A flange automatic magnetic leakage detection system and equipment

Abstract

The present invention discloses an automated magnetic flux leakage detection system and equipment for flanges. An automated magnetic flux leakage detection equipment for flanges comprises a frame, a ring bracket is arranged in the middle of the frame, and a plurality of magnetic sensors are arranged in the ring bracket; a rotating shaft is fixedly connected to one side of the ring bracket; a driven wheel is fixedly sleeved in the middle of the rotating shaft, and a transmission belt is arranged on the driven wheel; a driving wheel is matched at one end of the transmission belt, and a driving shaft is passed through the middle of the driving wheel; a stepping motor is connected to one end of the driving shaft; a support is fixedly connected to the driving shaft, and a ring coil is arranged inside the support; an automated magnetic flux leakage detection system for flanges is applied to automated magnetic flux leakage detection equipment for flanges. The present invention has the following beneficial effects: the automated magnetic flux leakage detection system and equipment for flanges can adjust the detection direction of the magnetic sensor according to the magnetization direction, ensure that the detection direction of the sensor is at the best measurement angle, and accurately judge and evaluate defects through algorithm processing, thereby improving the accuracy of defect judgment.

Description

Technical Field

[0001] The invention relates to the field of flange nondestructive testing, and in particular to an automatic flange magnetic leakage testing system and equipment thereof. Background Art

[0002] Flanges are important connecting components in industrial piping systems, and their quality is directly related to the safety and reliability of the entire system. Magnetic flux leakage detection technology is one of the most widely used non-destructive testing methods. It detects the changes in the magnetic field leakage on the flange surface to find defects inside the material. Existing flange magnetic flux leakage detection mainly uses single-direction magnetization and fixed-angle detection.

[0003] The existing flange magnetic flux leakage detection process usually includes the following steps: first, the flange is magnetized by an excitation device; then the magnetic field intensity distribution on the flange surface is detected by a magnetic sensor; finally, the collected magnetic field signal is analyzed by a signal processing system to determine whether there is a defect. This detection method is simple to operate and has been widely used in practical applications.

[0004] However, existing magnetic flux leakage detection systems have obvious technical limitations. The main problem is that when the magnetic flux lines are parallel to the defect, the magnetic flux lines will not be significantly distorted at the defect, resulting in extremely weak leakage magnetic field signals, making it difficult for the system to detect such defects. This phenomenon often occurs in actual detection, seriously affecting the reliability and integrity of the detection.

[0005] In order to solve this problem, the flange needs to be magnetized in multiple directions, that is, the angle of the magnetic flux lines of the ring coil needs to be changed so that the magnetic flux lines can intersect with defects in various directions. However, this method brings new technical challenges: the measurement accuracy of the magnetic sensor will change significantly with the change of the measurement angle. When the detection direction of the sensor deviates from the optimal angle, its sensitivity will decrease, the signal-to-noise ratio will decrease, and the accuracy of the detection results will be affected.

[0006] In addition, during the multi-directional magnetization process, the magnetization intensity at different positions and directions may be different, resulting in inconsistent detection sensitivity. The mutual influence of magnetic fields in multiple directions will produce a complex magnetic field superposition effect, which increases the difficulty of signal processing. At the same time, the traditional system also has shortcomings in terms of automation. Operators need to manually adjust the magnetization direction and sensor position, which not only reduces the detection efficiency, but also may introduce human errors. Summary of the invention

[0007] The purpose of the present invention is to provide a flange automatic magnetic flux leakage detection system and equipment thereof. The flange automatic magnetic flux leakage detection system and equipment thereof can adjust the detection direction of the magnetic sensor according to the magnetization direction, ensure that the detection direction of the sensor is at the best measurement angle, and accurately judge and evaluate defects through algorithm processing, thereby improving the accuracy of defect judgment.

[0008] The above technical objectives of the present invention are achieved through the following technical solutions:

[0009] A flange automatic magnetic leakage detection device comprises a frame, a workbench is arranged in the middle of the frame, and a positioning seat is arranged on the top of the workbench; an annular bracket is arranged on the outer periphery of the positioning seat, and a plurality of magnetic sensors are arranged in a circumferential direction on the inner side of the annular bracket, which is used to detect the magnetic leakage distribution on the flange surface; a rotating shaft is fixedly connected to one side of the annular bracket, and one end of the rotating shaft is connected to the frame through a bearing to realize the rotation movement of the annular bracket; a driven wheel is fixedly sleeved in the middle of the rotating shaft, and a transmission belt is arranged outside the driven wheel; a driving wheel is matched at one end of the transmission belt, and a driving shaft is fixedly penetrated in the middle of the driving wheel to drive the driven wheel to rotate; a stepping motor is connected to one end of the driving shaft, and the rotation angle and speed of the annular bracket are controlled by the stepping motor; the driving shaft is fixedly connected to a support, and an annular coil is arranged inside the support to generate a magnetic field to magnetize the flange; the support and the annular bracket always remain in a parallel state during the rotation process, so that the measuring direction of the magnetic sensor is perpendicular to the direction of the magnetic field generated by the annular coil, so as to improve the detection accuracy.

