An aeroelastic model simulation test system and method for truss bridge main beam

Abstract

The present application discloses an aeroelastic model simulation test system and method for the main beam of a truss beam bridge. An aeroelastic model simulation test system for the main beam of a truss beam bridge includes: a wind tunnel environment module, a wind speed control module, an aeroelastic model, and an evaluation module; a control module electrically connected to each elastic adjustment member for adjusting the elastic coefficient of the elastic adjustment member in real time; a test generation module with a built-in truss beam bridge structural performance degradation model, generating a sequence of elastic coefficient changes of the elastic adjustment member, and outputting multiple test schemes; a data execution module, linked with the test generation module and the wind speed control module, sequentially loading the elastic coefficient corresponding to each test scheme, triggering the wind tunnel test and recording the critical wind speed theoretical value. The technical solution provided by the present application can provide a refined analysis basis for the wind-induced vibration safety assessment of existing truss beam bridges, and effectively solve the technical problem of the mismatch between the traditional homogenized model and the time-varying characteristics of the actual bridge.

Description

Technical Field

[0001] The present application relates to the technical field of wind tunnel testing, and in particular to an aeroelastic model simulation test system and method for a truss bridge main beam. Background Art

[0002] Large-span bridges are prone to aerodynamic instabilities such as flutter, vortex-induced vibration, and buffeting under natural wind loads. In severe cases, these can lead to structural dynamic instability. To ensure the safety of bridge structures, aeroelastic model wind tunnel testing is necessary. This test involves building a scaled-down bridge model and simulating different wind speeds and angles of attack in a wind tunnel environment. The test then determines the critical instability wind speed for the structure, thereby evaluating the compatibility of the design with the wind environment parameters at the construction site.

[0003] Current wind tunnel tests of aeroelastic models typically use idealized assumptions: each model connection is set to a homogenized state for testing. However, in actual engineering, bridge structures exhibit dynamic characteristics during operation. First, the aging rates of different connection components vary significantly, and key locations such as welds and anchor nodes are prone to non-uniform structural damage. Second, during periodic inspections and maintenance, differences in the implementation and effectiveness of maintenance measures can lead to changes in the distribution of the structural elastic coefficients. Ultimately, these factors combine to produce a time-varying characteristic in the overall wind resistance of the bridge.

[0004] A limitation of existing testing methods is that they assume that the elastic modulus degradation of each model connection node is synchronized and that the structural tensile and shear strengths remain homogenous. This idealized assumption differs fundamentally from the evolution of the mechanical properties of actual bridges. This leads to significant deviations between the theoretical critical wind speed values obtained from experiments and the failure wind speed thresholds of actual operating bridges, directly impacting the accuracy of wind-induced vibration safety assessments. Summary of the Invention

[0005] The content of this application is used to briefly introduce concepts that will be described in detail in the detailed description section below. The content of this application is not intended to identify key features or essential features of the technical solution for which protection is sought, nor is it intended to limit the scope of the technical solution for which protection is sought.

[0006] As a first aspect of the present application, in order to solve the technical problems mentioned in the background technology section above, some embodiments of the present application provide an aeroelastic model simulation test system for a truss bridge main beam, including:

[0007] A wind tunnel environment module is used to construct a test environment simulating a natural wind field, and the wind tunnel environment module includes an adjustable angle of attack device and a turbulence generation unit;

[0008] A wind speed control module is connected to the wind tunnel environment module and is used to adjust the wind speed of the wind tunnel through a variable frequency motor and provide real-time feedback of wind speed data;

[0009] an aeroelastic model, comprising a plurality of segment modules distributed longitudinally along the main beam, and a connecting component connecting two adjacent segment modules;

[0010] An evaluation module, including a laser displacement meter and a strain sensor arranged on each segment module, for measuring the dynamic deformation data of the aeroelastic model;

[0011] in,

[0012] The connection component includes:

[0013] Two free end connecting parts are respectively fixed to the end surfaces of adjacent segment modules;

[0014] The elastic adjusting member is an elastic damper with an adjustable elastic coefficient, and both ends are rigidly connected to the free end connecting portion;

[0015] The aeroelastic model simulation test system also includes:

[0016] A control module electrically connected to each elastic adjustment member and configured to adjust the elastic coefficient of the elastic adjustment member in real time;

[0017] The test generation module has a built-in truss bridge structure performance degradation model, generates a sequence of elastic coefficient changes for elastic adjustment parts, and outputs multiple test plans;

[0018] The data execution module is linked with the test generation module and the wind speed control module to sequentially load the elastic coefficient variation sequence corresponding to each test scheme, trigger the wind tunnel test and record the critical wind speed theoretical value.

[0019] This technical solution, by constructing a structural degradation model for truss girder bridges, can quantitatively simulate the elastic coefficient degradation characteristics of connection components at different service stages. Based on this, a system of connection elastic coefficients with time-varying characteristics is generated. By establishing a multi-dimensional structural degradation database, the differentiated aging states of various connection nodes throughout the bridge's lifecycle can be accurately reproduced in wind tunnel testing, enabling multi-condition coupled wind resistance testing. Compared to traditional methods, this solution offers the following technical advantages: by introducing a time-varying structural mechanical model, it can systematically reflect the nonlinear elastic coefficient degradation patterns of key connection locations, such as weld fatigue damage and bolt preload relaxation. During the test, the elastic constraints of each connection node can be dynamically adjusted, establishing a quantitative mapping between aging degree and wind resistance performance, significantly improving the completeness of the test data and the coverage of the required operating conditions. This technology can provide a refined analytical basis for the wind-induced vibration safety assessment of existing truss girder bridges, effectively resolving the technical challenge of the mismatch between traditional homogenized models and the time-varying characteristics of actual bridges.

