Method for acquiring sample data during analysis of dynamic reliability of ballastless track

Abstract

The invention discloses a method for acquiring sample data during ballastless track dynamic reliability analysis, which comprises the following steps: establishing a ballastless track structure fault tree model, and converting the ballastless track structure fault tree model into a ballastless track Bayesian network; a three-dimensional explicit dynamic numerical model of the ballastless track structure is established, and the failure time of each damaged part in the ballastless track Bayesian network model in different service periods is calculated; taking the compressive strength of the deteriorated concrete material at different service moments and the corresponding failure time as a sample database, and constructing a regression model of the compressive strength of the material and the corresponding failure time by adopting a machine learning method; according to the method, the compressive strength of a concrete material is randomly sampled by using a Monte Carlo method, the corresponding failure time of the compressive strength is solved through a regression model, and sufficient failure sample data of each component are obtained. And the analysis result of the dynamic reliability of the ballastless track structure is more accurate.

Description

Technical Field

[0001] The present invention relates to the technical field of high-speed railways, and in particular to a method for obtaining sample data when analyzing the dynamic reliability of a ballastless track. Background Art

[0002] Ballastless track has the characteristics of high smoothness, good uniformity of rigidity, long-lasting geometric shape and position, and less maintenance work. All railways in my country with speeds of 200km or more use various types of ballastless track. The track structure of railways under construction in plateaus, high ground stress and active fault zones is also planned to adopt ballastless track.

[0003] The crustal plate activities and geological disasters in the above-mentioned areas will cause large deformation of the sub-track foundation conditions such as uneven subgrade settlement, frost heave arching, active faults, etc. In addition, extreme environments such as freeze-thaw will also cause the degradation of track component material performance. Since large deformation of the sub-track foundation and component material performance degradation accompany the entire life cycle of ballastless track, the reliability of the ballastless track system in the above-mentioned areas during normal service is an issue that needs to be clarified urgently.

[0004] At present, the existing research on the reliability of ballastless track structures only focuses on track structures in non-plateau, non-high-ground stress and active fault zones. The track component damage sample data required for reliability analysis are mainly obtained through the subjective experience of industry experts. The failure mode of the reliability analysis method is simple (two-state system) and the failure state is simple. The reliability research results cannot reflect the dynamic changes of system reliability over time.

[0005] Dynamic Bayesian Network is an efficient tool for analyzing the dynamic reliability of a system. However, when using dynamic Bayesian Network to analyze the time-varying reliability of the ballastless track structure in the above-mentioned areas, a large amount of failure sample data of each fault root node is required. Since most of the railways in the above-mentioned areas are under construction, the statistical data of the failure sample data of each component of the ballastless track structure in the area are almost blank. Summary of the invention

[0006] In view of the above-mentioned deficiencies in the prior art, the present invention provides a method for obtaining sample data when analyzing the dynamic reliability of ballastless track. The method solves the problem of difficulty in obtaining fault data of root node events in the Bayesian network when analyzing the dynamic reliability of ballastless track in plateaus, high ground stress areas and active fault zones.

[0007] In order to achieve the above-mentioned invention object, the technical solution adopted by the present invention is: a method for obtaining sample data when analyzing the dynamic reliability of ballastless track, comprising:

[0008] S1. Establish a ballastless track structure fault tree model and transform it into a ballastless track Bayesian network;

[0009] S2. Establish a three-dimensional explicit dynamic numerical model of the ballastless track structure and calculate the failure time of each damaged component in the ballastless track Bayesian network model at different service periods;

[0010] S3. Taking the compressive strength of deteriorated concrete materials at different service times and their corresponding failure times as sample databases, a regression model of the compressive strength of materials and their corresponding failure times is constructed using machine learning methods;

[0011] S4. Use the Monte Carlo method to randomly sample the compressive strength of concrete materials, solve the regression model to obtain the corresponding failure time, and obtain sufficient failure sample data for each component.

