A multi-modal intelligent collaborative suppression method and system for wind-induced vibration during the construction period of a cable-stayed bridge

Abstract

The present invention provides a multi-modal intelligent collaborative suppression method and system for wind-induced vibration during the construction period of a cable-stayed bridge, which relates to the field of cooperative control technology. It includes mapping three-dimensional wind speed vector field data and strain fields to a unified three-dimensional spatio-temporal grid coordinate system, and screening and generating a composite risk focus coordinate set; defining weighted Voronoi tessellation weights according to the main girder stiffness distribution and path probability density to generate a responsibility partition set; establishing an empirical mapping model of damping force-current-vibration response, and solving the optimal current sequence through model predictive control; monitoring the bridge deck acceleration response data after vibration suppression through a MEMS accelerometer array to form an optimized control strategy; using MAML meta-learning to update model parameters online, and achieving cross-section migration and deployment of the model through lightweight adjustment, so as to complete the multi-modal intelligent collaborative suppression of wind-induced vibration during the construction period of the cable-stayed bridge. The present invention realizes the multi-modal intelligent collaborative suppression of wind-induced vibration during the construction period of the cable-stayed bridge, ensuring the stability and safety of the bridge structure.

Description

Technical Field

[0001] The present invention relates to the technical field of cooperative control, and more particularly, to a method and system for intelligent cooperative suppression of multi-modal wind-induced vibration during the construction period of a cable-stayed bridge. Background Art

[0002] In recent years, with the rapid development of the construction of long-span cable-stayed bridges, the problem of wind-induced vibration control during the construction period has become increasingly prominent. Traditional vibration suppression methods mainly rely on passive dampers or temporary counterweights. Although they can alleviate some vibrations, they have significant limitations. For example, during construction, the bridge structure is vulnerable to wind-induced vibration, leading to structural safety and construction progress problems.

[0003] Existing wind-induced vibration control technologies, on the one hand, passive control cannot adapt to the time-varying structural stiffness and transient wind field characteristics during the construction period, resulting in low vibration suppression efficiency; on the other hand, existing technologies mostly adopt single-modal control strategies and are difficult to deal with the common multi-modal coupled vibration problems in the construction of cable-stayed bridges. For example, in the existing method based on a fixed-parameter tuned mass damper, although it can suppress vibrations at specific frequencies, it cannot be dynamically adjusted to adapt to the stiffness changes during the cantilever erection process and is prone to failure in a strong turbulent wind field. In addition, existing solutions lack systematic optimization of the cooperative control of multiple actuators, resulting in uneven resource allocation, insufficient vibration suppression in some areas, and excessive energy consumption in other areas. Summary of the Invention

[0004] The purpose of the present invention is to provide a method and system for intelligent cooperative suppression of multi-modal wind-induced vibration during the construction period of a cable-stayed bridge to improve the above problems. To achieve the above purpose, the technical solutions adopted by the present invention are as follows:

[0005] In a first aspect, the present application provides a method for intelligent cooperative suppression of multi-modal wind-induced vibration during the construction period of a cable-stayed bridge, including:

[0006] Collect three-dimensional wind speed vector field data of the construction area using a lidar array, and simultaneously synchronously monitor the strain field of the main girder joints through distributed fiber Bragg grating sensors. Through high-precision time synchronization technology, map the three-dimensional wind speed vector field data and the strain field to a unified three-dimensional space-time grid coordinate system, calculate the strain gradient amplitude based on spatial difference, and combine the local vorticity component of the wind speed field and the double-threshold criterion to screen and generate a set of composite risk focus coordinates, denoted as the target area for damper dynamic deployment;

[0007] Construct a bionic pheromone concentration field based on the set of composite risk focus coordinates, set update rules, generate the path probability density distribution of the damper in combination with the strain gradient amplitude, define the weighted Voronoi tessellation weight according to the main girder stiffness distribution and the path probability density, generate a set of responsibility partitions, verify the standard deviation of the vortex-induced vibration energy in each partition, and if not satisfied, iteratively adjust the tessellation weight until it meets the standard;

[0008] The local average wind speed is obtained by processing the three-dimensional wind speed vector field data in space and time, and combined with the characteristic height of the main girder, the target intervention frequency is inversely deduced through the Strouhal number criterion; according to the responsibility partition set, a micro-aerodynamic flap array is deployed inside the compound risk focus coordinate set, and based on the target intervention frequency, a real-time angle of attack command is generated, and an empirical mapping model of damping force-current-vibration response is established, and the optimal current sequence is solved through model predictive control; based on the time slot allocation protocol of the time-sensitive network, control commands are issued, where the control commands are to send the angle of attack command and the current spectrum to the edge computing node, and then control the aerodynamic flap and the magnetorheological damper;

[0009] According to the control command, the acceleration response data of the bridge deck after vibration suppression is monitored by the MEMS accelerometer array, the energy attenuation rate is calculated, and the angle of attack of the aerodynamic flap is adjusted accordingly. At the same time, the digital twin technology is used to generate a predicted response, and the control parameters are dynamically adjusted to form an optimized control strategy;

[0010] The optimized control strategy is adopted to classify and store the historical data according to the wind field characteristic parameters, and a multi-modal data set is constructed; the MAML meta-learning is used to update the model parameters online, the matching wind field parameters are loaded for the new construction section to initialize the meta-model, and the cross-section migration deployment of the model is realized through lightweight adjustment, so as to complete the multi-modal intelligent collaborative suppression of the wind vibration during the construction period of the cable-stayed bridge.

[0011] Preferably, the three-dimensional wind speed vector field data of the construction area is collected by the lidar array, and at the same time, the strain field of the main girder joint is synchronously monitored by the distributed fiber Bragg grating sensor. Through the high-precision time synchronization technology, the three-dimensional wind speed vector field data and the strain field are mapped to a unified three-dimensional space-time grid coordinate system, and the strain gradient amplitude is calculated based on the spatial difference, and combined with the local vorticity component of the wind speed field and the double-threshold criterion, the composite risk focus coordinate set is screened and generated, which is recorded as the target area for the dynamic deployment of the damper, including:

[0012] The three-dimensional wind speed vector field data of the construction area is obtained by the lidar array with a spatial resolution of 0.5 m and a time sampling interval of 0.1 s, and the strain field of the main girder joint is monitored in real time by the distributed fiber Bragg grating sensor network with a micro-strain sensitivity of 1 and a sampling frequency of 200 Hz; the carrier phase differential technology and the 1588 time protocol are used to map the data of the three-dimensional wind speed vector field data and the strain field to a unified three-dimensional space-time grid coordinate system to generate a fusion data set;

[0013] Perform spatial difference operations on the strain field, calculate the strain gradient vector of each spatial point in the strain field of the main girder joint, and extract the Euclidean norm in the strain gradient vector as the gradient amplitude. Determine whether the gradient amplitude exceeds 50 με / m. If it exceeds, mark it as a potential stress concentration area; if it does not exceed, do not mark it. Calculate the local vorticity component in the three-dimensional wind speed vector field data, set a threshold to determine whether the local vorticity component is greater than the threshold. If it exceeds, mark it as a strong rotation wind field area; if it does not exceed, do not mark it.

[0014] Traverse all spatio-temporal grid points, screen out the points that simultaneously include potential stress concentration areas and strong rotation wind field areas, and determine the grid points that meet the conditions as high-risk points of wind-structure coupling. Integrate the grid points of all high-risk points of wind-structure coupling to generate a composite risk focus coordinate set, denoted as the priority target area for damper dynamic deployment and aerodynamic flap regulation.

[0015] Preferably, construct a bionic pheromone concentration field based on the composite risk focus coordinate set, set an update rule, generate the path probability density distribution of the damper in combination with the strain gradient amplitude, define the weighted Voronoi partition weight according to the main girder stiffness distribution and the path probability density, generate a set of responsibility partitions, verify the standard deviation of the vortex-induced vibration energy in each partition. If it does not meet the requirements, iteratively adjust the partition weight until it reaches the standard, including:

[0016] In the composite risk focus coordinate set, assign an initial pheromone concentration value to all coordinate points, construct a bionic pheromone concentration field using the Gaussian kernel density estimation method, and perform dynamic update using a dual-time scale mechanism suitable for the cantilever construction of cable-stayed bridges to obtain the updated pheromone concentration field. The dual-time scale mechanism includes short-time scale update and long-time scale reset;

[0017] According to the updated pheromone concentration field and the strain gradient amplitude, set a dynamic update rule and calculate the path probability density of the damper;

[0018] Combine the path probability density of the damper and the time-varying model of the main girder stiffness, apply the weighted Voronoi partition algorithm to generate a set of responsibility partitions, use the MEMS accelerometer array to collect data at a sampling rate of 200 Hz, and calculate the standard deviation of the acceleration response in each partition. If the standard deviation of any partition exceeds 8%, adjust the partition weight and re-partition until the standard deviation of the acceleration response in all partitions meets the standard, thereby obtaining a set of partitions that meet the standards.

[0019] Preferably, based on the set of responsibility partitions, deploy a micro-aerodynamic flap array inside the composite risk focus coordinate set, generate real-time angle-of-attack commands based on the target intervention frequency, and establish an empirical mapping model of damping force-current-vibration response, and solve the optimal current sequence through model predictive control, including:

[0020] Deploy a micro pneumatic flap array in the form of an equally spaced matrix within the composite risk focus coordinate set according to the risk assessment within the responsibility partition set;

[0021] Utilize the target intervention frequency to calculate the real-time angle of attack of each pneumatic flap at time t;

[0022] Based on the relationship between the damping force, current, and vibration response, construct an empirical mapping model, and obtain the optimal current sequence through model predictive control.