[0020] When a truss girder bridge experiences structural deformation under strong wind loads, the mechanical degradation of its connection structure exhibits a continuous, progressive damage accumulation process, rather than the discrete "normal-yield-fracture" transitions assumed in traditional models. Existing modeling methods, which simplify each connection element into discrete performance states, have fundamental flaws: first, they fail to account for nonlinear elastic coefficient degradation characteristics such as bolt loosening and contact surface microslip; second, they ignore the coupling effects between different damage modes and the time-varying nature of damage rates; and third, they sever the dynamic correlation between local component damage and the overall structural response. This discretization of the continuous damage process results in a loss of process simulation accuracy in structural performance degradation models, ultimately causing the dynamic response data obtained from wind tunnel tests to deviate from the true evolutionary law, reducing the credibility of service safety assessment results.

[0021] To this end, this application provides the following technical solutions:

[0022] The truss beam bridge structural performance degradation model includes a plurality of hidden states corresponding to aging information and probability transition information between the hidden states.

[0023] The technical solution provided in this application describes the aging process of a truss bridge connection using several continuous hidden states, and then uses probabilistic transition information to describe the changes in the connection. This allows for a dynamic display of the accumulation of metal damage in the connection, allowing simulation experiments to accurately reflect the evolution of actual dynamic response data.

[0024] The composite loads borne by the connection components include three dynamic components: aerodynamic tension directly induced by the wind speed time history, a continuous axial force generated by the structure's own weight, and inertial alternating stresses caused by bridge vibration. The time-domain coupling effect of this multi-source excitation results in: first, a nonlinear correlation between the bolt preload relaxation rate and wind speed fluctuation characteristics; second, a positive feedback mechanism between the contact surface microslip displacement and the structural swing amplitude; and third, a dual control of the corrosion damage process by stress amplitude and the number of cycles. Static equivalent load simulation methods cannot replicate the time-varying characteristics of the tension-shear-bending composite stress field in actual wind-induced vibrations; and conventional dynamic loading devices have difficulty accurately decoupling the phase relationship between inertial and aerodynamic forces. This directly leads to significant differences between the component stress spectra obtained in the test and the actual service conditions of the bridge, causing the simulation results of the damage accumulation rate at the connection node to deviate from the actual evolution law, ultimately affecting the reliability of the structural wind safety assessment.

[0025] Furthermore, the structural performance degradation model of the truss beam bridge is constructed as follows:

[0026] S1: Set a number of consecutive aging states q, the number of aging states is Q, and q represents the index of the aging state;

[0027] S2: Set the state variable R t , R t represents the state variable at time t, R t ={ra t rb t rc t}, where ra t It represents the accumulated displacement in time t;

[0028] , Indicates the displacement length X of the connected component from time 0 to time t h Integration over time;

[0029] rb t Indicates the number of times the connected component is subjected to force within time t;

[0030] rc t Indicates the time during which the force on the connected component exceeds the force threshold within time t;

[0031] S3: Set the initial state probability distribution P of the aging state, P={p1, ...p q …p Q}, q∈[1,Q], where p q is the probability of the qth aging state occurring;

[0032] S4: Establish the probability transfer matrix W of the aging state, W={w qs},w qs represents the probability of transitioning from aging state q to aging state s, s∈[1,N]; s>q;

[0033] S5: Construct the state variable probability matrix E, E={b q (R t )} represents the probability density function of the connected component under each aging state;

[0034] ;

[0035] Among them, M represents the number of mixed Gaussian components contained in each aging state, g qm represents the mth Gaussian distribution weight of the aging state q; represents the mth Gaussian distribution function of the aging state q, u qm and L qm are the mean vector and covariance vector respectively;

[0036] ;

[0037] in, represents the matrix conversion symbol, P represents the initial state probability distribution, and exp represents the power operation of the base e of the natural logarithm;

[0038] S6: The probability transfer matrix W and the state variable probability matrix E are collaboratively modeled to generate a truss bridge structural performance degradation model.

[0039] This technical solution achieves a computable representation of the structural damage evolution mechanism by constructing a multi-level degradation model based on stochastic process theory. Specifically, data assimilation technology is used to establish a coupling relationship between the state transition probability matrix W and the observation state probability matrix E, forming a dual-matrix driven damage evolution analysis system:

[0040] 1) The state transition probability matrix W uses the Markov chain modeling method to quantitatively characterize the microscopic defects of the connection components under different stress levels;

[0041] 2) The observed state probability matrix E is based on a dynamic mapping relationship between the component's macroscopic mechanical response (including measurable parameters such as strain energy density and residual displacement) and the implicit damage state. This method can transform complex damage paths at connected nodes (such as the competing mechanisms of stress corrosion crack propagation and fatigue damage) into computable stochastic processes, significantly improving the simulation accuracy of structural responses under multi-physics coupling in wind tunnel testing.

[0042] Furthermore, S6 includes the following steps:

[0043] S61: Construct the initial truss bridge structural performance degradation model λ;

[0044] λ = (P, W, E);

[0045] Where P = {1, 0, 0, 0},

[0046] W=[w qs ] Q×Q , Q=4, the constraints of W are:

[0047] (1) w qs > 0, s > q, that is, the aging state can only be transferred to the aging state with a larger number;

[0048] (2) w qs =0, s≤q, that is, the aging state cannot be transferred to an aging state with a smaller number;

[0049] (3) , that is, the probability of the aging state q being transferred out is 1;

[0050] E={b q (R t )};

[0051] S62: Train the initial truss bridge structural performance degradation model λ to find the optimal model parameters so that the likelihood probability of the observation sequence is maximized.

