Method and device for identifying modal frequencies of a railway vehicle body

Abstract

The application discloses a kind of railway vehicle body modal frequency identification method and device, belong to rail transit vehicle state monitoring technical field.The present application aims at solving the problems of high cost in prior art, departure from real operating conditions.The method comprises: obtaining the vibration response data of the target vehicle in the process of meeting car;Vibration response data is preprocessed and effective data segment is intercepted;The effective data segment after interception is transformed in frequency domain;At least one characteristic frequency spectrum peak is extracted from the frequency energy distribution characteristics;The frequency corresponding to the characteristic frequency spectrum peak is determined as the inherent modal frequency of the target vehicle body under operating conditions.The present application uses the meeting car event frequently occurred in normal train operation as natural excitation source, without additional test, can long-term, on-line, low-costly obtain the inherent frequency of vehicle body that truly reflects the vehicle operating dynamics characteristics, provides efficient and reliable new means for vehicle state monitoring and health assessment.

Description

Technical Field

[0001] This invention relates to the field of rail transit vehicle condition monitoring technology, and in particular to a method and device for identifying the modal frequencies of railway vehicle bodies. 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] The dynamic characteristics of railway vehicle body structures (mainly characterized by modal parameters such as natural frequency, damping ratio, and mode shape) are key indicators for evaluating vehicle running stability, safety, and structural health. These dynamic characteristics are manifested in "rigid body modes" (such as roll and pitch) which include the influence of the suspension system, and "elastic modes" (such as first-order bending and torsion) which reflect the elasticity of the body structure itself.

[0004] Currently, traditional methods for obtaining these modal parameters have the following limitations: (1) Experimental modal analysis method: In the laboratory, controllable excitation is applied to the stationary vehicle body by a vibrator or a hammer to identify modal parameters. Although this method has high accuracy, it has the following problems: the test cost is high and the cycle is long, and it cannot be implemented on vehicles in online operation; the fixed boundary conditions in the laboratory are different from the vehicle's operating state on the real line, which makes the results unable to accurately reflect the "operational mode" exhibited by the vehicle in actual operation.

[0005] (2) Simulation analysis method: This type of method (including multibody system dynamics simulation for analyzing rigid body modes and finite element simulation for analyzing elastic modes) heavily relies on the accuracy of the model and the precision of the parameters. However, in the modeling process, simplification and assumptions about complex suspension characteristics, connection stiffness and structural damping are inevitable, resulting in inherent deviations between the simulation results and the actual dynamic characteristics of the vehicle.

[0006] In summary, existing technologies suffer from inherent drawbacks in acquiring complete modal information of in-service vehicles under actual operating conditions, including high cost and insufficient accuracy. Existing methods for acquiring the inherent frequencies of the vehicle body have inherent flaws, resulting in calculation results that cannot accurately reflect the dynamic characteristics of the vehicle under actual operating conditions. Specifically, they face the following technical bottlenecks: (1) high cost and limited computational accuracy; (2) inability to reproduce boundary conditions during operation, failing to accurately reflect the "operating modes" exhibited by the vehicle in actual operation.

[0007] Therefore, the core technical problem to be solved by the embodiments of the present invention is: how to use the frequent passing events that occur during normal train operation as a natural excitation source, and obtain the vehicle body's natural frequency that truly reflects the vehicle's operational dynamics characteristics in a long-term, online, and low-cost manner without additional experiments. Summary of the Invention

[0008] This invention provides a method for identifying the modal frequencies of railway vehicle bodies. This method utilizes frequent train encounters during normal operation as a natural excitation source, enabling long-term, online, and low-cost acquisition of the vehicle body's inherent frequencies that accurately reflect the vehicle's operational dynamics without additional testing. This provides an efficient and reliable new means for vehicle condition monitoring and health assessment. The method includes: Acquire vehicle body vibration response data generated by the target vehicle during a passing event; The vehicle vibration response data is preprocessed, and a valid data segment containing the complete oncoming vehicle excitation effect is extracted. The effective data segment is subjected to frequency domain transformation to obtain frequency domain features characterizing its energy distribution; Extract at least one characteristic spectral peak from the frequency domain features; The frequency corresponding to the peak value of the characteristic spectrum is identified as the inherent modal frequency of the target vehicle body in the operating state.

[0009] This invention also provides a railway vehicle body modal frequency identification device, which utilizes frequent train passing events during normal operation as a natural excitation source. Without additional testing, it can acquire the vehicle body's natural frequencies that truly reflect the vehicle's operational dynamics in a long-term, online, and low-cost manner, providing an efficient and reliable new method for vehicle condition monitoring and health assessment. The device includes: The data acquisition module is used to acquire vehicle vibration response data generated by the target vehicle during a passing event; The data processing module is used to preprocess the vehicle vibration response data and extract the effective data segment containing the complete oncoming vehicle excitation effect. The frequency domain analysis module is used to perform frequency domain transformation processing on the effective data segment to obtain frequency domain characteristics that characterize its energy distribution; Peak extraction module, used to extract at least one characteristic spectral peak from the frequency domain features; The frequency identification module is used to identify the frequency corresponding to the peak value of the characteristic spectrum as the inherent modal frequency of the target vehicle body in the operating state.