[0010] The present invention is further configured as follows: two ends of the annular coil are connected with connecting wires, and the connecting wires are connected with a power supply component for energizing the annular coil.

[0011] A flange automatic magnetic flux leakage detection system, applied to flange automatic magnetic flux leakage detection equipment, comprises the following steps:

[0012] S1: Signal processing stage: including collecting the original signal generated by the magnetization of the ring coil, eliminating high-frequency noise through average filtering, correcting the baseline drift by polynomial fitting, and implementing adaptive signal gain adjustment;

[0013] S2: Defect feature processing stage: calculate the first-order and second-order derivative characteristic parameters of the signal, confirm the defect by setting the adaptive threshold, use polar coordinate transformation to achieve precise positioning, and quantitatively estimate the defect size;

[0014] S3: Data processing and evaluation stage: The information from each detection angle is integrated through multi-angle data fusion technology, and the weight coefficient is calculated for weighted fusion. Finally, the credibility of the detection results is evaluated and the detection evaluation conclusion is given.

[0015] The present invention is further configured as follows: the signal processing stage includes: original signal acquisition: when the annular coil magnetizes the flange at different angles, the magnetic sensor will collect a leakage magnetic signal S(θ, t) containing defect information, and an interference signal N(θ, t) will also be mixed in. The original signal containing the effective signal and noise is obtained through the formula S(θ, t)=M(θ, t)+N(θ, t).

[0016] The present invention is further configured as follows: the signal processing stage also includes: signal filtering processing: the support and the annular bracket are flipped for a cycle, and the formula S'(θ,t)=(1 / n)∑[S(θ,ti)] is used for average filtering to eliminate high-frequency noise; baseline drift correction: the baseline is obtained by using a polynomial fitting method: B(θ,t)=a0+a1t+a2t²+...+a n t n ; Correct the signal baseline using the formula Sc(θ,t)=S'(θ,t)-B(θ,t).

[0017] The present invention is further configured as follows: the signal processing stage further includes: signal gain adaptive adjustment: G(θ)=V ref / max|Sc(θ,t)|; realize signal gain correction. The signal formula after gain correction is Sa(θ,t)=G(θ)·Sc(θ,t), thereby adjusting the signal amplitude.

[0018] The present invention is further configured as follows: the defect feature processing stage includes:

[0019] Feature extraction: Calculate the first-order derivative of the signal separately , second-order derivative ; and peak characteristics: P max (θ)=max[Sa(θ,t)];P min (θ)=min[Sa(θ,t)]; ΔP(θ)=P max (θ)-P min (θ) is used to characterize the defect characteristics; defect judgment: the flange defect judgment standard is set according to the adaptive threshold Th(θ)=k·σ(θ); when the three conditions of ΔP(θ)>Th(θ), |D1(θ,t)|>D1th, and |D2(θ,t)|>D2th are met, the defect is confirmed to exist.

[0020] The present invention is further configured as follows: the defect feature processing stage also includes: defect positioning: for the detected defect point, the defect position is determined by the polar coordinate conversion formula X=R·cos(θ), Y=R·sin(θ); defect size estimation: the defect length is calculated using L=v·Δt, and the defect depth is estimated by the formula H=α·ΔP(θ).

[0021] The present invention is further configured as follows: the data processing evaluation stage includes:

[0022] Multi-angle data fusion: for the same defect at different angles θ i The obtained eigenvalues ​​are weighted fused: F=∑[w i ·f(θ i )] / ∑w i , where w i is the weight coefficient, w i =SNR(θ i ) / ∑SNR(θ i ).

[0023] The present invention is further configured as follows: the data processing evaluation stage further includes: defining a credibility index: C = (ΔP-Th) / Th·SNR / SNR ref ; When C>C th , it is determined to be a reliable detection result.