[0052] Furthermore, S62 includes the following steps:

[0053] S621: Initialize training parameters and the maximum number of iterations. The training parameters include the initial probability transfer matrix W, the initial state variable probability matrix E, and the change threshold.

[0054] S622: Update the probability transfer matrix W:

[0055] S623: Update the state variable probability matrix E;

[0056] S624: Repeat S622 and S623 to repeat the iteration until the likelihood probability of the observation sequence is less than the change threshold or the maximum number of iterations is reached.

[0057] Furthermore, the pneumatic model connection components are divided into four aging states according to the ratio of the effective connection length under conventional test conditions.

[0058] In the technical solution provided in the present application, the aging state is divided according to the elastic coefficient of the connection component, which can accurately divide each state and avoid the problem of poor distinction between adjacent states, which makes the model too redundant.

[0059] Furthermore, the stress threshold is the tension of the connection component within the maximum elastic limit.

[0060] When subjected to a tensile force within the stress threshold, the connection component can recover. When subjected to a tensile force outside the stress threshold, the recovery ability of the connection component gradually decreases. This can gradually describe the aging process of the connection component.

[0061] When a truss beam bridge suffers structural damage in strong winds, it will experience a significant short-board effect. That is, as soon as a connection component in a certain area breaks, the stability of the truss beam bridge will collapse rapidly. Therefore, when conducting a wind tunnel test on a truss beam bridge, the actual test is to test the minimum wind environment that the truss beam bridge can withstand during its normal service life. However, current wind tunnel tests are all simulation tests that consider the structural performance of the connection components to be consistent and the structure to remain unchanged. In actual work, the accuracy is not high. To this end, this application provides the following technical solutions:

[0062] Furthermore, the test generation module includes:

[0063] Model building unit with built-in truss bridge structure performance degradation model;

[0064] The solution generation unit sets an aging state for each connected component based on the particle swarm algorithm to generate a simulation solution;

[0065] The simulation unit presets the elastic coefficient under each aging state; sets the aging state of each connection component in the simulation scheme to the aging state to generate a test scheme;

[0066] The probability fusion unit incorporates the transition probability of each aging state into the test plan based on the truss bridge structure performance degradation model.

[0067] This technical solution achieves a refined simulation of the degradation process of bridge connection nodes by constructing a parameterized dynamic elastic coefficient model and a probabilistic fusion mechanism. The specific implementation process includes two core modules:

[0068] 1) Based on the damage state-elastic coefficient mapping criterion, the aging state of each connection component in the simulation model is quantified as a time-varying elastic coefficient, where the elastic coefficient is dynamically adjusted to accurately characterize the nonlinear response process of the connection interface from micro-slip to macro-displacement;

[0069] 2) Establishing a probabilistic fusion analysis unit, by integrating state transition probabilities, constructs a multi-scale damage evolution prediction model that dynamically corrects the elastic coefficient degradation path of each node under different environmental excitations. This method effectively overcomes the errors caused by the idealized treatment of the elastic coefficients of connection nodes in conventional wind tunnel testing, enabling the test model to accurately reproduce the slip displacement time history curve and hysteretic energy dissipation characteristics of the connection under strong winds, significantly improving the engineering credibility of determining the critical point of structural dynamic instability.

[0070] The main beam of a truss bridge has multiple connection components. The degree of aging of each connection component during use is individually unpredictable. If the aging state of each connection component is set to the lowest, it will obviously not conform to the actual operation and maintenance conditions of the truss bridge main beam. If all possible aging state combinations are traversed, it will lead to an excessive number of wind tunnel tests, reducing test efficiency. To this end, this application provides the following technical solutions:

[0071] Furthermore, an aging score is set for each aging state, and a score threshold is set;

[0072] The scheme generation unit continuously generates simulation schemes on the premise that the sum of the aging scores in the simulation schemes is higher than the score threshold.

[0073] Setting the score threshold can simulate the actual aging of the main beam of a girder bridge during normal service, thereby obtaining wind tunnel test results that best match actual operating conditions while reducing the number of simulations.

[0074] As a second aspect of the present application, a method for simulating the aeroelastic model of a truss beam bridge main beam is provided, wherein the aeroelastic model simulation test system for the truss beam bridge main beam mentioned above is used to perform a simulation test of the aeroelastic model of the truss beam bridge main beam. BRIEF DESCRIPTION OF THE DRAWINGS

[0075] The drawings constituting a part of this application are used to provide a further understanding of this application and make other features, purposes and advantages of this application more apparent. The drawings and descriptions of the exemplary embodiments of this application are used to explain this application and do not constitute an improper limitation on this application.

[0076] In addition, throughout the drawings, the same or similar reference numerals represent the same or similar elements. It should be understood that the drawings are schematic and that the elements and components are not necessarily drawn to scale.

[0077] In the attached figure:

[0078] Figure 1 Schematic diagram of the structure of the aeroelastic model simulation test system for the main beam of a truss bridge.

[0079] Figure 2 A logic diagram for generating theoretical critical wind speed values for an aeroelastic model simulation test system for a truss bridge girder.

[0080] Figure 3 This is the nominal stress-strain curve of Q235 steel.

