Traction rod eddy current automatic detection method based on multiple mechanical arms

Abstract

The invention discloses a traction rod eddy current automatic detection method based on multiple mechanical arms. The method comprises the following steps: setting a scanning path of a traction rod and a preset threshold value of a defect amplitude value of the traction rod to be detected in a host; conveying the traction rod to be tested to a designated position; the automatic detection system is started for detection, the detection system comprises a detection mechanical arm, a carrying mechanical arm, a 3D structured light camera and a high-frequency eddy current instrument, the carrying mechanical arm is used for carrying the to-be-detected traction rod to a detection station, the detection mechanical arm carries an eddy current probe to scan the to-be-detected traction rod according to a scanning path, and the carrying mechanical arm carries the 3D structured light camera; the three-dimensional scanning module is used for performing three-dimensional point cloud scanning on a to-be Judging the defect existence of the traction rod area under the scanning path, and completing defect positioning and size analysis of the traction rod; and repeating the operation to complete defect detection of each batch of traction rods to be detected. According to the invention, subjective errors are reduced, and full-automatic high-precision detection of the complex curved surface workpiece is realized.

Description

Technical Field

[0001] The present invention relates to the technical field of non-destructive testing of traction rods, and in particular to an automatic eddy current testing method for traction rods based on multiple mechanical arms. Background Art

[0002] During high-speed train operation, the core load-bearing component connecting the car body and bogie is the traction rod of the high-speed train bogie. This component plays a key role in transmitting traction, braking force, and lateral loads, and typically requires meticulous daily maintenance and regular inspection. Made of high-strength alloy steel such as 42CrMo, it comprises the rod body, thin-walled end tube surface, and welded connection areas. During train operation, the traction rod is subjected to long-term alternating loads due to the train's starts, stops, accelerations, and decelerations. This can easily cause tiny fatigue cracks to develop in the heat-affected zone of the weld, the thin-walled end tube surface, and the threaded connection, which are difficult to detect with the naked eye. If these cracks are not detected promptly, they can cause the traction rod to break, leading to bogie instability or even train derailment, resulting in casualties and a serious threat to driving safety. Therefore, regular nondestructive testing of traction rods is a core component of the high-speed rail operation and maintenance system.

[0003] Currently, existing technologies for inspecting traction rods primarily rely on magnetic particle testing, penetrant testing, ultrasonic testing, and single-point eddy current testing, but all have significant drawbacks. Magnetic particle testing involves spraying a magnetic suspension and applying a magnetic field, using magnetic traces to reveal surface and near-surface defects. However, magnetic particle testing and penetrant testing require pre-treatment of the traction rod surface by stripping, polishing, and cleaning, which can take several hours and generate chemical waste. This violates the principles of green maintenance and can only detect large, open surface defects, insensitive to tiny cracks beneath the paint layer (typically less than 0.5mm in depth). Furthermore, ultrasonic testing relies on manual interpretation, resulting in poor repeatability and inability to achieve automated testing. While ultrasonic testing can identify internal defects, it suffers from poor coupling with complex curved surfaces, making it inefficient for inspecting irregular surfaces. Furthermore, its low scanning speed (typically less than 50mm / s) makes it difficult to meet the needs of large-scale inspections. Although single-point eddy current testing can penetrate non-conductive paint layers, it relies on manual operation and is inefficient. Path consistency also relies on operator experience, and the probe is sensitive to lift-off effects, often leading to fluctuations in paint layer thickness and signal drift. It lacks multi-dimensional data fusion capabilities and cannot distinguish between crack depth and tilt angle. Summary of the Invention

[0004] The present invention proposes an eddy current automatic detection method based on the collaboration of multiple robotic arms, aiming to completely eliminate the paint stripping pretreatment link through eddy current technology and realize fully automatic and high-precision detection of complex curved surface workpieces.

[0005] A method for automatically detecting eddy currents of a traction rod based on multiple robotic arms comprises the following steps: Step 1: Set the scanning path of the traction rod and the preset threshold value of the traction rod defect amplitude to be tested in the host; Step 2: Guide the traction rod to be tested to the designated position; Step 3: Start the automatic inspection system for inspection. The inspection system includes an inspection robot arm, a transport robot arm, a 3D structured light camera, and a high-frequency eddy current instrument. The transport robot arm is used to transport the traction rod to be tested to the inspection station. The inspection robot arm carries an eddy current probe to scan the traction rod according to the scanning path. The 3D structured light camera is used to perform a three-dimensional point cloud scan of the traction rod. Step 4: Determine the presence of defects in the traction rod area along the scanning path, and complete defect location and size analysis of the traction rod; Step 5: Repeat the above steps to complete defect detection of each batch of traction rods to be tested.

