Railway bridge damage and driving performance mapping method and device and storage medium

Abstract

The invention relates to a railway bridge damage and driving performance mapping method and device and a storage medium, and is applied to the technical field of traffic safety, and the method comprises the steps: building a mapping relation between rail-bridge key component earthquake-induced damage and rail irregularity, and comprehensively considering the influence of rail-bridge system earthquake-induced damage on driving safety performance; a deep learning method is adopted, a deep learning model is trained through a component earthquake-induced damage numerical value of the rail-bridge system and a train-rail-bridge coupling dynamic response result, and then a quantitative mapping relation between structural earthquake-induced damage of the rail-bridge system and driving performance is established. Therefore, safety and stability indexes of driving on the railway bridge track after key components or parts of the railway bridge track are damaged due to earthquakes can be quickly and accurately provided.

Description

Technical Field

[0001] The present invention relates to the field of traffic safety technology, and in particular to a method, device and storage medium for mapping railway bridge damage and vehicle performance. Background Art

[0002] As a major national strategic and pioneering infrastructure, railways have an irreplaceable position in my country's economic and social development. According to statistics, as of the end of 2021, the national high-speed railway operating mileage has reached 40,000 kilometers, and bridge structures account for more than 50% of most high-speed rail lines. With the continuous and rapid development of railways, it is inevitable to pass through special areas such as earthquake zones and poor geological conditions. Some railways even cross multiple Holocene active fault zones such as Longmen Mountain and Yala River. This makes it impossible for high-density trains to avoid driving on bridges during and after earthquakes in time and space, which ultimately leads to serious earthquake threats to bridges and their driving performance in high-intensity areas.

[0003] At present, the research on the driving performance on bridges mainly considers the influence of conventional factors such as bridge pier settlement and temperature load, while the research on earthquake effects is relatively scarce. However, earthquake effects are extremely random and time-varying, and their impact mechanism and action law on the driving performance on bridges are significantly different from those of the above factors. The post-seismic damage caused by the track-bridge system and components will be mapped to the track surface through the interlayer interaction between structural components, resulting in the degradation of the track smoothness, which in turn threatens driving safety and may even cause serious consequences such as train derailment. Summary of the invention

[0004] In view of this, the purpose of the present invention is to provide a method, device and storage medium for mapping railway bridge damage and driving performance, so as to solve the problem in the prior art that there is a lack of research on the impact of earthquake-induced component damage in the track-bridge system on driving safety performance, and the mapping relationship between the two has not yet been established.

[0005] According to a first aspect of an embodiment of the present invention, a method for mapping railway bridge damage and driving performance is provided, the method comprising:

[0006] The train models of different train models are established by using multi-rigid body dynamics theory, and the track model and bridge model are established by using finite element theory; the train-track-bridge coupling system calculation model is established by using the train model, track model and bridge model;

[0007] Obtain the earthquake-induced damage values ​​of components in the track-bridge system;

[0008] The basic assumptions of the theoretical model of the track-bridge structure are set, and the equilibrium differential equations of the mechanical model of the bridge and track structure are established according to the mechanical characteristics of the track-bridge structure. Based on the equilibrium differential equations, the mapping relationship between the seismic damage of the key components of the track-bridge and the track irregularity is obtained;

[0009] Acquire initial track irregularity samples and additional track irregularity samples respectively, and superimpose the initial track irregularity samples and the additional track irregularity samples to obtain vibration-induced random track irregularity samples;

[0010] The seismic-induced random track irregularity samples and the initial track irregularity samples are respectively input into the train-track-bridge coupling system calculation model to obtain two sets of train-track-bridge coupling dynamic response results;

[0011] The earthquake-induced damage value of the components of the track-bridge system is used as the first parameter; the train-track-bridge coupled dynamic response result is obtained by subtracting the two sets of train-track-bridge coupled dynamic response results, and the train-track-bridge coupled dynamic response result is used as the second parameter;

[0012] The pre-built model framework is trained by using the first parameter and the second parameter, thereby establishing a quantitative mapping relationship between the structural seismic damage and the driving performance of the track-bridge system.

[0013] Preferably,

[0014] The method of using multi-rigid body dynamics theory to establish train models of different train models and using finite element theory to respectively establish track models and bridge models includes:

[0015] The train model is obtained by using a rigid body to simulate the train body, front and rear bogies and four wheel sets, and using a spring-damping system to simulate the primary and secondary suspension systems of the train;

[0016] The track model is obtained by using a spatial beam unit or a plate unit to simulate the base plate and the track plate of the track, using a spatial beam unit to simulate the rail of the track, and using a spring-damping unit to simulate the interlayer connection components of the track;

[0017] The spatial beam unit is used to simulate the foundation, abutment, piers and main beam of the bridge, and the spring-damping is used to simulate the support of the bridge to obtain the bridge model.

[0018] Preferably,

[0019] The establishing of the train-track-bridge coupling system calculation model by using the train model, track model and bridge model comprises:

[0020] The wheel-rail interaction between the train and the track, between the track and the bridge, and the bridge-rail interaction are obtained through the train model, the track model, and the bridge model, and a train-track-bridge coupling system calculation model is established based on the wheel-rail interaction between the train and the track, between the track and the bridge, and the bridge-rail interaction;

[0021] The wheel-rail interaction includes: normal action and tangential action between the wheelset and the rail;

[0022] The normal action is obtained by using nonlinear Hertz contact theory, and the tangential action is obtained by using Kaller theory;

[0023] The bridge rails interact with each other to connect the main beam and the base plate.

[0024] Preferably,

[0025] The method of obtaining the earthquake-induced damage value of the components of the track-bridge system includes:

[0026] Generate random seismic waves that take site characteristics into consideration by adopting an artificially synthesized seismic wave method or a natural seismic wave amplitude modulation method, input the random seismic waves into the train-track-bridge coupling system calculation model, and obtain the seismic damage values ​​of components of the track-bridge system;

[0027] or,

[0028] The earthquake-induced damage data of the track-bridge system after the historical earthquake is obtained and used as the earthquake-induced damage value of the components of the track-bridge system.

