Rail periodic scratch identification method and device

Abstract

The application discloses a rail periodic scratch identification method and device, and the method comprises the following steps: determining a signal envelope of a vehicle axle box vibration acceleration signal according to vehicle dynamic response data of a rail to be detected; performing a trend item filtering treatment on the signal envelope to obtain a target signal; decomposing the target signal by using a CEEMDAN algorithm to obtain a multi-order intrinsic mode component of the target signal; determining a reconstructed signal of the vehicle dynamic response signal according to the multi-order intrinsic mode component; and performing a CWD analysis on the reconstructed signal to determine whether the rail to be detected has periodic scratches. The application can realize effective identification of rail periodic scratches and improve the accuracy and identification efficiency of rail periodic scratch identification.

Description

Technical Field

[0001] This invention relates to the field of railway engineering technology, and in particular to a method and device for identifying periodic rail abrasions. Background Technology

[0002] This section is intended to provide background or context for the embodiments of the invention set forth in the claims. The description herein is not an admission that it is prior art simply because it is included in this section.

[0003] Rails guide the wheels of locomotives and rolling stock and bear their immense pressure, serving as a primary component of the track structure. Their condition directly impacts the smoothness and stability of train operation. Rail surface abrasion is a significant form of rail damage; if not addressed promptly, it can lead to cracks and spalling, severely affecting track performance. Most rail abrasion is caused by wheel-rail friction and can be categorized into two main types: one is point abrasion or localized abrasion occurring in localized sections of the rail, exhibiting non-periodic characteristics; the other type occurs at intervals equal to the wheel diameter, exhibiting typical periodic characteristics. Periodic rail abrasion generates periodic excitation on train wheelsets, bogies, and the track structure, potentially leading to loosening or cracking of structural components and reducing their service life. Accurate and effective diagnostic technology for periodic rail abrasion can promptly identify abrasion sections, guiding on-site personnel in maintenance and repair, and providing reliable assurance for railway transportation.

[0004] Current methods for diagnosing periodic rail abrasion are based on machine vision and image processing technologies, which cannot be applied to existing high-speed integrated inspection trains that are not equipped with track image acquisition systems. Furthermore, these methods are not highly specific for identifying periodic rail abrasion and cannot accurately identify it, resulting in low accuracy and efficiency in identifying periodic rail abrasion. Summary of the Invention

[0005] This invention provides a method for identifying periodic rail abrasions, which effectively identifies periodic rail abrasions, improving the accuracy and efficiency of identification. The method includes:

[0006] Based on the vehicle dynamic response data of the rail to be tested, determine the signal envelope of the vehicle axle box vibration acceleration signal;

[0007] The target signal is obtained by filtering out the trend term from the signal envelope.

[0008] The target signal is decomposed using the CEEMDAN algorithm to obtain its multi-order intrinsic mode components.

[0009] Based on the multi-order intrinsic mode components, the reconstructed signal of the vehicle dynamic response signal is determined;

[0010] CWD analysis was performed on the reconstructed signal to determine whether there were periodic scratches on the rail to be inspected.

[0011] This invention also provides a rail periodic abrasion identification device to effectively identify rail periodic abrasions, improving the accuracy and efficiency of rail periodic abrasion identification. The device includes:

[0012] The signal envelope determination module is used to determine the signal envelope of the vehicle axle box vibration acceleration signal based on the vehicle dynamic response data of the rail to be tested.

[0013] The trend removal processing module is used to remove the trend term from the signal envelope to obtain the target signal.

[0014] The CEEMDAN decomposition module is used to decompose the target signal using the CEEMDAN algorithm to obtain the multi-order intrinsic mode components of the target signal.

[0015] The reconstruction module is used to determine the reconstructed signal of the vehicle dynamic response signal based on the multi-order intrinsic mode components.

[0016] The CWD analysis module is used to perform CWD analysis on the reconstructed signal to determine whether there are periodic scratches on the rail to be inspected.

[0017] This invention also provides a computer device, including a memory, a processor, and a computer program stored in the memory and executable on the processor. When the processor executes the computer program, it implements the above-described method for identifying periodic rail abrasions.

[0018] This invention also provides a computer-readable storage medium storing a computer program that, when executed by a processor, implements the above-described method for identifying periodic rail scratches.

[0019] This invention also provides a computer program product, which includes a computer program that, when executed by a processor, implements the above-described method for identifying periodic rail scratches.

[0020] In this embodiment of the invention, the signal envelope of the vehicle axle box vibration acceleration signal is determined based on the vehicle dynamic response data of the rail to be tested. The signal envelope is then filtered to remove the trend term, yielding the target signal. The target signal is decomposed using the CEEMDAN algorithm to obtain its multi-order intrinsic mode components. Based on these multi-order intrinsic mode components, the reconstructed signal of the vehicle dynamic response signal is determined. CWD analysis is performed on the reconstructed signal to determine whether the rail to be tested exhibits periodic abrasion. Compared with existing technologies that diagnose rail periodic abrasion based on machine vision and image processing, this invention, using the CEEMDAN algorithm, can determine the multi-order intrinsic mode components and thus the reconstructed signal. This enables data feature mining of rail periodic abrasion sections and, through CWD analysis of the reconstructed signal, effectively determines whether the rail to be tested exhibits periodic abrasion, achieving effective identification of rail periodic abrasion. This solves the problem of existing technologies being unable to identify rail periodic abrasion due to the lack of a track image acquisition system, improving the accuracy and efficiency of rail periodic abrasion identification and providing scientific guidance for rail maintenance. Attached Figure Description

[0021] 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. In the drawings:

[0022] Figure 1 This is a schematic diagram of a method for identifying periodic scratches on rails according to an embodiment of the present invention;

[0023] Figure 2 This is a specific example diagram of a raw axle box acceleration signal in an embodiment of the present invention;

[0024] Figure 3 This is a specific example diagram of a multi-order intrinsic mode component in an embodiment of the present invention;

[0025] Figure 4 This is a specific example diagram illustrating the instantaneous amplitude ratio of each order of IMF to the original signal in an embodiment of the present invention;

[0026] Figure 5 This is a schematic diagram of the structure of a reconstructed signal CWD and its frequency and time marginal spectrum in an embodiment of the present invention;

[0027] Figure 6 This is a specific example diagram of the temporal marginal spectrum trend term of a sensitive frequency component in an embodiment of the present invention;

[0028] Figure 7 This is a flowchart illustrating a method for identifying periodic scratches on rails according to an embodiment of the present invention.

[0029] Figure 8 This is a schematic diagram of a method for identifying periodic scratches on rails according to an embodiment of the present invention;

[0030] Figure 9 This is a structural example diagram of a rail periodic scratch identification device according to an embodiment of the present invention;

[0031] Figure 10 This is a schematic diagram of a computer device used for identifying periodic scratches on rails in an embodiment of the present invention. Detailed Implementation

[0032] To make the objectives, technical solutions, and advantages of the embodiments of the present invention clearer, the embodiments of the present invention will be further described in detail below with reference to the accompanying drawings. Here, the illustrative embodiments of the present invention and their descriptions are used to explain the present invention, but are not intended to limit the present invention.

[0033] In this document, the term "and / or" merely describes a relationship, indicating that three relationships can exist. For example, A and / or B can represent three cases: A alone, A and B simultaneously, and B alone. Furthermore, the term "at least one" in this document means any combination of at least two of any one or more elements. For example, including at least one of A, B, and C can mean including any one or more elements selected from the set consisting of A, B, and C.

