Method for forecasting bridge floor wind of river-crossing bridge and risk early warning service system

Abstract

The invention relates to the technical field of traffic engineering, in particular to a river-crossing bridge floor wind forecasting method and a risk early warning service system.The method comprises the steps that numerical weather forecasting data, three-dimensional terrain data, bridge structure three-dimensional model data and real-time traffic monitoring data are obtained, and a forecasting initial data set is constructed; simulating disturbance of a river canyon and a bridge structure to a background wind field based on a computational fluid dynamics method, and generating a bridge floor wind field analysis result; and further combining real-time bridge floor condition monitoring data and traffic flow data, querying a risk threshold matrix, dynamically correcting a risk level based on a deflection effect index, calculating to obtain a dynamic comprehensive risk level, and automatically generating graded early warning information and a traffic control decision instruction matched with the risk level and spatial distribution. According to the method, the refinement degree and the risk assessment accuracy of the bridge floor wind forecast of the river-crossing bridge can be improved, and the real-time early warning and active control of the bridge floor wind risk are realized.

Description

Technical Field

[0001] This invention relates to the field of transportation engineering technology, and in particular to a method for forecasting wind on the bridge deck of a cross-river bridge and a risk warning service system. Background Technology

[0002] With the continuous increase in the number of cross-river bridges and the increasing span of bridge structures, the impact of bridge deck winds on bridge traffic safety is becoming increasingly prominent. Due to the openness of the river surface, the undulating terrain of the valley, and the disturbance effect of the bridge structure itself on airflow, complex airflow phenomena often occur on the bridge deck area of ​​cross-river bridges, such as amplified wind speeds, sudden changes in wind direction, and uneven wind field distribution. Especially under severe convective weather or large-scale wind field conditions, this poses a significant threat to the driving stability of large vehicles and vehicles sensitive to crosswinds. Therefore, refined forecasting and risk assessment of bridge deck winds for cross-river bridges have become an important technical requirement for ensuring safe bridge traffic operation and improving the level of refined bridge management.

[0003] However, existing bridge wind forecasting and risk assessment technologies still have significant shortcomings. On the one hand, existing methods largely rely on regional-scale numerical weather prediction or single-point wind speed monitoring data, making it difficult to fully consider the disturbance effects of river and canyon topography and bridge structures on local wind fields, resulting in insufficient spatial resolution and accuracy in bridge wind forecasts. On the other hand, existing risk assessment methods often use wind speed thresholds as the main criterion, ignoring the impact of wind direction deflection and crosswinds on vehicle driving safety, and lack a dynamic assessment mechanism that combines bridge wind characteristics with real-time bridge conditions and traffic flow characteristics, making it difficult for risk assessment results to reflect actual operating conditions. In addition, existing technologies lack effective coordination between forecast results and traffic control measures, making it difficult to directly transform risk analysis results into actionable tiered early warning information and proactive control instructions, thus restricting the real-time response capability and overall safety level of cross-river bridges under complex weather conditions. Summary of the Invention

[0004] This invention provides a method for forecasting wind on the bridge deck of a cross-river bridge and a risk warning service system to solve the problems in existing bridge deck wind forecasting, such as the difficulty in accurately depicting the impact of river and canyon topography and bridge structure on local wind field disturbances, the difficulty in quantifying and assessing the risk of wind direction deflection, and the difficulty in effectively linking forecast results with bridge deck traffic operation status and active control measures.

[0005] A method for forecasting wind on the bridge deck of a cross-river bridge and a risk early warning service system, comprising the following steps:

[0006] S1: Obtain numerical weather forecast data, 3D terrain data, 3D model data of the bridge structure, and real-time traffic monitoring data for the area where the target cross-river bridge is located, and construct the initial forecast dataset;

[0007] S2: Based on the forecast initial dataset, simulate the disturbance of the background wind field by river canyons and bridge structures, and generate bridge deck wind field analysis results; the bridge deck wind field analysis results include wind direction, wind speed at each characteristic location on the bridge deck, and a deflection effect index used to quantify the degree of horizontal deflection of wind direction caused by terrain.

[0008] S3: Receive the bridge deck wind field analysis results, and simultaneously access real-time bridge deck condition monitoring data and real-time traffic flow data; match the wind direction and wind speed with a preset risk threshold matrix based on different bridge deck conditions and vehicle types to calculate the dynamic comprehensive risk level; based on the dynamic comprehensive risk level and the deflection effect index, automatically generate and output corresponding graded early warning information and decision instructions for the bridge active traffic control system.

