A method and system for automatically identifying modal parameters of a turnout area

Abstract

The application provides a turnout area modal parameter automatic identification method and system, relates to the technical field of turnout area modal testing, and comprises the following steps: acquiring first information, wherein the first information comprises size information of a target turnout area and a connection line length of a testing device; performing regional division processing on the target turnout area according to the first information to obtain a testing region division result; sequentially performing excitation and response measurement on each testing region according to the testing region division result and a partitioned MRIT modal testing scheme to obtain excitation-response data; performing matrix construction processing according to the excitation-response data to obtain a modal parameter stability diagram of each testing region; performing clustering processing according to the modal parameter stability diagram to obtain a clustering result; and obtaining a modal parameter identification result based on the clustering result. The application is based on a density-based spatial clustering algorithm, can automatically remove false modal and interference modal clusters, greatly improves the accuracy and efficiency of modal parameter analysis, and finally generates a modal parameter distribution of a whole turnout system.

Description

Technical Field

[0001] This invention relates to the field of modal testing technology for turnout areas, and more specifically, to an automatic identification method and system for modal parameters in turnout areas. Background Technology

[0002] As a crucial component of railway systems, turnouts are primarily used to change the direction of train travel, facilitating a smooth transition from one track to another. Their complex structure includes key components such as switch rails, stock rails, frogs, and connecting rods. During high-speed train operation, turnout systems are subjected to frequent vibrations and impacts, making the connection points of various components highly susceptible to wear and fatigue damage. Modal testing is an important method for studying structural dynamics. By obtaining the modal parameters of the turnout system, its response characteristics under various operating conditions can be revealed, providing a scientific basis for structural optimization design, dynamic performance control, vibration characteristic analysis, and health monitoring. However, existing modal testing methods are mostly applied to ordinary railway sections, and research on complex turnout systems remains lacking. Traditional laboratory testing methods (such as vibrators and axle drop tests) can effectively excite structural modes, but the equipment is large-scale and difficult to apply to actual field conditions. Conventional hammer tests, while convenient, have a limited excitation range and cannot fully cover the complex structural characteristics of turnouts, resulting in inaccurate distribution of obtained modal parameters. Furthermore, existing modal parameter identification methods rely on manual intervention, requiring engineers to subjectively judge the stability of modal parameters. This is not only time-consuming and labor-intensive, but also prone to inaccurate results due to subjective errors, making it difficult to meet the needs of efficient and accurate analysis of complex systems.

[0003] Based on the shortcomings of the existing technology, there is an urgent need for an automatic identification method and system for turnout area modal parameters. Summary of the Invention

[0004] The purpose of this invention is to provide a method and system for automatically identifying modal parameters in turnout areas, in order to improve the aforementioned problems. To achieve the above objective, the technical solution adopted by this invention is as follows:

[0005] Firstly, this application provides a method for automatically identifying modal parameters in a turnout area, including:

[0006] Obtain first information, which includes the dimensions of the target turnout area and the length of the connection line of the test equipment;

[0007] Based on the first information, the target turnout area is divided into regions. Based on the coherence principle of excitation-response data, each rail of the turnout is divided into multiple modal test areas to obtain the test area division results.

[0008] Based on the test area division results and the preset partitioned MRIT modal test scheme, each test area is subjected to excitation and response measurements in sequence, and common measurement points are retained between adjacent test areas to obtain excitation-response data;

[0009] The stimulus-response data is used to construct a matrix and calculate the modal parameters of each test region to obtain the modal parameter stability map of each test region.

[0010] Clustering is performed based on the modal parameter stability graph. By removing spurious modes and mode clusters of interference and filtering stable modal parameters, the clustering results are obtained.

[0011] Based on the clustering results, the mode shapes with frequencies within a preset range in adjacent regions are normalized using common measurement points as a reference, and the modal parameter identification results are spliced ​​together to obtain the modal parameter identification results. The modal parameter identification results represent the overall modal parameter distribution of the turnout system.

[0012] Secondly, this application also provides an automatic identification system for modal parameters in the turnout area, including:

[0013] The acquisition module is used to acquire first information, which includes the size information of the target turnout area and the length of the connection line of the test equipment.

[0014] The partitioning module is used to partition the target turnout area according to the first information, and divide each rail of the turnout into multiple modal test areas based on the coherence principle of the excitation-response data, so as to obtain the test area partitioning result.

[0015] The testing module is used to sequentially perform excitation and response measurements on each test area according to the test area division results and the preset partitioned MRIT modal testing scheme, and retain common measurement points between adjacent test areas to obtain excitation-response data;

[0016] The construction module is used to perform matrix construction processing based on the excitation-response data, calculate the modal parameters of each test region, and obtain the modal parameter stability map of each test region;

[0017] The clustering module is used to perform clustering processing based on the modal parameter stability graph. By removing spurious modes and mode cluster interference and filtering stable modal parameters, the clustering results are obtained.

