Railway switch ultrasonic guided wave nondestructive testing method and testing device

Abstract

The application discloses a railway point ultrasonic guided wave nondestructive detection method and a detection device, relates to the technical field of railway point nondestructive detection, and compares and judges the obtained real-time difference signals by using an abnormality analysis and judgment module; if there is no abnormality, the detection data is stored in a detection data database; if there is abnormality, the judgment result is sent to a prewarning reminding module, and meanwhile, the detection data is sent to an abnormality feedback verification model; the difference signals are predicted by using the abnormality feedback verification model, the predicted difference signals are obtained, the predicted difference signals are compared and analyzed with the real-time difference signals, and if the comparison is consistent, the comparison result is sent to the prewarning reminding module; when receiving the instructions of the abnormality analysis and judgment module and the difference signal analysis module, the prewarning reminding module performs point abnormality prewarning, the analysis result is verified by using the abnormality feedback verification model, and the influence of environmental factors on data acquisition is avoided to prevent misjudgment.

Description

Technical Field

[0001] This invention relates to the field of non-destructive testing technology for railway switch rails, specifically to an ultrasonic guided wave non-destructive testing method and device for railway switch rails. Background Technology

[0002] The railway switch rail is a core component of railway turnouts, responsible for guiding the train wheelset from one track to another. Its condition directly affects the safety of train operation and the lifespan of the track system. Therefore, structural health inspection of railway switch rails is particularly important.

[0003] In the prior art, Chinese invention patent (CN118731164A) discloses a "turnout switch rail crack detection system, method and device", which specifically discloses "using ultrasonic guided wave detection technology, by setting a temperature sensor to monitor real-time temperature data and setting a temperature range, comparing the real-time differential signal with the reference differential signal of different temperature ranges according to different real-time temperature data, judging whether the real-time differential signal is abnormal, and realizing non-destructive detection of internal defects of railway switch rails";

[0004] However, when the above-mentioned disclosed technology uses ultrasonic guided wave testing technology to perform non-destructive testing on railway switch rails, although it can reduce the influence of other factors on the judgment results by using real-time differential signals as the standard for judging internal defects in the switch rails, there will still be misjudgments of defects in the railway system due to the large number of influencing factors and the inaccuracy of judging by temperature range. In severe cases, this will affect the normal operation of the railway system.

[0005] Therefore, there is an urgent need for an intelligent, comprehensive, and automated method and device for ultrasonic guided wave nondestructive testing of railway switch rails to solve the above-mentioned technical problems. Summary of the Invention

[0006] The purpose of this invention is to provide a method and device for non-destructive testing of railway switch rails using ultrasonic guided waves, in order to solve the problems raised in the prior art.

[0007] To achieve the above objectives, the present invention provides the following technical solution: a non-destructive testing method for railway switch rails using ultrasonic guided waves, which uses a real-time data detection module to collect various information data of the switch rails and uses a number of symmetrically arranged ultrasonic guided wave transducer groups to collect real-time signals;

[0008] The differential signal calculation module is used to calculate the real-time signals collected by several ultrasonic guided wave transducer groups to obtain several real-time differential signals.

[0009] The anomaly analysis and judgment module is used to compare and judge the obtained real-time differential signal to determine whether there is an anomaly. If there is no anomaly, the information data collected by the real-time data detection module and the real-time signal collected by the ultrasonic guided wave transducer group are packaged and stored in the detection data database.

[0010] The historical detection data in the detection data database is retrieved by the historical data retrieval unit and sent to the differential signal analysis module;

[0011] The differential signal analysis module trains the model based on the retrieved historical detection data to obtain an anomaly feedback verification model;

[0012] If the anomaly analysis and judgment module determines that there is an anomaly in the real-time differential signal, it sends the judgment result to the early warning and reminder module. At the same time, it sends the various information data collected in real-time detection and the real-time differential signal to the anomaly feedback verification model. The anomaly feedback verification model uses the various information data as input features to make predictions and obtains the predicted differential signal. The predicted differential signal is compared and analyzed with the real-time differential signal. If the comparison is consistent, the comparison result is sent to the early warning and reminder module.

[0013] When the early warning module receives instructions from the anomaly analysis and judgment module and the differential signal analysis module, it issues an early warning for anomalies detected in the switch rail.

[0014] According to the above technical solution, the differential signal analysis module includes a historical data processing unit and a verification model training unit. The historical data processing unit is used to process the historical detection data retrieved by the historical data retrieval unit to obtain a training dataset and a test dataset.