[0009] Optionally, S1 includes:

[0010] S11: Parallel acquisition of numerical weather forecast data, 3D terrain data, 3D model data of bridge structure, and real-time traffic monitoring data of the target cross-river bridge area to form initial, unstandardized numerical weather forecast data, 3D terrain data, 3D model data of bridge structure, and real-time traffic monitoring data;

[0011] S12: Perform spatiotemporal benchmark unification and standardization processing on the initial, unstandardized numerical weather forecast data, three-dimensional terrain data, bridge structure three-dimensional model data and real-time traffic monitoring data, and output standardized numerical weather forecast data, three-dimensional terrain data, bridge structure three-dimensional model data and real-time traffic monitoring data.

[0012] S13: Align, fuse, and encapsulate the standardized numerical weather forecast data, three-dimensional terrain data, three-dimensional bridge structure model data, and real-time traffic monitoring data within a unified spatiotemporal framework to construct an initial forecast dataset containing multi-dimensional data layers with consistent internal relationships.

[0013] Optionally, the numerical weather forecast data is derived from the fusion product of a mesoscale weather forecast model and a regional high-resolution numerical model; the three-dimensional terrain data is acquired through airborne lidar scanning; the three-dimensional model data of the bridge structure is obtained by lightweight conversion of the bridge design BIM model; and the real-time traffic monitoring data is acquired through video vehicle detectors and microwave vehicle detectors deployed at the bridgehead and on the bridge deck.

[0014] Optionally, the alignment, fusion, and encapsulation within a unified spatiotemporal framework refers to: establishing a quasi-linear spatial reference system extending along the bridge deck, using the centerline of the target cross-river bridge as a spatial reference baseline; mapping the standardized numerical weather forecast data, three-dimensional terrain data, three-dimensional bridge structure model data, and real-time traffic monitoring data to the corresponding segments of this reference system according to their spatial locations; and finally encapsulating and generating the initial forecast dataset with a unified index, capable of quickly retrieving and extracting all related data based on the bridge deck mileage location.

[0015] Optionally, S2 includes:

[0016] S21: Based on the forecast initial dataset, a high-resolution computational fluid dynamics model covering the target cross-river bridge and surrounding river canyon areas is established. The physical geometry and computational grid are constructed using the three-dimensional terrain data and the three-dimensional model data of the bridge structure, and the numerical weather forecast data is interpolated and set as the model inlet boundary condition, thereby generating a customized high-resolution computational fluid dynamics model.

[0017] S22: Run the customized high-resolution computational fluid dynamics model to simulate the flow around, downrush and funneling effects generated when airflow passes through the space defined by the three-dimensional terrain data and the three-dimensional bridge structure model data, and solve for the three-dimensional wind field including the bridge deck space within the specified forecast period in the future.

[0018] S23: Extract airflow information of the height layer of each characteristic location on the bridge deck from the three-dimensional wind field, directly obtain the wind direction and wind speed at that location, and simultaneously calculate the absolute value of the horizontal angle between the wind direction at that location and the background wind direction provided by the numerical weather forecast data. Define the absolute value of the horizontal angle as the deflection effect index at that location. Integrate the wind direction, wind speed and deflection effect index of all characteristic locations to finally generate a complete bridge deck wind field analysis result.

[0019] Optionally, setting the numerical weather forecast data interpolation as the model inlet boundary condition specifically includes: spatially interpolating the horizontal wind field and temperature field data of the upwind boundary of the target area in the numerical weather forecast data to the corresponding inlet boundary grid points of the customized high-resolution computational fluid dynamics model, and assigning the surface roughness parameter represented by the three-dimensional terrain data to the bottom boundary of the model.

[0020] Optionally, the three-dimensional wind field has a grid resolution of not less than 10 meters in the horizontal direction and is layered into no less than 5 layers in the vertical direction near the bridge deck.

[0021] Optionally, S3 includes:

[0022] S31: Receive the bridge deck wind field analysis results transmitted by S2, and continuously access real-time bridge deck condition monitoring data from the bridge deck sensor network and real-time traffic flow data from the traffic monitoring system through the data interface to provide real-time data input for risk assessment;

[0023] S32: For each characteristic location on the bridge deck, the wind direction and wind speed corresponding to it in the bridge deck wind field analysis results are matched with the current risk scenario jointly determined by the real-time bridge deck condition monitoring data and real-time traffic flow data. The risk threshold matrix is ​​queried to obtain the basic risk level. Then, the basic risk level is dynamically corrected based on the deflection effect index corresponding to the location in the bridge deck wind field analysis results. Finally, the dynamic comprehensive risk level of each characteristic location is calculated.

[0024] S33: Based on the spatial distribution of the dynamic comprehensive risk level of all characteristic locations and its corresponding deflection effect index, combined with the preset early warning rule base and control strategy base, automatically generate hierarchical early warning information and decision instructions that match the content and level for different audiences and control equipment, and output the hierarchical early warning information and decision instructions to the designated early warning release platform and the control terminal of the bridge active traffic control system through the communication network.

