Turnout area driving state monitoring method and system and storage medium

Abstract

The embodiment of the application provides a kind of switch area driving state monitoring method, system and storage medium, belong to rail transit operating state monitoring technical field.The method comprises: based on the vibration signal generated by rail in the process of train passing switch area along the line laid vibration sensor of rail, to obtain the wide frequency vibration signal corresponding to the process of train passing switch;The wide frequency vibration signal is preprocessed to obtain the effective vibration signal for state analysis;Effective vibration signal is analyzed in time-frequency, determine the time information of each wheel of train through measuring point, and determine the running speed of train through switch area;Effective vibration signal is analyzed in frequency domain, extract low-frequency vibration characteristics, and judge whether there is car shaking phenomenon when train passes through switch area by combining the running speed of train through switch area, to obtain the corresponding driving state monitoring result.The scheme of the application realizes the automatic identification of car shaking phenomenon when train passes through switch area and driving state monitoring.

Description

Technical Field

[0001] This invention relates to the field of rail transit operation status monitoring technology, specifically to a method for monitoring train operation status in a turnout area, a system for monitoring train operation status in a turnout area, and a storage medium. Background Technology

[0002] Railway turnouts, as key infrastructure for enabling train switching and cross-track operation, have complex structures, including various critical components such as switch rails, stock rails, and frog rails. Their stress states and geometric relationships constantly change with track conditions and train operation. When trains pass through the turnout area, the wheel-rail relationship becomes significantly more complex, easily leading to wheel-rail impact, increased vibration, and changes in train operating conditions. The operating condition of the turnout not only affects the service stability of the track structure but also has a significant impact on train running smoothness and operational safety. Therefore, effective monitoring of train operating conditions in the turnout area is one of the important research directions in railway operation and maintenance.

[0003] As railway transportation develops towards high speed, heavy load, and intelligent operation, the operating conditions in switch areas are becoming more complex, and the dynamic impact and swaying phenomena of trains passing through switches are becoming more prominent. Traditional methods relying on manual inspection and experience-based judgment are insufficient to reflect the operational characteristics of trains passing through switches in a timely and objective manner, gradually revealing problems such as low detection efficiency, strong subjectivity, and insufficient real-time performance.

[0004] To improve the automation and informatization of turnout area operation and maintenance, various detection methods for track or turnout structural conditions have been proposed in existing technologies. One type of existing technology mainly relies on optical sensing methods to collect geometric information of the track or turnout structure, such as using lidar to acquire track contour or spatial point cloud data to achieve detection and analysis of track geometry. This type of method has the advantages of being non-contact and having high measurement accuracy, but it has high requirements for equipment installation conditions, environmental adaptability, and system calibration accuracy. Moreover, in situations where the turnout area structure is complex and there are many obstructions, data processing and effective feature extraction are quite difficult.

[0005] Another type of existing technology is based on machine vision principles. It uses camera devices to collect image information of key parts of the turnout and combines image processing or deep learning algorithms to identify and judge the key structural state of the turnout. This type of method automates the detection process to a certain extent, but its detection effect is easily affected by on-site factors such as changes in lighting, pollution, and vibration. It places high demands on image quality and algorithm robustness, and its stability still has certain limitations in complex driving environments.

[0006] In addition, some existing technologies analyze train operation status from the perspective of train operation response by collecting acceleration signals or car body vibration characteristics during the train's passage through turnouts. For example, by analyzing the lateral or vertical vibration characteristics of the vehicle, it can determine whether there is abnormal shaking in the train. This type of method can reflect the dynamic response characteristics of the train when passing through a turnout to a certain extent, but the detection results are usually affected by multiple factors such as train type, operating speed, and track conditions, and the stability still needs to be further improved.

[0007] With the development of information technology, some existing technologies attempt to introduce multimodal data fusion methods to jointly analyze image information, sensor data, and historical operation and maintenance data to improve the accuracy and reliability of turnout anomaly identification. However, such methods usually rely on the collaborative work of multiple hardware devices, resulting in a relatively complex system structure, high requirements for data synchronization and fusion algorithms, and relatively high engineering implementation and maintenance costs, which is not conducive to long-term and large-scale promotion and application in turnout areas.

[0008] In summary, existing turnout area condition monitoring technologies have made some progress in track geometry detection, visual inspection, and train vibration response analysis. However, under complex operating conditions, there are still problems such as insufficient stability of detection results and the need to improve engineering practicality in order to stably, in real time, and accurately identify the operating status of trains passing through turnout areas. Summary of the Invention

[0009] The purpose of this invention is to provide a method, system, and storage medium for monitoring train operation status in turnout areas, so as to at least solve the problem that existing technologies cannot accurately identify train operation status in turnout areas based on rail vibration signals.

[0010] To achieve the above objectives, the first aspect of the present invention provides a method for monitoring train operation status in a turnout area. The method includes: collecting vibration signals generated by the rails during a train passing through a turnout based on vibration sensors deployed along the rails in the turnout area, to obtain a broadband vibration signal corresponding to the train passing through the turnout; preprocessing the broadband vibration signal to obtain an effective vibration signal for status analysis; performing time-frequency analysis on the effective vibration signal to determine the time information of each train wheel passing through the measuring point, and determining the train's operating speed through the turnout area based on the time information; performing frequency domain feature analysis on the effective vibration signal to extract low-frequency vibration features, and determining whether there is a train swaying phenomenon when the train passes through the turnout area in combination with the train's operating speed through the turnout area, so as to obtain the corresponding train operation status monitoring result.

