Method and system for inverting out-of-smoothness of subway train wheels through tunnel wall vibration source intensity data

Abstract

The invention provides a method and system for inverting metro train wheel out-of-smoothness through tunnel wall vibration source intensity data, and belongs to the technical field of accurate quantitative inversion of train wheel out-of-smoothness, and the method comprises the steps: obtaining actual measurement data of wheel out-of-smoothness, track irregularity and tunnel wall vibration source intensity; on the basis of the probability mapping relation between the out-of-roundness of the wheels and the source strength of the tunnel wall, actually measured source strength data of the tunnel wall are input into a model by utilizing a pre-trained probability machine learning model, so that probability statistics of the out-of-roundness of the train wheels is output, and accurate inversion of the out-of-roundness state of the wheels is realized. According to the method, random excitation of a wheel track system and a dynamic coupling mechanism of a vehicle-track-tunnel-stratum system are considered; meanwhile, wheel out-of-roundness probability statistics of a group of whole-line trains is quantitatively inverted by considering a probability mapping relation between wheel out-of-roundness excitation and tunnel wall source intensity response.

Description

Technical Field

[0001] The present invention relates to the technical field of accurate quantitative inversion of train wheel out-of-roundness, and particularly relates to a method and system for inverting the out-of-roundness of subway train wheels through tunnel wall vibration source strength data. Background Technique

[0002] With the rapid advancement of urbanization and the continuous increase in the density of rail transit networks, the environmental vibration problems caused by the operation of subway trains have become a key bottleneck restricting the construction of green and intelligent transportation. The vibration energy generated when the train passes can be transmitted to the surrounding strata through the tunnel structure, causing significant interference to adjacent buildings and residents' lives. To build an environmentally friendly rail transit system, it is necessary to establish a high-precision vibration prediction model during the line planning stage to evaluate environmental vibration and optimize the track vibration reduction design accordingly. The selection of wheel out-of-roundness parameters is the core element determining the reliability of the prediction model. During the operation and maintenance stage, it is necessary to grasp the key excitation parameters of the vehicle in real time, especially the dynamic changes of wheel out-of-roundness with significant random characteristics, to achieve vibration source tracing and precise control. Therefore, quantitatively obtaining the out-of-roundness of train wheels is the key prerequisite for improving the accuracy of environmental vibration prediction and formulating scientific vibration reduction designs, and it is also the scientific basis for vibration source tracing and achieving precise maintenance and repair.

[0003] Currently, there are mainly the following types of methods for detecting the out-of-roundness state of wheels:

[0004] (1) Contact displacement sensor detection: This method uses a mechanical contact sensor to directly measure the geometric parameters of the wheel tread, or obtains data by contacting the wheel surface through a roller mechanical device. The measurement result of this method is accurate and suitable for fine analysis of wheel out-of-roundness. However, since it is necessary to stop the vehicle at the depot for detection, it takes a long time and has a high cost, and it cannot be carried out for the wheels of all vehicles on the entire line. In addition, the sensor is prone to wear during long-term use, resulting in a decrease in accuracy.

[0005] (2) Non-contact laser detection: This method uses multiple laser displacement sensors (such as a three-point laser system) to non-contact collect the wheel tread data, combines the least squares method to fit the center of the circle and wavelet filtering for denoising, and restores the wheel contour. Its advantages are avoiding contact wear, low maintenance cost, and improving efficiency through multi-point synchronous measurement, which is suitable for complex on-site environments. However, this method has a high equipment cost and needs to process high-frequency noise interference, and the accuracy cannot reach the level of contact displacement sensor detection. At the same time, it also needs to stop the vehicle at the depot for detection and cannot be carried out for the wheels of all vehicles on the entire line.

[0006] (3) On-vehicle detection method: This method belongs to an indirect test method and inverses the out-of-roundness state of the wheel through the detection signals of vehicle components. Although it supports dynamic monitoring, it is necessary to install sensors on the entire train, with a high deployment cost and complex data fusion. For the quantitative detection of the out-of-roundness of the wheels of all trains on the entire line, the cost is too high.

[0007] (4)Trackside vibration monitoring method: This method belongs to the indirect testing method. By using sensors arranged beside the track to capture the characteristics of wheel-rail interaction, it has unique advantages of global coverage, economy and high efficiency. However, most of the existing technologies are limited to qualitative diagnosis and comparing the state advantages and disadvantages of different train wheels, and no analysis technology that can quantitatively obtain wheel out-of-roundness has been developed yet.