[0024] In summary, the present invention has the following beneficial effects:

[0025] The flange automatic magnetic flux leakage detection equipment adopts a rotatable ring bracket and a ring coil combination structure to achieve all-round and multi-angle detection of the flange. The stepper motor drives the ring bracket to rotate through the transmission mechanism, so that the magnetic sensor can scan the flange at different angles, while keeping the magnetization direction and the detection direction always perpendicular, significantly improving the detection sensitivity and accuracy, overcoming the detection blind area problem existing in the traditional single-direction magnetization detection, and can effectively identify defects in all directions. The flange automatic magnetic flux leakage detection system adopts a systematic signal processing and analysis method. Starting from the original signal acquisition, it eliminates environmental interference and equipment noise through multi-level filtering and correction, and establishes a complete signal processing process. First, average filtering is used to eliminate high-frequency noise, baseline drift correction is performed through polynomial fitting, and then adaptive gain adjustment is used to normalize signals of different intensities. In the feature extraction link, the system simultaneously considers the first-order derivative, second-order derivative and peak characteristics of the signal, and constructs a multi-dimensional feature space for defect identification. Defect judgment adopts the method of combining adaptive threshold with multiple criteria, which not only ensures the sensitivity of detection, but also effectively reduces the misjudgment rate. Precise positioning is achieved through polar coordinate conversion, and a quantitative relationship between signal characteristics and defect size is established, realizing quantitative evaluation of defects. Multi-angle data fusion technology is adopted, and the flange circumference characteristics are used to improve the detection reliability through weighted fusion. At the same time, a credibility evaluation mechanism is introduced to quantitatively evaluate the detection results. The above process significantly improves the accuracy and reliability of detection, has good environmental adaptability and anti-interference ability, and realizes precise positioning and quantitative evaluation of defects. BRIEF DESCRIPTION OF THE DRAWINGS

[0026] Figure 1This is a schematic diagram of the structure of the flange automatic magnetic leakage detection equipment;

[0027] Figure 2 This is a schematic diagram of the structure of the flange automatic magnetic leakage detection equipment from an upward perspective;

[0028] Figure 3 It is a side view structural diagram of the flange automatic magnetic leakage detection equipment;

[0029] Figure 4 It is a schematic diagram of the detection state of the ring coil and the magnetic sensor;

[0030] Figure 5 for Figure 4 Schematic diagram of the detection state after flipping;

[0031] Figure 6 This is the flow chart of the flange automated magnetic flux leakage detection system.

[0032] Figure numerals: 1. frame; 2. workbench; 3. positioning seat; 4. ring bracket; 5. magnetic sensor; 6. rotating shaft; 7. driven wheel; 8. transmission belt; 9. driving wheel; 10. driving shaft; 11. stepping motor; 12. support; 13. ring coil; 14. connecting line; 15. power supply component; 16. magnetization direction; 17. measurement direction; 18. flange. DETAILED DESCRIPTION

[0033] In the description of the present invention, it should be understood that the terms "center", "longitudinal", "lateral", "length", "width", "thickness", "up", "down", "front", "back", "left", "right", "vertical", "horizontal", "page", "bottom", "inside", "outside", "clockwise", "counterclockwise" and the like indicating orientations or positional relationships are based on the orientations or positional relationships shown in the accompanying drawings, and are only for the convenience of describing the present invention and simplifying the description, and do not indicate or imply that the referred device or element must have a specific orientation, be constructed and operated in a specific orientation, and therefore should not be understood as limiting the present invention.

[0034] In addition, the terms "first" and "second" are used for descriptive purposes only and should not be understood as indicating or implying relative importance or implicitly indicating the number of the indicated technical features. Thus, the features defined as "first" and "second" may explicitly or implicitly include one or more of the features.

[0035] In the description of the present invention, "multiple" means two or more than two, unless otherwise clearly defined. In the present invention, unless otherwise clearly defined and defined, the terms "installed", "connected", "connected", "fixed" and the like should be understood in a broad sense. For example, it can be a fixed connection, a detachable connection, or an integral connection; it can be a mechanical connection or an electrical connection; it can be a direct connection, or it can be indirectly connected through an intermediate medium, or it can be internal communication between two elements. For those of ordinary skill in the art, the specific meanings of the above terms in the present invention can be understood according to the specific circumstances.

[0036] The present invention is further described in detail below in conjunction with the accompanying drawings.

[0037] like Figure 1 As shown, a flange automatic magnetic leakage detection device includes a frame 1. A workbench 2 is arranged in the middle of the frame 1. A positioning seat 3 for placing a flange 18 is arranged on the top of the workbench 2, and the flange 18 is placed on the positioning seat 3 and fixed. The fixing method can be fixing with a fixing member, and the above is the prior art and will not be repeated here.

[0038] like Figure 2 , Figure 3 As shown, in order to achieve the purpose of multi-directional non-destructive detection of internal defects of flange 18, the present invention is provided with an annular bracket 4 at the outer peripheral position of flange 18. A plurality of magnetic sensors 5 are arranged circumferentially on the inner side of the annular bracket 4 corresponding to the position of flange 18. The number of magnetic sensors 5 is set according to the measurable range of the magnetic sensors 5.

