Method and device for extracting weak characteristics of wheel-rail force electromagnetic interference of high-speed railway

Abstract

The invention discloses a method and a device for extracting weak characteristics of wheel-rail force electromagnetic interference of a high-speed railway, which can be used in the technical field of railway engineering and safety, and the method comprises the following steps: determining an interference area according to a wheel-rail force signal of the high-speed railway, and carrying out modal decomposition on the wheel-rail force signal of the interference area to obtain a plurality of connotation modal components; calculating the power spectral density of each connotation mode component, and determining the trend term of the wheel-rail force signal according to the power spectral density of each connotation mode component; performing modal decomposition on the wheel-rail force signal after the trend term is filtered out to obtain a plurality of connotation modal components without interference terms; calculating the kurtosis value of the connotation mode component of each non-interference item, and determining the non-impact item of the wheel-rail force signal; through short-time Fourier transform, a moving time window edge spectrum is obtained, and weak electromagnetic interference characteristics are extracted from wheel-rail force signals. According to the method, the high-speed railway wheel-rail force electromagnetic interference can be accurately identified, and reliable data support is provided for railway operation.

Description

Technical Field

[0001] The present invention relates to the technical field of railway engineering and safety, and particularly to a method and device for extracting weak characteristics of electromagnetic interference of wheel-rail forces in high-speed railways. Background Art

[0002] This section aims to provide background or context for the embodiments of the present invention. The descriptions herein are not admitted to be prior art just because they are included in this section.

[0003] Railway transportation is a major mode of transportation and plays a very important role in promoting economic development. The state of the track directly affects the safe and comfortable operation of the vehicle-track system. With the continuous increase in the operating mileage of high-speed railways, the identification and state prediction of track diseases have become key issues that need to be solved urgently. The detection of wheel-rail forces is an effective means for diagnosing track diseases. When a train passes through a phase separation section, transient electromagnetic interference will be generated on the wheel-rail forces. If the interference is not filtered, evaluation indexes calculated from the wheel-rail forces, such as derailment coefficient, load reduction rate, effective value, etc., are prone to false over-limit alarms.

[0004] Although a lot of research work has been carried out in the prior art on extracting the characteristics of interference signals, both the abnormal values and normal values of the wheel-rail force signals in high-speed railways have impact characteristics. There is still a lack of corresponding solutions on how to propose effective identification and filtering methods for the internal characteristics of the abnormal values of the wheel-rail force signals in high-speed railways. Summary of the Invention

[0005] An embodiment of the present invention provides a method for extracting weak characteristics of electromagnetic interference of wheel-rail forces in high-speed railways, which is used to accurately identify the electromagnetic interference of wheel-rail forces in high-speed railways and provide reliable data support for railway operation. The method includes:

[0006] Collect the wheel-rail force signals of high-speed railways, and determine the abnormal positions of the wheel-rail force signals according to the wheel-rail force signals of high-speed railways;

[0007] Determine the interference regions according to the abnormal positions, and perform modal decomposition on the wheel-rail force signals in the interference regions to obtain a plurality of intrinsic mode components;

[0008] Calculate the power spectral density of each intrinsic mode component, and determine the trend term of the wheel-rail force signal according to the power spectral density of each intrinsic mode component;

[0009] Perform modal decomposition on the wheel-rail force signal after filtering the trend term to obtain a plurality of intrinsic mode components without interference terms;

[0010] Calculate the kurtosis value of each intrinsic mode component without interference terms, and determine the non-impact term of the wheel-rail force signal according to the kurtosis value of each intrinsic mode component without interference terms;

[0011] By performing short-time Fourier transform on the wheel-rail force signal after filtering out non-impulse terms, a moving time-window marginal spectrum is obtained. Based on the moving time-window marginal spectrum, weak electromagnetic interference features are extracted from the wheel-rail force signal.

[0012] An embodiment of the present invention further provides a device for extracting weak electromagnetic interference features of high-speed railway wheel-rail forces, which is used to accurately identify the electromagnetic interference of high-speed railway wheel-rail forces and provide reliable data support for railway operation. The device includes:

[0013] An acquisition module, which is used to acquire the wheel-rail force signal of a high-speed railway and determine the abnormal position of the wheel-rail force signal according to the wheel-rail force signal of the high-speed railway;

[0014] A modal decomposition module, which is used to determine the interference area according to the abnormal position, perform modal decomposition on the wheel-rail force signal in the interference area, and obtain multiple intrinsic mode components;

[0015] A first calculation module, which is used to calculate the power spectral density of each intrinsic mode component and determine the trend term of the wheel-rail force signal according to the power spectral density of each intrinsic mode component;

[0016] A secondary decomposition module, which is used to perform modal decomposition on the wheel-rail force signal after filtering out the trend term to obtain multiple intrinsic mode components without interference terms;

[0017] A second calculation module, which is used to calculate the kurtosis value of each intrinsic mode component without interference terms and determine the non-impulse terms of the wheel-rail force signal according to the kurtosis value of each intrinsic mode component without interference terms;

[0018] An extraction module, which is used to perform short-time Fourier transform on the wheel-rail force signal after filtering out non-impulse terms to obtain a moving time-window marginal spectrum, and extract weak electromagnetic interference features from the wheel-rail force signal according to the moving time-window marginal spectrum.