[0008] The current dynamic detection technology for wheel out-of-roundness mainly focuses on qualitative diagnosis, that is, determining whether there are disease types such as polygonal wear and flat spots on the wheels. Although such methods can achieve fault alarms, they cannot quantify the harmonic amplitudes of wheel out-of-round orders and are difficult to meet the more refined detection requirements. The existing trackside dynamic detection technology has significant limitations: on the one hand, this technology generally ignores the influence of the dynamic interaction between the vehicle and the track on vibration transmission, resulting in insufficient detection accuracy and being unable to meet the requirements of environmental vibration prediction, vibration source tracing, and vibration precise control; on the other hand, the existing technology only considers the one-to-one mapping relationship between wheel out-of-roundness and trackside signals, that is, using the deterministic inversion method. This method can neither quickly capture the group statistical laws of wheel out-of-roundness of trains along the whole line nor be compatible with the random factors of the dynamic change of the wheel-rail contact position in the system.

[0009] Li Minghang et al. proposed a comprehensive evaluation method for wheel out-of-roundness based on vibration source intensity monitoring. This method realizes the identification of passing train numbers and the identification of vehicles with wheel out-of-roundness exceeding the limit. This method obtains the normal distribution curve of the vibration source intensity test results through testing, and takes the samples higher than the distribution curve as the wheel over-limit samples for early warning identification. In essence, this method can only find the wheels with poor states and cannot quantitatively describe the overall state of wheel out-of-roundness of each vehicle. The specific technical disadvantages include: unable to quantify the harmonic amplitudes of wheel out-of-round orders; unable to consider the dynamic interaction between the vehicle and the track system, resulting in inaccurate prediction accuracy; unable to accurately capture the group quantitative statistical laws of wheel out-of-roundness of trains along the whole line; and difficult to be compatible with the random factors of the dynamic change of the wheel-rail contact position in the system.

[0010] In summary, the existing trackside dynamic detection technology for wheel out-of-roundness generally ignores the influence of the dynamic interaction between the vehicle and the track on vibration transmission. The lack of this physical coupling mechanism will inevitably lead to inaccurate prediction accuracy, and ultimately can only achieve the state warning of wheel out-of-roundness but cannot achieve quantitative measurement. At the same time, the existing technology only considers the one-to-one mapping relationship between wheel out-of-roundness and trackside signals, that is, using the deterministic inversion method. This method can neither quickly capture the group statistical laws of wheel out-of-roundness of trains along the whole line nor be compatible with the random factors of the dynamic change of the wheel-rail contact position in the system. Summary of the Invention

[0011] The object of the present invention is to provide a method and system for inverting the out-of-roundness of subway train wheels through the vibration source intensity data of the tunnel wall, so as to solve at least one of the technical problems existing in the above-mentioned background technology.

[0012] To achieve the above object, the present invention adopts the following technical solutions:

[0013] In a first aspect, the present invention provides a method for inverting the out-of-roundness of subway train wheels through the vibration source intensity data of the tunnel wall, including:

[0014] Obtaining the measured data of wheel out-of-roundness, track unevenness, and tunnel wall vibration source intensity;

[0015] Based on the probability mapping relationship between wheel out-of-roundness and tunnel wall source intensity, using a pre-trained probability machine learning model, inputting the measured tunnel wall source intensity data into the model, so as to output the probability statistics of train wheel out-of-roundness, and realizing the accurate inversion of the wheel out-of-roundness state; among them, using the measured data of wheel out-of-roundness, track unevenness, and tunnel wall vibration source intensity, and establishing a vehicle-track-tunnel-stratum coupling dynamics model considering wheel-rail random excitation, a training data set for establishing the mapping relationship between wheel out-of-roundness and tunnel wall source intensity is established.

[0016] As a further limitation of the first aspect of the present invention, the dynamics model of the vehicle-track-tunnel-stratum coupling system is disassembled into two parts: a vehicle-track coupling model and a tunnel-stratum dynamics model; among them, the establishment of the vehicle-track coupling model is divided into three parts: a vehicle model, a track model, and wheel-rail contact;

[0017] All components of the vehicle model are simplified into rigid bodies, the primary and secondary suspensions are simplified into spring-damping units, each carriage is simplified into a mass-spring-damping system with 10 degrees of freedom, and the frequency control equation of the train is expressed as:

[0018]

[0019] In the formula, ω is the circular frequency; M m , C m , K m are the mass, damping, and stiffness matrices of the m-th vehicle respectively; is the displacement of the m-th vehicle; is the dynamic excitation force vector acting on the m-th vehicle.