[0010] 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 railway vehicle body modal frequency recognition method.

[0011] This invention also provides a computer-readable storage medium storing a computer program that, when executed by a processor, implements the above-described railway vehicle body modal frequency recognition method.

[0012] This invention also provides a computer program product, which includes a computer program that, when executed by a processor, implements the above-described railway vehicle body modal frequency recognition method.

[0013] Compared with existing technologies that suffer from high costs and low accuracy in obtaining modal parameters, the railway vehicle body modal frequency identification scheme based on aerodynamic excitation during train encounters provided in this invention involves: acquiring the vibration response data of the target vehicle during train encounters; preprocessing the vibration response data and extracting effective data segments; performing frequency domain transformation on the extracted effective data segments; extracting at least one characteristic spectral peak from the frequency domain energy distribution characteristics; and determining the frequency corresponding to the characteristic spectral peak as the inherent modal frequency of the target vehicle body under operating conditions. This invention utilizes the frequent train encounter events during normal train operation as a natural excitation source, enabling long-term, online, and low-cost acquisition of the vehicle body's inherent frequencies that truly reflect the vehicle's operational dynamics without additional testing. This provides an efficient and reliable new method for vehicle condition monitoring and health assessment. Attached Figure Description

[0014] 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: Figure 1 This is a flowchart illustrating the method for identifying the modal frequency of a railway vehicle body based on aerodynamic excitation during train passing, as described in an embodiment of the present invention. Figure 2 This is a schematic diagram illustrating the principle of the railway vehicle body modal frequency identification method based on aerodynamic excitation during train passing in an embodiment of the present invention. Figure 3 This is a schematic diagram of the original waveform of the lateral acceleration of the vehicle body in the passing state in an embodiment of the present invention; Figure 4 This is a schematic diagram of the lateral acceleration power spectral density of the CR400AF EMU body in an embodiment of the present invention; Figure 5 This is a schematic diagram of the lateral acceleration power spectral density of the CR400BF EMU body in an embodiment of the present invention; Figure 6 This is a schematic diagram of the structure of the railway vehicle body modal frequency recognition device based on aerodynamic excitation during train passing, as described in an embodiment of the present invention. Detailed Implementation

[0015] 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.

[0016] The acquisition, storage, use, and processing of data in this application comply with relevant laws and regulations.

[0017] The inventors discovered that the strong broadband aerodynamic excitation generated when trains meet at high speed on double tracks can effectively excite both the rigid body modes and elastic modes of the vehicle body simultaneously. This provides a natural and powerful excitation source for obtaining the vehicle's true operational modal parameters. Currently, there is a lack of a specific analytical method for extracting vehicle body modal parameters (especially low-frequency rigid body modes and mid-to-high-frequency elastic modes) from complex response signals based on train meeting data. Therefore, this invention proposes a vehicle body modal frequency identification scheme based on train meeting aerodynamic excitation. This scheme utilizes vibration response data generated by naturally occurring train meeting events during vehicle operation to identify the inherent modal frequencies of the in-service vehicle body.

[0018] Figure 1 This is a flowchart illustrating the vehicle body modal frequency recognition method based on aerodynamic excitation during oncoming traffic, as described in an embodiment of the present invention. Figure 1 As shown, the method includes the following steps: Step 101: Obtain vehicle vibration response data generated by the target vehicle during the oncoming traffic event; Step 102: Preprocess the vehicle vibration response data and extract the effective data segment containing the complete oncoming vehicle excitation effect; Step 103: Perform frequency domain transformation on the effective data segment to obtain frequency domain features characterizing its energy distribution; Step 104: Extract at least one characteristic spectral peak from the frequency domain features; Step 105: Identify the frequency corresponding to the peak value of the characteristic spectrum as the inherent modal frequency of the target vehicle body in the operating state.

[0019] The vehicles in this embodiment of the invention can be railway vehicles.

[0020] The method for identifying the modal frequencies of railway vehicle bodies based on aerodynamic excitation during train passing provided in this invention involves the following steps: acquiring vibration response data of the target vehicle body generated during a train passing event; preprocessing the vibration response data and extracting effective data segments containing the complete excitation effect of the train passing event; performing frequency domain transformation processing on the effective data segments to obtain frequency domain features characterizing their energy distribution; extracting at least one characteristic spectral peak from the frequency domain features; and identifying the frequency corresponding to the characteristic spectral peak as the inherent modal frequencies of the target vehicle body under operating conditions.