[0081] Figure 4 This is a flow chart for generating a theoretical value of critical wind speed in Example 2. DETAILED DESCRIPTION

[0082] The following will describe embodiments of the present application in more detail with reference to the accompanying drawings. Although certain embodiments of the present application are shown in the accompanying drawings, it should be understood that the present application can be implemented in various forms and should not be construed as being limited to the embodiments described herein. On the contrary, these embodiments are provided to provide a more thorough and complete understanding of the present application. It should be understood that the drawings and embodiments of the present application are for illustrative purposes only and are not intended to limit the scope of protection of the present application.

[0083] It should also be noted that, for ease of description, only the parts related to the invention are shown in the drawings. In the absence of conflict, the embodiments and features in the embodiments of this application can be combined with each other.

[0084] The present application will be described in detail below with reference to the accompanying drawings and in combination with embodiments.

[0085] Example 1:

[0086] Reference Figures 1 and 2The aeroelastic model simulation test system for the main beam of a truss bridge includes: a wind tunnel environment module, a wind speed control module, an aeroelastic model, an evaluation module, a control module, a test generation module, and a data execution module. The wind tunnel environment module is used to construct a test environment simulating a natural wind field, and the wind tunnel environment module includes an adjustable angle of attack device and a turbulence generation unit; the wind speed control module is connected to the wind tunnel environment module and is used to adjust the wind tunnel wind speed through a variable frequency motor and provide real-time feedback of wind speed data; the aeroelastic model includes a plurality of segment modules distributed longitudinally along the main beam, and a connecting component connecting two adjacent segment modules; the evaluation module includes a laser displacement meter and a strain sensor arranged on each segment module, which is used to measure the dynamic deformation data of the aeroelastic model and generate a critical wind speed theoretical value when the dynamic deformation data reaches a preset maximum value. The control module is electrically connected to each elastic adjustment member and is used to adjust the elastic coefficient of the elastic adjustment member in real time. The test generation module has a built-in truss beam bridge structural performance degradation model, generates a sequence of elastic coefficient changes of the elastic adjustment member, and outputs multiple test schemes. The data execution module is linked with the test generation module and the wind speed control module, sequentially loads the elastic coefficient change sequence corresponding to each test scheme, triggers the wind tunnel test and records the critical wind speed theoretical value.

[0087] The wind tunnel environment module, wind speed control module, aeroelastic model, and evaluation module are all existing technologies. The key to this application is to adjust the aging state of the aeroelastic model to derive the theoretical minimum critical wind speed value for a truss girder bridge within its normal service life.

[0088] The following describes how to adjust the aging state of the aeroelastic model and how to obtain the theoretical minimum critical wind speed value within the normal service life of a truss bridge.

[0089] The connection assembly includes a free-end connection and an elastic connector. The elastic connector is a pressure-adjustable dual-shaft cylinder, with the free-end connection serving as the connection between the two connecting shafts of the elastic connector. Adjusting the air pressure within the cylinder adjusts the elastic coefficient of the elastic connector. The key to this application lies in transforming the rigid connector in the original aeroelastic model into an elastic connector with adjustable elasticity.

[0090] like Figure 3 As shown: Figure 3 The nominal stress-strain curve for Q235 steel is shown in Figure 2. This steel exhibits distinct deformation characteristics under different loads. When the load is less than 24°, Q235 steel is within its elastic limit. When the load exceeds 38°, Q235 steel begins to gradually yield and loses its load-bearing capacity. The elastic connector in this solution simulates the elastic behavior of connectors in an aeroelastic model. Given a known load and stretch, the curve can be simulated by controlling the displacement of the free end through hydraulic or pneumatic pressure.

[0091] The test generation module includes a model building unit, a scenario generation unit, a simulation unit, and a probability fusion unit. The model building unit has a built-in truss bridge structural performance degradation model. The scenario generation unit uses a particle swarm algorithm to assign an aging state to each connection component to generate a simulation scenario. The simulation unit presets the elastic coefficients for each aging state. The aging state of each connection component in the simulation scenario is set to the corresponding elastic coefficient to generate a test scenario. The probability fusion unit incorporates the transition probability of each aging state into the test scenario based on the truss bridge structural performance degradation model.

[0092] For example, an aeroelastic model has 10 connected components. By setting elastic coefficients for each of these components and testing them in a wind tunnel environment module, a theoretical critical wind speed value can be obtained. Repeated testing can yield multiple theoretical values, and the optimal value can be selected as the theoretical critical wind speed value.

[0093] In this application, the model building unit is based on a truss beam bridge structural performance degradation model and can dynamically reflect the performance degradation characteristics of the connection components during use. The simulation unit can generate the degraded elastic coefficient in real time when the connection components degrade, thereby realizing dynamic adjustment of the simulation scheme. The core of this scheme is to establish a corresponding relationship between the elastic coefficient of the elastic adjustment member and the stress characteristics of the connection components under different aging conditions. The specific implementation plan is as follows:

[0094] Specifically: the truss beam bridge structure performance degradation model includes a plurality of hidden states corresponding to aging information, and probability transition information between the hidden states.

[0095] The structural performance degradation model of a truss bridge is constructed as follows:

[0096] S1: Set a number of consecutive aging states q, where the number of aging states is Q and q represents the index of the aging state.