[0020] As a further limitation of the first aspect of the present invention, according to the "infinite-period" theory, the track structure is regarded as a discrete support infinite-period structure with a fastener spacing L as the period; under the action of a unit moving load with a frequency of ω l , the vibration equation of the rail beam within a characteristic period is:

[0021]

[0022] In the formula, is the vertical displacement response of the rail beam in the frequency domain, abbreviated as is the elastic modulus including the damping characteristics of the rail material, E r is the real elastic modulus of the rail, η r is the material loss factor; I r is the moment of inertia of the rail; m is the mass of the rail per unit length; is the position of the moving load at the initial moment; x n is the coordinate of the nth fastener support point; L is the fastener spacing.

[0023] As a further limitation of the first aspect of the present invention, the wheel-rail contact relationship is simulated by a linear spring, and it is assumed that the wheel and the rail always remain in contact. The wheel-rail coupling irregularity is based on the improved trigonometric series fitting method. The excitation frequency experienced by the kth axle corresponds to ω l (l = -N R ,..., -1, 1,..., N R ) The irregularity is expressed as:

[0024]

[0025] In the formula, is the excitation amplitude; θ lk is the combined phase difference considering the irregularities of different wavelengths and the irregularities experienced between different axles.

[0026] As a further limitation of the first aspect of the present invention, the mathematical mapping relationship between wheel roundness and tunnel wall source strength is expressed as:

[0027] x r = f(x w )

[0028] In the formula, represents the first p w order amplitudes of the wheel roundness of the whole vehicle; represents the tunnel wall source strength vibration response at the center frequencies of p r 1 / 3 octaves.

[0029] As a further limitation of the first aspect of the present invention, the training of the probability machine learning model includes the principal component analysis method, the kernel density estimation method, the diffusion mapping method, the solution of differential equations, and statistical constraints.

[0030] The dataset of the wheel roundness - tunnel wall source strength mapping relationship required for the training of the probability machine learning model is expressed as:

[0031] X train = (X w , Xr )

[0032] wherein (abbreviated as X train,p×q ) is a p×q dimensional matrix, where p = p w + p r , and q is the number of groups of samples; X w and X r are the sample matrices corresponding to the vectors x w and x r respectively.

[0033] In the principal component analysis method, the training data set is first standardized and normalized into the matrix [X] p×q , and then its covariance matrix C p×p is calculated, expressed as:

[0034]

[0035] Through eigenvalue decomposition, the covariance matrix can be expressed as:

[0036] C = VΛV T ;

[0037] wherein is the orthogonal basis matrix; Λ is the diagonal matrix of eigenvalues. Finally, the principal component analysis method projects the training data into the principal component space, expressed as:

[0038] H = XV;

[0039] wherein is the independent variable matrix after decoupling by the principal component analysis method.

[0040] The kernel density estimation method uses the Gaussian kernel function to perform non-parametric estimation on the joint probability density distribution ρ H (h) of the matrix H, expressed as:

[0041]

[0042] wherein represents the principal component features of a sample; σ is the kernel bandwidth parameter. The solution of the joint probability density distribution provides a probability basis for subsequent conditional sampling.

[0043] The diffusion mapping method converts the matrix H into a stochastic matrix through non-linear dimensionality reduction technology and its expression is:

[0044] Z = Hg(g T g) -1 ;

[0045] wherein is the diffusion mapping basis, and w represents the number of data clusters.

[0046] The solution of the Ito stochastic differential equation is to construct a reduced-order stochastic differential equation based on the Galerkin projection method, which is expressed as:

[0047] dZ(t) = v(t)dt;

[0048]

[0049] where v(t) is the rate of change of the system state with time t; the potential function U(Z) = -lnρ H (Z); W(t) is the Wiener process; f0 is the dissipation coefficient of the Wiener process. By solving the Ito stochastic differential equation, a new sample matrix Z new is generated and inversely mapped to the original coordinate system to obtain a new sample matrix

[0050] In the statistical constraint, the statistical samples of the measured tunnel wall source strength are embedded in the solution of the stochastic differential equation to generate samples of wheel out-of-roundness that satisfy the statistical characteristics of the finite-dimensional Euclidean space

[0051] In a second aspect, the present invention provides a system for inverting the out-of-roundness of subway train wheels through tunnel wall vibration source strength data, including:

[0052] An acquisition module for acquiring measured data of wheel out-of-roundness, track unevenness, and tunnel wall vibration source strength;

[0053] A processing module for, based on the probability mapping relationship between wheel out-of-roundness and tunnel wall source strength, inputting the measured tunnel wall source strength data into a pre-trained probabilistic machine learning model using the model to output the probability statistics of train wheel out-of-roundness, thereby achieving accurate inversion of the wheel out-of-roundness state; wherein, using the measured data of wheel out-of-roundness, track unevenness, and tunnel wall vibration source strength and the established vehicle-track-tunnel-ground coupling dynamics model considering wheel-rail random excitation, a training data set for establishing the mapping relationship between wheel out-of-roundness and tunnel wall source strength is established.