[0012] Further: S1 includes:

[0013] S11. Determine the failure mode of the ballastless track structure based on the structural stress characteristics and damage form statistics of the ballastless track;

[0014] S12. Dividing the ballastless track structure system into different ballastless track structure subsystems according to the failure modes of the ballastless track components;

[0015] S13, analyzing the possibility of failure of the ballastless track structure subsystem caused by each failure mode, and determining the failure path of the ballastless track structure fault tree;

[0016] S14. Establish a fault tree model of the ballastless track structure according to the failure mode and failure path of the ballastless track structure;

[0017] S15. Convert the established ballastless track structure fault tree model into a ballastless track Bayesian network.

[0018] Further: In S11, the failure modes of the ballastless track structure include: cracking of the track slab in the middle of the span, cracking of the track slab at 1 / 4 of the distance from the end, falling of the track slab, crushing of the self-compacting concrete, cracking of the base slab in the middle of the span, cracking of the base slab at 1 / 4 of the distance from the end, large-scale falling of the base slab, cracking and falling of the limit boss, and crushing of the track slab limit groove.

[0019] Further: in S12, the subsystems of the ballastless track structure system include a track plate vertical support system, a base plate vertical support system, a track plate and base plate lateral restraint system, and a track plate longitudinal restraint system.

[0020] Further: In S13, the failure modes that lead to failure of the vertical support system of the track slab include: cracking of the track slab mid-span, cracking of the track slab 1 / 4 of the distance from the end, falling of the track slab, and crushing of the self-compacting concrete;

[0021] The failure modes that lead to failure of the vertical support system of the base plate include: cracking of the base plate in the middle of the span, cracking of the base plate at 1 / 4 of the end, and large-scale falling of the base plate;

[0022] The failure modes that lead to the failure of the lateral restraint system of the track plate and base plate include: cracking and falling of the limit boss, and crushing of the limit groove of the track plate;

[0023] The failure mode that causes the failure of the track plate longitudinal restraint system is: the track plate limit groove is crushed.

[0024] Further: S2 includes:

[0025] S21. Establish a three-dimensional explicit dynamic numerical model of ballastless track structure;

[0026] S22. Consider the effect of freeze-thaw environment on the deterioration of compressive strength of concrete materials, determine the deterioration rate of concrete materials, and calculate the compressive strength of deteriorated concrete materials at different service times;

[0027] S23, using the compressive strength of the concrete material after deterioration at different service times as a calculation parameter, inputting it into a three-dimensional explicit dynamic numerical model of the ballastless track structure, and simulating the deterioration state of the track structure at different service times;

[0028] S24, determining the value range of the applied load, designing corresponding calculation conditions according to the degradation state of the track structure at different service times through the failure mode of the root node event in the Bayesian network, and calculating the number of applied loads when the track structure components fail under the calculation conditions at different service times through the three-dimensional explicit dynamics calculation model of the ballastless track structure;

[0029] S25. Determine the mapping relationship between the number of applied loads and the actual time when the track structure components fail in the ballastless track structure numerical model, and obtain the actual failure time of the failure mode corresponding to the root node event in the Bayesian network model under each calculation condition.

[0030] Further: S22 includes:

[0031] S221. Considering the influence of freeze-thaw environment on the degradation of concrete materials, the relationship between the dynamic elastic modulus and Poisson's ratio of the material and the degradation of the number of freeze-thaw cycles is determined, and the expression is:

[0032]

[0033] E c =1.25E d -19

[0034]

[0035] Among them, E0 and ν are the initial elastic modulus and Poisson's ratio, E and ν are the elastic modulus and Poisson's ratio after degradation, E c is the static elastic modulus, E d is the dynamic elastic modulus, N is the number of freeze-thaw cycles experienced by the material;

[0036] S222. According to the relationship between the dynamic elastic modulus and Poisson's ratio of concrete materials and the degradation of the number of cycles, determine the compressive strength f of concrete materials that deteriorates with the number of cycles. cu , whose expression is:

[0037] f cu =(((E0(1-0.00007417N)+19) / 1.25-14) / 7.8) 3 ;

[0038] S223, based on the compressive strength f of the concrete material deteriorating with the number of cycles cu , calculate the compressive strength of deteriorated concrete materials at different service times.