[0011] Optionally, the vibration sensor is deployed in the wheel-load transition area of ​​the rail in the turnout area; wherein, the wheel-load transition area includes the base rail position in the switch area and the wing rail position in the frog area; the vibration sensor is fixedly installed at the web or bottom of the corresponding rail, and at least one measuring point is deployed along the train running direction to collect the vibration signals generated when the train wheels pass through each measuring point in sequence.

[0012] Optionally, the broadband vibration signal is preprocessed to obtain a valid vibration signal for condition analysis, including: performing baseline correction processing on the broadband vibration signal to eliminate the DC component in the signal; performing filtering processing on the vibration signal after baseline correction processing; segmenting the filtered vibration signal according to the time range of the train passing the measuring point to extract the vibration signal segment corresponding to the train passing process; and using the vibration signal segment as a valid vibration signal for condition analysis.

[0013] Optionally, time-frequency analysis is performed on the effective vibration signal to determine the time information of each train wheel passing through the measuring point, including: performing synchronous compressed wavelet transform processing on the effective vibration signal to obtain time-frequency distribution information of the vibration signal in the time and frequency dimensions; extracting wavelet coefficient amplitudes within a preset frequency band based on the time-frequency distribution information, and performing energy integration calculation on the wavelet coefficient amplitudes within the preset frequency band to obtain the wavelet energy index; performing peak detection on the wavelet energy index in the time dimension to obtain the time sequence information of the time when each train wheel passes through the measuring point, as the time information of each train wheel passing through the measuring point.

[0014] Optionally, synchronous compressed wavelet transform processing is performed on the effective vibration signal to obtain time-frequency distribution information of the vibration signal in the time and frequency dimensions, including: performing continuous wavelet transform processing on the effective vibration signal to obtain wavelet coefficients corresponding to each time position and each frequency position; calculating instantaneous frequency information corresponding to each time position and each frequency position based on the wavelet coefficients; performing frequency redistribution processing on the wavelet coefficients according to the instantaneous frequency information to map the energy to the corresponding instantaneous frequency position; and constructing a time-frequency distribution matrix of the vibration signal based on the redistributed wavelet coefficients to obtain time-frequency distribution information of the vibration signal in the time and frequency dimensions.

[0015] Optionally, determining the train's speed through the turnout area based on the time information includes: determining the time sequence of each wheel of the train passing through the measuring point in sequence according to the time information; determining the time when the first wheel of the train passes through the measuring point and the time when the last wheel of the train passes through the measuring point based on the time sequence, and calculating the corresponding time difference; obtaining the spatial distance between the first wheel and the last wheel of the train; and calculating the train's speed through the turnout area based on the spatial distance and the time difference.

[0016] Optionally, frequency domain feature analysis is performed on the effective vibration signal to extract low-frequency vibration features, including: performing Fourier transform processing on the effective vibration signal to obtain the spectral information of the vibration signal in the frequency dimension; selecting spectral data corresponding to a preset low-frequency band from the spectral information; and accumulating the spectral amplitude within the preset low-frequency band to obtain the low-frequency vibration features of the corresponding train passing through the measuring point.

[0017] Optionally, the system combines the train's operating speed through the switch area to determine whether there is train swaying when the train passes through the switch area, in order to obtain corresponding train operation status monitoring results. This includes: determining the low-frequency vibration intensity parameters of the corresponding train passing through the measuring point based on the low-frequency vibration characteristics; performing speed normalization processing on the low-frequency vibration intensity parameters based on the train's operating speed through the switch area to obtain normalized vibration characteristic parameters; comparing the normalized vibration characteristic parameters with a preset swaying judgment threshold; and determining that there is train swaying when the train passes through the switch area when the normalized vibration characteristic parameters are greater than the preset swaying judgment threshold, and generating corresponding train operation status monitoring results.

[0018] A second aspect of the present invention provides a train operation status monitoring system for a turnout area. The system includes: a data acquisition unit, used to acquire vibration signals generated by the rails during the train's passage through the turnout based on vibration sensors deployed along the rails in the turnout area, to obtain a broadband vibration signal corresponding to the train's passage through the turnout; a preprocessing unit, used to preprocess the broadband vibration signal to obtain an effective vibration signal for status analysis; a speed determination unit, used to perform time-frequency analysis on the effective vibration signal to determine the time information of each wheel of the train passing through the measuring point, and to determine the train's operating speed through the turnout area based on the time information; and a result output unit, used to perform frequency domain feature analysis on the effective vibration signal, extract low-frequency vibration features, and, in conjunction with the train's operating speed through the turnout area, determine whether there is a train swaying phenomenon when the train passes through the turnout area, so as to obtain the corresponding train operation status monitoring result.

[0019] On the other hand, the present invention provides a computer-readable storage medium storing instructions that, when executed on a computer, cause the computer to perform the above-described method for monitoring the train operation status in the turnout area.

[0020] Through the above technical solution, this invention collects vibration signals generated during train crossings by deploying vibration sensors along the rails in the turnout area. The collected broadband vibration signals are preprocessed, subjected to time-frequency analysis, and frequency domain feature analysis to extract wheel passage time information, train speed, and low-frequency vibration characteristics from the vibration signals. Based on this, and combined with the train speed, the invention determines the swaying phenomenon during train crossings, thereby achieving automatic identification and monitoring of train operation status in the turnout area. By utilizing rail vibration signals for operation status analysis, it eliminates the need for visual equipment or complex detection systems. It can reliably acquire train operation information even in complex turnout structures and vibration environments, improving the accuracy and engineering adaptability of turnout area operation status monitoring.