[0054] In a third aspect, the present invention provides a non-transitory computer-readable storage medium for storing computer instructions, which when executed by a processor, implement the method for inverting the out-of-roundness of subway train wheels through tunnel wall vibration source strength data as described in the first aspect.

[0055] Fourthly, the present invention provides a computer device, including a memory and a processor, the processor and the memory communicate with each other, the memory stores program instructions executable by the processor, and the processor calls the program instructions to execute the method for inverting the out-of-roundness of subway train wheels through the vibration source intensity data of the tunnel wall as described in the first aspect.

[0056] Fifthly, the present invention provides an electronic device, including: a processor, a memory and a computer program; wherein, the processor is connected to the memory, the computer program is stored in the memory, and when the electronic device runs, the processor executes the computer program stored in the memory so that the electronic device executes the instruction for implementing the method for inverting the out-of-roundness of subway train wheels through the vibration source intensity data of the tunnel wall as described in the first aspect.

[0057] Advantages of the present invention: The random coupling dynamics mechanism of the vehicle-track-tunnel-stratum system is organically combined with the probabilistic machine learning method, and the accurate quantitative inversion of the out-of-roundness of the wheels of in-service trains is realized through the measured statistical data of the source intensity of the tunnel wall. Description of the Drawings

[0058] In order to more clearly illustrate the technical solutions of the embodiments of the present invention, the following will briefly introduce the drawings required for the description of the embodiments. Obviously, the drawings in the following description are only some embodiments of the present invention. For those of ordinary skill in the art, other drawings can be obtained based on these drawings without creative efforts.

[0059] Figure 1 It is a flow chart of the full-system integrated tracking test method for the train vibration source intensity described in the embodiment of the present invention.

[0060] Figure 2 It is a schematic diagram of the vehicle-track analytical dynamics model described in the embodiment of the present invention.

[0061] Figure 3 It is a schematic diagram of the finite element model of the tunnel-soil layer system described in the embodiment of the present invention.

[0062] Figure 4 It is a flow chart of the training of the probabilistic machine learning model described in the embodiment of the present invention. Detailed Embodiments

[0063] The following details the embodiments of the present invention. The examples of the embodiments are shown in the drawings, where the same or similar reference numerals represent the same or similar elements or elements with the same or similar functions from beginning to end. The embodiments described below with reference to the drawings are exemplary and are only used to explain the present invention and should not be construed as a limitation of the present invention.

[0064] Those skilled in the art of the present technology can understand that, unless otherwise defined, all terms used herein (including technical terms and scientific terms) have the same meaning as the general understanding of those of ordinary skill in the art to which the present invention pertains.

[0065] It should also be understood that terms such as those defined in a general dictionary should be understood to have a meaning consistent with their meaning in the context of the prior art, and will not be interpreted with an idealized or overly formal meaning unless defined as herein.

[0066] Those skilled in the art of the present technology can understand that, unless specifically stated, the singular forms "a", "an", "the", and "said" used herein may also include the plural forms. It should be further understood that the term "comprising" used in the description of the present invention means the presence of the described features, integers, steps, operations, elements, and / or components, but does not exclude the presence or addition of one or more other features, integers, steps, operations, elements, and / or their groups.

[0067] In the description of this specification, the description with reference to terms such as "one embodiment", "some embodiments", "example", "specific example", or "some examples" etc. means that the specific features, structures, materials, or characteristics described in connection with the embodiment or example are included in at least one embodiment or example of the present invention. Moreover, the specific features, structures, materials, or characteristics described can be combined in a suitable manner in any one or more embodiments or examples. Without contradiction, those skilled in the art can combine and combine the different embodiments or examples described in this specification and the features of different embodiments or examples.

[0068] For the convenience of understanding the present invention, the following further explains the present invention with specific embodiments in conjunction with the accompanying drawings, and the specific embodiments do not constitute a limitation to the embodiments of the present invention.

[0069] Those skilled in the art should understand that the drawings are only schematic diagrams of the embodiments, and the components in the drawings are not necessarily essential for implementing the present invention.

[0070] The present invention proposes a method for quantitatively inversing the probability of wheel out-of-roundness based on tunnel wall vibration source intensity data. By using the wheel out-of-roundness - track irregularity - tunnel wall vibration source intensity tracking detection data and the vehicle - track - tunnel - soil layer system dynamics model, the dynamic correlation between the train wheel out-of-roundness and the tunnel wall vibration source intensity signal is established. By designing a popular probability machine learning model, the probability mapping relationship between the wheel out-of-roundness and the tunnel wall vibration source intensity signal is learned, so as to inversely calculate the wheel out-of-roundness of in-service trains according to the actually measured tunnel wall vibration source intensity data. The present invention provides an online wheel out-of-roundness quantitative detection solution that can be applied in engineering for the rail transit system through the collaborative architecture of dynamic mechanism and probability machine learning.