[0034] In the description of this specification, the terms "comprising," "including," "having," and "containing" are open-ended terms, meaning that they include but are not limited to. The terms "an embodiment," "a specific embodiment," "some embodiments," and "for example," etc., refer to specific features, structures, or characteristics described in connection with that embodiment or example that are included in at least one embodiment or example of this application. In this specification, the illustrative expressions of the above terms do not necessarily refer to the same embodiment or example. Furthermore, the specific features, structures, or characteristics described can be combined in any suitable manner in one or more embodiments or examples. The order of steps involved in the various embodiments is used to illustrate the implementation of this application, and the order of steps is not limited and can be adjusted appropriately as needed.

[0035] Rails guide the wheels of locomotives and rolling stock and bear their immense pressure, serving as a primary component of the track structure. Their condition directly impacts the smoothness and stability of train operation. Rail surface abrasion is a significant form of rail damage; if not addressed promptly, it can lead to cracks and spalling, severely affecting track performance. Most rail abrasion is caused by wheel-rail friction and can be categorized into two main types: one is point abrasion or localized abrasion occurring in localized sections of the rail, exhibiting non-periodic characteristics; the other type occurs at intervals equal to the wheel diameter, exhibiting typical periodic characteristics. Periodic rail abrasion generates periodic excitation on train wheelsets, bogies, and the track structure, potentially leading to loosening or cracking of structural components and reducing their service life. Accurate and effective diagnostic technology for periodic rail abrasion can promptly identify abrasion sections, guiding on-site personnel in maintenance and repair, and providing reliable assurance for railway transportation.

[0036] Current methods for diagnosing periodic rail abrasion are based on machine vision and image processing technologies, which cannot be applied to existing high-speed integrated inspection trains that are not equipped with track image acquisition systems. Furthermore, these methods are not highly specific for identifying periodic rail abrasion and cannot accurately identify it, resulting in low accuracy and efficiency in identifying periodic rail abrasion.

[0037] To address the aforementioned problems, this invention provides a method for identifying periodic rail abrasions, thereby achieving effective identification of periodic rail abrasions and improving the accuracy and efficiency of identification. (See also...) Figure 7 The method may include:

[0038] Step 701: Determine the signal envelope of the vehicle axle box vibration acceleration signal based on the vehicle dynamic response data of the rail to be tested;

[0039] Step 702: Filter out the trend term from the signal envelope to obtain the target signal;

[0040] Step 703: Decompose the target signal using the CEEMDAN algorithm to obtain the multi-order intrinsic mode components of the target signal;

[0041] Step 704: Determine the reconstructed signal of the vehicle dynamic response signal based on the multi-order intrinsic mode components;

[0042] Step 705: Perform CWD analysis on the reconstructed signal to determine whether there are periodic scratches on the rail to be inspected.

[0043] In this embodiment of the invention, the signal envelope of the vehicle axle box vibration acceleration signal is determined based on the vehicle dynamic response data of the rail to be tested. The signal envelope is then filtered to remove the trend term, yielding the target signal. The target signal is decomposed using the CEEMDAN algorithm to obtain its multi-order intrinsic mode components. Based on these multi-order intrinsic mode components, the reconstructed signal of the vehicle dynamic response signal is determined. CWD analysis is performed on the reconstructed signal to determine whether the rail to be tested exhibits periodic abrasion. Compared with existing technologies that diagnose rail periodic abrasion based on machine vision and image processing, this invention, using the CEEMDAN algorithm, can determine the multi-order intrinsic mode components and thus the reconstructed signal. This enables data feature mining of rail periodic abrasion sections and, through CWD analysis of the reconstructed signal, effectively determines whether the rail to be tested exhibits periodic abrasion, achieving effective identification of rail periodic abrasion. This solves the problem of existing technologies being unable to identify rail periodic abrasion due to the lack of a track image acquisition system, improving the accuracy and efficiency of rail periodic abrasion identification and providing scientific guidance for rail maintenance.

[0044] In practice, the signal envelope of the vehicle axle box vibration acceleration signal is first determined based on the vehicle dynamic response data of the rail to be tested.

[0045] In this embodiment, the vehicle dynamic response data includes axle box vibration acceleration detection data; the vehicle dynamic response signal includes axle box vibration acceleration signal.

[0046] Based on the vehicle dynamic response data of the rail to be tested, the signal envelope of the axle box vibration signal is determined, including:

[0047] The vibration acceleration signal of the axle box is determined based on the vibration acceleration detection data of the axle box of the rail to be tested.

[0048] The Hilbert transform is performed on the axle box vibration acceleration signal to obtain the signal envelope of the axle box vibration acceleration signal.

[0049] In the above embodiments, by determining the signal envelope of the vehicle axle box vibration acceleration signal, it is helpful to determine the multi-order intrinsic mode components in subsequent steps using the CEEMDAN algorithm, thereby determining the reconstructed signal and realizing the data feature mining of the rail periodically scratched section.

[0050] For example, the vehicle dynamic response signal (i.e., the axle box vibration acceleration signal) can be denoted as x0(t), t = 1, 2, ..., M; where M is the number of sampling points. Performing a Hilbert transform on the above vehicle dynamic response signal yields the signal envelope of the corresponding vehicle dynamic response signal.

[0051] Furthermore, the signal x0 can be subjected to a Hilbert transform H(x0) according to the following formula (1):

[0052]

[0053] Where H(x0(t)) is the Hilbert transform of signal x0; x0(t) is the vehicle dynamic response signal; P is the Cauchy principal component; t represents the t-th sampling point, t = 1, 2, ..., M; where M is the number of sampling points.

[0054] To give another example, consider the vibration acceleration detection data of axle boxes in a section of a high-speed railway. Figure 2 As shown, the vibration signal of the axle box in this section exhibits a periodic increase, with a wavelength of approximately 2.75m.

[0055] In practice, after determining the signal envelope of the vehicle axle box vibration acceleration signal based on the vehicle dynamic response data of the rail to be tested, the trend term is filtered out from the signal envelope to obtain the target signal.

[0056] For example, the trend term can be filtered out from the signal envelope of the vehicle dynamic response signal x0(t) to obtain the target signal x(t).

[0057] In practice, after filtering out the trend term from the signal envelope to obtain the target signal, the CEEMDAN algorithm is used to decompose the target signal to obtain the multi-order intrinsic mode components of the target signal.

[0058] In this embodiment, the target signal is decomposed using the CEEMDAN algorithm to obtain the multi-order intrinsic mode components of the signal envelope, including:

[0059] Gaussian white noise is added to the target signal multiple times to obtain the processed signal;

[0060] The processed signal is decomposed into EMD components to obtain the first-order EMD eigenmode components.

[0061] The average value of the first-order EMD eigenmode components is taken as the first-order eigenmode component of the signal envelope using the CEEMDAN algorithm.

[0062] The target signal with the first-order intrinsic mode components removed is used as the residual signal;

[0063] Gaussian white noise is added multiple times to the residual signal to obtain the updated signal; EMD decomposition is performed on the updated signal to obtain the first-order EMD intrinsic mode components; the average value of the first-order EMD intrinsic mode components of the updated signal is taken as the second-order intrinsic mode component of the signal envelope; the residual signal after removing the second-order intrinsic mode components is taken as the updated residual signal; the above steps of determining the updated signal and the updated residual signal are repeated until the obtained updated residual signal is a monotonic function.