[0019] An embodiment of the present invention further provides a computer device, including a memory, a processor, and a computer program stored on the memory and executable on the processor. When the processor executes the computer program, the above method for extracting weak electromagnetic interference features of high-speed railway wheel-rail forces is implemented.

[0020] An embodiment of the present invention further provides a computer-readable storage medium, which stores a computer program. When the computer program is executed by a processor, the above method for extracting weak electromagnetic interference features of high-speed railway wheel-rail forces is implemented.

[0021] An embodiment of the present invention further provides a computer program product, which includes a computer program. When the computer program is executed by a processor, the above method for extracting weak electromagnetic interference features of high-speed railway wheel-rail forces is implemented.

[0022] In an embodiment of the present invention, the wheel-rail force signal of high-speed railway is collected. According to the wheel-rail force signal of high-speed railway, the abnormal position of the wheel-rail force signal is determined; according to the abnormal position, the interference area is determined, and the wheel-rail force signal in the interference area is subjected to modal decomposition to obtain a plurality of intrinsic mode components; the power spectral density of each intrinsic mode component is calculated, and according to the power spectral density of each intrinsic mode component, the trend term of the wheel-rail force signal is determined; the wheel-rail force signal after filtering the trend term is subjected to modal decomposition to obtain a plurality of intrinsic mode components without interference terms; the kurtosis value of each intrinsic mode component without interference terms is calculated, and according to the kurtosis value of each intrinsic mode component without interference terms, the non-impulse term of the wheel-rail force signal is determined; by performing short-time Fourier transform on the wheel-rail force signal after filtering the non-impulse term, the moving time window marginal spectrum is obtained, and according to the moving time window marginal spectrum, the weak electromagnetic interference feature is extracted from the wheel-rail force signal. In this way, the wheel-rail force signal is decomposed, the average frequency algorithm of power spectral density is used to judge the trend term; the kurtosis index is used to diagnose the non-impulse term; the moving time window marginal spectrum method is proposed to identify the weak feature signal of electromagnetic interference, accurately identify the weak electromagnetic interference of the wheel-rail force of high-speed railway, and provide reliable data support for railway operation. BRIEF DESCRIPTION OF THE DRAWINGS

[0023] In order to more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the following will briefly introduce the drawings required for the description of the embodiments or the prior art. Obviously, the following drawings are only some embodiments of the present invention. For those of ordinary skill in the art, without creative efforts, other drawings can be obtained according to these drawings. In the drawings:

[0024] Figure 1 It is a flowchart of the method for extracting weak electromagnetic interference features of wheel-rail force of high-speed railway provided in the embodiment of the present invention;

[0025] Figure 2 It is a schematic diagram of the wheel-rail force signal provided in the embodiment of the present invention;

[0026] Figure 3 It is a schematic diagram of a specific intrinsic mode component and its power spectrum provided in the embodiment of the present invention;

[0027] Figure 4 It is a schematic diagram of the original signal, the signal after removing the trend term, the impact signal and their STFT results provided in the embodiment of the present invention;

[0028] Figure 5 It is a schematic diagram of the moving time window - marginal spectrum provided in the embodiment of the present invention;

[0029] Figure 6 It is a schematic diagram of the comparison between the original signal and the signal after removing electromagnetic interference provided in the embodiment of the present invention;

[0030] Figure 7 It is a schematic diagram of the device for extracting weak characteristics of electromagnetic interference of wheel-rail forces provided in the embodiments of the present invention;

[0031] Figure 8 It is a structural block diagram of the electronic device provided in the embodiments of the present invention. Specific embodiments

[0032] To make the objectives, technical solutions and advantages of the embodiments of the present invention clearer and more understandable, the following further describes the embodiments of the present invention in detail with reference to the accompanying drawings. Herein, the illustrative embodiments of the present invention and their descriptions are used to explain the present invention, but not to limit the present invention.

[0033] The term "and / or" in this article only describes an association relationship and means that three relationships can exist. For example, A and / or B can represent three situations: A exists alone, A and B exist simultaneously, and B exists alone. In addition, the term "at least one" in this article means any one of a plurality or any combination of at least two of a plurality. For example, including at least one of A, B, and C can mean including any one or more elements selected from the set composed of A, B, and C.

