A holographic identification method for rail damage

Abstract

The present invention discloses a holographic rail damage identification method, comprising the following steps: calling B-type and A-type inspection data, rail surface images, and profile matching data, and displaying them in multiple windows on the same screen or adjacent to multiple screens; S2) determining whether the B display window calls periodic historical B-type inspection data, and if so, proceeding to step S3); if not, jumping to step S5); S3) the B display window calls periodic historical B-type inspection data for multiple windows on the same screen or adjacent to multiple screens; S4) aligning the positions of the multi-periodic B-type inspection data; S5) matching the mileage of the multi-window display data and playing them sequentially; S6) performing damage identification based on the played multi-window display data, including B-type and A-type inspection data, rail surface and profile matching data; and S7) generating a damage report. The present invention can solve the technical problems of existing damage identification systems, such as a single detection method, low recognition accuracy, high false alarm rate, and low work efficiency.

Description

Technical Field

[0001] The present invention relates to the technical field of rail ultrasonic detection, and in particular to a damage identification method for in-service rails based on a high-speed holographic flaw detection mode. Background Art

[0002] As my country's railway mileage continues to grow, rail transport speeds continue to accelerate, and urban rail transit becomes increasingly developed, the requirements for engineering machinery and equipment to ensure railway and urban rail transit safety are also becoming increasingly stringent. Rail flaw detection vehicles, based on ultrasonic testing principles, are currently widely used in railway engineering systems and urban rail transit.

[0003] With the continuous improvement of rail flaw detection technology and the continuous development of auxiliary technologies such as automatic centering and mileage positioning, large-scale flaw detection vehicles have become fully adapted to high-speed inspections and diverse environmental conditions. However, many deficiencies remain in the field of rail damage analysis and identification, especially in the field of high-speed flaw detection. Due to the huge amount of A-display data, previous large-scale flaw detection vehicles only displayed the A-display data without storing it. Damage identification relied solely on the ultrasonic reflection wave threshold (binarization) and the fusion of the data into B-type image information to identify internal rail damage (system initial screening, manual supplementary identification). Due to the lack of auxiliary information such as the ultrasonic energy information represented by the A-display information and the track status (including precise mileage positioning), damage identification suffers from certain omissions and a large number of false alarms. Therefore, rail flaw detection using large-scale flaw detection vehicles also requires the use of hand-push rail flaw detectors for inspection and verification. In addition, the hand-push flaw detector has a low detection speed and efficiency, and the detection effect is too dependent on human subjective initiative (human technical level and operating status, etc.), and there are obvious limitations in flaw detection on large elevated roads, long bridges and tunnels on high-speed railways, passenger-dedicated lines and other lines.

[0004] In the prior art, the technical solutions more relevant to the present invention mainly include:

[0005] Prior art 1 is a Chinese invention application filed by the present applicant, Zhuzhou Times Electronic Technology Co., Ltd., on June 24, 2019, and published on September 17, 2019, with publication number CN110246134A. This invention discloses a rail damage classification device, which mainly describes damage classification based on the features of the damage image. Prior art 2 is a Chinese invention application filed by Liang Fan and Yu Yang on May 31, 2019, and published on August 27, 2019, with publication number CN110175422A. This invention discloses a multi-cycle rail damage trend prediction method based on data mining. Prior art 3 is a Chinese invention application filed by Liang Fan and Yu Yang on May 31, 2019, and published on August 30, 2019, with publication number CN110188777A. This invention discloses a multi-cycle rail damage data alignment method based on data mining. Prior Art 2 and 3 both describe methods for using Type A data for deep learning identification and periodic data alignment, thereby achieving applicability to the field of small instrument flaw detection. Prior Art 4 is a Chinese invention application filed by Xi'an University of Technology on March 3, 2020, and published on July 10, 2020, with publication number CN111398431A. This invention discloses a multi-gate rail flaw detection system and method with adaptive height, describing a damage identification method that uses Type A data for adaptive gate adjustment.

[0006] The above-mentioned existing technologies 1-4 either classify damage based on damage image features, or use deep learning judgment on Type A data to achieve periodic data alignment, or require adaptive gate adjustment for Type A data to achieve damage judgment. All of them have technical problems such as a single detection method, low damage identification accuracy resulting in a high false alarm rate, and low work efficiency. Summary of the Invention

[0007] In view of this, the purpose of the present invention is to provide a holographic identification method for rail damage to solve the technical problems of the existing rail damage identification system, such as a single detection method, low damage identification accuracy resulting in a high false alarm rate, and low work efficiency.

[0008] In order to achieve the above-mentioned object of the invention, the present invention specifically provides a technical implementation scheme of a rail damage holographic identification method, which comprises the following steps:

[0009] S1) Turn on the rail damage identification mode, call B-type detection data, A-type detection data, rail surface image data and profile matching data, and perform multi-window same-screen or multi-screen adjacent display.

[0010] The B-type data display window is used to display B-type detection data, the A-type data display window is used to display A-type detection data, the video image data display window is used to display rail surface image data, and the profile data display window is used to display profile matching data.

[0011] S2) Determine whether the B-type data display window calls periodic historical B-type detection data. If yes, continue to execute step S3); if not, jump to step S5).

[0012] S3) The B-type data display window calls the periodic historical B-type detection data to display multiple windows on the same screen or multiple screens adjacent to each other;

[0013] S4) Perform multi-cycle B-type detection data position alignment.

[0014] S5) Multi-window display data is matched with mileage and played sequentially.

[0015] S6) performing damage identification based on the multi-window display data including the played B-type detection data, the A-type detection data, the rail surface image data, and the profile matching data, and generating a damage report.

[0016] Furthermore, in step S1), the B-type test data is used to determine whether the rail is damaged. The A-type test data is used to reproduce the ultrasonic echo display state corresponding to each signal point in the B-type test data, assisting in identifying the authenticity of rail damage. The rail surface image data is used to intuitively display the surface state of the rail, eliminating the interference of surface damage and surface state on the B-type test data. The profile matching data is used to provide information on the wear and alignment status of the inner side of the rail, assisting in determining whether the test status is good, serving as a basis for determining the source of poor images in the B-type test data, and assisting in damage identification.

[0017] Furthermore, in step S5), the Type A inspection data, profile matching data, and Type B inspection data with multi-cycle data comparison and rail surface image data are displayed adjacently on the same screen or multiple screens. The Type A inspection data, Type B inspection data, rail surface image data, and profile matching data are positionally calibrated and aligned using synchronous pulse signals from the odometer encoder counts, and then displayed synchronously. Damage analysis and identification are performed based on the characteristics reflected by the various signals, and a damage report is generated.

