Power supply, 9-Fluorescence sensor, 10-Data processor, 11-Signal transmitter, 12-Fiber optic transceiver, 13-Fiber optic cable, 14-Nylon cable, 15-Internal detection device. Detailed Implementation

[0018] To make the objectives, technical solutions, and advantages of the embodiments of the present invention clearer, the technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. The components of the embodiments of the present invention described and shown in the accompanying drawings can generally be arranged and designed in various different configurations.

[0019] A method for detecting defects in small-diameter non-metallic composite pipes, such as Figure 1 As shown, it includes the following steps: S1. Data information is collected when the internal detection device 15 moves inside the composite tube through the fiber optic cable winch 3 and the nylon cable winch 4, and the collected data information is preprocessed. Place an optical fiber winch 3 outside one end of the composite pipe and a nylon wire winch 4 outside the other end of the composite pipe. Connect the nylon wire 14 in the nylon wire winch to the cleaning device and put it into the composite pipe. Use air pressure or hydraulic pressure to drive the cleaning device to carry the nylon wire 14 through the entire composite pipe, and then remove the nylon wire 14 from the cleaning device.

[0020] The front end of the internal detection device is connected to the nylon wire 14, and the rear end is connected to the fiber optic wire 13 in the fiber optic cable reel. Rotating the nylon cable reel 4 to retract the nylon wire 14 allows the internal detection device 15 to move forward within the composite tube. During this time, the fiber optic wire 13 in the fiber optic cable reel is in a feeding state, meaning the fiber optic wire 13 is continuously fed out as the internal detection device 15 moves forward. Rotating the fiber optic cable reel 3 to retract the fiber optic wire 13 allows the internal detection device 15 to move backward within the composite tube. During this time, the nylon wire 14 in the nylon cable reel is in a feeding state, meaning the nylon wire 14 is continuously fed out as the internal detection device 15 moves backward. This allows the internal detection device 15 to move forward or backward within the composite tube and stop flexibly at any position. Simultaneously, the fiber optic cable reel 3 records the distance traveled by the internal detection device while feeding and retracting the fiber optic wire 13.

[0021] When the internal detection device 15 moves in the composite tube at a speed of 0.1 m / s to 1 m / s, it collects the leakage magnetic signal in real time through the leakage magnetic sensor 9, collects video information in real time through the camera, and collects the position information in real time through the locator. The position information is then associated with the leakage magnetic signal and the video information, and the mileage of the internal detection device is recorded through the fiber optic cable winch 3.

[0022] Data processor 10 performs data preprocessing on the received magnetic flux leakage signal, video information, location information, and mileage. The data preprocessing includes: Data cleaning processes remove noisy, duplicate, and invalid data from the data. The filtering process involves using Gaussian filtering to remove noise from the video information and using wavelet transform to filter the magnetic leakage signal and eliminate interference signals. Standardization processes convert magnetic leakage signals, video information, location information, and mileage into a unified format and range, eliminating differences in dimensions, units, or orders of magnitude, reducing environmental interference, and improving the comparability, stability, and algorithm adaptability of the data.

[0023] S2. Detect the metal joint based on the pre-processed magnetic flux leakage signal. When the metal joint is detected, obtain the corrosion depth and corrosion location based on the magnetic flux leakage signal and location information. The magnetic flux leakage signal includes signals from four magnetic flux leakage probes. Based on the preprocessed signals, a majority voting mechanism is used to identify metal joints. Specifically, this involves determining whether the fluctuation amplitude of each magnetic flux leakage probe signal exceeds a threshold, which can be 5mV. If at least three magnetic flux leakage probe signals exceed the threshold, a metal joint is detected, and corrosion defect detection is triggered. If no at least three magnetic flux leakage probe signals exceed the threshold, no metal joint is detected, and corrosion defect detection is not triggered. This majority voting mechanism effectively reduces the false alarm rate for metal joints.