[0023] Preferably, according to the control instruction, monitor the acceleration response data of the bridge deck after vibration suppression through the MEMS accelerometer array, calculate the energy decay rate, and accordingly adjust the angle of attack of the pneumatic flap. At the same time, use digital twin technology to generate a predicted response and dynamically adjust the control parameters to form an optimized control strategy, including:

[0024] Deploy the MEMS accelerometer array at the key positions of the bridge according to the control instruction, where the key positions include the main girder, bridge tower, and stay cable structural components, and record the acceleration data of each sensor;

[0025] Analyze the collected acceleration data, and calculate the energy decay rate by comparing the energy changes of the system before and after vibration suppression;

[0026] Based on the energy decay rate, adjust the angle of attack of the pneumatic flap to obtain the adjusted angle of attack direction;

[0027] Use the collected real-time wind speed vector field data and bridge structure response data to conduct simulations in the digital twin model to predict the response of the bridge structure in the future period of time. Compare the predicted response obtained from the simulation with the actual monitoring data. If there are differences, optimize them, and combine the optimization results with the adjusted angle of attack direction to evaluate the current control strategy;

[0028] Dynamically adjust the control parameters in response to the energy decay rate and the current control strategy, where adjusting the control parameters includes iterative solution of model predictive control, optimization of the pneumatic flap angle of attack adjustment strategy, and magnetorheological damper current control sequence, so as to achieve the collaborative suppression of wind vibration and adapt to the current wind field conditions and bridge structure response.

[0029] In a second aspect, the present application also provides a multi-modal intelligent collaborative wind vibration suppression system for cable-stayed bridges during construction, including:

[0030] Mapping calculation module: It is used to collect three-dimensional wind speed vector field data of the construction area by using a lidar array, and simultaneously monitor the strain field of the main girder joints through distributed fiber Bragg grating sensors. Through high-precision time synchronization technology, the three-dimensional wind speed vector field data and the strain field are mapped to a unified three-dimensional space-time grid coordinate system. Based on spatial difference calculation, the strain gradient amplitude is obtained, and combined with the local vorticity component of the wind speed field and the double-threshold criterion, a set of composite risk focus coordinates is screened and generated, which is recorded as the target area for damper dynamic deployment;

[0031] Generation module: It is used to construct a bionic pheromone concentration field based on the set of composite risk focus coordinates, set update rules, generate the path probability density distribution of the damper in combination with the strain gradient amplitude, define the weighted Voronoi partition weight according to the main girder stiffness distribution and the path probability density, generate a set of responsibility partitions, verify the standard deviation of the vortex-induced vibration energy in each partition, and if not satisfied, iteratively adjust the partition weight until it meets the standard;

[0032] Establishment module: It is used to obtain the local average wind speed by processing the three-dimensional wind speed vector field data in space and time, and combined with the characteristic height of the main girder, inversely deduce the target intervention frequency through the Strouhal number criterion; According to the set of responsibility partitions, deploy a micro-aerodynamic wing array inside the set of composite risk focus coordinates, generate real-time angle-of-attack commands based on the target intervention frequency, and establish an empirical mapping model of damping force - current - vibration response, and solve the optimal current sequence through model predictive control; Based on the time slot allocation protocol of the time-sensitive network, issue control commands, where the control commands are to send the angle-of-attack commands and current spectra to the edge computing nodes, and then control the aerodynamic wings and magnetorheological dampers;

[0033] Adjustment and optimization module: It is used to monitor the bridge deck acceleration response data after vibration suppression through a MEMS accelerometer array according to the control commands, calculate the energy attenuation rate, and accordingly adjust the angle of attack of the aerodynamic wings. At the same time, use digital twin technology to generate predicted responses and dynamically adjust control parameters to form an optimized control strategy;

[0034] Construction module: It is used to adopt the optimized control strategy, classify and store historical data according to the wind field characteristic parameters, and construct a multi-modal data set; Use MAML meta-learning to update model parameters online, load matching wind field parameters for a new construction section to initialize the meta-model, and achieve cross-section migration and deployment of the model through lightweight adjustment, so as to complete the multi-modal intelligent collaborative suppression of wind-induced vibration during the construction period of the cable-stayed bridge.

[0035] Thirdly, the present application also provides a multi-modal intelligent collaborative suppression device for wind-induced vibration during the construction period of a cable-stayed bridge, including:

[0036] A memory, used to store computer programs;

[0037] A processor, which is used to implement the steps of the multi-modal intelligent collaborative suppression method for wind-induced vibration during the construction period of the cable-stayed bridge when executing the computer program.

[0038] In a fourth aspect, the present application also provides a readable storage medium, on which a computer program is stored. When the computer program is executed by a processor, the steps of the above-mentioned multi-modal intelligent collaborative suppression method for wind-induced vibration during the construction period of the cable-stayed bridge are implemented.

[0039] The beneficial effects of the present invention are as follows:

[0040] The present invention uses a lidar array and distributed fiber Bragg grating sensors to collect wind speed vector field and strain field data, maps them to a unified coordinate system through high-precision time synchronization technology, calculates the strain gradient amplitude, and combines vorticity components and double thresholds to screen out the risk focus coordinate set; constructs a bionic pheromone concentration field based on the risk focus coordinate set, generates the probability density distribution of the damper path in combination with the strain gradient amplitude, generates a responsibility partition set through weighted Voronoi tessellation, and verifies the standard deviation of the vortex-induced vibration energy.

[0041] According to the responsibility partition set, the present invention dynamically adjusts the attack angle of the micro-aero foil and the current of the magnetorheological damper, solves the optimal current sequence through model predictive control, realizes the continuous and effective suppression of wind-induced vibration, generates predicted responses using digital twin technology, compares with the actual monitoring data, evaluates the effect of the control strategy, and provides a basis for adjusting the control parameters. Based on the optimized control strategy, a multi-modal data set is constructed, the model parameters are updated online using MAML meta-learning, and cross-section migration and deployment of the model are achieved through lightweight adjustment, completing the multi-modal intelligent collaborative suppression of wind-induced vibration during the construction period of the cable-stayed bridge.

[0042] Other features and advantages of the present invention will be described in the subsequent specification, and some of them will become obvious from the specification, or can be understood by implementing the embodiments of the present invention. The objectives and other advantages of the present invention can be achieved and obtained through the structures specifically pointed out in the written specification, claims, and drawings. Description of the Drawings

[0043] In order to more clearly illustrate the technical solutions of the embodiments of the present invention, the following will briefly introduce the drawings required in the embodiments. It should be understood that the following drawings only show some embodiments of the present invention, and therefore should not be regarded as limiting the scope. For those of ordinary skill in the art, other related drawings can be obtained based on these drawings without creative efforts.

[0044] Figure 1 It is a schematic flow chart of the multi-modal intelligent collaborative suppression method for wind-induced vibration during the construction period of the cable-stayed bridge described in the embodiments of the present invention;

[0045] Figure 2Schematic diagram of the multi-modal intelligent collaborative suppression system for wind-induced vibration during the construction period of the cable-stayed bridge described in the embodiments of the present invention;

[0046] Figure 3 Schematic diagram of the multi-modal intelligent collaborative suppression equipment for wind-induced vibration during the construction period of the cable-stayed bridge described in the embodiments of the present invention. In the figure: 701, mapping calculation module; 702, generation module; 703, establishment module; 704, adjustment and optimization module; 705, construction module; 800, multi-modal intelligent collaborative suppression equipment for wind-induced vibration during the construction period of the cable-stayed bridge; 801, processor; 802, memory; 803, multimedia component; 804, I / O interface; 805, communication component. Specific implementation manners

[0047] To make the objectives, technical solutions and advantages of the embodiments of the present invention clearer, the technical solutions in the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings in the embodiments of the present invention. Obviously, the described embodiments are some, but not all, of the embodiments of the present invention. Usually, the components of the embodiments of the present invention described and illustrated herein can be arranged and designed in various different configurations. Therefore, the following detailed description of the embodiments of the present invention provided in the accompanying drawings is not intended to limit the scope of the claimed invention, but merely represents selected embodiments of the present invention. All other embodiments obtained by those of ordinary skill in the art based on the embodiments of the present invention without creative efforts shall fall within the protection scope of the present invention.

[0048] It should be noted that similar reference numerals and letters denote similar items in the following drawings. Therefore, once an item is defined in one drawing, it does not need to be further defined and explained in subsequent drawings. At the same time, in the description of the present invention, the terms "first", "second", etc. are only used for descriptive distinction and cannot be construed as indicating or implying relative importance.

[0049] Embodiment 1:

[0050] This embodiment provides a multi-modal intelligent collaborative suppression method for wind-induced vibration during the construction period of a cable-stayed bridge.

[0051] See Figure 1 , which shows that this method includes step S100, step S200, step S300, step S400 and step S500.

[0052] S100. Use a lidar array to collect three-dimensional wind speed vector field data of the construction area, and simultaneously monitor the strain field of the main girder joints through a distributed fiber Bragg grating sensor. Through high-precision time synchronization technology, map the three-dimensional wind speed vector field data and the strain field to a unified three-dimensional space-time grid coordinate system. Calculate the strain gradient amplitude based on spatial difference, and combine the local vorticity component of the wind speed field and the double-threshold criterion to screen and generate a composite risk focus coordinate set, denoted as the target area for damper dynamic deployment.