[0039] Correspondingly, a rotating shaft 6 is fixedly connected to one side of the annular bracket 4. One end of the rotating shaft 6 is connected to the frame 1 by a bearing. A driven wheel 7 is fixedly sleeved in the middle of the rotating shaft 6. A transmission belt 8 is provided outside the driven wheel 7. A driving wheel 9 is provided at the other end of the transmission belt 8. A driving shaft 10 is fixedly passed through the middle of the driving wheel 9. A stepping motor 11 is fixedly connected to one end of the driving shaft 10. When the stepping motor 11 is working, it drives the driving shaft 10 to rotate in steps, and then drives the driven wheel 7 to rotate through the matching relationship of the transmission belt 8, and then drives the annular bracket 4 to perform step-by-step flipping.

[0040] Through the above arrangement, the measuring orientation of the magnetic sensor 5 can be changed during the flipping of the annular bracket 4, thereby realizing multi-directional non-destructive detection of internal defects of the flange 18.

[0041] In order to achieve the purpose of magnetizing the flange 18. The drive shaft 10 of the present invention is fixedly connected to a support 12 at one end. A ring coil 13 is arranged inside the support 12. Connecting wires 14 are arranged at both ends of the ring coil 13 leading out of the support 12. The connecting wire 14 is connected to a power supply component 15. The power supply component 15 is used to energize the ring coil 13. When the ring coil 13 is energized, the magnetic field generated by the inner circle of the ring coil 13 is directed downward, and then the magnetic flux lines pass through the flange 18, and the flange 18 is magnetized in the corresponding direction (the magnetic field direction is downward).

[0042] In order to prevent the problem that the magnetic flux lines parallel to the internal defects of the flange 18 cannot detect defects, the present invention adopts the above arrangement so that when the stepper motor 11 is working, it can drive the driving shaft 10 to rotate, and then drive the support 12 to flip. During the flipping process of the support 12, the driven wheel 7 is driven to rotate through the transmission belt 8, and then the rotating shaft 6 is driven to rotate, so that the annular bracket 4 rotates by the same angle.

[0043] Under the above settings, if Figure 4 , Figure 5 As shown, the solid arrow direction is the magnetization direction 16 of the flange 18, that is, the direction of the magnetic flux lines, and the hollow arrow is the measurement direction 17 of the magnetic sensor 5. The support 12 and the annular support 4 always remain parallel when rotating. When the direction of the magnetic field generated by the annular coil 13 in the support 12 enters the magnetization direction of the flange 18 for adjustment, the magnetic sensor 5 on the annular support 4 adjusts the direction accordingly, so that the measurement orientation and the magnetic field direction always remain perpendicular, thereby ensuring the accuracy of the measurement data of the magnetic sensor 5 (the accuracy of the magnetic sensor 5 is highest when it is perpendicular to the direction of the magnetic flux lines).

[0044] Through the above arrangement, the internal defects of the flange 18 can be magnetized and magnetic leakage detected in multiple directions within a range of 270 degrees, and in this process the magnetic sensor 5 always maintains a perpendicular detection state to the magnetic flux lines, thereby improving the accuracy of the detection results.

[0045] In order to eliminate the influence of environmental factors on the measurement results and to perform calculation and evaluation on the measurement results. Figure 6 As shown, the present invention also provides a flange automatic magnetic leakage detection system, in which an algorithm chip is set in a power supply component, and includes the following steps:

[0046] S1: Signal processing stage; including collecting the original signal generated by the magnetization of the ring coil, eliminating high-frequency noise through average filtering, correcting the baseline drift by polynomial fitting, and implementing adaptive signal gain adjustment to ensure signal quality and stability.

[0047] S2: Defect feature processing stage; the system calculates the first-order and second-order derivatives of the signal and other characteristic parameters, confirms the defects by setting adaptive thresholds and multiple criteria, uses polar coordinate transformation to achieve precise positioning, and makes quantitative estimates of the defect size.

[0048] S3: Data processing and evaluation stage; through multi-angle data fusion technology, the information of each detection angle is integrated, the weight coefficient is calculated for weighted fusion, and finally the credibility of the detection result is evaluated to give the detection evaluation conclusion.

[0049] 1. Original signal acquisition

[0050] When the annular coil magnetizes the flange at different angles, the magnetic sensor will collect the leakage magnetic signal S(θ, t) containing defect information, and will also be mixed with interference signals N(θ, t) such as equipment vibration and environmental magnetic field fluctuations.

[0051] The original signal collected by the magnetic sensor at the angle θ is expressed as:

[0052] S(θ,t)=M(θ,t)+N(θ,t)

[0053] Among them, M(θ,t) is the effective magnetic leakage signal, and N(θ,t) is the noise signal;

[0054] S(θ,t): the total signal collected at angle θ and time t;

[0055] M(θ,t): effective magnetic flux leakage signal component;

[0056] N(θ,t): interference noise signal component;

[0057] θ: angular position of the sensor with respect to the horizontal plane;

[0058] t: sampling time.

[0059] The original signal acquisition stage can ensure the synchronization of signal sampling and ensure that a complete signal sequence can be obtained at each angular position.