[0018] Furthermore, in step S3), the periodic historical B-type detection data is the B-type detection data of the previous cycle or the most recent several cycles. When two or more B-type detection data files are called in the B-type data display window, the called B-type detection data files must be rail detection files of the same tested line. If two B-type detection data files are called, they are the current B-type detection data file and the B-type detection data file of the previous cycle. If multiple B-type detection data files are called, they are the current B-type detection data file and several B-type detection data files closest to the current detection time.

[0019] Furthermore, in step S5), the data is synchronously played in the B-type data display window as the main display window. In step S6), when suspected damage or other suspicious signals appear in the B-type inspection data, a comprehensive judgment is made by combining the periodic historical B-type inspection data, A-type inspection data, rail surface image data, and profile matching data at the same location, and the judgment result is marked.

[0020] Furthermore, in step S4), the display direction of two or more B-type test data files is adjusted based on the increase or decrease in line mileage and left / right track information to maintain the same direction. The two or more B-type test data are first roughly aligned based on the line mileage information, and then precisely aligned based on the characteristic image displayed by the B-type test data. In step S6), when the B-type image corresponding to a certain position in the current B-type test data shows suspected damage, the B-type image signal characteristics of the periodic historical B-type test data at the same position are identified. If similar damage signal characteristics are also displayed, the identification is performed and a development trend table is generated.

[0021] Furthermore, step S6) includes: for rail surface damage with continuous characteristics, combining the characteristics in the B-type diagram including the rail head signal and the rail bottom loss wave signal, regional record identification is performed according to the line mileage, and a damage report is automatically generated to focus on monitoring the lines in the area.

[0022] Furthermore, the step S6) includes: indirectly feeding back line abnormalities caused by misaligned rails, high and low joints, and slopes through the rail head position signal in the B-type diagram combined with the rail surface light band in the video image data display window, recording the line mileage section and automatically generating a damage report.

[0023] Furthermore, the step S6) includes: in the B-type data display window, when the B-display signal corresponding to the B-type detection data appears at a position including the rail head and a suspected weld position but cannot be clearly identified, the rail surface image at the same position in the video image data display window can be viewed to assist in judging rail damage including surface scratches, surface fish scales, peeling blocks, and weld bars, and the damage results of the rail damage identification are integrated to automatically generate a damage report.

[0024] Furthermore, step S6) includes: when the data of a certain channel in the B-type detection data is incomplete or the number of points is small, or the channel gain is not adjusted in time due to the change of speed level, the gain of the channel is amplified by the A-type detection data, and the B-type detection data is reconstructed, and then the position of the damage echo appearing in the fixed gate and the relative relationship between the damage echo amplitude and the threshold are used to assist in damage identification.

[0025] Furthermore, step S6) includes: when a suspected damage is found through the B-type detection data and a judgment cannot be made, the signal corresponding to the A-type detection data can be checked, and whether the damage echo appears within a specific gate range, and whether the damage echo appears within the primary wave gate or the secondary wave gate, it can be judged whether it is a reflection echo from other specific positions of a non-damage defect.

[0026] Furthermore, the step S6) includes: feeding back the working state of the damage detection device through the position deviation of the damage echo in the A-type detection data, and adjusting the damage detection device in the next cycle accordingly.

[0027] Furthermore, step S6) includes: using profile matching data to provide feedback on the wear status of the inner side of the rail, recording track sections with significant wear, and automatically generating a damage report. The rail profile matching data is then used to generate segment statistics using synchronous positioning based on line mileage, and a wear area threshold is set. Sections where the rail wear area exceeds the threshold are monitored.

[0028] Furthermore, step S6) includes: performing comprehensive identification by combining profile matching data with rail surface image data and type B detection data, recording line sections with poor detection results and automatically generating a damage report as a basis for repeated detection or key monitoring. At the same time, it can also indirectly feedback the working status of the detection equipment, provide a reference for parameter correction of the detection equipment including centering reference and inclination, and provide original data reference for rail grinding.

[0029] Furthermore, the ultrasonic inspection unit generates Type B and Type A inspection data for point-to-point alignment based on the mileage encoder counts. The video image inspection unit generates rail surface image data based on the synchronous pulse signals from the mileage encoder counts. The laser automatic centering unit generates profile matching data based on the synchronous pulse signals from the mileage encoder counts.

[0030] By implementing the technical solution of the rail damage holographic identification method provided by the present invention, the following beneficial effects are achieved:

[0031] (1) The present invention provides a holographic rail damage identification method, establishing a high-speed holographic flaw detection mode. The B-type (detection) data generated by the detection is integrated with the existing A-type (detection) data storage and display technology, surface video data, and profile matching data to build a damage data analysis and identification system, thus realizing the analysis and identification of detection data in the high-speed holographic flaw detection mode.

[0032] (2) The rail damage holographic identification method of the present invention fully utilizes the mileage synchronization function based on the mileage encoder (which can be calibrated with radio frequency tags), integrates the detection data of multiple dimensions such as ultrasonic detection technology, high-speed video imaging technology, and laser profile detection technology, and has a great possibility of reproducing the ultrasonic detection status and line status;

[0033] (3) The rail damage holographic identification method of the present invention imports multi-dimensional information such as periodic historical data, which can be compared with periodic historical data to analyze damage development trends, greatly improving the accuracy of damage identification, reducing false alarm rates, strengthening line monitoring, and improving work efficiency;

[0034] (4) The present invention's holographic rail damage identification method utilizes point-to-point A-display signals to assist in damage identification. By storing and calling the A-display signals, the position and amplitude of the damage wave in the A-display image are reproduced during damage identification, thereby improving the accuracy of damage identification. At the same time, it can also selectively adjust local gain, making up for the shortcomings of the B-display signal.

[0035] (5) In the holographic rail damage identification method of the present invention, the periodic data is B-type (detection) data. By importing the B-type (detection) data to be compared, and then aligning it through the mileage positioning signal based on the encoder + B-type image feature signal, referring to the B-type data and using the A-type display fusion to judge the damage, there is no need to adaptively adjust the gate parameters;

[0036] (6) The holographic identification method for rail damage of the present invention adopts an alignment method based on B-type data to perform multi-cycle data comparison and fusion judgment, thereby improving the damage recognition rate; by calling periodic historical B-type data, based on the B-type data signal, alignment is performed by detecting the direction, mileage increase or decrease, mileage value, B-type map feature information of special positions of the line, and then multi-cycle data comparison is performed on the suspected damage position, which can greatly improve the accuracy of damage identification, generate a damage development trend table, strengthen line monitoring, and provide a large amount of information on the service life of the track in different environments and different operating conditions. It is of great significance to improve the comprehensiveness and accuracy of rail damage identification and detection and improve the safety of railway operation. BRIEF DESCRIPTION OF THE DRAWINGS

[0037] To more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the following briefly introduces the drawings required for use in the embodiments or the description of the prior art. Obviously, the drawings described below are only some embodiments of the present invention. For those skilled in the art, other embodiments can be derived from these drawings without inventive effort.