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

[0035] In order to effectively filter out the electromagnetic interference of wheel-rail forces, it is necessary to identify the weak characteristics of electromagnetic interference. The weak characteristics of electromagnetic interference refer to the abnormal characteristics that exist in the marginal spectrum of the moving time window for some weak impacts after the trend term and non-impulse term are excluded from the wheel-rail force signal. Figure 5 It is a schematic diagram of the moving time window - marginal spectrum provided in the embodiments of the present invention. As Figure 5 shown, the circles on the moving time window - marginal spectrum indicate the positions where the weak characteristics of electromagnetic interference of the wheel-rail force signal appear. The energy of the moving time window - marginal spectrum represents the abnormal degree of the weak characteristics, and only the impacts with energy greater than the threshold will be identified as electromagnetic interference.

[0036] Abnormal data processing is a relatively common problem and is often encountered in the fields of machine learning, sound signal processing, and image processing. In the past decade or so, many scholars have conducted in-depth research on this and proposed a large number of feasible methods. The existing technologies respectively summarize different methods in the field of outlier detection research. The outlier detection methods are classified into the following four categories: detection methods based on statistical prediction, detection methods based on distance, detection methods based on density, and detection methods based on machine learning. The detection methods based on statistical prediction may cause large deviations in the results because they need to first assume that the detected data conforms to a certain type of distribution. The detection methods based on distance do not require the distribution type of the data, but due to the need to calculate the distances between data points, missed detection phenomena will occur when the data density changes. The detection methods based on density can complete outlier detection only by considering the local data distribution around the data objects. The detection methods based on machine learning need to establish a prediction model in advance. If the accuracy of the prediction model cannot be guaranteed, it will lead to incorrect outlier detection results. According to the characteristics of the sudden change of the burr outliers of the geometric irregularity of the high-speed railway track, the existing technology designs an improved fuzzy burr removal method, defines the points with suddenly increased or decreased differences as the starting points of the burr outliers, defines the points with suddenly reversed differences close to the starting points as the ending points of the burr outliers, selects the two endpoints of the smallest interval containing the burr outliers as the interpolation points, and uses their approximate linear interpolation to replace the original burr outliers.

[0037] Compared with the above research, the present invention focuses more on the problem of electromagnetic interference during the acquisition of the high-speed railway wheel-rail force waveform. For this problem, an adaptive time-frequency analysis technology is used to explore the weak feature extraction method of the electromagnetic interference of the high-speed railway wheel-rail force.

[0038] Based on this, an embodiment of the present invention provides a method for extracting weak features of electromagnetic interference of high-speed railway wheel-rail force, as Figure 1 shown, including:

[0039] Step 101: Collect the wheel-rail force signal of the high-speed railway, and determine the abnormal position of the wheel-rail force signal according to the wheel-rail force signal of the high-speed railway;

[0040] Step 102: Determine the interference region according to the abnormal position, and perform modal decomposition on the wheel-rail force signal in the interference region to obtain a plurality of intrinsic mode components;

[0041] Step 103: Calculate the power spectral density of each intrinsic mode component, and determine the trend term of the wheel-rail force signal according to the power spectral density of each intrinsic mode component;

[0042] Step 104: Perform modal decomposition on the wheel-rail force signal after filtering out the trend term to obtain a plurality of intrinsic mode components without interference terms;

[0043] Step 105: Calculate the kurtosis values of the intrinsic mode components of each interference-free item, and determine the non-impulse items of the wheel-rail force signal according to the kurtosis values of the intrinsic mode components of each interference-free item;

[0044] Step 106: Perform short-time Fourier transform on the wheel-rail force signal after filtering out the non-impulse items to obtain the moving time-window marginal spectrum, and extract the weak electromagnetic interference features from the wheel-rail force signal according to the moving time-window marginal spectrum.

[0045] The method for extracting weak electromagnetic interference features of high-speed railway wheel-rail force proposed in the embodiment of the present invention decomposes the wheel-rail force signal by using the adaptive chirp mode decomposition (ACMD) algorithm; uses the average frequency algorithm of power spectral density to judge the trend items; uses the kurtosis index to diagnose the non-impulse items; and proposes a moving time-window marginal spectrum method to identify the weak feature signals of electromagnetic interference.

[0046] The peak value of the wheel-rail force reflects the severity of the wheel-rail interaction. Evaluation indexes calculated from the wheel-rail force, such as derailment coefficient, axle lateral force, effective value, etc., are often used to evaluate the state of the line. If the high-speed railway wheel-rail force contains abnormal values caused by electromagnetic interference, directly calculating the evaluation indexes may lead to misjudgment.

[0047] Figure 2 It is a schematic diagram of the wheel-rail force signal provided in the embodiment of the present invention. The measured wheel-rail force is as Figure 2 shown. This signal contains abnormal values caused by electromagnetic interference and cannot be directly extracted. In order to effectively and conveniently extract and filter out the abnormal values caused by electromagnetic interference, the embodiment of the present invention proposes to determine the weak electromagnetic interference feature region based on the moving time-window - marginal spectrum of the short-time Fourier transform, and replace it with a Gaussian random signal.

