A method for reconstructing B-mode images in rail flaw detection

Abstract

The present invention discloses a method for reconstructing a B-scan diagram for rail flaw detection, and the method includes at least one of the following processing procedures: S1) calibrating the coordinate unit of the detected B-scan diagram according to a rail model; S2) correcting the spatial conversion parameters of a probe wafer according to the B-scan diagram data of a rail joint; S3) using the A-scan amplitude as the B-scan diagram brightness display to manually select a suitable display amplitude and perform amplitude filtering processing to filter out interference detection points with inappropriate detected amplitude settings; S4) restoring the position of a damage mirror image generated at a rail joint to avoid false alarms of damage positions; S5) filtering out rail jaw detection points according to the centering deviation record to eliminate false alarms of damage; S6) providing a warning for the lack of a detection channel according to the B-scan diagram data at a rail joint. The present invention can solve the technical problems that the parameters of the existing rail flaw detection vehicle need to be manually calibrated, the detected B-scan diagram data is not normalized, detection errors are likely to occur, and the damage recognition rate is not high.

Description

Technical Field

[0001] The present invention relates to the technical field of railway engineering machinery, and in particular to a method for reconstructing a detection B-scan diagram for a rail flaw detector, which is used to automatically calibrate and standardize the detection B-scan diagram and improve the flaw recognition rate. Background Art

[0002] In the field of railway engineering and maintenance vehicles, ultrasonic rail flaw detectors 50 are widely used for the flaw detection and maintenance of rails. As shown in the Figure 1 accompanying drawings, an ultrasonic rail flaw detection system generally includes: a rail flaw detection system 10 provided on the rail flaw detector, a rail flaw analysis system 20, and a probe wheel 30 provided under the rail flaw detector. When rail flaw detection operation is required, the rail flaw detector controls the probe wheel 30 to press on the upper surface of the rail 40. The rail flaw detection system 10 sends an ultrasonic excitation pulse signal to the probe wheel 30. Under the action of an alternating electric field, the ultrasonic wafer of the probe wheel 30 generates mechanical vibration synchronous with the electric field, thereby emitting an ultrasonic signal. This process is called the inverse piezoelectric effect. The ultrasonic wafer can also generate an alternating electric field when subjected to an alternating pressure, realizing the reception of ultrasonic echoes, which is called the direct piezoelectric effect. The principle of ultrasonic flaw detection applied in rail flaw detection is: according to the characteristics of ultrasonic waves, once encountering the surface of two different media, ultrasonic waves will be reflected, thus forming a certain ultrasonic echo signal. When ultrasonic waves enter the air from steel, 100% reflection will occur, so it has a good detection effect on rail defects.

[0003] As shown in the Figure 5 accompanying drawings, the probe wheel 30 usually adopts a wheel-shaped structure, and a plurality of ultrasonic wafers 5 with different detection angles are installed on the axle center frame. The inner membrane of the tire of the probe wheel 30 is filled with a coupling liquid. When the rail flaw detector runs, the probe wheel 30 rolls along the rail 40, and the moving direction of the ultrasonic wafer 5 is parallel to the rail 40. When performing rail flaw detection operations, the ultrasonic signals emitted by the ultrasonic wafers of the probe wheel 30 usually reflect at the contact surface between the probe wheel module 30 and the rail 40, the flaw surface in the rail 40, and the lower surface of the rail 40. The ultrasonic wafers of the probe wheel 30 receive the ultrasonic echo signals, and can further convert the acoustic signals into electrical signals through the piezoelectric effect and transmit them to the rail flaw detection system 10. The rail flaw detection system 10 transmits the detection data to the rail flaw analysis system 20 through Ethernet, and the rail flaw analysis system 20 generates a rail flaw detection B-scan diagram and conducts diagnostic analysis.

[0004] Ultrasonic waves are used for rail flaw detection, which mainly includes the following characteristics: 1) Penetration: Ultrasonic waves can penetrate steel with a thickness of up to several meters; 2) Directionality: It has the property of linear propagation similar to light. The higher the frequency, the better the directionality, and it is easy to determine the location of defects; 3) Reflection characteristics: When ultrasonic waves encounter the surface of two different media during propagation, ultrasonic waves will be reflected, generating ultrasonic echo signals. When ultrasonic waves enter the air from steel, 100% reflection will occur, so it has a good detection effect on rail defects. When the defect size is larger than the ultrasonic wavelength, ultrasonic waves will be reflected back from the defect surface. Rail flaw detection mainly forms A-scan display diagrams and B-scan display diagrams based on detection. During the detection process, water spraying coupling is required between the probe wheel and the rail surface to ensure that ultrasonic waves reach the rail. Due to the high speed of the locomotive, the coupling effect is not ideal. In addition, at the curve of the line, the probe wheel deviates from the center of the rail, resulting in many false alarms and missed detections. Damage identification still largely depends on manual identification, and there are many defects in automatic damage identification.

[0005] As shown in the Figure 1 attachment, the rail flaw detection system 10 is composed of ultrasonic wafers with up to more than 30 independent ultrasonic channels to perform multi-angle and multi-directional flaw detection on the rail 40. The detection system 10 is equipped with a probe wheel frame, which is symmetric left and right. Three probe wheels 30 are installed on each rail respectively. The probe wheel 30 adopts a wheel structure. The probe wheel frame is equipped with piezoelectric ceramic transducers with multiple different detection angles, and the outer membrane of the probe wheel is filled with coupling liquid. When the rail flaw detection vehicle 50 runs, the probe wheel 30 rolls along the rail, and the moving direction of the transducer is parallel to the rail. The front and rear probe wheels are of the same type and are installed in opposite directions. There are 6 transducers in the probe wheel, namely, a three-transducer array of 0 degrees, 37 degrees, and 70 degrees and a side-scan transducer; there are two transducers with a deflection angle of 70 degrees pointing to the inner side of the rail and one 0-degree transducer in the middle probe wheel.

[0006] The detection system 10 sends an excitation electrical signal to the piezoelectric wafer to generate ultrasonic waves. The ultrasonic wave signals are reflected on the contact surface between the probe wheel 30 and the rail 40, the flaw surface in the rail 40, and the bottom surface of the rail 40. Then, through the piezoelectric effect of the wafer, the ultrasonic echo signal is converted into an electrical signal and transmitted to the rail flaw detection system 10. The detection system 10 transmits the wafer type and path transmission time parameters to the analysis system 20 through Ethernet. The analysis system 20 calculates the spatial position of the damage point based on the type, position, and path transmission time of the wafer, forms a rail ultrasonic B-scan display diagram, and the analysis system 20 automatically identifies the graph and frames the damage with a yellow square, providing a modern detection means for rail maintenance.

[0007] An angle encoder is installed at the axle end of the wheel of the rail flaw detection vehicle. When the wheel travels one week, N pulses are generated. The B-scan diagram of the flaw detection uses the number of pulses as the X coordinate. When it becomes length, it needs to be calculated according to the wheel diameter. Due to wheel wear, it is necessary to regularly measure and determine the wheel diameter proportionality coefficient.

[0008] In the B-mode diagram, according to the reflection surface of ultrasonic waves, the fault characteristic points shown in the appendix are defined. The normal direction thereof is the ultrasonic wave transmission direction. Each fault characteristic point can be detected by the ultrasonic piezoelectric wafer at the corresponding angle, corresponding to the detection channel number of the ultrasonic echo parameter data. As shown in the appendix Figure 2 As shown, it is a schematic diagram of the 37-degree rear piezoelectric wafer 1 detecting the rail bolt hole 2. According to the directivity of ultrasonic waves, ultrasonic waves propagate linearly within the detection range close to the wafer. Taking the rail surface as the reference, the condition for having an ultrasonic echo signal is: Figure 4 As shown, it is a schematic diagram of the 37-degree rear piezoelectric wafer 1 detecting the rail bolt hole 2. According to the directivity of ultrasonic waves, ultrasonic waves propagate linearly within the detection range close to the wafer. Taking the rail surface as the reference, the condition for having an ultrasonic echo signal is:

[0009]

[0010] Among them, X and Y are the coordinates of the fault characteristic point A in the rail, a is the incident angle of ultrasonic waves in the rail 40, W is the width of the ultrasonic wafer. The rail flaw detection system 20 and the probe wheel 30 are both installed on the rail flaw detection vehicle 50. X1 is the running position of the rail flaw detection vehicle 50, and X2 is the relative coordinate of the ultrasonic wafer with respect to the rail flaw detection vehicle 50.