[0029] Preferably,

[0030] The said establishing a balanced differential equation of the mechanical model of the bridge and track structure according to the mechanical characteristics of the track-bridge structure, and obtaining the mapping relationship between the earthquake-induced damage of the key components of the track-bridge and the track irregularity based on the balanced differential equation includes:

[0031] Establishing deformation expressions of rails, track plates and base plates according to the track structure mechanics model, and obtaining matrix expressions of rail deformation, track plate deformation and base plate deformation according to the deformation expressions of rails, track plates and base plates respectively;

[0032] Establishing a deformation expression of the bridge girder according to the bridge structure mechanics model;

[0033] Obtain the spring force expression of each interlayer component of the track structure under the action of horizontal deformation, substitute the matrix expression of rail deformation, the matrix expression of track plate deformation, the matrix expression of base plate deformation and the deformation expression of bridge main beam into the spring force expression of each interlayer component of the track structure under the action of horizontal deformation, and solve to obtain the fastener spring force matrix under horizontal action;

[0034] The fastener spring force matrix under the horizontal action is substituted into the matrix expression of the rail deformation to obtain the mapping relationship between the seismic damage of the track-bridge key components and the track irregularity.

[0035] Preferably,

[0036] The obtaining of the initial track irregularity sample and the additional track irregularity sample comprises:

[0037] Acquire an initial track irregularity power spectrum, and generate the initial track irregularity sample by a numerical method based on the initial track irregularity power spectrum;

[0038] The seismic damage values ​​of the components of the track-bridge system are input into the mapping relationship between the seismic damage of the key components of the track-bridge and the track irregularity to obtain additional track irregularity samples.

[0039] Preferably,

[0040] The method of training the pre-built model framework by using the first parameter and the second parameter to establish a quantitative mapping relationship between the structural earthquake-induced damage and the driving performance of the track-bridge system includes:

[0041] Preprocessing the first parameter and the second parameter, and selecting a deep learning model according to the characteristics and type of the data;

[0042] Using the first parameter and the second parameter to construct a data set, and using the constructed data set to train the selected deep learning model, and using the trained model as a quantitative mapping relationship model between structural earthquake-induced damage and driving performance of the track-bridge system;

[0043] Inputting the earthquake-induced damage values ​​of the components of the track-bridge system to be detected into a quantitative mapping relationship model between the structural earthquake-induced damage and the driving performance of the track-bridge system, and outputting the train-track-bridge coupled dynamic response results of the track-bridge system;

[0044] Driving safety and stability indicators are selected, and the seismic damage threshold of the track-bridge structure is determined based on the obtained train-track-bridge coupled dynamic response results of the track-bridge system and the driving performance indicator limit values ​​specified in the specifications.

[0045] According to a second aspect of an embodiment of the present invention, a railway bridge damage and driving performance mapping device is provided, the device comprising:

[0046] Railway bridge coupling calculation module: used to establish train models of different train models using multi-rigid body dynamics theory, and to establish track models and bridge models using finite element theory; and to establish a train-track-bridge coupling system calculation model through the train model, track model and bridge model;

[0047] Component seismic damage numerical acquisition module: used to obtain the component seismic damage numerical value of the track-bridge system;

[0048] Track irregularity mapping module: used to set the basic assumptions of the track-bridge structure theoretical model, and establish the equilibrium differential equations of the bridge and track structure mechanical models according to the mechanical characteristics of the track-bridge structure, and obtain the mapping relationship between the seismic damage of key track-bridge components and track irregularity based on the equilibrium differential equations;

[0049] An irregularity sample acquisition module is used to respectively acquire an initial track irregularity sample and an additional track irregularity sample, and superimpose the initial track irregularity sample and the additional track irregularity sample to obtain a vibration-induced random track irregularity sample;

[0050] A coupled dynamic response module is used to input the vibration-induced random track irregularity samples and the initial track irregularity samples into the train-track-bridge coupled system calculation model, respectively, to obtain two sets of train-track-bridge coupled dynamic response results;

[0051] A training parameter acquisition module is used to use the earthquake-induced damage value of the components of the track-bridge system as the first parameter; subtract the two sets of train-track-bridge coupled dynamic response results to obtain the train-track-bridge coupled dynamic response result, and use the train-track-bridge coupled dynamic response result as the second parameter;

[0052] Mapping relationship acquisition module: used to train the pre-built model architecture through the first parameter and the second parameter, and then establish a quantitative mapping relationship between the structural seismic damage and driving performance of the track-bridge system.

[0053] According to a third aspect of an embodiment of the present invention, there is provided a storage medium, wherein the storage medium stores a computer program, and when the computer program is executed by a main controller, each step in the above method is implemented.

[0054] The technical solution provided by the embodiments of the present invention may have the following beneficial effects:

[0055] This application comprehensively considers the impact of seismic damage to the track-bridge system on driving safety performance by establishing a mapping relationship between seismic damage to key components of the track-bridge and track irregularities; adopts a deep learning method to train the deep learning model through the seismic damage values ​​of the components of the track-bridge system and the train-track-bridge coupled dynamic response results, and then establishes a quantitative mapping relationship between the structural seismic damage to the track-bridge system and the driving performance, thereby quickly and accurately providing safety and stability indicators for driving on the bridge after the key components or parts of the railway bridge track are damaged by an earthquake.

[0056] It is to be understood that the foregoing general description and the following detailed description are exemplary and explanatory only and are not restrictive of the invention. BRIEF DESCRIPTION OF THE DRAWINGS

[0057] The accompanying drawings, which are incorporated in and constitute a part of this specification, illustrate embodiments consistent with the invention and, together with the description, serve to explain the principles of the invention.