[0038] Figure 1 This is a schematic diagram of the B-type display interface of the present invention;

[0039] Figure 2 This is a schematic diagram of the A-type display and trigger signal waveform of the 0-degree channel ultrasonic chip in the present invention;

[0040] Figure 3 This is a system structure diagram of a specific embodiment of the rail damage holographic identification system based on the method of the present invention;

[0041] Figure 4 This is a flowchart of a specific implementation of the rail damage holographic identification method of the present invention;

[0042] Figure 5 This is a schematic diagram of a damage identification display interface in a specific embodiment of the rail damage holographic identification method of the present invention;

[0043] Figure 6 This is a waveform diagram of a rail damage holographic identification method according to a specific embodiment of the present invention, in which damage identification is assisted by an A-type display;

[0044] Figure 7 This is a schematic diagram of a periodic data alignment and synchronous playback interface in a specific embodiment of the rail damage holographic identification method of the present invention;

[0045] Figure 8 2. It is a schematic diagram of a video image-assisted damage identification interface in a specific embodiment of the rail damage holographic identification method of the present invention;

[0046] Figure 9 This is a schematic diagram of a profile matching information display interface in a specific embodiment of the rail damage holographic identification method of the present invention;

[0047] In the figure: 1-onboard system, 11-ultrasonic detection unit, 12-video image detection unit, 13-laser automatic centering unit, 2-damage identification system, 20-damage identification display unit, 21-B type data display window, 22-A type data display window, 23-video image data display window, 24-profile data display window, 25-damage identification processing unit, 26-damage report, 3-rail. DETAILED DESCRIPTION

[0048] For the purpose of reference and clarity, the technical terms, abbreviations or abbreviations used below are recorded as follows:

[0049] Type A display: It is a display method that displays analog ultrasonic signals through an oscilloscope (or virtual oscilloscope). The horizontal direction is the time quantity of the signal, and the vertical direction is the amplitude of the ultrasonic reflection signal of the reflector.

[0050] Type B display: It is a display method that intuitively displays the information of ultrasonic reflection points inside the rail through images. The horizontal direction is the mileage position of the ultrasonic reflection point, and the vertical direction is the burial depth of the reflection point.

[0051] To make the purpose, technical solutions, and advantages of the embodiments of the present invention more clear, the technical solutions in the embodiments of the present invention will be clearly and completely described below in conjunction with the accompanying drawings of the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without making creative efforts are within the scope of protection of the present invention.

[0052] As attached Figure 1 To the attached Figure 9 As shown, a specific embodiment of the rail damage holographic identification method of the present invention is given, and the present invention is further described below in conjunction with the drawings and specific embodiments.

[0053] The rail flaw detection system of the rail flaw detection vehicle generally has two graphic windows, A-type display and B-type display, for flaw judgment. Among them, the A-type display displays the ultrasonic echo simulation signal through an oscilloscope (or virtual oscilloscope), as shown in the attached Figure 2 The B-type display intuitively displays the information of the ultrasonic signal reflection points inside the rail through images, as shown in the attached figure. Figure 1As shown. Type A display has many channels and a large amount of data, so it is only displayed and not saved. However, Type A display contains a lot of useful information, such as the amplitude of the echo signal, whether the excitation signal has a tail, and the time domain value of the excitation signal from the interface reflection signal, which is of great significance for flaw detectors to judge the rail surface condition, whether the alignment is good, whether the wheel is filled with enough fluid, and whether the wheel is broken. As shown in the attached Figure 2 As shown in the waveform in the upper middle part, taking the ultrasonic chip with an incident angle of 0 degree as an example, a typical A-type display wave signal includes ultrasonic excitation signal B, interface wave gate C, monitoring gate E and rail bottom gate F, among which an ultrasonic echo signal D may be received in the monitoring gate E. Figure 2 The lower middle portion shows the waveform of trigger signal A, with the time interval t between the falling edge of trigger signal A and the rising edge of ultrasonic excitation signal B. By introducing a gate, only ultrasonic echo signals within the gate interval are collected. This gate interval corresponds to the sound length of a single ultrasonic echo or a specific rail flaw detection area. The detectable rail depth is half the product of the ultrasonic signal's propagation velocity and propagation time in the rail, minimizing the impact of noise introduced by inactive areas in the time domain. The gate delay and width parameters correspond to the detectable rail interval and can be set using the gate parameter setting module on the flaw detection vehicle's host computer. Type B displays, corresponding to the flaw detection system, display a longitudinal cross-section of the inspected rail and indicate the approximate size and relative position of ultrasonic reflectors within the rail. The horizontal axis of the cross-section represents the mileage of the flaw, while the vertical axis represents the flaw's depth below the rail surface.

[0054] An embodiment of a holographic rail damage identification system, based on the holographic rail damage identification method of the present invention, specifically comprises: an onboard system 1 and a damage identification system 2. The damage identification system 2 comprises a workstation (or a computer system), the core of which is a damage identification software module (including a display and identification component, namely, a damage identification display unit 20 and a damage identification processing unit 25). The onboard system 1 includes an ultrasonic detection unit 11, a video image detection unit 12, and a laser automatic centering unit 13, and may also include other expandable systems. The damage identification system 2 includes a damage identification display unit 20, which includes a B-type data display window 21, an A-type data display window 22, a video image data display window 23, and a profile data display window 24. The ultrasonic detection unit 11, the video image detection unit 12, and the laser automatic centering unit 13 achieve mileage synchronization by sharing a mileage encoder signal (which may be supplemented by mileage calibration using radio frequency tags). The ultrasonic detection unit 11 generates point-to-point aligned B-type and A-type detection data based on mileage encoder counts. The video image detection unit 12 generates rail surface image data based on the synchronous pulse signal counted by the mileage encoder. The laser automatic centering unit 13 generates profile matching data based on the synchronous pulse signal counted by the mileage encoder, and can calculate the rail profile wear. Type B and Type A detection data, rail surface image data and profile matching data are sent to the damage identification display unit 20, and are displayed in the Type B data display window 21, the Type A data display window 22, the video image data display window 23 and the profile data display window 24 respectively. All data information files in the holographic flaw detection mode are sent to the damage identification system 2 (industrial computer) through the network (or dump device), and the Type B and Type A detection data, rail surface image data and profile matching data are analyzed and processed to generate a damage report 26. As shown in the attached figure Figure 3 As shown, W1 is the B-type data display window 21, W2 is the A-type data display window 22, W3 is the video image data display window 23, W4 is the profile data display window 24, and W5 is the B-display reconstruction window based on the A-type inspection data. The damage identification system 2 also includes a damage identification processing unit 25. The damage identification processing unit 25 and / or manual damage judgment methods analyze and process the B-type and A-type inspection data, rail surface image data, and profile matching data to generate a damage report 26.