[0064] Obtain the multi-order intrinsic mode components of the signal envelope obtained in the above steps.

[0065] In one embodiment, Gaussian white noise v can be added to the signal to be decomposed (i.e. the target signal mentioned above) multiple times to obtain a new signal (i.e. the processed signal mentioned above). The new signal can be decomposed into EMD according to the following formula (2) to obtain the first-order EMD Intrinsic Mode Function (IMF) C1.

[0066]

[0067] Where E(·) represents the EMD calculation; j represents the number of noise additions, j = 1, 2, ..., N; q represents the sign parameter of the added white noise, q = 1, 2; ε represents the white noise standard table; r represents the residual component; v j (t) represents the added white noise signal.

[0068] In one embodiment, the N first-order modal components C1 generated by adding N white noise signals can be calculated according to the following formula (3). j The first-order IMF of the CEEMDAN decomposition can be obtained by averaging the overall values ​​of the first-order EMD intrinsic mode components mentioned above.

[0069]

[0070] in, t is the first-order intrinsic mode component of the signal envelope; N is the number of added white noise signals; j is the number of times noise is added, j = 1, 2, ..., N; C1(t) is the first-order EMD intrinsic mode component.

[0071] In one embodiment, the residual (i.e., the residual signal mentioned above) after removing the first-order IMF can be calculated according to the following formula (4):

[0072]

[0073] Where r1(t) is the residual after removing the first-order IMF; x(t) is the target signal; These are the first-order intrinsic mode components;

[0074] In one embodiment, N Gaussian white noises can be added to r1(t) according to the following formula (5) to obtain a new irregular signal (i.e., the updated processing signal mentioned above). The new irregular signal is used as a carrier for EMD decomposition to obtain the corresponding first-order mode component D1. Thus, the second IMF of CEEMDAN decomposition can be obtained:

[0075]

[0076] in, denoted as the second-order intrinsic mode component of the signal envelope; N is the number of added white noise signals; j is the number of times noise is added, j = 1, 2, ..., N; D1 is the first-order EMD intrinsic mode component.

[0077] In the embodiment, D1 refers to the first-order IMF of the residual signal, and C1 in front is the first-order IMF of the original envelope signal.

[0078] In one embodiment, the residual after removing the second modal component (i.e., the updated residual signal mentioned above) can be calculated according to the following formula (6):

[0079]

[0080] Where r2(t) is the residual after removing the second-order IMF; r1(t) is the residual after removing the first-order IMF; It represents the second-order intrinsic mode component of the signal envelope.

[0081] In one embodiment, the algorithm terminates by repeatedly executing the steps of determining the update processing signal and the update residual signal until the obtained update residual signal is a monotonic function and cannot be further decomposed.

[0082] The number of IMFs obtained at this point is K, and the original signal (i.e., the target signal) is decomposed as shown in formula (7):

[0083]

[0084] Where x(t) is the target signal; K is the number of intrinsic mode components of the signal envelope, i.e., the number of CEEMDAN decompositions; r k (t) represents the residual after removing the k-th order IMF; Let be the k-th eigenmode component of the signal envelope.

[0085] For example, the CEEMDAN method can be used to decompose the filtered envelope of the signal to obtain IMFs of various orders, such as... Figure 3 As shown.

[0086] In practice, after decomposing the target signal using the CEEMDAN algorithm to obtain the multi-order intrinsic mode components of the target signal, the reconstructed signal of the vehicle dynamic response signal is determined based on the multi-order intrinsic mode components.

[0087] In this embodiment, the reconstructed signal of the axle box vibration signal is determined based on the multi-order intrinsic mode components, including:

[0088] The target signal and each of its intrinsic mode components are subjected to Hilbert transform to obtain the analytical form of the target signal and each of its intrinsic mode components.

[0089] Based on the analytical form of the target signal and each intrinsic mode component, calculate the instantaneous amplitude of the target signal and each intrinsic mode component for each sampling point.

[0090] The reconstructed signal of the axle box vibration signal is determined based on the target signal and the instantaneous amplitude of each intrinsic mode component at each sampling point.

[0091] In one embodiment, Hilbert transform can be performed on the original signal (i.e., the target signal mentioned above) and the IMFs (i.e., each order of intrinsic mode components of the target signal mentioned above) obtained by CEEMDAN decomposition using formula (8), to obtain its analytical form as follows:

[0092]

[0093] Among them, Z j (t) represents the signal after the Hilbert transform; Let i be the real part of the j-th order IMF; i is the imaginary unit. for The imaginary part after Hilbert transform; a j The instantaneous amplitude of the j-th order IMF; It represents the instantaneous phase of the j-th order IMF.

[0094] In one embodiment, the instantaneous amplitude of each order IMF and the original signal can be calculated using formula (9). instantaneous amplitude a j for:

[0095]

[0096] Among them, a j (t) represents the instantaneous amplitude of the j-th intrinsic mode component at the t-th sampling point; Let the real part of the j-th order IMF be denoted as 'j'. for The imaginary part after Hilbert transform.

[0097] In one embodiment, the reconstructed signal of the axle box vibration signal is determined based on the target signal and the instantaneous amplitude of each intrinsic mode component at each sampling point, including:

[0098] For each intrinsic mode component, the instantaneous amplitude at each sampling point is calculated as follows: the ratio of the instantaneous amplitude of the intrinsic mode component at that sampling point to the instantaneous amplitude of the target signal at that sampling point, which is used as the first data of the intrinsic mode component at that sampling point; the average value of the first data of the intrinsic mode component at each sampling point is calculated.

[0099] The eigenmode components whose average value is greater than the first preset value are taken as signal reconstruction sub-items;

[0100] The reconstructed components of the above signals are added together to obtain the reconstructed signal of the axle box vibration signal.

[0101] In one embodiment, the average value p of the ratio of the instantaneous amplitude of each order IMF to the instantaneous amplitude of the original signal can be calculated using formula (10). j :

[0102]

[0103] Where, p j This represents the first data point of the j-th eigenmode component at this sampling point; a j (t) represents the instantaneous amplitude of the j-th intrinsic mode component at the t-th sampling point; a0(t) represents the instantaneous amplitude of the original envelope signal; t represents the t-th sampling point, t = 1, 2, ..., M; where M is the number of sampling points.

[0104] In the above embodiments, the signal envelope of the vehicle axle box vibration acceleration signal is determined based on the vehicle dynamic response data of the rail to be detected; the trend term is filtered out from the signal envelope to obtain the target signal; the target signal is decomposed using the CEEMDAN algorithm to obtain the multi-order intrinsic mode components of the target signal; and the reconstructed signal of the vehicle dynamic response signal is determined based on the multi-order intrinsic mode components, thus realizing vehicle dynamic response data mining based on CEEMDAN.

[0105] In practice, after determining the reconstructed signal of the vehicle dynamic response signal based on the multi-order intrinsic mode components, CWD analysis is performed on the reconstructed signal to determine whether there are periodic scratches on the rail to be tested.