[0060] 2. Signal filtering: First, the support and the annular bracket are flipped for a cycle for average filtering to eliminate high-frequency noise, as follows:

[0061] S'(θ,t)=(1 / n)∑[S(θ,ti)],i=0 to (n-1)

[0062] Where, S'(θ,t): the filtered signal

[0063] n: filter window size, indicating the number of sampling points involved in averaging

[0064] i: index variable of the summation item

[0065] The roughness and rust on the flange surface will generate high-frequency noise. The surface interference can be effectively suppressed by flipping the average filter.

[0066] Among them, the window size n is selected according to actual needs. If it is too large, the edge of the defect signal will be blurred, and if it is too small, the noise suppression effect is not ideal.

[0067] 3. Baseline drift correction:

[0068] (1) Use polynomial fitting method to obtain the baseline:

[0069] B(θ,t)=a0+a1t+a2t²+...+a n t n

[0070] (2) Corrected signal:

[0071] Sc(θ,t)=S'(θ,t)-B(θ,t)

[0072] Where, B(θ,t): baseline function obtained by fitting;

[0073] a0,a1,a2...a n : Polynomial fitting coefficients;

[0074] Sc(θ,t): signal after baseline correction.

[0075] The baseline drift correction solves the problem of baseline drift in the magnetic flux leakage signal due to factors such as flange material inhomogeneity and temperature changes. The polynomial fitting method tracks the slow drift trend without affecting the characteristics of the defect signal.

[0076] 4. Signal gain adaptive adjustment

[0077] G(θ)=V ref / max|Sc(θ,t)|

[0078] Signal after gain correction:

[0079] Sa(θ,t)=G(θ)·Sc(θ,t)

[0080] Where, G(θ): gain coefficient;

[0081] V ref : reference voltage value or standardized amplitude;

[0082] Sa(θ,t): signal after gain correction.

[0083] Signal gain adaptive adjustment is used to address the problem of large differences in signal amplitudes for defects at different locations and depths during flange inspection: adaptive gain adjustment is used to normalize signals of different strengths to the same range, making it easier to process and compare them later. It is mainly used to detect shallow defects.

[0084] 5. Feature extraction

[0085] Compute the first derivative of a signal:

[0086]

[0087] Compute the second derivative of a signal:

[0088]

[0089] Compute the peak characteristics of a signal:

[0090] P max (θ)=max[Sa(θ,t)]

[0091] P min (θ)=min[Sa(θ,t)]

[0092] ΔP(θ)=P max (θ)-P min (θ)

[0093] Where, D1(θ,t): the first-order derivative of the signal, indicating the rate of change of the signal;

[0094] D2(θ,t): The second-order derivative of the signal, which represents the rate of change of the signal’s rate of change;

[0095] P max (θ): maximum value of the signal;

[0096] P min (θ): minimum value of the signal;

[0097] ΔP(θ): Peak value, indicating the amplitude of signal fluctuation.

[0098] As mentioned above, the feature extraction stage shows the characteristic performance of typical defects: among them, the first-order derivative reflects the rate of change of the signal, which helps to detect the edge of crack-like defects; the second-order derivative reflects the curvature of the signal, which is suitable for identifying point defects; the peak feature is mainly used to evaluate the depth information of the defect.

[0099] 6. Defect judgment

[0100] Set adaptive threshold: Th(θ)=k·σ(θ)

[0101] Determine flange defects: When ΔP(θ)>Th(θ), |D1(θ,t)|>D1th, |D2(θ,t)|>D2th, and all three conditions are met, it is determined to be a defect point.

[0102] Where, Th(θ): adaptive threshold;

[0103] k: threshold adjustment coefficient, usually 2-3;

[0104] σ(θ): standard deviation of the angle position signal;

[0105] D1th: first-order derivative threshold;

[0106] D2th: Second-order derivative threshold.

[0107] D1th and D2th are usually determined based on experimental calibration.

[0108] As mentioned above, defect judgment uses multiple criteria to improve detection reliability:

[0109] The amplitude threshold Th(θ) is adaptively adjusted according to the signal fluctuation at the current angle to avoid misjudgment caused by a fixed threshold; the derivative threshold is used to confirm the steepness of the signal change and distinguish between real defects and slow material changes; the combined judgment of multiple conditions can effectively reduce the false alarm rate.

[0110] 7. Defect location

[0111] For the detected defect points, their polar coordinates are marked as:

[0112] X = R·cos(θ)

[0113] Y=R·sin(θ)

[0114] Where R is the flange radius.

[0115] The defect location algorithm uses the specific circular features of the flange to accurately locate the defect position through polar coordinate transformation.

[0116] 8. Defect size estimation

[0117] Defect length estimation: L=v·Δt

[0118] Defect depth estimation: H=α·ΔP(θ)

[0119] Where, L: defect length;

[0120] v: scanning speed;

[0121] Δt: signal duration;

[0122] H: Defect depth;

[0123] α: Depth conversion factor.