[0055] Among them, the B-type detection data is an image display form generated by the system after processing the ultrasonic echo. It is used to judge whether the rail is damaged and is an important identification basis for the system to judge whether the rail 3 is damaged. The A-type detection data is used to reproduce the ultrasonic echo display status corresponding to each signal point of the B-type detection data, and assist in identifying the authenticity of the rail damage (assisting in identifying whether it is real damage information) to improve the recognition accuracy. The rail surface image data is used to intuitively display the surface status of the rail and eliminate the interference of surface damage and surface status on the B-type detection data. The (laser centering) profile matching data is used to provide the wear status and centering status information of the inner side of the rail 3, assist in judging whether the detection status is good, serve as the basis for judging the source of bad images in the B-type detection data, and assist in damage identification.

[0056] The damage identification display unit 20 displays type A detection data, profile matching data, and type B detection data and rail surface image data with multi-cycle data comparison in the same display window. Type A detection data, type B detection data, rail surface image data and profile matching data are position calibrated, aligned and synchronously displayed through the synchronous pulse signal based on the mileage encoder count. Damage analysis and identification are performed based on the characteristics reflected by different signals, and finally the damage identification processing unit 25 analyzes and processes to automatically generate a damage report 26, and damage identification can be performed through manual intervention. The type B data display window 21 is also used to display periodic historical type B detection data, which is the type B detection data of the previous cycle or the most recent several cycles. Periodic historical type B detection data is historical detection data that is relatively recent, and generally the detection data of the previous cycle or the most recent two cycles are preferred.

[0057] The B-type data display window 21 is used as the main display window to play data synchronously. When suspected damage or other suspicious signals appear in the B-type detection data, a comprehensive judgment is made by combining the periodic historical B-type detection data, A-type detection data, rail surface image data and profile matching data at the same position, and the judgment result is marked. The display direction of two or more B-type detection data files is adjusted and kept consistent by the increase or decrease of line mileage and left / right rail information. First, two or more B-type detection data are roughly aligned according to the line mileage information, and then accurately aligned by the characteristic image displayed by the B-type detection data. When the B-type image corresponding to a certain position of the current B-type detection data is displayed as suspected damage, by identifying the B-type image signal characteristics of the periodic historical B-type detection data at the same position, if similar damage signal characteristics are also displayed, identification is performed and a development trend table is generated.

[0058] When performing detection data playback and identification, the display playback function module of the damage identification software can be used to call the periodic historical data for simultaneous or multi-screen adjacent display, and through direction identification + mileage positioning based on encoder signals (RFID tag mileage calibration can be used) + B-type graph characteristic signal points and other methods, the suspected damage position can be accurately located, the damage development trend can be identified, and a damage development trend table can be automatically generated. When two or more B-type detection data files are called in the B-type data display window 21, the called B-type detection data files must be rail detection files of the same tested line. If two B-type detection data files are called, they are the current B-type detection data file and the B-type detection data file of the previous cycle. If multiple B-type detection data files are called, they are the current B-type detection data file and several B-type detection data files closest to the time of this detection.

[0059] The rail damage holographic identification system described in this embodiment adopts a high-speed holographic flaw detection mode for intelligent damage identification, and performs high-speed holographic flaw detection of in-service rails based on mileage synchronization of encoder signals (which can be calibrated with radio frequency tag mileage), integrating ultrasonic detection, co-storage and co-display of type A and type B detection signals, automatic laser centering, profile matching, and high-speed video recording of the rail surface. The obtained ultrasonic detection type B detection data file, ultrasonic detection type A detection data file, surface video information file, (laser centering) profile matching file, and periodic historical type B data file are integrated into the data playback and analysis system of the damage identification software for fusion damage judgment, and a damage report 26 can be automatically generated. The rail damage holographic identification system described in this embodiment proposes a high-speed holographic flaw detection mode for in-service rails, using a large-scale flaw detection vehicle as the core. This system leverages existing technologies and integrates them to build a rail damage identification system. This system improves damage identification accuracy through theoretical and technical support, replaces the inspection cycle of manual flaw detectors with large-scale flaw detection vehicles, and significantly reduces (or even eliminates, once damage identification confidence is improved) the review work required for manual flaw detectors.

[0060] For rail surface damage with continuous characteristics, the system combines features such as rail head signals and rail bottom loss signals in the B-type image to identify regional damage based on line mileage, automatically generating a damage report 26 to focus on monitoring the line in that area. By combining the rail head signals in the B-type image with the rail surface light bands in the video image data display window 23, it indirectly reports line anomalies such as misaligned rails, high and low joints, and slope, recording line mileage sections and automatically generating a damage report 26.

[0061] When replaying and identifying test data, suspected damage points displayed on the B-type image are identified as true damage waveforms (excluding phantom waves and grass noise) by the location of the corresponding damage waveform in the A-type image within the gate. This determination automatically generates a damage report 26. If a channel in the B-type test data is incomplete or has a small number of points, or if the channel gain is not adjusted in a timely manner due to changes in velocity levels, the amplitude of the corresponding A-type damage waveform within the gate can be used to determine the damage. The gain of the channel can be locally amplified using the A-type test data, and the B-type test data can be reconstructed. The location of the damage echo within the fixed gate and the relative relationship between the damage echo amplitude and the threshold are then used to comprehensively determine whether the damage is true. This determination automatically generates a damage report 26.

[0062] When a suspected damage is found through the B-type detection data but a judgment cannot be made, the signal corresponding to the A-type detection data can be checked, and the damage echo can be judged by whether it appears within a specific gate range, and whether the damage echo appears within the primary wave gate or the secondary wave gate to determine whether it is a reflection echo from another specific location that is not a damage defect. The position deviation of the damage echo in the A-type detection data is used to feedback the working status of the damage detection equipment, and the damage detection equipment is adjusted accordingly for the next cycle. The wear status of the inner side of the rail 3 is fed back through the profile matching data, and the line sections with greater wear are recorded and a damage report 26 is automatically generated. The profile matching data of the rail 3 is synchronously positioned through the line mileage for section statistics, and a wear area threshold is set. Sections where the rail wear area exceeds the wear area threshold are monitored in detail.