[0058] Figure 1 It is a flow chart of a method for mapping railway bridge damage and driving performance according to an exemplary embodiment;

[0059] Figure 2 is a schematic diagram of a train model according to another exemplary embodiment;

[0060] Figure 3 is a front schematic diagram of a track-bridge calculation model according to another exemplary embodiment;

[0061] Figure 4 is a schematic side view of a track-bridge calculation model according to another exemplary embodiment;

[0062] Figure 5 is a seismic response spectrum diagram shown according to another exemplary embodiment;

[0063] Figure 6 is a flowchart for calculating the mapping relationship between earthquake-induced damage to key components of a track-bridge and track irregularity according to another exemplary embodiment;

[0064] Figure 7 is a schematic diagram of a sample of vibration-induced rail unevenness according to another exemplary embodiment;

[0065] Figure 8 is a schematic diagram of a deep learning analysis process according to another exemplary embodiment;

[0066] Fig. 9 is a system schematic diagram of a railway bridge damage and driving performance mapping device according to another exemplary embodiment;

[0067] In the attached figure: 1-railway bridge coupling calculation module, 2-component seismic damage numerical acquisition module, 3-track irregularity mapping module, 4-irregularity sample acquisition module, 5-coupled dynamic response module, 6-training parameter acquisition module, 7-mapping relationship acquisition module. DETAILED DESCRIPTION

[0068] Exemplary embodiments will be described in detail herein, examples of which are shown in the accompanying drawings. When the following description refers to the drawings, the same numbers in different drawings represent the same or similar elements unless otherwise indicated. The embodiments described in the following exemplary embodiments do not represent all embodiments consistent with the present invention. Instead, they are merely examples of devices and methods consistent with some aspects of the present invention as detailed in the appended claims.

[0069] Embodiment 1

[0070] Figure 1 is a flow chart of a method for mapping railway bridge damage and driving performance according to an exemplary embodiment. Figure 1 As shown, the method includes:

[0071] S1, using multi-rigid body dynamics theory to establish train models of different train models, using finite element theory to establish track models and bridge models respectively; and establishing a train-track-bridge coupling system calculation model through the train model, track model and bridge model;

[0072] S2, obtain the earthquake-induced damage values ​​of the components of the track-bridge system;

[0073] S3, setting the basic assumptions of the track-bridge structure theoretical model, and establishing the equilibrium differential equation of the bridge and track structure mechanical model according to the mechanical characteristics of the track-bridge structure, and obtaining the mapping relationship between the seismic damage of the key components of the track-bridge and the track irregularity based on the equilibrium differential equation;

[0074] S4, respectively obtaining an initial track irregularity sample and an additional track irregularity sample, and superimposing the initial track irregularity sample and the additional track irregularity sample to obtain a vibration-induced random track irregularity sample;

[0075] S5, inputting the seismic-induced random track irregularity samples and the initial track irregularity samples into the train-track-bridge coupling system calculation model, respectively, to obtain two sets of train-track-bridge coupling dynamic response results;

[0076] S6, using the earthquake-induced damage value of the components of the track-bridge system as the first parameter; subtracting the two sets of train-track-bridge coupled dynamic response results to obtain a train-track-bridge coupled dynamic response result, and using the train-track-bridge coupled dynamic response result as the second parameter;

[0077] S7, training the pre-built model framework by using the first parameter and the second parameter, thereby establishing a quantitative mapping relationship between the structural earthquake-induced damage and the driving performance of the track-bridge system;

[0078] Understandably,

[0079] S1, establishing a train-track-bridge coupling system calculation model including: train model, track model and bridge model;

[0080] The train model includes a car body and a bogie. A single bogie consists of four wheel sets, a primary suspension and a secondary suspension. The single-section train model is specifically divided into a car body, two frames and four wheel sets, a total of seven rigid bodies. Each rigid body has five degrees of freedom: lateral movement, sinking and floating, rolling, nodding and shaking, a total of 35 degrees of freedom, as shown in the attached figure. Figure 2 As shown in the figure, d0 is the fixed distance of the vehicle; d1 is the fixed wheelbase; W and T are the wheelset and frame respectively; k p and C p are suspension stiffness and damping respectively; k s , C s They are the stiffness and damping of the secondary suspension respectively; the establishment of the train dynamics model is realized by the multi-body dynamics software SIMPACK. When modeling, the parametric modeling method is adopted. The train body, front and rear bogies, and four wheel sets are simulated by rigid bodies. The primary and secondary suspension systems are simulated by spring-damping. The mechanical and geometric parameters of different components of the studied model are input into the pre-defined variable group. The purpose of quickly changing different models can be achieved by changing the values ​​of the parameter variable group according to the research needs of different models.

[0081] The track model, as shown in the attached Figure 3 As shown, it is mainly a CRTSⅠ type double-block ballastless track, which includes a base plate, an isolation layer and a limit groove, a track plate, a fastener layer and a rail from bottom to top; the track can be established using ANSYS finite element software, in which the base plate, track plate and rail can be simulated using the Timoshenko beam BEAM188 unit, and the isolation layer, limit groove and fasteners and other interlayer connection components can be simulated using the nonlinear spring COMBIN39 unit;

[0082] The bridge model is shown in the attached Figure 4As shown, it is mainly a simply supported beam bridge, which includes foundation, abutment, pier, support and main beam from bottom to top; the bridge can be simulated by ANSYS finite element software, in which the foundation, abutment, pier and main beam are simulated by Timoshenko beam BEAM188 unit, and the bridge support is simulated by nonlinear spring COMBIN39 and COMBIN40 unit;

[0083] The train-track-bridge coupling system calculation model is formed by the train model, the track model and the bridge model according to the wheel-rail interaction and the bridge-rail interaction between the train and the track, and between the track and the bridge, wherein the wheel-rail interaction is divided into the normal action and the tangential action between the wheelset and the rail, the wheel-rail normal action is obtained by adopting the nonlinear Hertz contact theory, and the wheel-rail tangential creep action is calculated by adopting the Kaller theory. The specific calculation formula is as follows:

[0084] According to Hertz contact theory, the maximum contact pressure p in the contact spot between two smooth surface objects is max It can be obtained from formula (1):