[0106] In this embodiment, CWD analysis is performed on the reconstructed signal to determine whether the rail under test has periodic scratches, such as... Figure 8 As shown, it includes:

[0107] Step 801: Perform CWD analysis on the reconstructed signal and calculate the CWD distribution signal of the reconstructed signal;

[0108] Step 802: Calculate the frequency marginal spectrum and full-frequency time marginal spectrum of the CWD distribution signal;

[0109] Step 803: Extract the periodic scratch frequency components from the frequency marginal spectrum;

[0110] Step 804: Calculate the time margin spectrum of the periodic abrasion frequency components based on the periodic abrasion frequency components and the full frequency time margin spectrum;

[0111] Step 805: Determine whether the rail to be tested has periodic scratches based on the time marginal spectrum of the periodic scratch frequency components.

[0112] In one embodiment, extracting the periodic scratch frequency components from the frequency marginal spectrum includes:

[0113] Determine the maximum values ​​of different maxima points in the frequency marginal spectrum and the corresponding frequency values ​​of different maxima points;

[0114] For each maximum value point, if the maximum value of the maximum value point is greater than the second preset value, and the remainder between the frequency value corresponding to the maximum value point and the preset frequency value corresponding to the periodic abrasion of the rail is less than the third preset value, the maximum value point is determined as the target maximum value point.

[0115] The frequency value of the target maximum point is used as the periodic abrasion frequency component.

[0116] In one embodiment, the CWD of the reconstructed signal y(t) can be calculated according to the following formula (11):

[0117]

[0118] Among them, CWD y (t,ω) is the CWD distribution signal of the reconstructed signal.

[0119] In the embodiment, CWD y In (t,ω): t represents time, ω represents frequency, and the subscript y represents the original signal, which generally does not need to be specified separately.

[0120] In one embodiment, the frequency marginal spectrum s(ω) and the full-frequency time marginal spectrum s(t) of the CWD distributed signal can be calculated according to the following formulas (12) and (13):

[0121]

[0122]

[0123] In one embodiment, determining the maxima of different maxima of the frequency marginal spectrum and the corresponding frequency values ​​of the different maxima may include:

[0124] The maximum points s(ω) of the frequency boundary spectrum x(ω) are obtained. i max ), i = 1, 2, ..., N max N max This represents the number of maximum points.

[0125] In one embodiment, for each maximum point, if the maximum value of the maximum point is greater than a second preset value, and the remainder between the frequency value corresponding to the maximum point and the preset frequency value corresponding to the periodic abrasion of the rail is less than a third preset value, the maximum point is determined as a target maximum point, which may include:

[0126] When s(ω) i max The marginal spectrum at this point is greater than the set threshold T. s Furthermore, the remainder after dividing the frequency at this point by the frequency ω0 corresponding to the periodic abrasion of the rail is less than the threshold T. ω Record the frequency components ω that satisfy the above conditions. k k = 1, 2, ..., N ω N ω The number of maximum points to satisfy the above conditions.

[0127] In one embodiment, calculating the time margin spectrum of the periodic abrasion frequency components based on the periodic abrasion frequency components and the full-frequency time margin spectrum includes:

[0128] Calculate the sum of the time margin spectra of each periodic abrasion frequency component within a preset frequency range;

[0129] The sum of the above time marginal spectra is subjected to low-pass filtering to obtain the trend term of the sum of the time marginal spectra;

[0130] When the trend term of the sum of the above time marginal spectra is greater than the fourth preset value, the ratio of the sum of the above time marginal spectra to the above frequency marginal spectra is taken as the energy proportion.

[0131] When the energy percentage mentioned above is greater than the fifth preset value, it is determined that the rail to be tested has periodic abrasions.

[0132] In one embodiment, each ω can be calculated according to formula (14). k A certain frequency range (i.e., the preset frequency range mentioned above, such as ω in the following formula) d The sum of the time marginal spectra within ) s p (t):

[0133]

[0134] In one embodiment, s can be used p (t) is low-pass filtered to obtain its trend term s pl (t), when s pl (t) is greater than the set threshold T sp Then, its energy percentage r can be calculated according to formula (15). s When r s Greater than the set threshold T r If so, it is considered that there is periodic rail abrasion in that section.

[0135]

[0136] In the above embodiments, CWD analysis is performed on the reconstructed signal to determine whether there are periodic scratches on the rail to be detected, thus realizing the extraction of periodic scratches on the rail based on CWD.

[0137] The following is a specific embodiment to illustrate the specific application of the method of the present invention.

[0138] The main objective of this invention is to solve the problem of identifying periodic scratches on rail surfaces, providing more reliable data and technical support for rail maintenance and repair operations. This involves the following two specific problems:

[0139] (1) How to extract information on periodic abrasion of vehicle rails from axle box acceleration detection data;

[0140] (2) How to use relevant indicators to accurately identify the characteristics of rail periodic scratch sections.

[0141] See Figure 1 The method provided in the specific embodiments of the present invention may include two parts, which can be described in detail for the above-mentioned specific problems:

[0142] Part 1: Feature mining of rail periodic abrasion sections based on CEEMDAN;

[0143] Part Two: Identification of Periodic Abrasion Sections in Rails Based on CWD.

[0144] This embodiment may specifically include the following steps:

[0145] Part 1: Feature mining of rail periodic abrasion sections based on CEEMDAN:

[0146] (1) Let the vehicle dynamic response signal be x0(t), t = 1, 2, ..., M, where M is the number of sampling points. Perform a Hilbert transform on it to obtain the signal envelope. Filter out the trend term to obtain the signal x(t). The Hilbert transform H(x0) of signal x0 is:

[0147]

[0148] Where P is the Cauchy principal component.

[0149] (2) Add Gaussian white noise v to the signal to be decomposed multiple times to obtain a new signal. Perform EMD decomposition on the new signal to obtain the first-order intrinsic mode component (IMF) C1.

[0150]

[0151] Where E(·) is the EMD calculation, j = 1, 2, ..., N is the number of noise additions, q = 1, 2, ε is the white noise standard table, and r is the residual component.

[0152] (3) For the N first-order modal components C1 generated by adding N white noise signals respectively, j The first-order IMF of the CEEMDAN decomposition is obtained by performing an overall average.

[0153]

[0154] (4) Calculate the residuals after removing the first-order IMF:

[0155]

[0156] (5) Add N Gaussian white noises to r1(t) to obtain a new irregular signal. Use the new signal as a carrier to perform EMD decomposition to obtain the corresponding first-order mode component D1. From this, the second IMF of CEEMDAN decomposition can be obtained:

[0157]

[0158] (6) Calculate the residual after removing the second modal component:

[0159]

[0160] (7) Repeat the above steps until the obtained residual signal is a monotonic function and cannot be further decomposed, at which point the algorithm ends. At this point, the number of IMFs obtained is K, and the original signal is decomposed into:

[0161]

[0162] (8) Using formula (1), perform Hilbert transform on the original signal and the IMFs of each order obtained from CEEMDAN decomposition to obtain their analytical forms:

[0163]

[0164] (9) Calculate the instantaneous amplitude of each order of IMF and the original signal. instantaneous amplitude a j for:

[0165]

[0166] (10) Calculate the average value p of the ratio of the instantaneous amplitude of each order IMF to the instantaneous amplitude of the original signal. j :

[0167]

[0168] (11) Select p j The IMF corresponding to values ​​greater than the set threshold is used as a component of the signal reconstruction, and these components are added together to obtain the reconstructed signal y(t).

[0169] Part Two: Identification of Periodic Abrasion Sections in Rails Based on CWD.

