CN116756833A - Dynamic identification and evaluation method, device, equipment and storage medium for railway bridge parameters - Google Patents

Abstract

The present invention relates to the technical field of railway bridge structures and sensor monitoring, and in particular to a method, device, equipment and storage medium for dynamic identification and evaluation of railway bridge parameters, including the use of long gauge strain sensors to obtain the information of a high-speed train passing through a railway simply-supported girder bridge. The long gauge strain response of the selected sample data is solved to obtain the long gauge strain time history curve, and then the area surrounded by the long gauge strain time history curve of all long gauge sensors is solved. The peak size and spacing are used as the parameters to be considered. To identify parameters, by establishing a coupled vibration model and conducting numerical simulations with a sufficient amount of different railway operating parameters, a sample library is obtained and used to train the neural network. The railway operating parameters can be obtained by substituting the parameters to be identified. The invention can realize rapid identification of train operating parameters on the upper part of the railway bridge without affecting operating traffic, and greatly improves monitoring efficiency.

Description

Railway bridge parameter dynamic identification and evaluation methods, devices, equipment and storage media

Technical field

The invention relates to the technical fields of railway bridge structures and sensor monitoring, and in particular to a method, device, equipment and storage medium for dynamic identification and evaluation of railway bridge parameters.

Background technique

High-speed railway bridges play an important role in the transportation system. While they provide people with safe, comfortable and convenient travel conditions, they are constantly subjected to the load of high-speed trains and the damage from the external environment, as well as the continuous improvement of the material properties of the bridge itself. Degradation, the bridge structure will suffer varying degrees of damage, causing its actual service life to be far less than its designed service life. Therefore, it is particularly urgent to carry out health monitoring research on existing high-speed railway bridges. Maintaining the health of bridges is crucial to the safe operation of high-speed railway trains.

Based on this, most of the existing technologies implement bridge health monitoring by testing bridge vibration, displacement, and strain. This method is usually used for large-span bridges. Due to the expensive installation and maintenance costs of the health monitoring system, it is not suitable for small- and medium-span bridges. There are few studies on bridges, and railway bridges are mainly medium- and small-span. Compared with highway bridges, the main characteristics of the load borne by railway bridges are that the load action position is fixed, the wheelbase and vehicle speed remain at the same level for a long time, and the axle load is larger. Based on these characteristics, by knowing the actual load acting on the structure in time, we can understand the working status of the bridge more intuitively and truly. We can also use experience or design specifications based on the characteristics of the load to estimate the bearing capacity of the bridge and the stability of the track. Make timely analysis and evaluation of flatness. In addition, for complex multi-parameter problems, since it is impossible to establish an ideal mathematical model, the parameters measured by the sensor obviously contain information about the railway operating parameters. Therefore, combined with the powerful ability of the BP neural network to solve nonlinear mapping problems, it can be easily It is good to identify the railway operating parameters from the obtained parameters. From the use of sensors, it can be found that because traditional strain sensors are "point-type" measurements, the measurement range is too local and cannot obtain the strain of the complete structure. Especially for long-span structures such as bridges, it is obviously unrealistic to cover all strain gauges, so There is an urgent need for a dynamic identification and evaluation method, device, equipment and storage medium for railway bridge parameters to solve the above problems.

Contents of the invention

The present invention aims to improve at least one of the technical problems existing in the prior art. To this end, the present invention proposes a method, device, equipment and storage medium for dynamic identification and evaluation of railway bridge parameters.

A method for dynamic identification and evaluation of railway bridge parameters according to the first embodiment of the present invention is characterized by including:

Step S1, build a dynamic identification and evaluation scenario for railway bridge parameters, and arrange several long-gauge strain sensors on the bridge to be monitored;

Step S2, collect the long-gauge strain response when the train load passes through the bridge to be monitored, and obtain the long-gauge strain response data;

Step S3: Solve the measured long-gauge strain response data, extract the quasi-static long-gauge strain signal, and obtain the long-gauge strain time history curve, thereby further obtaining the long-gauge strain influence line of the railway bridge to be identified;

Step S4: Solve the long-gauge strain time-history area, peak size and spacing of the long-gauge sensor based on the long-gauge strain time-history curve to form the input layer data of the railway bridge operating parameters to be identified;

Step S5, establish a train-track-bridge coupling vibration model, conduct numerical simulations with sufficient trains with different operating parameters, extract the long-gauge strain time history curve of the bridge sensor position under each train load, and obtain the enclosed long-gauge strain time Coverage area, vehicle speed, axle load and wheel-rail contact force curve;

Step S6: Use the simulated long gauge strain time history area, peak size and spacing as the input layer, and the vehicle speed, axle load, wheel-rail contact force curve as the output layer to establish a neural network model and train it to obtain the trained neural network. network model;

Step S7: Substitute the long-gauge strain time history area, peak size and spacing of the railway bridge to be identified as input layers into the trained neural network model to identify vehicle speed, axle load, and wheel-rail contact force curves;

Step S8: Analyze the long-gauge strain influence lines of the railway bridge to be identified, substitute them into the trained neural network model to identify railway operating parameters, and complete the dynamic identification and evaluation process of railway bridge parameters.