[0018] The output module, based on the clustering results, normalizes the mode shapes with frequencies within a preset range in adjacent regions using common measurement points as a reference, and splices them together to obtain the modal parameter identification results. The modal parameter identification results represent the overall modal parameter distribution of the turnout system.

[0019] The beneficial effects of this invention are as follows:

[0020] Based on the complex structure of turnouts, this invention employs a region partitioning strategy, dividing each rail of the turnout into multiple modal testing areas. By retaining common measuring points between adjacent areas, modal parameters are normalized and spliced. Furthermore, combined with a density-based spatial clustering algorithm, this method can automatically remove spurious modes and interfering mode clusters, significantly improving the accuracy and efficiency of modal parameter analysis. Ultimately, it generates the overall modal parameter distribution of the turnout system, providing reliable data support for long-term health monitoring, fault diagnosis, and optimized design of the turnout. Attached Figure Description

[0021] To more clearly illustrate the technical solutions of the embodiments of the present invention, the accompanying drawings used in the embodiments will be briefly introduced below. It should be understood that the following drawings only show some embodiments of the present invention and should not be regarded as a limitation on the scope. For those skilled in the art, other related drawings can be obtained based on these drawings without creative effort.

[0022] Figure 1 This is a schematic diagram of the automatic identification method for turnout area modal parameters described in this embodiment of the invention;

[0023] Figure 2 This is a schematic diagram of the test areas for each mode in the turnout area;

[0024] Figure 3 This is a schematic diagram illustrating the principle of partitioned MRIT modal testing.

[0025] Figure 4 Here is a flowchart of the DBSCAN clustering algorithm;

[0026] Figure 5 This is a schematic diagram of the modal parameter identification results. Detailed Implementation

[0027] To make the objectives, technical solutions, and advantages of the embodiments of the present invention clearer, the technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some, not all, of the embodiments of the present invention. The components of the embodiments of the present invention described and shown in the accompanying drawings can generally be arranged and designed in various different configurations. Therefore, the following detailed description of the embodiments of the present invention provided in the accompanying drawings is not intended to limit the scope of the claimed invention, but merely to illustrate selected embodiments of the invention. All other embodiments obtained by those skilled in the art based on the embodiments of the present invention without inventive effort are within the scope of protection of the present invention.

[0028] It should be noted that similar reference numerals and letters in the following figures indicate similar items; therefore, once an item is defined in one figure, it does not need to be further defined and explained in subsequent figures. Furthermore, in the description of this invention, terms such as "first," "second," etc., are used only to distinguish descriptions and should not be construed as indicating or implying relative importance.

[0029] Example 1:

[0030] This embodiment provides a method for automatically identifying modal parameters in turnout areas.

[0031] See Figure 1 The figure shows that the method includes steps S100, S200, S300, S400, S500 and S600.

[0032] Step S100: Obtain first information, which includes the size information of the target turnout area and the length of the connection line of the test equipment;

[0033] Understandably, the dimensional information of the target turnout area includes the length, width, location of connection points, and related geometric characteristics of each rail in the turnout area. This information is a prerequisite for dividing the test area and determining the layout of the measuring points. The length of the connecting line of the test equipment is an important constraint affecting the test range, especially in the field test environment, where the length of the connecting line directly determines the reachability of the sensor due to the physical limitations of the equipment.

[0034] Step S200: Divide the target turnout area into regions based on the first information. Based on the coherence principle of the excitation-response data, divide each rail of the turnout into multiple modal test areas to obtain the test area division results.

[0035] Further, step S200 includes steps S210 to S230.

[0036] Step S210: Based on the size information of the target turnout area in the first information and the length of the connection line of the test equipment, perform preliminary division of the area. By limiting the distance between the connection point of the test equipment and the farthest excitation point in the target area to within the reach of the equipment, the turnout area is divided into several preliminary test intervals along the longitudinal direction of the rail using a geometric segmentation method to obtain the preliminary test range.

[0037] Step S220: Based on the preliminary test range, calculate the layout of excitation points and response points. By calculating the spatial distribution relationship between each excitation point and the corresponding response point, and screening out the area range where the coherence coefficient is greater than the preset threshold, the effective excitation point and response point distribution of each rail is obtained.

[0038] Step S230: Perform partitioning and numbering based on the distribution of effective excitation and response points. By assigning a unique identifier to each modal test area and establishing a mapping relationship between the area number and the excitation and response points, the test area division result is generated.