[0021] Other features and advantages of the embodiments of the present invention will be described in detail in the following detailed description section. Attached Figure Description

[0022] The accompanying drawings are provided to further illustrate embodiments of the present invention and form part of the specification. They are used together with the following detailed description to explain the embodiments of the present invention, but do not constitute a limitation thereof. In the drawings: Figure 1 This is a flowchart of the steps of a turnout area train operation status monitoring method provided by one embodiment of the present invention; Figure 2 This is a schematic diagram of the layout of vibration sensors in the turnout area provided by one embodiment of the present invention; Figure 3 This is a schematic diagram illustrating the determination of the time taken for each wheel to pass by using the WEI energy index, provided by one embodiment of the present invention. Figure 4 This is a system structure diagram of a turnout area train operation status monitoring system provided in one embodiment of the present invention. Detailed Implementation

[0023] The specific embodiments of the present invention will be described in detail below with reference to the accompanying drawings. It should be understood that the specific embodiments described herein are for illustration and explanation only and are not intended to limit the present invention.

[0024] Figure 1 This is a flowchart illustrating the steps of a turnout area train operation status monitoring method according to one embodiment of the present invention. Figure 1 As shown, an embodiment of the present invention provides a method for monitoring the train operation status in a turnout area, the method comprising: Step S10: Based on the vibration sensors deployed along the rails in the turnout area, collect the vibration signals generated by the rails during the train's passage through the turnout to obtain the broadband vibration signal corresponding to the train's passage through the turnout.

[0025] Specifically, the vibration sensor is deployed in the wheel-load transition area of ​​the rail in the turnout area; wherein, the wheel-load transition area includes the base rail position in the switch area and the wing rail position in the frog area; the vibration sensor is fixedly installed at the web or bottom of the corresponding rail, and at least one measuring point is deployed along the train running direction to collect the vibration signal generated when the train wheels pass through each measuring point in sequence.

[0026] In this embodiment of the invention, vibration sensors are deployed along the rails in the turnout area to collect the vibration response generated by the rails during the train's passage through the turnout, thereby obtaining the broadband vibration signal corresponding to the train's passage through the turnout. Compared to ordinary track sections, turnout structures have a more complex wheel-rail relationship. When the train wheels enter the switch area and frog area, the wheel load transitions between different rails, and the wheel-rail contact state changes significantly. This process is often accompanied by impact vibration, enhanced structural response, and the superposition of multiple frequency vibration components. It is by utilizing this physical characteristic that the operating status of the train during the turnout passage can be indirectly reflected by collecting and analyzing the rail vibration signals.

[0027] Considering the structural characteristics of the turnout area, this embodiment deploys vibration sensors in the wheel-load transition zone of the rails. The wheel-load transition zone typically corresponds to the area where the train wheel transitions from the stock rail / switcher rail or the wing rail / spot rail. At this location, the wheel-rail contact relationship changes significantly, and the vibration signal response is more sensitive. In specific implementations, vibration sensors can be deployed at the stock rail position in the switch area and the wing rail position in the frog area. These locations are typical areas of concentrated dynamic response when the train passes through the turnout, thus enabling relatively stable acquisition of signals reflecting wheel-rail impact and structural vibration characteristics.

[0028] In terms of installation, the vibration sensor is fixedly installed on the web or base of the corresponding rail. The web and base areas possess good structural rigidity and signal transmission characteristics, enabling them to comprehensively reflect the vibration response of the rail under train loads. This also facilitates reliable sensor installation and minimizes the impact on train safety. In practical engineering applications, methods such as bolt fixing, clamping structures, adhesive bonding, or dedicated mounting brackets can be used to stably install the vibration sensor on the rail structure, ensuring a stable mechanical coupling between the sensor and the rail during signal acquisition.

[0029] To capture the vibration changes generated as train wheels pass by in sequence, at least one measuring point is set up along the rail in the direction of train travel. Using this signal data, the timing of each wheel's passage through the measuring point can be further identified, providing fundamental data support for subsequent train speed calculations and train status identification.

[0030] Vibration signals are synchronously acquired using data acquisition equipment. Considering that the train's passage through the turnout involves both low-frequency vibration responses and high-frequency vibration components caused by wheel-rail impact, the sampling frequency of the data acquisition system is set to cover the aforementioned vibration frequency range, thereby acquiring broadband vibration signals. In this embodiment, the acquired vibration signals include at least vertical acceleration signals and lateral acceleration signals. Vertical vibration reflects wheel-rail impact and track structure response, while lateral vibration is related to train running stability and swaying. These two types of signals together constitute an important source of information regarding the vibration characteristics of the train passing through the turnout.

[0031] By employing the aforementioned deployment method and signal acquisition technique, the rail vibration response during train crossing of turnouts can be obtained without interfering with normal train operation, providing a reliable data foundation for subsequent train operation status identification based on vibration signals. The entire acquisition process is simple in structure, highly adaptable to various field conditions, and can be stably implemented in various turnout scenarios, including tunnels, bridges, and open-air lines.