[0021] Compared with existing technologies that suffer from high costs and low accuracy in obtaining modal parameters, the railway vehicle body modal frequency identification method based on aerodynamic excitation during train encounters provided in this invention can utilize frequent train encounter events during normal operation as a natural excitation source. Without additional testing, it can obtain the vehicle body's natural frequencies that truly reflect the vehicle's operational dynamics in a long-term, online, and low-cost manner, providing an efficient and reliable new method for vehicle condition monitoring and health assessment. The following is a detailed description of the railway vehicle body modal frequency identification method based on aerodynamic excitation during train encounters provided in this invention.

[0022] This invention provides a method for identifying the natural frequency of railway vehicle bodies based on passing vehicle data. This method addresses the technical problems of existing technologies being costly, detached from real-world service conditions, and unable to achieve long-term online monitoring. It achieves low-cost, high-efficiency, and high-reliability acquisition of the natural frequency of the vehicle body that truly reflects the vehicle's operational dynamics. The following is a related explanation. Figures 2 to 5 A detailed introduction will be provided.

[0023] In specific implementation, in step 101 above, firstly, referring to the implementation of patent publication number CN118494558A (patent application number: 202410707060.0), the oncoming vehicle event of the target vehicle can be identified. Secondly, the vehicle body vibration response data generated by the target vehicle in the oncoming vehicle event is obtained. In one embodiment, the vehicle body vibration response data can be the lateral acceleration data of the vehicle body. The oncoming vehicle event can be automatically identified and screened to eliminate interference from non-oncoming vehicle conditions. In specific implementation, any automatic identification and screening method that can eliminate interference from non-oncoming vehicle conditions can be adopted. For example... Figure 2 The figures shown are “measured lateral acceleration of the vehicle body” and “automatic identification of oncoming traffic conditions based on data features”. Figure 2 This is a schematic diagram illustrating the principle of the vehicle body modal frequency recognition method based on oncoming aerodynamic excitation in an embodiment of the present invention.

[0024] In one embodiment, the above-mentioned railway vehicle body modal frequency identification method may further include: determining whether the lateral acceleration data of the vehicle body has preset lateral acceleration data characteristics for the time of passing oncoming trains, based on the lateral acceleration data characteristics of the vehicle body during passing (see the description in the example below for details). Figure 2 The image shows "Automatic recognition of meeting conditions based on data features".

[0025] In practice, automatic identification of oncoming traffic conditions based on data features can also include verifying the oncoming traffic events of the target vehicle body to eliminate false oncoming traffic events caused by crosswind interference and low-frequency periodic swaying. The obtained verified oncoming traffic events can further improve the accuracy of oncoming traffic event identification and further improve the accuracy of subsequent vehicle body modal frequency identification.

[0026] In specific implementation, the step of verifying the oncoming events of the target vehicle body can be further implemented by adopting a signal post-processing method for improving the accuracy of oncoming vehicle recognition, as disclosed in patent application number 202610125551.3, to verify the oncoming events of the target vehicle body, so as to eliminate false alarm conditions such as crosswind interference and low-frequency periodic shaking, and ensure that the obtained oncoming data is true and effective. The signal post-processing method for improving the accuracy of oncoming vehicle recognition disclosed in patent application number 202610125551.3 is as follows: Based on the candidate oncoming vehicle events initially identified by existing methods, the candidate events are first screened by time-domain waveform to eliminate events whose time-domain waveforms do not conform to a preset impact pattern; then, for the events that pass the screening, the frequency domain features of their vibration signals are extracted; finally, a second screening is performed based on the comparison results of the frequency domain features and a preset threshold to distinguish the real oncoming vehicle events. This signal post-processing method for improving the accuracy of oncoming vehicle recognition, through a two-stage screening architecture of first time domain and then frequency domain, can effectively capture and utilize the essential differences between transient oncoming vehicle impacts and persistent interference in time-domain waveforms and frequency-domain energy distribution, thereby significantly improving the accuracy and reliability of oncoming vehicle recognition and producing a high-purity oncoming vehicle event dataset.

[0027] In one embodiment, in step 102 above, the preprocessing may include bandpass filtering of the data to preserve the vehicle body modal frequency components and suppress high-frequency noise and low-frequency trend terms.

[0028] In specific implementation, in step 102 above, such as Figure 2 The "bandpass filtering based on domain prior knowledge" shown applies bandpass filtering to the data after automatic identification of passing conditions based on data features, resulting in a bandpass-filtered passing event. This preserves the vehicle body modal frequency components and suppresses high-frequency noise and low-frequency trend terms, which can further improve the accuracy of passing event identification and the accuracy of subsequent vehicle body modal frequency identification.