[0039] Further: S24 includes:

[0040] S241, establishing a vehicle-track coupling explicit dynamics model, calculating and determining the amplitude range of the wheel-rail force;

[0041] S242, taking the maximum wheel-rail force as a reference and based on the acceleration experiment theory, determining the load value that causes the track structure in the three-dimensional explicit dynamic model of ballastless track to fail prematurely;

[0042] S243, designing corresponding calculation conditions through failure modes of track structure components in the Bayesian network;

[0043] S244. Taking the load value that causes premature failure of the track structure in the three-dimensional explicit dynamic model of the ballastless track as the loading condition, the three-dimensional explicit numerical model of the ballastless track structure is used to calculate the corresponding calculation conditions to obtain the number of load actions on the track structure components of different service periods when they fail.

[0044] Further: S25 includes:

[0045] S251, determine the time T1 of the first failure of the ballastless track structure after it is put into use under the condition of large foundation deformation, the number of loads applied when the numerical model fails when the parameters of the concrete compressive strength have not been degraded, and the number of loads applied when the three-dimensional explicit dynamic numerical model of the ballastless track structure fails at different service times, N1;

[0046] S252. Determine the mapping relationship between the number of loads applied when the track structure components fail in the three-dimensional numerical calculation model of the ballastless track structure and their failure time in reality based on T1, n1 and N1. The expression is:

[0047] t=N1×T1 / n1

[0048] Wherein, t is the component failure time calculated by the three-dimensional explicit dynamic numerical model of ballastless track structure;

[0049] S253. According to the mapping relationship between the number of applied loads when the track structure components fail and their actual failure times in the three-dimensional explicit dynamic numerical model of the ballastless track structure, the specific failure times of the track structure components in each calculation condition at different service times are calculated through the three-dimensional explicit dynamic numerical model of the ballastless track structure.

[0050] Further: S4 includes:

[0051] S41. The sampling interval is determined by taking the initial compressive strength of concrete material as the maximum value and the compressive strength after deterioration during 60 years of service as the minimum value;

[0052] S42. Using the Monte Carlo random sampling method, randomly sample the compressive strength of concrete materials in the determined sampling interval;

[0053] S43. Based on the extracted compressive strength data of concrete materials, the corresponding failure time is obtained by solving the regression model, and sufficient failure sample data of each component is obtained.

[0054] The beneficial effects of the present invention are:

[0055] 1. The present invention adopts a numerical simulation method to establish a three-dimensional explicit dynamic numerical model. Different three-dimensional explicit dynamic numerical models can be established according to different working conditions, covering most of the ballastless track structure disease forms. The obtained data is relatively comprehensive, with low calculation cost and high efficiency;

[0056] 2. The present invention uses machine learning methods to establish a regression model of compressive strength and failure time, and adopts the Monte Carlo method for sampling to obtain sufficient failure sample data. This method is simple to operate, has low calculation cost, controllable amount of failure sample data and high result accuracy;

[0057] 3. The method for obtaining sample data when analyzing the dynamic reliability of a ballastless track structure proposed in the present invention can take into account more forms of track structure defects, thereby establishing a sufficient number of representative failure sample libraries, making the dynamic reliability analysis of the ballastless track structure more accurate. BRIEF DESCRIPTION OF THE DRAWINGS

[0058] Figure 1The present invention is a flow chart of a method for obtaining sample data when analyzing the dynamic reliability of ballastless track.

[0059] Figure 2 Schematic diagram of the fault tree model of ballastless track structure.

[0060] Figure 3 This is the Bayesian network diagram of ballastless track structure.

[0061] Figure 4 Schematic diagram of the three-dimensional explicit dynamic numerical model of ballastless track structure.

[0062] Figure 5 Schematic diagram of the three-dimensional explicit dynamics numerical model of vehicle-track coupling. DETAILED DESCRIPTION

[0063] The specific implementation modes of the present invention are described below so that those skilled in the art can understand the present invention. However, it should be clear that the present invention is not limited to the scope of the specific implementation modes. For those of ordinary skill in the art, as long as various changes are within the spirit and scope of the present invention as defined and determined by the attached claims, these changes are obvious, and all inventions and creations utilizing the concept of the present invention are protected.