[0032] In one specific implementation, such as Figure 2 As shown, multiple vibration sensors are installed along the rails in the turnout area to collect vibration response signals generated by the rails when a train passes through the turnout. Taking the switch area as an example, multiple measuring points SP1, SP2, SP3, SP4, SP6, and SP7 are installed along the main rail in the direction of train operation, and measuring point SP5 is set at the corresponding position on the other rail; measuring points SP8 and SP9 are set at a predetermined distance from the top surface of the rail. Among the above measuring points, SP1, SP2, SP3, SP5, SP6, and SP8 are equipped with vertical vibration sensors to collect the vertical vibration response of the rails; SP4, SP7, and SP9 are equipped with lateral vibration sensors to collect the lateral vibration response of the rails, reflecting possible lateral swaying or instability when a train passes through the turnout. The vibration sensors are installed at the web or bottom of the rails and are stably connected to the rails through a fixed installation structure to ensure that the vibration signal can be effectively transmitted to the sensor.

[0033] Furthermore, multiple vibration acquisition points are also set up in the frog area. Points CP1, CP2, and CP3 (not shown in the figure) are deployed on the wing rail and its adjacent rail area, and vibration sensors are installed at the corresponding locations to collect the vibration response generated when the train wheels pass through the frog area. By deploying multiple acquisition points in the switch and frog areas, vibration signals at different locations can be obtained as the train passes each acquisition point sequentially along the track direction. The vibration data collected from each acquisition point are synchronously recorded by data acquisition equipment, thus forming continuous vibration data of the train passing through the frog, providing basic data for subsequent identification of wheel passage time, calculation of train speed, and determination of whether the train is swaying.

[0034] Step S20: Preprocess the broadband vibration signal to obtain an effective vibration signal for state analysis.

[0035] Specifically, baseline correction processing is performed on the broadband vibration signal to eliminate the DC component in the signal; filtering processing is performed on the vibration signal after baseline correction processing; the filtered vibration signal is segmented according to the time range of the train passing the measuring point to extract the vibration signal segments corresponding to the train passing process; the vibration signal segments are used as valid vibration signals for state analysis.

[0036] In this embodiment of the invention, the acquired broadband vibration signal is preprocessed to obtain an effective vibration signal for subsequent state analysis. When a train passes a turnout, the rail vibration signal typically contains baseline drift, environmental noise, and frequency components unrelated to the analysis target. If directly used for feature extraction, it can easily affect the accuracy of subsequent speed identification and sway judgment. Therefore, before performing time-frequency analysis on the vibration signal, it is necessary to preprocess the original signal to improve signal quality and enhance the consistency between data from different measurements. Let the original vibration signal acquired by the vibration sensor be... ,in, Represents a time variable. This represents the acceleration value at the corresponding moment. Due to the influence of acquisition equipment or environmental factors, the raw signal usually contains a DC component. To eliminate this influence, baseline correction processing is performed on the raw signal, that is, the DC component is eliminated by removing the average value of the signal. The processing procedure can be expressed as:

[0037] in,

[0038] Indicates the time interval The average value of the internal signal, This represents the vibration signal after baseline correction. Through the above processing, baseline drift in the signal can be effectively eliminated, causing the vibration signal to be distributed around zero.

[0039] After baseline correction, the vibration signal is filtered to suppress noise components. Considering that the vibration signal during train turnout crossing contains both low-frequency vibration components and higher-frequency vibration components generated by wheel-rail impact, a low-pass filter is typically used to suppress high-frequency noise during the filtering process. The filtered signal can be expressed as:

[0040] Among them, symbols This represents the convolution operation. This represents the unit impulse response function of the filter. This represents the vibration signal after filtering.

[0041] After obtaining the filtered vibration signal, the signal is segmented according to the time range of the train passing the measuring point to extract the vibration signal segments corresponding to the train's passage. Let the starting time of the train passing the measuring point be... The end time is The corresponding vibration signal segment can be represented as:

[0042] in, This represents a segment of the captured vibration signal. This signal segment covers at least the entire process of the train wheel entering, passing through, and leaving the measuring point, thus fully reflecting the vibration response characteristics of the rail during the train's passage through the turnout.

[0043] Through the baseline correction, filtering, and signal segmentation processes described above, an effective vibration signal reflecting the train's turnout passage process can be extracted from the original broadband vibration signal. This effective vibration signal serves as the basic data input for subsequent time-frequency analysis, wheel passage time identification, and train operation status judgment, thereby ensuring the stability and reliability of the subsequent analysis results.

[0044] Step S30: Perform time-frequency analysis on the effective vibration signal to determine the time information of each train wheel passing through the measuring point, and determine the train's running speed through the switch area based on the time information.

[0045] Specifically, time-frequency analysis is performed on the effective vibration signal to determine the time information of each train wheel passing through the measuring point. This includes: performing synchronous compressed wavelet transform processing on the effective vibration signal to obtain time-frequency distribution information of the vibration signal in the time and frequency dimensions; extracting wavelet coefficient amplitudes within a preset frequency band based on the time-frequency distribution information, and performing energy integration calculation on the wavelet coefficient amplitudes within the preset frequency band to obtain the wavelet energy index; and performing peak detection on the wavelet energy index in the time dimension to obtain the time sequence information of the time when each train wheel passes through the measuring point, which serves as the time information of each train wheel passing through the measuring point.

[0046] Furthermore, synchronous compressed wavelet transform processing is performed on the effective vibration signal to obtain time-frequency distribution information of the vibration signal in the time and frequency dimensions. This includes: performing continuous wavelet transform processing on the effective vibration signal to obtain wavelet coefficients corresponding to each time position and each frequency position; calculating instantaneous frequency information corresponding to each time position and each frequency position based on the wavelet coefficients; performing frequency redistribution processing on the wavelet coefficients according to the instantaneous frequency information to map the energy to the corresponding instantaneous frequency position; and constructing a time-frequency distribution matrix of the vibration signal based on the redistributed wavelet coefficients to obtain time-frequency distribution information of the vibration signal in the time and frequency dimensions.