[0011] The flaw detection B-mode diagram is obtained by detecting with multiple piezoelectric wafers at different positions. The result of the corresponding detection value needs to be calculated according to the position of the piezoelectric wafer, that is, the space conversion parameter. To avoid measurement errors, it needs to be debugged and set through an artificial calibration test. Summary of the Invention

[0012] In view of this, the purpose of the present invention is to provide a method for reconstructing the B-mode diagram of rail flaw detection, which automatically normalizes the data of the detected B-mode diagram, so as to solve the technical problems that it is difficult to set the space conversion parameter in the artificial calibration test, easy to make mistakes, and resulting in a low damage recognition rate of the B-mode diagram.

[0013] In order to achieve the above invention purpose, the present invention specifically provides a technical implementation scheme of a method for reconstructing the B-mode diagram of rail flaw detection. The method for reconstructing the B-mode diagram of rail flaw detection includes at least one of the following processing procedures for the B-mode diagram of rail flaw detection:

[0014] S1) Calibrate the coordinate unit of the detected B-mode diagram according to the rail model;

[0015] S2) Correct the space conversion parameter of the probe wheel wafer according to the B-mode diagram data of the rail joint;

[0016] S3) Use the amplitude of the A-scan display as the brightness display of the detected B-mode diagram, for manual selection of an appropriate display amplitude, and perform amplitude filtering processing to filter out the interference detection points with inappropriate detected amplitudes;

[0017] S4) Restore the position of the damage mirror image generated at the rail joint to avoid false alarms of the damage position;

[0018] S5) Filter out the rail jaw detection points according to the centering deviation record of the detection wheel to eliminate false alarms of damage.

[0019] S6) Warn of the missing detection channels according to the detection B-scan data at the rail joint.

[0020] Further, the process S1) includes the following steps:

[0021] S101) Judge the rail joint according to the detection B-scan.

[0022] S102) Calibrate the length of the detected rail by adjusting the wheel diameter ratio coefficient k. Before adjustment, the value of k is taken as 1.

[0023] S103) Select a standard-length rail according to the rail joint, and calculate and determine the adjusted wheel diameter ratio coefficient k.

[0024] S104) According to the adjusted wheel diameter ratio coefficient k, take the length of the rail corresponding to the measured encoder pulse number n as the calibrated rail length coordinate.

[0025] Further, the step S101) includes:

[0026] Judge the normal bolt holes according to the A-shaped feature map in the detection B-scan, judge the rail joint according to the Y-shaped feature map. The bolt holes on both sides of the rail joint are symmetrically distributed and conform to the standard rail model size.

[0027] Further, the step S102) includes:

[0028] The detection B-scan uses the encoder pulse number n installed at the axle end of the wheel set as the coordinate unit, and calibrates the length S of the detected rail by adjusting the wheel diameter ratio coefficient k n Calibration: where π is the pi, and the wheel diameter ratio coefficient D0 is the standard wheel set diameter, D1 is the actual wheel set diameter, and N is the number of pulses output by the encoder when it rotates one week.

[0029] Further, the step S103) includes:

[0030] Select a standard-length rail without truncation according to the detection B-scan: Calculate the length of the rail according to the spacing of the rail joint as where D0 is the standard wheel set diameter, N is the number of pulses output by the corresponding encoder when it rotates one week, and n1 is the number of pulses corresponding to the measured rail. Let the length of the standard-length rail be S, and δ1 be the rail wheel diameter error. When the length S1 of the detected rail satisfies (1 - δ1)S1 ≤ S ≤ (1 + δ1)S1, determine that the detected rail with length S1 is a standard-length rail without truncation. Then select and calculate the length of another detected rail as Among them, D0 is the standard wheel diameter, N is the number of pulses output by the corresponding encoder for one revolution, and n2 is the number of pulses corresponding to the measured rail. When (1 - δ1)S2 ≤ S ≤ (1 + δ1)S2 is satisfied, it is determined that the detected rail with length S2 is a standard-length rail without truncation. The relative error between the lengths of the two selected rails is within δ2, that is Then the wheel diameter proportionality coefficient k is determined by the average value of the encoder pulses corresponding to the two measured rails:

[0031] Further, in the step S104), substitute the k value obtained in step S103) into the following formula in step S102) to determine the calibrated rail length coordinate S n :

[0032]

[0033] Among them, n is the measured number of encoder pulses, and k is the adjusted wheel diameter proportionality coefficient.

[0034] Further, the process S2) includes the following steps:

[0035] S201) The spatial conversion parameters are determined by the wafer position. Define the farthest point detected by the rear probe wheel ultrasonic as the reference point. The probe wheel is based on the wheel axle center. The distance between the rear probe wheel and the reference point is L, the distance between the center probe wheel and the rear probe wheel is L1, and the distance between the front probe wheel and the rear probe wheel is L2. The front and rear probe wheels are symmetrically installed. The distance between the 0-degree wafer of the rear probe wheel and the wheel axle is l1, the distance between the 37-degree wafer of the rear probe wheel and the wheel axle is l2, and the distance between the 70-degree wafer of the rear probe wheel and the wheel axle is l3. The distance between the 0-degree wafer of the center probe wheel and the wheel axle is l4, the distance between the 70-degree wafer with a rear offset of the center probe wheel and the wheel axle is l5, and the distance between the 70-degree wafer with a front offset of the center probe wheel and the wheel axle is l6. The wafer spatial conversion parameters are:

[0036] Rear probe wheel 0-degree wafer: L + l1;

[0037] Rear probe wheel 37-degree wafer: L - l2;

[0038] Rear probe wheel 70-degree wafer: L + l3;

[0039] Center probe wheel 0-degree wafer: L + L1 + l4;

[0040] Center probe wheel 70-degree wafer with a rear offset: L + L1 - l5;

[0041] Center probe wheel 70-degree wafer with a front offset: L + L1 + l6;

[0042] Front probe wheel 0-degree wafer: L + L2 - l1;

[0043] Front probe wheel 37-degree wafer: L + L2 + l2;

[0044] Front detection wheel 70-degree chip: L + L2 - l3;

[0045] The detection wheel is of a fixed structure and has been checked. The structural parameters l1 to l6 of the detection wheel are constant values.

[0046] S202) Select the rear detection wheel as the reference, and do not correct the spatial conversion parameters of the rear detection wheel.

[0047] S203) Select the rail joint position where the B-scan is detected during the low-speed operation of the rail flaw detection vehicle, and calculate the difference in the detection patterns of the 0-degree chips of the front and rear detection wheels in the B-scan as ΔL2. Then the corrected spatial conversion parameter of the front detection wheel is:

[0048] Front detection wheel 0-degree chip: L + L2 + ΔL2 - l1;

[0049] Front detection wheel 37-degree chip: L + L2 + ΔL2 + l2;

[0050] Front detection wheel 70-degree chip: L + L2 + ΔL2 - l3.

[0051] S204) According to the detected B-scan, calculate the difference in the distance between the center of the detection pattern of the 70-degree chip when the center detection wheel is deflected and the center of the Y-shaped characteristic pattern as ΔL1. Then the corrected spatial conversion parameter of the center detection wheel is:

[0052] Center detection wheel 0-degree chip: L + L1 + ΔL1 + l4;

[0053] Center detection wheel rear-deflected 70-degree chip: L + L1 + ΔL1 - l5;

[0054] Center detection wheel front-deflected 70-degree chip: L + L1 + ΔL1 + L6.

[0055] Furthermore, the process S3) includes the following steps:

[0056] S301) Detect the position of the detected damage point in the B-scan, and use the A-scan amplitude corresponding to the ultrasonic echo size of the damage point as the display brightness of the corresponding detection point in the B-scan.