[0025] Optionally, the real-time bridge surface condition monitoring data includes dry / wet / ice condition codes and friction coefficients collected by road surface sensors, and the real-time traffic flow data includes cross-sectional traffic flow, average vehicle speed, and the proportion of large vehicles obtained by video vehicle detectors.

[0026] A risk early warning service system is used to provide early warning services for the aforementioned method of wind forecasting on the surface of a cross-river bridge, and includes the following modules:

[0027] Data acquisition and preprocessing module: used to acquire numerical weather forecast data, three-dimensional terrain data, three-dimensional model data of bridge structure and real-time traffic monitoring data in parallel for the area where the target cross-river bridge is located, and to perform spatiotemporal benchmark unification and standardization processing on the above data, as well as alignment, fusion and encapsulation under a unified spatial reference system to construct the initial forecast dataset;

[0028] Bridge deck wind field simulation and analysis module: connected to the data acquisition and preprocessing module, used to establish a computational fluid dynamics model covering the target cross-river bridge and surrounding river canyon areas based on the forecast initial dataset, and generate bridge deck wind field analysis results including wind direction, wind speed and deflection effect index at various characteristic locations on the bridge deck by simulating the disturbance of the background wind field by the river canyon and bridge structure.

[0029] Real-time status access module: It is used to continuously access real-time bridge surface condition monitoring data from the bridge surface sensor network and real-time traffic flow data from the traffic monitoring system through the data interface, and to perform time synchronization and validity verification of the real-time data to provide real-time input for risk assessment.

[0030] Dynamic risk assessment module: connected to the bridge deck wind field simulation analysis module and the real-time status access module, used to match the wind direction and wind speed at each characteristic location of the bridge deck with a preset risk threshold matrix based on different bridge deck conditions and vehicle types, calculate the basic risk level, and dynamically correct the basic risk level by combining the deflection effect index of the corresponding characteristic location to obtain the dynamic comprehensive risk level of each characteristic location of the bridge deck.

[0031] Early warning decision generation module: connected to the dynamic risk assessment module, used to automatically generate graded early warning information matching the risk level and decision instructions for the bridge active traffic control system based on the spatial distribution of the dynamic comprehensive risk level of each characteristic location on the bridge surface and its corresponding deflection effect index, by calling the preset early warning rule library and control strategy library.

[0032] Early warning release and control interface module: connected to the early warning decision generation module, used to release the graded early warning information to the designated early warning release platform through the communication network, and send the decision instructions to the control terminal of the bridge active traffic control system, so as to realize active intervention and risk prevention and control of bridge traffic.

[0033] The beneficial effects of this invention are:

[0034] 1. This invention rigorously unifies, standardizes, and integrates numerical weather prediction data, high-precision 3D topographic data, 3D bridge structure model data, and real-time traffic monitoring data using a unified spatiotemporal reference. A quasi-linear spatial reference system extending along the bridge is constructed using the bridge deck centerline as a reference, ensuring precise alignment of meteorological background, topographic relief, bridge structure, and traffic operation status within the same spatial framework. This data organization method avoids the wind field misjudgment problems caused by inconsistent data scales and spatial misalignment in traditional bridge deck wind assessments. It enables subsequent computational fluid dynamics simulations to accurately reflect the local amplification, deflection, and uneven distribution characteristics of the bridge deck wind field due to river and canyon topography and bridge structure, thereby significantly improving the spatial precision of bridge deck wind forecasts.

[0035] 2. This invention not only acquires refined wind direction and speed at various characteristic locations on the bridge deck, but also constructs a deflection effect index by comparing the horizontal angle between the refined wind direction and the background wind direction. This index is used to quantify the degree of horizontal wind direction deflection caused by the river canyon terrain and bridge structure on the background wind field. Furthermore, the deflection effect index is introduced into the risk assessment process. When the deflection effect index exceeds a preset angle threshold, the basic risk level is upgraded and corrected. This mechanism avoids the problem of judging risk solely based on wind speed while ignoring the lateral force risk caused by sudden changes in wind direction. This makes the risk assessment of large vehicles and crosswind-sensitive vehicles more consistent with actual working conditions, effectively improving the reliability and safety margin of bridge deck wind risk identification.