[0029] In one embodiment, in step 102 above, the rule for extracting valid data segments can be: based on the multiple characteristic peaks that appear in the vibration response data during the passing process, starting from a first preset position before the first characteristic peak and ending at a second preset position after the last characteristic peak, extract the data in between.

[0030] In specific implementation, in step 102 above, such as Figure 2 The "effective data sequence selection" shown extracts effective passing data segments from the bandpass-filtered passing events based on the peak values ​​of the vehicle's lateral acceleration curve. When a passing event occurs, the vehicle's lateral acceleration will have four peak values, such as... Figure 3 As shown, for example: a) If the data from the train meeting comes from a high-speed comprehensive inspection train (spatial sampling method), the valid data selection rule is: take the first peak point. x The first 15 meters are taken as the starting point, and the last 15 meters after the fourth peak are taken as the ending point. The lateral acceleration of the vehicle body between the starting point and the ending point is recorded as valid data.

[0031] b) If the data from passing trains comes from other railway vehicles or other testing equipment capable of collecting lateral acceleration of the train body (time sampling method), the valid data selection rule is: take the data 15 minutes before the first peak point. v The time is taken as the starting time, 15 / after the fourth peak. v The moment is the end moment. v The train's speed is expressed in m / s.

[0032] As can be seen from the above, in specific implementation, the first preset position before the first characteristic peak can be a preset distance position before the first characteristic peak. The preset distance position is preferably in the range of 14 to 16 meters, more preferably 15 meters. The first preset position before the first characteristic peak can also be the position corresponding to a preset time before the first characteristic peak. The preset time position is preferably in the range of 14 meters before the first peak point. v Up to 16 meters / v The time corresponds to the position. The value of the second preset position after the last feature peak as the endpoint can be determined by referring to the implementation of the first preset position before the first feature peak mentioned above.

[0033] In one embodiment, in step 103 above, the frequency domain transformation process may include calculating the power spectral density of the effective data segment, such as... Figure 2The "power spectral density calculation" shown in the figure. In specific implementation, this embodiment of the invention uses power spectral density for analysis because it can effectively smooth random noise, provide a more stable spectrum estimate, and further improve the accuracy of subsequent vehicle body modal frequency identification. Of course, this embodiment of the invention can also use other time-frequency analysis methods for frequency domain transformation processing to obtain frequency domain features characterizing its energy distribution; this embodiment of the invention does not impose specific limitations.

[0034] In specific implementations, embodiments of the present invention may also employ the Welch average periodogram method to calculate the power spectral density of the effective data segment. That is, when calculating the power spectral density, embodiments of the present invention preferably use the Welch average periodogram method to further improve the accuracy of subsequent vehicle body modal frequency identification.

[0035] In one embodiment, step 104 above may include: calculating the first-order forward difference of the frequency domain feature, identifying local maxima where the sign changes from positive to negative, and filtering the local maxima based on a preset amplitude threshold to obtain the feature spectrum peak, such as... Figure 2 The terms "first-order forward difference", "peak extraction", and "effective peak selection" are shown in the diagram.

[0036] In specific implementation, in step 104 above, extracting at least one characteristic spectral peak from the frequency domain features may specifically include: extracting spectral peak information from the power spectral density curve, which may include: Calculate the first-order forward difference at the current peak point based on the power spectral density curve; Find frequency points where the first-order difference of the previous peak point is greater than 0 and the first-order forward difference of the current peak point is less than 0 to form a candidate peak set; Based on a preset filtering threshold (preset amplitude threshold), the final set of peaks is extracted from the candidate peak set; the preset filtering threshold is used to filter fluctuations caused by noise, and its specific value is adjusted according to the actual data signal-to-noise ratio.

[0037] In practice, the specific implementation method for extracting the peak information of each spectrum in the power spectral density curve described above can improve the accuracy and efficiency of power spectral density curve extraction, and thus further improve the accuracy and efficiency of vehicle encounter recognition.

[0038] In one embodiment, in step 105 above, such as Figure 2 The "natural modal frequencies of the vehicle body" shown herein can include both rigid body modal frequencies and elastic modal frequencies of the vehicle body. Of course, other forms of modal frequencies are also included in this invention, but are not limited thereto.

[0039] To facilitate understanding of how this invention is implemented, an example is given below for detailed explanation.

[0040] This invention uses measured lateral acceleration data of the train body during the passing phase as the research object, and employs power spectral density analysis to determine the natural frequencies of the railway vehicle body under operating conditions. The complete technical solution of this invention is as follows: Figure 2 As shown, the embodiments of the present invention mainly consist of three parts: the first part is intelligent recognition of train meeting conditions based on time-domain features; the second part is data preprocessing technology for calculating natural frequencies; and the third part is the calculation of the natural frequencies of the railway vehicle body in its operating state.

[0041] This invention takes the lateral acceleration data of a railway vehicle collected in real time during operation as the research object, and the specific implementation method is as follows.