According to the dynamic identification and evaluation method of railway bridge parameters according to the embodiment of the present invention, the long gauge strain response of a high-speed train passing through a railway simply supported girder bridge is obtained by using a long gauge strain sensor, and the selected sample data is solved to obtain the long gauge strain time history curve, and then solve for the area enclosed by the long gauge strain time history curve of all long gauge sensors. The peak size and spacing are used as parameters to be identified. By establishing a train-track-bridge coupled vibration model, a sufficient amount of different railways The operating parameters are numerically simulated to obtain a rich sample library and used to train the neural network, and then substituted into the parameters to be identified to obtain the railway operating parameters. The invention can quickly identify the train operating parameters on the upper part of the railway bridge without affecting the operating traffic, greatly improves the monitoring efficiency, and provides guarantee for the operational safety of the bridge.

In a possible implementation of the first aspect, the long gauge sensor is a long gauge fiber grating sensor, or

Compared with traditional point strain gauges, long-gauge resistance strain sensors can complete the full-span layout of the bridge and realize bridge strain influence line monitoring.

In a possible implementation of the first aspect, in step S6, the long gauge strain time history area, peak size, spacing, etc. are used as the input layer, and the train speed, axle load, and wheel-rail contact force curve are established as the output layer Neural network model and training, including:

Record the strain time history area of each gauge sensor at the bottom of the bridge of the lth train as , the peak size is recorded as/> , the peak distance is/> , the vehicle speed is/> , the axle weight is/> , the wheel-rail contact force curve is/> , of which/> ,/> is the number of trains;

Establish a training model through a single hidden layer neural network, determine the range of the number of hidden layer neurons through empirical formulas, then design a BP network with a variable number of hidden layer neurons, compare the training errors of each network structure, and obtain the optimal The appropriate number of neurons, the empirical formula is as follows:

,

in, is the number of neurons in the hidden layer,/> is the number of neurons in the input layer,/> is the number of neurons in the output layer, for/> constant between;

The number of neurons in the input layer and output layer of the neural network is directly determined by the input parameters and output parameters of the network, where the input parameters of the input layer are ,/> is the input data, and the output parameters of the output layer are/> ,/> for output data. .

In a possible implementation of the first aspect, in step S6, the input layer parameters are all or part of the data obtained by long gauge sensor analysis, and the output layer is all or part of the data of the railway operating parameters, which can be trained at once. For all parameters, the network model of a single parameter can also be trained separately multiple times.

In a possible implementation of the first aspect, in step S6, when constructing a neural network for training the model, the neural network used includes a multi-layer feedforward network, and the transfer function used is between the input layer and the hidden layer. Choose a logarithmic function or a tangent function between them, and choose a linear function between the hidden layer and the output layer, where:

The logarithmic function is: ;

The tangent function is: ;

The linear function is: .

In a possible implementation of the first aspect, in step S4, direct integration is used to calculate the long gauge strain time history area by calculating the area enclosed by the curve and the abscissa; the peak size is based on the long gauge strain time history curve. Features, take the average of multiple peaks; the peak spacing takes the maximum value among adjacent peaks as the reference point to calculate the peak spacing.

In a possible implementation of the first aspect, the railway bridge includes a concrete bridge or a steel bridge.

A device for dynamic identification and evaluation of railway bridge parameters according to the second embodiment of the present invention, wherein the device for dynamic identification and evaluation of parameters includes:

The acquisition module is used to collect the long-gauge strain response of the train load passing through the bridge to be monitored, and obtain the long-gauge strain response data;

The calculation module solves the measured long-gauge strain response data, extracts the quasi-static long-gauge strain signal, and obtains the long-gauge strain time history curve, thereby further obtaining the long-gauge strain influence line of the railway bridge to be identified; Based on the long-gauge strain time history curve, the long-gauge strain time-history area, peak size and spacing of the long-gauge sensor are calculated to form the input layer data of the railway bridge operating parameters to be identified;

Build a module to establish a train-track-bridge coupled vibration model, conduct numerical simulations with a sufficient number of trains with different operating parameters, extract the long-gauge strain time history curve of the bridge sensor position under each train load, and obtain the enclosed long-gauge strain time Coverage area, vehicle speed, axle load and wheel-rail contact force curve;

The training module uses the simulated long gauge strain time history area, peak size and spacing as the input layer, and the vehicle speed, axle load, wheel-rail contact force curve as the output layer to establish a neural network model and train it, and obtain the trained neural network network model;

The identification module uses the long gauge strain time history area, peak size and spacing of the railway bridge to be identified as the input layer and substitutes it into the trained neural network model to identify vehicle speed, axle load and wheel-rail contact force curves;

The analysis output module analyzes the long-gauge strain influence lines of the railway bridge to be identified, and substitutes them into the trained neural network model to identify the railway operating parameters and output the railway bridge operating parameters.