[0039] It should be noted that due to the limited length of the test equipment's connecting cable and the significant attenuation of rail vibration during longitudinal propagation, areas far from the excitation point cannot acquire effective response signals. Therefore, the effective test range in actual field measurements is usually limited, making it difficult to cover the entire turnout area. To address this issue, considering the length of the test equipment's connecting cable and the propagation characteristics of rail vibration, a region division method based on excitation-response data coherence is adopted. A coherence coefficient (e.g., ≥0.8) is used as a threshold for evaluating signal quality to determine the practically usable measurement range. Based on this, and according to the turnout's geometric characteristics, the rail is divided into multiple modal test areas, such as... Figure 2 As shown, the straight main rail, curved tip rail, curved guide rail, wing rail, long center rail, and short center rail are divided into different test areas, with different test ranges marked in different colors. Furthermore, for the straight main rail section, the area is divided at equal intervals according to the equipment coverage; for complex structures such as curved tip rails and curved guide rails, the area division is refined based on curvature variations to ensure that the test covers key parts. Common test points are set between adjacent areas for subsequent modal parameter normalization and splicing operations. Ultimately, this division method enables complete modal testing of complex turnout structures under equipment and signal constraints, providing high-quality test data support for subsequent modal parameter identification.

[0040] Step S300: Based on the test area division results and the preset partitioned MRIT modal test scheme, excitation and response measurements are performed sequentially on each test area, and common measurement points are retained between adjacent test areas to obtain excitation-response data;

[0041] Further, step S300 includes steps S310 to S330.

[0042] Step S310: Based on the test area division results and the preset partitioned MRIT modal test scheme, determine the layout of measurement points in each test area. By arranging multiple excitation points and response measurement points in each test area, and ensuring that adjacent areas share at least one common measurement point, a measurement point layout scheme is obtained.

[0043] Step S320: Based on the measurement point layout scheme, force hammer excitation is performed on each excitation point in each test area, and data is collected through sensors to obtain excitation point time series data and response point time series data;

[0044] Step S330: Perform data integration processing based on the time series data of excitation points and the time series data of response points. By recording the signal correspondence between all excitation points and response points as an excitation-response matrix, the excitation-response data of each test area is obtained.

[0045] It should be noted that the partitioned MMRIT modal testing scheme is an efficient modal testing method for complex structures such as turnout systems, aiming to address the shortcomings of traditional testing methods in terms of coverage, data integrity, and testing efficiency. This method is based on divided test areas, treating each area as an independent test unit. Multi-reference point impact testing (MRIT) technology is used to comprehensively excite and measure the modal characteristics within each test area. Specifically, multiple accelerometers are placed within each test area, with their positions avoiding modal nodes to ensure the capture of the complete vibration characteristics of the structure. During testing, a hammer sequentially excites all excitation points within the test area, and the sensor response signals are recorded simultaneously, forming the frequency response function matrix of that test area. Each element of the matrix represents the frequency response function between the excitation point and the response point, reflecting the dynamic characteristics of the area. To ensure the continuity and integrity of the test data, adjacent test areas share at least one common measurement point. Figure 3 As shown in the figure, [H(ω)] is the frequency response function matrix of the entire rail. 1, a, c, g, h, j, l, and m represent the region numbers. Each element in the matrix is ​​the frequency response function between each excitation point and the response measurement point, such as H. ij (ω) represents the frequency response function of excitation at point i and measurement at point j. The measurement regions in the figure represent the frequency response function matrices of each test region of the rail. For example, measurement region I in the figure indicates that accelerometers were placed at points 1, a, and c on the rail, and the hammer excites the rail sequentially from point 1 to point c. The frequency response function of each excitation data and the response data from the three sensors is shown. During testing, a common measurement point is retained between adjacent regions; for example, the last measurement point in measurement region I is point c, and the first measurement point in measurement region II is also point c. This design not only enables seamless stitching of modal data between regions but also provides a benchmark for subsequent normalization of modal parameters. The advantage of multi-reference point hammer impact testing is that it effectively avoids the situation where a single sensor misses a signal due to being at a modal node, and improves the coverage and reliability of modal testing. Furthermore, this method can establish the frequency response function matrix of each test region through multiple excitations and multi-point measurements, laying the foundation for modal parameter identification. Compared to traditional single-point excitation methods, the partitioned MMRIT scheme is more suitable for the modal testing needs of complex multi-component structures like turnouts. It can accurately capture the dynamic characteristics of the system and support subsequent modal parameter identification and splicing, providing high-quality basic data for the health monitoring and optimization design of turnout systems.

[0046] Step S400: Perform matrix construction processing based on the excitation-response data, and calculate the modal parameters of each test region to obtain the modal parameter stability map of each test region;

[0047] It should be noted that this step is based on excitation-response data, and performs matrix construction, modal parameter extraction and stability analysis in sequence. Through the modal parameter stability diagram, the modal characteristics of each test area are systematically displayed, providing accurate data support for modal identification and laying the foundation for the unified splicing of the overall modal parameters of the turnout system.

[0048] Further, step S400 includes steps S410 to S440.