[0071] Example 1

[0072] In this Example 1, first, a system for inverting the out-of-roundness of subway train wheels through the vibration source intensity data of the tunnel wall is provided, including: an acquisition module for acquiring the measured data of wheel out-of-roundness, track unevenness, and tunnel wall vibration source intensity. A processing module for, based on the probability mapping relationship between wheel out-of-roundness and tunnel wall source intensity, using a pre-trained probability machine learning model, inputting the measured tunnel wall source intensity data into the model, and thus outputting the probability statistics of train wheel out-of-roundness to achieve accurate inversion of the wheel out-of-roundness state; wherein, using the measured data of wheel out-of-roundness, track unevenness, and tunnel wall vibration source intensity and the established vehicle-track-tunnel-stratum coupling dynamics model considering wheel-rail random excitation, a training data set for establishing the mapping relationship between wheel out-of-roundness and tunnel wall source intensity is established.

[0073] In this example, using the above system, a method for inverting the out-of-roundness of subway train wheels through the vibration source intensity data of the tunnel wall is realized.

[0074] In this method, first, the acquisition module is used to acquire the measured data of wheel out-of-roundness, track unevenness, and tunnel wall vibration source intensity. For example, the above data can be obtained through the full-system integrated tracking test of the train vibration source intensity and then manually input into the acquisition module. As Figure 1 shown, the full-system integrated tracking test of the train vibration source intensity includes the following process:

[0075] Using the measured tunnel wall source intensity statistical data and the trained probability machine learning model, quantitatively invert the out-of-roundness of in-service train wheels and complete the probability statistical inversion of in-service train wheel out-of-roundness.

[0076] Specifically, establish a vehicle-track-tunnel-stratum coupling dynamics model considering wheel-rail random excitation and verify the accuracy of the dynamics model based on the measured statistical data of wheel out-of-roundness, track unevenness, and tunnel wall vibration source intensity obtained through experiments.

[0077] As Figure 2 , Figure 3 shown, the dynamic modeling of the vehicle-track-tunnel-stratum coupling system can be disassembled into two parts: a vehicle-track coupling model and a tunnel-stratum dynamics model. Among them, the establishment of the vehicle-track coupling model is divided into three parts: a vehicle model, a track model, and wheel-rail contact.

[0078] All components of the vehicle model are simplified as rigid bodies, and the primary and secondary suspensions are simplified as spring-damper units. Each carriage is simplified as a mass-spring-damper system with 10 degrees of freedom. The frequency control equation of the train can be expressed as:

[0079]

[0080] where ω is the circular frequency; M m , C m , K m are the mass, damping, and stiffness matrices of the m-th vehicle section, respectively; is the displacement of the m-th vehicle section; is the dynamic excitation force vector acting on the m-th vehicle section.

[0081] According to the "infinite - periodic" theory, the track structure is regarded as a discrete - supported infinite - periodic structure with a period of the fastener spacing L. Under the action of a unit moving load with a frequency of ω l , the vibration equation of the rail - beam within one characteristic period is:

[0082]

[0083] where is the vertical displacement response of the rail - beam in the frequency domain, abbreviated as is the elastic modulus including the damping characteristics of the rail material, E r is the real elastic modulus of the rail, η r is the material loss factor; I r is the moment of inertia of the rail; m is the mass of the rail per unit length; is the position of the moving load at the initial moment; x n is the coordinate of the n - th fastener support point; L is the fastener spacing.

[0084] The wheel - rail contact relationship is simulated by a linear spring, and it is assumed that the wheel and the rail always remain in contact. The wheel - rail coupling irregularity is based on the improved trigonometric series fitting method. The excitation frequency experienced by the k - th axle corresponding to ω l ( l =-N R ,...,-1,1,...,N R ) of the irregularity can be expressed as:

[0085]

[0086] where is the excitation amplitude; θ lk is the combined phase difference considering irregularities of different wavelengths and irregularities experienced between different axles.

[0087] The finite - element method is used to establish the tunnel - soil dynamics model. Among them, the rail fastener reaction force in the vehicle - track system acts on the tunnel - formation system to form a vehicle - track - tunnel - formation coupling system.

[0088] The mathematical mapping relationship between wheel roundness irregularity and tunnel wall source strength is established, expressed as:

[0089] x r = f(x w )

[0090] wherein, represents the first p w order amplitude of the non-roundness of the vehicle wheels; represents the vibration response of the tunnel wall source strength at the center frequency of p r 1 / 3 octave bands.