[0064] like Figure 1 As shown, in one embodiment of the present invention, a method for obtaining sample data when analyzing the dynamic reliability of a ballastless track is provided, comprising:

[0065] S1. Establish a ballastless track structure fault tree model and transform it into a ballastless track Bayesian network;

[0066] S2. Establish a three-dimensional explicit dynamic numerical model of the ballastless track structure and calculate the failure time of each damaged component in the ballastless track Bayesian network model at different service periods;

[0067] S3. Taking the compressive strength of deteriorated concrete materials at different service times and their corresponding failure times as sample databases, a regression model of the compressive strength of materials and their corresponding failure times is constructed using machine learning methods;

[0068] S4. Use the Monte Carlo method to randomly sample the compressive strength of concrete materials, solve the regression model to obtain the corresponding failure time, and obtain sufficient failure sample data for each component.

[0069] This application considers various forms of track structure damage, thereby establishing a sufficient and representative failure sample library, making the analysis results of the dynamic reliability of the ballastless track structure more accurate.

[0070] Specifically, Figure 2The figure shows a schematic diagram of a ballastless track structure fault tree model. Therefore, the method for establishing a ballastless track structure fault tree model in S1 includes:

[0071] S11. Determine the failure mode of the ballastless track structure based on the structural stress characteristics and damage form statistics of the ballastless track;

[0072] The failure modes of ballastless track structures include: cracking of the track slab in the middle of the span, cracking of the track slab at 1 / 4 of the end, falling of the track slab, crushing of self-compacting concrete, cracking of the base slab in the middle of the span, cracking of the base slab at 1 / 4 of the end, large-scale falling of the base slab, cracking and falling of the limit boss, and crushing of the track slab limit groove;

[0073] S12. Dividing the ballastless track structure system into different ballastless track structure subsystems according to the failure modes of the ballastless track components;

[0074] The subsystems of the ballastless track structure system include the track plate vertical support system, the base plate vertical support system, the track plate and base plate lateral restraint system, and the track plate longitudinal restraint system;

[0075] S13, analyzing the possibility of failure of the ballastless track structure subsystem caused by each failure mode, and determining the failure path of the ballastless track structure fault tree;

[0076] The failure modes that lead to failure of the vertical support system of the track slab include: cracking of the track slab mid-span, cracking of the track slab 1 / 4 of the distance from the end, falling of the track slab, and crushing of the self-compacting concrete;

[0077] The failure modes that lead to failure of the vertical support system of the base plate include: cracking of the base plate in the middle of the span, cracking of the base plate at 1 / 4 of the end, and large-scale falling of the base plate;

[0078] The failure modes that lead to the failure of the lateral restraint system of the track plate and base plate include: cracking and falling of the limit boss, and crushing of the limit groove of the track plate;

[0079] The failure mode that causes the longitudinal restraint system of the track plate to fail is: the track plate limit groove is crushed

[0080] S14. Establish a fault tree model of the ballastless track structure based on the failure mode and failure path of the ballastless track structure.

[0081] S15. The established ballastless track structure fault tree model is converted into a ballastless track Bayesian network. The ballastless track Bayesian network diagram is as follows: Figure 3 shown.

[0082] Specifically, S15 includes:

[0083] S151, mapping basic events in the fault tree to root nodes of the Bayesian network;

[0084] S152, mapping the intermediate events in the fault tree to the intermediate nodes of the Bayesian network;

[0085] S153, mapping the top event in the fault tree to a leaf node of the Bayesian network;

[0086] S154. Determine the directed edges and conditional probability table connecting the nodes in the Bayesian network based on the logical relationship between the events in the fault tree.