[0085]

[0086] Where: P is the normal load of the contact spot; m, n, A, B are constants related to the wheel-rail contact geometry; G* is the material physical parameter, which is determined by the following formula (2):

[0087]

[0088] Where, E1 is the elastic modulus of the steel wheel; E2 is the elastic modulus of the rail; v1 is the Poisson's ratio of the steel wheel; v2 is the Poisson's ratio of the rail;

[0089] The wheel-rail longitudinal and transverse creep effects are analyzed using the Kaller theory;

[0090] The calculation formulas for wheel-rail creep rate and creep force are:

[0091] Transverse creep rate:

[0092]

[0093] Longitudinal creep rate:

[0094]

[0095] Creep force:

[0096] F r = -fξ (5)

[0097] In the above formula, v x is the actual forward speed of the wheel; v xrv is the pure rolling forward speed of the wheel; y Actual lateral speed of the wheel; v yr is the pure rolling lateral speed of the wheel; b is half of the rolling circle spacing between the left and right wheels; r0 is the nominal radius; λ0 is the equivalent slope of the tread; ψ, is the wheel set shaking angle; Ω is the forward speed of the wheelset;

[0098] The bridge-rail interaction is mainly the connection between the main beam and the base plate, which is a rigid connection;

[0099] The train-track-bridge coupling system calculation model is implemented through the joint simulation of ANSYS and SIMPACK; the specific steps are as follows:

[0100] ① Generation of CDB files of geometric mesh information of track-bridge finite element model: Generate separate CDB files for the finite element models of main components such as rails, track plates, base plates, main beams and bridge piers in ANSYS;

[0101] ② Generation of SUB files for substructure analysis of track-bridge finite element model: Based on the CDB files of each component, perform substructure analysis solution settings in ANSYS. The solution results should include the finite element model stiffness and mass matrix and concentrated mass matrix; based on the modal superposition theory, set the main nodes and main degrees of freedom of the finite element model; generate SUB files for each component after solution;

[0102] ③ Generation of flexible body FBI files of finite element models: Based on the CDB and SUB files of each component, generate flexible body FBI files of each component in SIMPACK, which contain information such as node degrees of freedom, super unit mass, stiffness matrix, etc.;

[0103] ④ Prepare the FTR flexible track basic file: When preparing the FTR flexible track basic file, the track structure is spliced ​​in sections in the FTR file by setting the call and assembly rules of the FBI files of individual components such as rails, roadbed plates, and base plates. In addition, the transition section between the elastic rails on the main beam and the rigid rails in SIMPACK should be set in the FTR file, and the stiffness and damping should be set for it; and the tolerance, stiffness and quality screening values ​​should be set;

[0104] ⑤Import the track structure FTR file into the SIMPACK model with the vehicle model already built;

[0105] ⑥ Import the FBI files of the main beam, piers and other components into the vehicle-track coupling model of SIMPACK and locate them;

[0106] ⑦ Create marker points on each component in SIMPACK;

[0107] ⑧Set force elements between the marker points of different components to achieve the connection between bridge structures (such as fasteners, supports, etc.);

[0108] S2, numerical acquisition of earthquake-induced damage of track-bridge system components includes: numerical calculation method and post-earthquake data query method;

[0109] The numerical calculation method is: by adopting the method of artificially synthesizing seismic waves or the method of amplitude modulation of natural seismic waves, random seismic waves that take into account the site characteristics are generated. The site type of the generated seismic waves must match the building site, and according to the requirements of the "Code for Seismic Design of Buildings", the error between the average value of several seismic wave response spectra and the standard response spectrum is within ±20%, as shown in the attached Figure 5 As shown; the generated seismic waves are then substituted into the established train-track-bridge coupling calculation model, and the seismic damage values ​​of the components of the track-bridge system can be obtained through calculation;

[0110] The post-earthquake data query method is as follows: the earthquake damage data of the track-bridge system recorded after the earthquakes in previous years can be queried and used as the earthquake damage value of the track-bridge system components that can be input;

[0111] S3, the mapping relationship between earthquake-induced damage of key track-bridge components and track irregularity is obtained, including:

[0112] First, the basic assumptions of the theoretical model of the track-bridge structure are determined. Then, based on the mechanical characteristics of the track-bridge structure, the equilibrium differential equations of the mechanical models of the bridge structure and the track structure are constructed, and then the deformation of each component is solved. Then, the deformation equations of the track component and the bridge component are linked through the forces of the system interlayer components, and the mapping relationship between the earthquake-induced damage deformation of different interlayer components and the rail deformation can be solved.

[0113] The basic assumptions of the theoretical model of the track-bridge structure mainly include:

[0114] ① Ignore the influence of track on bridge deformation;

[0115] ② The track interlayer components such as the fastener system are assumed to be springs evenly distributed along the centerline of the rail;

[0116] ③ Consider that the track of the roadbed section has sufficient track extension sections, and assume that the track boundary of the extension section is simply supported;

[0117] ④ It is considered that the deformation of the base plate is consistent with that of the bridge;

[0118] According to the track structure mechanics model, the mechanical equilibrium differential equation of the rail is established, and the expression (6) of the rail deformation is obtained by solving it, as shown below:

[0119]

[0120] Y ri is the deformation at the rail fastener position; E r I r is the bending stiffness of the rail; l MNK+1 is the total length of the rail; F j is the fastener spring force, l i is the distance from the rail end point to the i-th fastener force; M is the number of bridge spans; N is the number of main beam track plates per span; K is the number of fasteners per track plate.