[0024] When a metal joint is detected, corrosion defect detection of the metal joint is triggered. The corrosion depth is obtained using a Fast Fourier Transform algorithm with a Hanning window based on the magnetic flux leakage signal. The corrosion location is obtained based on the location information associated with the magnetic flux leakage signal, specifically including: The time-domain signal is obtained from the leakage magnetic field signal. The time-domain signal is truncated or padded with zeros to ensure its length matches the length of the Hanning window. The time-domain signal is then multiplied point-by-point by the Hanning window to obtain the windowed time-domain signal. A Fast Fourier Transform (FFT) is performed on the windowed time-domain signal to convert it into a frequency domain spectrum. Spectral analysis is then performed on the frequency domain spectrum to extract the characteristic frequency offset. The corrosion depth of the metal joint is obtained based on the characteristic frequency offset. Finally, the corrosion location, i.e., the location of the metal joint, is obtained based on the location information associated with the leakage magnetic field signal.

[0025] In the field of signal processing, the Fast Fourier Transform (FFT) is an efficient tool for converting time-domain signals into frequency-domain spectra. However, it is prone to spectral leakage when truncating signals. The Hanning window, as a commonly used window function, can effectively suppress this spectral leakage. Therefore, the Fast Fourier Transform algorithm using the Hanning window can improve the accuracy of spectral analysis of magnetic leakage signals.

[0026] A large number of magnetic flux leakage signals at different corrosion depths are acquired using the magnetic flux leakage sensor 9. Based on these signals, a large number of characteristic frequency offsets are obtained. It can be seen from these characteristic frequency offsets at different corrosion depths that there is a linear relationship between corrosion depth and characteristic frequency offset. The formula for calculating corrosion depth is: , Where d is the corrosion depth, Δf is the characteristic frequency offset, and k is the calibration coefficient.

[0027] S3. Based on the preprocessed video information, the flattening deformation and perforation diameter of the composite pipe are obtained through the defect identification model. The flattening position and perforation position of the composite pipe are obtained based on the position information associated with the video information. Composite pipe defects include flattening defects and perforation defects. Based on the pre-processed video information, the flattening deformation amount of the flattening defect and the perforation diameter of the perforation defect are obtained through the defect recognition model.

[0028] The formula for calculating the flattening deformation is: , Where T is the amount of flattening deformation, D1 is the standard diameter of the composite pipe, and D2 is the measured diameter of the flattened part of the composite pipe.

[0029] The construction of the defect recognition model includes: acquiring a large number of defect videos of composite pipes under different working conditions using various detection devices; performing image enhancement and defect annotation on the defect videos; adjusting brightness, contrast, and color balance of the defect videos to improve image clarity and quality and enhance the recognizability of defect features; marking defect types, delimiting defect bounding boxes, and annotating defect dimensions in the image-enhanced defect videos to form a training dataset; inputting the training dataset into the YOLOv8 model for defect recognition model training; using a focus loss function to optimize and update the defect recognition model during training; and obtaining a defect recognition model for composite pipe flattening and perforation defects after training.

[0030] The expression for the focus loss function is: , in, Predict probabilities for the model; This is the modulation factor coefficient, typically set to 0.25; The weight parameters are used to balance the positive and negative samples.

[0031] S4. Based on the mileage recorded by the fiber optic cable winch, the corrosion location, flattening location, and perforation location are corrected to obtain the corrected corrosion location, corrected flattening location, and corrected perforation location. The mileage pulse signals of the camera, locator, magnetic leakage sensor 9 and fiber optic cable winch use the same crystal oscillator clock source to ensure that the timestamps of the mileage data and the location information data are consistent.

[0032] The calculation formula for the correction process is as follows: , in, The coordinates are for the corrected position, which includes the corrected corrosion position, the corrected flattening position, and the corrected perforation position. The coordinates of the position before correction include the corrosion location, the flattening location, and the perforation location; The azimuth angle of the composite pipe; The mileage recorded by the fiber optic cable winch; The mileage is calculated based on the coordinates of the location before correction. =0.8.

[0033] The corrosion depth was correlated with the corrected corrosion location, the flattening deformation was correlated with the corrected flattening location, and the perforation diameter was correlated with the corrected perforation location.

[0034] S5. When the corrosion depth, flattening deformation amount, and perforation diameter are obtained, the trigger signal transmitter emits a defect signal. The operator determines the excavation location based on the defect signal received by the positioning receiver and the corrected corrosion location, flattening location, and perforation location.