[0087] Specifically, S2 includes:

[0088] S21. Establish a three-dimensional explicit dynamic numerical model of ballastless track structure;

[0089] S22. Consider the effect of freeze-thaw environment on the deterioration of compressive strength of concrete materials, determine the deterioration rate of concrete materials, and calculate the compressive strength of deteriorated concrete materials at different service times;

[0090] S23, using the compressive strength of the concrete material after deterioration at different service times as a calculation parameter, inputting it into a three-dimensional explicit dynamic numerical model of the ballastless track structure, and simulating the deterioration state of the track structure at different service times;

[0091] S24, determining the value range of the applied load, designing corresponding calculation conditions according to the degradation state of the track structure at different service times through the failure mode of the root node event in the Bayesian network, and calculating the number of applied loads when the track structure components fail under the calculation conditions at different service times through the three-dimensional explicit dynamics calculation model of the ballastless track structure;

[0092] S25. Determine the mapping relationship between the number of applied loads and the actual time when the track structure components fail in the ballastless track structure numerical model, and obtain the actual failure time of the failure mode corresponding to the root node event in the Bayesian network model under each calculation condition.

[0093] like Figure 4 As shown, in particular, in this embodiment, ANSYS / LS-DYNA software is used to establish a three-dimensional explicit dynamic numerical model of the ballastless track structure, specifically including:

[0094] S211, using the APDL language of ANSYS software to write a command stream for establishing a 3D numerical model of the ballastless track, and saving the command stream as a K file that can be read by LS-DYNA software through ANSYS software;

[0095] S212. By reading the K file into LS-DYNA software, a three-dimensional numerical model of ballastless track based on LS-DYNA software for explicit dynamic analysis is established;

[0096] S213. In the LS-DYNA software, the material type and properties of each track component in the ballastless track numerical model established in step S212 are modified to the materials in the LS-DYNA software material library, wherein the beam unit simulating the rail is modified to an isotropic elastic material, the spring unit simulating the fastener stiffness is modified to a linear elastic discrete beam material, the spring unit simulating the CA mortar stiffness is modified to a linear elastic discrete beam material, the solid unit simulating the track slab is modified to a continuous cap beam concrete material, and the beam unit simulating the steel bars in the track slab adopts an elastic-plastic material, so that the constitutive relationship of the reinforced concrete material in the numerical model has the function of simulating mechanical behaviors such as material damage accumulation and structural cracking.

[0097] In particular, the large deformation condition of the foundation in the numerical model is implemented using simply supported constraints.

[0098] S22. The deterioration effect of concrete material strength over time under freeze-thaw environment is considered in the numerical model. The implementation method is to determine the deterioration rate of concrete material and calculate the compressive strength of concrete material after deterioration at different service times.

[0099] Specifically, S22 includes:

[0100] S221. Considering the influence of freeze-thaw environment on the degradation of concrete materials, the relationship between the dynamic elastic modulus and Poisson's ratio of the material and the degradation of the number of freeze-thaw cycles is determined, and the expression is:

[0101]

[0102] E c =1.25E d -19

[0103]

[0104] Among them, E0 and ν are the initial elastic modulus and Poisson's ratio, E and ν are the elastic modulus and Poisson's ratio after degradation, E c is the static elastic modulus, E d is the dynamic elastic modulus, N is the number of freeze-thaw cycles experienced by the material;

[0105] S222. According to the relationship between the dynamic elastic modulus and Poisson's ratio of concrete materials and the degradation with the number of cycles, determine the compressive strength f of concrete materials and the degradation with the number of cycles. cu , whose expression is:

[0106] f cu =(((E0(1-0.00007417N)+19) / 1.25-14) / 7.8) 3 ;

[0107] S223, based on the compressive strength of concrete material and the deterioration with the number of cycles cu , calculate the compressive strength of concrete materials after deterioration at different service times.