[0053] It can be understood that in this step S100, it includes S101, S102, and S103, where:

[0054] S101. Use a lidar array to obtain three-dimensional wind speed vector field data of the construction area with a spatial resolution of 0.5 m and a time sampling interval of 0.1 s, and monitor the strain field of the main girder joints in real time through a distributed fiber Bragg grating sensor network with a sensitivity of 1 microstrain and a sampling frequency of 200 Hz; use carrier phase differential technology and the 1588 time protocol to map the three-dimensional wind speed vector field data and the strain field data to a unified three-dimensional space-time grid coordinate system to generate a fusion data set;

[0055] It should be noted that a lidar array is used to obtain three-dimensional wind speed vector field data of the construction area, where the three-dimensional wind speed vector field data includes wind speed components υ x (x, y, z, t), υ y (x, y, z, t), and υ z (x, y, z, t), a spatial resolution of 0.5 m, and a time sampling interval of 0.1 s. The strain field ε(x, y, t) of the main girder joints is monitored in real time through a distributed fiber Bragg grating sensor network, with a sensitivity of 1 microstrain and a sampling frequency of 200 Hz. Set up a Beidou RTK reference station in the bridge construction area to provide real-time differential positioning signals for the lidar array and fiber Bragg grating sensor nodes (roving stations), and correct the spatial coordinates of each sensor to a unified construction coordinate system (such as the center of the bridge tower as the origin), with a plane positioning error ≤ 1 cm and an elevation error ≤ 2 cm; the lidar point positions obtain absolute coordinates (x lidar , y lidar , z lidar ) through RTK, and the fiber sensor node coordinates are marked as (x fiber , y fiber , z fiber ). Deploy an IEEE 1588 Precision Time Protocol (PTP) master clock, distribute synchronization signals to all sensor nodes through a fiber optic network to achieve full-system clock synchronization, with a time jitter < 100 nanoseconds, and the lidar wind speed data timestamp is t v , and the fiber strain data timestamp is t ε, the absolute value of the laser radar wind speed data timestamp minus the optical fiber strain data time is less than 1 microsecond. Then, with the longitudinal direction of the main beam as the X-axis, the transverse direction as the Y-axis, and the vertical direction as the Z-axis, a spatial grid is constructed: the grid resolution covering the main beam and the tower construction area is Δx = 0.5m, Δy = 0.5m, Δz = 0.3m, and the maximum vortex frequency corresponding to the time step is 10HzΔt = 0.1s, where the grid point coordinates are (x i ,y j , z k , t m ), where i, j, k are spatial indexes and m is a time index. Then, data interpolation and fusion are performed, where the laser radar discrete point cloud data is mapped to the grid through inverse distance weighted interpolation, and the fiber optic sensor data is assigned to the grid based on the nearest neighbor principle. If there is a sensor node q within r = 0.2m around a grid point p, then ∈ grid (p) = ∈(q), where ∈ grid (p) is the mapped strain value at the grid point p, ∈(q) is the measured strain value at the physical point q, otherwise linear interpolation is used for filling, and the time step t m Align wind speed and strain data to generate a fused dataset D(a1, b2, z k , t m )=[υ x , v y , z , ε], where a1, b2 and z k is the spatial coordinate (along the longitudinal, transverse and vertical directions of the bridge), t m is a time series point, υ x ,υ y and z is the three-dimensional wind speed component, and ε is the strain value. Finally, the residual δ between the interpolated data and the original data is calculated as ||D grid -D raw ||, where δ is the difference norm between the grid interpolation data and the original data, D grid is the gridded data matrix, D raw is the original sensor data matrix. If the maximum value of δ is greater than the threshold, such as wind speed residual > 0.2 m / s or strain residual > 5 με), local grid refinement is triggered (resolution is increased to 0.2 m), and at time tm, the gradient mutation of adjacent grid points is checked (such as / / / / >3m / s / m or / / / / >100με / m), the abnormal area is marked and the sensor status is manually reviewed.

[0056] S102. Perform spatial difference operations on the strain field, calculate the strain gradient vector at each spatial point of the strain field at the main girder joint, and extract the Euclidean norm in the strain gradient vector as the gradient amplitude. Determine whether the gradient amplitude exceeds 50 με / m. If it exceeds, mark it as a potential stress concentration area; if it does not exceed, do not mark it. Calculate the local vorticity component in the three-dimensional wind speed vector field data, set a threshold to determine whether the local vorticity component is greater than the threshold. If it exceeds, mark it as a strong rotating wind field area; if it does not exceed, do not mark it.

[0057] S103. Traverse all spatio-temporal grid points (a m , b n ), where a m is the coordinate of the m-th measuring point in the longitudinal direction of the bridge, and b n is the coordinate of the n-th measuring point in the transverse direction of the bridge. Select the points that simultaneously include the potential stress concentration area and the strong rotating wind field area, and determine the grid points that meet the conditions as the high-risk points of wind-structure coupling. Integrate the grid points of all wind-structure coupling high-risk points to generate a composite risk focus coordinate set, denoted as the priority target area for damper dynamic deployment and aerodynamic flap regulation.

[0058] It should be noted that for each spatial point, the partial derivatives of its strain in the x, y (and z directions if three-dimensional) are calculated by the central difference method to obtain the strain gradient vector of this point, which is used to identify the stress concentration area at the main girder joint (such as high-gradient areas like the anchorage end and welded joint). The calculation object of the strain gradient vector obtained in this step is the spatial strain change rate of each discrete measurement point in the strain field. The partial derivatives in each direction are directly obtained through difference operations and combined into a vector.

[0059] In this step, ω thresh = 0.8 s -1 , and the set threshold for the vorticity criterion of the wind field is: ω thresh = 50 με / m, to screen the areas with significant rotation intensity in the local wind field. The set threshold for the strain gradient amplitude is: Identify the stress concentration risk area at the main girder joint, and perform screening under composite conditions. Screen the points that simultaneously meet the following conditions:

[0060] Among them, ω z is the vorticity in the z direction (perpendicular to the bridge deck direction), and then generate a composite risk focus coordinate set, that is, Ω risk = {(x i , y j | the double-threshold condition is satisfied)}, and the output Ω risk is used as the priority target area for damper dynamic deployment and aerodynamic flap regulation to ensure that the suppression resources accurately cover the strong coupling area of wind vibration energy and structural response, and at the same time balance the sensitivity and false alarm rate through the double-threshold values.

[0061] S200. Construct a bionic pheromone concentration field based on the composite risk focus coordinate set, set an update rule, generate the path probability density distribution of the damper by combining the strain gradient amplitude, define the weighted Voronoi tessellation weight according to the main girder stiffness distribution and the path probability density, generate a set of responsibility partitions, verify the standard deviation of the vortex-induced vibration energy in each partition, and if not satisfied, iteratively adjust the tessellation weight until it meets the standard.

[0062] It can be understood that in this step S200, it includes S201, S202, and S203, where:

[0063] S201. In the composite risk focus coordinate set, assign an initial pheromone concentration value to all coordinate points, construct a bionic pheromone concentration field using the Gaussian kernel density estimation method, and perform dynamic update using a double-time-scale mechanism applicable to the cantilever construction of cable-stayed bridges, so as to obtain the updated pheromone concentration field, where the double-time-scale mechanism includes short-time-scale update and long-time-scale reset;

[0064] It should be noted that in the composite risk focus coordinate set, an initial value τ ij (0) = 1 is assigned to all coordinate points, and the dynamic evolution formula of the pheromone concentration is defined as where is the strain gradient amplitude, suppressing the pheromone accumulation in high-stress regions, and outputting the dynamic pheromone field τ ij (t) characterizing the path attractiveness distribution of the risk area. The short-time-scale update in this step is:

[0065] where ΔE vortex represents the increment of vortex-induced vibration energy, represents the strain gradient amplitude of the temporary anchorage area, φ(t) is the pheromone concentration field at time t, λ is the pheromone decay coefficient, and Δt = 0.1 s is the short-time scale; long-time-scale reset (for each assembled segment): when the main girder stiffness distribution K(x, y) changes by more than 15% due to segment assembly, re-initialize the concentration field.

[0066] S202. According to the updated pheromone concentration field and the strain gradient amplitude, set a dynamic update rule and calculate the path probability density of the damper. The calculation formula is as follows:

[0067]

[0068] In the formula, P damp (x, y) is the priority of deploying the damper at the coordinate (x, y), φ is the bionic pheromone concentration field, is the strain gradient amplitude, H(ω - ω th ) is the vorticity threshold function, ω thThresholds set for the cantilever construction of the steel box girder The global integral normalization factor is used to integrate the entire bridge deck area

[0069] In this step, this probability density reflects the layout priority of dampers at different positions during the wind vibration suppression of the cable-stayed bridge during the construction period, and is preferentially arranged in high-risk areas (high vorticity, high strain gradient areas). Through normalization, it is ensured that the sum of the probability density on the bridge deck is 1, providing a scientific basis for the optimal layout of dampers

[0070] S203. Combine the damper path probability density and the time-varying model of the main girder stiffness, apply the weighted Voronoi tessellation algorithm to generate the set of responsibility partitions, use the MEMS accelerometer array to collect data at a sampling rate of 200 Hz, and calculate the standard deviation of the acceleration response within each partition; if the standard deviation of any partition exceeds 8%, adjust the tessellation weight and re-tessellate until the standard deviation of the acceleration response of all partitions meets the standard, thus obtaining a set of partitions that meet the standards

[0071] It should be noted that in this step, the acquisition and calculation of the time-varying model of the main girder stiffness come from real-time monitoring data and computational analysis. The real-time monitoring data comes from obtaining the strain distribution and deformation data of the main girder through the distributed optical fiber monitoring system. At the same time, the vehicle dynamic weighing system is used to obtain the load information, and combined with the strain sensor data, the change of the main girder stiffness is evaluated; while the computational analysis is based on the finite element method, considering the geometric shape, material properties and initial prestress distribution of the main girder, and through the least squares linear fitting of the monitoring data, a first-order linear equation of the main girder stiffness changing with time is obtained, so as to obtain the attenuation rate coefficient of the stiffness, and further obtain the optimized time-varying model of the stiffness

[0072] S300. Obtain the local average wind speed by processing the three-dimensional wind speed vector field data in space and time, and combine the characteristic height of the main girder to inversely deduce the target intervention frequency through the Strouhal number criterion; based on the set of responsibility partitions, deploy the micro-aerodynamic fin array inside the composite risk focus coordinate set, generate the real-time angle of attack command based on the target intervention frequency, and establish an empirical mapping model of damping force - current - vibration response, and solve the optimal current sequence through model predictive control; based on the time slot allocation protocol of the time-sensitive network, issue control commands, where the control commands are to send the angle of attack command and the current spectrum to the edge computing node, and then control the aerodynamic fins and magnetorheological dampers

[0073] It should be noted that in this step, the pneumatic fins are small devices installed on the bridge structure. By adjusting their angle of attack, the aerodynamic force distribution on the bridge surface is changed, thereby suppressing the vibration of the bridge under the action of wind. This method is particularly important in long-span cable-stayed bridges because such bridges are prone to vortex-induced vibration (VIV) under the action of wind, which affects the safety and service life of the bridge. The dynamic angle-of-attack control of the pneumatic fins is achieved by real-time monitoring of the wind field and structural response, and generating real-time angle-of-attack commands according to the target intervention frequency to suppress the local aeroservoelastic coupling effect. The current sequence is related to the magnetorheological damper, which is an intelligent damper whose damping force can be adjusted by changing the current. In the multi-modal intelligent collaborative suppression method for wind-induced vibration during the construction period of cable-stayed bridges, an empirical mapping model of damping force-current-vibration response is established, and the optimal current sequence is solved using model predictive control to achieve effective control of bridge vibration. This method can accurately adjust the damping force to adapt to different wind field conditions and the dynamic response of the bridge. In summary, the pneumatic fins and the current sequence play important roles in the multi-modal intelligent collaborative suppression method for wind-induced vibration during the construction period of cable-stayed bridges. Through intelligent control technology, they act together on the bridge structure to effectively suppress wind-induced vibration and ensure the safety and stability of the bridge.