[0042] Part 1: Intelligent Recognition of Meeting Conditions Based on Time-Domain Features.

[0043] This part is the basic algorithm, which provides a solid and reliable data foundation for subsequent calculation of the vehicle's natural modal frequencies under operating conditions.

[0044] Figure 3 This is a schematic diagram of the original waveform of the lateral acceleration of the train body when two trains meet, as shown in this embodiment of the invention. It clearly illustrates the characteristics of the lateral acceleration data of the train body during the meeting process. (1) There are 4 significant peak points and 2 peak-to-peak points (the front and rear of the adjacent vehicles pass the target vehicle in succession). (2) Between two peaks that form a peak-to-peak value, the amplitude of the latter peak is greater than that of the former peak (2>1, 4>3). (3) During the meeting phase, the peak values ​​of the two peaks are in opposite directions (1 and 4 have the same number, 2 and 3 have the same number); (4) The distance between two peaks that make up the same peak value is relatively fixed, about 6 to 9 meters; (5) The distance between the first and fourth peak points can be quantified, and its value is related to the length of the adjacent EMU (distance between the front and rear of the train) and the relative difference in operating speed.

[0045] Therefore, the automatic identification of the passing conditions during operation is transformed into "determining whether the lateral acceleration data of the car body has the aforementioned data characteristics". That is, in one embodiment, the above-mentioned railway vehicle car body modal frequency identification method may further include: based on the lateral acceleration data characteristics of the car body during passing, determining whether the lateral acceleration data of the car body has the preset lateral acceleration data characteristics of the car body during passing as described in steps (1) to (5) above, such as... Figure 2 The image shows "Automatic recognition of meeting conditions based on data features".

[0046] Before designing the indicators, the "quantification process of the distance between the 1st and 4th peak points" is explained first. The detailed implementation process of step (5) above is as follows.

[0047] The meeting point begins when the front of the target vehicle and the adjacent vehicle meet, and ends when their rear ends meet. This process can be viewed as the target vehicle and the adjacent vehicle meeting at... t The length of the adjacent vehicles moved together within a certain time period.

[0048] Therefore, the motion expression during the passing process is as follows:

[0049] In the formula, v 1 represents the target vehicle's operating speed; v 2 represents the speed of vehicles traveling on the adjacent track; L The length of the vehicle on the adjacent track.

[0050] Table 1 shows the body lengths of commonly used railway vehicles currently in service.

[0051] Table 1. Common Railway Vehicle Body Lengths Currently in Service

[0052] To simplify the calculations, railway vehicles can be divided into two categories based on their length: approximately 200m and approximately 400m. The travel time can then be calculated. t satisfy: (1) Railway vehicles on high-speed railways typically operate at stable speeds of 250 km / h, 300 km / h, and 350 km / h. The acceleration and deceleration processes during station entry are relatively short, but the stress and vibration conditions during this phase are complex. When calculating the departure distance of the target vehicle, speeds of 250 km / h and 350 km / h are used to simulate the speed difference between the two railway vehicles during a meeting. (2) (1) When the target vehicle is traveling at 350 km / h and the adjacent vehicle is traveling at 250 km / h and has a body length of 200 meters, the distance between the first and fourth peak points is approximately 116 m.

[0053] (2) When the target vehicle is traveling at 350 km / h and the adjacent vehicle is traveling at 250 km / h and has a body length of 400 meters, the distance between the first and fourth peak points is approximately 233 m.

[0054] (3) When the target vehicle is traveling at 250 km / h and the adjacent vehicle is traveling at 350 km / h and has a body length of 200 meters, the distance between the first and fourth peak points is approximately 83.3 m.

[0055] (4) When the target vehicle is traveling at 250 km / h and the adjacent vehicle is traveling at 350 km / h and has a body length of 400 meters, the distance between the first and fourth peak points is approximately 167 m.

[0056] Based on the above analysis results, the distance between the 1st and 4th peak points is quantified to a range of 80~250m.

[0057] Based on the characteristics of the vehicle's lateral acceleration data, an automatic identification method for oncoming traffic conditions is developed. Figure 2 The automatic identification of meeting conditions based on data features is shown below: (1) Determine whether the current lateral acceleration amplitude of the vehicle body is greater than 0.03g. If so, record it as the possible first peak point. a 1. Record the corresponding coordinate information. L .

[0058] (2) Using the first possible peak point as a reference, and referring to the distance between the first and second peak points, extract the lateral acceleration data segment of the vehicle body. A 2.

[0059] (3)

[0060] In the formula, This represents the minimum possible distance between the second peak point and the first peak point, which is set to 6 meters here. This represents the maximum possible distance between the second peak point and the first peak point, which is set to 9 meters here. A This is the original sequence of lateral acceleration data for the vehicle body.