[0108] Optionally, the compressive strength of the concrete material that deteriorates with the number of cycles can be converted into the compressive strength of the concrete material that deteriorates with time f c ' u , thus determining the compressive strength of concrete materials after deterioration at different service times, the expression is:

[0109] f c ' u =(((E0(1-0.00007417T×N′)+19 / 1.25-14) / 7.8) 3

[0110] Among them, f c ' u is the compressive strength, T is the service time, and N′ is the number of freeze-thaw cycles in the area where the ballastless track structure is located in one year;

[0111] S23, using the compressive strength of deteriorated concrete materials at different service times as a calculation parameter of the three-dimensional numerical model of the ballastless track structure to simulate the deterioration state of the track structure at different service times;

[0112] S24, determining the value range of the applied load, designing corresponding calculation conditions according to the degradation state of the track structure at different service times through the failure mode of the root node event in the Bayesian network, and calculating the number of applied loads when the track structure components fail under the calculation conditions at different service times through the established three-dimensional explicit dynamics calculation model of the ballastless track structure;

[0113] Specifically, S24 includes:

[0114] S241, establishing a vehicle-track coupling explicit dynamics model, calculating and determining the amplitude range of the wheel-rail force;

[0115] The vehicle-track coupling explicit dynamics model constructed using LS-DYNA software is shown in the following figure: Figure 5 As shown;

[0116] S242, taking the maximum wheel-rail force as a reference and based on the acceleration experiment theory, determining a load value that causes premature failure of the track structure in the three-dimensional explicit dynamic model of ballastless track; the load value is greater than the reference wheel-rail force;

[0117] S243, designing corresponding calculation conditions through the failure mode corresponding to the root node event in the Bayesian network model;

[0118] S244, taking the load value that causes premature failure of the track structure in the three-dimensional explicit dynamic model of the ballastless track as a loading condition, using the three-dimensional explicit numerical model of the ballastless track structure to calculate the corresponding calculation conditions, and obtaining the number of load actions on the track structure components of different service periods when they fail;

[0119] S25, determining the mapping relationship between the number of applied loads and the actual time when the track structure components fail in the ballastless track structure numerical model, and obtaining the actual failure time of the failure mode corresponding to the root node event in the Bayesian network model under each calculation condition;

[0120] Specifically, S25 includes:

[0121] S251, determine the time T1 of the first failure of the ballastless track structure after it is put into use under the condition of large foundation deformation, the number of loads applied when the numerical model fails when the parameters of the concrete compressive strength have not been degraded, and the number of loads applied when the three-dimensional explicit dynamic numerical model of the ballastless track structure fails at different service times, N1;

[0122] S252. Determine the mapping relationship between the number of loads applied when the track structure components fail in the three-dimensional numerical calculation model of the ballastless track structure and their failure time in reality based on T1, n1 and N1. The expression is:

[0123] t=N1×T1 / n1

[0124] Wherein, t is the component failure time calculated by the three-dimensional explicit dynamic numerical model of ballastless track structure;

[0125] S253. According to the mapping relationship between the number of applied loads when the track structure components fail and their actual failure times in the three-dimensional explicit dynamic numerical model of the ballastless track structure, the specific failure times of the track structure components in each calculation condition at different service times are calculated through the three-dimensional explicit dynamic numerical model of the ballastless track structure.

[0126] In particular, in this embodiment, S3 includes:

[0127] S31, divide 4 / 5 of the samples in the sample database into a training set, and use the remaining 1 / 5 of the samples in the sample database as a validation set;

[0128] S32. Based on the data in the training set, a random forest algorithm in the machine learning method is used to establish a regression model with the compressive strength of the concrete material as the independent variable and the failure time of the component as the dependent variable;

[0129] S33. Based on the data in the validation set, the random forest algorithm in machine learning method is used to verify the accuracy of the regression model.

[0130] S34. The established regression model can be used to calculate the failure time of each damaged component of the ballastless track at different service periods.

[0131] Furthermore, S4 includes:

[0132] S41. The sampling interval is determined by taking the initial compressive strength of concrete material as the maximum value and the compressive strength after deterioration during 60 years of service as the minimum value;

[0133] S42. Using the Monte Carlo random sampling method, randomly sample the compressive strength of concrete materials in the determined sampling interval;

[0134] S43. Based on the extracted compressive strength data of concrete materials, the corresponding failure time is obtained by solving the regression model, and sufficient failure sample data of each component is obtained.