A device for dynamic identification and evaluation of railway bridge parameters according to a third embodiment of the present invention, wherein the device for dynamic identification and evaluation of railway bridge parameters includes: a memory, a processor, and a device stored on the memory and capable of running on the processor. A dynamic identification and evaluation program for railway bridge parameters. When the dynamic identification and evaluation program for railway bridge parameters is executed by the processor, the steps of the above-mentioned dynamic identification and evaluation method for railway bridge parameters are implemented.

The storage medium according to the fourth embodiment of the present invention, wherein a railway bridge parameter dynamic identification and evaluation program is stored on the storage medium,

When the railway bridge parameter dynamic identification and evaluation program is executed by the processor, the steps of the railway bridge parameter dynamic identification and evaluation method as described above are implemented.

Additional aspects and advantages of the invention will be set forth in part in the description which follows, and in part will be obvious from the description, or may be learned by practice of the invention.

Description of the drawings

In order to more clearly illustrate the technical solutions of the embodiments of the present invention, the drawings required to be used in the embodiments will be briefly introduced below. Obviously, the drawings in the following description are only some embodiments of the present invention and are not relevant to the present invention. For those of ordinary skill in the art, other drawings can also be obtained based on these drawings without exerting creative efforts.

Figure 1 is a flow chart of a method for dynamic identification and evaluation of railway bridge parameters according to an embodiment of the present invention;

Figure 2 is a schematic diagram of the arrangement position of long gauge sensors of simply supported beams under train load in the dynamic identification and evaluation method of railway bridge parameters according to an embodiment of the present invention;

Figure 3 is a schematic diagram of the train-track-bridge finite element model in the dynamic identification and evaluation method of railway bridge parameters according to an embodiment of the present invention;

Figure 4 is a partial enlarged schematic diagram of the train-track-bridge finite element in Figure 3;

Figure 5 is a schematic diagram of the coupling of rails, fasteners, and track plate nodes;

Figure 6 is a train wheelbase diagram in the dynamic identification and evaluation method of railway bridge parameters according to an embodiment of the present invention;

Figure 7 is a macro strain time history curve diagram of a long gauge sensor in the dynamic identification and evaluation method of railway bridge parameters according to an embodiment of the present invention;

Figure 8 is a wheel-rail contact force curve of a certain wheel pair of the train obtained in the dynamic identification and evaluation method of railway bridge parameters according to the embodiment of the present invention;

Figure 9 is a comparison diagram between the wheel-rail contact force curve identified in the dynamic identification and evaluation method of railway bridge parameters according to the embodiment of the present invention and the real wheel-rail force curve;

Figure 10 is the relative error between the wheel-rail contact force curve identified in the dynamic identification and evaluation method of railway bridge parameters according to the embodiment of the present invention and the real wheel-rail force curve.

Detailed ways

The embodiments of the present invention are described in detail below. The embodiments described with reference to the drawings are exemplary. It should be understood that the specific embodiments described here are only used to explain the present application and are not used to limit the present application.

It should be noted that when an element is referred to as being "fixed" to another element, it can be directly on the other element or intervening elements may also be present. When an element is said to be "connected" to another element, it can be directly connected to the other element or there may also be intervening elements present.

Unless otherwise defined, all technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the technical field to which the invention belongs. The terminology used herein in the description of the invention is for the purpose of describing specific embodiments only and is not intended to limit the invention. As used herein, the term "and/or" includes any and all combinations of one or more of the associated listed items.

The terms "first", "second", "third", etc. in the description, claims, and drawings of this application are used to distinguish different objects and are not used to describe a specific sequence. Furthermore, the terms "including" and "having" and any variations thereof are intended to cover non-exclusive inclusion. For example, a series of steps or units are included, or optionally, steps or units that are not listed, or optionally other steps or units that are inherent to these processes, methods, products or devices.