[0061] (3) Retain A 2 sequences satisfying a · a Acceleration data points with the condition 1<0 are denoted as the new sequence. A 2' and record the maximum absolute value as the possible second peak point. a 2: (4) In the formula, sgn(·) retains the positive or negative sign.

[0062] (4) If a 2>0.06g, continue; if a 2 < 0.06g, the current data does not meet the requirements for meeting data.

[0063] (5) Using the first possible peak point as a reference, and referring to the distance between the first and fourth peak points, extract the lateral acceleration data segment of the vehicle body. A 4.

[0064] (5)

[0065] In the formula, This represents the minimum possible distance between the fourth peak point and the first peak point, which is set to 50 meters here. This represents the maximum possible distance between the third peak point and the first peak point, which is set to 250 meters here.

[0066] (6) Retain A 4 sequences satisfying a · a Acceleration data points with the condition 1>0 are denoted as the new sequence. A 4' and record the maximum absolute value as the possible fourth peak point. a 4: (6) (7) If a 4 > 0.03g, record the 4th peak point. a 4. Location information corresponding to the position L 4. Simultaneously extract the lateral acceleration data segment of the vehicle body corresponding to the oncoming traffic. A m (7) For a more specific implementation of train meeting event recognition (automatic recognition of train meeting conditions), please refer to the applicant’s earlier patent applications CN202410707060.0 (Method and Apparatus for Recognizing Train Meeting) and CN202610125551.3 (Signal Post-processing Method and Apparatus for Improving the Accuracy of Train Meeting Recognition), the full text of which is incorporated herein by reference.

[0067] Part Two: Data Preprocessing Techniques for Natural Frequency Calculation

[0068] This part is the key algorithm, which ensures the validity, scientific nature, and reliability of the subsequent calculation results of the vehicle body's natural modal frequencies.

[0069] (1) Noise Removal: The lateral acceleration data of the vehicle body during oncoming traffic, which has been filtered through the above steps, is first subjected to filtering, such as the bandpass filtering preprocessing in step 102 above. The filtered signal components should cover the main modal frequency components of the vehicle body. Based on the research results in this field, the filtering frequency band can be selected as 0.1 to 20 Hz. The filtering process can be carried out with reference to the following formula: (8) (9) (2) The basis for selecting the data sequence for natural frequency calculation, namely, the effective data segment containing the complete excitation effect of the meeting mentioned in step 102 above: a) If the data from the train meeting comes from a high-speed comprehensive inspection train (spatial sampling method), the valid data selection rule is: take the first peak point. L The first 15 meters is taken as the starting point, the 4th peak point. L The point 15 meters after the start time is taken as the end time. The lateral acceleration of the vehicle body between the start time and the end time is recorded as valid data.

[0070] b) If the data from passing trains comes from other railway vehicles or other testing equipment capable of collecting lateral acceleration of the train body (time sampling method), the valid data selection rule is: take the data 15 / 100 seconds before the first peak point. v The time is taken as the starting time, 15 / after the 4th peak point. v The moment is the end moment. v The train's speed is expressed in m / s.

[0071] This data will subsequently be used for calculating the vehicle body's natural modal frequencies. For convenience, the data sequence will be denoted as... A c .

[0072] Part Three: Calculation of the inherent frequency of railway vehicle body under operating conditions.

[0073] This part is the core algorithm, which analyzes the inherent frequency of the vehicle body under real operating conditions based on the data provided in the first part.

[0074] (1) Frequency domain energy distribution calculation: For the preprocessed effective passing data segment, perform spectrum analysis to obtain its frequency domain energy distribution. The spectrum analysis can be achieved using power spectral density, short-time Fourier transform marginal spectrum, etc. Here, we take power spectral density as an example to introduce the implementation process: Lateral acceleration sensors were installed on a typical CR400AF and a certain CR400BF EMU to collect operational data. First, the lateral acceleration data of the car body at the time of passing was obtained using the time-domain feature-based intelligent identification method for passing conditions mentioned in Part I. Then, the data preprocessing technique for natural frequency calculation mentioned in Part II was used to complete the preprocessing of the passing data. Finally, the power spectral density was calculated using the preprocessed passing data, employing the Welch average periodogram method. The calculation results are shown below. Figure 4 and Figure 5 As shown.

[0075] from Figure 4 and Figure 5 As can be seen, power spectral density analysis will be as follows: Figure 3 The lateral acceleration data of the vehicle body shown has been transformed from abstract to concrete: the power spectral density plot shows several dominant frequencies with obvious peaks. These frequencies are the vehicle body's natural mode frequencies that this embodiment of the invention aims to solve.

[0076] Based on the above analysis results, the calculation scheme for the natural modal frequencies of the vehicle body can be summarized as follows: (1) Based on the measured lateral acceleration data of the vehicle body, intelligently obtain the time / position information of the target vehicle when it passes another vehicle. L and L 4.