[0057] Furthermore, the step S301) includes:

[0058] Calculate the mean value A of the A-scan amplitudes of the detection points in the same channel of the chips at the bolt holes and rail joints m , m is the chip channel number, manually adjust the amplitude filtering coefficient K, take 0 < K < 1, KA m as the filtered amplitude. In the normal area of the wheel alignment, that is, when the alignment deviation ΔS is within the allowable deviation δ3, ΔS ≤ δ3, the detection points below the filtered amplitude are not displayed, and a new detected B-scan is generated. Artificial-assisted damage identification is performed according to the displayed pattern, and a suitable amplitude filtering coefficient K is selected.

[0059] Further, the process S4) includes the following steps:

[0060] S401) The mirror position for detecting the B-mode diagram is determined by the coordinates of the rail joint. The ultrasonic wave emitted by the 37-degree rear wafer is reflected by the rail joint to detect the point P(x, y), thereby forming the mirror detection point P'(x', y'). The actual damage is restored to point P based on the position of point P'. Let the coordinates of the rail joint be x0, and θ be the angle of the 37-degree rear wafer. Then, when the point P'(x', y') satisfies the condition 0 ≤ x' - x0 ≤ y' * tanθ, the coordinates of the mirror restoration point P(x, y) are:

[0061]

[0062] Further, the process S4) includes the following steps:

[0063] S402) The ultrasonic wave emitted by the 37-degree front wafer is reflected by the rail joint to detect the point P'(x', y'), thereby forming the mirror detection point P(x, y). The actual damage is restored to point P' based on the position of point P. Let the coordinates of the rail joint be x0, and θ be the angle of the 37-degree front wafer. Then, when P(x, y) satisfies 0 ≥ x - x0 ≥ -y * tanθ, the coordinates of the mirror restoration point P'(x', y') are:

[0064]

[0065] Further, the process S5) includes the following steps:

[0066] When the probe wheel deviates from the rail at a curve on the line, the automatic centering system is used to detect, control, and record the deviation value ΔS at the corresponding position of the detected B-mode diagram. When the probe wheel is not centered and the deviation value is greater than the set value δ4, and when the ultrasonic wave bottom echo of the rail from the 0-degree wafer is not lost, the detection reflection points of the rail jaw part are filtered out in the detected B-mode diagram to achieve non-centering filtering. Let the height of the rail jaw from the rail surface be h, and the measurement error limit value be δ5. When y is the vertical coordinate of the B-mode diagram, the non-centering filtering needs to satisfy the following three conditions simultaneously:

[0067] ⅰ) The bottom echo is not lost;

[0068] ⅱ) ΔS > δ4;

[0069] ⅲ) h - δ5 ≤ y ≤ h + δ5.

[0070] Further, the process S6) includes the following steps:

[0071] When passing through the area with rail joints and bolt holes, except for the side-beam channel, corresponding feature points are displayed in the detected B-mode diagram, and the missing part shows the corresponding channel missing.

[0072] By implementing the technical solution of the method for reconstructing the B-scan diagram in rail flaw detection provided by the present invention as described above, the following beneficial effects are achieved:

[0073] (1) In the method for reconstructing the B-scan diagram in rail flaw detection of the present invention, the reconstruction of the B-scan diagram data according to the rail model standardizes the detection graph, lays a foundation for computer damage recognition, improves the damage recognition efficiency, and reduces false alarms;

[0074] (2) In the method for reconstructing the B-scan diagram in rail flaw detection of the present invention, the rail joints are identified according to the detected B-scan diagram, and the wheel diameter is automatically calibrated by the standard rail length, which can improve the detection accuracy and provide a means for accurate damage positioning and trend comparison analysis;

[0075] (3) In the method for reconstructing the B-scan diagram in rail flaw detection of the present invention, the spatial parameters of the probe wheel chips of the rail flaw detection vehicle are automatically adjusted according to the actually detected B-scan diagram, which can effectively avoid errors and deviations caused by manual settings, improve the automation level of rail flaw detection, and reduce the dependence of the detection system on the operator's level;

[0076] (4) In the method for reconstructing the B-scan diagram in rail flaw detection of the present invention, the amplitude of the A-scan is used as the brightness representation of the B-scan, which is convenient for manual observation and analysis. By setting an appropriate filtering amplitude, detection interference points with smaller amplitudes can be removed;

[0077] (5) In the method for reconstructing the B-scan diagram in rail flaw detection of the present invention, the use of the centering deviation for auxiliary judgment can effectively remove false alarms of damage in the rail web part, and at the same time can warn of the detection channel missing fault, reminding the operator to repair in time to ensure the integrity of the detection graph. BRIEF DESCRIPTION OF THE DRAWINGS

[0078] In order to more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the following will briefly introduce the drawings required for use in the description of the embodiments or the prior art. Obviously, the drawings in the following description are only some embodiments of the present invention, and those of ordinary skill in the art can also obtain other embodiments according to these drawings without creative efforts.

[0079] Figure 1 is the system structure block diagram of the flaw detection system of the existing rail flaw detection vehicle;

[0080] Figure 2 is the definition schematic diagram of the fault feature points in the B-scan diagram of rail flaw detection;

[0081] Figure 3 is the schematic diagram of the "A"-shaped detection graph of the bolt hole in the B-scan diagram of rail flaw detection;

[0082] Figure 4 It is a schematic diagram of the principle of using an ultrasonic piezoelectric wafer to detect rail damage in the prior art;

[0083] Figure 5 It is a schematic diagram of the detection process of an ultrasonic piezoelectric wafer in the prior art;

[0084] Figure 6 It is a schematic diagram of the "Y"-shaped detection pattern at the rail joint in the B-mode diagram of rail flaw detection;

[0085] Figure 7 It is a typical B-mode diagram for detecting rail joints in the prior art;

[0086] Figure 8 It is a schematic diagram of the spatial conversion parameters of the B-mode diagram for rail flaw detection;

[0087] Figure 9 It is a schematic diagram of the dimensional structure of the rail cross-section in the prior art;

[0088] Figure 10 It is a schematic diagram of the dimensional structure of the side of the rail end in the prior art;

[0089] Figure 11 It is a schematic diagram of the principle of damage mirror restoration processing in a specific embodiment of the method for reconstructing the B-mode diagram of rail flaw detection according to the present invention;

[0090] Figure 12 It is a schematic diagram of the principle of non-alignment filtering processing in a specific embodiment of the method for reconstructing the B-mode diagram of rail flaw detection according to the present invention;

[0091] Figure 13 It is a block diagram of the structural composition of the coordinate unit calibration module in a specific embodiment of the device for reconstructing the B-mode diagram of rail flaw detection based on the method of the present invention;

[0092] Figure 14 It is a block diagram of the structure of the spatial conversion parameter correction module in a specific embodiment of the device for reconstructing the B-mode diagram of rail flaw detection based on the method of the present invention;

[0093] Figure 15 It is a block diagram of the structure of the amplitude filtering module in a specific embodiment of the device for reconstructing the B-mode diagram of rail flaw detection based on the method of the present invention;

[0094] Figure 16 It is a block diagram of the structure of the damage mirror restoration module in a specific embodiment of the device for reconstructing the B-mode diagram of rail flaw detection based on the method of the present invention;

[0095] Figure 17It is the structural block diagram of the misalignment filtering module in a specific embodiment of the B-scan image reconstruction device for rail flaw detection based on the method of the present invention;

[0096] Figure 18 It is the structural block diagram of the detection channel missing warning module in a specific embodiment of the B-scan image reconstruction device for rail flaw detection based on the method of the present invention;

[0097] Figure 19 It is the program flow chart of a specific embodiment of the method for reconstructing the B-scan image of rail flaw detection of the present invention;

[0098] In the figure: 1 - Coordinate unit calibration module, 2 - Spatial transformation parameter correction module, 3 - Amplitude filtering module, 4 - Damage mirror reduction module, 5 - Misalignment filtering module, 6 - Detection channel missing warning module, 10 - Rail flaw detection system, 11 - Rail joint judgment unit, 12 - Wheel diameter ratio coefficient adjustment unit, 13 - Rail length coordinate calibration unit, 20 - Rail flaw detection vehicle analysis system, 30 - Probe wheel, 40 - Rail, 50 - Rail flaw detection vehicle, 60 - Chip, 70 - Damage feature point, 80 - Bolt hole, 90 - Rail joint, 100 - Rail jaw detection reflection point. Specific embodiment

[0099] Mileage pulse number: A rotary encoder is installed at the axle end of the locomotive wheel set. When a wheel with a diameter of D rotates one week, the traveling distance is πD, corresponding to the number of pulses N output by the encoder when it rotates one week. Since the encoder has high precision and is used for the distance measurement of the flaw detection vehicle, the pulse interval (πD / N) is used as the basic unit for positioning.