[0048] In one embodiment, determining the abnormal position of the wheel-rail force signal according to the wheel-rail force signal of the high-speed railway includes:

[0049] Determine the impact positions where the extreme values in the wheel-rail force signal of the high-speed railway reach a preset value;

[0050] Determine the abnormal position of the wheel-rail force signal according to the impact positions where the extreme values in the wheel-rail force signal reach a preset value.

[0051] In one embodiment, determining the interference region according to the abnormal position includes:

[0052] Determine the start point and end point of the interference signal according to the abnormal position;

[0053] Determine the interference region of the wheel-rail force signal according to the start point and end point of the interference signal.

[0054] In one embodiment, modal decomposition of the wheel-rail force signal in the interference region includes:

[0055] Perform modal decomposition on the wheel-rail force signal in the interference region using the adaptive chirp modal decomposition algorithm.

[0056] Specifically, the ACMD algorithm is a technique for signal processing, specifically for analyzing non-linear and non-stationary signals. It decomposes the signal into a set of intrinsic mode components with physical significance, which can effectively capture the instantaneous frequency and amplitude changes of the signal.

[0057] Decompose the wheel-rail force signal using the ACMD algorithm to obtain the intrinsic mode components. Then calculate the power spectral density and its average frequency of the intrinsic mode components. Use the average frequency to identify the trend term caused by the static wheel load of the train.

[0058] Specifically, the detailed algorithm description is as follows:

[0059] (1) Denote the wheel-rail force signal as {F i , i = 1, 2, … N}, where F i is the wheel-rail force, N is the number of sampling points, and the sampling frequency is Fs, with a value of 5000;

[0060] (2) Find the impacts in the high-speed railway wheel-rail force signal that meet the condition of F i ≥ 200 kN as the abnormally large values, that is, the aforementioned abnormal positions, and record the positions w x of the abnormally large values;

[0061] (3) According to the position w x of the abnormally large value, demarcate the interference region with a length of [w ps , w pe , where the starting point w ps of the interference region = w x - 1000, the ending point w pe of the interference signal = w c + 1000. Denote the interference region signal as {p i , i = 1, …, N p}, N p = w pe - w ps + 1;

[0062] (4) Perform modal decomposition on the interference region signal {p i , i = 1, …, N p} using the ACMD algorithm, and the signal is decomposed into

[0063]

[0064] In the above formula, K represents the number of modes, A j，iis the instantaneous amplitude (IA), f j，i is the instantaneous frequency (IF) of the signal, is the initial phase, q j，i is the intrinsic mode component signal;

[0065] (5) Calculate the power spectral density (PSD) {Pxx j，i , i = 1, …, N p} of each intrinsic mode component {q j，i , i = 1, …, N p};

[0066] (6) Calculate the average frequency of the power spectral density of each intrinsic mode component as

[0067]

[0068] In the above formula, N p is the number of signal points in the interference region, Pxx j，i is the power spectral density, ω i is the frequency.

[0069] (7) Select the components that meet the condition and define them as the trend terms, and remove these components.

[0070] In one example, the wheel-rail force signal in the interference region is decomposed by the ACMD algorithm to obtain several intrinsic mode functions (IMFs). Then, the average frequency algorithm of the power spectral density is used to judge the trend terms, and the kurtosis index is used to diagnose the non-impulse terms. The results are as Figure 3 shown,[[]]END]] Figure 3 is a schematic diagram of the specific intrinsic mode components and their power spectra provided in the embodiment of the present invention, where Figure 3 in (a), Figure 3 in (c), Figure 3 in (e) respectively show the trend terms, non-impulse terms, and impulse terms. The characteristics of the above three types of intrinsic mode components are observed as follows: the average value of the trend terms is significantly higher than zero, and there is only a low-frequency component close to zero on its power spectrum; the power density of the non-impulse terms is very small, and there is no impact characteristic, and the signal has a long-term stable oscillation; the impact amplitude of the impulse terms is basically the same as that of the original signal, and the power density is large. It can be seen that the algorithm in the embodiment of the present invention can effectively identify the trend terms, non-impulse terms, and impulse terms in the intrinsic mode components.

[0071] In one embodiment, the kurtosis values of the intrinsic mode components of each interference-free term are calculated, and according to the kurtosis values of the intrinsic mode components of each interference-free term, the non-impulse term of the wheel-rail force signal is determined, including:

[0072] Calculate the kurtosis values of the intrinsic mode components of each interference-free term;

[0073] Sort the kurtosis values of the intrinsic mode components of each interference-free term from small to large, and select a preset number of kurtosis values in order from the sorted kurtosis values as the target kurtosis values;

[0074] Determine the components in the wheel-rail force signal corresponding to the target kurtosis values as the non-impulse term of the wheel-rail force signal.