[0077] (2) Obtain the lateral acceleration data of the vehicle body at the time of the meeting based on the meeting information. A m And complete data preprocessing; (3) The power spectral density of the lateral acceleration of the vehicle body is calculated using the Welch average periodogram method.

[0078] The specific steps are as follows: acceleration signal A c Divided into segments with a 50% overlap rate H There are data segments, each with a length of . Apply a Hanning window to each data segment. To suppress spectral leakage, the discrete Fourier transform of that segment is then calculated. . No. The power spectrum of the segment is estimated as follows Finally, for all The power spectral density is obtained by averaging the power spectral estimates of the segments. ={ } and map it to the corresponding frequency axis superior.

[0079] (4) Calculate the first-order forward difference of the power spectrum sequence: (10) Indicates at frequency point The trend of spectral value changes at that location >0 indicates an increase in spectral value; <0 indicates a decrease in spectral value; =0 means the spectral value remains unchanged.

[0080] (5) Peak point identification: The peak point (local maximum) satisfies the following condition: it increases on the left and decreases on the right, meaning the difference value changes from positive to negative. At this point, the peak point index is... m satisfy: (11) (6) Determination of valid peak points, i.e. Figure 2 The effective peak filtering shown: The power spectral density value at the peak point is greater than a set threshold if and only if P th When the time is reached, it is recorded as the effective peak point.

[0081] (12)

[0082] P th It should be greater than 0 to filter out minute fluctuations caused by noise; its specific value can be adjusted based on the actual signal-to-noise ratio of the data. In this invention, it is set to 1.0 × 10⁻⁶. -10 .

[0083] (7) Calculation of the natural modal frequencies of the vehicle body: Based on the above calculations, the set of valid peak point indices can be written in the following form: (13) Extract the peak frequency corresponding to the effective peak point; this frequency is the vehicle body's natural modal frequency. (14) In the formula, f s Sampling frequency, l This represents the number of points in the Fast Fourier Transform.

[0084] Compared with the prior art, the vehicle body modal frequency recognition method based on oncoming aerodynamic excitation provided in this embodiment of the invention has the following significant advantages: (1) The calculation results are real and reliable: The modal frequencies calculated in the embodiments of the present invention are derived from the dynamic response of the vehicle under the actual operating conditions, including the influence of all actual factors such as track, suspension, and aerodynamics. The calculation results are more representative of the real dynamic characteristics of the vehicle, overcoming the drawbacks of laboratory calculations and simulation calculations being divorced from reality.

[0085] (2) The calculation process is made online and normalized: The embodiments of the present invention use the frequent passing events that occur during normal train operation as excitation. Without additional test equipment and production stoppage testing, the long-term, online and normalized calculation of the inherent frequency of the car body can be realized, providing unprecedented convenience for structural health monitoring.

[0086] (3) Superior calculation excitation source: The aerodynamic excitation generated by high-speed passing in the embodiment of the present invention has the outstanding advantages of strong energy and wide bandwidth, which can effectively excite the main modes of the vehicle body from low-frequency rigid body mode to mid-to-high frequency elastic mode, ensuring the integrity of the calculation results and high signal-to-noise ratio.

[0087] This invention also provides a vehicle body modal frequency recognition device based on oncoming vehicle aerodynamic excitation, as described in the following embodiments. Since the principle by which this device solves the problem is similar to the vehicle body modal frequency recognition method based on oncoming vehicle aerodynamic excitation, the implementation of this device can refer to the implementation of the vehicle body modal frequency recognition method based on oncoming vehicle aerodynamic excitation; repeated details will not be elaborated further.

[0088] Figure 6 This is a schematic diagram of the vehicle body modal frequency recognition device based on aerodynamic excitation during oncoming traffic, as described in an embodiment of the present invention. Figure 6 As shown, the device includes: Data acquisition module 01 is used to acquire vehicle vibration response data generated by the target vehicle during a passing event; The data processing module 02 is used to preprocess the vehicle vibration response data and extract the effective data segment containing the complete oncoming vehicle excitation effect. Frequency domain analysis module 03 is used to perform frequency domain transformation processing on the effective data segment to obtain frequency domain characteristics that characterize its energy distribution; Peak extraction module 04 is used to extract at least one characteristic spectral peak from the frequency domain features; The frequency identification module 05 is used to identify the frequency corresponding to the peak value of the characteristic spectrum as the inherent modal frequency of the target vehicle body in the operating state.

[0089] In one embodiment, the vehicle vibration response data can be the vehicle lateral acceleration data; the oncoming traffic event can be automatically identified and screened to eliminate interference from non-oncoming traffic conditions.

