A steel rail ultrasonic flaw detection practical training simulation method and system based on a damage sample library

Abstract

The application discloses a kind of based on rail ultrasonic flaw detection practical training simulation method and system of sample library of injury and loss, belong to railway rail nondestructive testing practical training technical field;The method comprises: constructing sample library of injury and loss, the sample library of injury and loss is stored with the structured injury and loss sample data including ultrasonic flaw detection waveform data, injury and loss type, injury and loss position, injury and loss quantification data;Obtain the real-time running displacement and change amplitude data in practical training process;According to the real-time running displacement data and the displacement information in the sample library of injury and loss are matched, real-time generates the ultrasonic flaw detection simulation waveform data corresponding to current running displacement state;The present application replaces entity test block practical training mode with rail flaw detection digital simulation practical training mode, provides comprehensive, quantifiable evaluation efficient practical training means, greatly improves the practical training range, effect, quality, efficiency and the skill level of practical training personnel guarantee line safety, while reducing the cost of practical training.

Description

Technical Field

[0001] This invention relates to the field of railway nondestructive testing training technology, and in particular to a simulation method and system for ultrasonic flaw detection training of rails based on a damage sample library. Background Technology

[0002] Rails are a critical component of railway tracks, and timely detection of internal damage and defects is essential for ensuring transportation safety. Ultrasonic testing, due to its high sensitivity and strong penetrating power, has become the mainstream method for non-destructive testing of rails. The effective application of this technology highly depends on the operator's ability to identify and interpret complex waveform characteristics; therefore, professional and efficient practical training is crucial.

[0003] Currently, the mainstream training method for ultrasonic flaw detection of rails relies on hands-on operation on real or simulated damaged rail test blocks. This method is essentially a physical-dependent and experience-transfer-based training model. While it can provide realistic physical feedback, it suffers from the following insurmountable technical bottlenecks: (1) Training samples are scarce and costly, making it difficult to cover all working conditions. Fabricating physical test blocks that cover various typical and atypical damage types, such as railhead cracks, rail web bolt hole cracks, and longitudinal cracks at the rail base, requires extremely high material, processing, and time costs. Some damage test blocks (such as large cracks) pose safety hazards and are not convenient to store and reuse. This severely limits the diversity, comprehensiveness, and safety of the training sample library, preventing trainees from accessing all necessary damage cases.

[0004] (2) The training process is difficult to standardize and scale up, and the evaluation system is subjective. Due to the limited number of physical test blocks, it is difficult to support multiple people to train at the same time and at high frequency, resulting in low training efficiency. At the same time, the inherent differences in damage parameters (position, angle, size) between different test blocks, as well as the subjectivity that is difficult to avoid in manual scoring, lead to inconsistent assessment conditions and difficulty in quantifying evaluation standards, making it impossible to objectively and consistently evaluate the skills of trainees.

[0005] (3) The training data is not reproducible and the process is not traceable, making it difficult to achieve refined teaching. The operation on the physical test block is one-time and irreversible. The system cannot fully record key behavioral data such as the probe movement trajectory, waveform observation time, and damage labeling logic during the flaw detection process. Therefore, instructors cannot accurately replay and analyze the root causes of the students' operation process, making it difficult to provide personalized and targeted guidance, resulting in slow improvement in training effectiveness.

[0006] In summary, existing training models based on physical test blocks have systemic limitations in terms of sample coverage, training efficiency, assessment standardization, and process data analysis. With the development of railway inspection technology and the expansion of talent training, the industry urgently needs a new training method for ultrasonic flaw detection of rails that can overcome these limitations and provide high-fidelity, high-flexibility, and quantifiable evaluation. Summary of the Invention

[0007] The purpose of this invention is to provide a simulation method and system for ultrasonic flaw detection of rails based on a damage sample library in order to solve the above problems.

[0008] The present invention achieves the above objectives through the following technical solutions: This invention provides a simulation training method for ultrasonic flaw detection of rails based on a damage sample library, including: A damage sample library is constructed, which stores structured damage sample data including ultrasonic flaw detection waveform data, damage type, damage location, and damage quantification data. Acquire real-time data on the displacement and change amplitude of the flaw detector during the training process; The ultrasonic flaw detection simulation waveform data corresponding to the current walking displacement state is generated in real time by matching the real-time walking displacement data with the displacement information in the damage sample library.

[0009] Preferably, real-time generation of simulation waveform data specifically includes: Based on real-time travel displacement data, the corresponding multi-channel A-waveform data is extracted from the damage sample library.

[0010] Preferably, when the real-time walking displacement value is located between adjacent displacement points in the sample library, interpolation is performed based on the waveform data of the adjacent displacement points to generate transition waveform data.