The drawings show only part but not all of the content relevant to the present application. Before discussing example embodiments in more detail, it should be mentioned that some example embodiments are described as processes or methods depicted as flowcharts. Although flowcharts describe various operations (or steps) as a sequential process, many of the operations may be performed in parallel, concurrently, or simultaneously. Additionally, the order of operations can be rearranged. The process may be terminated when its operation is completed, but may also have additional steps not included in the figures. The processing may correspond to a method, function, procedure, subroutine, subroutine, or the like.

The terms "component", "module", "system", "unit", etc. used in this specification are used to refer to computer-related entities, hardware, firmware, combinations of hardware and software, software or software in execution. For example, a unit may be, but is not limited to, a process running on a processor, a processor, an object, an executable file, a thread of execution, a program and/or distributed between two or more computers. Additionally, these units can execute from various computer-readable media having various data structures stored thereon. A unit may, for example, respond to a signal with one or more data packets (eg data from a second unit interacting with another unit, a local system, a distributed system and/or a network. For example, the Internet interacting with other systems via signals) Communicate through local and/or remote processes.

Example 1

Referring to Figure 1, this embodiment provides a method for dynamic identification and evaluation of railway bridge parameters, which is characterized by including:

Step S1, build a dynamic identification and evaluation scenario for railway bridge parameters, and arrange several long-gauge strain sensors on the bridge to be monitored;

Step S2, collect the long-gauge strain response when the train load passes through the bridge to be monitored, and obtain the long-gauge strain response data;

Step S3: Solve the measured long-gauge strain response data, extract the quasi-static long-gauge strain signal, and obtain the long-gauge strain time history curve, thereby further obtaining the long-gauge strain influence line of the railway bridge to be identified;

Step S4: Solve the long-gauge strain time-history area, peak size and spacing of all long-gauge sensors based on the long-gauge strain time-history curve to form the input layer data of the railway bridge operating parameters to be identified;

Step S5, establish a train-track-bridge coupling vibration model, conduct numerical simulations with sufficient trains with different operating parameters, extract the long-gauge strain time history curve of the bridge sensor position under each train load, and obtain the enclosed long-gauge strain time Train operating parameters such as track area, vehicle speed, axle load and wheel-rail contact force curve solve the problem of being unable to obtain train load parameters to form a complete sample in actual situations;

Step S6: Use the simulated long gauge strain time history area, peak size and spacing as the input layer, and the vehicle speed, axle load, wheel-rail contact force curve as the output layer to establish a neural network model and train it to obtain the trained neural network. network model;

Step S7: Substitute the long-gauge strain time history area, peak size and spacing of the railway bridge to be identified as the input layer into the trained neural network model to identify vehicle speed, axle load, and wheel-rail contact force curves;

Step S8: Analyze the long-gauge strain influence lines of the railway bridge to be identified, substitute them into the trained neural network model to identify railway operating parameters, and complete the dynamic identification and evaluation process of railway bridge parameters.

According to the dynamic identification and evaluation method of railway bridge parameters according to the embodiment of the present invention, the long gauge strain response of a high-speed train passing through a railway simply supported girder bridge is obtained by using a long gauge strain sensor, and the selected sample data is solved to obtain the long gauge strain time history curve, and then solve for the area enclosed by the long gauge strain time history curve of all long gauge sensors. The peak size and spacing are used as parameters to be identified. By establishing a train-track-bridge coupled vibration model, a sufficient amount of different railways The operating parameters are numerically simulated to obtain a rich sample library and used to train the neural network, and then substituted into the parameters to be identified to obtain the railway operating parameters. The invention can quickly identify the train operating parameters on the upper part of the railway bridge without affecting the operating traffic, greatly improves the monitoring efficiency, and provides guarantee for the operational safety of the bridge.

Optionally, railway bridges include concrete or steel bridges, prestressed bridges and non-prestressed bridges.

It should be noted that the long gauge sensor is a long gauge fiber grating sensor, or

Compared with traditional point strain gauges, long-gauge resistance strain sensors can complete the full-span layout of the bridge and realize bridge strain influence line monitoring.

It should be noted that in step S2, the long-gauge strain response after the train load passes through the bridge under test needs to be collected; the train load can be any train load, including high-speed trains, ordinary railway passenger cars, railway freight cars, etc., and no skylight time is required Operation.

It should be noted that in step S6, the long gauge strain time history area, peak size and spacing are used as the input layer, and the train speed, axle load, and wheel-rail contact force curve are used as the output layer to establish a neural network model and train it. Specifically, include:

Record the strain time history area of each gauge sensor at the bottom of the bridge of the lth train as , the peak size is recorded as/> , the peak distance is/> , the vehicle speed is/> , the axle weight is/> , the wheel-rail contact force curve is/> , of which/> ,/> is the number of trains;

Establish a training model through a single hidden layer BP neural network, determine the approximate range of the number of hidden layer neurons through empirical formulas, and then design a BP network with a variable number of hidden layer neurons through program code, and compare the training of each network structure Error, the most appropriate number of neurons can be obtained, where the empirical formula is as follows:

,

in, is the number of neurons in the hidden layer,/> is the number of neurons in the input layer,/> is the number of neurons in the output layer, for/> constant between;

The number of neurons in the input layer and output layer of the neural network is directly determined by the input parameters and output parameters of the network, where the input parameters of the input layer are ,/> is the input data, and the output parameters of the output layer are/> ,/> for output data. .