[0170] (1) Calculate the CWD of the reconstructed signal y(t):

[0171]

[0172] (2) Calculate CWD y (t, ω) Frequency marginal spectrum s(ω) and time marginal spectrum s(t):

[0173]

[0174]

[0175] (3) Obtain the maximum points s(ω) of the frequency boundary spectrum x(ω). i max ), i = 1, 2, ..., N max N max This represents the number of maximum points.

[0176] (4) When s(ω i max The marginal spectrum at this point is greater than the set threshold T. s Furthermore, the remainder after dividing the frequency at this point by the frequency ω0 corresponding to the periodic abrasion of the rail is less than the threshold T. ω Record the frequency components ω that satisfy the above conditions. k k = 1, 2, ..., N ω N ω The number of maximum points to satisfy the above conditions.

[0177] (5) Calculate each ω kThe sum of the time marginal spectrum within a certain frequency range, s p (t):

[0178]

[0179] (6) For s p (t) is low-pass filtered to obtain its trend term s pl (t), when s pl (t) is greater than the set threshold T sp Then, its energy percentage r is calculated according to formula (15). s When r s Greater than the set threshold T r If so, it is considered that there is periodic rail abrasion in that section.

[0180]

[0181] Here is an example of the above specific embodiments:

[0182] Vibration acceleration test data of axle boxes in a certain section of high-speed railway is as follows: Figure 2 As shown, the vibration signal of the axle box in this section exhibits a periodic increase, with a wavelength of approximately 2.75m.

[0183] First, the CEEMDAN method in the scheme can be used to decompose the envelope of the filtered signal to obtain the IMFs of each order, such as... Figure 3 As shown, the amplitudes of each IMF order are compared to the amplitudes of the original signal, for example... Figure 4 As shown in Table 1, the average amplitude of each order of IMF is compared with the original signal.

[0184] Table 1

[0185] IMF order Average amplitude percentage 1 0.149 2 0.064 3 0.269 4 0.322 5 0.310 6 0.289 7 0.491 8 0.252 9 0.228 10 0.231 11 0.247 12 0.183 13 0.018 14 0.025

[0186] Subsequently, signals from IMFs of all orders (i.e., 4th, 5th, and 7th order IMFs) with an average amplitude percentage greater than 0.3 were extracted as reconstruction terms to reconstruct the original envelope signal. The CWD distribution of the reconstructed signal and its frequency marginal spectrum, along with the corresponding time marginal spectra of the sensitive frequency components, were calculated. The results are as follows: Figure 5 As shown, the top figure is the filtered envelope signal, the middle figure is its CWD distribution, the left figure is the frequency boundary spectrum, and the bottom figure is the time boundary spectrum corresponding to the sensitive frequency components.

[0187] from Figure 5 As shown in the left figure, there are six main frequency components in the frequency boundary spectrum of the CWD distribution of this signal: 26Hz, 79Hz, 106Hz, 132Hz, 185Hz and 212Hz. Among them, 26Hz is the dominant frequency, and the rest are harmonics. The train speed in this section is 258km / h, which corresponds to a wavelength of about 2.75m.

[0188] Figure 5 The figure below shows the time marginal spectrum corresponding to the above six main frequency components. Figure 6 As can be seen from its trend term, the temporal marginal spectrum of this section (within the red rectangle) is significantly higher than that on both sides, and the energy proportion reaches 0.8, both exceeding the set threshold. On-site verification revealed periodic rail abrasion in this section, with a wavelength of approximately 2.7–2.8 m.

[0189] In summary, this invention, based on vehicle dynamic response detection technology, applies Complete Efficient Mode Decomposition with Adaptive Noise (CEEMDAN) and Choi-Williams Distribution (CWD) to effectively identify periodic rail abrasions, providing scientific guidance for rail maintenance.

[0190] Of course, it is understood that there may be other variations of the above detailed process, and all such variations should fall within the protection scope of this invention.

[0191] In this embodiment of the invention, the signal envelope of the vehicle axle box vibration acceleration signal is determined based on the vehicle dynamic response data of the rail to be tested. The signal envelope is then filtered to remove the trend term, yielding the target signal. The target signal is decomposed using the CEEMDAN algorithm to obtain its multi-order intrinsic mode components. Based on these multi-order intrinsic mode components, the reconstructed signal of the vehicle dynamic response signal is determined. CWD analysis is performed on the reconstructed signal to determine whether the rail to be tested exhibits periodic abrasion. Compared with existing technologies that diagnose rail periodic abrasion based on machine vision and image processing, this invention, using the CEEMDAN algorithm, can determine the multi-order intrinsic mode components and thus the reconstructed signal. This enables data feature mining of rail periodic abrasion sections and, through CWD analysis of the reconstructed signal, effectively determines whether the rail to be tested exhibits periodic abrasion, achieving effective identification of rail periodic abrasion. This solves the problem of existing technologies being unable to identify rail periodic abrasion due to the lack of a track image acquisition system, improving the accuracy and efficiency of rail periodic abrasion identification and providing scientific guidance for rail maintenance.

[0192] This invention also provides a rail periodic abrasion identification device, as described in the following embodiments. Since the principle behind this device is similar to that of the rail periodic abrasion identification method, its implementation can be referenced from the implementation of the rail periodic abrasion identification method; repeated details will not be elaborated further.

[0193] This invention provides a rail periodic abrasion identification device to effectively identify rail periodic abrasions, improving the accuracy and efficiency of rail periodic abrasion identification. Figure 9 As shown, the device includes:

[0194] The signal envelope determination module 901 is used to determine the signal envelope of the vehicle axle box vibration acceleration signal based on the vehicle dynamic response data of the rail to be tested.

[0195] The trend removal processing module 902 is used to remove the trend term from the signal envelope to obtain the target signal.

[0196] CEEMDAN decomposition module 903 is used to decompose the target signal using the CEEMDAN algorithm to obtain the multi-order intrinsic mode components of the target signal.

[0197] The reconstruction module 904 is used to determine the reconstructed signal of the vehicle dynamic response signal based on the multi-order intrinsic mode components.

[0198] The CWD analysis module 905 is used to perform CWD analysis on the reconstructed signal to determine whether there are periodic scratches on the rail to be inspected.

[0199] In one embodiment, the aforementioned vehicle dynamic response data includes axle box vibration acceleration detection data;

[0200] The signal envelope determination module is specifically used for:

[0201] The vibration acceleration signal of the axle box is determined based on the vibration acceleration detection data of the axle box of the rail to be tested.

[0202] The Hilbert transform is performed on the axle box vibration acceleration signal to obtain the signal envelope of the axle box vibration acceleration signal.

[0203] In one embodiment, the CEEMDAN decomposition module is specifically used for:

[0204] Gaussian white noise is added to the target signal multiple times to obtain the processed signal;

[0205] The processed signal is decomposed into EMD components to obtain the first-order EMD eigenmode components.

[0206] The average value of the first-order EMD eigenmode components is taken as the first-order eigenmode component of the signal envelope using the CEEMDAN algorithm.

[0207] The target signal with the first-order intrinsic mode components removed is used as the residual signal;

[0208] Gaussian white noise is added multiple times to the residual signal to obtain the updated signal; EMD decomposition is performed on the updated signal to obtain the first-order EMD intrinsic mode components; the average value of the first-order EMD intrinsic mode components of the updated signal is taken as the second-order intrinsic mode component of the signal envelope; the residual signal after removing the second-order intrinsic mode components is taken as the updated residual signal; the above steps of determining the updated signal and the updated residual signal are repeated until the obtained updated residual signal is a monotonic function.