[0074] It is understood that the optimal current sequence, together with the angle-of-attack commands of the pneumatic fins, is sent to the edge computing node based on the time-sensitive network (TSN) time-slot allocation protocol to ensure that the control delay is less than 10 ms. After being sent, these commands will be used to control the pneumatic fins and the magnetorheological damper, thereby achieving intelligent collaborative suppression of the wind-induced vibration of the cable-stayed bridge.

[0075] In this embodiment, the local average wind speed and the characteristic height of the main girder are important parameters for calculating the target intervention frequency and conducting wind-induced vibration control. The local average wind speed can be measured in the construction area by devices such as a lidar array, while the characteristic height D of the main girder usually refers to a representative dimension of the bridge main girder, such as height or width. The acquisition of these parameters is crucial for understanding and predicting hydrodynamic phenomena, such as in the wind load analysis of buildings and bridges. After obtaining the local average wind speed and the characteristic height of the main girder, the Strouhal number can be used to calculate the target intervention frequency. The Strouhal number is a dimensionless number used to describe the frequency of vortex shedding in the wake of an object in fluid flow, which relates the vortex shedding frequency to the flow velocity and the characteristic length of the object. As for the optimal current sequence, it is obtained through the nonlinear MPC control of the magnetorheological damper and is used to adjust the damping force to suppress bridge vibration. This sequence, together with the angle-of-attack command of the aerodynamic flap, is sent to the edge computing node through the time-slot allocation protocol of the time-sensitive network (TSN) to ensure that the control delay is less than 10 ms. After sending these commands, the edge computing node will control the aerodynamic flap and the magnetorheological damper, thereby achieving intelligent collaborative suppression of the wind-induced vibration of the cable-stayed bridge. This process involves precise control strategies to adapt to different wind field conditions and the dynamic response of the bridge.

[0076] Among them, the three-dimensional wind speed vector field data usually includes the magnitude and direction information of the wind speed, which can be used to calculate the local average wind speed. The local average wind speed refers to the average value of the wind speed in a specific area or time period, and this parameter is crucial for understanding and predicting the impact of wind on structures (such as cable-stayed bridges). In practical applications, the local average wind speed can be obtained by performing spatial and temporal averaging on the wind speed vector field data.

[0077] Among them, the local average wind speed can be obtained by statistically analyzing the collected wind speed data within a certain time and space range. For example, multiple wind speed sensors can be deployed in the bridge construction area to collect wind speed data at various locations, and then these data can be averaged to obtain the local average wind speed. This method can provide in-depth understanding of the wind field characteristics in the construction area, thus providing an important basis for bridge design and construction. As for the characteristic height D of the main girder, this is a parameter related to bridge design, usually given in the bridge design drawings or specifications, and it represents the characteristic dimensions such as the height or width of the bridge main girder. In wind engineering analysis, this parameter is used to calculate wind load-related indicators such as the Strouhal number.

[0078] It can be understood that in this step S300, it includes S301, S302, and S303, where:

[0079] S301. According to the risk assessment within the responsibility partition set, deploy a micro-aerodynamic flap array in the form of an equidistant matrix within the composite risk focus coordinate set;

[0080] It is understandable that in this step, according to the risk assessment within the responsibility partition set, a micro pneumatic flap array is deployed in a matrix form with equal spacing within the composite risk focus coordinate set. This deployment method can ensure that the flap array can cover the entire risk area without interference, thereby effectively controlling wind vibration, and then adjusting the airflow in a timely manner when wind vibration occurs, reducing the impact on the bridge structure.

[0081] S302. Calculate the real-time angle of attack of each pneumatic flap at time t using the target intervention frequency. The calculation formula is as follows:

[0082]

[0083] In the formula, θ i (t) is the real-time angle of attack of the i-th pneumatic flap, θ0 is the set value of the initial angle of attack, f v is the target intervention frequency, t is the time, K P is the proportional control gain, ε(x i , t) is the strain value measured at position x i and time t, H(·) is the Heaviside step function, is the strain gradient amplitude, 0.5 με / m is the threshold of the strain gradient, which is used as the condition to trigger the angle of attack adjustment; the technical effect of this step is that it can dynamically adjust the angle of attack of the pneumatic flap according to the real-time wind field and structural response, thereby effectively suppressing wind vibration.

[0084] S303. Based on the relationship between the damping force, current, and vibration response, construct an empirical mapping model and obtain the optimal current sequence through model predictive control. The calculation formula is as follows:

[0085]

[0086] In the formula, I j is the current control sequence of the j-th magnetorheological damper, T is the total time range, that is, the upper limit of the time step, ρ(x, y) is the trajectory density distribution, λ is the energy consumption penalty coefficient, I j (t) is the current control sequence at time t, ü(t) is the structural vibration acceleration response, and F is the relationship between the damping force, current, and vibration response.

[0087] It should be noted that in this step, by solving the optimal current sequence through model predictive control, it is possible to minimize energy consumption while ensuring the suppression effect, adapt to different wind field conditions and bridge structure responses, and improve the adaptability and robustness of the system. By solving the optimal current sequence through model predictive control, it can ensure that the magnetorheological damper can provide the best damping force under different wind field conditions, further enhancing the stability and safety of the bridge structure.

[0088] S400. According to the control instruction, monitor the acceleration response data of the bridge deck after vibration suppression through the MEMS accelerometer array, calculate the energy decay rate, and accordingly adjust the angle of attack of the aerodynamic flap. At the same time, use the digital twin technology to generate a predicted response and dynamically adjust the control parameters to form an optimized control strategy.

[0089] It can be understood that in this step S400, it includes S401, S402, S403, S404 and S405, where:

[0090] S401. According to the control instruction, deploy the MEMS accelerometer array at the key positions of the bridge, where the key positions include the main girder, the bridge tower and the stay cable structural components, and record the acceleration data of each sensor.

[0091] It can be understood that through these sensors, accurately monitor the acceleration changes of the bridge under the influence of wind vibration, record the acceleration values of each sensor, and reflect the actual dynamic behavior of the bridge structure under the action of wind vibration.

[0092] S402. Analyze the collected acceleration data, and calculate the energy decay rate by comparing the energy changes of the system before and after vibration suppression. The calculation formula is as follows:

[0093]

[0094] In the formula, η(t) is the energy decay rate at time t, ∑ x is the accumulation of the energy changes of all monitoring points on the entire bridge structure, is the velocity response before vibration suppression, α is the strain energy weight coefficient, is the strain history data before vibration suppression, is the acceleration response after vibration suppression, is the real-time strain after vibration suppression, is to calculate the energy change from the initial moment to the current moment, dr is the tiny time interval;

[0095] It should be noted that this formula calculates the energy change of the system before and after vibration suppression, and evaluates the effect of the vibration suppression measures by comparing the energy before and after vibration suppression. The closer the energy decay rate η(t) is to 1, the better the vibration suppression effect, and the more the system energy decreases. This step can quantitatively evaluate the effectiveness of the vibration suppression measures and provide a basis for the subsequent adjustment of the control strategy.

[0096] S403. Based on the energy decay rate, adjust the angle of attack of the aerodynamic flap to obtain the adjusted angle of attack direction. The calculation formula is as follows:

[0097]

[0098] In the formula, is the gradient direction sign function, and θ i (t) is the angle of attack of the i-th micro pneumatic fin at time t, and x i is the position coordinate of the i-th micro pneumatic fin on the bridge structure, and t is the current time point;

[0099] It should be noted that during the implementation of this method, based on the energy decay rate results calculated above, the angle of attack of the micro pneumatic fins will be adjusted to further optimize the wind vibration suppression effect. Specifically, the adjustment strategy of the angle of attack will be based on real-time monitoring data to dynamically change the angle of attack of the pneumatic fins to ensure that the direction of the counteracting force generated by the pneumatic fins is opposite to the direction of vorticity propagation, thereby effectively suppressing wind vibration.

[0100] In specific operations, the angle of attack of each pneumatic fin will be adjusted according to its position and the strain gradient at the current moment. If the strain gradient exceeds a preset threshold, it indicates that there may be a large stress concentration at this position. At this time, the angle of attack of the pneumatic fin will be increased or decreased accordingly to generate sufficient counteracting force to offset the vibration caused by wind vibration. In this way, the pneumatic fin can quickly respond to local wind field changes and achieve real-time suppression of the vibration of the bridge structure. That is to say, if the direction indicated by the gradient vector is the same as the current angle of attack direction, the angle of attack is increased by π radians (180 degrees) to change the aerodynamic direction of the pneumatic fin; if the directions are opposite, the angle of attack is decreased by π radians. This adjustment helps to optimize the performance of the pneumatic fin to suppress wind vibration.