[0036] 3. This invention, based on the spatial distribution of dynamic comprehensive risk levels at various characteristic locations on the bridge deck, and combined with an early warning rule base and a control strategy base, automatically generates tiered early warning information and decision-making instructions that match the risk level and spatial situation, and directly outputs them to the early warning release platform and the control terminal of the bridge active traffic control system. In this way, the bridge deck wind risk assessment results no longer remain at the analysis level, but can be transformed in real time into specific control measures such as variable message sign prompts, variable speed limits, and lane control suggestions. This enables proactive intervention and coordinated control of bridge deck traffic, thereby effectively reducing the risk of traffic accidents under severe bridge deck wind conditions and improving the overall operational safety and management efficiency of cross-river bridges under complex weather conditions. Attached Figure Description

[0037] To more clearly illustrate the technical solutions in this invention or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are only for this invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0038] Figure 1 This is a schematic diagram of the method flow according to an embodiment of the present invention;

[0039] Figure 2 This is a schematic diagram of the system flow according to an embodiment of the present invention. Detailed Implementation

[0040] The present invention will now be described in detail with reference to the accompanying drawings and specific embodiments. It should also be noted that, to make the embodiments more comprehensive, the following embodiments are the best and preferred embodiments, and those skilled in the art can use other alternative methods to implement some well-known technologies; moreover, the accompanying drawings are only for more specific description of the embodiments and are not intended to specifically limit the present invention.

[0041] It should be noted that the use of terms such as "an embodiment," "an embodiment," "an exemplary embodiment," and "some embodiments" in the specification indicates that the described embodiment may include a specific feature, structure, or characteristic, but not every embodiment necessarily includes that specific feature, structure, or characteristic. Furthermore, when a specific feature, structure, or characteristic is described in connection with an embodiment, implementing such a feature, structure, or characteristic in conjunction with other embodiments (whether explicitly described or not) should be within the knowledge of those skilled in the art.

[0042] Generally, terms can be understood at least partly from their use in context. For example, depending at least partly on the context, the term "one or more" as used herein can be used to describe any feature, structure, or characteristic in a singular sense, or a combination of features, structures, or characteristics in a plural sense. Additionally, the term "based on" can be understood not necessarily to convey an exclusive set of factors, but rather, alternatively, depending at least partly on the context, to allow for the presence of other factors that are not necessarily explicitly described.

[0043] like Figure 1 As shown, a method for forecasting wind on the bridge deck of a cross-river bridge includes the following steps:

[0044] S1: Obtain numerical weather forecast data, high-precision 3D terrain data, 3D model data of the bridge structure, and real-time traffic monitoring data for the area where the target cross-river bridge is located, and construct the initial forecast dataset. The specific steps are as follows:

[0045] S11: Parallel acquisition of numerical weather forecast data, high-precision 3D terrain data, bridge structure 3D model data, and real-time traffic monitoring data for the target cross-river bridge area, forming initial, unstandardized data, specifically:

[0046] In the data aggregation and preprocessing module, four independent parallel data acquisition tasks are first created, and each task is allocated independent computing threads or process resources. During the task initialization phase, the same acquisition boundary conditions are uniformly written for the four tasks. These boundary conditions include: the spatial coverage parameters of the target cross-river bridge and the time window parameters corresponding to the current forecast service. The spatial coverage parameters are based on the center location of the bridge, extending upstream and downstream to form a spatial range encompassing the river canyon area; the time window parameters are defined as the forecast start time and the real-time acquisition period corresponding to the current forecast time.

[0047] After completing the above configuration, four parallel execution tasks will be started simultaneously, allowing various types of data to be acquired in parallel within the same acquisition window.

[0048] In the corresponding parallel execution task, the system first reads the access address, access permission information, and data indexing rules of the fusion product of the mesoscale weather forecast model and the regional high-resolution numerical model from the system configuration. Then, it converts the spatial coverage parameters into geographic enclosing range request parameters supported by the fusion product, and converts the time window parameters into the corresponding combination request of the reporting time and forecast lead time. Afterward, it initiates a data retrieval operation to the fusion product according to the converted request parameters and receives the returned raw meteorological data files. For each received meteorological data file, only file integrity and field existence checks are performed; no data resampling or coordinate transformation is performed. After successful checks, the data is written to the raw meteorological data cache according to the original grid structure, retaining the original timestamp, grid coordinate definitions, and variable fields as initial, unstandardized numerical weather forecast data.

[0049] In the corresponding parallel execution task, the point cloud index information of the airborne lidar scanning results is first read. The index information includes at least the spatial coverage range and storage identifier of each point cloud block. Then, based on the spatial coverage range parameter, point cloud blocks that have spatial intersection with the target area are selected from the point cloud index. For each selected point cloud block, a point cloud file reading operation is performed, parsing only its original 3D coordinates and point attribute fields, without performing coordinate system transformation or filtering. After reading, each point cloud block, along with its original coordinate reference information, is written to the terrain raw data cache to form initial, unstandardized high-precision 3D terrain data.