[0209] Obtain the multi-order intrinsic mode components of the signal envelope obtained in the above steps.

[0210] In one embodiment, the refactoring module is specifically used for:

[0211] The target signal and each of its intrinsic mode components are subjected to Hilbert transform to obtain the analytical form of the target signal and each of its intrinsic mode components.

[0212] Based on the analytical form of the target signal and each intrinsic mode component, calculate the instantaneous amplitude of the target signal and each intrinsic mode component for each sampling point.

[0213] The reconstructed signal of the axle box vibration signal is determined based on the target signal and the instantaneous amplitude of each intrinsic mode component at each sampling point.

[0214] In one embodiment, the refactoring module is specifically used for:

[0215] For each intrinsic mode component, the instantaneous amplitude at each sampling point is calculated as follows: the ratio of the instantaneous amplitude of the intrinsic mode component at that sampling point to the instantaneous amplitude of the target signal at that sampling point, which is used as the first data of the intrinsic mode component at that sampling point; the average value of the first data of the intrinsic mode component at each sampling point is calculated.

[0216] The eigenmode components whose average value is greater than the first preset value are taken as signal reconstruction sub-items;

[0217] The reconstructed components of the above signals are added together to obtain the reconstructed signal of the axle box vibration signal.

[0218] In one embodiment, the CWD analysis module is specifically used for:

[0219] Perform CWD analysis on the reconstructed signal and calculate the CWD distribution signal of the reconstructed signal;

[0220] Calculate the frequency marginal spectrum and full-frequency time marginal spectrum of the CWD distributed signal;

[0221] Extract the periodic scratch frequency components from the frequency marginal spectrum;

[0222] Calculate the time margin spectrum of the periodic abrasion frequency components based on the periodic abrasion frequency components and the full-frequency time margin spectrum;

[0223] Based on the time marginal spectrum of the frequency components of periodic abrasion, it is determined whether the rail to be inspected has periodic abrasion.

[0224] In one embodiment, the CWD analysis module is specifically used for:

[0225] Determine the maximum values ​​of different maxima points in the frequency marginal spectrum and the corresponding frequency values ​​of different maxima points;

[0226] For each maximum value point, if the maximum value of the maximum value point is greater than the second preset value, and the remainder between the frequency value corresponding to the maximum value point and the preset frequency value corresponding to the periodic abrasion of the rail is less than the third preset value, the maximum value point is determined as the target maximum value point.

[0227] The frequency value of the target maximum point is used as the periodic abrasion frequency component.

[0228] In one embodiment, the CWD analysis module is specifically used for:

[0229] Calculate the sum of the time margin spectra of each periodic abrasion frequency component within a preset frequency range;

[0230] The sum of the above time marginal spectra is subjected to low-pass filtering to obtain the trend term of the sum of the time marginal spectra;

[0231] When the trend term of the sum of the above time marginal spectra is greater than the fourth preset value, the ratio of the sum of the above time marginal spectra to the above frequency marginal spectra is taken as the energy proportion.

[0232] When the energy percentage mentioned above is greater than the fifth preset value, it is determined that the rail to be tested has periodic abrasions.

[0233] This invention provides an embodiment of a computer device for implementing all or part of the above-described method for identifying periodic rail abrasions. Specifically, the computer device includes the following components:

[0234] The computer device comprises a processor, memory, a communications interface, and a bus; wherein the processor, memory, and communications interface communicate with each other via the bus; the communications interface is used to realize information transmission between related devices; the computer device can be a desktop computer, tablet computer, or mobile terminal, etc., and this embodiment is not limited to these. In this embodiment, the computer device can be implemented with reference to the embodiments for implementing the method for identifying periodic scratches on rails and the embodiments for implementing the device for identifying periodic scratches on rails, the contents of which are incorporated herein by reference, and repeated details will not be described again.

[0235] Figure 10 This is a schematic block diagram illustrating the system configuration of the computer device 1000 according to an embodiment of this application. Figure 10 As shown, the computer device 1000 may include a central processing unit 1001 and a memory 1002; the memory 1002 is coupled to the central processing unit 1001. It is worth noting that... Figure 10 This is an example; other types of structures can also be used to supplement or replace this structure to achieve telecommunications functions or other functions.

[0236] In one embodiment, the rail periodic scratch recognition function can be integrated into the central processing unit 1001. The central processing unit 1001 can be configured to perform the following control:

[0237] Based on the vehicle dynamic response data of the rail to be tested, determine the signal envelope of the vehicle axle box vibration acceleration signal;

[0238] The target signal is obtained by filtering out the trend term from the signal envelope.

[0239] The target signal is decomposed using the CEEMDAN algorithm to obtain its multi-order intrinsic mode components.

[0240] Based on the multi-order intrinsic mode components, the reconstructed signal of the vehicle dynamic response signal is determined;

[0241] CWD analysis was performed on the reconstructed signal to determine whether there were periodic scratches on the rail to be inspected.

[0242] In another embodiment, the rail periodic scratch identification device can be configured separately from the central processing unit 1001. For example, the rail periodic scratch identification device can be configured as a chip connected to the central processing unit 1001, and the rail periodic scratch identification function can be realized through the control of the central processing unit.

[0243] like Figure 10As shown, the computer device 1000 may further include: a communication module 1003, an input unit 1004, an audio processor 1005, a display 1006, and a power supply 1007. It is worth noting that the computer device 1000 does not necessarily need to include... Figure 10 All components shown; in addition, the computer device 1000 may also include Figure 10 For components not shown, please refer to existing technologies.

[0244] like Figure 10 As shown, the central processing unit 1001, sometimes also referred to as a controller or operation control, may include a microprocessor or other processor device and / or logic device. The central processing unit 1001 receives input and controls the operation of various components of the computer device 1000.

[0245] The memory 1002 may be, for example, one or more of a cache, flash memory, hard drive, removable medium, volatile memory, non-volatile memory, or other suitable device. It may store the aforementioned failure-related information, and also store a program for executing that information. The central processing unit 1001 may execute the program stored in the memory 1002 to perform information storage or processing, etc.

[0246] Input unit 1004 provides input to central processing unit 1001. This input unit 1004 may be, for example, a keypad or touch input device. Power supply 1007 provides power to computer device 1000. Display 1006 displays images, text, and other display objects. This display may be, for example, an LCD display, but is not limited to this.

[0247] The memory 1002 can be a solid-state memory, such as a read-only memory (ROM), random access memory (RAM), a SIM card, etc. It can also be a memory that retains information even when power is off, can be selectively erased, and contains more data; examples of this type of memory are sometimes referred to as EPROMs, etc. The memory 1002 can also be some other type of device. The memory 1002 includes a buffer memory 1021 (sometimes referred to as a buffer). The memory 1002 may include an application / function storage unit 1022 for storing application programs and function programs or processes for executing operations of the computer device 1000 via the central processing unit 1001.

[0248] The memory 1002 may also include a data storage unit 1023 for storing data, such as contacts, digital data, pictures, sounds, and / or any other data used by the computer device. The driver storage unit 1024 of the memory 1002 may include various drivers for the computer device for communication functions and / or for performing other functions of the computer device (such as messaging applications, address book applications, etc.).