[0101] The technical effect of this step is reflected in the ability to flexibly adjust the working state of the pneumatic fins according to the actual vibration state of the bridge structure and the wind field conditions. This can not only improve the adaptability and effectiveness of wind vibration suppression measures, but also maintain the stability and safety of the structure during the entire bridge construction period. Through this dynamic adjustment strategy, the pneumatic fins can participate more intelligently in the wind vibration control process, providing key technical support for the multi-modal intelligent collaborative suppression of wind vibration during the bridge construction period.

[0102] S404. Use the collected real-time wind speed vector field data and bridge structure response data to perform simulations in the digital twin model to predict the response of the bridge structure in the next period of time. Compare the predicted response obtained from the simulation with the actual monitoring data. If there are differences, optimize them, and combine the optimization results with the adjusted angle of attack direction to evaluate the current control strategy;

[0103] It can be understood that in this step, using digital twin technology, the collected data is input into the digital twin model for simulation to predict the response of the bridge structure in the future for a period of time. The digital twin model can simulate the equipment model, upload the status data to the virtual model for simulation operation, run synchronously with the real equipment, simulate the future operation conditions, and solve the problems of high cost of continuous operation of actual equipment and poor information accuracy. Compare the predicted response obtained from the simulation with the actual monitoring data. This step involves comparing and analyzing the simulation results with the actual acceleration response data, and evaluating the effect of the current control strategy through quantitative analysis such as regression analysis and trend analysis. If there are significant differences between the predicted response and the actual monitoring data, it indicates that the current control strategy may need to be adjusted. According to the comparison results, evaluate the current control strategy and provide a basis for subsequent adjustment of control parameters. This step involves continuously iteratively solving the model predictive control problem to achieve continuous and effective suppression of wind-induced vibration. Its control strategies include the angle of attack adjustment strategy of the aerodynamic fin and the current control sequence of the magnetorheological damper, which improves the adaptability and accuracy of the control strategy and provides key technical support for the multi-modal intelligent collaborative suppression of wind-induced vibration during the bridge construction period.

[0104] S405. Dynamically adjust the control parameters in response to the energy decay rate and the current control strategy, where adjusting the control parameters includes iteratively solving the model predictive control, optimizing the angle of attack adjustment strategy of the aerodynamic fin and the current control sequence of the magnetorheological damper, so as to achieve collaborative suppression of wind-induced vibration and adapt to the current wind field conditions and bridge structure response.

[0105] It should be noted that through the above steps, multi-modal intelligent collaborative suppression of wind-induced vibration during the construction period of the cable-stayed bridge is achieved, ensuring the stability and safety of the bridge structure. This involves continuously iteratively solving the model predictive control problem to ensure that the control strategy adapts to the changing wind field conditions and bridge response, and achieving the optimal wind-induced vibration suppression effect.

[0106] S500. Adopt an optimized control strategy, classify and store the historical data according to the wind field characteristic parameters, construct a multi-modal data set; use MAML meta-learning to online update the model parameters, load the matching wind field parameters for the new construction section to initialize the meta-model, and achieve cross-section migration and deployment of the model through lightweight adjustment, thereby completing the multi-modal intelligent collaborative suppression of wind-induced vibration during the construction period of the cable-stayed bridge.

[0107] It can be understood that in this step, an optimized control strategy is adopted, and the historical data is classified and stored according to the wind field characteristic parameters (such as Froude number Fr and Reynolds number Re) to construct a multi-modal data set, using the predicted response generated in the above steps and the calculated energy decay rate to provide a data basis for subsequent model parameter update.

[0108] It should be noted that in step S500, it includes S501, S502, and S503, where:

[0109] S501. Adopt an optimization control strategy, classify and store historical data according to wind field characteristic parameters (such as Froude number Fr and Reynolds number Re), and store the classified data as a multi-modal data set for subsequent model training and verification. The predicted response generated in the above steps and the calculated energy decay rate are utilized to provide a data basis for subsequent model parameter updates.

[0110] S502. Utilize the multi-modal data set to online update model parameters through the Model-Agnostic Meta-Learning (MAML) algorithm. For a new construction section, load the matching wind field parameters to initialize the meta-model, and update the trajectory generator parameters through few-shot incremental learning to adapt to the new wind field conditions.

[0111] Among them, MAML is a meta-learning method that updates the parameters of the trajectory generator through few-shot incremental learning, thereby improving the generalization ability of the model. This step utilizes the multi-modal data set obtained above to update model parameters through online learning to adapt to the new wind field conditions.

[0112] S503. Adopt lightweight technology to adjust the meta-model, reduce the model complexity and computational amount, and deploy the adjusted model to the new construction section to achieve cross-section model migration and application.

[0113] It can be understood that lightweight is a technology to reduce model complexity and computational amount, which helps the model to migrate and be deployed between different sections, ensuring that the control strategy can adapt to the current wind field conditions and bridge structure responses, and realizing wind vibration suppression.

[0114] Therefore, through the above steps, the multi-modal intelligent collaborative suppression of wind vibration during the construction period of the cable-stayed bridge is completed. It not only utilizes historical data and current control strategies to optimize control parameters, but also realizes the rapid adaptation and deployment of the model through online learning and lightweight adjustment, thereby improving the effect and efficiency of wind vibration suppression.

[0115] Embodiment 2:

[0116] As Figure 2 shown, this embodiment provides a multi-modal intelligent collaborative suppression system for wind vibration during the construction period of a cable-stayed bridge. Refer to Figure 2 The system includes:

[0117] Mapping calculation module 701: It is used to collect three-dimensional wind speed vector field data of the construction area by using a lidar array, and simultaneously monitor the strain field of the main girder joint through a distributed fiber Bragg grating sensor. Through high-precision time synchronization technology, the three-dimensional wind speed vector field data and the strain field are mapped to a unified three-dimensional space-time grid coordinate system. Based on spatial difference calculation, the strain gradient amplitude is obtained, and combined with the local vorticity component of the wind speed field and the double-threshold criterion, a set of composite risk focus coordinates is screened and generated, which is recorded as the target area for damper dynamic deployment;

[0118] Generation module 702: It is used to construct a bionic pheromone concentration field based on the set of composite risk focus coordinates, set an update rule, generate the path probability density distribution of the damper by combining the strain gradient amplitude, define the weighted Voronoi tessellation weight according to the main girder stiffness distribution and the path probability density, generate a set of responsibility partitions, verify the standard deviation of the vortex-induced vibration energy in each partition, and if not satisfied, iteratively adjust the tessellation weight until it meets the standard;

[0119] Establishment module 703: It is used to obtain the local average wind speed by processing the three-dimensional wind speed vector field data in space and time, and combined with the characteristic height of the main girder, inversely deduce the target intervention frequency through the Strouhal number criterion; according to the set of responsibility partitions, deploy a micro airfoil array inside the set of composite risk focus coordinates, generate a real-time angle of attack command based on the target intervention frequency, and establish an empirical mapping model of damping force - current - vibration response, and solve the optimal current sequence through model predictive control; based on the time-sensitive network's time slot allocation protocol, send control instructions, where the control instructions are to send the angle of attack command and the current spectrum to the edge computing node, and then control the airfoil and the magnetorheological damper;

[0120] Adjustment and optimization module 704: It is used to monitor the bridge deck acceleration response data after vibration suppression through a MEMS accelerometer array according to the control instructions, calculate the energy decay rate, and accordingly adjust the angle of attack of the airfoil. At the same time, use digital twin technology to generate a predicted response, dynamically adjust the control parameters, and form an optimized control strategy;

[0121] Construction module 705: It is used to adopt an optimized control strategy, classify and store historical data according to the wind field characteristic parameters, and construct a multi-modal data set; use MAML meta-learning to update the model parameters online, load the matching wind field parameters for a new construction section to initialize the meta-model, and achieve cross-section migration and deployment of the model through lightweight adjustment, so as to complete the multi-modal intelligent collaborative suppression of wind vibration during the construction period of the cable-stayed bridge.

[0122] Specifically, the mapping calculation module 701 includes:

[0123] The first generation unit: It is used to obtain the three-dimensional wind speed vector field data of the construction area with a spatial resolution of 0.5 m and a time sampling interval of 0.1 s by using a lidar array, and to monitor the strain field of the main girder joint in real time through a distributed fiber optic grating sensor network with a sensitivity of 1 microstrain and a sampling frequency of 200 Hz; by using carrier phase differential technology and the 1588 time protocol, map the three-dimensional wind speed vector field data and the strain field data to a unified three-dimensional spatio-temporal grid coordinate system to generate a fusion data set;

[0124] The judgment unit: It is used to perform spatial difference operations on the strain field, calculate the strain gradient vector of each spatial point in the strain field of the main girder joint, and extract the Euclidean norm in the strain gradient vector as the gradient amplitude, and judge whether the gradient amplitude exceeds 50 με / m. If it exceeds, it is marked as a potential stress concentration area; if it does not exceed, no marking is performed; calculate the local vorticity component in the three-dimensional wind speed vector field data, set a threshold to judge whether the local vorticity component is greater than the threshold. If it exceeds, it is marked as a strong rotating wind field area; if it does not exceed, no marking is performed;

[0125] The second generation unit: It is used to traverse all spatio-temporal grid points, screen out the points that simultaneously include potential stress concentration areas and strong rotating wind field areas, and determine the grid points that meet the conditions as wind-structure coupling high-risk points, integrate all grid points of wind-structure coupling high-risk points, generate a composite risk focus coordinate set, which is recorded as the priority target area for damper dynamic deployment and aerodynamic flap regulation.