[0047] Furthermore, determining the train's speed through the turnout area based on the time information includes: determining the time sequence of each wheel of the train passing through the measuring point in sequence according to the time information; determining the time when the first wheel of the train passes through the measuring point and the time when the last wheel passes through the measuring point based on the time sequence, and calculating the corresponding time difference; obtaining the spatial distance between the first wheel and the last wheel of the train; and calculating the train's speed through the turnout area based on the spatial distance and the time difference.

[0048] In this embodiment of the invention, after obtaining the effective vibration signal for state analysis, time-frequency analysis is performed on the effective vibration signal to determine the time information of each train wheel passing through the measuring point, and further, the train's operating speed when passing through the switch zone is calculated based on this time information. Due to characteristics such as wheel load transition, wheel-rail impact, and structural discontinuities, the switch zone causes significant transient vibration responses from the train wheels when passing over the rails. This vibration signal exhibits a clear energy concentration phenomenon in the time-frequency domain. Utilizing this characteristic, features can be extracted from the vibration signal using time-frequency analysis methods, thereby identifying the specific time when the wheels pass through the measuring point.

[0049] In this embodiment, the effective vibration signal Synchrosqueezed Wavelet Transform (SWT) is performed to obtain the time-frequency distribution information of the vibration signal in both the time and frequency dimensions. SWT is a time-frequency analysis method that redistributes frequency based on continuous wavelet transform, effectively improving the energy concentration of the time-frequency representation. Specifically, a continuous wavelet transform is first performed on the effective vibration signal, and its expression can be given as:

[0050] in, Represents the continuous wavelet transform coefficients; Indicates the scale factor; Indicates the time shift factor; Represents the mother wavelet function; The complex conjugate operation is represented by t, which represents the time variable. Through the continuous wavelet transform described above, wavelet coefficients of the vibration signal at different time positions and scales can be obtained.

[0051] Furthermore, the corresponding instantaneous frequency information is calculated based on the continuous wavelet transform coefficients. The instantaneous frequency can be determined by the rate of change of the wavelet coefficients in the time direction, and its expression is:

[0052] in, Indicates the instantaneous frequency corresponding to the time and scale position; These are partial derivatives; The unit is the imaginary unit. Subsequently, based on the instantaneous frequency information, the continuous wavelet transform coefficients undergo frequency redistribution processing, remapping the energy originally distributed along the scale dimension to the corresponding instantaneous frequency positions, thereby obtaining synchronously compressed wavelet coefficients with more concentrated energy. After completing the frequency redistribution described above, a time-frequency distribution matrix of the vibration signal in the time and frequency dimensions can be constructed, thereby obtaining the time-frequency distribution information of the vibration signal.

[0053] After obtaining the time-frequency distribution information, in order to identify the specific time when the wheel passes through the measuring point, a preset frequency band corresponding to the wheel-rail impact characteristics is selected from the time-frequency distribution information. In this embodiment, the frequency band can be selected as 120-200Hz, which typically contains significant wheel-rail impact energy. It should be noted that this frequency band is not the only possible value; depending on different line conditions or vehicle types, adjacent frequency bands, such as 110-220Hz, can be selected based on the actual situation.

[0054] Within the preset frequency band, the energy integral of the wavelet coefficient amplitudes obtained by synchronous compressed wavelet transform is calculated to obtain the wavelet-based energy index (WEI). Its expression is:

[0055] in, Indicates time The corresponding wavelet energy index; This represents the absolute value of the wavelet coefficients obtained after synchronous compressed wavelet transform processing; and These represent the lower and upper frequency limits of the selected frequency band, respectively. By summing the wavelet coefficients within this frequency band, the variation curve of vibration energy over time can be obtained.

[0056] like Figure 3As shown, when the train wheels pass the measuring points in sequence, the vibration energy generated by the wheel-rail impact will... A distinct peak is formed on the curve. Through analysis... Peak detection on the curve can obtain the time series information of the corresponding wheel passing through the measuring point. For example, when A wheel passing event can be considered detected when the value exceeds a preset threshold (e.g., 15). For example, if 16 maximum points are identified during a detection process, it means that 16 wheels passed the measurement point. If each carriage contains 4 wheels, then the corresponding train consists of 4 carriages.

[0057] After obtaining the time information of each wheel passing through the measuring point, the train's operating speed when passing through the switch area can be further determined. Specifically, the time sequence of each wheel passing through the measuring point is determined based on the time information. .in, This indicates the time when the first wheel passes the measuring point. This indicates the time when the last wheel passes the measuring point. The time difference between the two is calculated as follows:

[0058] This allows us to obtain the total time interval between the train passing the measuring point. Furthermore, we can obtain the spatial distance between the first and last wheels of the train. This distance can be determined by the train formation structure parameters or the wheelbase information of standard carriages. From this, the train's speed passing through the turnout area can be calculated.

[0059] in, This indicates the train's operating speed. Using the method described above, the energy characteristics of rail vibration signals in the time-frequency domain can be used to accurately identify the time of wheel passage and calculate the train's operating speed accordingly. Compared to traditional methods that rely solely on trackside speed measurement equipment or vehicle signaling systems, this method directly utilizes the rail vibration response to obtain train operating information, eliminating the need for additional speed measurement equipment. It features simple structure, convenient installation, and strong environmental adaptability. Furthermore, by using synchronous compressed wavelet transform to enhance the concentration of time-frequency energy, the wheel passage event becomes clearer in the energy curve, thereby improving the accuracy of wheel identification and speed calculation.