[0097] Under conventional test conditions, aerodynamic model connectors are classified into four aging states based on the effective connection length ratio. Specifically, the fundamental cause of truss bridge collapse due to airflow is the fracture and failure of connectors. Connectors continue to age during service, resulting in a gradual decrease in tensile strength. This solution divides the aging process of connectors into four aging states. To quantitatively characterize these states, the grading is based on the effective connection length ratio. For example, using a steel connector with a standard width of 10 cm, by prefabricating standardized notches of 1 cm, 2 cm, and 3 cm (or retaining effective lengths of 9 cm, 8 cm, and 7 cm), typical operating conditions with effective connection lengths of 100%, 90%, 80%, and 70% are achieved. The design principle of this solution is based on the mechanism of material degradation: connector aging is essentially an irreversible deterioration of the material's load-bearing properties, manifested as a continuous reduction in the volume of the steel structure with effective connection capacity. By prefabricating standardized notches, multiple sets of sample data with different aging states can be rapidly generated under test conditions, significantly improving data collection efficiency for subsequent model training.

[0098] S2: Set the state variable R t , R t represents the state variable at time t, R t ={ra t rb t rc t}, where ra t Represents the accumulated displacement in time t;

[0099] , Indicates the displacement length X of the connected component from time 0 to time t h Integration over time;

[0100] rb t Indicates the number of times a connected component is subjected to force within time t; the count is triggered by the strain sensor threshold.

[0101] rc t It indicates the time when the stress of the connected component exceeds the stress threshold within time t, and the statistical stress amplitude exceeds the design threshold.

[0102] In this scheme, the state variables are physical quantities that can be measured during aerodynamic tests and are directly measured by the laser displacement meter and strain sensor on the evaluation module.

[0103] State variable R tThe method for obtaining σ is as follows: a uniaxial quasi-static tensile test is performed on the connection components of the standard aeroelastic model, and a step-by-step incremental load is applied through a servo loading system: the load range covers the elastic stage (0≤σ≤240MPa), the plastic development stage (240MPa<σ≤380MPa) and the failure stage (σ>380MPa), where σ is the load.

[0104] For example, after applying a standard tensile load for a specific period of time to a certain connection component, the corresponding state variable value can be measured through mechanical analysis. When the component is in the second aging state, this state variable is classified as the second aging state characterization parameter. After accumulating data through a large number of experiments, sufficient sample data can be obtained for each aging state. For each aging state, the probability distribution characteristics of its state variable are the Markov transition characteristics that characterize the state. Therefore, the state variable R t Essentially, it carries the implicit feature information of the connection component migrating from aging state 1 to aging state 2.

[0105] S3: Set the initial state probability distribution P of the aging state, P={p1, ...p q …p Q}, q∈[1,Q], where p q is the probability of the qth aging state, and the initial state probability distribution P = {1, 0, 0, 0}.

[0106] As connected components age, they gradually transition from their first aging state to their final aging state. This process is irreversible and will inevitably occur, but the timing of individual transitions cannot be accurately predicted. Therefore, using transition probabilities to describe this state transition process provides greater accuracy.

[0107] S4: Establish the probability transfer matrix W of the aging state, W={w qs},w qs represents the probability of transitioning from aging state q to aging state s, s∈[1,Q]; s>q;

[0108] S5: Construct the state variable probability matrix E, E={b q (R t )} represents the probability density function of the connected component under each aging state;

[0109] ;

[0110] Where M represents the number of mixed Gaussian components contained in each aging state, m represents the index of the Gaussian mixture model component, m∈[1.M], g qm represents the mth Gaussian distribution weight of the aging state q; represents the mth Gaussian distribution function of the aging state q, u qm and Lqm are the mean vector and covariance vector respectively;

[0111] ;

[0112] in, Represents the matrix transformation symbol, P represents the initial state probability distribution, and exp represents the power operation of the base e of the natural logarithm.

[0113] The state variable is just an observation; there's no clear correspondence between the state variable and the aging state. However, according to aging laws, the distribution of the state variable corresponding to each aging state follows a regular pattern. The probability density function in this solution actually describes the distribution of the state variable under the corresponding aging state.

[0114] S6: The probability transfer matrix W and the state variable probability matrix E are collaboratively modeled to generate a truss bridge structural performance degradation model.

[0115] The probability transfer matrix W describes the implicit change information between aging states, while the state variable probability matrix E describes the intrinsic connection between the observed value of the state variable and the aging state. To this end, S6 specifically includes the following steps:

[0116] S61: Construct the initial truss bridge structural performance degradation model λ;

[0117] λ = (P, W, E);

[0118] Where P = {1, 0, 0, 0},

[0119] W=[w qs ] Q×Q , Q=4, the constraints of W are:

[0120] (1) w qs > 0, s > q, that is, the aging state can only be transferred to the aging state with a larger number;

[0121] (2) w qs =0, s≤q, that is, the aging state cannot be transferred to an aging state with a smaller number;

[0122] (3) , that is, the probability of the aging state q being transferred out is 1;

[0123] E={b q (R t )};

[0124] S62: Train the initial truss bridge structural performance degradation model λ to find the optimal model parameters so that the likelihood probability of the observation sequence is maximized;

[0125] Specifically, S62 includes the following steps:

[0126] S621: Initialize training parameters and the maximum number of iterations. The training parameters include the initial state distribution, the probability transfer matrix W, the state variable probability matrix E, and the change threshold.

[0127] S622: Update the probability transfer matrix W:

[0128] Calculate the forward vector using the forward algorithm ;

[0129] ;

[0130] Among them, R1…R t represents the data sequence of state variables from time 1 to time t, | represents the conditional probability symbol;

[0131] q t =q means that the time t is in the aging state q;

[0132] Calculate the likelihood probability P(R1;T|λ) of the data sequence;

[0133] ;

[0134] Where T represents the total duration of the data sequence, represents the forward vector at the final time T;

[0135] Calculate the backward variable ;

[0136] , is a backward variable, which represents the conditional probability of observing the data sequence from time t to T;

[0137] ;

[0138] represents the probability of the aging state q transferring to the aging state s after the updated probability transfer matrix is updated;

[0139] represents the probability of transitioning from aging state q to aging state s at time t;

[0140] ;

[0141] represents the probability of being in the aging state q at time t;

[0142] ;

[0143] according to Update the probability transfer matrix W;

[0144] S623: Update the state variable probability matrix E.