[0100] A-scan display: The ultrasonic echo is displayed in the form of a waveform, with the vertical coordinate representing the amplitude of the reflected wave and the horizontal coordinate representing the wave transmission time.

[0101] B-scan display: The position of the damage is displayed in the form of an image.

[0102] Damage identification: In the B-scan image, the damage area is framed with a warning color (such as yellow).

[0103] To make the objectives, technical solutions, and advantages of the embodiments of the present invention clearer, the technical solutions in the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings in the embodiments of the present invention. Obviously, the described embodiments are only a part of the embodiments of the present invention, rather than all of the embodiments. All other embodiments obtained by those of ordinary skill in the art based on the embodiments of the present invention without creative efforts shall fall within the scope of protection of the present invention.

[0104] As shown in the appended Figure 1 to the appended Figure 19As shown, a specific embodiment of the method for reconstructing the B-scan diagram of rail flaw detection according to the present invention is given. The present invention will be further described below in conjunction with the accompanying drawings and specific embodiments.

[0105] Embodiment 1

[0106] As shown in the attached Figure 19 drawing, an embodiment of the method for reconstructing the B-scan diagram of rail flaw detection according to the present invention includes at least one (i.e., one or more than two) processing processes for the B-scan diagram of rail flaw detection as follows:

[0107] S1) Calibrate the coordinate unit of the detected B-scan diagram according to the rail model;

[0108] S2) Correct the spatial conversion parameters of the probe wheel chip according to the B-scan diagram data of the rail joint 90;

[0109] S3) Use the amplitude of the A-scan display as the brightness display of the detected B-scan diagram for manual selection of the appropriate display amplitude, perform amplitude filtering processing, and filter out the interference detection points with inappropriate detected amplitude settings;

[0110] S4) Restore the position of the damage mirror image generated at the rail joint 90 to avoid false alarms of the damage position;

[0111] S5) Filter out the rail jaw detection points according to the recorded centering deviation of the probe wheel to eliminate false alarms of damage;

[0112] S6) Warn of the lack of detection channels according to the B-scan diagram data at the rail joint 90.

[0113] Process S1) further includes the following steps:

[0114] S101) Judge the rail joint 90 according to the detected B-scan diagram;

[0115] S102) Calibrate the length of the detected rail 40 by adjusting the wheel diameter ratio coefficient k. The k value is taken as 1 before adjustment;

[0116] S103) Select a standard length rail according to the rail joint 90, calculate and determine the adjusted wheel diameter ratio coefficient k;

[0117] S104) Take the length of the rail corresponding to the measured encoder pulse number n as the calibrated rail length coordinate according to the adjusted wheel diameter ratio coefficient k.

[0118] Step S101) further includes:

[0119] Judge the normal bolt hole 80 according to the A-shaped feature diagram in the detected B-scan diagram data, judge the rail joint 90 according to the Y-shaped feature diagram, and the bolt holes 80 on both sides of the rail joint 90 are symmetrically distributed and conform to the standard rail model size. As shown in the attachedFigure 9 As shown in the figure. Automatically calibrating the wheel diameter according to the standard length rail model can improve the measurement accuracy of the detection B-mode diagram, providing an effective means for accurate positioning of damage and trend comparison analysis.

[0120] A rotary encoder is installed at the axle end of the wheels of the rail flaw detector 50. Assuming the standard wheel diameter is D0, the distance traveled by the wheel set in one revolution is πD0, and the number of pulses output by the encoder in one revolution is N. Since the rotary encoder has high precision and is used for the distance measurement of the rail flaw detector 50, this number of pulses is used as the coordinate measurement unit for the detection B-mode diagram, and the corresponding spacing is πD0 / N. Assuming the number of pulses for measuring the standard rail with the standard wheel diameter is n0, and the standard rail length is S. After the rail flaw detector 50 has run for a period of time, wear will occur on the tread of the wheel hub. Let D1 be the wheel diameter with wear, and the measured number of pulses corresponding to the standard length rail is n1. Then:

[0121]

[0122] Let k be the wheel diameter ratio coefficient, representing the change in the wheel diameter, which should meet the requirements of operation management error, and the error is δ1.

[0123] Step S102) further includes:

[0124] The detection B-mode diagram uses the number of pulses n of the encoder installed at the axle end of the wheel set as the coordinate unit, and calibrates the length S of the detected rail 40 by adjusting the wheel diameter ratio coefficient k: n Calibration: where π is the pi, and the wheel diameter ratio coefficient D0 is the standard wheel set diameter, D1 is the actual wheel set diameter, and N is the number of pulses output by the encoder in one revolution.

[0125] Step S103) further includes:

[0126] Select the standard length rail without truncation according to the detection B-mode diagram: According to the spacing of the rail joints 90, calculate the length of the rail 40 as where D0 is the standard wheel set diameter, N is the number of pulses output by the corresponding encoder in one revolution, and n1 is the number of pulses corresponding to the measured rail. Assuming the length of the standard length rail is S, and δ1 is the rail wheel diameter error. When the length S1 of the detected rail 40 satisfies (1 - δ1)S1 ≤ S ≤ (1 + δ1)S1, it is determined that the detected rail 40 with length S1 is a standard length rail without truncation. Then select and calculate the length of another detected rail 40 as Among them, D0 is the standard wheel diameter, N is the number of pulses output by the corresponding encoder for one revolution, and n2 is the number of pulses corresponding to the measured rail. When (1 - δ1)S2 ≤ S ≤ (1 + δ1)S2 is satisfied, it is determined that the detected rail 40 with length S2 is a standard-length rail without truncation. The relative error between the lengths of two rails is within δ2, that is Then the wheel diameter proportionality coefficient k is determined by the average value of the encoder pulse numbers corresponding to the two measured rails:

[0127] In step S104), substitute the k value obtained in step S103) into the following formula in step S102) to determine the calibrated rail length coordinate S n :

[0128]

[0129] Among them, n is the measured encoder pulse number, and k is the adjusted wheel diameter proportionality coefficient.

[0130] The flaw detection system uses multiple detections. A fault feature point 70 can be detected multiple times. The fault feature points A, B, and C of the bolt hole 80 can be detected by different (piezoelectric) wafers 60, thus forming a detection pattern in the shape of "A" as shown in the appendix Figure 3 Shown. Among them, there is no rail bottom reflection of the 0-degree ultrasonic wave at the bolt hole, forming a missing bottom wave.

[0131] As shown in the appendix Figure 6 Shown, there is a fixed detection pattern at the rail joint 90. At the perpendicular intersection of the rail joint 90 and the rail jaw surface, the ultrasonic wave L1 can be reflected to form a rail joint pattern. In addition, the ultrasonic wave L2 is reflected by the rail joint 90, and the wave type is converted from longitudinal wave to transverse wave to detect the bolt hole 80. Since the transverse wave sound velocity (3230 m / s) is much smaller than the longitudinal wave sound velocity (5860 m / s), the displayed distance of the detection point is shorter than the actual one, forming a corner of the "Y" character feature map.

[0132] A typical B-mode diagram of rail joint detection is shown in the appendix Figure 7 Shown. Currently, the rail 40 mainly has five types listed in Table 1. The rail height can be measured using 0-degree ultrasonic wave to determine the type of the rail 40 and obtain the corresponding model parameters, as shown in the appendix Figure 9 and the appendix Figure 10 Shown. The centering record file uses the same coordinates as the B-mode diagram, and records are made every 1 mm change in deviation.