[0091] Establish a training and validation mapping data set: Using the monitored statistical data obtained through experiments and the above-established dynamic coupling model, generate a training data set and a validation data set that satisfy the mathematical mapping relationship between wheel non-roundness and tunnel wall source strength, and train a probabilistic machine learning model. As shown in Figure 4 , the training method is divided into five small steps in total: principal component analysis method, kernel density estimation method, diffusion mapping method, solution of differential equations, and statistical constraints. The specific machine learning training process is as follows:

[0092] The data set of the wheel non-roundness - tunnel wall source strength mapping relationship required for the training of the probabilistic machine learning model is expressed as:

[0093] X train = (X w , X r );

[0094] wherein, (abbreviated as X train,p×q ) is a p×q-dimensional matrix, where p = p w + p r , q is the number of sample groups; X w and X r are the sample matrices corresponding to the vectors x w and x r , respectively.

[0095] In the principal component analysis method, the training data set is first standardized and normalized into the matrix [X] p×q , and then its covariance matrix C p×p is calculated, which is expressed as:

[0096]

[0097] Through eigenvalue decomposition, the covariance matrix can be expressed as:

[0098] C = VΛV T ;

[0099] wherein, is the orthogonal basis matrix; Λ is the diagonal matrix of eigenvalues. Finally, the principal component analysis method projects the training data into the principal component space, which is expressed as:

[0100] H = XV;

[0101] Wherein, is the independent variable matrix decoupled by the principal component analysis method.

[0102] The kernel density estimation method uses a Gaussian kernel function to perform non-parametric estimation on the joint probability density distribution ρ H (h) of the matrix H, expressed as:

[0103]

[0104] Wherein, represents the principal component features of a sample; σ is the kernel bandwidth parameter. The solution of the joint probability density distribution provides a probability basis for subsequent conditional sampling.

[0105] The diffusion mapping method converts the matrix H into a stochastic matrix through a non-linear dimensionality reduction technique The expression thereof is:

[0106] Z = Hg(g T g) -1 ;

[0107] Wherein, is the diffusion mapping basis, and w represents the number of data clusters.

[0108] The solution of the Ito stochastic differential equation is to construct a reduced-order stochastic differential equation based on the Galerkin projection method, expressed as:

[0109] dZ(t) = v(t)dt;

[0110]

[0111] Wherein, v(t) is the change rate of the system state with time t; the potential function U(Z) = -lnρ H (Z); W(t) is the Wiener process; f0 is the dissipation coefficient of the Wiener process. By solving the Ito stochastic differential equation, a new sample matrix Z new is generated, and inverse mapping is performed to the original coordinate system to obtain a new sample matrix

[0112] In the statistical constraint, the statistical sample of the measured tunnel wall source strength is embedded in the solution of the stochastic differential equation to generate a sample of wheel out-of-roundness that satisfies the statistical characteristics of the finite-dimensional Euclidean space

[0113] Example 2

[0114] Embodiment 2 provides a non-transitory computer-readable storage medium for storing computer instructions. When the computer instructions are executed by a processor, the method for inverting the out-of-roundness of subway train wheels through the tunnel wall vibration source intensity data as described above is implemented. The method includes:

[0115] Obtain the measured data of wheel out-of-roundness, track unevenness, and tunnel wall vibration source intensity. Based on the probability mapping relationship between wheel out-of-roundness and tunnel wall source intensity, use a pre-trained probability machine learning model to input the measured tunnel wall source intensity data into the model, and then output the probability statistics of train wheel out-of-roundness, so as to achieve accurate inversion of the wheel out-of-roundness state; among them, use the measured data of wheel out-of-roundness, track unevenness, and tunnel wall vibration source intensity, and the established vehicle-track-tunnel-stratum coupling dynamics model considering wheel-rail random excitation to establish a training data set for the mapping relationship between wheel out-of-roundness and tunnel wall source intensity.

[0116] Embodiment 3

[0117] Embodiment 3 provides a computer device, including a memory and a processor. The processor and the memory communicate with each other. The memory stores program instructions executable by the processor. The processor calls the program instructions to execute the method for inverting the out-of-roundness of subway train wheels through the tunnel wall vibration source intensity data as described above. The method includes:

[0118] Obtain the measured data of wheel out-of-roundness, track unevenness, and tunnel wall vibration source intensity. Based on the probability mapping relationship between wheel out-of-roundness and tunnel wall source intensity, use a pre-trained probability machine learning model to input the measured tunnel wall source intensity data into the model, and then output the probability statistics of train wheel out-of-roundness, so as to achieve accurate inversion of the wheel out-of-roundness state; among them, use the measured data of wheel out-of-roundness, track unevenness, and tunnel wall vibration source intensity, and the established vehicle-track-tunnel-stratum coupling dynamics model considering wheel-rail random excitation to establish a training data set for the mapping relationship between wheel out-of-roundness and tunnel wall source intensity.