[0006] Furthermore, the step 1 includes: In advance, in the host computer of the detection control and data processing unit, the scanning path of the traction rod is compiled according to the 3D model of the traction rod, and the scanning starting point, scanning end point and each point in the scanning path are determined; The traction rod artificial defect test block is placed on the traction rod artificial defect test block placement platform, and is inspected using a detection robot arm carrying an eddy current probe. The amplitude of the corresponding defect is determined based on the detection results, and the amplitude is set as a preset threshold and input into the high-frequency eddy current analysis software.

[0007] Furthermore, the step 2 includes: Place the entire batch of traction rods on the workpiece conveying unit, which automatically transports the traction rods to the designated testing position.

[0008] Wherein, the designated detection position is located within the working range of the transport robot arm.

[0009] Furthermore, the step three includes: The automatic detection system is started by the host computer to perform self-test procedures and initialization operations on the detection robot arm, handling robot arm, 3D structured light camera and high-frequency eddy current instrument; The detection robot automatically grabs the eddy current probe and moves to the standby area near the detection station. The transport robot carries the mechanical gripper and moves to the designated detection position above the workpiece conveying unit and enters the standby state. The detection station is arranged between the detection robot arm and the transport robot arm, and the standby area is a safe location near the detection station to avoid collision and interference between the detection robot arm and the transport robot arm; The 3D structured light camera is used to perform 3D point cloud scanning on the traction rod, and its spatial coordinates and posture data are analyzed based on the coordinate transformation algorithm, and the data is transmitted to the host. The transport robot grabs the traction rod according to the analyzed data and moves to the inspection station. Then the inspection robot carries the eddy current probe to the preset scanning starting point, starts scanning according to the pre-compiled and debugged robot scanning path, and completes the real-time acquisition and AD conversion of the detection signal through the high-frequency eddy current meter until the scan reaches the preset scanning end point, completing the scanning process.

[0010] Furthermore, the step 4 includes: Through the detection control and data processing unit, eddy current imaging technology is applied to the traction rod area under the scanning path of the detection robot arm. Based on the acquired impedance and phase difference, combined with the preset threshold, the existence of defects is determined, and the defect features are mapped to the 3D model of the rod through image fusion technology to realize the defect location and size analysis of the traction rod.

[0011] Furthermore, the step five includes: After the inspection is completed, the transport robot arm puts it back to its original position and returns to the state to be grasped. The host controls the traction rod to position the tooling conveying mechanism, and drives the downward traction rod to move to the bottom of the transport robot arm. The transport robot arm starts the grasping operation. The above process is repeated to obtain the inspection data of each traction rod in turn.

[0012] Furthermore, during the detection process, high-frequency eddy current technology is used to penetrate the non-conductive paint layer to obtain metal substrate defect signals.

[0013] Furthermore, a 3D structured light camera is used to perform a three-dimensional point cloud scan of the traction rod. The specific process is as follows: the spatial coordinates and posture data of the traction rod in the 3D structured light camera coordinate system are obtained and transmitted to the host in a timely manner; the host applies a coordinate transformation algorithm based on the received spatial coordinates and posture data in the 3D structured light camera coordinate system, converts them into coordinates and postures in the coordinate system of the transport robot arm, and sends them to the transport robot arm to control the transport robot arm to accurately grasp the traction rod and move it to the inspection station.

[0014] Furthermore, the detection control and data processing unit applies eddy current imaging technology to the traction rod area within the detection robot's scanning path. Based on the acquired impedance and phase difference, combined with a preset threshold, the presence of defects is determined. The specific process is as follows: the detection robot carries an eddy current probe to scan the detection area. The high-frequency eddy current instrument completes real-time acquisition and A / D conversion of the detection signal. The high-frequency eddy current instrument sends the acquired signal to the host computer, which processes the signal through the high-frequency eddy current software. After processing, the signal is displayed as a waveform on the software's A-scan interface. At the same time, the signal amplitude is compared with the preset threshold. Parts exceeding the preset threshold are determined to be defect signals and mapped to the three-dimensional model for defect imaging. The acquired eddy current signal is fused with the position information of the points passed by the probe end. When the detection robot performs the scan, the coordinates of the points currently scanned by the probe are sent to the host computer at a fixed frequency. The high-frequency eddy current software also receives it. The software's C-scan imaging interface displays the traction rod's three-dimensional model. The received points are displayed in real time on the surface of the traction rod's 3D model. The points are marked by color changes, where areas without defects are displayed in blue and areas with defects are marked in red.