[0060] Step S40: Perform frequency domain feature analysis on the effective vibration signal, extract low-frequency vibration features, and combine the train's running speed through the switch area to determine whether there is a swaying phenomenon when the train passes through the switch area, so as to obtain the corresponding train operation status monitoring results.

[0061] Specifically, frequency domain feature analysis is performed on the effective vibration signal to extract low-frequency vibration features, including: performing Fourier transform processing on the effective vibration signal to obtain the spectral information of the vibration signal in the frequency dimension; selecting spectral data corresponding to a preset low-frequency band from the spectral information; and accumulating the spectral amplitude within the preset low-frequency band to obtain the low-frequency vibration features of the corresponding train passing through the measuring point.

[0062] Furthermore, the system determines whether train swaying occurs when the train passes through the switch area based on the train's operating speed, thereby obtaining corresponding train operation status monitoring results. This includes: determining the low-frequency vibration intensity parameters of the corresponding train passing through the measuring point based on the low-frequency vibration characteristics; performing speed normalization processing on the low-frequency vibration intensity parameters based on the train's operating speed through the switch area to obtain normalized vibration characteristic parameters; comparing the normalized vibration characteristic parameters with a preset swaying judgment threshold; and determining that train swaying occurs when the normalized vibration characteristic parameters are greater than the preset swaying judgment threshold, and generating corresponding train operation status monitoring results when the train passes through the switch area.

[0063] In this embodiment of the invention, when a train passes through a turnout area, due to changes in wheel-rail contact, discontinuities in the track structure, and transitions in wheel load between different rails, the train body and bogie may experience a certain degree of lateral swaying or enhanced vibration. This phenomenon typically manifests as a significant increase in low-frequency vibration energy in the rail vibration signal; therefore, the swaying state can be identified by analyzing the frequency domain characteristics of the vibration signal.

[0064] In this embodiment, the effective time interval for each carriage (or bogie) to pass the measuring point is first determined based on the wheel passing time information obtained in step S3. Let the first... The effective time interval for each carriage to pass the measuring point is: Within this time interval, the corresponding acceleration signal segment is extracted from the original vibration signal to obtain:

[0065] in, This indicates the rail vibration acceleration signal. Indicates the first The vibration signal segment corresponding to the passing of each train car through the measuring point. This signal segment reflects the vibration response of the rail structure when the train car passes through the switch area.

[0066] Subsequently, a Fourier transform is performed on the vibration signal segment to obtain the spectral information of the vibration signal in the frequency dimension. Its expression is:

[0067] in, Indicates the first The vibration signal of each carriage is at the frequency Spectral amplitude at that location Indicates the number of signal sampling points. Represents discrete frequency points. This represents the amplitude of the nth sampling point in the vibration signal segment captured when the i-th carriage passes the measuring point. It is the imaginary unit.

[0068] Considering that train swaying typically manifests as increased vibration in the lower frequency range, a preset low-frequency band is selected from the spectral information. Analysis is performed. In this embodiment, the low-frequency band can be set as f1=1Hz, f2=20Hz, and the corresponding discrete frequency set is:

[0069] Within this frequency band, the spectral amplitudes are accumulated to obtain the low-frequency vibration intensity parameters of the corresponding carriage:

[0070] in, Indicates the first The low-frequency vibration intensity parameter of each carriage when it passes the measuring point. This parameter reflects the energy distribution of the vibration signal in the low-frequency range.

[0071] Furthermore, to reduce the impact of train speed on vibration intensity, the low-frequency vibration intensity parameters can be normalized based on the train's speed when passing through the switch area. Let the train speed be... Then the normalized vibration characteristic parameters can be expressed as:

[0072] in, This represents the vibration characteristic parameters after speed normalization. By introducing speed normalization, the consistency of vibration characteristic evaluation can be maintained under different train operating speed conditions, thereby improving the reliability of train sway detection.

[0073] After obtaining the normalized vibration characteristic parameters, they are compared with the preset sway detection threshold. Compare. When satisfied... At that time, it can be determined that the first The train carriages exhibited significant swaying when passing through the switch area. This determination process can also be expressed as an indicator function:

[0074] in, This indicates that a vehicle shaking phenomenon was detected. This indicates that no significant vehicle swaying was detected.

[0075] Using the above method, the energy variation characteristics of vibration signals in the low-frequency band can be used to determine the operational stability of trains passing through switch areas. Since swaying is usually accompanied by an increase in low-frequency vibration energy, statistical analysis of the spectral amplitude within the 1-20Hz band, combined with normalization based on train speed, can more stably identify abnormal vibration states when trains pass through switch areas, thus obtaining corresponding train operation status monitoring results. This method can identify swaying using only rail vibration signals, without relying on onboard equipment or additional trackside detection systems, and features simple structure, strong environmental adaptability, and good monitoring continuity.

[0076] In another possible implementation, during actual line operation, the turnout structure will gradually experience wear, gauge changes, or local structural loosening as the service time increases. These changes may cause a slow drift in the overall energy level of the vibration signal, thereby affecting the accuracy of the fixed threshold determination method.