[0145] The specific steps include:

[0146] Gaussian distribution weight g qm Make updates;

[0147] ;

[0148] in, is the updated Gaussian distribution weight, represents the posterior probability that the state variable at time t belongs to the mth Gaussian component under the aging state q;

[0149] ;

[0150] ;

[0151] ;

[0152] represents a multivariate Gaussian distribution, represents the index of the Gaussian mixture model component, ∈[1.M], Indicates the first The mean vector of the Gaussian components; Indicates the first The covariance vector of the Gaussian components; is the updated mean vector, is the updated covariance vector;

[0153] according to Update the state variable probability matrix E.

[0154] S624: Repeat S622 and S623 to repeat the iteration until the likelihood probability of the observation sequence is less than the change threshold or the maximum number of iterations is reached.

[0155] Once the probability transfer matrix W and the state variable probability matrix E are constructed, a mapping relationship between the two can be established. Based on this, the aging state with the highest corresponding probability can be analyzed by inputting the state variables, thereby revealing the implicit association mechanism between the state variables and the aging state.

[0156] Furthermore, the simulation unit sets an elastic coefficient for each aging state. A dedicated elastic parameter system is configured for each aging state: when the load on the connection component is within the stress threshold, the system simulates normal elastic deformation of the steel structure. As the load increases, the connection component simulates plastic deformation, and as the load continues to increase, the connection component simulates yielding. Throughout the simulation, only the elastic coefficient needs to be adjusted. This solution sets four aging states, each of which includes a stress threshold, a load threshold, and a collapse threshold. When the load on the connection component is within the stress threshold, the connection component maintains its initial elastic coefficient. When the load on the connection component enters the load threshold range, the elastic coefficient is attenuated. When the load on the connection component exceeds the collapse threshold, the connection component's stretch is negatively correlated with its load-bearing capacity. When the load completely exceeds the collapse threshold, the elastic coefficient is reset to zero, the connection constraint is released, and free end motion is achieved. In this way, the simulation unit only needs to record the aging state set for each connection component and then map the corresponding stress threshold, load threshold, and collapse threshold to the elastic coefficient of the connection component.

[0157] The aging state of the connected component is an initial setting. During the wind tunnel test, it is necessary to continue simulating the aging transfer process of the connected component. To this end, the probability fusion unit records the corresponding state variable R of each connected component during the wind tunnel test. t The most likely aging state of each connected component is generated based on the new state variables.

[0158] For example, a connection component is set to the first aging state, and then the aging state is input into the truss bridge structure performance degradation model to obtain the initial state variable R of the connection component. t During the wind tunnel test, the tension on the connecting component increases, so its state variable R t Start to increase, and increase R t Inputting this data into the girder bridge structural degradation model yields a 60% probability of this connection component transitioning to the second aging state. The probabilities of all connection components transitioning to the next aging state are then calculated. Based on these probabilities, the aging states of all connection components are then adjusted, along with the force thresholds, load thresholds, and collapse thresholds for all connection components. This allows the aging process of a bridge to be simulated in a wind tunnel test.

[0159] Example 2:

[0160] Example 1 provides a method for simulating the wind resistance test of a truss bridge girder during aging by using a connection component with a controllable elastic coefficient. However, in practice, there are a large number of connection components on a truss bridge girder, and there are a large number of possible test schemes. In order to find the theoretical value of the lowest critical wind speed of a truss bridge girder during normal operation, the present invention also provides Example 2, referring to Figure 4 , Example 2 provides the following technical solutions:

[0161] An aging score is set for each aging state, and a score threshold is set; the scheme generating unit continuously generates simulation schemes on the premise that the sum of the aging scores in the simulation schemes is higher than the score threshold.

[0162] For example, the first aging state is set to 4 points, the second aging state is set to 3 points, the third aging state is set to 2 points, and the fourth aging state is set to 1 point.

[0163] The threshold score should be determined based on actual project requirements and is typically tied to the maintenance standards for truss bridge girders. In this example, the specification requires that girder connection components be maintained at or above the third aging state. However, actual operational data indicates that their aging state distribution exhibits a near-normal distribution: primarily concentrated in the second and third aging states (approximately 68%), with a smaller percentage in the first and fourth aging states (approximately 16% each). Setting the threshold score based on this distribution pattern effectively eliminates extreme operating conditions (e.g., all components in the third or fourth aging states) that are unlikely to occur under standard maintenance conditions.

[0164] Considering the mechanical sensitivity of truss bridge girders, sudden fracture of key connection components may cause structural instability. To capture such nonlinear failure modes and determine the theoretical value of the lowest critical wind speed, this proposal adopts the following technical approach:

[0165] The solution generation unit generates a new simulation solution based on the following steps:

[0166] Z1: Generate a sorting sequence based on the geometric distance between each connection component and the center line of the main beam in ascending order;

[0167] The smaller the central support structure of a truss bridge's main girder, the greater the stress it is subjected to in practice. To this end, the connection components are sorted from low to high, and the serial number of the connection component can reflect its stress situation.