[0015] The data in the training and test datasets are divided into input and output data based on input and output features, and the input and output data are preprocessed to define a loss function L.

[0016] The verification model training unit trains the verification model using the training dataset, incorporates a defined loss function L, and tests the trained verification model using the test dataset to ultimately obtain the anomaly feedback verification model.

[0017] According to the above technical solution, the paired real-time signals are used to calculate the real-time differential signal V through the differential signal calculation module. diff (t i ), where i represents the i-th group of ultrasonic guided wave transducers, and the set of real-time differential signals obtained is V = {V diff (t1),V diff (t2),V diff (t3)…,V diff (t n)}, where n represents a total of n sets of ultrasonic guided wave transducers, generating n real-time differential signals;

[0018] The real-time differential signal threshold V is set through the anomaly analysis and judgment module. 阈 In the set of real-time differential signals, if there exists a real-time differential signal V diff (t i () Greater than or equal to the real-time differential signal threshold V 阈 If an anomaly is detected in the switch rail, the anomaly result is sent to the early warning module, and the real-time differential signal V indicating the anomaly is also sent. diff (t i The corresponding information data is sent to the differential signal analysis module.

[0019] According to the above technical solution, the corresponding information data are used as input features to the anomaly feedback verification model. The anomaly feedback verification model is then used for prediction to obtain the predicted differential signal V. pred (t i ), where i represents the predicted differential signal corresponding to the i-th group of ultrasonic guided wave transducers;

[0020] The differential signal analysis module further includes a difference calculation unit and a result judgment unit, used to analyze the real-time differential signal V. diff (t i ) and the predicted differential signal V pred (t i The difference ΔV(t) between the differential signals i The calculation is performed; the result judgment unit is used to determine the difference value ΔV(t) of the differential signal. i The result is compared with the set threshold. If it does not exceed the set threshold, the result is sent to the early warning module. If it exceeds the set threshold, the prediction result of the abnormal feedback verification model is verified again.

[0021] Based on the above technical solution, the prediction results of the anomaly feedback verification model are re-verified, including the following steps:

[0022] S101. In the set V of real-time differential signals, for the real-time differential signal V... diff (t i () less than the real-time differential signal threshold V 阈 The data packet, which includes various information data, is retrieved, and the real-time differential signal V is retrieved. diff (t i (Redefined as real-time verification differential signal V′) diff (t i );

[0023] S102. Use the information data from the data packet retrieved in step S101 as input features to the anomaly feedback verification model, and use the anomaly feedback verification model to output the prediction verification differential signal V′. pred (t i );

[0024] S103, Predict and verify the differential signal V′ pred (t i The real-time verification differential signal V′ in the retrieved data packet diff (t i Compare them;

[0025] S104. Judge the comparison results of step S103. If they are consistent, it indicates that the prediction result of the abnormal feedback verification model is normal. The judgment result of the result judgment unit is not sent to the early warning module. Since the early warning module has not received the instruction from the differential signal analysis module, it will not issue an early warning. If they are inconsistent, it indicates that the prediction result of the abnormal feedback verification model is abnormal and the prediction result of the abnormal feedback verification model is unreliable. The differential signal analysis module sends an instruction to the early warning module. The early warning module receives the instructions from the differential signal analysis module and the abnormal analysis judgment module and issues an early warning.

[0026] According to the above technical solution, the information data includes at least electromagnetic noise data, mechanical vibration data, and real-time temperature data of the switch rail.

[0027] According to the above technical solution, the ultrasonic guided wave transducer adopts a piezoelectric transducer.

[0028] A detection device, the detection device comprising a tensioning bracket, a probe moving seat, a probe, and an analysis system;

[0029] The tensioning brackets are symmetrically installed on both sides of a switch rail. A probe moving seat is provided on the tensioning bracket, and a probe is installed on the probe moving seat. The probe moving seat is used to adjust the position of the probe. An ultrasonic guided wave transducer is integrated inside the probe.

[0030] The detection device also includes an excitation module connected to an ultrasonic guided wave transducer for generating ultrasonic guided wave signals;

[0031] The analysis system is used to perform intelligent analysis on the information data collected by the detection module.