[0049] Step S410: Perform frequency domain analysis processing based on the excitation-response data. By processing the signals between the excitation point and the response measurement point, the frequency response function matrix of each test area is constructed.

[0050] Step S420: Perform data transformation processing based on the frequency response function matrix. Use the inverse transform algorithm to transform the frequency domain data into time domain data and generate the corresponding impulse response function matrix.

[0051] Step S430: Based on the impulse response function matrix, construct the generalized Hankel matrix of the test area using the system identification algorithm, and extract the modal parameters of the generalized Hankel matrix through matrix decomposition to obtain preliminary modal parameter data;

[0052] Step S440: Perform stability analysis based on the preliminary modal parameter data. By calculating the stability of modal parameters at different orders of Hankel matrix, a modal parameter stability diagram is constructed.

[0053] Specifically, the excitation and response data of all measurement points are divided and organized according to the test area, and the frequency response function matrix [H(ω)] of each test area is obtained based on the frequency response function H1 estimation formula (1).

[0054]

[0055] In the formula, H1(ω) is the frequency response function of region 1; S fx (ω) represents the cross-power spectral density of the excitation and response data; S ff (ω) represents the self-power spectral density of the excitation data.

[0056] Perform an inverse Fourier transform on all frequency response functions, as shown in equation (2), to obtain the impulse response function matrix for each test region.

[0057]

[0058] In the formula, x[n] is the nth sample of the time domain signal; X[k] is the kth sample of the frequency domain signal; N is the number of points in the signal; and i is the imaginary unit.

[0059] Based on the impulse response function matrix of each test region, construct the generalized Hankel matrix of each test region, and perform singular value decomposition (3) on each generalized Hankel matrix.

[0060]

[0061] In the formula, [U] is the generalized Hankel matrix; [∑] is the left singular vector matrix; [∑] is the singular value matrix; [V] is the right singular vector matrix.

[0062] Based on the basic formula (4) of the ERA method, the minimum implementation of the turnout system in each test area is obtained [[A1], [B1], [G]].

[0063]

[0064] In the formula, [A1] is the system matrix; [B1] is the control matrix; [G] is the observation matrix; [E] is the observation matrix. L ]、[E N ] is the construction matrix, [E N ] t =[[I N ][0 N ]…[0 N [E] L ] T =[[I L ][0 L ]…[0 L ]).

[0065] The system matrix [A1] of each test region is decomposed using Eigenvalue decomposition equation (5) to obtain the eigenvalue matrix [Z] and eigenvector [Ψ] of each test region.

[0066] [A1][Ψ]=[Ψ][Z] (5)

[0067] Based on the relationship between the system matrix eigenvalues ​​and system eigenvalues ​​(6), the system matrix eigenvalues ​​[Λ] of each test region are obtained, and the modal frequencies ω of each test region are further obtained according to the relationship between the system eigenvalues ​​and modal parameters (7). i And modal damping ratio ζ i

[0068]

[0069]

[0070] In the formula, σ i The attenuation coefficient of the i-th mode; Let λ be the eigenvalue of the i-th order system. i The real part; Δt is the data sampling time interval; Let λ be the eigenvalue of the i-th order system. i The imaginary part of ω; di Let be the damped natural frequency of the i-th mode.

[0071] Based on the relationship between the modal shape matrix and the system eigenvector matrix (8), the modal shape matrix [Φ] of each test region is obtained.

[0072] [Φ]=[G][Ψ] (8)

[0073] Repeat the above calculation process to obtain the modal parameters of each test region under different Hankel matrix orders, and calculate the stability of each order of modal parameters under different Hankel matrix orders to obtain the modal parameter stability diagram of each test region. This diagram intuitively shows the distribution of parameters such as modal frequency and damping ratio under different orders.

[0074] Step S500: Perform clustering processing based on the modal parameter stability diagram. By removing spurious modes and mode cluster interference and filtering stable modal parameters, the clustering results are obtained.

[0075] It should be noted that traditional modal analysis involves manually determining the stability axis from the stability diagram to obtain the parameters of the most stable modes of each order, thus distinguishing between true and spurious modes. This method is inefficient and prone to introducing subjective errors. Furthermore, in the stability diagram, there may be mode clusters near the true modes that have different frequencies but essentially the same damping ratio and mode shape. This can interfere with the selection of stable modes and affect the accuracy of the modal parameter identification results.

[0076] Further, step S500 includes steps S510 to S540.

[0077] Step S510: Perform parameter initialization processing based on the modal parameter stability diagram. The initialization parameter configuration is obtained by analyzing the density of modal points in the diagram. The initialization parameter configuration includes the neighborhood radius parameter and the minimum number of points parameter.

[0078] Step S520: Based on the initialization parameter configuration, traverse all data points in the modal parameter stability graph, calculate the number of points in the neighborhood of each data point, and mark the points with a neighborhood number greater than or equal to the core point as core points, and mark the points with a neighborhood number less than the core point as boundary points, thus obtaining the core point set and the boundary point set.