It should be noted that in step S6, the input layer parameters are all or part of the data obtained by long gauge sensor analysis, and the output layer is all or part of the data of railway operating parameters. All parameters can be trained at one time or multiple times. Train network models for individual parameters separately.

It should be noted that in step S6, when building a neural network for training the model, the neural network used is not limited to BP neural network, including other code toolboxes for deep learning. The transfer function used is between the input layer and Choose a logarithmic function or a tangent function between the hidden layers, and choose a linear function between the hidden layer and the output layer, where:

The logarithmic function is: ;

The tangent function is: ;

The linear function is: .

It should be noted that the railway bridge includes a concrete bridge or a steel bridge.

This embodiment is mainly based on the fact that the long gauge strain time history curve contains obvious train operating parameter information, combined with the powerful ability of the BP neural network to solve nonlinear mapping problems. On this basis, a method based on the long gauge is developed. The method of identifying railway operating parameters from distance sensors is explained below with specific examples.

Figure 2 shows a simply supported beam high-speed railway bridge, which is a double-track bridge. The total length of the beam is L and the stiffness of the whole beam is EI . It is assumed that the middle position of the bottom of the beam is covered with n long-gauge sensors with a gauge length of l _e . Record the strain time history area of each gauge sensor at the bottom of the bridge of the lth train as , the peak size is recorded as/> , the peak distance is/> , the vehicle speed is/> , the axle weight is/> , the wheel-rail contact force curve is/> , of which/> ,/> is the number of trains.

The long gauge strain time history area uses direct integration to calculate the area enclosed by the curve and the abscissa; the peak size is based on the characteristics of the long gauge strain time history curve, and the average value of multiple peaks is taken; the peak distance is the largest among adjacent peaks The value is used as the base point to calculate the peak distance.

The principle of BP neural network adopts the gradient fastest descent method, that is, through a large number of sample training, the connection weights of the hidden layer in the network are constantly adjusted, and the actual value is constantly brought close to the expected value to minimize the overall error. The entire algorithm consists of two basic propagation processes: forward propagation and backward propagation. In forward propagation, the input signal first enters the network structure from the input layer, is processed step by step through the hidden layer and is transmitted to the output layer, and the expected value and output value error are compared. If the error requirements are not met, back propagation is entered, that is, the error signal is returned along the original connection channel, and the error signal is minimized by modifying the weights of the neurons in each layer.

Use the sample set in the database to train the BP neural network, and the test set to test the trained network structure and obtain the error of this network. Then select the network structure with the smallest error as the final network structure, and finally use the verification set to verify the network structure. accuracy.

The refined train-track-bridge coupling vibration model in the present invention is the basis for calculating the sample data set. Only the bridge vibration response measured on site can be obtained based on simulating high-speed train crossing the bridge as sample data.

As shown in Figures 3 and 4, A in Figure 4 is the rail, B is the track plate, C is the cement asphalt mortar (CA mortar layer), D is the concrete base plate, E is the main beam, and F is Coupling diagram of rail and track plate nodes. This invention establishes a 32-meter-span track-bridge model. CRTSⅡ plate-type ballastless track is laid on the bridge. The rails are 60kg/m hot-rolled rails commonly used in high-speed railways. The BEAM188 beam element based on Timoshenko beam theory is used to simulate the rails and main beams. The commonly used fasteners for high-speed railway CRTS II plate ballastless track are WJ-8 type, the fastener spacing is 0.6 meters, and the COMBIN14 linear elastic spring-damper unit is used for simulation. The track slab, CA mortar layer, concrete base plate, bridge beam, etc. are simulated using the BEAM188 unit, and the model is established by giving the unit a combined section, as shown in Figure 5. This model simulates the connection between the rail and the track plate into a narrow rigid strip bridge model. The overall coordinate system of the bridge model is xyz, and the corresponding local coordinate system of the rail unit is defined as uvw. Since the rail unit has an x axis perpendicular to the yoz axis of the beam plane The freedom of axis rotation and the freedom of rotation around the z-axis perpendicular to the xoy axis of the beam plane, while the fastener spring-damper COMBIN14 unit has no rotational freedom and cannot form rotational constraints on the rail beam unit, so it does not conform to the real rails and fasteners. , The connection relationship between track plates is shown in Figure 5. To this end, the constraint equations between the rail beam unit nodes and the fastener spring-damping are established to constrain the excess degrees of freedom of the rail beam unit.