[0121] From formula (6), the matrix expression of rail deformation is as follows:

[0122] [Y r ]=[R][F] (7)

[0123] In the formula, [Y r ] is the deformation matrix of the rail; [R] is the spring force coefficient matrix of the rail fastener; [F] is the spring force matrix of the fastener;

[0124] According to the track structure mechanics model, the mechanical equilibrium differential equation of the track plate is established, and the expression (8) of the deformation of a single track plate is obtained by solving it, as shown below:

[0125]

[0126] Where Y smi is the deformation of the mth track plate at the fastener position; F mj is the spring force of the fastener at the jth fastener position of the mth track plate; Y bm (x) is the displacement of the main beam at the position of the mth track plate; l mi is the distance from the track plate end point to the i-th fastener force; d m1 With d m2 are the distances from the end point of the limiting groove to the left end point of the track plate; k d is the stiffness of the isolation layer; k g A is the stiffness of the elastic pad; m1 , A m2 With A m3 is a constant;

[0127] Given the deformation of each position of the mth track plate, the deformation of the M×N track plates on the M-span main beam can be organized into the following matrix form:

[0128] [Y s ]=[S][F]+[C] (9)

[0129] In the formula, [Y s] is the deformation matrix of M×N track plates; [S] is the fastener spring force coefficient matrix of the track plate; [F] is the fastener spring force matrix of the track plate; [C] is the influence matrix of the main beam deformation on the track plate deformation;

[0130] The displacement expression of the bridge girder is as follows:

[0131]

[0132] Where Y bi is the displacement of any position of the main beam of the i-th span; h i1 is the residual displacement of the left end support of the main beam of the i-th span; is the main beam deflection angle; L b L is the total length of a single span of a simply supported beam bridge; b0 is the distance from the support to the beam end; d is the distance between the expansion joints of the simply supported beams; X i is the distance from the point on the main beam of the i-th span to the coordinate origin;

[0133] Under the action of horizontal deformation, the fastener spring force can be expressed as:

[0134] [F] = k r ([Y s ]-[Y r ]) (11)

[0135] In the formula, k r is the fastener stiffness;

[0136] Substituting equation (7), equation (9) and equation (10) into equation (11), we can obtain:

[0137] [F]=([I]-k r [S]+k r [R]) -1 k r [C] (12)

[0138] Substituting formula (12) into formula (7), we can get the mapping deformation expression of the rail:

[0139] [Y r ]=[R]([I]-k r [S]+k r [R]) -1 k r [C] (13)

[0140] Since the mapping model between bridge structure deformation and track surface geometry has many nodes to solve and a large amount of computational workload, the track deformation can be solved using MATLAB. The solution process is as follows: Figure 6 As shown;

[0141] S4, acquisition of random track irregularity samples caused by earthquakes includes:

[0142] First, determine the initial track irregularity power spectrum, such as the German railway track irregularity power spectrum, the British railway track irregularity power spectrum, and the Chinese high-speed railway ballastless track irregularity power spectrum. This embodiment takes the Chinese high-speed railway ballastless track irregularity spectrum as an example. The power spectrum is expressed in the form of a piecewise power function, which can better describe the distribution characteristics of ballastless track irregularity at different frequencies:

[0143] S(f)=Af -k (14)

[0144] Where, f is the spatial frequency (1 / m); A is the fitting coefficient;

[0145] Secondly, based on the selected initial track irregularity power spectrum, the initial track geometric irregularity samples are generated by numerical methods. Here, the triangular series superposition method is used. This method can quickly generate track irregularity samples with specific power spectrum density. The calculation formula is as follows:

[0146]

[0147] Where w(x) represents the generated random sample sequence of track irregularity; x is the mileage from the origin to the track extension position; S(f k ) is the selected track irregularity power spectrum; f k Indicates the spatial frequency of the power spectrum. When k = 1, N, it indicates the upper and lower cutoff limits of the spatial frequency of the orbital spectrum; Φ k It represents the random phase corresponding to the k-th spatial frequency, and the value is randomly generated according to uniform distribution, ranging from 0 to 2π; Δf represents the bandwidth of the spatial frequency interval;

[0148] Substitute the earthquake-induced damage values ​​of the track-bridge system into the mapping relationship between earthquake-induced damage of key track-bridge components and track irregularity to obtain additional track irregularity samples;

[0149] Finally, the initial track irregularity samples and the additional track irregularity samples are superimposed to obtain the vibration-induced random track irregularity samples, such as Figure 7 As shown;

[0150] S5, two sets of train-track-bridge coupling dynamic response results are obtained, including:

[0151] First, the random track irregularity samples and initial track irregularity samples induced by earthquakes are obtained through the above steps, and then they are respectively substituted into the established train-track-bridge coupling calculation model; finally, two sets of calculation results of train-track-bridge coupling dynamic responses are calculated, namely, the calculation results of only inputting the initial track irregularity and the calculation results of inputting the superimposed track irregularity induced by earthquakes;

[0152] S6, obtaining the first parameter and the second parameter includes:

[0153] First, the track-bridge system seismic damage value obtained above is used as the first parameter; second, the two sets of train-track-bridge coupled dynamic response results obtained above are subtracted to obtain the train-track-bridge coupled dynamic response result corresponding to the track-bridge system seismic damage, which is used as the second parameter;

[0154] S7, training the pre-built model framework by using the first parameter and the second parameter, and then establishing a quantitative mapping relationship between the structural earthquake-induced damage and the driving performance of the track-bridge system, including:

[0155] Deep learning, an important implementation method in artificial neural networks, is used to establish a quantitative mapping relationship between earthquake-induced damage to track-bridge structures and driving safety and stability indicators. The value of earthquake-induced damage to track-bridge structures includes structural damage under different earthquakes. Finally, based on the established quantitative mapping relationship, train driving safety and stability indicators are predicted, and the earthquake-induced damage threshold of track-bridge structures is determined.