[0126] Specifically, the generation module 702 includes:

[0127] The acquisition unit: It is used to assign an initial pheromone concentration value to all coordinate points in the composite risk focus coordinate set, construct a bionic pheromone concentration field by using the Gaussian kernel density estimation method, and perform dynamic update by using a dual-time scale mechanism suitable for the cantilever construction of cable-stayed bridges, so as to obtain the updated pheromone concentration field, where the dual-time scale mechanism includes short-time scale update and long-time scale reset;

[0128] The first calculation unit: It is used to set dynamic update rules according to the updated pheromone concentration field and the strain gradient amplitude, and calculate the damper path probability density. The calculation formula is as follows:

[0129]

[0130] In the formula, P damp (x, y) is the priority of deploying a damper at the coordinate (x, y), φ is the bionic pheromone concentration field, is the strain gradient amplitude, H(ω - ω th ) is the vorticity threshold function, ω th is the threshold set for the cantilever construction of the steel box girder, is the global integral normalization factor, which integrates over the entire bridge deck area;

[0131] The second calculation unit: used to combine the damper path probability density and the time-varying model of the main girder stiffness, apply the weighted Voronoi tessellation algorithm to generate a set of responsibility partitions, use the MEMS accelerometer array to collect data at a sampling rate of 200 Hz, and calculate the standard deviation of the acceleration response within each partition; if the standard deviation of any partition exceeds 8%, adjust the tessellation weight and re-tessellate until the standard deviation of the acceleration response of all partitions meets the standard, so as to obtain a set of partitions that meet the standard.

[0132] Specifically, the establishment module 703 includes:

[0133] The first deployment unit: used to deploy the micro pneumatic flap array in the form of an equally spaced matrix within the composite risk focus coordinate set according to the risk assessment within the responsibility partition set;

[0134] The third calculation unit: used to calculate the real-time angle of attack of each pneumatic flap at time t using the target intervention frequency, and its calculation formula is as follows:

[0135]

[0136] In the formula, θ i (t) is the real-time angle of attack of the i-th pneumatic flap, θ0 is the set value of the initial angle of attack, f v is the target intervention frequency, t is the time, K P is the proportional control gain, ε(x i , t) is the strain value measured at position x i and time t, H(·) is the Heaviside step function, is the strain gradient amplitude, and 0.5 με / m is the threshold of the strain gradient;

[0137] The construction unit: used to construct an empirical mapping model based on the relationship between the damping force, current, and vibration response, and obtain the optimal current sequence through model predictive control, and its calculation formula is as follows:

[0138]

[0139] In the formula, I j is the current control sequence of the j-th magnetorheological damper, T is the total time range, that is, the upper limit of the time step, ρ(x, y) is the trajectory density distribution, λ is the energy consumption penalty coefficient, I j (t) is the current control sequence at time t, ü(t) is the structural vibration acceleration response, and F is the relationship between the damping force, current, and vibration response.

[0140] Specifically, the adjustment and optimization module 704 includes:

[0141] The second deployment unit: used to deploy the MEMS accelerometer array at the critical positions of the bridge according to the control instruction, where the critical positions include the main girder, bridge tower and stay cable structural components, and record the acceleration data of each sensor;

[0142] The fourth calculation unit: used to analyze the collected acceleration data, calculate the energy decay rate by comparing the energy changes of the system before and after vibration suppression, and the calculation formula is as follows:

[0143]

[0144] In the formula, η(t) is the energy decay rate at time t, ∑ x is to accumulate the energy changes of all monitoring points on the entire bridge structure, is the velocity response before vibration suppression, α is the strain energy weight coefficient, is the strain history data before vibration suppression, is the acceleration response after vibration suppression, is the real-time strain after vibration suppression, is to calculate the energy change from the initial moment to the current moment, dr is the small time interval;

[0145] The adjustment unit: used to adjust the angle of attack of the pneumatic fin based on the energy decay rate to obtain the adjusted angle of attack direction, and its calculation formula is as follows:

[0146]

[0147] In the formula, is the gradient direction sign function, θ i (t) is the angle of attack of the i-th micro pneumatic fin at time t, x i is the position coordinate of the i-th micro pneumatic fin on the bridge structure, and t is the current time point;

[0148] The simulation and optimization unit: used to perform simulations in the digital twin model using the collected real-time wind speed vector field data and bridge structure response data to predict the response of the bridge structure in the future for a period of time, compare the predicted response obtained from the simulation with the actual monitoring data, if there are differences, optimize, and combine the optimization results with the adjusted angle of attack direction to evaluate the current control strategy;

[0149] The adjustment and coordination unit: used to dynamically adjust the control parameters in response to the energy decay rate and the current control strategy, where the adjusted control parameters include iterative solution of model predictive control, optimization of the pneumatic fin angle of attack adjustment strategy, and magnetorheological damper current control sequence, so as to achieve coordinated suppression of wind vibration and adapt to the current wind field conditions and bridge structure response.

[0150] It should be noted that regarding the system in the above embodiments, the specific manners in which each module performs operations have been described in detail in the embodiments related to the method, and will not be elaborated herein.

[0151] Embodiment 3:

[0152] Corresponding to the above method embodiment, in this embodiment, a multi-modal intelligent collaborative suppression device for wind-induced vibration during the construction period of a cable-stayed bridge is also provided. The multi-modal intelligent collaborative suppression device for wind-induced vibration during the construction period of a cable-stayed bridge described below can be correspondingly referred to the multi-modal intelligent collaborative suppression method for wind-induced vibration during the construction period of a cable-stayed bridge described above.

[0153] Figure 3 It is a block diagram of a multi-modal intelligent collaborative suppression device 800 for wind-induced vibration during the construction period of a cable-stayed bridge shown according to an exemplary embodiment. As Figure 3 shown, the multi-modal intelligent collaborative suppression device 800 for wind-induced vibration during the construction period of a cable-stayed bridge includes: a processor 801 and a memory 802. The multi-modal intelligent collaborative suppression device 800 for wind-induced vibration during the construction period of a cable-stayed bridge further includes one or more of a multimedia component 803, an I / O interface 804, and a communication component 805.

[0154] Among them, the processor 801 is used to control the overall operation of the multi-modal intelligent collaborative vibration suppression device 800 during the construction period of the cable-stayed bridge, so as to complete all or part of the steps in the above-mentioned multi-modal intelligent collaborative vibration suppression method for the cable-stayed bridge during the construction period. The memory 802 is used to store various types of data to support the operation of the multi-modal intelligent collaborative vibration suppression device 800 during the construction period of the cable-stayed bridge. These data may include, for example, instructions for any application or method operating on the multi-modal intelligent collaborative vibration suppression device 800 during the construction period of the cable-stayed bridge, as well as application-related data, such as contact data, sent and received messages, pictures, audio, video, and so on. The memory 802 can be implemented by any type of volatile or non-volatile storage device or a combination thereof, such as static random access memory (SRAM), electrically erasable programmable read-only memory (EEPROM), erasable programmable read-only memory (EPROM), programmable read-only memory (PROM), read-only memory (ROM), magnetic memory, flash memory, magnetic disk or optical disc. The multimedia component 803 may include a screen and an audio component. Among them, the screen may be a touch screen, and the audio component is used to output and / or input audio signals. For example, the audio component may include a microphone, and the microphone is used to receive external audio signals. The received audio signals may be further stored in the memory 802 or sent through the communication component 805. The audio component further includes at least one speaker for outputting audio signals. The I / O interface 804 provides an interface between the processor 801 and other interface modules, and the above-mentioned other interface modules may be a keyboard, a mouse or buttons, etc. These buttons may be virtual buttons or physical buttons. The communication component 805 is used for wired or wireless communication between the multi-modal intelligent collaborative vibration suppression device 800 during the construction period of the cable-stayed bridge and other devices. Wireless communication, such as Wi-Fi, Bluetooth, near field communication (NFC), 2G, 3G or 4G, or a combination of one or more of them. Therefore, the corresponding communication component 805 may include: a Wi-Fi module, a Bluetooth module or an NFC module.

[0155] In an exemplary embodiment, the multi-modal intelligent collaborative suppression device 800 for wind-induced vibration during the construction period of a cable-stayed bridge can be implemented by one or more application specific integrated circuits (ASICs), digital signal processors (DSPs), digital signal processing devices (DSPDs), programmable logic devices (PLDs), field programmable gate arrays (FPGAs), controllers, microcontrollers, microprocessors, or other electronic components, and is used to execute the above-mentioned multi-modal intelligent collaborative suppression method for wind-induced vibration during the construction period of a cable-stayed bridge.

[0156] In another exemplary embodiment, a computer-readable storage medium including program instructions is further provided. When the program instructions are executed by a processor, the steps of the above-mentioned multi-modal intelligent collaborative suppression method for wind-induced vibration during the construction period of a cable-stayed bridge are implemented. For example, the computer-readable storage medium can be the above-mentioned memory 802 including program instructions, and the above program instructions can be executed by the processor 801 of the multi-modal intelligent collaborative suppression device 800 for wind-induced vibration during the construction period of a cable-stayed bridge to complete the above-mentioned multi-modal intelligent collaborative suppression method for wind-induced vibration during the construction period of a cable-stayed bridge.

[0157] Embodiment 4:

[0158] Corresponding to the above method embodiment, in this embodiment, a readable storage medium is further provided. A readable storage medium described below can be correspondingly referred to with a multi-modal intelligent collaborative suppression method for wind-induced vibration during the construction period of a cable-stayed bridge described above.

[0159] A computer program is stored on the readable storage medium. When the computer program is executed by a processor, the steps of the multi-modal intelligent collaborative suppression method for wind-induced vibration during the construction period of a cable-stayed bridge in the above method embodiment are implemented.

[0160] The readable storage medium can specifically be various readable storage media such as a USB flash drive, a mobile hard disk, a read-only memory (ROM), a random access memory (RAM), a magnetic disk, or an optical disc that can store program codes.

[0161] The above are only the preferred embodiments of the present invention and are not intended to limit the present invention. For those skilled in the art, the present invention can have various changes and modifications. Any modifications, equivalent replacements, improvements, etc. made within the spirit and principle of the present invention shall be included within the protection scope of the present invention.