Rotation angular displacement of rail beam element node i around the X direction It has no effect on the x-direction displacement at the node of the j-1, j, and j+1 fastener spring unit. Rotation angular displacement of beam element node i around the X direction/> It affects the y-direction displacement at the node of the j-1 fastener spring element, and its value is:

in: is the z-coordinate of node i,/> is the z-coordinate of node i-1,/> is the rotational angular displacement of node i around the X direction, h is the distance between the rail beam unit node and the fastener spring unit node, and l is the fastener distance.

Similarly, the rotational angular displacement of the rail beam unit i around the X direction It affects the z-direction displacement at the j+1 fastener spring element node, and its value is:

in: is the z-coordinate of node i,/> is the z-coordinate of node i+1,/> is the rotational angular displacement of node i around the X direction, h is the distance between the rail beam unit node and the fastener spring unit node, and l is the fastener distance.

In the same way, the influence of the rotational angular displacement rotz of the rail beam unit i around the Z direction on the x- and y-direction displacements at the j-1, j, and j+1 fastener spring unit nodes is approximately 0. The rotational angular displacement rotz of beam unit i around the Z direction has no effect on the z-direction displacement at the j-1, j, j+1 fastener spring unit nodes. In this way, the y-direction displacement constraint equation is:

in: is the vertical displacement of the fastener spring element node,/> is the vertical displacement of the element node of the rail beam unit,/> is the rotational angular displacement of node i-1 of the rail beam unit unit around the X direction,/> is the rotational angular displacement of node i+1 of the rail beam unit unit around the X direction, and l is the fastener distance.

The x- and z-direction displacement of the fastener spring unit node is determined by the displacement on the bottom surface of the beam unit it is connected to. Its displacement constraint equation is approximately:

in: is the lateral displacement of the fastener spring element node,/> is the lateral displacement of the element node of the rail beam unit,/> is the longitudinal displacement of the fastener spring element node,/> is the longitudinal displacement of the element node of the rail beam element.

The train model is based on the CRH3 EMU as a prototype, as shown in Figure 6. In the modeling of this invention, it is simplified into a 10-degree-of-freedom vehicle model. The train body, bogies and wheels are simulated using the MASS21 unit, taking into account the characteristics of each component. Degree of freedom, set the car body and bogie mass units to have rotational inertia, and the wheel mass unit to have no rotational inertia. The vertical connection between each component is simulated using the COMBIN14 linear elastic spring-damper unit, and the longitudinal connection is simulated using the BEAM4 unit.

After the model is established, enter the solution stage, and set the mass unit parameters of the train, the time step during solution, and the train movement distance according to the train axle weight and speed that need to be simulated. After solving, enter the post-processing stage, extract the wheel-rail contact force, axle load, speed, and macro-strain time history curve under train load, and establish a data set. The extracted macrostrain time history curve is shown in Figure 7, and the extracted wheel-rail contact force curve is shown in Figure 8.

The comparison between the identification value and the real value obtained after parameterizing the input layer in the BP network is shown in Figure 9. It can be seen from Figure 10 that the relative error of the trained neural network structure in identifying the wheel-rail force curve is less than 5%, and the effect is significant.

Example 2

This embodiment provides a device for dynamic identification and evaluation of railway bridge parameters, wherein the device for dynamic identification and evaluation of parameters includes:

The acquisition module is used to collect the long-gauge strain response of the train load passing through the bridge to be monitored, and obtain the long-gauge strain response data;

The calculation module solves the measured long-gauge strain response data, extracts the quasi-static long-gauge strain signal, and obtains the long-gauge strain time history curve, thereby further obtaining the long-gauge strain influence line of the railway bridge to be identified; Based on the long-gauge strain time history curve, the long-gauge strain time-history area, peak size and spacing of all long-gauge sensors are solved to form the input layer data of the railway bridge operating parameters to be identified;

Build a module to establish a train-track-bridge coupled vibration model, conduct numerical simulations with a sufficient number of trains with different operating parameters, extract the long-gauge strain time history curve of the bridge sensor position under each train load, and obtain the enclosed long-gauge strain time Train operating parameters such as track area, vehicle speed, axle load and wheel-rail contact force curve;

The training module uses the simulated long gauge strain time history area, peak size and spacing as the input layer, and the vehicle speed, axle load, wheel-rail contact force curve as the output layer to establish a neural network model and train it, and obtain the trained neural network network model;

The identification module uses the long gauge strain time history area, peak size and spacing of the railway bridge to be identified as input layers into the trained neural network model to identify the vehicle speed, axle load, and wheel-rail contact force curves;

The analysis output module analyzes the long-gauge strain influence lines of the railway bridge to be identified, and substitutes them into the trained neural network model to identify the railway operating parameters and output the railway bridge operating parameters.