[0135] The above embodiments are only used to illustrate the technical solutions of the present application, rather than to limit them. Although the present application has been described in detail with reference to the aforementioned embodiments, those skilled in the art should understand that they can still modify the technical solutions described in the aforementioned embodiments, or make equivalent replacements for some of the technical features therein. However, these modifications or replacements do not deviate the essence of the corresponding technical solutions from the spirit and scope of the technical solutions of the embodiments of the present application.

Claims

1. A method for obtaining sample data when analyzing the dynamic reliability of ballastless track, characterized in that: include: S1. Establish a ballastless track structure fault tree model and transform it into a ballastless track Bayesian network; S2. Establish a three-dimensional explicit dynamic numerical model of the ballastless track structure and calculate the failure time of each damaged component in the ballastless track Bayesian network model at different service periods; S3. Taking the compressive strength of deteriorated concrete materials at different service times and their corresponding failure times as sample databases, a regression model of the compressive strength of materials and their corresponding failure times is constructed using machine learning methods; S4. Use the Monte Carlo method to randomly sample the compressive strength of concrete materials, solve the regression model to obtain the corresponding failure time, and obtain sufficient failure sample data for each component.

2. The method for obtaining sample data when analyzing the dynamic reliability of ballastless track according to claim 1, characterized in that: S1 includes: S11. Determine the failure mode of the ballastless track structure based on the structural stress characteristics and damage form statistics of the ballastless track; S12. Dividing the ballastless track structure system into different ballastless track structure subsystems according to the failure modes of the ballastless track components; S13, analyzing the possibility of failure of the ballastless track structure subsystem caused by each failure mode, and determining the failure path of the ballastless track structure fault tree; S14. Establish a fault tree model of the ballastless track structure according to the failure mode and failure path of the ballastless track structure; S15. Convert the established ballastless track structure fault tree model into a ballastless track Bayesian network.

3. The method for obtaining sample data when analyzing the dynamic reliability of ballastless track according to claim 2, characterized in that: In S11, the failure modes of the ballastless track structure include: cracking of the track slab in the middle of the span, cracking of the track slab at 1 / 4 of the distance from the end, falling of the track slab, crushing of the self-compacting concrete, cracking of the base slab in the middle of the span, cracking of the base slab at 1 / 4 of the distance from the end, large-scale falling of the base slab, cracking and falling of the limit boss, and crushing of the track slab limit groove.

4. The method for obtaining sample data when analyzing the dynamic reliability of ballastless track according to claim 3, characterized in that: In S12, the subsystems of the ballastless track structure system include a track plate vertical support system, a base plate vertical support system, a track plate and base plate lateral restraint system, and a track plate longitudinal restraint system.

5. The method for obtaining sample data when analyzing the dynamic reliability of ballastless track according to claim 4, characterized in that: In S13, the failure modes that lead to failure of the vertical support system of the track slab include: cracking of the track slab mid-span, cracking of the track slab 1 / 4 of the distance from the end, falling of the track slab, and crushing of the self-compacting concrete; The failure modes that lead to failure of the vertical support system of the base plate include: cracking of the base plate in the middle of the span, cracking of the base plate at 1 / 4 of the end, and large-scale falling of the base plate; The failure modes that lead to the failure of the lateral restraint system of the track plate and base plate include: cracking and falling of the limit boss, and crushing of the limit groove of the track plate; The failure mode that causes the failure of the track plate longitudinal restraint system is: the track plate limit groove is crushed.

6. The method for obtaining sample data when analyzing the dynamic reliability of ballastless track according to claim 1, characterized in that S2 include: S21. Establish a three-dimensional explicit dynamic numerical model of ballastless track structure; S22. Consider the effect of freeze-thaw environment on the deterioration of compressive strength of concrete materials, determine the deterioration rate of concrete materials, and calculate the compressive strength of deteriorated concrete materials at different service times; S23, using the compressive strength of the concrete material after deterioration at different service times as a calculation parameter, inputting it into a three-dimensional explicit dynamic numerical model of the ballastless track structure, and simulating the deterioration state of the track structure at different service times; S24, determining the value range of the applied load, designing corresponding calculation conditions according to the degradation state of the track structure at different service times through the failure mode of the root node event in the Bayesian network, and calculating the number of applied loads when the track structure components fail under the calculation conditions at different service times through the three-dimensional explicit dynamics calculation model of the ballastless track structure; S25. Determine the mapping relationship between the number of applied loads and the actual time when the track structure components fail in the ballastless track structure numerical model, and obtain the actual failure time of the failure mode corresponding to the root node event in the Bayesian network model under each calculation condition.