[0168] Z2: Set the mean and standard deviation of the aging score;

[0169] The distribution of the aging status of each connected component in the simulation scenario must satisfy the mean and standard deviation of the aging score. The mean describes the concentration of aging status, while the standard deviation describes the spread of aging status away from the mean. The mean aging score is actually another way of expressing the score threshold.

[0170] In this solution, we need to find the theoretical minimum critical wind speed under normal conditions. Therefore, after setting the score threshold, we need to make the sum of the aging scores of each connected component exactly equal to the score threshold. To this end, directly setting the mean of the aging scores will be more representative.

[0171] The above three data are set as needed. In this solution, there are 20 connected components, the mean aging score is 3 points, the score threshold is 60 points, and the standard deviation can be any value between 1 and 2.

[0172] Z3: Based on the mean and standard deviation of the aging scores, the probability density of each aging state is obtained. The number of connected components J1, J2, J3, and J4 corresponding to each aging state is calculated based on the probability density.

[0173] After obtaining the mean and standard deviation, we can draw a normal distribution curve of the aging state distribution, and then we can get the probability density of each aging state. After the total number of connected components is known, we can get the number of connected components corresponding to each aging state.

[0174] Z4: Set J1 first particles, J2 second particles, J3 third particles, and J4 fourth particles. Randomly place the first, second, third, and fourth particles into a sorted sequence to generate a simulation plan. The positions of each first, second, third, and fourth particle do not overlap. Send the simulation plan to the simulation unit to generate a test plan and obtain a theoretical critical wind speed value.

[0175] The number of first particles corresponds to the number of first aging states, so placing the first particle in the sorted sequence indicates that the connected component corresponding to the first particle in the sorted sequence is in the first aging state. Similarly, placing all first, second, third, and fourth particles in the sorted sequence will generate a simulation solution.

[0176] In practice, in order to reduce the number of particles, the reverse filling method can be used - non-dominant particle positions are allocated first, and the remaining positions are automatically assigned to the maximum state.

[0177] For example, in this application, the number of second particles is the largest, so only the first particles, third particles and fourth particles can be set. After the first particles, third particles and fourth particles are diffused, the remaining positions in the sorting sequence can be set to the second aging state.

[0178] Z5: configure the objective function f(x);

[0179] ;

[0180] Where x is the input of the objective function f(x), Y1 and Y2 are preset weight coefficients, Y1<Y2, Y1+Y2=1, d represents the index of the connected component, D represents the total number of connected components, h d represents the burn-in score of the d-th connected component, c d represents the sequence number of the dth connected component in the sorted sequence, and F is the theoretical value of the critical wind speed of the wind tunnel test.

[0181] Thus, in the objective function f(x), the smaller the theoretical critical wind speed value, the larger f(x) becomes. Furthermore, the closer the severely deteriorated connection component is to the center of the truss bridge girder, the larger f(x) becomes. Therefore, during the iteration process, the objective function f(x) set in this solution guides severely deteriorated connection components toward the center of the truss bridge girder while striving to find the lowest theoretical critical wind speed value.

[0182] Z6: Set the maximum number of iterations and update the particle position based on the maximum objective function. Repeat steps Z4-Z5 for B consecutive iterations, where B is a positive integer, until the maximum number of iterations is reached or f(x) reaches a threshold. When updating the particle position, you can perform random updates by setting speed, update distance, etc.

[0183] In the technical solution provided in this application, a particle swarm algorithm is used for continuous iterative operations, which can quickly find the lowest theoretical critical wind speed value of the main beam of a truss bridge under daily maintenance work.

[0184] Example 3:

[0185] A method for simulating an aeroelastic model of a truss girder bridge main beam is provided. The method uses the simulating test system for the aeroelastic model of a truss girder bridge main beam described in Example 1 to perform a simulation test on the aeroelastic model of the truss girder bridge main beam.

[0186] The above descriptions are merely some preferred embodiments of the present application and illustrate the technical principles employed. Those skilled in the art should understand that the scope of the invention described in the embodiments of the present application is not limited to technical solutions formed by specific combinations of the aforementioned technical features, but also encompasses other technical solutions formed by any combination of the aforementioned technical features or their equivalents without departing from the aforementioned inventive concept. For example, a technical solution formed by replacing the aforementioned features with (but not limited to) technical features with similar functions disclosed in the embodiments of the present application.

Claims

1. An aeroelastic model simulation test system for a truss bridge main beam, comprising: Wind tunnel environment module, wind speed control module, aeroelastic model and evaluation module; an aeroelastic model, comprising a plurality of segment modules distributed longitudinally along the main beam, and a connecting component connecting two adjacent segment modules; Its characteristics are: The connection component includes: Two free end connecting parts are respectively fixed to the end faces of adjacent segment modules; The elastic adjusting member is an elastic damper with an adjustable elastic coefficient, and both ends are rigidly connected to the free end connecting portion; The aeroelastic model simulation test system for the truss beam bridge main beam also includes: A control module, electrically connected to each elastic adjustment member, for adjusting the elastic coefficient of the elastic adjustment member in real time; The test generation module has a built-in truss bridge structural performance degradation model, generates a sequence of elastic coefficient changes of connection components under different aging conditions, and outputs multiple test plans; The data execution module is linked with the test generation module and the wind speed control module to sequentially load the elastic coefficient variation sequence corresponding to each test scheme, trigger the wind tunnel test and record the critical wind speed theoretical value.

2. The aeroelastic model simulation test system for a truss bridge main beam according to claim 1, characterized in that: The truss beam bridge structural performance degradation model includes a plurality of hidden states corresponding to aging information and probability transition information between the hidden states.