[0035] When the internal inspection device 15 performs defect detection inside the composite pipe, the operator holds a positioning receiver 5 and moves along the composite pipe laying route with the internal inspection device at the construction site. When the data processor 10 obtains the corrosion depth, flattening deformation amount, and perforation diameter, it triggers the signal transmitter to emit a defect signal. The operator receives the defect signal emitted by the signal transmitter through the positioning receiver 5 at the construction site. When the positioning receiver 5 starts to receive the defect signal, it provides a defect location reminder based on the signal strength ratio displayed on the positioning receiver. The signal strength ratio is divided into 100%, 90%, 80%, 50%, and 20%. The higher the signal strength ratio, the closer the positioning receiver 5 is to the defect location. When the signal strength ratio reaches 100%, the operator combines the correction of corrosion location, correction of flattening location, and correction of perforation location to determine the excavation location, thereby further locating the defect location of the composite pipe and improving the accuracy of the excavation and repair location.

[0036] The internal detection device 15 detects metal joints, the corrosion depth is obtained through the fast Fourier transform algorithm of the Hanning window, the flattening deformation and perforation diameter of the composite pipe are obtained through the defect identification model, the defect position is corrected based on the mileage recorded by the fiber optic winch, and the excavation position is determined by the positioning receiver 5. This improves the accuracy and versatility of defect detection results for small-diameter non-metallic composite pipes and increases the maintenance efficiency of operators.

[0037] This invention also relates to a defect detection system for small-diameter non-metallic composite pipes, such as... Figure 2 As shown, the device includes: an internal detection device 15, a computing device 1, an optical fiber signal modulator 2, an optical fiber cable reel 3, a nylon cable reel 4, and a positioning receiver 5. The internal detection device 15 sequentially includes a camera locator 6, a data video acquisition device 7, a power supply 8, a magnetic leakage sensor 9, a data processor 10, a signal transmitter 11, and an optical fiber signal transceiver 12. The internal detection device 15 is placed inside the composite tube. The computing device 1, the optical fiber signal modulator 2, and the optical fiber cable reel 3 are located outside one port of the composite tube, and the nylon cable reel 4 is located outside the other port of the composite tube. The camera locator 6 is connected to the nylon cable 14 in the nylon cable reel. The optical fiber signal transceiver is connected to one end of the optical fiber in the optical fiber cable reel, and the other end of the optical fiber is connected to the input end of the optical fiber signal modulator. The output end of the optical fiber signal modulator is connected to the input end of the computing device. The optical fiber cable reel 3 is connected to the computing device 1, and the positioning receiver is wirelessly connected to the signal transmitter.

[0038] The internal detection device 15, consisting of a camera locator 6, a data video acquisition unit 7, a power supply 8, a magnetic leakage sensor 9, a data processor 10, a signal transmitter 11, and an optical fiber signal transceiver 12, is small in size and suitable for small-diameter composite pipes.

[0039] The camera locator 6 includes a camera and a locator. The camera is used to collect video information inside the composite pipe in real time. The video information is used to detect flattening defects and perforation defects in the composite pipe. The camera can be a 200W pixel camera with a resolution greater than or equal to 1080P and a frame rate of 30fps. The locator is used to collect position information in real time and associate the position information with the video information and the magnetic leakage signal respectively, so that each video information and each magnetic leakage signal corresponds to position information.

[0040] The leakage magnetic field sensor 9 is used to acquire leakage magnetic field signals in real time. The leakage magnetic field sensor 9 is an array-type Hall sensor, which includes a coil and a magnetic core. When the leakage magnetic field sensor 9 is close to the metal connector, the metal connector cuts the magnetic lines of force, causing a change in the magnetic flux in the magnetic circuit. The change in magnetic flux causes a change in the induced electromotive force of the coil, which in turn causes the electrical signal output by the leakage magnetic field sensor to fluctuate.

[0041] The magnetic flux leakage sensor 9 includes four magnetic flux leakage probes, which are arranged in a ring array with each probe overlapping the others at 30° angles, enabling 360-degree omnidirectional detection and ensuring that metal joints can be detected from any angle. Each magnetic flux leakage probe independently acquires its signal, and the metal joint is identified based on the four signals through a majority voting mechanism. The majority voting mechanism includes the following: if the fluctuation amplitude of at least three of the four magnetic flux leakage signals is greater than a threshold, the joint is identified as a metal joint; otherwise, it is not identified as a metal joint, thereby reducing the false alarm rate of metal joints.