[0015] Furthermore, the process of marking the defect is as follows: the impedance of the eddy current detection coil is in complex form, including resistance R and reactance X : Z = R + JX = R + j ( ωL -1 / ωC ) in, ω is the angular frequency, L is the inductor, C is the capacitance. Phase angle θ =arctan( X / R ) reflects the phase difference between voltage and current and is the core analysis object of the model. When a material defect exists, the eddy current path changes, causing a redistribution of the magnetic field, and the amplitude and phase of the coil impedance change accordingly. This phase difference is directly related to the defect depth, shape, and material conductivity. The applied high-frequency eddy current probe distinguishes the phase of the eddy current signal at the defect from that of the signal in a non-defective area. If the defect signal exceeds a preset threshold, the probe will mark it as a defect.

[0016] Compared with the prior art, the advantages and positive effects of the present invention are: (1) High detection efficiency. The system uses multiple robotic arms to work together. The transporting robotic arm quickly grabs and transfers the traction rod to be detected, and the detection robotic arm synchronously detects the traction rod to be detected. This greatly shortens the detection time of a single traction rod and realizes continuous detection of the entire batch of traction rods, thereby improving the overall detection efficiency. (2) High precision in micro-defect detection. High-frequency eddy current technology is used in the traction rod scanning process, and the traction rod special array eddy current probe can penetrate the non-conductive paint layer to obtain the metal matrix defect signal, and can accurately detect the micro defects on the surface and near the surface of the traction rod without the need for pre-treatment of the paint layer. At the same time, by pre-detecting the artificial defect test block of the traction rod and setting the preset threshold, combined with eddy current imaging technology, impedance and phase difference and image fusion technology, the existence of defects can be accurately determined and high-precision defect positioning and size analysis can be achieved, effectively ensuring the accuracy and reliability of the detection results; (3) High degree of automation. The entire testing process does not require direct human participation in the testing operation link. Only auxiliary work such as equipment startup and material loading is required, which reduces the labor intensity of workers; (4) The tooling has strong adaptability and guaranteed maintenance. The tooling uses multiple robotic arms, and the mechanical grippers can match different types such as two-finger pneumatic mechanical grippers and three-finger pneumatic mechanical grippers, which can adapt to traction rod workpieces of different diameters, thereby improving the adaptability of the tooling to diversified products. At the same time, the multi-robotic arm collaborative control unit adopts the manufacturer's own collaborative control algorithm, which is convenient for equipment debugging and maintenance, and to a certain extent, the working parameters and processes of the robotic arms can be flexibly adjusted according to actual production needs, thereby enhancing the versatility of the equipment; (5) The positioning accuracy of the traction rod to be tested is high. The traction rod placement slots and traction rod positioning fixtures of the workpiece conveying unit achieve precise conveying and positioning of the workpiece. The 3D structured light camera obtains the accurate position information of the traction rod through three-dimensional point cloud scanning before grasping, making the inspection process more orderly and accurate. The inspection control and data processing unit centrally processes and analyzes the inspection data, facilitating the company's comprehensive and systematic control of product quality, and promptly identifying and improving quality problems in the production process. BRIEF DESCRIPTION OF THE DRAWINGS

[0017] Figure 1 This is a schematic diagram of the structure of the traction rod eddy current automatic detection tooling; Figure 2 This is a flow chart of the automatic eddy current detection method for traction rods based on multiple robotic arms.

[0018] In the figure: 1. Robotic arm collaborative control unit; 2. Workpiece conveying unit; 3. Detection control and data processing unit; 101. Handling robot arm; 102. Detection robot arm; 103. Pneumatic quick-change plate; 201. Traction rod placement hole slot; 202. Traction rod positioning and placement tooling; 203. Traction rod positioning and placement tooling conveying mechanism; 3. Detection control and data processing unit; 301. Control box. Specific embodiments like Figure 1 As shown, the traction rod eddy current automatic detection tooling mainly includes a robot arm collaborative control unit 1, a detection control and data processing unit 3 and a workpiece conveying unit 2. The robot arm collaborative control unit 1 includes a detection robot arm 102 and a handling robot arm 101. The detection robot arm 102 detects the traction rod through an eddy current probe. The workpiece conveying unit 2 is arranged on one side of the handling robot arm 101. The handling robot arm 101 is arranged in the area between the detection robot arm 102 and the workpiece conveying unit 2. The workpiece conveying unit 2 is used for conveying the traction rod.