Example 3

This embodiment provides a device for dynamic identification and evaluation of railway bridge parameters, wherein the device for dynamic identification and evaluation of railway bridge parameters includes: a memory, a processor, and a railway bridge stored in the memory and operable on the processor. A parameter dynamic identification and evaluation program is provided. When the railway bridge parameter dynamic identification and evaluation program is executed by the processor, the steps of the railway bridge parameter dynamic identification and evaluation method as described above are implemented.

Example 4

This embodiment provides a storage medium, wherein a railway bridge parameter dynamic identification and evaluation program is stored on the storage medium,

When the railway bridge parameter dynamic identification and evaluation program is executed by the processor, the steps of the railway bridge parameter dynamic identification and evaluation method as described above are implemented.

In the description of the present invention, it should be understood that the terms "center", "longitudinal", "transverse", "length", "width", "thickness", "upper", "lower", "front", " "Back", "Left", "Right", "Vertical", "Horizontal", "Top", "Bottom", "Inside", "Outside", "Clockwise", "Counterclockwise", "Axis" The orientations or positional relationships indicated by "radial direction", "circumferential direction", etc. are based on the orientations or positional relationships shown in the drawings. They are only for the convenience of describing the present invention and simplifying the description, and do not indicate or imply the device or device referred to. Elements must have a specific orientation, be constructed and operate in a specific orientation and therefore are not to be construed as limitations on the invention.

In the description of this specification, reference to the terms "one embodiment," "some embodiments," "illustrative embodiments," "examples," "specific examples," or "some examples" or the like is intended to be incorporated into the description of the implementation. An example or example describes a specific feature, structure, material, or characteristic that is included in at least one embodiment or example of the invention. In this specification, schematic representations of the above terms do not necessarily refer to the same embodiment or example.

Obviously, the described embodiments are only some of the embodiments of the present application, but not all of the embodiments. Reference herein to "an embodiment" means that a particular feature, structure, or characteristic described in connection with the embodiment may be included in at least one embodiment of the present embodiment application. The appearances of this phrase in various places in the specification are not necessarily all referring to the same embodiment, nor are separate or alternative embodiments mutually exclusive of other embodiments. Those skilled in the art will understand, both explicitly and implicitly, that the embodiments described herein may be combined with other embodiments. Based on the embodiments in this application, all other embodiments obtained by those of ordinary skill in the art without creative efforts fall within the scope of protection of this application.

Although the embodiments of the present invention have been shown and described, those of ordinary skill in the art will appreciate that various changes, modifications, substitutions and variations can be made to these embodiments without departing from the principles and purposes of the invention. The scope of the invention is defined by the claims and their equivalents.

Claims (10)