[0079] Step S530: Perform clustering expansion processing based on the core point set and the boundary point set. By adding all directly density-reachable points in the core point and its neighborhood to the same cluster, and recursively processing other core points connected to the current core point density, a preliminary clustering result is obtained.

[0080] Step S540: Based on the preliminary clustering results, interference removal processing is performed. By analyzing the central characteristics and distribution density of each cluster, clusters that meet the consistency of modal frequency, damping ratio and mode shape are selected. At the same time, isolated false modal points and interference mode clusters are removed to obtain the final clustering results.

[0081] Specifically, this method determines the most stable mode based on the proportion of the total number of stable occurrences of each mode in the stability graph to the total number of calculations. Simultaneously, based on the characteristic of mode clusters appearing near the true mode, a clustering algorithm is used to automatically analyze and identify the stability graph data. There are many clustering algorithms; commonly used algorithms such as K-Means require a pre-defined number of clusters K, but in actual mode identification, prior information such as the order of structural modes is unknown, making them unsuitable. This method uses a density-based spatial clustering algorithm (DBSCAN), the algorithm flow of which is as follows: Figure 4 As shown in the figure, d(p,q) represents the Euclidean distance between data points p and q in the multidimensional space; i represents the dimension index of the data point; n represents the number of features of the data point; p i q represents the feature value of data point p in dimension i; i This represents the feature value of data point q in dimension i. The core idea of ​​this algorithm is to cluster data points based on the density around them, rather than assuming the number of clusters in advance as in traditional clustering algorithms. This makes it very suitable for analyzing stable graph data where clusters only appear near the true mode, but the number of clusters is unknown.

[0082] Step S600: Based on the clustering results, normalize the mode shapes with frequencies within a preset range in adjacent regions using common measurement points as a reference, and splice them together to obtain the modal parameter identification results. The modal parameter identification results represent the overall modal parameter distribution of the turnout system.

[0083] Further, step S600 includes steps S610 to S630.

[0084] Step S610: Based on the clustering results, normalization processing is performed. By normalizing the modal shape coefficients of the common reference measurement points to 1, normalized modal shape data is obtained.

[0085] Understandably, by normalizing the modal coefficients of common measurement points to 1, it ensures that the characteristics of modal shapes in each region remain consistent at the common measurement points during cross-regional stitching, eliminating the differences in modal shape scale caused by region division. The normalized modal shape data provides standardized input for subsequent modal matching and stitching.

[0086] Step S620: Perform matching processing based on the normalized modal data. By comparing and matching the modal data point by point, establish the cross-regional correspondence of the modal data, and mark each successfully matched modal pair to obtain the modal matching result.

[0087] Specifically, firstly, modal modes with frequencies within a preset range in each region are compared to select modes with similar frequencies. Then, based on these modal modes with similar frequencies, point-by-point comparisons are performed across all measurement points to establish cross-regional modal correspondences. During the matching process, it is necessary to ensure the continuity and consistency of modal modes at common measurement points, and mark successfully matched modal mode pairs to form modal mode matching results.

[0088] Step S630: Based on the modal shape matching results, perform splicing processing. By traversing the region by region, splice each pair of matched modal shapes according to the consistency of common measurement points to obtain the final modal parameter identification results.

[0089] It should be noted that during splicing, a common measuring point is used as a reference to unify and integrate the matched modal shapes, ensuring that the spliced ​​modal parameters have consistent frequency, mode shape, and damping characteristics in adjacent test areas. By progressively splicing the modal parameters of adjacent areas, a complete modal parameter distribution of the turnout system is generated, as shown below. Figure 5 The results of modal parameter identification for the turnout area are shown. This distribution includes the modal characteristics of all turnout components (such as straight main rails, curved tip rails, curved guide rails, etc.), providing complete data support for the overall dynamic analysis and health monitoring of the turnout system.

[0090] Example 2:

[0091] This embodiment provides an automatic modal parameter identification system for turnout areas, the system comprising:

[0092] The acquisition module is used to acquire first information, which includes the size information of the target turnout area and the length of the connection line of the test equipment.

[0093] The partitioning module is used to partition the target turnout area according to the first information. Based on the coherence principle of the excitation-response data, each rail of the turnout is divided into multiple modal test areas to obtain the test area partitioning results.

[0094] The testing module is used to sequentially perform excitation and response measurements on each test area according to the test area division results and the preset partitioned MRIT modal testing scheme, and retain common measurement points between adjacent test areas to obtain excitation-response data;

[0095] The module is used to perform matrix construction processing based on the stimulus-response data, calculate the modal parameters of each test region, and obtain the modal parameter stability map of each test region;

[0096] The clustering module is used to perform clustering based on the modal parameter stability graph. It obtains the clustering results by removing spurious modes and mode clusters of interference and filtering stable modal parameters.