[0124] As mentioned above, the defect size estimation establishes the relationship between the leakage magnetic signal characteristics and the actual defect size, estimates the defect length by the signal duration, and estimates the defect depth by the signal amplitude.

[0125] 9. Multi-angle data fusion

[0126] For the same defect at different angles θ i The obtained eigenvalues ​​are weighted fused:

[0127] F=∑[w i ·f(θ i )] / ∑w i

[0128] where w i is the weight coefficient, which is related to the signal-to-noise ratio:

[0129] w i =SNR(θ i ) / ∑SNR(θ i )

[0130] Among them, F: fusion feature value; f(θ i ) represents the i-th angle θ i The defect characteristic value detected at the position; w i : Weight coefficient; f(θ i ): eigenvalue of the i-th angle; SNR: signal-to-noise ratio; θ i : the ith angular position.

[0131] The above multi-angle data fusion is particularly suitable for the circumferential shape of the flange. When the same defect may have different signal responses at different angles, weighted fusion can be used to combine the detection results of each angle to improve the accuracy of detection. The choice of weight is related to the signal-to-noise ratio, and the angle data with good signal quality can be given priority.

[0132] 10. Evaluation of the credibility of test results

[0133] Defining credibility indicators:

[0134] C = (ΔP-Th) / Th·SNR / SNR ref

[0135] Where, C: credibility index; ΔP: peak value; Th: threshold; SNR: actual signal-to-noise ratio; SNR ref , usually 10-20dB: reference signal-to-noise ratio; C th : Credibility threshold, usually 0.8-0.9.

[0136] When C>C th , it is determined to be a reliable detection result.

[0137] The flange automatic magnetic flux leakage detection system adopts a systematic signal processing and analysis method. Starting from the original signal acquisition, it eliminates environmental interference and equipment noise through multi-level filtering and correction, and establishes a complete signal processing process. First, average filtering is used to eliminate high-frequency noise, baseline drift correction is performed through polynomial fitting, and then adaptive gain adjustment is used to normalize signals of different intensities. In the feature extraction link, the system simultaneously considers the first-order derivative, second-order derivative and peak characteristics of the signal, and constructs a multi-dimensional feature space for defect identification. Defect judgment adopts an adaptive threshold combined with multiple criteria, which not only ensures the sensitivity of detection, but also effectively reduces the misjudgment rate. Accurate positioning is achieved through polar coordinate conversion, and a quantitative relationship between signal characteristics and defect size is established to achieve quantitative evaluation of defects. Multi-angle data fusion technology is adopted, and the flange circumference characteristics are used to improve the detection reliability through weighted fusion. At the same time, a credibility evaluation mechanism is introduced to quantitatively evaluate the detection results. The above process significantly improves the accuracy and reliability of detection, has good environmental adaptability and anti-interference ability, and realizes accurate positioning and quantitative evaluation of defects.

[0138] The following example uses an example to calculate the data for the above algorithm process:

[0139] Let's use a specific flange inspection example to demonstrate the calculation process:

[0140] Assume the detection scene parameters:

[0141] Flange radius R = 150mm

[0142] Detection angle θ = 30°

[0143] Sampling time interval Δt=0.001s

[0144] Scanning speed v=10mm / s

[0145] Reference voltage V ref =5V

[0146] Filter window size n=5

[0147] 1. Original signal acquisition

[0148] Assume that 5 data points are collected continuously at the position of θ=30°:

[0149] S(30°,t)=[4.2,4.5,4.8,4.3,4.1]V

[0150] 2. Signal filtering

[0151] Average filter calculation:

[0152] S'(30°,t)=(4.2+4.5+4.8+4.3+4.1) / 5=4.38V

[0153] 3. Baseline drift correction

[0154] Assume the baseline is obtained by fitting a quadratic polynomial:

[0155] B(30°,t)=0.2+0.1t+0.05t²

[0156] When t=0.001s:

[0157] B(30°,0.001)=0.2+0.1(0.001)+0.05(0.001)²=0.2001V

[0158] Corrected signal:

[0159] Sc(30°,t)=4.380.2001=4.1799V

[0160] 4. Signal gain adaptive adjustment

[0161] G(30°)=5V / 4.1799V=1.196

[0162] Signal after gain correction:

[0163] Sa(30°,t)=1.196×4.1799=5V

[0164] 5. Feature extraction

[0165] Assume that at this location:

[0166] P max (30°)=5.0V

[0167] P min (30°)=3.2V

[0168] ΔP(30°)=5.0-3.2=1.8V

[0169] 6. Defect judgment

[0170] Assumptions: standard deviation σ(30°)=0.5; k=2.5

[0171] Threshold calculation:

[0172] Th(30°)=2.5×0.5=1.25V

[0173] Since ΔP(30°)=1.8V>Th(30°)=1.25V, it is initially determined to be a defective point.