[0249] The communication module 1003 is a transmitter / receiver 1003 that transmits and receives signals via the antenna 1008. The communication module (transmitter / receiver) 1003 is coupled to the central processing unit 1001 to provide input signals and receive output signals, which can be the same as in a conventional mobile communication terminal.

[0250] Based on different communication technologies, multiple communication modules 1003 can be configured in the same computer device, such as cellular network modules, Bluetooth modules, and / or wireless LAN modules. The communication module (transmitter / receiver) 1003 is also coupled to a speaker 1009 and a microphone 1010 via an audio processor 1005 to provide audio output via the speaker 1009 and receive audio input from the microphone 1010, thereby realizing typical telecommunications functions. The audio processor 1005 may include any suitable buffer, decoder, amplifier, etc. Furthermore, the audio processor 1005 is also coupled to a central processing unit 1001, enabling on-device recording via the microphone 1010 and on-device playback of stored sound via the speaker 1009.

[0251] This invention also provides a computer-readable storage medium storing a computer program that, when executed by a processor, implements the above-described method for identifying periodic rail scratches.

[0252] This invention also provides a computer program product, which includes a computer program that, when executed by a processor, implements the above-described method for identifying periodic rail scratches.

[0253] In this embodiment of the invention, the signal envelope of the vehicle axle box vibration acceleration signal is determined based on the vehicle dynamic response data of the rail to be tested. The signal envelope is then filtered to remove the trend term, yielding the target signal. The target signal is decomposed using the CEEMDAN algorithm to obtain its multi-order intrinsic mode components. Based on these multi-order intrinsic mode components, the reconstructed signal of the vehicle dynamic response signal is determined. CWD analysis is performed on the reconstructed signal to determine whether the rail to be tested exhibits periodic abrasion. Compared with existing technologies that diagnose rail periodic abrasion based on machine vision and image processing, this invention, using the CEEMDAN algorithm, can determine the multi-order intrinsic mode components and thus the reconstructed signal. This enables data feature mining of rail periodic abrasion sections and, through CWD analysis of the reconstructed signal, effectively determines whether the rail to be tested exhibits periodic abrasion, achieving effective identification of rail periodic abrasion. This solves the problem of existing technologies being unable to identify rail periodic abrasion due to the lack of a track image acquisition system, improving the accuracy and efficiency of rail periodic abrasion identification and providing scientific guidance for rail maintenance.

[0254] Those skilled in the art will understand that embodiments of the present invention can be provided as methods, systems, or computer program products. Therefore, the present invention can take the form of a completely hardware embodiment, a completely software embodiment, or an embodiment combining software and hardware aspects. Furthermore, the present invention can take the form of a computer program product embodied on one or more computer-usable storage media (including, but not limited to, disk storage, CD-ROM, optical storage, etc.) containing computer-usable program code.

[0255] This invention is described with reference to flowchart illustrations and / or block diagrams of methods, apparatus (systems), and computer program products according to embodiments of the invention. It will be understood that each block of the flowchart illustrations and / or block diagrams, and combinations of blocks in the flowchart illustrations and / or block diagrams, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, special-purpose computer, embedded processor, or other programmable data processing apparatus to produce a machine, such that the instructions, which execute via the processor of the computer or other programmable data processing apparatus, generate instructions for implementing the flowchart illustrations and / or block diagrams. Figure 1 One or more processes and / or boxes Figure 1 A device that provides the functions specified in one or more boxes.

[0256] These computer program instructions may also be stored in a computer-readable storage medium that can direct a computer or other programmable data processing device to function in a particular manner, such that the instructions stored in the computer-readable storage medium produce an article of manufacture including instruction means, which are implemented in a process Figure 1 One or more processes and / or boxes Figure 1 The function specified in one or more boxes.

[0257] These computer program instructions may also be loaded onto a computer or other programmable data processing equipment to cause a series of operational steps to be performed on the computer or other programmable equipment to produce a computer-implemented process, thereby providing instructions that execute on the computer or other programmable equipment for implementing the process. Figure 1 One or more processes and / or boxes Figure 1 The steps of the function specified in one or more boxes.

[0258] The specific embodiments described above further illustrate the purpose, technical solution, and beneficial effects of the present invention. It should be understood that the above descriptions are merely specific embodiments of the present invention and are not intended to limit the scope of protection of the present invention. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of the present invention should be included within the scope of protection of the present invention.

Claims

1. A method for identifying periodic scratches on rails, characterized in that, include: Based on the vehicle dynamic response data of the rail to be tested, determine the signal envelope of the vehicle axle box vibration acceleration signal; The target signal is obtained by filtering out the trend term from the signal envelope. The target signal is decomposed using the CEEMDAN algorithm to obtain its multi-order intrinsic mode components. Based on the multi-order intrinsic mode components, the reconstructed signal of the vehicle dynamic response signal is determined; CWD analysis was performed on the reconstructed signal to determine whether there were periodic scratches on the rail under test; CWD analysis was performed on the reconstructed signal to determine whether periodic scratches existed on the rail under inspection, including: Perform CWD analysis on the reconstructed signal and calculate the CWD distribution signal of the reconstructed signal; Calculate the frequency marginal spectrum and full-frequency time marginal spectrum of the CWD distributed signal; Extract the periodic scratch frequency components from the frequency marginal spectrum; Calculate the time margin spectrum of the periodic abrasion frequency components based on the periodic abrasion frequency components and the full-frequency time margin spectrum; Based on the time marginal spectrum of the frequency components of periodic abrasion, it is determined whether the rail to be inspected has periodic abrasion.

2. The method as described in claim 1, characterized in that, The vehicle dynamic response data includes axle box vibration acceleration detection data; Based on the vehicle dynamic response data of the rail to be tested, the signal envelope of the axle box vibration signal is determined, including: The vibration acceleration signal of the axle box is determined based on the vibration acceleration detection data of the axle box of the rail to be tested. The Hilbert transform is performed on the axle box vibration acceleration signal to obtain the signal envelope of the axle box vibration acceleration signal.

3. The method as described in claim 1, characterized in that, The target signal is decomposed using the CEEMDAN algorithm to obtain the multi-order intrinsic mode components of the signal envelope, including: Gaussian white noise is added to the target signal multiple times to obtain the processed signal; The processed signal is decomposed into EMD components to obtain the first-order EMD eigenmode components. The average value of the first-order EMD eigenmode components is taken as the first-order eigenmode component of the signal envelope using the CEEMDAN algorithm. The target signal with the first-order intrinsic mode components removed is used as the residual signal; Gaussian white noise is added multiple times to the residual signal to obtain the updated signal; EMD decomposition is performed on the updated signal to obtain the first-order EMD intrinsic mode components; the average value of the first-order EMD intrinsic mode components of the updated signal is taken as the second-order intrinsic mode component of the signal envelope; the residual signal after removing the second-order intrinsic mode components is taken as the updated residual signal; the above steps of determining the updated signal and the updated residual signal are repeated until the obtained updated residual signal is a monotonic function. Obtain the multi-order intrinsic mode components of the signal envelope obtained in the above steps.

4. The method as described in claim 1, characterized in that, Based on the multi-order intrinsic mode components, the reconstructed signal of the axle box vibration signal is determined, including: The target signal and each of its intrinsic mode components are subjected to Hilbert transform to obtain the analytical form of the target signal and each of its intrinsic mode components. Based on the analytical form of the target signal and each intrinsic mode component, calculate the instantaneous amplitude of the target signal and each intrinsic mode component for each sampling point. The reconstructed signal of the axle box vibration signal is determined based on the target signal and the instantaneous amplitude of each intrinsic mode component at each sampling point.