[0097] The output module, based on the clustering results, normalizes the mode shapes with frequencies within a preset range in adjacent regions using common measurement points as a reference, and splices them together to obtain the modal parameter identification results. The modal parameter identification results represent the overall modal parameter distribution of the turnout system.

[0098] In one specific embodiment disclosed in this application, the partitioning module includes:

[0099] The first division unit is used to perform preliminary division of the area based on the size information of the target turnout area in the first information and the length of the connection line of the test equipment. By limiting the distance between the connection point of the test equipment and the farthest excitation point of the target area to within the reach of the equipment, the turnout area is divided into several preliminary test intervals along the longitudinal direction of the rail using a geometric segmentation method to obtain the preliminary test range.

[0100] The first calculation unit is used to calculate the arrangement of excitation points and response points based on the preliminary test range. By calculating the spatial distribution relationship between each excitation point and the corresponding response point, and screening out the area range where the coherence coefficient is greater than the preset threshold, the effective excitation point and response point distribution of each rail is obtained.

[0101] The first numbering unit is used to perform partitioning and numbering based on the distribution of effective excitation and response points. By assigning a unique identifier to each modal test area and establishing a mapping relationship between the area number and the excitation and response points, the test area division result is generated.

[0102] In one specific embodiment disclosed in this application, the test module includes:

[0103] The first processing unit is used to determine the layout of measurement points in each test area based on the test area division results and the preset partitioned MRIT modal test scheme. By arranging multiple excitation points and response measurement points in each test area, while ensuring that adjacent areas share at least one common measurement point, a measurement point layout scheme is obtained.

[0104] The second processing unit, based on the measurement point layout scheme, performs force hammer excitation processing on each excitation point in each test area, and collects data through sensors to obtain time series data of excitation points and time series data of response points;

[0105] The first integration unit is used to integrate and process data based on the time series data of the excitation points and the time series data of the response points. By recording the signal correspondence between all excitation points and response points as an excitation-response matrix, the excitation-response data of each test area is obtained.

[0106] In one specific embodiment disclosed in this application, the construction module includes:

[0107] The first analysis unit is used to perform frequency domain analysis processing based on the excitation-response data. By processing the signal between the excitation point and the response measurement point, the frequency response function matrix of each test area is constructed.

[0108] The first conversion unit is used to perform data conversion processing based on the frequency response function matrix. It converts the frequency domain data into time domain data by using an inverse transformation algorithm to generate the corresponding impulse response function matrix.

[0109] The first construction unit is used to construct the generalized Hankel matrix of the test area based on the impulse response function matrix using the system identification algorithm, and to extract the modal parameters of the generalized Hankel matrix through matrix decomposition to obtain preliminary modal parameter data.

[0110] The second calculation unit is used to perform stability analysis based on the preliminary modal parameter data. By calculating the stability of modal parameters at different orders of Hankel matrix, a modal parameter stability map is constructed.

[0111] In one specific embodiment disclosed in this application, the clustering module includes:

[0112] The third processing unit is used to perform parameter initialization processing based on the modal parameter stability diagram. It obtains the initialization parameter configuration by analyzing the density of modal points in the diagram. The initialization parameter configuration includes the neighborhood radius parameter and the minimum number of points parameter.

[0113] The third calculation unit traverses all data points in the modal parameter stability graph based on the initialization parameter configuration, calculates the number of points in the neighborhood of each data point, and marks points with a neighborhood number greater than or equal to the core point as core points, and points with a neighborhood number less than the core point but connected to the core point as boundary points, thus obtaining the core point set and the boundary point set.

[0114] The fourth processing unit is used to perform cluster expansion processing based on the core point set and the boundary point set. It adds all directly density-reachable points in the core point and its neighborhood to the same cluster and recursively processes other core points connected to the current core point density to obtain preliminary clustering results.

[0115] The fifth processing unit is used to remove interference based on the preliminary clustering results. By analyzing the central characteristics and distribution density of each cluster, it selects clusters that meet the consistency of modal frequency, damping ratio and mode shape, and removes isolated false modal points and interference mode clusters to obtain the final clustering results.

[0116] In one specific embodiment disclosed in this application, the output module includes:

[0117] The sixth processing unit performs normalization processing based on the clustering results. By normalizing the modal coefficients of the common reference measurement points to 1, normalized modal data is obtained.

[0118] The first matching unit is used to perform matching processing based on the normalized modal data. By comparing and matching the modal data point by point, the cross-regional correspondence of the modal data is established, and each successfully matched modal pair is marked to obtain the modal matching result.

[0119] The seventh processing unit is used to perform splicing processing based on the modal shape matching results. By traversing the region one by one, each pair of matched modal shapes is spliced ​​according to the consistency of common measurement points to obtain the final modal parameter identification results.