[0133]

[0134] Table 1 Rail type names and main dimensions

[0135] The rail flaw detection vehicle 50 performs inspections through multiple piezoelectric wafers 60 at different positions. When its position deviates, the detected B-scan image also shows corresponding deviations and variations, thus affecting the automatic identification of rail damage. The detection wheels 30 of the rail flaw detection vehicle 50 include a front detection wheel, a rear detection wheel, and a center detection wheel. Since the detection wheels 30 use constant pressure to control the downward pressure and monitor the ultrasonic A-scan of the 0-degree wafer of the detection wheels 30, there is no deviation in the Y-axis direction. Therefore, the spatial transformation parameters only need to be adjusted for the X-axis direction parameters.

[0136] As shown in the appendix Figure 8 The process S2) further includes the following steps:

[0137] S201) The spatial transformation parameters are determined by the wafer position. The farthest point of ultrasonic detection of the rear detection wheel is defined as the reference point. The detection wheel 30 is based on the wheel axle center. The distance between the rear detection wheel and the reference point is L, the distance between the center detection wheel and the rear detection wheel is L1, and the distance between the front detection wheel and the rear detection wheel is L2. The front and rear detection wheels are symmetrically installed. The distance between the 0-degree wafer and the wheel axle is l1, the distance between the 37-degree wafer and the wheel axle is l2, and the distance between the 70-degree wafer and the wheel axle is l3. The distance between the 0-degree wafer of the center detection wheel and the wheel axle is l4, the distance between the 70-degree wafer with a rear offset of the center detection wheel and the wheel axle is l5, and the distance between the 70-degree wafer with a front offset of the center detection wheel and the wheel axle is l6. The wafer spatial transformation parameters are:

[0138] 0-degree wafer of the rear detection wheel: L + l1;

[0139] 37-degree wafer of the rear detection wheel: L - l2;

[0140] 70-degree wafer of the rear detection wheel: L + l3;

[0141] 0-degree wafer of the center detection wheel: L + L1 + l4;

[0142] 70-degree wafer with a rear offset of the center detection wheel: L + L1 - l5;

[0143] 70-degree wafer with a front offset of the center detection wheel: L + L1 + l6;

[0144] 0-degree wafer of the front detection wheel: L + L2 - l1;

[0145] 37-degree wafer of the front detection wheel: L + L2 + l2;

[0146] 70-degree wafer of the front detection wheel: L + L2 - l3;

[0147] The detection wheel 30 is a fixed structure and has been calibrated. The structural parameters l1 to l6 of the detection wheel 30 are constant values.

[0148] S202) Select the rear detection wheel as the reference, and the spatial transformation parameters of the rear detection wheel are not corrected;

[0149] S203) Select the 90 position of the rail joint in the detected B-scan image when the rail flaw detector runs at a low speed, and calculate the difference ΔL2 between the detection patterns of the 0-degree chips of the front and rear detection wheels in the detected B-scan image. Then, the corrected spatial conversion parameter of the front detection wheel is as follows:

[0150] For the 0-degree chip of the front detection wheel: L + L2 + ΔL2 - l1;

[0151] For the 37-degree chip of the front detection wheel: L + L2 + ΔL2 + l2;

[0152] For the 70-degree chip of the front detection wheel: L + L2 + ΔL2 - l3.

[0153] S204) According to the detected B-scan image, calculate the difference ΔL1 between the center distance of the detection pattern of the 70-degree chip of the center detection wheel and the center of the Y-shaped feature pattern. Then, the corrected spatial conversion parameter of the center detection wheel is as follows:

[0154] For the 0-degree chip of the center detection wheel: L + L1 + ΔL1 + l4;

[0155] For the 70-degree chip with rear deflection of the center detection wheel: L + L1 + ΔL1 - l5;

[0156] For the 70-degree chip with front deflection of the center detection wheel: L + L1 + ΔL1 + L6.

[0157] In Example 1, according to the actual detected B-scan image, the spatial parameters of the detection wheels of the rail flaw detector 50 are automatically adjusted, avoiding errors and deviations caused by manual settings, improving the automation level of rail flaw detection, and reducing the dependence of the detection system on the operator's level.

[0158] Process S3) further includes the following steps:

[0159] S301) The detected B-scan image shows the position of the detected damage point, and the A-scan amplitude corresponding to the ultrasonic echo size of the damage point is used as the display brightness of the corresponding detection point in the B-scan display.

[0160] Step S301) further includes:

[0161] Calculate the mean value A of the A-scan amplitudes of the detection points in the same channel of the chip 60 at the bolt hole 80 and the rail joint 90 m , where m is the chip channel number, manually adjust the amplitude filtering coefficient K, and take 0 < K < 1, KA m as the filtered amplitude. In the normal area of the 30 pairs of detection wheels, that is, when the alignment deviation ΔS is within the allowable deviation δ3, ΔS ≤ δ3, the detection points lower than the filtered amplitude are not displayed, and a new detected B-scan image is generated. Artificial-assisted damage identification is performed according to the displayed image, and a suitable amplitude filtering coefficient K is selected.

[0162] The rail joint 90 will produce a specular reflection on ultrasonic waves, causing an image of damage to occur in the nearby bolt holes 80, thereby leading to misalignment in damage identification. As shown in the appendix Figure 11 The process S4) further includes the following steps:

[0163] S401) The mirror image position of the B-scan is detected and determined by the coordinates of the rail joint 90. The ultrasonic waves emitted by the 37-degree rear wafer are reflected by the rail joint 90 to detect the point P(x, y), forming a mirror image detection point P'(x', y'). The actual damage restores the point P according to the position of the point P'. Let the coordinates of the rail joint 90 be x0, and θ be the angle of the 37-degree rear wafer. Then when the point P'(x', y') satisfies the condition 0 ≤ x' - x0 ≤ y'tanθ, the coordinates of the mirror image restoration point P(x, y) are:

[0164]

[0165] The process S4) further includes the following steps:

[0166] S402) The ultrasonic waves emitted by the 37-degree front wafer are reflected by the rail joint 90 to detect the point P'(x', y'), forming a mirror image detection point P(x, y). The actual damage restores the point P' according to the position of the point P. Let the coordinates of the rail joint 90 be x0, and θ be the angle of the 37-degree front wafer. Then when P(x, y) satisfies 0 ≥ x - x0 ≥ -ytanθ, the coordinates of the mirror image restoration point P'(x', y') are:

[0167]

[0168] In Embodiment 1, by reconstructing the detected B-scan of the damage image caused by the rail joint 90, the damage detection pattern can be standardized, avoiding mispositioning of damage, helping to improve the automatic damage recognition rate, and helping to achieve trend quantitative analysis through multiple damage detections, which is more intuitive compared to manual recognition.

[0169] The process S5) further includes the following steps:

[0170] When the probe wheel 30 has a deviation from the rail 40 at a line curve, the automatic centering system is used to detect, control, and record the deviation value ΔS at the corresponding position of the detected B-scan. When the probe wheel 30 is not centered and the deviation value is greater than the set value δ4, and when the ultrasonic wave bottom echo of the rail at the 0-degree wafer does not disappear, the detection reflection points 100 of the rail jaw part of the rail 40 are filtered out in the detected B-scan to achieve non-centering filtering. Let the height of the rail jaw from the rail surface be h, and the measurement error limit value be δ5. When y is the vertical coordinate of the B-scan, the non-centering filtering needs to simultaneously meet the following three conditions:

[0171] ⅰ) The bottom echo is not lost;

[0172] ii) ΔS > δ4;

[0173] iii) h - δ5 ≤ y ≤ h + δ5.

[0174] By using the centering deviation for auxiliary judgment, the false alarms of damage in the web part of the 40-rail can be effectively removed.

[0175] Process S6) further includes the following steps:

[0176] When passing through the areas with rail joints 90 and bolt holes 80, except for the side-beating channels, corresponding feature points are displayed in the detected B-scan images, and the missing parts show the corresponding channel losses. Warning of the detected channel loss fault can remind the operator to repair in time, thus ensuring the integrity of the detected images and the reliability of the system operation.