[0075] Specifically, kurtosis is an index that measures the sensitivity of a distribution to outliers. Components with low kurtosis are not sensitive to impulses in the signal, and the energy is relatively dispersed. Excluding components with lower kurtosis helps to observe the vertical ridge line of the signal in the later stage.

[0076] To ensure complete exclusion of non-impulse terms, select and remove the 7 components with the smallest peaks.

[0077] Use the ACMD algorithm to decompose the wheel-rail force signal after removing the trend term to obtain the intrinsic mode components. Then calculate the kurtosis index of the intrinsic mode components. Use the kurtosis index to diagnose non-impulse components. The detailed algorithm description is as follows:

[0078] (1) Decompose the wheel-rail force signal after removing the trend term to obtain the intrinsic mode components, denoted as {x j，i , i = 1,..., N p};

[0079] (2) Calculate the kurtosis index of each intrinsic mode component

[0080]

[0081] In the above formula, N p is the number of signal points in the interference area, x j，i is the intrinsic mode component of the wheel-rail force signal after removing the trend term, is the mean value of the intrinsic mode component.

[0082] (3) Sort the peak indexes of the intrinsic mode components from small to large;

[0083] (4) Find the components corresponding to the first 7 peak indexes after sorting as the non-impulse terms and filter them out.

[0084] In one embodiment, by performing a short-time Fourier transform on the wheel-rail force signal after filtering out the non-impulse term, a moving time-window marginal spectrum is obtained, and according to the moving time-window marginal spectrum, weak electromagnetic interference features are extracted from the wheel-rail force signal, including:

[0085] After filtering out the non-impulse terms from the wheel-rail force signal with the trend terms removed, an impact signal is obtained;

[0086] After performing a short-time Fourier transform on the impact signal, calculate the time marginal spectrum, determine the moving average value of the time marginal spectrum, and obtain the moving time window marginal spectrum;

[0087] Determine the weak characteristic electromagnetic interference points based on the moving time window marginal spectrum, and extract the weak characteristics of electromagnetic interference from the wheel-rail force signal.

[0088] In one embodiment, after extracting the weak characteristics of electromagnetic interference, it further includes:

[0089] Replace the region after extracting the weak characteristics of electromagnetic interference with Gaussian random signals.

[0090] During specific implementation, filter the interference region signal {p i , i = 1,..., N p} for trend terms and non-impulse terms, and denote the obtained signal as the impact signal {u i , i = 1,..., N p}. First, perform a short-time Fourier transform on the impact signal to obtain the time-frequency distribution, then calculate the moving time window - marginal spectrum, and use its maximum value to extract the weak characteristics of electromagnetic interference. The detailed algorithm description is as follows:

[0091] (1) Perform a short-time Fourier transform on the impact signal {u i , i = 1,..., N p}, and the calculation formula is as follows:

[0092]

[0093] In the above formula, w(m) is the Gaussian window function, and N represents the window function length;

[0094] (2) Calculate the time marginal spectrum S m

[0095]

[0096] (3) Calculate the moving average value of the time marginal spectrum S m to obtain the moving time window - marginal spectrum, as Figure 4 shown,

[0097]

[0098] In the above formula, w p represents that the window length is taken as 10;

[0099] (4) Calculate the maximum value E V (i) of Emax = max(E v (i));

[0100] (5) Calculate the maximum value of E v (i), {E e (i), i = 1, …, E p}}, all the maximum value points that satisfy the condition E e (i) > 2% E max are weak characteristic electromagnetic interference points, denoted as R;

[0101] (6) Replace the points R ± 10 of the weak characteristic electromagnetic interference points with Gaussian random signals.

[0102] In one example, the signal after removing the trend term and non-impulse term from the original signal is defined as the impulse signal. Figure 4 This is a schematic diagram of the original signal, the signal after removing the trend term, the impulse signal and their STFT results provided in the embodiments of the present invention. The time-frequency diagrams of the short-time Fourier transform of the original signal, the signal after removing the trend term, and the impulse signal are respectively as shown in Figure 4 (b) in Figure 4 (d) in Figure 4 (f) in. In order to distinguish the outliers caused by electromagnetic interference, it is necessary to extract the weak characteristics of the electromagnetic interference. As can be seen from Figure 4 (b), most of the energy of the original signal is concentrated in the low-frequency trend term, and it is difficult to identify the impulse characteristics caused by electromagnetic interference. As can be seen from Figure 4 (d), the time-frequency diagram of only the signal after removing the trend term has relatively obvious vertical ridge lines, and there are also horizontal ridge line interferences. As can be seen from Figure 4 (f), the impulse signal retains the vast majority of the energy of the vertical ridge lines, and at the same time, each impulse characteristic is clearly visible. The moving time window-edge spectrum calculated using Figure 4 (f) is as shown in Figure 5 . Figure 5 This is a schematic diagram of the moving time window-edge spectrum provided in the embodiments of the present invention. It can be seen that the moving time window-edge spectrum can accurately identify the weak characteristics of electromagnetic interference. Replace the signal in the weak characteristic area of electromagnetic interference with a Gaussian random signal to obtain clean wheel-rail force data.