[0063] In the B-type data display window 21, when B-display signals corresponding to B-type inspection data appear at locations including the rail head and suspected welds of the rail 3 but cannot be clearly identified, the rail surface image at the same location in the video image data display window 23 can be used to assist in determining rail damage, including surface scratches, surface fish scales, flaking, and weld reinforcement. This, combined with the damage identification and display unit 20's assessment, automatically generates a damage report 26. By combining profile matching data with rail surface image data and B-type inspection data for comprehensive identification, track sections with poor inspection results are recorded and automatically generated. This report serves as a basis for repeated inspections or focused monitoring. It also indirectly provides feedback on the operating status of the inspection equipment, providing a reference for parameter corrections, including alignment reference and inclination, and providing raw data reference for rail grinding. By comparing video image data with B-type inspection data and referencing rail profile matching information, rail surface damage is comprehensively identified and a damage report 26 is automatically generated. This report also records section-specific status information.

[0064] The above-mentioned embodiment of the present invention specifically provides a holographic rail flaw detection mode. The multi-dimensional data (ultrasonic A display, B display, video image, laser sensor centering status, etc.) files obtained by the vehicle-mounted system 1 can all be called and analyzed and processed by the damage identification system (based on the workstation) to generate a damage report 26.

[0065] The damage identification system 2 can display type A detection data, profile matching data, type B detection data with multi-cycle data comparison, and rail surface image data (i.e., video image data) in the same display window. All fused detection data (except for periodic historical type B detection data) are calibrated and displayed synchronously through synchronous pulse signals based on mileage encoder counts. Damage analysis and identification are performed based on the characteristics reflected by different signals, and a damage report is automatically generated (damage identification can be performed manually). An embodiment of the rail damage holographic identification method of the present invention based on the above-mentioned rail damage holographic identification system is shown in the attached flow chart of its display function. Figure 4 As shown, the method specifically includes the following steps:

[0066] S1) The rail damage identification mode is activated, and Type B inspection data, Type A inspection data, rail surface image data, and profile matching data are retrieved and displayed in multiple windows on the same screen or adjacent to each other. The Type B data display window 21 is used to display Type B inspection data, the Type A data display window 22 is used to display Type A inspection data, the video image data display window 23 is used to display rail surface image data, and the profile data display window 24 is used to display profile matching data.

[0067] In the damage identification system, the damage identification software is opened and the detection data file in the holographic flaw detection mode is called. The file includes the A-type detection data file and the B-type detection data file generated by the ultrasonic detection unit 11, the video image detection data file, and the profile matching data file. The files are displayed on the same screen or multiple screens adjacent to each other in the damage identification software display window (i.e., the damage identification display unit 20) and can be divided into different functional partitions for display, as shown in the attached figure. Figure 5 shown.

[0068] S2) Determine whether the B-type data display window calls periodic historical B-type detection data. If yes, continue to execute step S3); if not, jump to step S5).

[0069] S3) The B-type data display window calls the periodic historical B-type detection data to display multiple windows on the same screen or multiple screens adjacent to each other.

[0070] The periodic historical B-type detection data file is called according to demand, and the B-type data display window 21 of the damage identification display unit 20 is displayed in the same screen or multiple screens.

[0071] S4) Perform multi-cycle B-type detection data position alignment.

[0072] Multi-data position alignment is achieved through comprehensive identification based on factors such as detection direction, mileage increase or decrease, mileage position, and B-type map characteristic signals.

[0073] S5) Multi-window display data is matched with mileage and played sequentially.

[0074] The signals of all windows displayed by the damage identification display unit 20 (including the B-type data display window 21, the A-type data display window 22, the video image data display window 23 and the profile data display window 24) are aligned by mileage matching, and the data are played synchronously and sequentially with the B-type data display window 21 as the main display window.

[0075] S6) Damage identification is performed based on the multi-window display data including the played B-type detection data, A-type detection data, rail surface image data and profile matching data, and a damage report 26 is generated.

[0076] When suspected damage or other suspicious signals appear in the B-type (data) display, a comprehensive judgment is made by combining the periodic historical B-type detection data, A-type detection data, video image data (i.e., rail surface image data), profile data, and other auxiliary information at the same location, and the judgment result is marked. The periodic historical B-type detection data can be compared with the current B-type detection data to determine the development trend of rail damage, improve the accuracy of damage identification, and generate a damage trend development table, which can also be accompanied by comparison pictures. When the B-type detection data has incomplete signal channels in certain locations, abnormal signal display positions, abnormal clutter, etc., the position, amplitude, and gain adjustment functions of the damage waveform display in the A-type detection data can also be used to assist in judgment, reduce the missed detection rate of rail damage, and improve the accuracy of damage identification. The information displayed on the rail head in the B-type diagram can be combined with the video image data for judgment, which can improve the accuracy of identifying surface damage and weld damage, and can also mark surface damage in sections to provide a basis for later line monitoring. Profile data can reflect the degree of wear on the inner side of the rails and can be combined with B-type inspection data for analysis. Ultrasonic inspection results can be provided in sections for comprehensive damage identification and section marking to inform track status. A damage report 26 is automatically generated, incorporating various diagnostic information. Separate damage reports can also be generated based on specific or multiple types of conclusions.

[0077] Large rail flaw detection vehicles currently primarily identify damage using signals displayed on the B-type display. This lack of real-time detection status information (especially the A-display information that characterizes the real-time energy of the signal) often leads to false alarms of damage. In the following situations, a holographic rail damage identification method based on the aforementioned specific embodiment of the present invention uses A-type detection data to assist in rail damage identification, thereby improving the accuracy of damage identification. The method specifically includes the following steps:

[0078] S11) When the signal of a channel in the B-type image is incomplete, or the inspector fails to adjust the channel gain in time due to factors such as changes in speed level, the A-type display can be used to amplify the channel gain, and then the damage wave that appears in the fixed gate and the relative relationship between the damage wave amplitude and the threshold value can be used to assist in the identification of rail damage. For example, when the inspection speed increases from 30km / h to 70km / h, the inspector does not increase the gain value of a certain channel synchronously with the change in speed. In the B-type display image, the number of characteristic signal points of this channel is obviously small. If there is suspected damage in this section, the system will miss it due to the small number of points. At this time, the channel gain can be amplified (quantified according to the speed level) through the A-display data information, and the B-display can be reconstructed. Then, the number of reconstructed information points and the position can be used to comprehensively judge whether it is real damage.