[0119] Embodiment 4

[0120] Embodiment 4 of the present invention provides an electronic device, including: a processor, a memory, and a computer program; wherein, the processor is connected to the memory, the computer program is stored in the memory, and when the electronic device runs, the processor executes the computer program stored in the memory, so that the electronic device executes an instruction for implementing the method for inverting the out-of-roundness of a subway train wheel through the vibration source intensity data of the tunnel wall, and the method includes: obtaining the measured data of the wheel out-of-roundness, track irregularity, and tunnel wall vibration source intensity. Based on the probability mapping relationship between the wheel out-of-roundness and the tunnel wall source intensity, using a pre-trained probability machine learning model, inputting the measured tunnel wall source intensity data into the model, and thus outputting the probability statistics of the train wheel out-of-roundness, so as to achieve accurate inversion of the wheel out-of-roundness state; wherein, using the measured data of the wheel out-of-roundness, track irregularity, and tunnel wall vibration source intensity, and the established vehicle-track-tunnel-stratum coupling dynamics model considering wheel-rail random excitation, a training data set for establishing the mapping relationship between the wheel out-of-roundness and the tunnel wall source intensity is established.

[0121] Those skilled in the art should understand that the embodiments of the present invention can be provided as a method, a system, or a computer program product. Therefore, the present invention can take the form of a completely hardware embodiment, a completely software embodiment, or an embodiment combining software and hardware aspects. Moreover, the present invention can take the form of a computer program product implemented on one or more computer-usable storage media (including but not limited to disk storage, CD-ROM, optical storage, etc.) containing computer-usable program code.

[0122] The present invention is described with reference to the flowcharts and / or block diagrams of methods, devices (systems), and computer program products according to the embodiments of the present invention. It should be understood that each process and / or block in the flowcharts and / or block diagrams, and the combination of processes and / or blocks in the flowcharts and / or block diagrams, can be implemented by computer program instructions. These computer program instructions can be provided to the processor of a general-purpose computer, a special-purpose computer, an embedded processor, or other programmable data processing devices to generate a machine, so that the instructions executed by the processor of the computer or other programmable data processing devices generate a device for implementing the functions specified in Figure 1 one process or multiple processes and / or blocks Figure 1 a block or multiple blocks.

[0123] These computer program instructions can also be stored in a computer-readable memory that can direct a computer or other programmable data processing device to work in a specific manner, so that the instructions stored in the computer-readable memory generate a manufactured article including an instruction device, and the instruction device implements the functions specified in Figure 1 one process or multiple processes and / or blocks Figure 1 a block or multiple blocks.

[0124] These computer program instructions may also be loaded onto a computer or other programmable data processing apparatus, where they execute a series of operational steps to produce a computer-implemented process, so that the instructions executed on the computer or other programmable apparatus provide steps for implementing the functions specified in one process or a plurality of processes and / or blocks Figure 1 one process or a plurality of processes and / or blocks Figure 1 or steps for implementing the functions specified in one block or a plurality of blocks.

[0125] Although the specific embodiments of the present invention have been described above in conjunction with the accompanying drawings, they are not intended to limit the scope of protection of the present invention. Those skilled in the art should understand that, based on the technical solutions disclosed in the present invention, various modifications or variations that can be made by those skilled in the art without creative efforts should be covered within the scope of protection of the present invention.

Claims

1. A method for inverting the out-of-roundness of subway train wheels from the vibration source intensity data of the tunnel wall, characterized in that, Including: Obtaining the measured data of wheel out-of-roundness, track irregularity, and tunnel wall vibration source intensity; Based on the probability mapping relationship between wheel out-of-roundness and tunnel wall source intensity, using a pre-trained probabilistic machine learning model, inputting the measured tunnel wall source intensity data into the model, and thus outputting the probability statistics of train wheel out-of-roundness to achieve accurate inversion of the wheel out-of-roundness state; wherein, using the measured data of wheel out-of-roundness, track irregularity, and tunnel wall vibration source intensity and the established vehicle-track-tunnel-ground coupling dynamics model considering wheel-rail random excitation, a training data set for the mapping relationship between wheel out-of-roundness and tunnel wall source intensity is established.