[0050] In the corresponding parallel execution task, the three-dimensional model file obtained by lightweight conversion of the bridge design BIM model is first located; then the model parsing program is called to load the model file, and the geometric shape, topological connection relationship and component identification information of each component in the model are parsed in turn; during the parsing process, the original coordinate system of the model is kept unchanged, and only the bridge deck, main beam, bridge tower, cable surface and auxiliary components are extracted as geometric entity sets respectively; after the parsing is completed, the geometric entity sets and their hierarchical relationship are written into the original data cache of the structural model to form the initial, unstandardized three-dimensional model data of the bridge structure.

[0051] In the corresponding parallel execution task, communication connections are first established with each video vehicle detector and microwave vehicle detector deployed at the bridgehead and on the bridge surface. After the communication connection is established, a data subscription request is sent to each detection device, so that it continuously outputs monitoring data within the time period limited by the time window parameter. For each received data frame, the vehicle count, vehicle speed, vehicle type and the original sampling timestamp of the device are obtained according to the device type. After the parsing is completed, the data frame content and the corresponding device location identifier are written into the traffic raw data cache area to form the initial, unstandardized real-time traffic monitoring data.

[0052] S12: Perform spatiotemporal benchmark unification and standardization on the initial, unstandardized data, and output standardized data, specifically:

[0053] In the data aggregation and preprocessing module, Coordinated Universal Time (UTC) is first set as the sole time base. Then, the raw timestamp information from the meteorological raw data cache and the traffic raw data cache is read. For numerical weather prediction data, it is converted to the corresponding absolute UTC based on its start time and forecast lead time fields. For real-time traffic monitoring data, it is converted to UTC after time deviation correction based on the device sampling timestamps and device time synchronization records. After conversion, the unified timestamps are written back to the corresponding data records, forming a data object with a unified time base.

[0054] After unifying the time reference, a single geodetic datum and projected coordinate system are fixed as the spatial reference. Subsequently, the spatial coordinates of different types of data are processed sequentially. For numerical weather prediction data, the original grid coordinates are converted to planar coordinates in the projected coordinate system. For high-precision 3D terrain data, the original point cloud coordinates are batch-converted to 3D coordinates in the projected coordinate system. For 3D model data of bridge structures, the entire model is transformed from its original coordinate system to the projected coordinate system. For real-time traffic monitoring data, the coordinates of the deployment locations of each detection device are transformed to the projected coordinate system. All transformation processes use the same set of spatial transformation parameters, thereby achieving spatial reference unification.

[0055] After unifying the spatiotemporal reference, format standardization was performed on different types of data: numerical weather prediction data was converted from its original grid format into a general meteorological data format with a unified field structure; high-precision 3D terrain data and 3D bridge structure model data were converted into a standard 3D mesh format compatible with computational fluid dynamics methods; and real-time traffic monitoring data were converted into a structured time-series data format ordered by time. The format standardization process did not change the physical meaning of the data, only unifying the field structure and organization.

[0056] S13: Within a unified spatiotemporal framework, standardized data is aligned, fused, and encapsulated to construct the initial forecast dataset, specifically as follows:

[0057] In the data aggregation and preprocessing module, the geometric information of the bridge deck centerline is first read from the 3D model data of the bridge structure, which has already achieved spatial benchmark unification. If the bridge deck centerline exists in the form of multiple broken lines, the nodes of the broken lines are sequentially sorted according to the actual traffic direction of the bridge to form a continuous geometric curve of the centerline. Subsequently, taking the starting point of this centerline as the zero mileage point, the spatial distance between adjacent geometric nodes is calculated sequentially along the centerline direction, and the distances are accumulated to generate a "mileage value - spatial coordinate" correspondence table. Then, the centerline is discretized at a preset mileage sampling interval, and a unique mileage index identifier is generated at each discrete mileage position, thereby establishing a quasi-linear spatial reference system with the bridge deck centerline as the baseline and the mileage index as the main index.

[0058] For each geometric entity in the 3D model data of the bridge structure, calculate the minimum vertical projection distance from all nodes of the geometric entity to the centerline of the bridge deck; after completing the distance calculation, determine the position of the centerline projection point corresponding to the geometric entity, and read the mileage value of the projection point in the "mileage value-spatial coordinate" correspondence table; then, map the mileage value to the nearest mileage index identifier, and write the index identifier into the attribute field of the geometric entity, so that each bridge structure geometric entity is associated with a unique mileage index in the quasi-linear spatial reference system.

[0059] For each terrain point or terrain grid cell in the high-precision 3D terrain data that has completed spatial benchmark unification, calculate its nearest projected position to the bridge deck centerline in the projected coordinate system; read the corresponding mileage value based on the geometric position of the projected position on the centerline; then convert the mileage value into the nearest mileage index identifier and write the index identifier into the attribute field of the corresponding terrain point or terrain grid cell; through the above processing, the terrain data can be retrieved and sliced ​​according to the bridge deck mileage segment while maintaining the original 3D shape.