[0042] The data video acquisition unit 7 is used to convert corrosion depth, corrosion location, flattening deformation amount, flattening location, perforation diameter, perforation location, magnetic leakage signal, video information and location information into signals that can be transmitted by the optical fiber, and to convert the mileage transmitted by the optical fiber into information that can be recognized by the data processor.

[0043] Power supply 8 is used to provide power to the internal detection device 15.

[0044] The data processor 10 is used to preprocess the received magnetic flux leakage signal, video information, location information and mileage; associate the location information with the magnetic flux leakage signal and video information respectively; detect metal joints based on the magnetic flux leakage signal, and when a metal joint is detected, obtain the corrosion depth and corrosion location based on the magnetic flux leakage signal and location information; obtain the flattening deformation amount and perforation diameter of the composite pipe through a defect recognition model based on the video information, and obtain the flattening location and perforation location of the composite pipe based on the location information associated with the video information.

[0045] The signal transmitter is used to transmit defect signals; when the data processor 10 obtains the corrosion depth, flattening deformation amount, and perforation diameter, the signal transmitter is triggered to transmit defect signals.

[0046] The positioning receiver 5 is used to receive the defect signal emitted by the signal transmitter. When the defect signal is received, the signal strength ratio is displayed on the display screen of the positioning receiver. The defect location is reminded by the signal strength ratio. The signal strength ratio is divided into 100%, 90%, 80%, 50%, and 20%. The higher the signal strength ratio, the closer the positioning receiver 5 is to the defect location.

[0047] Fiber optic transceivers are used to transmit information such as corrosion depth, corrosion location, flattening deformation, flattening location, perforation diameter, perforation location, magnetic leakage signal, video information, and location information, and to receive the mileage recorded by the fiber optic cable reel.

[0048] Fiber optic cable 13 is used to transmit signals transmitted or received by the fiber optic transceiver; the transmission rate of the fiber optic cable can be greater than or equal to 1Gbps, the tensile strength of the fiber optic cable can be greater than or equal to 100N, and the length of the fiber optic cable can be 5.5km. Transmitting data signals through fiber optic cable 13 improves the accuracy of long-distance data signal transmission, avoids interference from the pipeline environment, and solves the defect of wireless transmission distance of less than 500 meters.

[0049] The fiber optic signal modulator 2 is used to convert signals transmitted through fiber optic lines into electrical signals that can be recognized by computing devices, and to convert mileage into signals that can be recognized by fiber optic lines.

[0050] The fiber optic cable winch 3 is used to wind up and unwind the fiber optic cable 13. The fiber optic cable 13 can be used to pull the internal detection device 15 to move inside the composite tube and record the mileage of the internal detection device.

[0051] The nylon wire winch 4 is used to wind up and unwind the nylon wire 14, and the nylon wire 14 pulls the internal detection device 15 to move inside the composite tube.

[0052] The computing device 1 is used to store data such as corrosion depth, corrosion location, flattening deformation, flattening location, perforation diameter, perforation location, magnetic flux leakage signal, video information, location information, and mileage.

[0053] The mileage pulse signals of the camera, locator, magnetic leakage sensor 9 and fiber optic cable winch all use the same crystal oscillator clock source to ensure that the timestamps of the mileage data and the location information data are consistent.

[0054] The above are merely preferred embodiments of the present invention and are not intended to limit the present invention. Various modifications and variations can be made to the present invention by those skilled in the art. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of the present invention should be included within the scope of protection of the present invention.

Claims

1. A method for detecting defects in small-diameter non-metallic composite pipes, characterized in that, Includes the following steps: S1. When the internal detection device moves inside the composite tube, it collects leakage magnetic signals, video information and position information through fiber optic cable twisters and nylon cable twisters, and associates the position information with the leakage magnetic signals and video signals respectively. S2. Detect metal joints based on magnetic flux leakage signals. When a metal joint is detected, obtain the corrosion depth using the fast Fourier transform algorithm of the Hanning window based on the magnetic flux leakage signal, and obtain the corrosion location based on the location information associated with the magnetic flux leakage signal. S3. Based on video information, obtain the flattening deformation amount and perforation diameter of the composite pipe through a defect identification model, and obtain the flattening position and perforation position of the composite pipe based on the location information associated with the video information. S4. Based on the mileage recorded by the fiber optic cable winch, the corrosion location, flattening location, and perforation location are corrected to obtain the corrected corrosion location, corrected flattening location, and corrected perforation location. S5. When the corrosion depth, flattening deformation amount, and perforation diameter are obtained, the trigger signal transmitter emits a defect signal. The operator determines the excavation location based on the defect signal received by the positioning receiver and the corrected corrosion location, flattening location, and perforation location.