[0162] As described above, these are only the specific implementation manners of the present invention, but the protection scope of the present invention is not limited thereto. Any person skilled in the art within the technical scope disclosed by the present invention can easily think of changes or replacements, which should all be covered within the protection scope of the present invention. Therefore, the protection scope of the present invention shall be subject to the protection scope of the claims.

Claims

1. A multi-modal intelligent collaborative suppression method for wind-induced vibration during the construction period of a cable-stayed bridge, characterized in that, Including: Collect three-dimensional wind speed vector field data of the construction area using a lidar array. At the same time, synchronously monitor the strain field of the main girder joints through distributed fiber Bragg grating sensors. Through high-precision time synchronization technology, map the three-dimensional wind speed vector field data and the strain field to a unified three-dimensional space-time grid coordinate system. Calculate the strain gradient amplitude based on spatial difference, and combine the local vorticity component of the wind speed field and the double-threshold criterion to screen and generate a set of composite risk focus coordinates, denoted as the target area for damper dynamic deployment; Construct a bionic pheromone concentration field based on the set of composite risk focus coordinates, set update rules, generate the path probability density distribution of the damper by combining the strain gradient amplitude, define the weighted Voronoi partition weight according to the main girder stiffness distribution and the path probability density, generate a set of responsibility partitions, verify the standard deviation of the vortex-induced vibration energy in each partition, and if not satisfied, iteratively adjust the partition weight until it meets the standard; Obtain the local average wind speed by processing the three-dimensional wind speed vector field data in space and time, and combine the characteristic height of the main girder to inversely deduce the target intervention frequency according to the Strouhal number criterion; Based on the set of responsibility partitions, deploy a micro-aerodynamic wing array inside the set of composite risk focus coordinates. Based on the target intervention frequency, generate real-time angle-of-attack commands, and establish an empirical mapping model of damping force-current-vibration response. Solve the optimal current sequence through model predictive control; Based on the time-slot allocation protocol of the time-sensitive network, issue control commands, where the control commands are to send the angle-of-attack commands and the current spectrum to the edge computing node, and then control the aerodynamic wings and magnetorheological dampers; According to the control commands, monitor the acceleration response data of the bridge deck after vibration suppression through the MEMS accelerometer array, calculate the energy attenuation rate, and accordingly adjust the angle of attack of the aerodynamic wings. At the same time, use digital twin technology to generate predicted responses and dynamically adjust the control parameters to form an optimized control strategy; Adopt the optimized control strategy, classify and store the historical data according to the wind field characteristic parameters, and construct a multi-modal data set; Use MAML meta-learning to update the model parameters online, load the matching wind field parameters for the new construction section to initialize the meta-model, and achieve cross-section migration and deployment of the model through lightweight adjustment, so as to complete the multi-modal intelligent collaborative suppression of wind-induced vibration during the construction period of the cable-stayed bridge; Among them, constructing a bionic pheromone concentration field based on the set of composite risk focus coordinates, setting update rules, generating the path probability density distribution of the damper by combining the strain gradient amplitude, defining the weighted Voronoi partition weight according to the main girder stiffness distribution and the path probability density, generating a set of responsibility partitions, verifying the standard deviation of the vortex-induced vibration energy in each partition, and if not satisfied, iteratively adjust the partition weight until it meets the standard, which includes: In the set of composite risk focus coordinates, assign an initial pheromone concentration value to all coordinate points, construct a bionic pheromone concentration field using the Gaussian kernel density estimation method, and perform dynamic update using a double-time-scale mechanism suitable for the cantilever construction of the cable-stayed bridge to obtain the updated pheromone concentration field, where the double-time-scale mechanism includes short-time-scale update and long-time-scale reset; According to the updated pheromone concentration field and strain gradient amplitude, set the dynamic update rule and calculate the path probability density of the damper. The calculation formula is as follows: In the formula, P damp (x, y) is the priority of deploying a damper at the coordinates (x, y), φ is the bionic pheromone concentration field, is the strain gradient amplitude, H(ω - ω th ) is the vorticity threshold function, ω th is the threshold set for the cantilever construction of the steel box girder, is the global integral normalization factor, which integrates over the entire bridge deck area; Combined with the path probability density of the damper and the time-varying model of the main girder stiffness, apply the weighted Voronoi tessellation algorithm to generate the set of responsibility partitions. Use the MEMS accelerometer array to collect data at a sampling rate of 200 Hz, and calculate the standard deviation of the acceleration response within each partition. If the standard deviation of any partition exceeds 8%, adjust the tessellation weight and re-tessellate until the standard deviation of the acceleration response of all partitions meets the standard, so as to obtain a set of partitions that meet the standard; Among them, the predicted response in the generation of the predicted response using the digital twin technology is the acceleration predicted response.

2. The multi-modal intelligent collaborative suppression method for wind-induced vibration during the construction period of the cable-stayed bridge according to claim 1, wherein Use the lidar array to collect the three-dimensional wind speed vector field data of the construction area, and at the same time synchronously monitor the strain field of the main girder joint through the distributed fiber Bragg grating sensor. Through the high-precision time synchronization technology, map the three-dimensional wind speed vector field data and the strain field to a unified three-dimensional space-time grid coordinate system, calculate the strain gradient amplitude based on spatial difference, and combine the local vorticity component of the wind speed field and the double-threshold criterion to screen and generate the composite risk focus coordinate set, denoted as the target area for the dynamic deployment of the damper, including: Use the lidar array to obtain the three-dimensional wind speed vector field data of the construction area with a spatial resolution of 0.5 m and a time sampling interval of 0.1 s, and use the distributed fiber Bragg grating sensor network with a sensitivity of 1 microstrain and a sampling frequency of 200 Hz to monitor the strain field of the main girder joint in real time; use the carrier phase differential technology and the 1588 time protocol to map the data of the three-dimensional wind speed vector field and the strain field to a unified three-dimensional space-time grid coordinate system to generate a fusion data set; Perform spatial difference operations on the strain field, calculate the strain gradient vector of each spatial point in the strain field of the main girder joint, and extract the Euclidean norm in the strain gradient vector as the gradient amplitude. Judge whether the gradient amplitude exceeds 50 με / m. If it exceeds, mark it as a potential stress concentration area. If it does not exceed, do not mark it; calculate the local vorticity component in the three-dimensional wind speed vector field data, set a threshold to judge whether the local vorticity component is greater than the threshold. If it exceeds, mark it as a strong rotating wind field area. If it does not exceed, do not mark it; Traverse all space-time grid points, screen out the points that include both the potential stress concentration area and the strong rotating wind field area at the same time, and determine the grid points that meet the conditions as the high-risk points of wind-structure coupling. Integrate all the grid points of the high-risk points of wind-structure coupling to generate the composite risk focus coordinate set, denoted as the priority target area for the dynamic deployment of the damper and the regulation of the pneumatic wing.

3. The multi-modal intelligent collaborative suppression method for wind-induced vibration during the construction period of the cable-stayed bridge according to claim 1, wherein Deploy the micro pneumatic wing array inside the composite risk focus coordinate set according to the set of responsibility partitions, generate the real-time angle of attack command based on the target intervention frequency, and establish an empirical mapping model of damping force - current - vibration response, and solve the optimal current sequence through model predictive control, including: Deploy a micro pneumatic flap array in the form of an equidistant matrix within the composite risk focus coordinate set according to the risk assessment within the responsibility partition set; use the target intervention frequency to calculate the real-time angle of attack of each pneumatic flap at time t, and its calculation formula is as follows: where θ i (t) is the real-time angle of attack of the i-th pneumatic fin, θ0 is the set value of the initial angle of attack, f v is the target intervention frequency, t is time, K P is the proportional control gain, ε(x i , t) is the strain value measured at position x i and time t, H(·) is the Heaviside step function, is the strain gradient amplitude, and 0.5 με / m is the threshold of the strain gradient; Based on the relationship between the damping force, current, and vibration response, construct an empirical mapping model, and obtain the optimal current sequence through model predictive control, and its calculation formula is as follows: where I j is the current control sequence of the j-th magnetorheological damper, T is the total time range, i.e., the upper limit of the time step, ρ(x, y) is the trajectory density distribution, λ is the energy consumption penalty coefficient, I j (t) is the current control sequence at time t, ü(t) is the structural vibration acceleration response, and F is the relationship among the damping force, current, and vibration response.

4. The multi-modal intelligent collaborative suppression method for wind-induced vibration during the construction period of a cable-stayed bridge according to claim 1, characterized in that According to the control instruction, monitor the acceleration response data of the bridge deck after vibration suppression through the MEMS accelerometer array, calculate the energy decay rate, and accordingly adjust the angle of attack of the pneumatic flap. At the same time, use digital twin technology to generate a predicted response and dynamically adjust the control parameters to form an optimized control strategy, which includes: According to the control instruction, deploy the MEMS accelerometer array at the key positions of the bridge, where the key positions include the main girder, bridge tower, and stay cable structural components, and record the acceleration data of each sensor; Analyze the collected acceleration data, and calculate the energy decay rate by comparing the energy changes of the system before and after vibration suppression. The calculation formula is as follows: is the energy decay rate at time t, ∑ x is to accumulate the energy changes at all monitoring points on the entire bridge structure, is the velocity response before vibration suppression, α is the strain energy weight coefficient, is the strain history data before vibration suppression, is the acceleration response after vibration suppression, is the real-time strain after vibration suppression, is to calculate the energy change from the initial time to the current time, dr is the small time interval; Based on the energy decay rate, adjust the angle of attack of the pneumatic flap to obtain the adjusted angle of attack direction, and its calculation formula is as follows: In the formula, is the gradient direction sign function, and θ i (x) is the angle of attack of the i-th micro-aerodynamic flap at time t, where x i is the position coordinate of the i-th micro-aerodynamic flap on the bridge structure, and t is the current time point; Use the collected real-time wind speed vector field data and bridge structure response data to conduct simulations in the digital twin model to predict the response of the bridge structure in the future for a period of time. Compare the predicted response obtained from the simulation with the actual monitoring data. If there are differences, optimize them, and combine the optimization results with the adjusted angle of attack direction to evaluate the current control strategy; Dynamically adjust the control parameters in response to the energy decay rate and the current control strategy, where adjusting the control parameters includes iteratively solving the model predictive control, optimizing the pneumatic flap angle of attack adjustment strategy, and the magnetorheological damper current control sequence, so as to achieve the collaborative suppression of wind vibration and adapt to the current wind field conditions and bridge structure response.