[0060] For each meteorological grid point in the numerical weather prediction data that has completed spatial benchmark unification, its plane coordinates in the projected coordinate system are read; then the vertical projection position of the grid point coordinates to the bridge deck centerline is calculated, and the corresponding mileage value is determined based on the geometric position of the projection position on the centerline; then the mileage value is converted into a mileage index identifier in a quasi-linear spatial reference system; after completing the index mapping, the original meteorological grid point variable values ​​are associated with the corresponding mileage index identifiers, so that the numerical weather prediction data can be organized and accessed according to the bridge deck mileage position.

[0061] For real-time traffic monitoring data that has achieved spatial benchmark unification, the deployment coordinates of each video vehicle detector and microwave vehicle detector are read; the projection calculation of the position coordinates onto the bridge centerline is performed to determine the projection point on the centerline; the corresponding mileage value is read according to the geometric position of the projection point and converted into a mileage index identifier; then the mileage index identifier is written into all traffic monitoring data records generated by the device, so that a one-to-one correspondence is established between real-time traffic monitoring data and bridge mileage position.

[0062] After completing the spatial mapping, Coordinated Universal Time (UTC) is used as the unified time axis, and a preset time alignment step size is set. Subsequently, numerical weather forecast data and real-time traffic monitoring data are resampled according to this time step size, so that each type of data has a corresponding data entry at each time step node. For data with a time resolution higher than the time step size, aggregation is performed according to the time window aggregation rules. For data with a time resolution lower than the time step size, interpolation is performed according to the adjacent time node interpolation rules, thereby ensuring that all types of data can be accessed simultaneously under the same mileage index and the same time index.

[0063] A two-layer index structure is constructed using mileage index identifiers and time index identifiers. Under this two-layer index structure, the 3D model data of the bridge structure is used as the structural data layer, the high-precision 3D terrain data as the terrain data layer, the numerical weather prediction data as the meteorological data layer, and the real-time traffic monitoring data as the traffic data layer, all of which are written into the same data container. During the writing process, cross-data-layer association pointers are established for each "mileage index-time index" combination, so that at any bridge mileage location and at any time node, the corresponding structural, terrain, meteorological, and traffic data can be retrieved simultaneously through the unified index. After the writing is completed, the data container is marked as the initial forecast dataset and output for subsequent steps.

[0064] S2: Based on the initial forecast dataset, simulate the disturbance of the background wind field by river canyons and bridge structures, and generate bridge deck wind field analysis results. The bridge deck wind field analysis results include wind direction and wind speed at various characteristic locations on the bridge deck, as well as a deflection effect index used to quantify the degree of horizontal deflection of wind direction caused by terrain. The specific steps are as follows:

[0065] S21: Based on the initial forecast dataset, establish a customized high-resolution computational fluid dynamics model covering the target cross-river bridge and surrounding river canyon areas, specifically:

[0066] First, high-precision 3D terrain data and 3D bridge structure model data with spatially unified precision are read from the initial forecast dataset. Using the centerline of the target cross-river bridge as a reference, the model is extended upstream and downstream along the bridge axis by a predetermined distance, and vertically to cover the near-ground boundary layer height, thus determining the overall computational domain of the computational fluid dynamics model. Next, the 3D terrain point cloud data within the computational domain is cropped, retaining only terrain data points falling within the computational domain. Simultaneously, the bridge deck, main beams, towers, and cable surface geometric entities located within the computational domain in the 3D bridge structure model are extracted. Finally, the cropped terrain data and bridge structure geometric entities are merged to form a complete physical geometric model for computational fluid dynamics modeling.

[0067] After constructing the physical geometric model, spatial discretization was performed on it. Specifically, unstructured tetrahedral meshes were used to fill and discretize the flow field space within the computational domain; multi-layered prismatic meshes were generated along the normal direction near the ground surface and bridge structure surface to characterize near-wall velocity gradients and shear effects. During mesh generation, the horizontal mesh scale of the bridge deck and its adjacent areas was controlled to be no greater than 10 meters, and at least 5 vertical mesh layers were set within the height range of the bridge deck. After mesh generation, a quality check was performed on the mesh, and mesh cells with high distortion or negative volume were removed, ultimately obtaining a hybrid computational mesh that meets the requirements of refined simulation.

[0068] After completing the computational grid construction, a Reynolds-averaged turbulence model was selected as the closed model for the computational fluid dynamics model. The Reynolds-averaged Navier-Stokes equations for incompressible flow were used as the governing equations, and the additional transport equations corresponding to the turbulence model were also loaded into the solver configuration. Subsequently, physical property parameters such as air density and dynamic viscosity were specified for the model, so that the model has the basic physical conditions for numerical solution.