[0077] Therefore, in this embodiment, a historical vibration feature database is introduced to record the vibration statistical characteristics of the same turnout area in different operating cycles. Specifically, after each train passes through the turnout, the low-frequency vibration intensity parameters and their normalized speed parameters are statistically analyzed to form a corresponding operating cycle feature vector. The system performs sliding statistical analysis on the feature vector within a preset time window to obtain a long-term benchmark level of the vibration response in the turnout area. When a continuous deviation between the current vibration characteristics and the historical benchmark level is detected, the sway judgment threshold is adaptively adjusted, thereby forming judgment parameters that are dynamically updated according to the turnout's service status.

[0078] This method enables the turnout sway recognition model to maintain stable performance under long-term turnout operation conditions, and can also indirectly reflect the changing trend of the turnout structure's service status, thus providing additional reference information for turnout maintenance and track condition assessment.

[0079] In another possible implementation, in the actual operating environment of the turnout area, due to the influence of factors such as wheel-rail contact state, track structure, and train load distribution, the vibration signals collected at different measuring points may have certain differences in amplitude and frequency distribution. To avoid the judgment results being affected by the local structural state or environmental noise interference of a single measuring point, this implementation uses multiple vibration measuring points deployed along the track direction to jointly analyze the same train passing process.

[0080] Specifically, the low-frequency vibration intensity parameters are calculated as the train passes each measuring point, and the vibration characteristics of each measuring point are normalized according to the train's speed. Then, the consistency of the vibration characteristic parameters from multiple measuring points is compared. When multiple measuring points show significant enhancement of low-frequency vibration during the passage of the same train carriage, the swaying phenomenon can be determined to have a holistic characteristic; if an anomaly occurs only at a single measuring point, it may be a local structural response or an occasional disturbance. This multi-measuring-point vibration consistency analysis method can reduce the probability of misjudgment, improve the stability of the train operation status monitoring results in the turnout area, and more accurately reflect the overall operating status of the train when passing through the turnout.

[0081] In one specific application embodiment, by performing feature analysis on vibration signals collected during train passage through a turnout area and obtaining corresponding train operation status monitoring results, the operational stability of the train during turnout passage can be objectively assessed. When the monitoring results show obvious train swaying, it indicates an abnormal increase in wheel-rail dynamic response when the train passes through the turnout area, which may be related to factors such as changes in the turnout structure, changes in track conditions, or changes in train operating conditions. By continuously recording and statistically analyzing the train operation status monitoring results, it can be further used to identify the changing trends of train operation status in the turnout area, providing a reference for line maintenance personnel. At the same time, the monitoring results can also serve as auxiliary information for turnout area operation status assessment and line maintenance decisions, such as determining key inspection sections, optimizing maintenance plans, or conducting further on-site inspections, thereby improving the targeting and efficiency of turnout area maintenance work.

[0082] Figure 4 This is a system structure diagram of a turnout area train operation status monitoring system provided in one embodiment of the present invention. Figure 4 As shown, this invention provides a turnout area train operation status monitoring system. The system includes: a data acquisition unit, used to acquire vibration signals generated by the rails during train passage through the turnout based on vibration sensors deployed along the rails in the turnout area, to obtain broadband vibration signals corresponding to the train passage process; a preprocessing unit, used to preprocess the broadband vibration signals to obtain effective vibration signals for status analysis; a speed determination unit, used to perform time-frequency analysis on the effective vibration signals to determine the time information of each train wheel passing through the measuring points, and to determine the train's operating speed through the turnout area based on the time information; and a result output unit, used to perform frequency domain feature analysis on the effective vibration signals, extract low-frequency vibration features, and, in conjunction with the train's operating speed through the turnout area, determine whether there is a train swaying phenomenon when the train passes through the turnout area, so as to obtain corresponding train operation status monitoring results.

[0083] The present invention also provides a computer-readable storage medium storing instructions that, when executed on a computer, cause the computer to perform the above-described method for monitoring the train operation status in the turnout area.

[0084] Those skilled in the art will understand that all or part of the steps in the methods of the above embodiments can be implemented by a program instructing related hardware. This program is stored in a storage medium and includes several instructions to cause a microcontroller, chip, or processor to execute all or part of the steps of the methods described in the various embodiments of the present invention. The aforementioned storage medium includes various media capable of storing program code, such as a USB flash drive, a portable hard drive, a read-only memory (ROM), a random access memory (RAM), a magnetic disk, or an optical disk.

[0085] The optional embodiments of the present invention have been described in detail above with reference to the accompanying drawings. However, the embodiments of the present invention are not limited to the specific details described above. Within the scope of the technical concept of the embodiments of the present invention, various simple modifications can be made to the technical solutions of the embodiments of the present invention, and these simple modifications all fall within the protection scope of the embodiments of the present invention. It should also be noted that the various specific technical features described in the above specific embodiments can be combined in any suitable manner without contradiction. To avoid unnecessary repetition, the embodiments of the present invention will not further describe the various possible combinations.

[0086] Furthermore, various different embodiments of the present invention can be combined in any way, as long as they do not violate the spirit of the embodiments of the present invention, they should also be regarded as the content disclosed by the embodiments of the present invention.

Claims

1. A method for monitoring train operation status in a turnout area, characterized in that, The method includes: Based on the vibration sensors deployed along the rails in the turnout area, the vibration signals generated by the rails during the train passing through the turnout are collected to obtain the broadband vibration signal corresponding to the train passing through the turnout. The broadband vibration signal is preprocessed to obtain an effective vibration signal for state analysis; Time-frequency analysis is performed on the effective vibration signal to determine the time information of each train wheel passing through the measuring point, and the train speed passing through the switch area is determined based on the time information. Frequency domain feature analysis is performed on the effective vibration signal to extract low-frequency vibration features. Combined with the train's operating speed when passing through the switch area, it is determined whether there is a swaying phenomenon when the train passes through the switch area, so as to obtain the corresponding train operation status monitoring results.