[0103] Figure 6 This is a schematic diagram of the comparison between the original signal and the signal after removing electromagnetic interference provided in the embodiments of the present invention. The original signal and the impulse signal after filtering out electromagnetic interference are respectively as shown in Figure 6 (a) in Figure 6 (b) in. As can be seen from Figure 6As can be seen from Figure (b), the impact signal after filtering out electromagnetic interference still has impact characteristics caused by short-wave irregularities of the track such as joints and abrasions, which belong to normal values. The method described in the embodiment of the present invention will not misjudge this impact characteristic as an abnormal value caused by electromagnetic interference, but effectively distinguishes normal values and abnormal values caused by weak electromagnetic interference features according to the moving time window - marginal spectrum, providing more reliable data support for track maintenance work.

[0104] An apparatus for extracting weak characteristics of electromagnetic interference in high-speed railway wheel-rail force is also provided in an embodiment of the present invention, as described in the following embodiments. Since the principle of solving problems by this apparatus is similar to that of the method for extracting weak characteristics of electromagnetic interference in high-speed railway wheel-rail force, the implementation of this apparatus can refer to the implementation of the method for extracting weak characteristics of electromagnetic interference in high-speed railway wheel-rail force, and the repeated parts will not be elaborated.

[0105] Figure 7 is a schematic diagram of the apparatus for extracting weak characteristics of electromagnetic interference in high-speed railway wheel-rail force provided in an embodiment of the present invention, as Figure 7 shown, the apparatus includes:

[0106] An acquisition module 701, configured to acquire the wheel-rail force signal of a high-speed railway, and determine the abnormal position of the wheel-rail force signal according to the wheel-rail force signal of the high-speed railway;

[0107] A modal decomposition module 702, configured to determine an interference region according to the abnormal position, perform modal decomposition on the wheel-rail force signal in the interference region, and obtain a plurality of intrinsic mode components;

[0108] A first calculation module 703, configured to calculate the power spectral density of each intrinsic mode component, and determine the trend term of the wheel-rail force signal according to the power spectral density of each intrinsic mode component;

[0109] A secondary decomposition module 704, configured to perform modal decomposition on the wheel-rail force signal after filtering out the trend term, and obtain a plurality of intrinsic mode components without interference terms;

[0110] A second calculation module 705, configured to calculate the kurtosis value of each intrinsic mode component without interference term, and determine the non-impact term of the wheel-rail force signal according to the kurtosis value of each intrinsic mode component without interference term;

[0111] An extraction module 706, configured to perform short-time Fourier transform on the wheel-rail force signal after filtering out the non-impact term to obtain a moving time window marginal spectrum, and extract weak electromagnetic interference characteristics from the wheel-rail force signal according to the moving time window marginal spectrum.

[0112] In one embodiment, the acquisition module 701 is specifically configured to:

[0113] Determine the impact position where the extreme value in the wheel-rail force signal of the high-speed railway reaches a preset value according to the wheel-rail force signal of the high-speed railway;

[0114] Determine the abnormal position of the wheel-rail force signal according to the impact position where the extreme value in the wheel-rail force signal reaches the preset value.

[0115] In one embodiment, the mode decomposition module 702 is specifically configured to:

[0116] Determine the start point and the end point of the interference signal according to the abnormal position;

[0117] Determine the interference region of the wheel-rail force signal according to the start point and the end point of the interference signal.

[0118] In one embodiment, the mode decomposition module 702 is specifically configured to:

[0119] Perform mode decomposition on the wheel-rail force signal in the interference region by using the adaptive chirp mode decomposition algorithm.

[0120] In one embodiment, the second calculation module 705 is specifically configured to:

[0121] Calculate the kurtosis values of the intrinsic mode components of each interference-free term;

[0122] Sort the kurtosis values of the intrinsic mode components of each interference-free term from small to large, and select a preset number of kurtosis values in order from the sorted kurtosis values as the target kurtosis values;

[0123] Determine the components in the wheel-rail force signal corresponding to the target kurtosis values as the non-impact terms of the wheel-rail force signal.

[0124] In one embodiment, the extraction module 706 is specifically configured to:

[0125] After filtering out the non-impact terms from the wheel-rail force signal with the trend term filtered out, obtain the impact signal;

[0126] After performing short-time Fourier transform on the impact signal, calculate the time marginal spectrum, determine the moving average value of the time marginal spectrum, and obtain the moving time window marginal spectrum;

[0127] Determine the weak characteristic electromagnetic interference points according to the moving time window marginal spectrum, and extract the weak characteristics of electromagnetic interference from the wheel-rail force signal.