[0032] Compared with the prior art, the beneficial effects of the present invention are:

[0033] This invention collects various information data of railway switch rails through a real-time data detection module, acquires real-time signals using an ultrasonic guided wave transducer array, trains a model using historical detection data to obtain an anomaly feedback verification model, and after analyzing the real-time signals of the ultrasonic guided wave transducer array, selectively uses the anomaly feedback verification model to verify the analysis results based on the analysis results. This avoids the impact of environmental factors on the data acquisition of the ultrasonic guided wave transducer array, which could lead to misjudgments, thereby ensuring the accuracy of non-destructive testing of railway switch rails.

[0034] Meanwhile, when the analysis results are inconsistent with the verification results of the abnormal feedback verification model, the abnormal feedback verification model will also be diagnosed to determine and rule out whether the inconsistency in the verification results is caused by anomalies in the abnormal feedback verification model. This will further ensure the accuracy of non-destructive testing of railway switch rails, avoid misdiagnosis of switch rail testing, and ensure the accuracy of the testing and diagnosis results. Attached Figure Description

[0035] Figure 1 This is a schematic diagram of the logic flow of an ultrasonic guided wave nondestructive testing method for railway switch rails according to the present invention;

[0036] Figure 2 This is a schematic diagram illustrating the logical flow of an ultrasonic guided wave nondestructive testing method for railway switch rails according to the present invention.

[0037] Figure 3 This is a schematic diagram of the ultrasonic guided wave transducer in the ultrasonic guided wave non-destructive testing device for railway switch rails according to the present invention;

[0038] Figure 4 This is a schematic diagram of the parameter specifications of the ultrasonic guided wave transducer in the ultrasonic guided wave non-destructive testing device for railway switch rails according to the present invention.

[0039] Figure 5 This is a schematic diagram of the installation of the ultrasonic guided wave transducer in the ultrasonic guided wave non-destructive testing device for railway switch rails according to the present invention. Detailed Implementation

[0040] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0041] Example 1: As Figures 1-2As shown, this invention provides a technical solution for a non-destructive testing method for railway switch rails using ultrasonic guided waves. The method utilizes a real-time data detection module to collect various information data of the switch rail. For example, it uses a near-field probe and a spectrum analyzer to collect electromagnetic noise data, a vibration sensor installed on the switch rail to collect mechanical vibration data, and a temperature sensor installed on the switch rail to collect real-time temperature data. A plurality of symmetrically arranged ultrasonic guided wave transducer groups are used to collect real-time signals. Each ultrasonic guided wave transducer group includes a first ultrasonic guided wave transducer and a second ultrasonic guided wave transducer, symmetrically installed on both sides of a switch rail. The first and second ultrasonic guided wave transducers in each group respectively receive ultrasonic guided wave excitation signals generated by an excitation module, and after receiving the ultrasonic guided wave excitation signals, they send and receive real-time signals to each other to obtain a first real-time signal and a second real-time signal.

[0042] like Figures 3-4 As shown, the ultrasonic guided wave transducer is a piezoelectric transducer;

[0043] In this embodiment, the real-time data detection module and the ultrasonic guided wave transducer group are encoded and bundled. Information data collected at a specific timestamp is bundled with a set of first and second real-time signals to form a data packet, and this data packet is defined as S. i-T Where T represents the timestamp and i represents the i-th bundled data;

[0044] The first real-time signal is The second real-time signal is The first real-time signal acquired by several ultrasonic guided wave transducer groups using the differential signal calculation module Second real-time signal The difference is calculated to obtain several real-time differential signals V. diff (t i ), where i represents the real-time differential signal acquired by the i-th group of ultrasonic guided wave transducers, and the set of real-time differential signals is V = {V diff (t1),V diff (t2),V diff (t3)…,V diff (t n )}, where n represents a total of n sets of ultrasonic guided wave transducers, generating n real-time differential signals;

[0045] The real-time differential signal threshold V is set through the threshold setting unit. 阈 The differential signal comparison unit is used to compare and analyze the real-time differential signal with the real-time differential signal threshold. If a real-time differential signal V exists in the set of real-time differential signals... diff (t i() Greater than or equal to the real-time differential signal threshold V 阈 If an anomaly is detected, the switch rail is determined to be faulty. The anomaly result is sent to the early warning module, and an anomaly feedback verification command is generated and sent to the encoded data packet. The real-time differential signal V indicating the anomaly is then sent. diff (t i ) and the corresponding bundled encoded data packet S i-T Send to the differential signal analysis module;

[0046] In the above embodiment, the data detection results of the switch rail indicate that there is an anomaly in the switch rail. Normally, when an anomaly is detected in the switch rail, a warning is issued directly. However, in the above embodiment, when an anomaly is detected in the switch rail, the detection result is sent to the warning module. However, the warning module does not issue a warning directly. Instead, it sends the abnormal detection data to the differential signal analysis module. The differential signal analysis module verifies the detection data. When the verification result shows that the detection data is indeed abnormal, the verification result is sent to the warning module. The warning module only issues a warning when it receives two instructions, thus avoiding the possibility that the detection data may have large errors due to external environmental factors, affecting the judgment result.