[0069] After completing the physical and numerical framework setup of the model, the horizontal wind and temperature fields corresponding to the upwind boundary of the target area are read from the initial forecast dataset of numerical weather prediction data. Then, the upwind side boundary of the computational fluid dynamics model is identified as the inlet boundary, and spatial mapping is performed on all grid nodes on this inlet boundary: using spatial interpolation, the wind speed components, wind direction information, and temperature values ​​at the corresponding height layers in the numerical weather prediction data are assigned to the inlet boundary grid nodes; during the interpolation process, based on the coordinates of the inlet boundary grid nodes in three-dimensional space, their relative positions in the original grid points of the numerical weather prediction data are calculated, and the interpolation assignment is completed accordingly. After interpolation, the interpolated wind and temperature fields are written into the model configuration as the model inlet boundary conditions.

[0070] After setting the inlet boundary conditions, the roughness parameters representing the surface type from the high-precision 3D terrain data are assigned to the bottom boundary of the computational fluid dynamics model, giving the bottom boundary realistic surface friction characteristics. Simultaneously, the top boundary of the model is set as a free-slip or outflow boundary, and the leeward boundary is set as the outlet boundary to ensure that airflow can naturally exit the computational domain. This completes the entire process of establishing the customized high-resolution computational fluid dynamics model.

[0071] S22: Run a customized high-resolution computational fluid dynamics model to obtain a refined three-dimensional wind field, specifically:

[0072] After establishing a customized high-resolution computational fluid dynamics model and setting the inlet boundary, bottom boundary, and other boundary conditions, the model is numerically solved to obtain a refined three-dimensional wind field for a specified forecast period.

[0073] First, at the start of the calculation, the initial velocity and temperature fields within the computational domain are set to a spatial distribution consistent with the model inlet boundary conditions. Then, using the mass conservation condition of the incompressible airflow as a fundamental constraint, a continuity condition is applied to the entire computational domain, mathematically expressed as follows:

[0074] ;

[0075] in, Let be the three-dimensional average wind speed vector in the Reynolds-averaged sense. This represents the refined three-dimensional wind field velocity solution to be solved in this invention, which, under the premise of satisfying the continuity condition, performs a momentum conservation solution for the airflow motion process within the computational domain. The momentum conservation is expressed in Reynolds-averaged form as follows:

[0076] ;

[0077] in, air density, For the Reynolds mean pressure field, The molecular dynamic viscosity of air. The turbulent viscosity is calculated based on the Reynolds-averaged turbulence model and is used to characterize the turbulent momentum exchange effect induced by the topography of river canyons and bridge structures.

[0078] In this invention, the above equations are spatially discretized using a hybrid mesh of unstructured tetrahedral and prism layers, and solved iteratively using a time-progression approach. During the time-progression process, the Reynolds-mean-momentum equation is discretized and updated according to a preset time step, causing the velocity field to gradually converge over time. The update relationship can be summarized as follows:

[0079] ;

[0080] in, and Let represent the wind speed solutions for adjacent time steps. The time step used for numerical solution. This represents a discrete residual operator composed of convection, pressure gradient, and viscous diffusion terms. Within each time step, the updated velocity field satisfies the continuity condition through coupling correction between the velocity and pressure fields.

[0081] In the above solution process, the airflow numerically interacts with the complex geometric boundaries defined by high-precision three-dimensional terrain data and bridge structure three-dimensional model data. The complex boundary conditions are reflected in the governing equations through velocity gradient terms and pressure gradient terms, thus naturally forming the flow around the flow, terrain downrush effect and canyon narrowing effect in the solution results, without the need to introduce additional empirical correction terms.

[0082] Once the numerical solution is completed within the specified forecast period, a three-dimensional wind speed vector field covering the entire computational domain is output from the computational fluid dynamics model. The three-dimensional wind speed vector field maintains its original distribution form on the hybrid computational grid in space and corresponds to the discrete time sequence within the forecast period in time, thus forming a refined three-dimensional wind field including the bridge deck space, and serving as input data for subsequent steps to extract bridge deck wind information and calculate the deflection effect index.

[0083] S23: Extract bridge deck wind information from the refined 3D wind field and calculate the deflection effect index to generate bridge deck wind field analysis results, specifically:

[0084] After obtaining a detailed three-dimensional wind field including the bridge deck space, wind information extraction and deflection effect index calculation are performed on each characteristic location of the bridge deck to form a complete bridge deck wind field analysis result.

[0085] First, the locations of bridge deck features and their corresponding height layers are determined. Specifically, the bridge deck geometry information is read from the 3D model data of the bridge structure, and multiple bridge deck feature locations are selected along the bridge deck centerline at preset intervals. For each bridge deck feature location, its corresponding bridge deck height is read, and a vertical grid layer covering that height is selected in the refined 3D wind field as the wind information extraction height layer for that location.