[0156] Specifically, the analysis process of deep learning is as follows: Figure 8 As shown, the basic steps are as follows:

[0157] (1) Data collection and preprocessing

[0158] First, based on the values ​​obtained in step B and step F, the data set required for deep learning is preliminarily established; secondly, the data set is cleaned and denoised to ensure that the earthquake-induced damage data of the track-bridge system and the train-track-bridge coupled dynamic response data are completely caused by the earthquake, including the extremely small deformation data generated by the track-bridge system under its own weight; and by preprocessing the data, the quality and availability of the data can be improved, making the subsequent analysis process more accurate and reliable; finally, the data is divided into training set, validation set and test set. This division helps to build, optimize and evaluate the machine learning model to ensure its generalization ability on unknown data; the general division ratio is: training set 60%-70%, validation set 15%-20%, test set 15%-20%;

[0159] (2) Design model architecture

[0160] First, according to the characteristics and types of data, select the appropriate deep learning model architecture, such as multi-layer perceptron (MLP), convolutional neural network (CNN), recurrent neural network (RNN), etc. According to the characteristics of this research direction, the deep learning model framework used is MLP as an example. The basic architecture of MLP includes input layer, hidden layer and output layer. The working process of MLP can be divided into two stages: forward propagation and backward propagation.

[0161] Secondly, design the structure of MLP, such as the number of layers, number of nodes, and activation function; ① Number of layers: Set 3-5 hidden layers. A shallower network has the advantages of fast training speed and easy debugging; ② Number of nodes: The number of nodes in the hidden layer is usually set according to the number of input features and the complexity of the task. Generally speaking, the number of nodes should not be less than the number of input features; the rule of thumb is to start with a larger number of nodes (for example, 128, 256, 512), and then adjust according to the performance of the model; ③ Activation function: Commonly used activation functions include Relu, Sigmoid, and Tanh, etc. Compared with the Sigmoid function that maps the input to the [0,1] interval, the Tanh function's mapping output range is [-1,1], which is more in line with the output range of the data, so this embodiment uses the Tanh function as the activation function, and the formula is as follows:

[0162]

[0163] (3) Model training

[0164] First, the designed deep learning model is trained using the training data set, and the model parameters are continuously adjusted through the back-propagation algorithm to minimize the loss function;

[0165] Then, consider training techniques such as optimization algorithms and learning rate adjustment strategies to improve the training efficiency and performance of the model. In order to improve the training efficiency and performance of deep learning models, the Adam optimization algorithm can be used in combination with momentum and adaptive learning rate to dynamically adjust the learning rate.

[0166] Finally, monitor indicators during the training process, such as loss value, accuracy, etc., so as to adjust the model or training parameters in time. Common methods include recording the loss value and accuracy of the training set and validation set at the end of each training cycle (epoch), and visually observing the performance changes of the model by drawing the loss curve and accuracy curve.

[0167] (4) Model evaluation and tuning

[0168] First, run the model through the validation set and calculate performance indicators such as loss value and accuracy;

[0169] Then, analyze the indicators and evaluate the results. According to the evaluation results, adjust the model structure (such as the number of layers and nodes) and training parameters (such as learning rate and regularization strength), and further optimize using techniques such as cross-validation and hyperparameter search.

[0170] Finally, through iterative evaluation and tuning, the performance and generalization ability of the model are gradually improved.

[0171] (5) Model preservation and application

[0172] Use the test set to test the final tuned model, evaluate its performance on unknown real data, analyze the test results, summarize the advantages and disadvantages of the model, and further improve and save the model as needed. Save the trained model as a formula or program and apply it to actual scientific problems for prediction and other tasks.

[0173] Assume that the earthquake-induced damage matrix of track-bridge components obtained by S2 is D, which includes the earthquake-induced damage of bridge piers, bearings and track components of the track-bridge system;

[0174] Assume the weight matrix obtained by deep learning is W i , W i Represented as the weight matrix of the i-th layer network structure;

[0175] Let the bias value obtained by deep learning be J i , J i Represented as the bias value of the i-th layer network structure;

[0176] Assuming that the dynamic response result of the train-track-bridge output by the model is V, the corresponding relationship between the seismic damage of the track-bridge system components and the driving response result is as follows:

[0177] V=T(T(T(DW1+J1)W2+J2)W3+J3)W4+J4 (17)

[0178] Where T(·) represents the Tanh function transformation of the matrix;

[0179] It is worth emphasizing that the premise of this formula is to set up 3 hidden layers and the output layer does not set an activation function. The specific situation can be changed accordingly as needed, and this application does not impose any restrictions on this.

[0180] Driving safety and stability indicators are selected, among which driving safety indicators can be divided into indicators based on wheel-rail force criteria and indicators based on wheel-rail relative displacement criteria, and driving stability indicators are mainly train acceleration indicators and Sperling indicators; and according to the "Locomotive and Train Dynamic Performance Evaluation and Experimental Appraisal Specifications" and "High-speed Railway Design Specifications", the train driving safety and stability indicator limits can be determined as shown in Table 1 below:

[0181] Table 1

[0182]

[0183] In Table 1, Q is the lateral horizontal force of the wheelset; P is the dynamic wheel weight; ΔP is the vehicle axle weight reduction; P0 is the static axle weight; a z is the vertical vibration acceleration of the vehicle body; a y is the lateral vibration acceleration of the vehicle body; g is the acceleration due to gravity, which is generally taken as 9.8m / s 2 .

[0184] Based on the dynamic response results of the train-track-bridge output by the model and the limit values ​​of the train driving safety and stability indicators, the seismic damage threshold of the track-bridge structure is determined.