[0090] In one embodiment, the preprocessing includes bandpass filtering the data to preserve the vehicle body modal frequency components and suppress high-frequency noise and low-frequency trend terms.

[0091] In one embodiment, the rule for extracting valid data segments is as follows: based on the multiple characteristic peaks that appear in the vibration response data during the passing process, the data in between is extracted from the first preset position before the first characteristic peak and the second preset position after the last characteristic peak.

[0092] In one embodiment, the frequency domain transformation process includes calculating the power spectral density of the effective data segment or employing other time-frequency analysis methods.

[0093] In one embodiment, the extraction of the characteristic spectral peak includes: calculating the first-order forward difference of the frequency domain feature, identifying the local maxima where the sign changes from positive to negative, and filtering the local maxima based on a preset amplitude threshold to obtain the characteristic spectral peak.

[0094] In one embodiment, the inherent modal frequencies include the rigid body modal frequencies and elastic modal frequencies of the vehicle body.

[0095] 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 vehicle body modal frequency recognition method based on oncoming aerodynamic excitation.

[0096] This invention also provides a computer-readable storage medium storing a computer program that, when executed by a processor, implements the above-described vehicle body modal frequency recognition method based on oncoming aerodynamic excitation.

[0097] This invention also provides a computer program product, which includes a computer program that, when executed by a processor, implements the above-described vehicle body modal frequency recognition method based on oncoming aerodynamic excitation.

[0098] Compared with existing technologies that suffer from high costs and low accuracy in obtaining modal parameters, the vehicle body modal frequency identification scheme based on aerodynamic excitation during train encounters provided in this invention involves: acquiring vehicle body vibration response data of the target vehicle during train encounters; preprocessing the vibration response data and extracting effective data segments; performing frequency domain transformation on the extracted effective data segments; extracting at least one characteristic spectral peak from the frequency domain energy distribution characteristics; and determining the frequency corresponding to the characteristic spectral peak as the inherent modal frequency of the target vehicle body under operating conditions. This invention utilizes the frequent train encounter events during normal train operation as a natural excitation source, enabling long-term, online, and low-cost acquisition of the vehicle body's inherent frequencies that truly reflect the vehicle's operational dynamics without additional testing. This provides an efficient and reliable new method for vehicle condition monitoring and health assessment.

[0099] 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.

[0100] 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.

[0101] 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.

[0102] 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.

[0103] 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 the modal frequencies of a railway vehicle body, characterized in that, include: Acquire vehicle body vibration response data generated by the target vehicle during a passing event; The vehicle vibration response data is preprocessed, and a valid data segment containing the complete oncoming vehicle excitation effect is extracted. The effective data segment is subjected to frequency domain transformation to obtain frequency domain features characterizing its energy distribution; Extract at least one characteristic spectral peak from the frequency domain features; The frequency corresponding to the peak value of the characteristic spectrum is identified as the inherent modal frequency of the target vehicle body in the operating state.

2. The method according to claim 1, characterized in that, The vehicle vibration response data is the vehicle lateral acceleration data; the passing events are automatically identified and screened to eliminate interference from non-passing conditions.

3. The method according to claim 1, characterized in that, The preprocessing includes bandpass filtering of the data to preserve the vehicle body modal frequency components and suppress high-frequency noise and low-frequency trend terms.

4. The method according to claim 1, characterized in that, The rule for extracting valid data segments is as follows: based on the multiple characteristic peaks that appear in the vibration response data during the passing process, the data in between is extracted from the first preset position before the first characteristic peak and the second preset position after the last characteristic peak.

5. The method according to claim 1, characterized in that, The frequency domain transformation process includes calculating the power spectral density of the effective data segment or using other time-frequency analysis methods.

6. The method according to claim 1, characterized in that, The extraction of the feature spectrum peak includes: calculating the first-order forward difference of the frequency domain feature, identifying the local maxima where the sign changes from positive to negative, and filtering the local maxima based on a preset amplitude threshold to obtain the feature spectrum peak.

7. The method according to any one of claims 1 to 6, characterized in that, The inherent modal frequencies include the rigid body modal frequencies and elastic modal frequencies of the vehicle body.

8. A railway vehicle body modal frequency identification device, characterized in that, include: The data acquisition module is used to acquire vehicle vibration response data generated by the target vehicle during a passing event; The data processing module is used to preprocess the vehicle vibration response data and extract the effective data segment containing the complete oncoming vehicle excitation effect. The frequency domain analysis module is used to perform frequency domain transformation processing on the effective data segment to obtain frequency domain characteristics that characterize its energy distribution; Peak extraction module, used to extract at least one characteristic spectral peak from the frequency domain features; The frequency identification module is used to identify the frequency corresponding to the peak value of the characteristic spectrum as the inherent modal frequency of the target vehicle body in the operating state.

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, it implements the method of any one of claims 1 to 7.

10. 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.

Images