[0011] Preferably, it also includes: Based on the echo signals exceeding a set threshold in the simulated waveform data, B-display image data is generated; The B-display image data is subjected to coordinate mapping and dynamic rendering, and then displayed on the display interface.

[0012] Preferably, a reference line corresponding to the geometric parameters of the rail is overlaid and drawn in the dynamically rendered B-display image.

[0013] Preferably, the structured sample data in the injury sample library also includes accurate injury attribute description information, i.e., standard answer information; the method further includes: Record user operational behavior data during the training process; The operational behavior data is compared with the standard answer information, and the evaluation result is output.

[0014] Preferably, the visibility of the standard answer information is controlled according to different training modes, including teaching mode, self-study mode, and examination mode.

[0015] Preferably, the sample data in the damage sample library is organized using a hierarchical classification structure, including at least a first-level classification based on rail location and a second-level classification based on defect type.

[0016] This invention also provides a training simulation system for ultrasonic flaw detection of rails, comprising: The memory is used to store computer programs and a damage sample library, wherein the damage sample library stores structured sample data containing rail damage waveform data and corresponding displacement information. The processor is used to execute the computer program and run the damage sample library to generate ultrasonic flaw detection simulation waveform data in real time based on the mapping relationship between real-time walking displacement data and displacement information in the damage sample library, and to realize training interaction and evaluation based on the simulation waveform data.

[0017] The communication interface is used to receive the real-time displacement data of the flaw detector from the displacement sensor on the running support structure.

[0018] The present invention also provides a computer-readable storage medium having a computer program stored thereon that, when executed by a processor, implements the method described above.

[0019] Compared with the prior art, the beneficial effects of the present invention are as follows: 1. Through simulation using a structured sample library built on real damage data, the generated A-waveform and B-image are highly consistent with the real equipment in terms of signal characteristics and dynamic response, providing trainees with an immersive training environment and ensuring effective experience transfer from virtual training to on-site practice; 2. The damage sample library can systematically integrate various typical and atypical defects and can flexibly combine them to generate complex training cases, completely overcoming the limitations of physical test blocks in terms of type, quantity, and safety, and realizing the infinite expansion and on-demand configuration of training scenarios; 3. Replacing reliance on physical rail test blocks, track resources, and a large amount of manpower with computer simulation avoids the safety hazards and time and space limitations of on-site training, supports large-scale training and assessment with multiple terminals, high concurrency, and repeatability, and greatly reduces the overall cost. 4. By automating the comparison between user actions and preset standard answers, the system achieves objectivity, standardization, and immediacy in assessment and scoring. At the same time, the system fully records the data of the operation process, providing a data foundation for teaching review and personalized skills diagnosis, thus contributing to precision teaching. 5. The adoption of a unified and open structured data architecture facilitates the continuous accumulation and sharing of damage samples, ensuring the advanced nature and vitality of the simulation content; the standardized data system also lays the foundation for the subsequent integration of intelligent diagnostic algorithms and the realization of automated evaluation and decision support for flaw detection skills. Attached Figure Description

[0020] To more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0021] Figure 1 This is a flowchart of the overall process for simulating training data on ultrasonic flaw detection of rails.

[0022] Figure 2 This is a schematic diagram of the flaw detector terminal structure.

[0023] Figure 3 This is a structural diagram of the rail damage sample library.

[0024] Explanation of reference numerals in the attached figures: 1. Traveling support structure; 2. Displacement sensor; 3. Limiter; 4. Flaw detector main unit. Detailed Implementation

[0025] To enable those skilled in the art to better understand the technical solution of the present invention, the present invention will be further described below with reference to the accompanying drawings and embodiments. It should be understood that the embodiments described herein are only for explaining the present invention and are not intended to limit the scope of protection of the present invention. like Figure 1 As shown, a simulation training method for ultrasonic flaw detection of rails based on a damage sample library includes the following steps: Step S1: Construct a damage sample library: Store structured sample data of several rail damage defects in the damage sample library; Step S2: Real-time displacement acquisition of the flaw detector during training: Acquire the real-time movement displacement and change amplitude data of the flaw detector during the training process; Step S3: Real-time generation of simulation data: Extract or generate corresponding multi-channel A-mode waveform data from the damage sample library based on real-time displacement data; Step S4: Generate A-waveform data sequence: Generate A-waveform data sequence for multiple ultrasonic flaw detection channels based on the extracted or interpolated data; Step S5: Dynamic generation and rendering of B display image: Based on the out-of-standard echo signal in the A display waveform, generate B display image data, and perform coordinate mapping, reference line overlay and dynamic scrolling rendering, and display it on the interface; Step S6: Practical Training Interaction and Evaluation: Present a simulation scenario according to the selected mode (teaching / self-study / exam), record all user operations, compare them with the standard answers, and generate feedback or scores.