2. The method for detecting defects in small-diameter non-metallic composite pipes according to claim 1, characterized in that, The method of detecting metal connectors based on magnetic flux leakage signals includes: The magnetic flux leakage signal includes four magnetic flux leakage probe signals; the metal connector is identified based on the four magnetic flux leakage probe signals through a majority voting mechanism.

3. The method for detecting defects in small-diameter non-metallic composite pipes according to claim 2, characterized in that, The method of identifying metal connectors based on the majority voting mechanism using signals from four magnetic flux leakage probes includes: The fluctuation amplitude of each magnetic flux leakage probe signal is determined to be greater than the threshold. If the result is that the fluctuation amplitude of at least three magnetic flux leakage probe signals is greater than the threshold, then a metal connector is detected; if the result is that the fluctuation amplitude of no at least three magnetic flux leakage probe signals is greater than the threshold, then no metal connector is detected.

4. The method for detecting defects in small-diameter non-metallic composite pipes according to claim 1, characterized in that, The step of obtaining the corrosion depth based on the leakage magnetic field signal using the Hanning window fast Fourier transform algorithm includes: The time-domain signal is obtained from the leakage magnetic field signal. The time-domain signal is truncated or padded with zeros to make the length of the time-domain signal the same as the length of the Hanning window. The time-domain signal is multiplied point by point by the Hanning window to obtain the windowed time-domain signal. The windowed time-domain signal is converted into a frequency domain spectrum by Fast Fourier Transform (FFT). The frequency domain spectrum is analyzed to extract the characteristic frequency offset. The corrosion depth is obtained based on the characteristic frequency offset.

5. The method for detecting defects in small-diameter non-metallic composite pipes according to claim 4, characterized in that, The formula for calculating the corrosion depth is: ， Where d is the corrosion depth, Δf is the characteristic frequency offset, and k is the calibration coefficient.

6. The method for detecting defects in small-diameter non-metallic composite pipes according to claim 1, characterized in that: The defect identification model is a YOLOv8 model that is optimized and updated using a focus loss function.

7. The method for detecting defects in small-diameter non-metallic composite pipes according to claim 1, characterized in that, The formula for calculating the flattening deformation is: ， Where T is the amount of flattening deformation, D1 is the standard diameter of the composite pipe, and D2 is the measured diameter of the flattened part of the composite pipe.

8. The method for detecting defects in small-diameter non-metallic composite pipes according to claim 1, characterized in that, It also includes a data preprocessing step, which includes data cleaning, filtering and standardization.

9. A defect detection system for small-diameter non-metallic composite pipes, characterized in that: The system includes an internal detection device, a computing device, a fiber optic signal modulator, a fiber optic cable reel, a nylon cable reel, and a positioning receiver. The internal detection device sequentially includes a camera locator, a data video acquisition unit, a power supply, a magnetic flux leakage sensor, a data processor, a signal transmitter, and a fiber optic transceiver. The internal detection device is placed inside a composite tube. The computing device, the fiber optic signal modulator, and the fiber optic cable reel are located outside one port of the composite tube, and the nylon cable reel is located outside the other port of the composite tube. The camera locator is connected to the nylon cable in the nylon cable reel. The fiber optic transceiver is connected to one end of the fiber optic cable in the fiber optic cable reel, and the other end of the fiber optic cable is connected to the input end of the fiber optic signal modulator. The output end of the fiber optic signal modulator is connected to the input end of the computing device. The fiber optic cable reel is connected to the computing device, and the positioning receiver is wirelessly connected to the signal transmitter.

10. A defect detection system for small-diameter non-metallic composite pipes according to claim 9, characterized in that: The magnetic flux leakage sensor is an array-type Hall sensor, which includes four magnetic flux leakage probes arranged in pairs overlapping each other at 30°.

Images