[0120] The above description is merely a specific embodiment of the present invention, but the scope of protection of the present invention is not limited thereto. Any variations or substitutions that can be easily conceived by those skilled in the art within the technical scope disclosed in the present invention should be included within the scope of protection of the present invention. Therefore, the scope of protection of the present invention should be determined by the scope of the claims.

Claims

1. A method for automatic identification of modal parameters in a turnout area, characterized in that, include: Obtain first information, which includes the dimensions of the target turnout area and the length of the connection line of the test equipment; Based on the first information, the target turnout area is divided into regions. Based on the coherence principle of excitation-response data, each rail of the turnout is divided into multiple modal test areas to obtain the test area division results. Based on the test area division results and the preset partitioned MRIT modal test scheme, each test area is subjected to excitation and response measurements in sequence, and common measurement points are retained between adjacent test areas to obtain excitation-response data; The stimulus-response data is used to construct a matrix and calculate the modal parameters of each test region to obtain the modal parameter stability map of each test region. Clustering is performed based on the modal parameter stability graph. By removing spurious modes and mode clusters of interference and filtering stable modal parameters, the clustering results are obtained. Based on the clustering results, the mode shapes with frequencies within a preset range in adjacent regions are normalized using common measurement points as a reference, and the modal parameter identification results are spliced ​​together to obtain the modal parameter identification results. The modal parameter identification results are the overall modal parameter distribution of the turnout system. Specifically, based on the test area division results and the preset partitioned MRIT modal testing scheme, each test area is sequentially subjected to excitation and response measurements, and common measurement points are retained between adjacent test areas to obtain excitation-response data, including: Based on the test area division results and the preset partitioned MRIT modal test scheme, the layout of measurement points in each test area is determined. By arranging multiple excitation points and response measurement points in each test area, while ensuring that adjacent areas share at least one common measurement point, a measurement point layout scheme is obtained. Based on the measurement point layout scheme, each excitation point in each test area is subjected to force hammer excitation processing, and data is collected through sensors to obtain time series data of excitation points and time series data of response points; Data integration processing is performed based on the time series data of the excitation points and the time series data of the response points. By recording the signal correspondence between all excitation points and response points as an excitation-response matrix, the excitation-response data of each test area is obtained. The clustering process, based on the modal parameter stability graph, involves removing spurious modes and mode cluster interference, and filtering stable modal parameters to obtain the clustering results, including: The parameter initialization process is performed based on the modal parameter stability graph. The initialization parameter configuration is obtained by analyzing the density of modal points in the graph. The initialization parameter configuration includes a neighborhood radius parameter and a minimum number of points parameter. Based on the initialization parameter configuration, traverse all data points in the modal parameter stability graph, calculate the number of points in the neighborhood of each data point, and mark the points with a neighborhood number greater than or equal to the minimum number of points parameter as core points, and mark the points with a neighborhood number less than the minimum number of points but connected to the core points as boundary points, thus obtaining the core point set and the boundary point set. Clustering expansion is performed based on the core point set and the boundary point set. This involves adding all directly density-reachable points within the core point and its neighborhood to the same cluster and recursively processing other core points that are density-connected to the current core point to obtain preliminary clustering results. Based on the preliminary clustering results, interference removal processing is performed. By analyzing the central characteristics and distribution density of each cluster, clusters that meet the consistency of modal frequency, damping ratio and mode shape are selected. At the same time, isolated false modal points and interference mode clusters are removed to obtain the final clustering results.

2. The automatic identification method for modal parameters in the turnout area according to claim 1, characterized in that, The stimulus-response data is used to construct a matrix, and the modal parameters of each test region are calculated to obtain the modal parameter stability map of each test region, including: Frequency domain analysis is performed on the excitation-response data. By processing the signals between the excitation point and the response measurement point, a frequency response function matrix for each test region is constructed. Data transformation processing is performed based on the frequency response function matrix. The frequency domain data is transformed into time domain data by using an inverse transform algorithm to generate the corresponding impulse response function matrix. Based on the impulse response function matrix, a generalized Hankel matrix of the test region is constructed using a system identification algorithm, and modal parameters of the generalized Hankel matrix are extracted through matrix decomposition to obtain preliminary modal parameter data; Stability analysis is performed based on the preliminary modal parameter data. By calculating the stability of modal parameters at different Hankel matrix orders, a modal parameter stability map is constructed.

3. The automatic identification method for modal parameters in the turnout area according to claim 1, characterized in that, Based on the clustering results, the modal modes with frequencies within a preset range in adjacent regions are normalized using common measurement points as a reference, and then spliced ​​together to obtain the modal parameter identification results. These modal parameter identification results represent the overall modal parameter distribution of the turnout system, including: Based on the clustering results, normalization processing is performed by normalizing the modal shape coefficients of the common reference measurement points to 1 to obtain the normalized modal shape data. The normalized modal data is used for matching. By comparing and matching the modal data point by point, a cross-regional correspondence between the modal data is established, and each successfully matched modal data pair is marked to obtain the modal matching result. The modal mode matching results are spliced ​​together by traversing each region one by one, according to the consistency of common measurement points, to obtain the final modal parameter identification results.