[0128] In one embodiment, the extraction module 706 is further configured to:

[0129] Replace the region after extracting the weak characteristics of electromagnetic interference with Gaussian random signals.

[0130] Based on the foregoing inventive concept, as Figure 8As shown in the figure, the present invention also provides a computer device 800, including a memory 810, a processor 820, and a computer program 830 stored on the memory 810 and operable on the processor 820. When the processor 820 executes the computer program 830, the foregoing method for extracting weak electromagnetic interference features of high-speed railway wheel-rail forces is implemented.

[0131] An embodiment of the present invention also provides a computer-readable storage medium. The computer-readable storage medium stores a computer program, and when the computer program is executed by a processor, the foregoing method for extracting weak electromagnetic interference features of high-speed railway wheel-rail forces is implemented.

[0132] An embodiment of the present invention also provides a computer program product. The computer program product includes a computer program, and when the computer program is executed by a processor, the foregoing method for extracting weak electromagnetic interference features of high-speed railway wheel-rail forces is implemented.

[0133] In summary, in the embodiments of the present invention, the wheel-rail force signal of the high-speed railway is collected, and according to the wheel-rail force signal of the high-speed railway, the abnormal position of the wheel-rail force signal is determined; the interference area is determined according to the abnormal position, and the wheel-rail force signal in the interference area is subjected to modal decomposition to obtain a plurality of intrinsic mode components; the power spectral density of each intrinsic mode component is calculated, and according to the power spectral density of each intrinsic mode component, the trend term of the wheel-rail force signal is determined; the wheel-rail force signal after filtering the trend term is subjected to modal decomposition to obtain a plurality of intrinsic mode components without interference terms; the kurtosis value of each intrinsic mode component without interference terms is calculated, and according to the kurtosis value of each intrinsic mode component without interference terms, the non-impulse term of the wheel-rail force signal is determined; by performing short-time Fourier transform on the wheel-rail force signal after filtering the non-impulse term, a moving time window marginal spectrum is obtained, and according to the moving time window marginal spectrum, weak electromagnetic interference features are extracted from the wheel-rail force signal. In this way, the wheel-rail force signal is decomposed, the average frequency algorithm of the power spectral density is used to judge the trend term; the kurtosis index is used to diagnose the non-impulse term; a moving time window marginal spectrum method is proposed to identify weak feature signals of electromagnetic interference, accurately identify the electromagnetic interference of high-speed railway wheel-rail forces, and provide reliable data support for railway operation.

[0134] Those skilled in the art should understand that the embodiments of the present invention can be provided as a method, a system, or a computer program product. Therefore, the present invention can be implemented in the form of a complete hardware embodiment, a complete software embodiment, or an embodiment combining software and hardware aspects. Moreover, the present invention can be implemented in the form of a computer program product implemented 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.

[0135] The present invention is described with reference to the flowcharts and / or block diagrams of methods, apparatus (systems), and computer program products according to embodiments of the present invention. It should be understood that each flow and / or block in the flowchart and / or block diagram, and combinations of flows and / or blocks in the flowchart and / or block diagram, can be implemented by computer program instructions. These computer program instructions can be provided to the processors of general-purpose computers, special-purpose computers, embedded processors, or other programmable data processing devices to produce a machine, such that the instructions executed by the processors of the computer or other programmable data processing devices produce means for implementing the functions specified in one flow Figure 1 one flow or multiple flows and / or blocks Figure 1 or in multiple blocks.

[0136] These computer program instructions can also be stored in a computer-readable memory that can direct a computer or other programmable data processing device to work in a specific manner, such that the instructions stored in the computer-readable memory produce a manufactured article including instruction means that implement the functions specified in one flow Figure 1 one flow or multiple flows and / or blocks Figure 1 or in multiple blocks.

[0137] These computer program instructions can also be loaded onto a computer or other programmable data processing device, such that a series of operation steps are executed on the computer or other programmable device to produce a computer-implemented process, and thus the instructions executed on the computer or other programmable device provide steps for implementing the functions specified in one flow Figure 1 one flow or multiple flows and / or blocks Figure 1 or in multiple blocks.

[0138] The specific embodiments described above further elaborate on the objectives, technical solutions, and beneficial effects of the present invention. It should be understood that the above are only specific embodiments of the present invention and are not used to limit the protection scope of the present invention. Any modifications, equivalent replacements, improvements, etc. made within the spirit and principles of the present invention shall be included in the protection scope of the present invention.