7. The method for obtaining sample data when analyzing the dynamic reliability of ballastless track according to claim 6, characterized in that: S22 includes: S221. Considering the influence of freeze-thaw environment on the degradation of concrete materials, the relationship between the dynamic elastic modulus and Poisson's ratio of the material and the degradation of the number of freeze-thaw cycles is determined, and the expression is: AND c =1.25E d -19 Among them, E0 and ν are the initial elastic modulus and Poisson’s ratio, respectively. and are the elastic modulus and Poisson’s ratio after degradation, E c is the static elastic modulus, E d is the dynamic elastic modulus, N is the number of freeze-thaw cycles experienced by the material; S222. According to the relationship between the dynamic elastic modulus and Poisson's ratio of concrete materials and the degradation of the number of cycles, determine the compressive strength f of concrete materials that deteriorates with the number of cycles. cu , whose expression is: in cu =(((E0(1-0.00007417N)+19) / 1.25-14) / 7.8) 3 ; S223, based on the compressive strength of concrete material and the deterioration with the number of cycles cu , calculate the compressive strength of concrete material after deterioration at different service times.

8. The method for obtaining sample data when analyzing the dynamic reliability of ballastless track according to claim 6, characterized in that: S24 includes: S241, establishing a vehicle-track coupling explicit dynamics model, calculating and determining the amplitude range of the wheel-rail force; S242, taking the maximum wheel-rail force as a reference and based on the acceleration experiment theory, determining the load value that causes the track structure in the three-dimensional explicit dynamic model of ballastless track to fail prematurely; S243, designing corresponding calculation conditions through failure modes of track structure components in the Bayesian network; S244. Taking the load value that causes premature failure of the track structure in the three-dimensional explicit dynamic model of the ballastless track as the loading condition, the three-dimensional explicit numerical model of the ballastless track structure is used to calculate the corresponding calculation conditions to obtain the number of load actions on the track structure components of different service periods when they fail.

9. The method for obtaining sample data when analyzing the dynamic reliability of ballastless track according to claim 6, characterized in that: S25 includes: S251, determine the time T1 of the first failure of the ballastless track structure after it is put into use under the condition of large foundation deformation, the number of loads applied when the numerical model fails when the parameters of the concrete compressive strength have not been degraded, and the number of loads applied when the three-dimensional explicit dynamic numerical model of the ballastless track structure fails at different service times, N1; S252. Determine the mapping relationship between the number of loads applied when the track structure components fail in the three-dimensional numerical calculation model of the ballastless track structure and their failure time in reality based on T1, n1 and N1. The expression is: t=N1×T1 / n1 Wherein, t is the component failure time calculated by the three-dimensional explicit dynamic numerical model of ballastless track structure; S253. According to the mapping relationship between the number of applied loads when the track structure components fail and their actual failure times in the three-dimensional explicit dynamic numerical model of the ballastless track structure, the specific failure times of the track structure components in each calculation condition at different service times are calculated through the three-dimensional explicit dynamic numerical model of the ballastless track structure.

10. The method for obtaining sample data when analyzing the dynamic reliability of ballastless track according to claim 1, characterized in that S4 include: S41. The sampling interval is determined by taking the initial compressive strength of concrete material as the maximum value and the compressive strength after deterioration during 60 years of service as the minimum value; S42. Using the Monte Carlo random sampling method, randomly sample the compressive strength of concrete materials in the determined sampling interval; S43. Based on the extracted compressive strength data of concrete materials, the corresponding failure time is obtained by solving the regression model, and sufficient failure sample data of each component is obtained.

Images