[0047] Example 2: Based on Example 1, if there is no real-time differential signal V in the set of real-time differential signals... diff (t i () Greater than or equal to the real-time differential signal threshold V 阈 Then, a data packaging and storage instruction is generated, which combines the various information data collected by the real-time data detection module, the real-time signals collected by the ultrasonic guided wave transducer group, and the real-time differential signal V. diff (t i The data is packaged and stored in the detection data database;

[0048] When there are no anomalies in the detection data, the detection data is directly stored in the detection data database. This allows for the continuous expansion of the amount of real data in the detection data database, thereby ensuring that the output results of the anomaly feedback verification model are more realistic and reliable when the detection data in the detection data database is used to train and optimize the model.

[0049] In this embodiment, when training the anomaly feedback verification model, historical detection data in the detection data database is retrieved through the historical data retrieval unit and sent to the differential signal analysis module.

[0050] The differential signal analysis module trains the model based on the retrieved historical detection data;

[0051] The differential signal analysis module includes a historical data processing unit and a verification model training unit. The historical data processing unit is used to process the historical detection data retrieved by the historical data retrieval unit to obtain a training dataset and a test dataset.

[0052] The data in the training and test datasets are divided into input and output data based on input and output features, and the input and output data are preprocessed to define a loss function L.

[0053] Specifically:

[0054] Input data includes: electromagnetic noise data, mechanical vibration data, real-time temperature data, etc.

[0055] The output data is: the predicted differential signal;

[0056] The input and output data are normalized to unify the units of measurement, which facilitates the subsequent model training. As for the specific normalization method, it is a conventional data processing technique and will not be elaborated on here.

[0057] Define loss function

[0058] Where N represents the total number of samples in the input data. Let represent the loss function for the k-th sample;

[0059] The verification model training unit trains the verification model using the training dataset, incorporates a defined loss function L, and tests the trained verification model using the test dataset to ultimately obtain the anomaly feedback verification model.

[0060] If the anomaly analysis and judgment module determines that there is an anomaly in the real-time differential signal, it sends the judgment result to the early warning module. Simultaneously, based on the generated anomaly feedback verification instruction, it sends all the information data collected from the real-time detection data acquisition, along with the first real-time signal, the second real-time signal, and the real-time differential signal, to the anomaly feedback verification model. The corresponding information data is used as input features into the anomaly feedback verification model, which then performs prediction to obtain the predicted differential signal V. pred (t i ), where i represents the predicted differential signal corresponding to the i-th group of ultrasonic guided wave transducers;

[0061] The differential signal analysis module further includes a difference calculation unit and a result judgment unit, used to analyze the real-time differential signal V. diff (t i ) and the predicted differential signal V pred (t i The difference ΔV(t) between the differential signals iThe calculation is performed; the result judgment unit is used to determine the difference value ΔV(t) of the differential signal. i The result is compared with a set threshold. If the result does not exceed the set threshold, the result is sent to the early warning module. When the early warning module receives instructions from the anomaly analysis and judgment module and the differential signal analysis module, it issues an early warning for the switch rail detection anomaly.

[0062] Example 3: Based on Example 2, if the set threshold is exceeded, the prediction results of the anomaly feedback verification model are verified again.

[0063] Specifically, re-validating the prediction results of the anomaly feedback validation model includes the following steps:

[0064] S101. In the set V of real-time differential signals, for the real-time differential signal V... diff (t i () less than the real-time differential signal threshold V 阈 The data packet, which includes various information data, is retrieved, and the real-time differential signal V is retrieved. diff (t i (Redefined as real-time verification differential signal V′) diff (t i );

[0065] In step S101, the retrieval of data packets must be performed using the real-time differential signal V. diff (t i () less than the real-time differential signal threshold V 阈 The data packets are because these types of data packets indicate that the detection data for the switch rail is normal and there is no abnormality in the detection data;

[0066] S102. Use the information data from the data packet retrieved in step S101 as input features to the anomaly feedback verification model, and use the anomaly feedback verification model to output the prediction verification differential signal V′. pred (t i );

[0067] S103, Predict and verify the differential signal V′ pred (t i The real-time verification differential signal V′ in the retrieved data packet diff (t i Compare them;

[0068] S104. Judge the comparison results of step S103. If they are consistent, it indicates that the prediction results of the abnormal feedback verification model are normal. Do not send the judgment results of the result judgment unit to the early warning module. If the early warning module does not receive the instruction from the differential signal analysis module, it will not issue an early warning.