5. The method as described in claim 4, characterized in that, Based on the target signal and the instantaneous amplitude of each intrinsic mode component at each sampling point, the reconstructed signal of the axle box vibration signal is determined, including: For each intrinsic mode component, the instantaneous amplitude at each sampling point is calculated as follows: the ratio of the instantaneous amplitude of the intrinsic mode component at that sampling point to the instantaneous amplitude of the target signal at that sampling point is used as the first data of the intrinsic mode component at that sampling point; the average value of the first data of the intrinsic mode component at each sampling point is calculated. The eigenmode components whose average value is greater than the first preset value are taken as signal reconstruction components; The reconstructed components of the signal are added together to obtain the reconstructed signal of the axle box vibration signal.

6. The method as described in claim 1, characterized in that, From the frequency marginal spectrum, the periodic scratch frequency components are extracted, including: Determine the maxima of different maxima of the frequency marginal spectrum, and the frequency values ​​corresponding to different maxima; For each maximum value point, if the maximum value of the maximum value point is greater than the second preset value, and the remainder obtained by dividing the frequency value of the maximum value point by the preset frequency value of the rail periodic abrasion is less than the third preset value, the maximum value point is determined as the target maximum value point. The frequency value of the target maximum point is used as the periodic abrasion frequency component.

7. The method as described in claim 1, characterized in that, Based on the periodic abrasion frequency components and the full-frequency time marginal spectrum, the time marginal spectrum of the periodic abrasion frequency components is calculated, including: Calculate the sum of the time margin spectra of all periodic abrasion frequency components within a preset frequency range; The sum of the time marginal spectra is subjected to low-pass filtering to obtain the trend term of the sum of the time marginal spectra; When the trend term of the sum of the time marginal spectra is greater than a fourth preset value, the ratio of the sum of the time marginal spectra to the frequency marginal spectra is taken as the energy percentage. When the energy percentage is greater than the fifth preset value, it is determined that the rail to be tested has periodic abrasions.

8. A rail periodic scratch identification device, characterized in that, include: The signal envelope determination module is used to determine the signal envelope of the vehicle axle box vibration acceleration signal based on the vehicle dynamic response data of the rail to be tested. The trend removal processing module is used to remove the trend term from the signal envelope to obtain the target signal. The CEEMDAN decomposition module is used to decompose the target signal using the CEEMDAN algorithm to obtain the multi-order intrinsic mode components of the target signal. The reconstruction module is used to determine the reconstructed signal of the vehicle dynamic response signal based on the multi-order intrinsic mode components. The CWD analysis module is used to perform CWD analysis on the reconstructed signal to determine whether there are periodic scratches on the rail to be inspected. The CWD analysis module is specifically used for: Perform CWD analysis on the reconstructed signal and calculate the CWD distribution signal of the reconstructed signal; Calculate the frequency marginal spectrum and full-frequency time marginal spectrum of the CWD distributed signal; Extract the periodic scratch frequency components from the frequency marginal spectrum; Calculate the time margin spectrum of the periodic abrasion frequency components based on the periodic abrasion frequency components and the full-frequency time margin spectrum; Based on the time marginal spectrum of the frequency components of periodic abrasion, it is determined whether the rail to be inspected has periodic abrasion.

9. The apparatus as claimed in claim 8, characterized in that, The vehicle dynamic response data includes axle box vibration acceleration detection data; The signal envelope determination module is specifically used for: The vibration acceleration signal of the axle box is determined based on the vibration acceleration detection data of the axle box of the rail to be tested. The Hilbert transform is performed on the axle box vibration acceleration signal to obtain the signal envelope of the axle box vibration acceleration signal.

10. The apparatus as claimed in claim 8, characterized in that, The CEEMDAN decomposition module is specifically used for: Gaussian white noise is added to the target signal multiple times to obtain the processed signal; The processed signal is decomposed into EMD components to obtain the first-order EMD eigenmode components. The average value of the first-order EMD eigenmode components is taken as the first-order eigenmode component of the signal envelope using the CEEMDAN algorithm. The target signal with the first-order intrinsic mode components removed is used as the residual signal; Gaussian white noise is added multiple times to the residual signal to obtain the updated processed signal; The updated signal is decomposed using EMD to obtain the first-order EMD intrinsic mode components. The average value of the first-order EMD intrinsic mode components of the updated signal is taken as the second-order intrinsic mode component of the signal envelope. The residual signal after removing the second-order intrinsic mode components is taken as the updated residual signal. The above steps of determining the updated signal and the updated residual signal are repeated until the obtained updated residual signal is a monotonic function. Obtain the multi-order intrinsic mode components of the signal envelope obtained in the above steps.

11. The apparatus as claimed in claim 8, characterized in that, The refactoring module is specifically used for: The target signal and each of its intrinsic mode components are subjected to Hilbert transform to obtain the analytical form of the target signal and each of its intrinsic mode components. Based on the analytical form of the target signal and each intrinsic mode component, calculate the instantaneous amplitude of the target signal and each intrinsic mode component for each sampling point. The reconstructed signal of the axle box vibration signal is determined based on the target signal and the instantaneous amplitude of each intrinsic mode component at each sampling point.

12. The apparatus as claimed in claim 11, characterized in that, The refactoring module is specifically used for: For each intrinsic mode component, the instantaneous amplitude at each sampling point will be calculated as the ratio of the instantaneous amplitude of the intrinsic mode component at that sampling point to the instantaneous amplitude of the target signal at that sampling point, and this ratio will be used as the first data of the intrinsic mode component at that sampling point. Calculate the average value of the first data for each sampling point for the eigenmode component of that order; The eigenmode components whose average value is greater than the first preset value are taken as signal reconstruction components; The reconstructed components of the signal are added together to obtain the reconstructed signal of the axle box vibration signal.

13. The apparatus as claimed in claim 8, characterized in that, The CWD analysis module is specifically used for: Determine the maxima of different maxima of the frequency marginal spectrum, and the frequency values ​​corresponding to different maxima; For each maximum value point, if the maximum value of the maximum value point is greater than the second preset value, and the remainder obtained by dividing the frequency value of the maximum value point by the preset frequency value of the rail periodic abrasion is less than the third preset value, the maximum value point is determined as the target maximum value point. The frequency value of the target maximum point is used as the periodic abrasion frequency component.

14. The apparatus as claimed in claim 8, characterized in that, The CWD analysis module is specifically used for: Calculate the sum of the time margin spectra of all periodic abrasion frequency components within a preset frequency range; The sum of the time marginal spectra is subjected to low-pass filtering to obtain the trend term of the sum of the time marginal spectra; When the trend term of the sum of the time marginal spectra is greater than a fourth preset value, the ratio of the sum of the time marginal spectra to the frequency marginal spectra is taken as the energy percentage. When the energy percentage is greater than the fifth preset value, it is determined that the rail to be tested has periodic abrasions.

15. A computer device comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, characterized in that, When the processor executes the computer program, it implements the method of any one of claims 1 to 7.

16. A computer-readable storage medium, characterized in that, The computer-readable storage medium stores a computer program that, when executed by a processor, implements the method of any one of claims 1 to 7.

17. A computer program product, characterized in that, The computer program product includes a computer program that, when executed by a processor, implements the method of any one of claims 1 to 7.

Images