[0174] 7. Defect location

[0175] X=150×cos(30°)=129.90mm

[0176] Y=150×sin(30°)=75mm

[0177] 8. Defect size estimation

[0178] Assumptions:

[0179] Signal duration Δt=0.5s

[0180] Depth conversion factor α=2mm / V

[0181] Defect Length:

[0182] L=10mm / s×0.5s=5mm

[0183] Defect depth:

[0184] H=2mm / V×1.8V=3.6mm

[0185] 9. Multi-angle data fusion

[0186] Assume the eigenvalues ​​and signal-to-noise ratios measured at three angles:

[0187] θ1(30°): f(θ1)=1.8, SNR(θ1)=15dB

[0188] θ2(60°): f(θ2)=1.6, SNR(θ2)=12dB

[0189] θ3(90°): f(θ3)=1.7, SNR(θ3)=18dB

[0190] Weight calculation:

[0191] Total SNR = 15 + 12 + 18 = 45

[0192] w1=15 / 45=0.333

[0193] w2=12 / 45=0.267

[0194] w3=18 / 45=0.4

[0195] Fusion eigenvalues:

[0196] F=(0.333×1.8+0.267×1.6+0.4×1.7)=1.703

[0197] 10. Evaluation of the credibility of test results

[0198] Assuming SNR ref =15dB

[0199] C = (1.8-1.25) / 1.25×15 / 15 = 0.44

[0200] After filtering and correction, the original signal is filtered to obtain a more reliable signal. The defect is confirmed by feature extraction and multiple criteria, and the detection reliability is improved by multi-angle data fusion. The calculated defect is located on the flange surface (129.90mm, 75mm), with an estimated length of 5mm and a depth of 3.6mm. The credibility index is 0.44, which is lower than the typical threshold of 0.8, indicating that the test result needs to be repeatedly verified.

[0201] This specific embodiment is merely an explanation of the present invention and is not a limitation of the present invention. After reading this specification, those skilled in the art may make non-creative modifications to the present embodiment as needed. However, as long as they are within the scope of the claims of the present invention, they are protected by the patent law.

Claims

1. A flange automatic magnetic flux leakage detection device, characterized in that: It comprises a frame, a workbench is arranged in the middle of the frame, and a positioning seat is arranged on the top of the workbench; An annular bracket is arranged on the outer periphery of the positioning seat, and a plurality of magnetic sensors are arranged along the circumferential direction on the inner side of the annular bracket for detecting the magnetic leakage distribution on the flange surface; A rotating shaft is fixedly connected to one side of the annular bracket, and one end of the rotating shaft is connected to the frame through a bearing to realize the rotation movement of the annular bracket; The fixed sleeve in the middle of the rotating shaft is provided with a driven wheel, and the driven wheel is provided with a transmission belt outside; One end of the transmission belt is matched with a driving wheel, and a driving shaft is fixedly penetrated in the middle of the driving wheel for driving the driven wheel to rotate; One end of the driving shaft is connected to a stepper motor, and the rotation angle and speed of the annular bracket are controlled by the stepper motor; The drive shaft is fixedly connected to a support, and a ring coil is arranged inside the support for generating a magnetic field to magnetize the flange; The support and the annular bracket always maintain a parallel state during the rotation process, so that the measuring direction of the magnetic sensor is perpendicular to the direction of the magnetic field generated by the annular coil, so as to improve the detection accuracy.

2. The flange automatic magnetic flux leakage detection equipment according to claim 1 is characterized in that: Two ends of the annular coil are connected with connecting wires, and the connecting wires are connected with a power supply component for energizing the annular coil.

3. A flange automatic magnetic flux leakage detection system, characterized in that: The flange automatic magnetic flux leakage detection device as claimed in any one of claims 1 to 2 comprises the following steps: S1: Signal processing stage: including collecting the original signal generated by the magnetization of the ring coil, eliminating high-frequency noise through average filtering, correcting the baseline drift by polynomial fitting, and implementing adaptive signal gain adjustment; S2: Defect feature processing stage: calculate the first-order and second-order derivatives and characteristic parameters of the signal, confirm the defect by setting the adaptive threshold, use polar coordinate transformation to achieve precise positioning, and make quantitative estimation of the defect size; S3: Data processing and evaluation stage: Through multi-angle data fusion, the information of each detection angle is integrated, the weight coefficient is calculated for weighted fusion, and finally the credibility of the detection results is evaluated to give the detection evaluation conclusion.