[0026] Steps S1-S6 form a complete closed loop for training data simulation, achieving full-process coverage from data preparation to training application. For example... Figure 3 As shown, the damage sample library adopts a multi-layered structured design, which is the data foundation for achieving high-fidelity simulation in this invention. Specifically, it includes: Sample identification information layer: unique number, damage name, category (such as rail head, rail web, rail bottom) and specific type; Rail geometric parameters layer: Rail model and corresponding key dimensions such as rail height, rail head height, and bolt hole height, used for B-view image coordinate mapping and reference line drawing; Metadata layer: Records the operating conditions during raw data acquisition, such as equipment model, probe configuration, detection time, vehicle speed, etc. Core waveform data layer: Contains displacement encoding sequences and multi-channel ultrasound A-waveform (amplitude-depth sequence) corresponding to each displacement point; this layer is the data source for real-time generation of simulation waveforms; Equipment configuration parameter layer: Stores parameters such as gain, gate, alarm threshold, and probe angle required to simulate real equipment, ensuring that the simulation signal is consistent with the logic of the real equipment; Damage feature parameter layer: Quantitatively describes the geometric and signal characteristics of defects, such as depth, length, and angle; Standard answer information layer: Clearly defines the type of injury, the precise location coordinates, and the judgment criteria for automated scoring.

[0027] The damage sample database adopts a hierarchical classification management system. The first level is divided according to rail parts (rail head, rail web, rail base); the second level is subdivided according to defect type (such as cracks, core damage, etc.); the third level can be further precisely located according to spatial position (such as distinguishing the left and right sides and specific bolt hole phases for bolt hole cracks). When adding new samples, they can be entered according to predefined structured formats (such as JSON) to achieve smooth and standardized expansion of the sample database.

[0028] This invention also discloses a training simulation system for ultrasonic flaw detection of rails, comprising: The memory is used to store computer programs and a damage sample library, wherein the damage sample library stores structured sample data containing rail damage waveform data and corresponding displacement information. The processor is used to execute the computer program and run the damage sample library to generate ultrasonic flaw detection simulation waveform data in real time according to the mapping relationship between real-time walking displacement data and displacement information in the damage sample library, and to realize training interaction and evaluation based on the simulation waveform data. The communication interface is used to receive the real-time displacement data of the flaw detector from the displacement sensor on the running support structure.

[0029] In some embodiments, the damage sample files are stored in a structured data format (e.g., JSON), which records information such as damage type identifier, classification field, key parameters and data volume index, so that the system can quickly read and call them. When a new damage type is added, information can be supplemented according to the preset field specifications to achieve smooth expansion of the sample library.

[0030] The exam management module is used to create exams, maintain test questions, assign candidates, and publish exams, and also provides proctoring functionality. The "My Exams" module allows candidates to access exams, submit answers, and check scores; the score query module allows administrators to view historical exam data and trace answer details; and the individual ability assessment module is used to statistically analyze and visualize the abilities of trainees, such as skill rankings and ability radar charts.

[0031] The linkage simulation mechanism of this system is as follows: When the trainee pushes the training flaw detector body to make slight movements on the traveling support structure, the displacement sensor outputs the displacement status signal in real time; the flaw detector host 4 sends the displacement status signal and the current training / examination mode information to the background server; the background server calls the corresponding sample data from the damage sample database according to the displacement status and task configuration, or combines, slices or interpolates the sample data according to the preset generation rules to form an ultrasonic data sequence consistent with the traveling process; then the ultrasonic A-display and B-display data are sent to the flaw detector host 4, which dynamically renders the data on the display interface to realize the synchronous simulation of "operation-data feedback".

[0032] In some embodiments, the system can set trigger conditions for flaw detection actions, such as marking damage actions, mode switching actions, or key button actions; when the trigger conditions are met, the system records the displacement state at the trigger moment and displays waveform segments for training playback, error localization, or exam scoring.

[0033] In examination scenarios, the support system can randomly select test papers from the sample library or customize them according to the specified damage type based on the test question configuration, and send the examination task to the flaw detector terminal; after the candidates complete the interpretation and answer within the specified time, they submit the results; the system automatically scores and archives the test questions and answers based on the interpretation rules, and can also generate performance statistics and ability profiles for targeted improvement training.

[0034] In terms of data security, the system can provide hierarchical permission management and a unified authorization mechanism, perform regular automatic or manual backups of important data, and support off-site storage; in the communication and storage links, encryption and integrity verification mechanisms can be adopted to reduce the risk of data leakage and tampering; in terms of network access, security can be improved through local area network isolation, terminal access control and port restrictions.

[0035] In terms of performance, this system can optimize the communication and data loading mechanism to ensure that the synchronous communication latency between the flaw detector and the platform meets the training requirements and supports concurrent examination scenarios for multiple terminals. The sample coverage, data loading speed and continuous operation capability of the system can be parameterized and expanded according to actual deployment needs.