[0185] Embodiment 2:

[0186] Fig. 9 : is a system schematic diagram of a railway bridge damage and driving performance mapping device according to another exemplary embodiment, the device comprising:

[0187] Railway bridge coupling calculation module 1: used to establish train models of different train models using multi-rigid body dynamics theory, and to establish track models and bridge models respectively using finite element theory; and to establish a train-track-bridge coupling system calculation model through the train model, track model and bridge model;

[0188] Component seismic damage numerical acquisition module 2: used to obtain the component seismic damage numerical value of the track-bridge system;

[0189] Track irregularity mapping module 3: used to set the basic assumptions of the track-bridge structure theoretical model, and establish the equilibrium differential equation of the bridge and track structure mechanical model according to the mechanical characteristics of the track-bridge structure, and obtain the mapping relationship between the seismic damage of the key components of the track-bridge and the track irregularity based on the equilibrium differential equation;

[0190] The irregularity sample acquisition module 4 is used to respectively acquire the initial track irregularity sample and the additional track irregularity sample, and superimpose the initial track irregularity sample and the additional track irregularity sample to obtain the vibration-induced random track irregularity sample;

[0191] Coupled dynamic response module 5: used to input the vibration-induced random track irregularity samples and the initial track irregularity samples into the train-track-bridge coupled system calculation model, respectively, to obtain two sets of train-track-bridge coupled dynamic response results;

[0192] Training parameter acquisition module 6: used to use the earthquake-induced damage value of the track-bridge system as the first parameter; subtract the two sets of train-track-bridge coupled dynamic response results to obtain the train-track-bridge coupled dynamic response result, and use the train-track-bridge coupled dynamic response result as the second parameter;

[0193] Mapping relationship acquisition module 7: used to train the pre-built model framework through the first parameter and the second parameter, and then establish a quantitative mapping relationship between the structural seismic damage and driving performance of the track-bridge system.

[0194] Embodiment three:

[0195] This embodiment provides a storage medium, wherein the storage medium stores a computer program, and when the computer program is executed by a main controller, each step in the above method is implemented;

[0196] It is understandable that the storage medium mentioned above can be a read-only memory, a magnetic disk or an optical disk, etc.

[0197] It can be understood that the same or similar parts of the above embodiments can be referenced to each other, and the contents not described in detail in some embodiments can refer to the same or similar contents in other embodiments.

[0198] It should be noted that, in the description of the present invention, the terms "first", "second", etc. are only used for descriptive purposes and cannot be understood as indicating or implying relative importance. In addition, in the description of the present invention, unless otherwise specified, the meaning of "plurality" refers to at least two.

[0199] Any process or method description in a flowchart or otherwise described herein may be understood to represent a module, segment or portion of code that includes one or more executable instructions for implementing the steps of a specific logical function or process, and the scope of the preferred embodiments of the present invention includes alternative implementations in which functions may not be performed in the order shown or discussed, including performing functions in a substantially simultaneous manner or in the reverse order depending on the functions involved, which should be understood by those skilled in the art to which the embodiments of the present invention belong.

[0200] It should be understood that the various parts of the present invention can be implemented by hardware, software, firmware or a combination thereof. In the above-mentioned embodiments, a plurality of steps or methods can be implemented by software or firmware stored in a memory and executed by a suitable instruction execution system. For example, if implemented by hardware, as in another embodiment, it can be implemented by any one of the following technologies known in the art or their combination: a discrete logic circuit having a logic gate circuit for implementing a logic function for a data signal, a dedicated integrated circuit having a suitable combination of logic gate circuits, a programmable gate array (PGA), a field programmable gate array (FPGA), etc.

[0201] A person skilled in the art may understand that all or part of the steps in the method for implementing the above-mentioned embodiment may be completed by instructing related hardware through a program, and the program may be stored in a computer-readable storage medium, which, when executed, includes one or a combination of the steps of the method embodiment.

[0202] In addition, each functional unit in each embodiment of the present invention may be integrated into a processing module, or each unit may exist physically separately, or two or more units may be integrated into one module. The above-mentioned integrated module may be implemented in the form of hardware or in the form of a software functional module. If the integrated module is implemented in the form of a software functional module and sold or used as an independent product, it may also be stored in a computer-readable storage medium.

[0203] The storage medium mentioned above can be a read-only memory, a magnetic disk or an optical disk, etc.

[0204] In the description of this specification, the description with reference to the terms "one embodiment", "some embodiments", "examples", "specific examples", or "some examples" means that the specific features, structures, materials or characteristics described in conjunction with the embodiment or example are included in at least one embodiment or example of the present invention. In this specification, the schematic representation of the above terms does not necessarily refer to the same embodiment or example. Moreover, the specific features, structures, materials or characteristics described may be combined in any one or more embodiments or examples in a suitable manner.

[0205] Although the embodiments of the present invention have been shown and described above, it is to be understood that the above embodiments are exemplary and are not to be construed as limitations of the present invention. A person skilled in the art may change, modify, replace and vary the above embodiments within the scope of the present invention.

Claims

1. A railway bridge damage and driving performance mapping method, characterized in that: The method comprises: The train models of different train models are established by using multi-rigid body dynamics theory, and the track model and bridge model are established by using finite element theory; the train-track-bridge coupling system calculation model is established by using the train model, track model and bridge model; Obtain the earthquake-induced damage values ​​of components in the track-bridge system; The basic assumptions of the theoretical model of the track-bridge structure are set, and the equilibrium differential equations of the mechanical model of the bridge and track structure are established according to the mechanical characteristics of the track-bridge structure. Based on the equilibrium differential equations, the mapping relationship between the seismic damage of the key components of the track-bridge and the track irregularity is obtained; Acquire initial track irregularity samples and additional track irregularity samples respectively, and superimpose the initial track irregularity samples and the additional track irregularity samples to obtain vibration-induced random track irregularity samples; The seismic-induced random track irregularity samples and the initial track irregularity samples are respectively input into the train-track-bridge coupling system calculation model to obtain two sets of train-track-bridge coupling dynamic response results; The earthquake-induced damage value of the components of the track-bridge system is used as the first parameter; the train-track-bridge coupled dynamic response result is obtained by subtracting the two sets of train-track-bridge coupled dynamic response results, and the train-track-bridge coupled dynamic response result is used as the second parameter; The pre-built model framework is trained by using the first parameter and the second parameter, thereby establishing a quantitative mapping relationship between the structural seismic damage and the driving performance of the track-bridge system.

2. The method according to claim 1, characterized in that The method of using multi-rigid body dynamics theory to establish train models of different train models and using finite element theory to respectively establish track models and bridge models includes: The train model is obtained by using a rigid body to simulate the train body, front and rear bogies and four wheel sets, and using a spring-damping system to simulate the primary and secondary suspension systems of the train; The track model is obtained by using a spatial beam unit or a plate unit to simulate the base plate and the track plate of the track, using a spatial beam unit to simulate the rail of the track, and using a spring-damping unit to simulate the interlayer connection components of the track; The spatial beam unit is used to simulate the foundation, abutment, piers and main beam of the bridge, and the spring-damping is used to simulate the support of the bridge to obtain the bridge model.