4. The flange automatic magnetic flux leakage detection system according to claim 3 is characterized in that: The signal processing stage includes: Original signal acquisition: When the annular coil magnetizes the flange at different angles, the magnetic sensor will collect the leakage magnetic signal S(θ, t) containing defect information, and will also be mixed with the interference signal N(θ, t). The original signal containing the effective signal and noise is obtained through the formula S(θ, t) = M(θ, t) + N(θ, t); Where, M(θ,t) is the effective magnetic leakage signal, N(θ,t) is the noise signal; S(θ,t) is the total signal collected at angle θ and time t; M(θ,t) is the effective magnetic leakage signal component; N(θ,t) is the interference noise signal component; θ is the angular position between the sensor and the horizontal plane; t is the sampling time.

5. The flange automatic magnetic flux leakage detection system according to claim 4 is characterized in that: The signal processing stage also includes: Signal filtering: The support and the annular bracket are flipped over for a cycle, and the formula S'(θ,t)=(1 / n)∑[S(θ,ti)] is used for average filtering to eliminate high-frequency noise; Among them, S'(θ,t): the filtered signal; n: the filter window size, indicating the number of sampling points involved in the average; i: the index variable of the summation term; Baseline drift correction: Use polynomial fitting method to obtain the baseline: B(θ,t)=a0+a1t+a2t²+...+a n t n ; Correct the signal baseline by the formula Sc(θ,t)=S'(θ,t)-B(θ,t); Where, B(θ,t): baseline function obtained by fitting; a0, a1, a2...a n : polynomial fitting coefficients; Sc(θ,t): signal after baseline correction.

6. The flange automatic magnetic flux leakage detection system according to claim 5, characterized in that: The signal processing stage also includes: signal gain adaptive adjustment: G(θ)=V ref / max|Sc(θ,t)|; realize signal gain correction. The signal formula after gain correction is Sa(θ,t)=G(θ)·Sc(θ,t), thereby adjusting the signal amplitude; where G(θ): gain coefficient; V ref : reference voltage value or standardized amplitude; Sa(θ,t): signal after gain correction.

7. The flange automatic magnetic flux leakage detection system according to claim 3, characterized in that: The defect feature processing stage includes: Feature extraction: Calculate the first-order derivative of the signal separately , second-order derivative ; And the peak characteristics: P max (θ)=max[Sa(θ,t)]; P min (θ)=min[Sa(θ,t)]; ΔP(θ)=P max (θ)-P min (θ), used to characterize defect characteristics; Where D1(θ,t): the first-order derivative of the signal, which indicates the rate of change of the signal; D2(θ,t): the second-order derivative of the signal, which indicates the rate of change of the rate of change of the signal; P max (θ): maximum signal value; P min (θ): minimum value of the signal; ΔP(θ): peak value, indicating the amplitude of signal fluctuation; Defect judgment: Set the flange defect judgment standard according to the adaptive threshold Th(θ)=k·σ(θ); When the three conditions of ΔP(θ)>Th(θ), |D1(θ,t)|>D1th, and |D2(θ,t)|>D2th are met, it is confirmed that there is a defect; Among them, Th(θ): adaptive threshold; k: threshold adjustment coefficient, which is 2-3; σ(θ): standard deviation of the angle position signal; D1th: first-order derivative threshold; D2th: second-order derivative threshold.

8. The flange automatic magnetic flux leakage detection system according to claim 7, characterized in that: The defect feature processing stage also includes: Defect location: For the detected defect point, the defect location is determined by the polar coordinate conversion formula X=R·cos(θ), Y=R·sin(θ); where R is the flange radius; Defect size estimation: The defect length is calculated using L=v·Δt, and the defect depth is estimated using the formula H=α·ΔP(θ); Where, L: defect length; v: scanning speed; Δt: signal duration; H: defect depth; α: depth conversion factor.

9. The flange automatic magnetic flux leakage detection system according to claim 3, characterized in that: The data processing and evaluation phase includes: Multi-angle data fusion: For the same defect at different angles θ i The obtained eigenvalues ​​are weighted fused: F=∑[w i ·f(θ i )] / ∑w i , where w i is the weight coefficient, w i =SNR(θ i ) / ∑SNR(θ i ); Among them, F: fusion feature value; f(θ i ) represents the i-th angle θ i The defect characteristic value detected at the position; w i : Weight coefficient; f(θ i ): eigenvalue of the i-th angle; SNR: signal-to-noise ratio; θ i : the ith angular position.

10. The flange automatic magnetic flux leakage detection system according to claim 9, characterized in that: The data processing and evaluation phase also includes: Defining credibility indicators: C=(ΔP-Th) / Th·SNR / SNR ref ; When C>C th When , it is determined to be a reliable test result; Where, C: credibility index; ΔP: peak value; Th: threshold; SNR: actual signal-to-noise ratio; SNR ref , the value is 10-20dB: reference signal-to-noise ratio; C th : Credibility threshold, set to 0.8-0.9.

Images