[0036] like Figure 2 As shown, the present invention also discloses a flaw detector terminal structure applicable to the above method, including a traveling support structure 1, a displacement sensor 2, a limiter 3, a flaw detector host 4, and a training auxiliary terminal.

[0037] In some embodiments, the traveling support structure 1 is used to simulate the traveling conditions of a rail on site. Preferably, a rail-like metal structure is used to provide a near-realistic contact surface and micro-dampening characteristics, giving the operator a realistic feel.

[0038] In some embodiments, the displacement sensor 2 is installed on the traveling support structure 1 to detect in real time the displacement, direction, speed and stopping status of the training flaw detector body along the track direction, and outputs the displacement status signal.

[0039] In some embodiments, the limiter 3 is used to limit the maximum displacement range of the flaw detector body to ensure equipment safety and signal validity.

[0040] In some embodiments, the system supports both online and offline operating modes: Offline mode: The flaw detector host communicates with the backend server and is responsible for acquiring, displacing, answering questions, jumping, and submitting signals. According to the training mode, it retrieves or generates corresponding ultrasonic simulation data from the damage sample library. In this mode, the flaw detector communicates with the backend server using protocols such as socket, http, and https, and also includes local communication between the flaw detector sensor and its own host.

[0041] Online mode: Runs on a computer, involving only communication between the front-end interactive interface and the back-end server. The front-end display interface renders waveforms (A) and images (B) in real time.

[0042] In some embodiments, the training auxiliary terminal (such as a PAD) is used to implement auxiliary interactive functions such as user authentication, training / exam task management, and proctoring. A biometric module (such as facial recognition) can be integrated to enhance the control of the examination process.

[0043] The foregoing has shown and described the basic principles, main features and advantages of the present invention. Those skilled in the art should understand that the present invention is not limited to the above embodiments. The embodiments and descriptions in the specification are only illustrative of the principles of the present invention. Various changes and modifications can be made to the present invention without departing from the spirit and scope of the present invention, and all such changes and modifications fall within the scope of the present invention as claimed.

Claims

1. A simulation training method for ultrasonic flaw detection of rails based on a damage sample library, characterized by: A damage sample library is constructed, which stores structured damage sample data including ultrasonic flaw detection waveform data, damage type, damage location, and damage quantification data. Acquire real-time data on the displacement and change amplitude of the flaw detector during the training process; The ultrasonic flaw detection simulation waveform data corresponding to the current walking displacement state is generated in real time by matching the real-time walking displacement data with the displacement information in the damage sample library.

2. The method according to claim 1, characterized in that, Real-time generation of simulation waveform data specifically includes: Based on real-time travel displacement data, the corresponding multi-channel A-waveform data is extracted from the damage sample library.

3. The method according to claim 2, characterized in that, When the real-time travel displacement value is located between adjacent displacement points in the sample library, interpolation is performed based on the waveform data of the adjacent displacement points to generate transition waveform data.

4. The method according to claim 1, characterized in that, Also includes: Based on the echo signals exceeding a set threshold in the simulated waveform data, B-display image data is generated; The B-display image data is subjected to coordinate mapping and dynamic rendering, and then displayed on the display interface.

5. The method according to claim 4, characterized in that, In the dynamically rendered B-screen image, reference lines corresponding to the geometric parameters of the rail are overlaid and drawn.

6. The method according to claim 1, characterized in that, The structured sample data in the injury sample database also includes accurate injury attribute description information, i.e., standard answer information; the method further includes: Record user operational behavior data during the training process; The operational behavior data is compared with the standard answer information, and the evaluation result is output.

7. The method according to claim 6, characterized in that, The visibility of the standard answer information is controlled according to different training modes, including teaching mode, self-study mode, and examination mode.

8. The method according to claim 1, characterized in that, The sample data in the damage sample library is organized using a hierarchical classification structure, including at least a first-level classification based on rail location and a second-level classification based on defect type.

9. A simulation training system for ultrasonic flaw detection of rails, characterized in that, include: The memory is used to store computer programs and a damage sample library, wherein the damage sample library stores structured sample data containing rail damage waveform data and corresponding displacement information. The processor is used to execute the computer program and run the damage sample library to generate ultrasonic flaw detection simulation waveform data in real time according to the mapping relationship between real-time walking displacement data and displacement information in the damage sample library, and to realize training interaction and evaluation based on the simulation waveform data. The communication interface is used to receive the real-time displacement data of the flaw detector from the displacement sensor on the running support structure.

10. A computer-readable storage medium having a computer program stored thereon, characterized in that, When the program is executed by the processor, it implements the method as described in any one of claims 1 to 8.

Images