[0020] The workpiece conveying unit 2 includes a traction rod placement hole slot 201, a traction rod positioning and placement tool 202, and a traction rod positioning and placement tool conveying mechanism 203. A plurality of evenly arranged traction rod positioning and placement toolings 202 are fixedly provided on the conveyor belt of the traction rod positioning and placement tool conveying mechanism 203. The traction rod positioning and placement tooling 202 is provided with a traction rod placement hole slot 201. The traction rod is placed in the traction rod placement hole slot 201. The clamping tool unit can grasp the traction rod in the traction rod placement hole slot 201. The workpiece conveying unit 2 is arranged within the working range of the handling robot arm 101. The detection control and data processing unit 3 includes a control box 301 and a high-frequency eddy current meter. The control box 301 includes a host and a central console. The central console is connected to the host. The high-frequency eddy current meter and the host are both arranged in the control box 301. The high-frequency eddy current meter is electrically connected to the eddy current probe. The high-frequency eddy current meter is equipped with high-frequency eddy current analysis software. The host is used to receive and process data from the detection robot 102, the handling robot 101, the eddy current probe and the 3D structured light camera.

[0021] like Figure 2 As shown, when using a multi-manipulator-based traction rod eddy current automatic detection tool to perform eddy current automatic detection on a traction rod to be tested, the following steps are included: (1) In the host computer of the detection control and data processing unit, a scanning path for the traction rod must be compiled in advance based on the 3D model of the traction rod, and the starting point and end point as well as the key points on the path must be clearly set; (2) Place the traction rod artificial defect test block on the traction rod artificial defect test block placement platform, operate the detection robot arm carrying the eddy current probe to detect it, accurately determine the amplitude of the corresponding defect based on the detection results, and use the amplitude as a preset threshold to input it into the high-frequency eddy current analysis software; (3) Place the entire batch of traction rods to be inspected neatly on the workpiece conveying unit, automatically transport the traction rods to the designated inspection position, start the automatic inspection system through the host, and quickly start the self-inspection program and initialization operation of the inspection robot arm, handling robot arm, 3D structured light camera and high-frequency eddy current meter; The self-test program usually checks whether the working status of each part of the tooling structure is normal, such as whether the mechanical arm, handling mechanical arm, high-frequency eddy current meter, 3D structured light camera and other components are electrically connected normally and in a start-up state. If they are in an abnormal state, such as if the missing detection mechanical arm or 3D structured light camera is detected to be abnormal, a fault prompt will pop up. Initialization clears the parameters set in the last operation and restores them to the default state.

[0022] (4) After the self-test is completed, the detection robot automatically grabs the eddy current probe and moves smoothly to the safe standby area near the detection station; at the same time, the transport robot carries the mechanical gripper and moves to the top of the designated detection position of the workpiece transport unit and enters the standby state; (5) The 3D structured light camera starts working, performs a three-dimensional point cloud scan on the traction rod, uses the coordinate transformation algorithm to quickly analyze its spatial coordinates and posture data, and transmits them to the host in a timely manner; the specific process of performing a three-dimensional point cloud scan on the traction rod is as follows: obtaining the spatial coordinates and posture data of the traction rod in the 3D structured light camera coordinate system, and transmitting them to the host in a timely manner; the host applies the coordinate transformation algorithm based on the spatial coordinates and posture data received in the 3D structured light camera coordinate system, converts them into coordinates and postures in the coordinate system of the transport robot, and sends them to the transport robot, controlling the transport robot to accurately grasp the traction rod and move it to the inspection station.