[0079] S12) When a suspected defect is found but a judgment cannot be made, the signal corresponding to the A-type display can be checked to see if the defect waveform appears within a specific gate range. It can also be determined whether it is a reflection from another specific location that is not a defect by finding whether it is within the primary wave gate or the secondary wave gate.

[0080] S13) At the same time, the wave position (such as the attached Figure 6 (As shown in Figure 1) The state of the deviation feedback device during detection is used to facilitate device adjustment in the next cycle, as shown in the attached Figure 6 shown.

[0081] S14) Mark the identification results in the damage identification software, and the system automatically generates a damage report 26.

[0082] For example, if a damage waveform detected by a primary wave on a channel has a matching damage echo (t1) within the first pulse signal's transmission time (T0), it is considered a true damage waveform. However, if this echo appears within the second pulse signal's transmission time (T0 + t2), it is not a true damage waveform but rather a reflection from another fixed source that was received by the monitoring gate of the second pulse and thus interpreted by the system as the true echo signal of the second pulse. False alarms such as these can be identified by calling the A-type display information from the B-type graph and comprehensively identifying the damage waveform's specific location within the gate on the A-display.

[0083] As attached Figure 7 As shown, based on the rail damage holographic identification method of the above-mentioned specific embodiment of the present invention, a method for comparing multi-cycle data and generating a development trend table is also provided. The method specifically includes the following steps:

[0084] S21) Two or more Type B test data files are called in the Type B data display window 21 of the damage identification display unit 20. The called data files must be test files for the same circuit under test. If two files are called, they are generally the current test data file and the previous test data file. If multiple files are called, they are the current test data file and several test data files with the most recent time.

[0085] S22) adjusting the display directions of the plurality of inspection data files according to information such as mileage increase / decrease, left / right track, and so on, and maintaining consistency.

[0086] S23) Data alignment is performed by first performing a rough alignment based on the line mileage information, and then performing a fine alignment using the characteristic images displayed on the B-type map. B-type map information with obvious features, such as switches, joints, conductor holes, splints, and welds, are preferred reference points.

[0087] S24) Identify rail damage and generate a development trend table. When the B-type image at a certain position in the current detection data file shows suspected damage, the system identifies the signal characteristics of the B-type image at the same position in the historical period data. If similar damage signal characteristics are also displayed (perhaps the number of signal points is small and the system did not identify it as damage this time), it will identify it and generate a development trend table. The table is accompanied by a partially enlarged B-type display image for easy viewing at any time.

[0088] As attached Figure 8 As shown, based on the rail damage holographic identification method of the specific embodiment of the present invention, a method for assisting damage identification by using high-speed video images and profile matching information is also provided. The method specifically includes the following steps:

[0089] S31) In the B-type data display window 21 of the damage identification display unit 20, when the displayed B-type signal appears at the rail head portion, suspected weld portion, etc. of the rail 3 and cannot be clearly identified, the video image of the rail at the same position (i.e., rail surface image data) can be viewed to assist in the judgment, such as surface abrasions, surface fish scales, peeling, weld bars, etc., and the damage is integrated and confirmed, and a damage report 26 is automatically generated.

[0090] S32) For some rail surface damage with continuous characteristics, combined with the rail head signal and rail bottom loss wave signal in the B-type diagram, regional record identification can be performed based on line mileage, and a report 26 can be automatically generated to facilitate key monitoring of the rails in the area.

[0091] S33) By combining the signal from the rail head of the B-type image with the rail surface light band in the rail video image, it is possible to indirectly feedback the abnormalities of the line rails, such as misaligned teeth, high and low joints, slope, etc., record the line mileage section and automatically generate a damage report 26.

[0092] For example: When replaying and analyzing a certain test data, the rail head at the left clamping plate position has suspected damage information (see the attached Figure 8 The elliptical frame in the middle indicates the location of the damage, but the specific situation is uncertain. In this case, the rail surface image of the location can be retrieved through the synchronized mileage to further identify the damage. From the rail surface image, it can be seen that the location is the rail surface crush damage, as shown in the attached figure. Figure 8 shown.

[0093] S34) Rail profile matching data can provide feedback on the wear of the inner side of the rails, record the sections with greater wear and automatically generate a damage report 26. It can also be combined with rail video images and B-type display data for comprehensive identification, record the sections with poor detection results and automatically generate a damage report 26, which serves as a basis for repeated detection or key monitoring. It can also indirectly provide feedback on the status of the equipment during detection, provide a reference for the correction of parameters such as the equipment centering reference and inclination, and provide original data reference for rail grinding. Figure 9 As shown, G shows the wear position and H shows the wear area. The profile matching data in the figure is synchronously positioned through line mileage for section statistics. The rail wear area threshold is set (which can be a reference standard value or an empirical value). The line section where the wear area exceeds the threshold can be monitored in detail.

[0094] The holographic identification method for rail damage described in the specific embodiment of the present invention above adopts a holographic flaw detection mode to perform multi-dimensional data fusion and judgment, and adopts different software to call the data and display it on the same display screen (or multiple display screens in adjacent positions), or adopts the same software to only reduce the number of window displays or display positions. Such display methods for realizing multi-dimensional fusion judgment are all within the scope of protection requested by the present invention. At the same time, the fusion judgment method described in the specific embodiment of the present invention, whether it is manual judgment or automatic judgment by the system, or a combination of manual and automatic judgment by the system, whether it is manual generation of damage reports or automatic generation of damage reports, are all within the scope of protection requested by the present invention. The following is a specific explanation by way of example.

[0095] In the specific embodiment of the rail damage holographic identification method, the main body of the display window is still the B-type data display window 21. Other data are based on the B-type diagram and are synchronized on the basis of mileage, and are used to further improve the accuracy of damage identification and the integrity of the damage report. In the process of test data playback, more other types of test data are integrated in, and the rail damage holographic identification method can adopt automatic identification + manual intervention for comprehensive processing. First, it is based on automatic identification, and secondly, for each test data, the manual intervention control button and the parameter setting control button are retained (some are input identification conditions, which are set according to specific data) for manual identification and confirmation, and finally a complete damage report 26 is generated automatically. The following is a detailed description of each data display window:

[0096] (1) For the display of type A detection data on the same screen or multiple screens. For example: during the manual playback of type B detection data, there is a 10-kilometer-long type B display signal, the number of points in the 45° channel is relatively small, and the type A detection data has a significantly lower amplitude (it was found that the speed in this section increased from 40km / h to 70km / h, and the operator did not increase the gain according to the speed increase). Then, when playing back the type B detection data, use the manual intervention control key to find the channel, manually increase 3dB, and enter the mileage range that needs intervention. The damage identification software automatically generates the type A detection data display result after intervention and generates a new type B diagram. For example: a type B signal at a certain position, such as the 70° phantom wave at the screw hole position, from the perspective of the type A display signal, the position of its damage waveform is incorrect and should not be judged as damage, but the type B diagram shows it as damage. At this time, after comprehensive judgment, the manual intervention judges the type B display information as "normal". After multiple manual interventions for this type of damage, it will be automatically judged as "normal" after machine learning. The final damage report 26 is automatically generated and the results after manual intervention are retained.