[0177] Embodiment 2

[0178] As shown in the Figure 13 accompanying drawings, an embodiment of a rail flaw detection B-scan image reconstruction device based on the method described in Embodiment 1 includes a coordinate unit calibration module 1, and the coordinate unit calibration module 1 further includes: a rail joint judgment unit 11, a wheel diameter ratio coefficient adjustment unit 12, and a rail length coordinate calibration unit 13. The rail joint judgment unit 11 judges the rail joint 90 according to the detected B-scan image. The rail length coordinate calibration unit 13 calibrates the length of the detected rail 40 by adjusting the wheel diameter ratio coefficient k, and the k value is taken as 1 before adjustment. The wheel diameter ratio coefficient adjustment unit 12 selects a standard length rail without truncation according to the detected B-scan image, calculates and determines the adjusted wheel diameter ratio coefficient k according to the rail joint 90. The rail length coordinate calibration unit 13 takes the length of the rail 40 corresponding to the measured encoder pulse number n as the calibrated rail length coordinate according to the adjusted wheel diameter ratio coefficient k.

[0179] The rail joint judgment unit 11 judges the normal bolt holes 80 according to the A-shaped feature image in the detected B-scan image, and judges the rail joint 90 according to the Y-shaped feature image. The bolt holes 80 on both sides of the rail joint 90 are symmetrically distributed and conform to the standard rail model size.

[0180] The detected B-scan image uses the encoder pulse number n installed at the axle end of the wheel set as the coordinate unit, and the rail length coordinate calibration unit 13 calibrates the length S of the detected rail 40 by adjusting the wheel diameter ratio coefficient k n Calibration: where π is the pi, and the wheel diameter ratio coefficient D0 is the standard wheel set diameter, D1 is the actual wheel set diameter, and N is the number of pulses output by the encoder when rotating one week.

[0181] The wheel diameter ratio coefficient adjustment unit 12 selects a standard length rail without truncation according to the detected B-scan image: calculates the length of the rail 40 according to the spacing of the rail joint 90 as Among them, D0 is the standard wheel set diameter, N is the number of pulses output when the corresponding encoder rotates one week, and n1 is the number of pulses corresponding to the measured rail. Let the length of the standard length rail be S, and δ1 be the rail wheel diameter error. When the length S1 of the detected rail 40 satisfies (1 - δ1)S1 ≤ S ≤ (1 + δ1)S1, it is determined that the detected rail 40 with length S1 is a standard length rail without truncation. Then, select and calculate the length of another detected rail 40 as Among them, D0 is the standard wheel set diameter, N is the number of pulses output when the corresponding encoder rotates one week, and n2 is the number of pulses corresponding to the measured rail. When (1 - δ1)S2 ≤ S ≤ (1 + δ1)S2 is satisfied, it is determined that the detected rail 40 with length S2 is a standard length rail without truncation. The relative error between the lengths of the two selected rails is within δ2, that is Then, the wheel diameter ratio coefficient adjustment unit 12 determines the wheel diameter ratio coefficient according to the average value of the encoder pulse numbers corresponding to the two measured rails

[0182] The rail length coordinate calibration unit 13 further calculates the calibrated rail length coordinate S according to the wheel diameter ratio coefficient k determined by the wheel diameter ratio coefficient adjustment unit 12 n :

[0183]

[0184] Among them, n is the measured encoder pulse number, and k is the adjusted wheel diameter ratio coefficient.

[0185] The detection wheels 30 of the rail flaw detection vehicle 50 include a front detection wheel, a rear detection wheel and a center detection wheel. As shown in the appendix Figure 14 shown, the B-type diagram reconstruction device for rail flaw detection and inspection further includes a space conversion parameter correction module 2. The space conversion parameters are determined by the wafer position. Define the farthest point of ultrasonic detection of the rear detection wheel as the reference point. The detection wheel 30 is based on the wheel axle center. The distance from the rear detection wheel to the reference point is L, the distance from the center detection wheel to the rear detection wheel is L1, and the distance from the front detection wheel to the rear detection wheel is L2. The front and rear detection wheels are symmetrically installed. The distance from the 0-degree wafer to the wheel axle is l1, the distance from the 37-degree wafer to the wheel axle is l2, and the distance from the 70-degree wafer to the wheel axle is l3. The distance from the 0-degree wafer of the center detection wheel to the wheel axle is l4, the distance from the 70-degree wafer with a rear offset of the center detection wheel to the wheel axle is l5, and the distance from the 70-degree wafer with a front offset of the center detection wheel to the wheel axle is l6. The detection wheel 30 is a fixed structure and has been calibrated. The structural parameters l1 to l6 of the detection wheel 30 are constant values. The space conversion parameter correction module 2 calculates the wafer space conversion parameters according to the following formula:

[0186] Rear detection wheel 0-degree wafer: L + l1;

[0187] Rear detection wheel 37-degree wafer: L - l2;

[0188] Rear detection wheel 70-degree wafer: L + l3;

[0189] Center detection wheel 0-degree wafer: L + L1 + l4;

[0190] Center detection wheel 70-degree rear offset wafer: L + L1 - l5;

[0191] Center detection wheel 70-degree front offset wafer: L + L1 + l6;

[0192] Front detection wheel 0-degree wafer: L + L2 - l1;

[0193] Front detection wheel 37-degree wafer: L + L2 + l2;

[0194] Front detection wheel 70-degree wafer: L + L2 - l3.

[0195] Select the rear detection wheel as the reference, and the spatial conversion parameters of the rear detection wheel are not corrected. The spatial conversion parameter correction module 2 selects the detection B-scan image at the 90th rail joint when the rail flaw detection vehicle is running at low speed, and the difference in the number of detection points of the 0-degree wafers of the front and rear detection wheels

[0196] ≤1. Since the detection wheel 30 has a fixed structure and has been calibrated, the structural parameters l1 to l6 of the detection wheel 30 are constant values.

[0197] The spatial conversion parameter correction module 2 calculates the difference in the detection pattern of the 0-degree wafer of the detection B-scan images of the front and rear detection wheels as ΔL2 according to the detection B-scan image. The spatial conversion parameter correction module 2 calculates and outputs the corrected spatial conversion parameters of the front detection wheel according to the following formula:

[0198] Front detection wheel 0-degree wafer: L + L2 + ΔL2 - l1;

[0199] Front detection wheel 37-degree wafer: L + L2 + ΔL2 + l2;

[0200] Front detection wheel 70-degree wafer: L + L2 + ΔL2 - l3.

[0201] The spatial conversion parameter correction module 2 calculates the difference in the distance between the center of the detection pattern of the 70-degree deflected wafer of the center detection wheel and the center of the Y-shaped feature point pattern as ΔL1 according to the detection B-scan image. The spatial conversion parameter correction module 2 calculates and outputs the corrected spatial conversion parameters of the center detection wheel according to the following formula:

[0202] Center detection wheel 0-degree wafer: L + L1 + ΔL1 + l4;

[0203] Center detection wheel 70-degree rear offset wafer: L + L1 + ΔL1 - l5;

[0204] Center detection wheel 70-degree front offset wafer: L + L1 + ΔL1 + L6.

[0205] As attachedFigure 15 As shown in the figure, the B-scan image reconstruction device for rail flaw detection further includes an amplitude filtering module 5. The B-scan image shows the positions of the detected damage points. The amplitude filtering module 5 uses the A-scan amplitude corresponding to the ultrasonic echo size of the damage points as the display brightness signal for the corresponding detection points in the B-scan display.

[0206] The amplitude filtering module 3 calculates the average value A of the A-scan amplitudes of the detection points in the same channel of the wafer 60 at the bolt holes 80 and the rail joints 90. m , where m is the wafer channel number. Manually adjust the amplitude filtering coefficient K, where 0 < K < 1, and KA m is used as the filtered amplitude. In the normal area of the pair of detection wheels 30, that is, when the alignment deviation ΔS is within the allowable deviation δ3, ΔS ≤ δ3, the detection points below the filtered amplitude are not displayed, and a new detection B-scan image is generated. Artificial-assisted flaw identification is performed based on the displayed image, and an appropriate amplitude filtering coefficient K is selected.