[0023] (6) The transport robot arm accurately grabs the traction rod and transfers it to the inspection station based on the analytical data received by the host. Then the inspection robot arm carries the eddy current probe and moves to the preset scanning starting point, and starts scanning according to the pre-programmed scanning path. During the scanning process, the high-frequency eddy current instrument completes the real-time acquisition and AD conversion of the detection signal until the preset scanning end point is scanned. The detection control and data processing unit processes the data of the traction rod area under the scanning path of the inspection robot arm, and uses eddy current imaging technology to determine the existence of defects based on the acquired impedance and phase difference combined with the preset threshold. The specific process is that the inspection robot arm carries the eddy current probe to scan the inspection area, and the high-frequency eddy current instrument completes the real-time acquisition and AD conversion of the detection signal. The high-frequency eddy current instrument sends the collected signal to the host, and processes the signal through the high-frequency eddy current software. After the processing is completed, the signal is displayed as a waveform on the A-scan interface of the software. At the same time, the signal amplitude is compared with the preset threshold. The part exceeding the preset threshold is judged as a defect signal and mapped to the three-dimensional model for defect imaging. The collected eddy current signals are integrated with the information of the points passed by the probe end. When the detection robot arm performs scanning, the coordinates of the points currently scanned by the probe are sent to the host at a fixed frequency and are also received by the high-frequency eddy current software. The three-dimensional model of the traction rod will be displayed in the C-scan imaging interface of the software. The surface of the three-dimensional model of the traction rod will display the received points in real time, and the points are marked by color changes. The parts without defects are displayed in blue, and the parts with defects are marked in red.

[0024] (7) After completing the inspection of a traction rod, the transport robot arm puts it back to its original position and then returns to the state to be grasped. The host controls the traction rod positioning and placement tooling conveying mechanism to work, and drives the traction rod to move to the bottom of the transport robot arm. The transport robot arm starts the grasping operation. The above process is repeated to obtain the inspection data of each traction rod in turn, thereby completing the automated, efficient and accurate inspection of the entire batch of traction rods.

[0025] The preferred embodiments of the present invention disclosed above are intended only to help illustrate the present invention. These preferred embodiments do not exhaustively describe all details, nor do they limit the present invention to specific embodiments. Obviously, many modifications and variations are possible based on the contents of this specification. These embodiments are selected and described in detail in this specification to better explain the principles and practical applications of the present invention, thereby enabling those skilled in the art to better understand and utilize the present invention. The present invention is limited only by the claims and their full scope and equivalents.

Claims

1. A method for automatic eddy current detection of a traction rod based on multiple robotic arms, characterized by: The following steps are involved: Step 1: Set the scanning path of the traction rod and the preset threshold value of the traction rod defect amplitude to be tested in the host; Step 2: transport the traction rod to be tested to the designated position; Step 3: Start the automatic detection system for detection. The detection system includes a detection robot arm, a transport robot arm, a 3D structured light camera and a high-frequency eddy current instrument. The transport robot arm is used to transport the traction rod to be tested to the detection station. The detection robot arm carries an eddy current probe to scan the traction rod according to the scanning path. The 3D structured light camera is used to perform a three-dimensional point cloud scan of the traction rod. Step 4: Determine the presence of defects in the traction rod area along the scanning path, and complete defect location and size analysis of the traction rod; Step 5: Repeat the above steps to complete defect detection of each batch of traction rods to be tested.

2. The method for automatic eddy current detection of a traction rod based on multiple robotic arms according to claim 1, characterized in that: The step one comprises: In advance, in the host computer of the detection control and data processing unit, the scanning path of the traction rod is compiled according to the 3D model of the traction rod, and the scanning starting point, scanning end point and each point in the scanning path are determined; The traction rod artificial defect test block is placed on the traction rod artificial defect test block placement platform, and is inspected using a detection robot arm carrying an eddy current probe. The amplitude of the corresponding defect is determined based on the detection results, and the amplitude is set as a preset threshold and input into the high-frequency eddy current analysis software.

3. The method for automatic eddy current detection of a traction rod based on multiple manipulators according to claim 1, characterized in that: The second step includes: Place the entire batch of traction rods on the workpiece conveying unit, and automatically transport the traction rods to the designated testing position; Wherein, the designated detection position is located within the working range of the transport robot arm.