3. The method according to claim 2, characterized in that The establishing of the train-track-bridge coupling system calculation model by using the train model, track model and bridge model comprises: The wheel-rail interaction between the train and the track, between the track and the bridge, and the bridge-rail interaction are obtained through the train model, the track model, and the bridge model, and a train-track-bridge coupling system calculation model is established based on the wheel-rail interaction between the train and the track, between the track and the bridge, and the bridge-rail interaction; The wheel-rail interaction includes: normal action and tangential action between the wheelset and the rail; The normal action is obtained by using nonlinear Hertz contact theory, and the tangential action is obtained by using Kaller theory; The bridge rails interact with each other to connect the main beam and the base plate.

4. The method according to claim 1, characterized in that: The method of obtaining the earthquake-induced damage value of the components of the track-bridge system comprises: Generate random seismic waves that take site characteristics into consideration by adopting an artificially synthesized seismic wave method or a natural seismic wave amplitude modulation method, input the random seismic waves into the train-track-bridge coupling system calculation model, and obtain the seismic damage values ​​of components of the track-bridge system; or, The earthquake-induced damage data of the track-bridge system after the historical earthquake is obtained and used as the earthquake-induced damage value of the components of the track-bridge system.

5. The method according to claim 4, characterized in that The said establishing a balanced differential equation of the mechanical model of the bridge and track structure according to the mechanical characteristics of the track-bridge structure, and obtaining the mapping relationship between the earthquake-induced damage of the key components of the track-bridge and the track irregularity based on the balanced differential equation includes: Establishing deformation expressions of rails, track plates and base plates according to the track structure mechanics model, and obtaining matrix expressions of rail deformation, track plate deformation and base plate deformation according to the deformation expressions of rails, track plates and base plates respectively; Establishing a deformation expression of the bridge girder according to the bridge structure mechanics model; Obtain the spring force expression of each interlayer component of the track structure under the action of horizontal deformation, substitute the matrix expression of rail deformation, the matrix expression of track plate deformation, the matrix expression of base plate deformation and the deformation expression of bridge main beam into the spring force expression of each interlayer component of the track structure under the action of horizontal deformation, and solve to obtain the fastener spring force matrix under horizontal action; The fastener spring force matrix under the horizontal action is substituted into the matrix expression of the rail deformation to obtain the mapping relationship between the seismic damage of the track-bridge key components and the track irregularity.

6. The method according to claim 1, characterized in that The obtaining of the initial track irregularity sample and the additional track irregularity sample comprises: Acquire an initial track irregularity power spectrum, and generate the initial track irregularity sample by a numerical method based on the initial track irregularity power spectrum; The seismic damage values ​​of the components of the track-bridge system are input into the mapping relationship between the seismic damage of the key components of the track-bridge and the track irregularity to obtain additional track irregularity samples.

7. The method according to claim 1, characterized in that The method of training the pre-built model framework by using the first parameter and the second parameter to establish a quantitative mapping relationship between the structural earthquake-induced damage and the driving performance of the track-bridge system includes: Preprocessing the first parameter and the second parameter, and selecting a deep learning model according to the characteristics and type of the data; Using the first parameter and the second parameter to construct a data set, and using the constructed data set to train the selected deep learning model, and using the trained model as a quantitative mapping relationship model between structural earthquake-induced damage and driving performance of the track-bridge system; Inputting the earthquake-induced damage values ​​of the components of the track-bridge system to be detected into a quantitative mapping relationship model between the structural earthquake-induced damage and the driving performance of the track-bridge system, and outputting the train-track-bridge coupled dynamic response results of the track-bridge system; Driving safety and stability indicators are selected, and the seismic damage threshold of the track-bridge structure is determined based on the obtained train-track-bridge coupled dynamic response results of the track-bridge system and the driving performance indicator limit values ​​specified in the specifications.

8. A railway bridge damage and driving performance mapping device, characterized in that: The device comprises: Railway bridge coupling calculation module: used to establish train models of different train models using multi-rigid body dynamics theory, and to establish track models and bridge models using finite element theory; and to establish a train-track-bridge coupling system calculation model through the train model, track model and bridge model; Component seismic damage numerical acquisition module: used to obtain the component seismic damage numerical value of the track-bridge system; Track irregularity mapping module: used to set the basic assumptions of the track-bridge structure theoretical model, and establish the equilibrium differential equations of the bridge and track structure mechanical models according to the mechanical characteristics of the track-bridge structure, and obtain the mapping relationship between the seismic damage of key track-bridge components and track irregularity based on the equilibrium differential equations; An irregularity sample acquisition module is used to respectively acquire an initial track irregularity sample and an additional track irregularity sample, and superimpose the initial track irregularity sample and the additional track irregularity sample to obtain a vibration-induced random track irregularity sample; A coupled dynamic response module is used to input the vibration-induced random track irregularity samples and the initial track irregularity samples into the train-track-bridge coupled system calculation model, respectively, to obtain two sets of train-track-bridge coupled dynamic response results; A training parameter acquisition module is used to use the earthquake-induced damage value of the components of the track-bridge system as the first parameter; subtract the two sets of train-track-bridge coupled dynamic response results to obtain the train-track-bridge coupled dynamic response result, and use the train-track-bridge coupled dynamic response result as the second parameter; Mapping relationship acquisition module: used to train the pre-built model architecture through the first parameter and the second parameter, and then establish a quantitative mapping relationship between the structural seismic damage and driving performance of the track-bridge system.

9. A storage medium, characterized in that: The storage medium stores a computer program, and when the computer program is executed by the main controller, each step in the method for mapping railway bridge damage and driving performance as described in any one of claims 1 to 7 is implemented.

Images