5. A multi-modal intelligent collaborative suppression system for wind-induced vibration during the construction period of a cable-stayed bridge, based on the multi-modal intelligent collaborative suppression method for wind-induced vibration during the construction period of a cable-stayed bridge described in claim 1, characterized in that, Including: Mapping calculation module: used to collect the three-dimensional wind speed vector field data of the construction area by using the lidar array, and at the same time synchronously monitor the strain field of the main girder joint through the distributed fiber Bragg grating sensor. Through the high-precision time synchronization technology, map the three-dimensional wind speed vector field data and the strain field to the unified three-dimensional space-time grid coordinate system, calculate the strain gradient amplitude based on spatial difference, and combine the local vorticity component of the wind speed field and the double-threshold criterion to screen and generate the composite risk focus coordinate set, denoted as the target area for damper dynamic deployment; Generation module: used to construct a bionic pheromone concentration field based on the composite risk focus coordinate set, set the update rule, generate the path probability density distribution of the damper in combination with the strain gradient amplitude, define the weighted Voronoi partition weight according to the main girder stiffness distribution and the path probability density, generate the responsibility partition set, and verify the standard deviation of the vortex-induced vibration energy in each partition. If it is not satisfied, iteratively adjust the partition weight until it meets the standard; Establishment module: It is used to obtain the local average wind speed by processing the three-dimensional wind speed vector field data in space and time, and combine it with the characteristic height of the main girder to inversely deduce the target intervention frequency according to the Strouhal number criterion; deploy a micro-aerodynamic wing array inside the composite risk focus coordinate set according to the responsibility partition set, generate real-time angle-of-attack commands based on the target intervention frequency, establish an empirical mapping model of damping force-current-vibration response, and solve the optimal current sequence through model predictive control; based on the time-sensitive network's time slot allocation protocol, issue control commands, where the control commands are to send the angle-of-attack commands and current spectra to the edge computing nodes, and then control the aerodynamic wings and magnetorheological dampers; Adjustment and optimization module: It is used to monitor the bridge deck acceleration response data after vibration suppression through the MEMS accelerometer array according to the control commands, calculate the energy decay rate, and adjust the angle of attack of the aerodynamic wings accordingly. At the same time, use digital twin technology to generate predictive responses and dynamically adjust the control parameters to form an optimized control strategy; Construction module: It is used to adopt the optimized control strategy, classify and store historical data according to the wind field characteristic parameters, and construct a multi-modal data set; use MAML meta-learning to update the model parameters online, load the matching wind field parameters for the new construction section to initialize the meta-model, and achieve cross-section migration and deployment of the model through lightweight adjustment, so as to complete the multi-modal intelligent collaborative suppression of wind-induced vibration during the construction period of the cable-stayed bridge; Among them, the generation module includes: Obtaining unit: It is used to assign initial pheromone concentration values to all coordinate points in the composite risk focus coordinate set, construct a bionic pheromone concentration field using the Gaussian kernel density estimation method, and perform dynamic update using the double-time scale mechanism applicable to the cantilever construction of the cable-stayed bridge, so as to obtain the updated pheromone concentration field, where the double-time scale mechanism includes short-time scale update and long-time scale reset; First calculation unit: It is used to set dynamic update rules according to the updated pheromone concentration field and the strain gradient amplitude, and calculate the damper path probability density, and its calculation formula is as follows: In the formula, F damp (x, y) is the priority of deploying a damper at the coordinates (x, y), φ is the concentration field of bionic pheromone, is the strain gradient amplitude, H(ω - ω th ) is the vorticity threshold function, ω th is the threshold set for the cantilever construction of the steel box girder, is the global integral normalization factor, which integrates over the entire bridge deck area; Second calculation unit: It is used to combine the damper path probability density and the time-varying model of the main girder stiffness, apply the weighted Voronoi tessellation algorithm to generate the responsibility partition set, use the MEMS accelerometer array to collect data at a sampling rate of 200Hz, and calculate the standard deviation of the acceleration response within each partition; if the standard deviation of any partition exceeds 8%, adjust the tessellation weight and re-tessellate until the standard deviation of the acceleration response of all partitions meets the standard, so as to obtain a qualified partition set; Among them, the predictive response in the adjustment and optimization module is the acceleration predictive response.

6. The multi-modal intelligent collaborative suppression system for wind-induced vibration during the construction period of the cable-stayed bridge according to claim 5, wherein, The mapping calculation module includes: The first generation unit: It is used to obtain the three-dimensional wind speed vector field data of the construction area with a spatial resolution of 0.5 m and a time sampling interval of 0.1 s by using a lidar array, and to monitor the strain field of the main girder joint in real time through a distributed fiber Bragg grating sensor network with a sensitivity of 1 microstrain and a sampling frequency of 200 Hz; using carrier phase differential technology and the 1588 time protocol, map the three-dimensional wind speed vector field data and the strain field data to a unified three-dimensional spatio-temporal grid coordinate system to generate a fusion data set; The judgment unit: It is used to perform spatial difference operations on the strain field, calculate the strain gradient vector of each spatial point in the strain field of the main girder joint, and extract the Euclidean norm in the strain gradient vector as the gradient amplitude, and judge whether the gradient amplitude exceeds 50 με / m. If it exceeds, it is marked as a potential stress concentration area. If it does not exceed, it is not marked; calculate the local vorticity component in the three-dimensional wind speed vector field data, set a threshold to judge whether the local vorticity component is greater than the threshold. If it exceeds, it is marked as a strong rotating wind field area. If it does not exceed, it is not marked; The second generation unit: It is used to traverse all spatio-temporal grid points, screen out the points that simultaneously include potential stress concentration areas and strong rotating wind field areas, and determine the grid points that meet the conditions as wind-structure coupling high-risk points, integrate the grid points of all wind-structure coupling high-risk points, and generate a composite risk focus coordinate set, denoted as the priority target area for damper dynamic deployment and aerodynamic flap regulation.

7. The multi-modal intelligent collaborative suppression system for wind-induced vibration during the construction period of the cable-stayed bridge according to claim 5, characterized in that The establishment module, which includes: The first deployment unit: It is used to deploy a micro-aerodynamic flap array in the form of an equally spaced matrix within the composite risk focus coordinate set according to the risk assessment within the responsibility partition set; The third calculation unit: It is used to calculate the real-time angle of attack of each aerodynamic flap at time t by using the target intervention frequency, and its calculation formula is as follows: where, θ i (t) is the real-time angle of attack of the i-th pneumatic fin, θ0 is the set value of the initial angle of attack, f v is the target intervention frequency, t is time, K P is the proportional control gain, ε(x i , t) is the strain value measured at position x i and time t, H(·) is the Heaviside step function, is the strain gradient amplitude, and 0.5 με / m is the threshold value of the strain gradient; The construction unit: It is used to construct an empirical mapping model based on the relationship between damping force, current and vibration response, and obtain an optimal current sequence through model predictive control, and its calculation formula is as follows: where I j is the current control sequence of the j-th magnetorheological damper, T is the total time range, i.e., the upper limit of the time step, ρ(x, y) is the trajectory density distribution, λ is the energy consumption penalty coefficient, I j (t) is the current control sequence at time t, is the structural vibration acceleration response, and F is the relationship among the damping force, current, and vibration response.

8. The multi-modal intelligent collaborative suppression system for wind-induced vibration during the construction period of a cable-stayed bridge according to claim 5, wherein The adjustment and optimization module, which includes: The second deployment unit: It is used to deploy a MEMS accelerometer array at key positions of the bridge according to the control instruction, where the key positions include main girder, bridge tower and stay cable structural components, and record the acceleration data of each sensor; The fourth calculation unit: It is used to analyze the collected acceleration data, calculate the energy decay rate by comparing the energy changes of the system before and after vibration suppression, and the calculation formula is as follows: where η(t) is the energy decay rate at time t, and ∑ x is the accumulation of the energy changes at all monitoring points on the entire bridge structure, is the velocity response before vibration suppression, α is the strain energy weight coefficient, is the strain history data before vibration suppression, is the acceleration response after vibration suppression, is the real-time strain after vibration suppression, is to calculate the energy change from the initial time to the current time, and dr is the tiny time interval; The adjustment unit: It is used to adjust the angle of attack of the aerodynamic flap based on the energy decay rate to obtain the adjusted angle of attack direction, and its calculation formula is as follows: In the formula, is the gradient direction sign function, and θ i (t) is the angle of attack of the i-th micro-aerodynamic flap at time t, and x i is the position coordinate of the i-th micro-aerodynamic flap on the bridge structure, and t is the current time point; The simulation and optimization unit: It is used to perform simulations in the digital twin model by using the collected real-time wind speed vector field data and bridge structure response data to predict the response of the bridge structure in the next period of time, compare the predicted response obtained from the simulation with the actual monitoring data, optimize if there are differences, and combine the optimization results with the adjusted angle of attack direction to evaluate the current control strategy; Adjustment and coordination unit: It is used to respond to the energy decay rate and the current control strategy by dynamically adjusting control parameters. The adjustment of control parameters includes iterative solution of model predictive control, optimization of the airfoil angle of attack adjustment strategy, and current control sequence of magnetorheological dampers, so as to achieve coordinated suppression of wind vibration and adapt to the current wind field conditions and bridge structure responses.

Images