[0207] As shown in the appendix Figure 16 As shown in the figure, the B-scan image reconstruction device for rail flaw detection further includes a damage mirror image restoration module 4. The mirror image position of the B-scan image is determined by the coordinates of the rail joint 90. The ultrasonic wave emitted by the rear 37-degree wafer is reflected by the rail joint 90 to detect the point P(x, y). The point P is located in the detection blind area of the front 37-degree wafer, that is, the ΔABC area, and a mirror image detection point P'(x', y') is formed. The actual damage restores the point P according to the position of the point P'. Let the coordinates of the rail joint 90 be x0, and θ be the angle of the rear 37-degree wafer. Then when the point P'(x', y') satisfies the condition 0 ≤ x' - x0 ≤ y'tanθ, the coordinates of the point P(x, y) restored by the mirror image restoration module 4 are:

[0208]

[0209] The ultrasonic wave emitted by the front 37-degree wafer is reflected by the rail joint 90 to detect the point P'(x', y'). The point P' is located in the detection blind area of the rear 37-degree wafer, that is, the ΔABD area, and a mirror image detection point P(x, y) is formed. The actual damage restores the point P' according to the position of the point P. Let the coordinates of the rail joint 90 be x0, and θ be the angle of the front 37-degree wafer. Then when P(x, y) satisfies 0 ≥ x - x0 ≥ -ytanθ, the coordinates of the point P'(x', y') restored by the mirror image restoration module 4 are:

[0210]

[0211] As shown in the appendix Figure 17As shown in the figure, the B-scan diagram reconstruction device for rail flaw detection further includes an off-center filtering module 5. When the detection wheel 30 deviates from the rail 40 at a curve on the track, the automatic centering system is used to detect, control, and record the deviation value ΔS at the corresponding position of the detection B-scan diagram. When the detection wheel 30 is off-center and the deviation value is greater than the set value δ4, and when the ultrasonic wave bottom echo of the rail from the 0-degree wafer is not lost, the off-center filtering module 5 filters out the detection reflection points 100 of the rail jaw part in the detection B-scan diagram to achieve off-center filtering. Let the height of the rail jaw from the rail surface be h, and the measurement error limit value be δ5. When y is the ordinate of the B-scan diagram, the off-center filtering module 5 needs to simultaneously meet the following three conditions for off-center filtering:

[0212] ⅰ) The bottom echo is not lost;

[0213] ⅱ) ΔS > δ4;

[0214] ⅲ) h - δ5 ≤ y ≤ h + δ5.

[0215] As shown in the attachment Figure 18 As shown in the figure, the B-scan diagram reconstruction device for rail flaw detection further includes a detection channel missing warning module 6. When passing through the area with rail joints 90 and bolt holes 80, except for the side-scan channel, corresponding feature points are displayed in the detection B-scan diagram, and the missing part is indicated by the detection channel missing warning module 6 to show the corresponding channel missing.

[0216] In the description of the present application, it should be noted that when an element is referred to as "fixed to" or "disposed on" another element, it can be directly on the other element or indirectly disposed on the other element; when an element is referred to as "connected to" another element, it can be directly connected to the other element or indirectly connected to the other element.

[0217] It should be understood that the orientation or positional relationship indicated by the terms "center", "longitudinal", "transverse", "length", "width", "thickness", "upper", "lower", "front", "rear", "left", "right", "vertical", "horizontal", "top", "bottom", "inner", "outer", etc. is based on the orientation or positional relationship shown in the drawings, and is only for the convenience of describing the present application and simplifying the description, rather than indicating or implying that the device or element referred to must have a specific orientation, be constructed and operated in a specific orientation, and thus should not be construed as a limitation to the present application.

[0218] In addition, the terms "first" and "second" are only used for descriptive purposes and cannot be construed as indicating or implying relative importance or implicitly specifying the quantity of the indicated technical features. Thus, the features defined with "first" and "second" may explicitly or implicitly include one or more of such features. In the description of the present application, the meaning of "a plurality of" and "several" is two or more, unless otherwise specifically defined.

[0219] It should be noted that the structures, ratios, sizes, etc. shown in the attached drawings of this specification are only used to cooperate with the content disclosed in the specification for those familiar with this technology to understand and read, and are not used to limit the implementation conditions of this application. Therefore, they do not have substantial technical significance. Any modification of the structure, change in the proportional relationship, or adjustment of the size, without affecting the effects that this application can produce and the purposes that can be achieved, should still fall within the scope covered by the technical content disclosed in this application.

[0220] By implementing the technical solution of the method for reconstructing the B-scan diagram of rail flaw detection described in the specific embodiments of the present invention, the following technical effects can be achieved:

[0221] (1) The method for reconstructing the B-scan diagram of rail flaw detection described in the specific embodiments of the present invention, through the reconstruction of the B-scan diagram data of the detection, standardizes the detection graphics, lays a foundation for computer flaw recognition, and improves the flaw recognition efficiency;

[0222] (2) The method for reconstructing the B-scan diagram of rail flaw detection described in the specific embodiments of the present invention, identifies the rail joints according to the B-scan diagram of the detection, and automatically calibrates the wheel diameter according to the standard rail length, which can improve the detection accuracy and provide means for accurate flaw positioning and comparison trend analysis;

[0223] (3) The method for reconstructing the B-scan diagram of rail flaw detection described in the specific embodiments of the present invention, automatically adjusts the spatial parameters of the detection wheels of the rail flaw detection vehicle according to the actual B-scan diagram of the detection, which can effectively avoid errors and deviations caused by manual settings, improve the automation level of rail flaw detection, and reduce the dependence of the detection system on the operator's level;

[0224] (4) The method for reconstructing the B-scan diagram of rail flaw detection described in the specific embodiments of the present invention, uses the amplitude of the A-scan as the brightness representation of the B-scan, which is convenient for manual observation and analysis. By setting an appropriate filtering amplitude, detection interference points with smaller amplitudes can be removed;

[0225] (5) The method for reconstructing the B-scan diagram of rail flaw detection described in the specific embodiments of the present invention, uses the centering deviation for auxiliary judgment, which can effectively remove false alarms of flaws in the rail jaw part, and at the same time can warn of the detection channel missing fault, reminding the operator to repair in time to ensure the integrity of the detection graphics.

[0226] The various embodiments in this specification are described in a progressive manner. Each embodiment focuses on the differences from other embodiments. For the same or similar parts among the various embodiments, reference can be made to each other.

[0227] The above are only the preferred embodiments of the present invention, and do not impose any form of limitation on the present invention. Although the present invention has been disclosed above with the preferred embodiments, it is not intended to limit the present invention. Any person skilled in the art can make many possible changes and modifications to the technical solution of the present invention by using the methods and technical contents disclosed above, or modify it into equivalent embodiments with equivalent changes. Therefore, any simple modification, equivalent replacement, equivalent change and modification made to the above embodiments based on the technical essence of the present invention without departing from the technical solution of the present invention still fall within the scope of protection of the technical solution of the present invention.

Claims

1. A method for reconstructing the B-mode diagram in rail flaw detection, characterized in that Including at least one processing procedure for the B-scan diagram of rail flaw detection, as follows: S1) Calibrate the coordinate unit of the detected B-scan diagram according to the rail model; S2) Modify the spatial conversion parameters of the probe wheel chip according to the B-scan diagram data of the rail joint (90); S3) Use the amplitude of the A-scan display as the brightness display of the detected B-scan diagram for manual selection of an appropriate display amplitude, and perform amplitude filtering processing to filter out interference detection points with inappropriate detected amplitude settings; S4) Restore the position of the damage mirror image generated at the rail joint (90) to avoid false alarms of the damage position; S5) Filter out the rail jaw detection points according to the record of the probe wheel centering deviation to eliminate false alarms of damage; S6) Warn of the missing detection channels according to the B-scan diagram data at the rail joint (90); The process S1) further includes the following steps: S101) Judge the rail joint (90) according to the detected B-scan diagram; S102) Calibrate the length of the detected rail (40) by adjusting the wheel diameter ratio coefficient Before adjustment The value is taken as 1; S103) Select standard-length rails according to the rail joint (90), calculate and determine the adjustment wheel diameter ratio coefficient ; S104) According to the adjusted wheel diameter ratio coefficient , use the measured encoder pulse count corresponding to the length of the rail (40) as the calibrated rail length coordinate; The step S101) further includes: Judge the normal bolt holes (80) according to the A-shaped feature diagram in the detected B-scan diagram, and judge the rail joint (90) according to the Y-shaped feature diagram. The bolt holes (80) on both sides of the rail joint (90) are symmetrically distributed and conform to the standard rail model size.