3. The aeroelastic model simulation test system for a truss bridge main beam according to claim 2, characterized in that: The structural performance degradation model of a truss bridge is constructed as follows: S1: Set a number of consecutive aging states q, the number of aging states is Q, and q represents the index of the aging state; S2: Set the state variable R t , R t represents the state variable at time t, R t ={ra t rb t rc t }, where ra t It represents the accumulated displacement in time t; , Indicates the displacement length X of the connected component from time 0 to time t h Integration over time; rb t Indicates the number of times the connected component is subjected to force within time t; rc t Indicates the time during which the force on the connected component exceeds the force threshold within time t; S3: Set the initial state probability distribution P of the aging state, P={p1, ...p q …p Q }, q∈[1,Q], where p q is the probability of the qth aging state occurring; S4: Establish the probability transfer matrix W of the aging state, W={w qs },w qs represents the probability of transitioning from aging state q to aging state s, s∈[1,Q]; s>q; S5: Construct the state variable probability matrix E, E={b q (R t )} represents the probability density function of the connected component under each aging state; ； Among them, M represents the number of mixed Gaussian components contained in each aging state, g qm represents the mth Gaussian distribution weight of the aging state q; represents the mth Gaussian distribution function of the aging state q, u qm and L qm are the mean vector and covariance vector respectively; ； in, represents the matrix conversion symbol, P represents the initial state probability distribution, and exp represents the power operation of the base e of the natural logarithm; S6: The probability transfer matrix W and the state variable probability matrix E are collaboratively modeled to generate a truss bridge structural performance degradation model.

4. The aeroelastic model simulation test system for a truss bridge main beam according to claim 3, characterized in that: S6 includes the following steps: S61: Construct the initial truss bridge structural performance degradation model λ; λ = (P, W, E); Where P = {1, 0, 0, 0}, W=[w qs ] Q×Q , Q=4, the constraints of W are: w qs > 0, s > q, that is, the aging state can only be transferred to the aging state with a larger number; w qs =0, s≤q, that is, the aging state cannot be transferred to an aging state with a smaller number; , that is, the probability of the aging state q being transferred out is 1; E={b q (R t )}; S62: Train the initial truss bridge structural performance degradation model λ to find the optimal model parameters so that the likelihood probability of the observation sequence is maximized.

5. The aeroelastic model simulation test system for a truss bridge main beam according to claim 4, characterized in that: S62 includes the following steps: S621: Initialize training parameters and the maximum number of iterations. The training parameters include the initial probability transfer matrix W, the initial state variable probability matrix E, and the change threshold. S622: Update the probability transfer matrix W: S623: Update the state variable probability matrix E; S624: Repeat S622 and S623 to repeat the iteration until the likelihood probability of the observation sequence is less than the change threshold or the maximum number of iterations is reached.

6. The aeroelastic model simulation test system for a truss bridge main beam according to claim 5, characterized in that: S622: Specifically including; Calculate the forward vector using the forward algorithm ; ； Among them, R1…R t represents the data sequence of state variables from time 1 to time t, | represents the conditional probability symbol; q t =q means that the time t is in the aging state q; Calculate the likelihood probability P(R1;T|λ) of the data sequence; ； Where T represents the total duration of the data sequence, represents the forward vector at the final time T; Calculate the backward variable ; ； is a backward variable, which represents the conditional probability of observing the data sequence from time t to T; Iteratively update the probability transfer matrix W and the state variable probability matrix E through expectation maximization; ； represents the probability of the aging state q transferring to the aging state s after the updated probability transfer matrix is updated; ； represents the probability of transitioning from aging state q to aging state s at time t; ； represents the probability of being in the aging state q at time t; according to Update the probability transfer matrix W; S623: Specifically including: Gaussian distribution weight g qm Make updates; ； in, is the updated Gaussian distribution weight, represents the posterior probability that the state variable at time t belongs to the mth Gaussian component under the aging state q; ； ； ； represents a multivariate Gaussian distribution, represents the index of the Gaussian mixture model (GMM) component, Indicates the first The mean vector of the Gaussian components; Indicates the first The covariance vector of the Gaussian components; is the updated mean vector, is the updated covariance vector; according to Update the state variable probability matrix E.

7. The aeroelastic model simulation test system for a truss bridge main beam according to claim 3, characterized in that: The pneumatic model connection components are divided into four aging states according to the ratio of the effective connection length under conventional test conditions.

8. The aeroelastic model simulation test system for a truss bridge main beam according to any one of claims 1 to 7, characterized in that: The test generation module includes: Model building unit with built-in truss bridge structure performance degradation model; The solution generation unit sets an aging state for each connected component based on the particle swarm algorithm to generate a simulation solution; A simulation unit, which presets elastic coefficients under various aging conditions and sets the aging conditions of various connection components in the simulation scheme to corresponding elastic coefficients to generate a test scheme; The probability fusion unit incorporates the transition probability of each aging state into the test plan based on the truss bridge structure performance degradation model.

9. The aeroelastic model simulation test system for a truss bridge main beam according to claim 8, characterized in that: Set an aging score for each aging state and set a score threshold; The scheme generation unit continuously generates simulation schemes on the premise that the sum of the aging scores in the simulation schemes is higher than the score threshold.

10. A method for simulating an aeroelastic model test of a truss bridge main beam, characterized in that: A simulation test of the aeroelastic model of a truss girder bridge main beam is performed using the aeroelastic model simulation test system of a truss girder bridge main beam according to any one of claims 1 to 9.

Images