2. The method for inverting the out-of-roundness of subway train wheels based on the vibration source intensity data of the tunnel wall according to claim 1, wherein The dynamic model of the vehicle-track-tunnel-ground coupling system is disassembled into two parts: a vehicle-track coupling model and a tunnel-ground dynamics model; among them, the establishment of the vehicle-track coupling model is divided into three parts: a vehicle model, a track model, and wheel-rail contact. All components of the vehicle model are simplified to rigid bodies, the primary and secondary suspensions are simplified to spring-damper units, and each carriage is simplified to a mass-spring-damper system with 10 degrees of freedom. The frequency control equation of the train is expressed as: where ω is the circular frequency; M m , C m , K m are respectively the mass, damping, and stiffness matrices of the m-th vehicle section; is the displacement of the m-th vehicle section; is the dynamic excitation force vector acting on the m-th vehicle section.

3. The method for inverting the out-of-roundness of subway train wheels based on the vibration source intensity data of the tunnel wall according to claim 2, wherein, According to the "infinite - periodic" theory, the track structure is regarded as a discrete - supported infinite - periodic structure with a period of the fastener spacing L; under the action of a unit moving load with a frequency of ω l , the vibration equation of the rail beam within a characteristic period is as follows: In the formula, is the vertical displacement response of the rail beam in the frequency domain, abbreviated as is the elastic modulus including the damping characteristics of the rail material, E r is the real elastic modulus of the rail, η r is the material loss factor; I r is the moment of inertia of the rail; m is the mass of the rail per unit length; is the position of the moving load at the initial moment; x n is the coordinate of the nth fastener support point; L is the fastener spacing.

4. The method for inverting the out-of-roundness of subway train wheels based on the vibration source intensity data of the tunnel wall according to claim 3, characterized in that, The wheel-rail contact relationship is simulated using a linear spring, and it is assumed that the wheel and the rail are always in contact. The wheel-rail coupling irregularity is based on the improved trigonometric series fitting method, and the excitation frequency experienced by the k-th axle corresponds to ω l (l = -N R ,..., -1, 1,..., N R ) The irregularity expression is: wherein, is the excitation amplitude; θ lk is the combined phase difference considering the wavelength irregularities of different wavelengths and the irregularities experienced between different axles.

5. The method for inverting the out-of-roundness of subway train wheels from the vibration source intensity data of the tunnel wall according to claim 4, characterized in that The mathematical mapping relationship between wheel out-of-roundness and tunnel wall source intensity is expressed as: x r = f(x w ) In the formula, represents the first p w order amplitude of the out-of-roundness of the vehicle's wheels; represents the vibration response of the tunnel wall source strength at the center frequencies of p r 1 / 3 octave bands.

6. The method for inverting the out-of-roundness of subway train wheels based on the vibration source intensity data of the tunnel wall according to claim 1, wherein The training of the probabilistic machine learning model includes principal component analysis method, kernel density estimation method, diffusion mapping method, solution of differential equations, and statistical constraints.

7. A system for inverting the out-of-roundness of subway train wheels through the vibration source intensity data of the tunnel wall, characterized in that, Including: An acquisition module for obtaining the measured data of wheel out-of-roundness, track irregularity, and tunnel wall vibration source intensity; A processing module for, based on the probability mapping relationship between wheel out-of-roundness and tunnel wall source intensity, using a pre-trained probabilistic machine learning model, inputting the measured tunnel wall source intensity data into the model, and thus outputting the probability statistics of train wheel out-of-roundness to achieve accurate inversion of the wheel out-of-roundness state; wherein, using the measured data of wheel out-of-roundness, track irregularity, and tunnel wall vibration source intensity and the established vehicle-track-tunnel-ground coupling dynamics model considering wheel-rail random excitation, a training data set for the mapping relationship between wheel out-of-roundness and tunnel wall source intensity is established.

8. A non-transitory computer-readable storage medium, characterized in that, The non-transitory computer-readable storage medium is used to store computer instructions, and when the computer instructions are executed by a processor, the method for inverting the out-of-roundness of subway train wheels through tunnel wall vibration source intensity data as described in any one of claims 1-6 is realized.

9. A computer device, characterized in that, Including a memory and a processor, the processor and the memory communicate with each other, the memory stores program instructions executable by the processor, and the processor calls the program instructions to execute the method for inverting the out-of-roundness of subway train wheels through tunnel wall vibration source intensity data as described in any one of claims 1-6.

10. An electronic device, characterized in that, Including: A processor, a memory, and a computer program; wherein, the processor is connected to the memory, the computer program is stored in the memory, and when the electronic device runs, the processor executes the computer program stored in the memory so that the electronic device executes the instructions for realizing the method for inverting the out-of-roundness of subway train wheels through tunnel wall vibration source intensity data as described in any one of claims 1-6.

Images