[0097] (2) Display of video image data. The video image data display window 23 can automatically identify surface damage based on the captured video image, and automatically classify and count it (such as surface scratches, chipping, fish scales, etc.). The automatically counted and classified surface damage table has the function of manual reconfirmation, such as: based on the information displayed by the B-type detection data and the careful identification of the rail surface image, it is found that this is not a scratch but a weld, then manual intervention is performed to identify it as a "weld". For example: the surface damage has obvious signal characteristics on the B-type image (if you only look at the B-type image, you cannot identify the specific type of surface damage). When the surface damage is confirmed by combining the video image data, the waveform information on the B-type image can be automatically qualitatively identified. The manual intervention control button is also retained (that is, the identification control button is set). After machine learning, the automatic recognition level is higher. Finally, the damage report 26 is automatically generated, and the results after manual intervention are retained.

[0098] (3) For profile data display. Here, the main purpose is to use the profile information to calculate rail wear. The damage identification software can set several wear levels, for example: wear greater than 10mm 2 , it is automatically recorded in the B-type data display window 21. When replaying B-type test data, the system comprehensively considers the causes of noise or abnormal B-type signals. When identifying suspected damage in B-type test data, it can add some manual intervention damage identification and notes. Finally, in the automatically generated damage report 26, the notes of increased wear are automatically added, and the results after manual intervention are retained.

[0099] (4) Comparison of periodic historical B-type inspection data. The periodic historical inspection data can be replayed synchronously, which can be targeted at B-type inspection data only, or further targeted at rail surface image data. The damage recognition software first automatically performs mileage alignment and retains manual adjustments (if it is found that some positions of the system are not aligned, it can be manually revised again). For the suspected damage identified this time, the damage recognition software automatically generates a damage report 26, which includes a picture of the current B-type inspection data display, an enlarged picture of the B-type inspection data display, and a comparison picture of the development trend of the last B-type inspection data display picture. Manual intervention operations are retained. If a suspected damage is not desired to be compared, it can be manually canceled through the control button. Finally, a damage report 26 with a picture of the development trend of rail damage is automatically generated, and the results after manual intervention are retained.

[0100] (5) Damage identification processing. All data information can be combined to perform damage assessment, either automatically or with manual intervention. Finally, the damage identification software generates a damage report 26, which contains all the information that needs to be reflected or expected to be reflected in the inspection data.

[0101] By implementing the technical solution of the rail damage holographic identification method described in the specific embodiment of the present invention, the following technical effects can be achieved:

[0102] (1) The rail damage holographic identification method described in the specific embodiment of the present invention establishes a high-speed holographic flaw detection mode, integrates the B-type (detection) data generated by the detection with the existing A-type (detection) data storage and display technology, surface video data and profile matching data, and builds a damage data analysis and identification system, realizing the analysis and identification of detection data in the high-speed holographic flaw detection mode;

[0103] (2) The rail damage holographic identification method described in the specific embodiment of the present invention fully utilizes the mileage synchronization function based on the mileage encoder (which can be calibrated with radio frequency tags), integrates the detection data of multiple dimensions such as ultrasonic detection technology, high-speed video imaging technology, and laser profile detection technology, and has a great possibility of reproducing the ultrasonic detection status and line status;

[0104] (3) The rail damage holographic identification method described in the specific embodiment of the present invention imports multi-dimensional information such as periodic historical data, which can be compared with the periodic historical data to analyze the damage development trend, greatly improving the accuracy of damage identification, reducing the false alarm rate, strengthening line monitoring, and improving work efficiency;

[0105] (4) The rail damage holographic identification method described in the specific embodiment of the present invention utilizes point-to-point A-display signals to assist damage identification. By storing and calling the A-display signals, the position and amplitude of the damage wave in the A-display image are reproduced during damage identification, thereby improving the accuracy of damage identification. At the same time, local gain adjustment can be selectively performed to compensate for the shortcomings of the B-display signal.

[0106] (5) In the rail damage holographic identification method described in the specific embodiment of the present invention, the periodic data is B-type (detection) data. By importing the B-type (detection) data to be compared, and then aligning it through the mileage positioning signal + B-type image feature signal based on the encoder, referring to the B-type data and using the A-type display fusion to judge the damage, there is no need to adaptively adjust the gate parameters;

[0107] (6) The holographic identification method for rail damage described in the specific embodiment of the present invention adopts an alignment method based on B-type data to perform multi-cycle data comparison and fusion damage judgment, thereby improving the damage recognition rate; by calling periodic historical B-type data, based on the B-type data signal, alignment is performed by detecting the direction, mileage increase or decrease, mileage value, B-type map feature information of special positions of the line, and then multi-cycle data comparison is performed on the suspected damage position, which can greatly improve the accuracy of damage identification, generate a damage development trend table, strengthen line monitoring, and provide a large amount of information on the service life of the track in different environments and different operating conditions. It is of great significance to improve the comprehensiveness and accuracy of rail damage identification and detection and improve the safety of railway operation.

[0108] The various embodiments in this specification are described in a progressive manner, and each embodiment focuses on the differences from other embodiments. The same or similar parts between the various embodiments can be referenced to each other.