[0069] Because the data packets retrieved are those with normal detection data, using these normal data packets to verify the anomaly feedback verification model ensures real-time verification of the differential signal V′. diff (t i If the data is normal, then the abnormal feedback verification model output prediction verification difference signal V′ is correct. pred (t i ) and real-time verification differential signal V′ diff (t i If the values ​​are inconsistent, then the output of the abnormal feedback verification model must be the prediction verification difference signal V′. pred (t i An anomaly exists, meaning the anomaly feedback verification model has an anomaly;

[0070] If they are inconsistent, it indicates that the prediction result of the anomaly feedback verification model is abnormal and unreliable. This indirectly reflects that the first and second real-time signals collected by the ultrasonic guided wave transducer group are correct. In this case, if the anomaly analysis and judgment module analyzes that there is an anomaly in the real-time differential signal, then the real-time differential signal is indeed abnormal. The differential signal analysis module sends an instruction to the early warning and reminder module, and the early warning and reminder module receives the instructions from the differential signal analysis module and the anomaly analysis and judgment module and issues an early warning and reminder.

[0071] At the same time, early warnings are issued for abnormal situations in the abnormal feedback verification model, and the abnormal feedback verification model is strengthened through training to improve its verification accuracy.

[0072] Example 4: Figure 5 As shown, the detection device includes a tensioning bracket, a probe moving seat, a probe, and an analysis system;

[0073] The tensioning brackets are symmetrically installed on both sides of a switch rail. A probe moving seat is provided on the tensioning bracket, and a probe is installed on the probe moving seat. The probe moving seat is used to adjust the position of the probe, thereby ensuring that the detection device can perform more comprehensive data detection on the switch rail. An ultrasonic guided wave transducer is integrated inside the probe.

[0074] The detection device also includes an excitation module connected to an ultrasonic guided wave transducer for generating ultrasonic guided wave signals;

[0075] The analysis system is used to perform intelligent analysis on the information data collected by the detection module.

[0076] It will be apparent to those skilled in the art that the present invention is not limited to the details of the exemplary embodiments described above, and that the invention can be implemented in other specific forms without departing from its spirit or essential characteristics. Therefore, the embodiments should be considered in all respects as exemplary and non-limiting, and the scope of the invention is defined by the appended claims rather than the foregoing description. Thus, all variations falling within the meaning and scope of equivalents of the claims are intended to be included within the present invention. No reference numerals in the claims should be construed as limiting the scope of the claims.

Claims

1. A method for non-destructive testing of railway switch rails using ultrasonic guided waves, characterized in that: The real-time data detection module is used to collect various information data of the switch rail, and several symmetrically arranged ultrasonic guided wave transducer groups are used to collect real-time signals. The differential signal calculation module is used to calculate the real-time signals collected by several ultrasonic guided wave transducer groups to obtain several real-time differential signals. The anomaly analysis and judgment module is used to compare and judge the obtained real-time differential signal to determine whether there is an anomaly. If there is no anomaly, the information data collected by the real-time data detection module and the real-time signal collected by the ultrasonic guided wave transducer group are packaged and stored in the detection data database. The historical detection data in the detection data database is retrieved by the historical data retrieval unit and sent to the differential signal analysis module; The differential signal analysis module trains the model based on the retrieved historical detection data to obtain an anomaly feedback verification model; If the anomaly analysis and judgment module determines that there is an anomaly in the real-time differential signal, it sends the judgment result to the early warning and reminder module. At the same time, it sends the various information data collected in real-time detection and the real-time differential signal to the anomaly feedback verification model. The anomaly feedback verification model uses the various information data as input features to make predictions and obtains the predicted differential signal. The predicted differential signal is compared and analyzed with the real-time differential signal. If the comparison is consistent, the comparison result is sent to the early warning and reminder module. When the early warning module receives instructions from the anomaly analysis and judgment module and the differential signal analysis module, it issues an early warning for anomalies in the switch rail detection. The differential signal analysis module includes a historical data processing unit and a verification model training unit. The historical data processing unit is used to process the historical detection data retrieved by the historical data retrieval unit to obtain a training dataset and a test dataset. The data in the training and test datasets are divided into input and output data based on input and output features, and the input and output data are preprocessed to define a loss function L. The verification model training unit trains the verification model using the training dataset, incorporates a defined loss function L, and tests the trained verification model using the test dataset to ultimately obtain the anomaly feedback verification model.