4. An automatic mode parameter identification system for turnout areas, characterized in that, include: The acquisition module is used to acquire first information, which includes the size information of the target turnout area and the length of the connection line of the test equipment. The partitioning module is used to partition the target turnout area according to the first information, and divide each rail of the turnout into multiple modal test areas based on the coherence principle of the excitation-response data, so as to obtain the test area partitioning result; The testing module is used to sequentially perform excitation and response measurements on each test area according to the test area division results and the preset partitioned MRIT modal testing scheme, and retain common measurement points between adjacent test areas to obtain excitation-response data; The construction module is used to perform matrix construction processing based on the excitation-response data, calculate the modal parameters of each test region, and obtain the modal parameter stability map of each test region; The clustering module is used to perform clustering processing based on the modal parameter stability graph. By removing spurious modes and mode cluster interference and filtering stable modal parameters, the clustering results are obtained. The output module, based on the clustering results, normalizes the mode shapes with frequencies within a preset range in adjacent regions using common measurement points as a reference, and splices them together to obtain the modal parameter identification results. The modal parameter identification results are the overall modal parameter distribution of the turnout system. The test module includes: The first processing unit is used to determine the layout of measurement points in each test area according to the test area division results and the preset partitioned MRIT modal test scheme. By arranging multiple excitation points and response measurement points in each test area, while ensuring that adjacent areas share at least one common measurement point, a measurement point layout scheme is obtained. The second processing unit, based on the measurement point layout scheme, performs force hammer excitation processing on each excitation point in each test area, and collects data through sensors to obtain time series data of excitation points and time series data of response points; The first integration unit is used to perform data integration processing based on the excitation point time series data and the response point time series data, and obtain the excitation-response data of each test area by recording the signal correspondence between all excitation points and response points as an excitation-response matrix. The clustering module includes: The third processing unit is used to perform parameter initialization processing based on the modal parameter stability graph. The initialization parameter configuration is obtained by analyzing the density of modal points in the graph. The initialization parameter configuration includes a neighborhood radius parameter and a minimum number of points parameter. The third calculation unit traverses all data points in the modal parameter stability graph based on the initialization parameter configuration, calculates the number of points in the neighborhood of each data point, and marks points with a neighborhood number greater than or equal to the minimum number of points parameter as core points, and marks points with insufficient neighborhood points but connected to core points as boundary points, thus obtaining a core point set and a boundary point set. The fourth processing unit is used to perform clustering expansion processing based on the core point set and the boundary point set. By adding all directly density-reachable points in the core point and its neighborhood to the same cluster, and recursively processing other core points connected to the current core point density, a preliminary clustering result is obtained. The fifth processing unit is used to perform interference removal processing based on the preliminary clustering results. By analyzing the central characteristics and distribution density of each cluster, it selects clusters that meet the consistency of modal frequency, damping ratio and mode shape, and removes isolated false modal points and interference mode clusters to obtain the final clustering results.

5. The automatic mode parameter identification system for turnout area according to claim 4, characterized in that, The building module includes: The first analysis unit is used to perform frequency domain analysis processing based on the excitation-response data. By processing the signal between the excitation point and the response measurement point, a frequency response function matrix for each test area is constructed. The first conversion unit is used to perform data conversion processing based on the frequency response function matrix, and to convert the frequency domain data into time domain data by using an inverse transformation algorithm to generate the corresponding impulse response function matrix. The first construction unit is used to construct a generalized Hankel matrix of the test region based on the impulse response function matrix using a system identification algorithm, and extract the modal parameters of the generalized Hankel matrix through matrix decomposition to obtain preliminary modal parameter data. The second calculation unit is used to perform stability analysis based on the preliminary modal parameter data. By calculating the stability of modal parameters at different orders of Hankel matrix, a modal parameter stability map is constructed.

6. The automatic mode parameter identification system for turnout area according to claim 4, characterized in that, The output module includes: The sixth processing unit performs normalization processing based on the clustering results, and obtains normalized modal shape data by normalizing the modal shape coefficients of the common reference measurement points to 1. The first matching unit is used to perform matching processing based on the normalized modal data. By comparing and matching the modal data point by point, a cross-regional correspondence relationship of the modal data is established, and each successfully matched modal data pair is marked to obtain the modal matching result. The seventh processing unit is used to perform splicing processing based on the modal shape matching results. By traversing the region one by one, each pair of matched modal shapes is spliced ​​according to the consistency of common measurement points to obtain the final modal parameter identification results.

Images