1. A method for dynamic identification and evaluation of railway bridge parameters, which is characterized by including: Step S1, build a dynamic identification and evaluation scenario for railway bridge parameters, and arrange several long-gauge strain sensors on the bridge to be monitored; Step S2, collect the long-gauge strain response when the train load passes through the bridge to be monitored, and obtain the long-gauge strain response data; Step S3: Solve the measured long-gauge strain response data, extract the quasi-static long-gauge strain signal, and obtain the long-gauge strain time history curve, thereby further obtaining the long-gauge strain influence line of the railway bridge to be identified; Step S4: Solve the long-gauge strain time-history area, peak size and spacing of the long-gauge sensor based on the long-gauge strain time-history curve to form the input layer data of the railway bridge operating parameters to be identified; Step S5, establish a train-track-bridge coupling vibration model, conduct numerical simulations with sufficient trains with different operating parameters, extract the long-gauge strain time history curve of the bridge sensor position under each train load, and obtain the enclosed long-gauge strain time Coverage area, vehicle speed, axle load and wheel-rail contact force curve; Step S6: Use the simulated long gauge strain time history area, peak size and spacing as the input layer, and the vehicle speed, axle load, wheel-rail contact force curve as the output layer to establish a neural network model and train it to obtain the trained neural network. network model; Step S7: Substitute the long-gauge strain time history area, peak size and spacing of the railway bridge to be identified as input layers into the trained neural network model to identify vehicle speed, axle load, and wheel-rail contact force curves; Step S8: Analyze the long-gauge strain influence lines of the railway bridge to be identified, substitute them into the trained neural network model to identify railway operating parameters, and complete the dynamic identification and evaluation process of railway bridge parameters. 2. The dynamic identification and evaluation method of railway bridge parameters according to claim 1, characterized in that the long gauge sensor is a long gauge fiber grating sensor or a long gauge resistance strain sensor. 3. The dynamic identification and evaluation method of railway bridge parameters according to claim 1, characterized in that, In step S6, a neural network model is established and trained using the long gauge strain time history area, peak size and spacing as the input layer, and the train speed, axle load, wheel-rail contact force curve as the output layer, including: Record the strain time history area of each gauge sensor at the bottom of the bridge of the lth train as , the peak size is recorded as/> , the peak distance is/> , the vehicle speed is/> , the axle weight is/> , the wheel-rail contact force curve is/> , of which/> ,/> is the number of trains; Establish a training model through a single hidden layer neural network, determine the range of the number of hidden layer neurons through empirical formulas, then design a BP network with a variable number of hidden layer neurons, compare the training errors of each network structure, and obtain the optimal The appropriate number of neurons, the empirical formula is as follows: , in, is the number of neurons in the hidden layer,/> is the number of neurons in the input layer,/> is the number of neurons in the output layer,/> for constant between; The number of neurons in the input layer and output layer of the neural network is directly determined by the input parameters and output parameters of the network, where the input parameters of the input layer are ,/> is the input data, and the output parameters of the output layer are/> ,/> for output data. 4. The dynamic identification and evaluation method of railway bridge parameters according to claim 1, characterized in that, in step S6, the input layer parameters are all or part of the data obtained by long gauge sensor analysis, and the output layer is all the railway operating parameters. Data or part of the data, train all parameters at once or train a single parameter network model multiple times. 5. The dynamic identification and evaluation method of railway bridge parameters according to claim 1, characterized in that, In step S6, when constructing a neural network for training the model, the neural network used includes a multi-layer feedforward network, and the transfer function used selects a logarithmic function or a tangent function between the input layer and the hidden layer. Choose a linear function between the layer and the output layer. 6. The dynamic identification and evaluation method of railway bridge parameters according to claim 1, characterized in that, in step S4, the area surrounded by the curve and the abscissa is calculated by direct integration to solve the long gauge strain time history area; the peak size According to the characteristics of the long gauge strain time history curve, the average value of multiple peaks is taken; the peak spacing is calculated by taking the maximum value among adjacent peaks as the reference point. 7. The dynamic identification and evaluation method of railway bridge parameters according to claim 1, characterized in that the railway bridge includes a concrete bridge or a steel bridge. 8. A device for dynamic identification and evaluation of railway bridge parameters, characterized in that the device for dynamic identification and evaluation of parameters includes: The acquisition module is used to collect the long-gauge strain response of the train load passing through the bridge to be monitored, and obtain the long-gauge strain response data; The calculation module solves the measured long-gauge strain response data, extracts the quasi-static long-gauge strain signal, and obtains the long-gauge strain time history curve, thereby further obtaining the long-gauge strain influence line of the railway bridge to be identified; Based on the long-gauge strain time history curve, the long-gauge strain time-history area, peak size and spacing of the long-gauge sensor are calculated to form the input layer data of the railway bridge operating parameters to be identified; Build a module to establish a train-track-bridge coupled vibration model, conduct numerical simulations with a sufficient number of trains with different operating parameters, extract the long-gauge strain time history curve of the bridge sensor position under each train load, and obtain the enclosed long-gauge strain time Coverage area, vehicle speed, axle load and wheel-rail contact force curve; The training module uses the simulated long gauge strain time history area, peak size and spacing as the input layer, and the vehicle speed, axle load, wheel-rail contact force curve as the output layer to establish a neural network model and train it, and obtain the trained neural network network model; The identification module uses the long gauge strain time history area, peak size and spacing of the railway bridge to be identified as the input layer and substitutes it into the trained neural network model to identify vehicle speed, axle load and wheel-rail contact force curves; The analysis output module analyzes the long-gauge strain influence lines of the railway bridge to be identified, and substitutes them into the trained neural network model to identify the railway operating parameters and output the railway bridge operating parameters. 9. A device for dynamic identification and evaluation of railway bridge parameters, characterized in that the device for dynamic identification and evaluation of railway bridge parameters includes: a memory, a processor, and a railway bridge stored in the memory and operable on the processor. The parameter dynamic identification and evaluation program of the railway bridge, when executed by the processor, implements the steps of the railway bridge parameter dynamic identification and evaluation method as described in any one of claims 1 to 7. 10. A storage medium, characterized in that a railway bridge parameter dynamic identification and evaluation program is stored on the storage medium, When the railway bridge parameter dynamic identification and evaluation program is executed by the processor, the steps of the railway bridge parameter dynamic identification and evaluation method according to any one of claims 1 to 7 are implemented.