2. The method for non-destructive testing of railway switch rails using ultrasonic guided waves according to claim 1, characterized in that: The real-time differential signal is obtained by calculating the real-time differential signal through the differential signal calculation module using the paired real-time signals. Where i represents the i-th group of ultrasonic guided wave transducers, and the set of real-time differential signals obtained. Where n represents a total of n sets of ultrasonic guided wave transducers, generating n real-time differential signals; The real-time differential signal threshold is set through the anomaly analysis and judgment module. In the set of real-time differential signals, if a real-time differential signal exists... Greater than or equal to the real-time differential signal threshold If an anomaly is detected in the switch rail, the anomaly result is sent to the early warning module, and the real-time differential signal indicating the anomaly is also sent. The corresponding information data is sent to the differential signal analysis module.

3. The method for non-destructive testing of railway switch rails using ultrasonic guided waves according to claim 2, characterized in that: The corresponding information data are used as input features to the anomaly feedback verification model, and prediction is performed through the anomaly feedback verification model to obtain the predicted differential signal. , where i represents the predicted differential signal corresponding to the i-th group of ultrasonic guided wave transducers; The differential signal analysis module further includes a difference calculation unit and a result judgment unit, used for analyzing real-time differential signals. Differential signal with prediction The difference between the differential signals The calculation is performed; the result judgment unit is used to determine the difference between the differential signals. The result is compared with a set threshold. If the result does not exceed the set threshold, the result is sent to the early warning module. If the result exceeds the set threshold, the prediction result of the abnormal feedback verification model is verified again.

4. The method for non-destructive testing of railway switch rails using ultrasonic guided waves according to claim 3, characterized in that: The steps to re-validate the prediction results of the anomaly feedback validation model include: S101, in the set of real-time differential signals In the context of real-time differential signals Less than the real-time differential signal threshold The data packets, which include various information data, are retrieved, and the real-time differential signal is transmitted. Redefining it as a real-time verification differential signal ; S102. Input the information data from the data packet retrieved in step S101 into the anomaly feedback verification model as input features, and use the anomaly feedback verification model to output the prediction verification differential signal. ; S103, Verify the predicted differential signal Real-time verification differential signal in the retrieved data packet Perform a comparison; S104. Judge the comparison results of step S103. If they are consistent, it indicates that the prediction result of the abnormal feedback verification model is normal. The judgment result of the result judgment unit is not sent to the early warning module. Since the early warning module has not received the instruction from the differential signal analysis module, it will not issue an early warning. If they are inconsistent, it indicates that the prediction result of the abnormal feedback verification model is abnormal and the prediction result of the abnormal feedback verification model is unreliable. The differential signal analysis module sends an instruction to the early warning module. The early warning module receives the instructions from the differential signal analysis module and the abnormal analysis judgment module and issues an early warning.

5. A method for non-destructive testing of railway switch rails using ultrasonic guided waves according to any one of claims 1-4, characterized in that: The information data includes at least electromagnetic noise data, mechanical vibration data, and real-time temperature data of the switch rail.

6. The method for non-destructive testing of railway switch rails using ultrasonic guided waves according to any one of claims 1-4, characterized in that: The ultrasonic guided wave transducer is a piezoelectric transducer.

7. A testing device for implementing the ultrasonic guided wave nondestructive testing method for railway switch rails according to any one of claims 1-4, characterized in that: The detection device includes a tensioning bracket, a probe moving seat, a probe, and an analysis system; The tensioning brackets are symmetrically installed on both sides of a switch rail. A probe moving seat is provided on the tensioning bracket, and a probe is installed on the probe moving seat. The probe moving seat is used to adjust the position of the probe. An ultrasonic guided wave transducer is integrated inside the probe. The detection device also includes an excitation module connected to an ultrasonic guided wave transducer for generating ultrasonic guided wave signals; The analysis system is used to perform intelligent analysis on the information data collected by the detection module.

Images