2. The method for reconstructing the B-mode diagram of rail flaw detection according to claim 1, wherein The step S102) further includes: The detection B-type diagram uses the number of encoder pulses installed at the axle end of the wheel set as the coordinate unit, and adjusts the wheel diameter ratio coefficient to detect the length of the rail (40) Calibration: ; where is pi, the wheel diameter ratio coefficient , is the standard wheel set diameter, is the actual wheel set diameter, is the number of pulses output by the encoder for one revolution.

3. The rail flaw detection B-type diagram reconstruction method according to claim 1 or 2, characterized in that The step S103) further includes: Select a standard-length rail without truncation according to the detected B-type diagram: Calculate the length of the rail (40) based on the spacing of the rail joints (90) as ; where is the standard wheel set diameter, is the number of pulses output for one revolution of the corresponding encoder, is the number of pulses corresponding to the measured rail; Let the length of the standard-length rail be , is the rail wheel diameter error. When the length of the detected rail (40) satisfies , determine that the detected rail (40) with length is a standard-length rail without truncation; Then select and calculate the length of another detected rail (40) as , where is the standard wheel set diameter, is the number of pulses output for one revolution of the corresponding encoder, is the number of pulses corresponding to the measured rail; When is satisfied, determine that the detected rail (40) with length is a standard-length rail without truncation; The relative error between the lengths of the two selected rails is within , that is , then the wheel diameter proportionality coefficient is determined by the average value of the encoder pulses corresponding to the two measured rails: .

4. The rail flaw detection B-type diagram reconstruction method according to claim 3, characterized in that In the step S104), substitute the value obtained in the step S103) into the following formula in the step S102) to determine the calibrated rail length coordinates : ​ Among them, is the measured encoder pulse count, is the adjusted wheel diameter ratio coefficient.

5. The method for reconstructing the B-type diagram of rail flaw detection according to claim 1, 2 or 4, characterized in that, The process S2) further includes the following steps: S201) The space conversion parameters are determined by the wafer position. After defining the farthest point detected by the rear probe wheel's ultrasonic wave as the reference point, with the probe wheel (30) taking the center of the wheel axle as the benchmark, the distance between the rear probe wheel and the reference point is , the distance between the center probe wheel and the rear probe wheel is , and the distance between the front probe wheel and the rear probe wheel is ; The front and rear probe wheels are symmetrically installed. The distance between the 0-degree wafer and the wheel axle is , the distance between the 37-degree wafer and the wheel axle is , and the distance between the 70-degree wafer and the wheel axle is ; The distance between the 0-degree wafer of the center probe wheel and the wheel axle is , the distance between the 70-degree wafer with a 70-degree rear offset of the center probe wheel and the wheel axle is , the distance between the 70-degree wafer with a 70-degree front offset of the center probe wheel and the wheel axle is . The space conversion parameters of the wafer are as follows: Rear detection wheel 0-degree wafer: ; Rear detection wheel 37-degree wafer: ; Rear exploration wheel 70-degree wafer: ; Central search wheel 0-degree wafer: ; Center Probe Wheel Rear Bias 70 Degrees Wafer: ; Center probe wheel front offset 70-degree wafer: ; Front exploration wheel 0-degree wafer: ; Front exploration wheel 37-degree wafer: ; 70-degree wafer of the front exploration wheel: ; The inspection wheel (30) is of a fixed structure and has been checked, and the structural parameters of the inspection wheel (30) are constant values; S202) Select the rear probe wheel as the reference, and do not modify the spatial conversion parameters of the rear probe wheel; S203) Select the position of the rail joint (90) in the B-scan diagram detected during the low-speed operation of the rail flaw detector, and calculate the difference in the detection patterns of the 0-degree wafers in the B-scan diagrams of the front and rear detection wheels as , then the corrected spatial conversion parameter of the front detection wheel is: Front exploration wheel 0-degree wafer: ; Front exploration wheel 37-degree wafer: ; 70-degree wafer of the front exploration wheel: ; S204) According to the detected B-type pattern, calculate that the distance difference between the center of the wafer detection pattern when the central search wheel deflects by 70 degrees and the center of the Y-shaped feature pattern is , then the corrected spatial conversion parameters of the central search wheel are: Central search wheel 0-degree wafer: ; Center probe wheel 70 degrees backward wafer: ; Center detecting wheel is deflected 70 degrees forward for the wafer: .

6. The method for reconstructing the B-type diagram of rail flaw detection according to claim 5, wherein The process S3) further includes the following steps: S301) The detected B-scan diagram shows the positions of the detected damage points, and use the A-scan display amplitude corresponding to the ultrasonic echo size of the damage points as the display brightness of the corresponding detection points in the B-scan display.

7. The method for reconstructing the B-mode diagram of rail flaw detection according to claim 6, wherein The step S301) further includes: Calculate the mean value of the A-scan amplitude of the wafer (60) at the bolt hole (80) and the rail joint (90) with respect to the channel detection points , is the wafer channel number, and the amplitude filtering coefficient is adjusted manually , take , as the filtered amplitude; in the normal area where the probe wheel (30) is centered, that is, the centering deviation is within the allowable deviation , , the detection points below the filtered amplitude are not displayed, and the detection B-scan image is regenerated. Manual-assisted damage identification is performed based on the displayed image, and an appropriate amplitude filtering coefficient is selected.

8. The method for reconstructing the B-type diagram of rail flaw detection according to claim 1, 2, 4, 6 or 7, characterized in that, The process S4) further includes the following steps: The mirror position of the detected B-type pattern is determined by the coordinates of the rail joint (90). The ultrasonic wave emitted by the 37-degree rear wafer is reflected by the rail joint (90) to detect a point to form a mirror detection point , and the actual damage is restored according to the point position ; Let the coordinates of the rail joint (90) be , be the angle of the 37-degree rear wafer. Then when the point satisfies the condition , the coordinates of the mirror restoration point are: 。 9. The method for reconstructing the B-type diagram of rail flaw detection according to claim 8, characterized in that, The process S4) further includes the following steps: The ultrasonic wave emitted by the first 37-degree wafer is detected by reflection through the rail joint (90) at point to form a mirror detection point . The actual damage is restored according to the point position; Let the coordinates of the rail joint (90) be . , is the angle of the first 37-degree wafer. Then when satisfies , the coordinates of the mirror restoration point are: 。 10. The method for reconstructing the B-mode diagram of rail flaw detection according to claim 1, 2, 4, 6, 7 or 9, characterized in that, The process S5) further includes the following steps: When the probe wheel (30) deviates from the rail (40) at a curve of the line, the automatic centering system is used to detect, control and record the deviation value at the corresponding position of the detected B-mode diagram. ; When the probe wheel (30) is not centered and the deviation value is greater than the set value and when the ultrasonic bottom echo of the rail from the 0-degree wafer is not lost, the detected reflection points (100) of the rail jaw part of the rail (40) are filtered out in the detected B-mode diagram to achieve off-center filtering; Let the height of the rail jaw from the rail surface be , and the measurement error limit value is When is the ordinate of the B-mode diagram, off-center filtering needs to meet the following three conditions simultaneously: ⅰ) The bottom wave is not lost; ⅱ） ； ⅲ） 。 11. The method for reconstructing the B-mode diagram of rail flaw detection according to claim 10, wherein, The process S6) further includes the following steps: When passing through the area with rail joints (90) and bolt holes (80), except for the side-scan channel, corresponding feature points are displayed in the detected B-scan diagram, and the missing part shows the corresponding channel loss.

Images