Claims

1. A method for extracting weak features of electromagnetic interference of wheel-rail force of high-speed railway, characterized in that: include: Collect wheel-rail force signals of high-speed railways, and determine the abnormal positions of the wheel-rail force signals according to the wheel-rail force signals of high-speed railways; The interference area is determined according to the abnormal position, and the wheel-rail force signal in the interference area is modally decomposed to obtain multiple intrinsic modal components; Calculate the power spectrum density of each intrinsic modal component, and determine the trend term of the wheel-rail force signal according to the power spectrum density of each intrinsic modal component; Perform modal decomposition on the wheel-rail force signal after filtering out the trend term to obtain multiple intrinsic modal components without interference terms. Calculate the kurtosis value of the intrinsic modal component of each non-interference item, and determine the non-impact item of the wheel-rail force signal according to the kurtosis value of the intrinsic modal component of each non-interference item; By performing short-time Fourier transform on the wheel-rail force signal after filtering out non-impact terms, a moving time window edge spectrum is obtained. Based on the moving time window edge spectrum, weak features of electromagnetic interference are extracted from the wheel-rail force signal.

2. The method according to claim 1, characterized in that According to the wheel-rail force signal of the high-speed railway, the abnormal position of the wheel-rail force signal is determined, including: According to the wheel-rail force signal of the high-speed railway, an impact position where the extreme value of the wheel-rail force signal reaches a preset value is determined; The abnormal position of the wheel-rail force signal is determined according to the impact position where the extreme value in the wheel-rail force signal reaches a preset value.

3. The method according to claim 1, characterized in that Determine the interference area based on the abnormal location, including: Determine the starting point and ending point of the interference signal according to the abnormal position; The interference area of ​​the wheel-rail force signal is determined according to the starting point and the ending point of the interference signal.

4. The method according to claim 1, characterized in that Modal decomposition of the wheel-rail force signal in the interference area includes: The wheel-rail force signal in the interference area is modally decomposed using the adaptive chirp modal decomposition algorithm.

5. The method according to claim 1, characterized in that Calculate the kurtosis value of the intrinsic modal component of each non-interference item, and determine the non-impact item of the wheel-rail force signal according to the kurtosis value of the intrinsic modal component of each non-interference item, including: Calculate the kurtosis value of each intrinsic modal component without interference terms; The kurtosis values ​​of the connotation modal components without interference items are sorted from small to large, and a preset number of kurtosis values ​​are selected in order from the sorted kurtosis values ​​as target kurtosis values; The component in the wheel-rail force signal corresponding to the target kurtosis value is determined as the non-impact term of the wheel-rail force signal.

6. The method according to claim 1, characterized in that By performing short-time Fourier transform on the wheel-rail force signal after filtering out non-impact terms, the moving time window edge spectrum is obtained. According to the moving time window edge spectrum, weak electromagnetic interference features are extracted from the wheel-rail force signal, including: After filtering out non-impact items from the wheel-rail force signal with trend items removed, an impact signal is obtained; After performing short-time Fourier transform on the impact signal, the time edge spectrum is calculated, the moving average of the time edge spectrum is determined, and the moving time window edge spectrum is obtained; The weak characteristic electromagnetic interference points are determined according to the edge spectrum of the moving time window, and the weak characteristics of electromagnetic interference are extracted from the wheel-rail force signal.

7. The method according to claim 1, characterized in that After extracting weak features of electromagnetic interference, it also includes: The area after the weak features of electromagnetic interference are extracted is replaced with a Gaussian random signal.

8. A device for extracting weak features of electromagnetic interference of wheel-rail force of high-speed railway, characterized in that: include: A collection module is used to collect wheel-rail force signals of the high-speed railway, and determine the abnormal position of the wheel-rail force signals according to the wheel-rail force signals of the high-speed railway; The modal decomposition module is used to determine the interference area according to the abnormal position, perform modal decomposition on the wheel-rail force signal in the interference area, and obtain multiple intrinsic modal components; The first calculation module is used to calculate the power spectrum density of each intrinsic modal component, and determine the trend term of the wheel-rail force signal according to the power spectrum density of each intrinsic modal component; The secondary decomposition module is used to perform modal decomposition on the wheel-rail force signal after filtering out the trend term, and obtain multiple intrinsic modal components without interference terms; The second calculation module is used to calculate the kurtosis value of the intrinsic modal component of each non-interference item, and determine the non-impact item of the wheel-rail force signal according to the kurtosis value of the intrinsic modal component of each non-interference item; The extraction module is used to obtain a moving time window edge spectrum by performing a short-time Fourier transform on the wheel-rail force signal after filtering out non-impact terms, and extract weak features of electromagnetic interference from the wheel-rail force signal according to the moving time window edge spectrum.

9. 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, the method according to any one of claims 1 to 7 is implemented.

10. A computer-readable storage medium, characterized in that: The computer-readable storage medium stores a computer program, and when the computer program is executed by a processor, the method according to any one of claims 1 to 7 is implemented.

11. A computer program product, characterized in that The computer program product comprises a computer program, and when the computer program is executed by a processor, the method according to any one of claims 1 to 7 is implemented.

Images