2. The method for monitoring train operation status in the turnout area according to claim 1, characterized in that, The vibration sensors are installed in the wheel-load transition area of ​​the rails in the turnout area; wherein, the wheel-load transition area includes the position of the main rail in the switch area and the position of the wing rail in the frog area; The vibration sensor is fixedly installed on the web or bottom of the corresponding rail, and at least one measuring point is arranged along the direction of train operation to collect the vibration signals generated when the train wheels pass through each measuring point in sequence.

3. The method for monitoring train operation status in the turnout area according to claim 1, characterized in that, The broadband vibration signal is preprocessed to obtain an effective vibration signal for condition analysis, including: Baseline correction processing is performed on the broadband vibration signal to eliminate the DC component in the signal; The vibration signal after baseline correction is filtered. The filtered vibration signal is segmented according to the time range of the train passing the measuring point in order to extract the vibration signal segments corresponding to the train passing process. The vibration signal segment is used as the effective vibration signal for state analysis.

4. The method for monitoring train operation status in the turnout area according to claim 1, characterized in that, Time-frequency analysis is performed on the effective vibration signal to determine the time information of each train wheel passing the measuring point, including: The effective vibration signal is subjected to synchronous compressed wavelet transform processing to obtain the time-frequency distribution information of the vibration signal in the time and frequency dimensions. Based on the time-frequency distribution information, the amplitude of wavelet coefficients within a preset frequency band is extracted, and the energy integral of the wavelet coefficient amplitudes within the preset frequency band is calculated to obtain the wavelet energy index. Peak detection is performed on the wavelet energy index in the time dimension to obtain the time sequence information of the time when each train wheel passes the measuring point, which is used as the time information of each train wheel passing the measuring point.

5. The method for monitoring train operation status in the turnout area according to claim 4, characterized in that, Perform synchronous compressed wavelet transform processing on the effective vibration signal to obtain time-frequency distribution information of the vibration signal in the time and frequency dimensions, including: The effective vibration signal is subjected to continuous wavelet transform processing to obtain wavelet coefficients corresponding to each time position and each frequency position; Based on the wavelet coefficients, calculate the instantaneous frequency information corresponding to each time position and each frequency position; The wavelet coefficients are frequency redistributed based on the instantaneous frequency information to concentrate energy and map it to the corresponding instantaneous frequency position. The time-frequency distribution matrix of the vibration signal is constructed based on the redistributed wavelet coefficients, thereby obtaining the time-frequency distribution information of the vibration signal in the time and frequency dimensions.

6. The method for monitoring train operation status in the turnout area according to claim 4, characterized in that, Determining the train's speed through the switch zone based on the aforementioned time information includes: Based on the aforementioned time information, determine the time sequence in which each wheel of the train passes the measuring point in sequence; Based on the time series, determine the time when the first wheel of the train passes the measuring point and the time when the last wheel passes the measuring point, and calculate the corresponding time difference; Obtain the spatial distance between the first and last wheels of the train; The train's speed through the turnout area is calculated based on the spatial distance and the time difference.

7. The method for monitoring train operation status in a turnout area according to claim 1, characterized in that, Frequency domain feature analysis is performed on the effective vibration signal to extract low-frequency vibration features, including: Perform Fourier transform processing on the effective vibration signal to obtain the spectral information of the vibration signal in the frequency dimension; Select the spectrum data corresponding to the preset low-frequency band from the spectrum information; The spectral amplitudes within the preset low-frequency band are accumulated and calculated to obtain the low-frequency vibration characteristics of the corresponding train passing through the measuring point.

8. The method for monitoring train operation status in a turnout area according to claim 7, characterized in that, Based on the train's operating speed as it passes through the switch zone, the system determines whether train swaying occurs when the train passes through the switch zone, in order to obtain corresponding train operation status monitoring results, including: The low-frequency vibration intensity parameters of the corresponding train passing through the measuring point are determined based on the low-frequency vibration characteristics. The low-frequency vibration intensity parameters are normalized based on the train's operating speed through the turnout area to obtain normalized vibration characteristic parameters. The normalized vibration characteristic parameters are compared with the preset swaying judgment threshold; When the normalized vibration characteristic parameter is greater than the preset swaying judgment threshold, it is determined that there is swaying when the train passes through the switch area, and the corresponding train operation status monitoring result is generated.

9. A turnout area train operation status monitoring system, characterized in that, The system includes: The acquisition unit is used to acquire vibration signals generated by the rails during the train passing through the turnout based on vibration sensors deployed along the rails in the turnout area, so as to obtain the broadband vibration signal corresponding to the train passing through the turnout. The preprocessing unit is used to preprocess the broadband vibration signal to obtain an effective vibration signal for state analysis. The speed determination unit is used to perform time-frequency analysis on the effective vibration signal, determine the time information of each wheel of the train passing the measuring point, and determine the running speed of the train passing through the switch area based on the time information. The result output unit is used to perform frequency domain feature analysis on the effective vibration signal, extract low-frequency vibration features, and determine whether there is a swaying phenomenon when the train passes through the switch area based on the train's running speed, so as to obtain the corresponding train operation status monitoring results.

10. A computer-readable storage medium, characterized in that, The computer-readable storage medium stores instructions that, when executed on a computer, cause the computer to perform the turnout area traffic status monitoring method according to any one of claims 1-8.

Images