[0109] The above description is only a preferred embodiment of the present invention and does not constitute any formal limitation to the present invention. Although the present invention has been disclosed as above in terms of a preferred embodiment, it is not intended to limit the present invention. Any person skilled in the art can, without departing from the spirit and technical solution of the present invention, use the methods and technical contents disclosed above to make many possible changes and modifications to the technical solution of the present invention, or modify it into an equivalent embodiment of 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 content of 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 rail damage holographic identification method, characterized in that: The following steps are involved: S1) Start rail damage identification mode, call B-type detection data, A-type detection data, rail surface image data and profile matching data, and display them in multiple windows on the same screen or multiple screens adjacent to each other; The B-type data display window is used to display the B-type detection data, the A-type data display window is used to display the A-type detection data, the video image data display window is used to display the rail surface image data, and the profile data display window is used to display the profile matching data; S2) determining whether the B-type data display window calls periodic historical B-type detection data, if yes, proceeding to step S3), if not, jumping to step S5); S3) The B-type data display window calls the periodic historical B-type detection data to display multiple windows on the same screen or multiple screens adjacently; S4) performing multi-cycle B-type detection data position alignment; S5) Multi-window display data for mileage matching and sequential playback; S6) performing damage identification based on the multi-window display data including the played B-type inspection data, A-type inspection data, rail surface image data, and profile matching data, and generating a damage report; In step S5), the A-type detection data, the profile matching data, and the B-type detection data with multi-cycle data comparison and the rail surface image data are displayed adjacently on the same screen or multiple screens; the A-type detection data, the B-type detection data, the rail surface image data and the profile matching data are position-calibrated, aligned and synchronously displayed using a synchronous pulse signal based on the mileage encoder count; Damage analysis and identification are performed based on the characteristics reflected by different signals, and a damage report is generated.

2. The rail damage holographic identification method according to claim 1, characterized in that: In step S1), the B-type detection data is used to determine whether the rail is damaged; the A-type detection data is used to reproduce the ultrasonic echo display state corresponding to each signal point in the B-type detection data to assist in identifying the authenticity of the rail damage; the rail surface image data is used to intuitively display the rail surface state, eliminating the interference of surface damage and surface state on the B-type detection data; The profile matching data is used to provide information on the wear and centering status of the inner side of the rail, assist in determining whether the inspection status is good, serve as a basis for determining the source of poor images in the B-type inspection data, and assist in damage identification.

3. The rail damage holographic identification method according to claim 1 or 2, characterized in that: In step S3), the periodic historical B-type detection data is the B-type detection data of the previous cycle or the most recent several cycles before that; when two or more B-type detection data files are called in the B-type data display window, the called B-type detection data files must be rail detection files of the same tested line; if two B-type detection data files are called, they are the current B-type detection data file and the B-type detection data file of the previous cycle; if multiple B-type detection data files are called, they are the current B-type detection data file and several B-type detection data files closest to the current detection time.

4. The rail damage holographic identification method according to claim 3, characterized in that: In the step S5), the data is synchronously played with the B-type data display window as the main display window; In step S6), when suspected damage or other suspicious signals appear in the B-type detection data, a comprehensive judgment is made by combining the periodic historical B-type detection data, A-type detection data, rail surface image data and profile matching data of the same position, and the judgment result is marked.

5. The rail damage holographic identification method according to claim 1, 2 or 4, characterized in that: In step S4), the display directions of two or more B-type detection data files are adjusted and kept consistent based on the increase or decrease of line mileage and left / right track information; the two or more B-type detection data files are first roughly aligned based on the line mileage information, and then accurately aligned based on the characteristic images displayed by the B-type detection data; In step S6), when the B-type graph corresponding to a certain position of the current B-type detection data shows suspected damage, by identifying the B-type graph signal characteristics of the periodic historical B-type detection data at the same position, if similar damage signal characteristics are also shown, identification is performed and a development trend table is generated.

6. The rail damage holographic identification method according to claim 5, characterized in that: The step S6) further comprises: For rail surface damage with continuous characteristics, combined with the characteristics of the B-type diagram including the rail head signal and the rail bottom loss wave signal, regional record identification is carried out according to the line mileage, and a damage report is automatically generated to focus on monitoring the lines in the area.

7. The rail damage holographic identification method according to claim 6, characterized in that: The step S6) further comprises: By combining the rail head position signal in the B-type diagram with the rail surface light band in the video image data display window, indirect feedback is given on line anomalies caused by rail misalignment, high and low joints, and slope, and the line mileage section is recorded and a damage report is automatically generated.

8. The rail damage holographic identification method according to claim 7, characterized in that: The step S6) further comprises: In the B-type data display window, when the B-display signal corresponding to the B-type detection data appears at a position including the rail head and a suspected weld position but cannot be clearly identified, the rail surface image at the same position in the video image data display window can be viewed to assist in judging rail damage including surface scratches, surface fish scales, peeling and falling pieces, and weld bars, and the damage report is automatically generated by integrating the judgment results of rail damage identification.

9. The rail damage holographic identification method according to claim 1, 2, 4, 6, 7 or 8, characterized in that: Said step S6) further comprises: When the data of a channel in the B-type detection data is incomplete or has few points, or the channel gain is not adjusted in time due to changes in speed level, the gain of the channel is amplified using the A-type detection data, and the B-type detection data is reconstructed. The position of the damage echo within the fixed gate and the relative relationship between the damage echo amplitude and the threshold are then used to assist in damage identification.

10. The rail damage holographic identification method according to claim 9, characterized in that: Said step S6) further comprises: When a suspected defect is found through the B-type test data and a judgment cannot be made, the signal corresponding to the A-type test data can be checked. The damage echo can be judged by whether it appears within the set gate range and whether the damage echo appears within the primary wave gate or the secondary wave gate to determine whether it is a reflection echo from other set positions that are not damage defects.

11. The rail damage holographic identification method according to claim 10, characterized in that: The step S6) further comprises: The position deviation of the damage echo in the A-type detection data is used to feedback the working status of the damage detection equipment, and the damage detection equipment is adjusted for the next cycle accordingly.

12. The rail damage holographic identification method according to claim 1, 2, 4, 6, 7, 8, 10 or 11, characterized in that: The step S6) further comprises: The wear status of the inner side of the rail is fed back through profile matching data, and the line sections with large wear are recorded and a damage report is automatically generated; the profile matching data of the rail is synchronously positioned through line mileage for section statistics, and a wear area threshold is set, and sections where the rail wear area exceeds the wear area threshold are monitored in detail.

13. The rail damage holographic identification method according to claim 12, characterized in that: The step S6) further comprises: Comprehensive identification is performed by combining profile matching data with rail surface image data and Type B inspection data. Line sections with poor inspection results are recorded and damage reports are automatically generated to serve as the basis for repeated inspections or key monitoring. At the same time, it can also indirectly feedback the working status of the inspection equipment, provide a reference for the correction of parameters of the inspection equipment, including centering reference and inclination, and provide original data reference for rail grinding.

14. The rail damage holographic identification method according to claim 1, 2, 4, 6, 7, 8, 10, 11 or 13, characterized in that: The ultrasonic detection unit generates type B and type A detection data based on point-to-point alignment counted by the mileage encoder; the video image detection unit generates rail surface image data based on the synchronous pulse signal counted by the mileage encoder; and the laser automatic centering unit generates profile matching data based on the synchronous pulse signal counted by the mileage encoder.

Images