4. The method for automatic eddy current detection of a traction rod based on multiple manipulators according to claim 1, characterized in that: The step three includes: The automatic detection system is started by the host computer to perform self-test procedures and initialization operations on the detection robot arm, handling robot arm, 3D structured light camera and high-frequency eddy current instrument; The detection robot automatically grabs the eddy current probe and moves to the standby area near the detection station. The transport robot carries the mechanical gripper and moves to the designated detection position above the workpiece conveying unit and enters the standby state. The detection station is arranged between the detection robot arm and the transport robot arm, and the standby area is a safe location near the detection station to avoid collision and interference between the detection robot arm and the transport robot arm; The 3D structured light camera is used to perform 3D point cloud scanning on the traction rod, and its spatial coordinates and posture data are analyzed based on the coordinate transformation algorithm, and the data is transmitted to the host. The transport robot grabs the traction rod according to the analyzed data and moves to the inspection station. Then the inspection robot carries the eddy current probe to the preset scanning starting point, starts scanning according to the pre-compiled and debugged robot scanning path, and completes the real-time acquisition and AD conversion of the detection signal through the high-frequency eddy current meter until the scan reaches the preset scanning end point, completing the scanning process.

5. The method for automatic eddy current detection of a traction rod based on multiple robotic arms according to claim 1, characterized in that: The fourth step includes: Through the detection control and data processing unit, eddy current imaging technology is applied to the traction rod area under the scanning path of the detection robot arm. Based on the acquired impedance and phase difference, combined with the preset threshold, the existence of defects is determined, and the defect features are mapped to the 3D model of the rod through image fusion technology to realize the defect location and size analysis of the traction rod.

6. The method for automatic eddy current detection of a traction rod based on multiple manipulators according to claim 1, characterized in that: The step five includes: After the inspection is completed, the transport robot arm puts it back to its original position and returns to the state to be grasped. The host controls the traction rod to position the tooling conveying mechanism, and drives the downward traction rod to move to the bottom of the transport robot arm. The transport robot arm starts the grasping operation. The above process is repeated to obtain the inspection data of each traction rod in turn.

7. The method for automatic eddy current detection of a traction rod based on multiple robotic arms according to claim 1, characterized in that: During the detection process, high-frequency eddy current technology is used to penetrate the non-conductive paint layer to obtain metal substrate defect signals.

8. The method for automatic eddy current detection of a traction rod based on multiple robotic arms according to claim 4, characterized in that: The 3D point cloud scanning of the traction rod is performed by using a 3D structured light camera. The specific process is as follows: The spatial coordinates and posture data of the traction rod in the 3D structured light camera coordinate system are obtained and transmitted to the host in a timely manner. The host applies a coordinate transformation algorithm based on the spatial coordinates and posture data received in the 3D structured light camera coordinate system, converts them into coordinates and postures in the coordinate system of the transport robot arm, and sends them to the transport robot arm, controlling the transport robot arm to accurately grasp the traction rod and move it to the inspection station.

9. The method for automatic eddy current detection of a traction rod based on multiple robotic arms according to claim 5, characterized in that: The detection control and data processing unit applies eddy current imaging technology to the traction rod area under the scanning path of the detection robot arm. Based on the acquired impedance and phase difference, the presence of defects is determined in combination with a preset threshold. The specific process is as follows: the detection robot arm carries an eddy current probe to scan the detection area, and the high-frequency eddy current instrument completes real-time acquisition and AD conversion of the detection signal. The high-frequency eddy current instrument sends the acquired signal to the host, and the signal is processed by the high-frequency eddy current software. After the processing is completed, the signal is displayed as a waveform on the software A-scan interface. At the same time, the signal amplitude is compared with the preset threshold. The part exceeding the preset threshold is determined as a defect signal and mapped to the three-dimensional model for defect imaging; the acquired eddy current signal is integrated with the point information passed by the probe end. When the detection robot arm performs scanning, the coordinates of the point currently scanned by the probe are sent to the host at a fixed frequency and are also received by the high-frequency eddy current software. The three-dimensional model of the traction rod is displayed on the C-scan imaging interface of the software. The received points are displayed in real time on the surface of the three-dimensional model of the traction rod, and the points are marked by color changes, where the parts without defects are displayed in blue and the parts with defects are marked in red.

10. The method for automatic eddy current detection of a traction rod based on multiple robotic arms according to claim 9, characterized in that: The specific process of marking the defect is as follows: the impedance Z of the eddy current detection coil is: Z = R + JX = R + j ( ωL -1 / ωC ) Where R is resistance, X is reactance, ω is the angular frequency, L is the inductor, C is the capacitor; Phase Angle θ =arctan( X / R ); When there is a defect in the material, the high-frequency eddy current probe distinguishes the phase of the eddy current signal at the defect and the signal at the non-defective area. If the signal at the defect exceeds the preset threshold, it will be marked as a defect.

Images