[0086] Subsequently, refined wind speed and direction at key bridge deck locations are obtained. Specifically, in the refined three-dimensional wind field, for each key bridge deck location... Read the three-dimensional wind speed vector within the grid cell of its corresponding height level:

[0087] ;

[0088] in, and The horizontal velocity component is... The vertical velocity component is represented by the horizontal velocity component. Based on the horizontal velocity component, the refined wind speed at the characteristic location on the bridge deck is calculated:

[0089] ;

[0090] And calculate the refined wind direction at this location based on the direction of the horizontal velocity component:

[0091] ;

[0092] in, Indicates the location of bridge deck features The fine-grained wind speed, This indicates the refined wind direction angle at that location, and all wind direction angles are uniformly expressed under the same angular reference system.

[0093] After obtaining the refined wind direction, the corresponding background wind direction is further acquired. Specifically, for each bridge surface feature location... The horizontal coordinates of the data are read in a unified projected coordinate system and then back-mapped to the original spatial grid of the numerical weather prediction data. In the original spatial grid, the four adjacent grid points containing the back-mapped location are determined, and the wind direction value at that location is calculated using bilinear interpolation, expressed as:

[0094] ;

[0095] in, Location of bridge deck features The background wind direction, , , The wind direction values ​​at adjacent grid points in the raw spatial grid of numerical weather prediction data. and This represents the normalized relative coordinates of the bridge deck feature location within the grid cell.

[0096] After obtaining the refined wind direction and background wind direction, the deflection effect index is calculated. Specifically, for each characteristic location on the bridge deck... , will refine wind direction Corresponding background wind direction The orientation angles are uniformly expressed as those in the same angular reference frame, and the absolute value of the horizontal angle between them is calculated:

[0097] ;

[0098] in, Defined as the deflection effect index at the characteristic location of the bridge deck, it is used to quantify the degree of horizontal deflection caused by the terrain and bridge structure to the background wind direction.

[0099] Finally, refine the wind direction corresponding to all bridge surface feature locations. Fine-grained wind speed and the deflection effect index The data is collected and organized in order of bridge deck mileage to form a bridge deck wind field analysis result that includes wind direction, wind speed and deflection effect index at various characteristic locations on the bridge deck, and is output for use in subsequent steps.

[0100] S3: Receives bridge deck wind field analysis results, and simultaneously integrates real-time bridge deck condition monitoring data and real-time traffic flow data; matches wind direction and speed with a preset risk threshold matrix based on different bridge deck conditions and vehicle types to calculate the dynamic comprehensive risk level; based on the dynamic comprehensive risk level and deflection effect index, automatically generates and outputs corresponding graded early warning information and decision instructions for the bridge active traffic control system. The specific steps are as follows:

[0101] S31: Receives the bridge deck wind field analysis results transmitted from S2, and simultaneously continuously receives real-time bridge deck condition monitoring data from the bridge deck sensor network and real-time traffic flow data from the traffic monitoring system via the data interface, providing real-time data input for risk assessment, specifically:

[0102] After S2 completes and generates the bridge deck wind field analysis results, it first receives the bridge deck wind field analysis results output by S2 through the internal data channel, and parses the bridge deck wind field analysis results into structured data objects indexed by bridge deck feature locations. Each bridge deck feature location corresponds to its refined wind direction, refined wind speed, and deflection effect index.

[0103] At the same time, the system continuously accesses real-time bridge surface condition monitoring data collected by the bridge surface sensor network through a pre-configured data interface; it performs data parsing and state discrimination processing on the accessed bridge surface condition monitoring data, converts the raw signals collected by the road surface sensors into standardized bridge surface condition codes, and simultaneously extracts the friction coefficient values ​​corresponding to the bridge surface condition, so that each bridge surface feature location corresponds to a unique bridge surface condition type and its physical characteristic parameters at the current moment.

[0104] Meanwhile, real-time traffic flow data is continuously accessed through the data interface established with the traffic monitoring system, and cross-sectional analysis is performed on the traffic flow data to extract the traffic flow, average vehicle speed, and proportion of large vehicles at the corresponding cross-section of the bridge. Furthermore, the traffic flow and vehicle composition are analyzed according to preset statistical rules to determine the dominant traffic parameters that reflect the current traffic operation characteristics.

[0105] After receiving and parsing the data, the bridge wind field analysis results, real-time bridge condition monitoring data, and real-time traffic flow data are time-aligned according to a unified time benchmark. Spatial correspondences are established according to the bridge surface feature locations, so that the same bridge surface feature location has wind field parameters, bridge surface condition parameters, and traffic operation parameters at the same assessment time, thereby constructing a real-time data input set for risk assessment.

[0106] S32: For each characteristic location on the bridge deck, the wind direction and speed corresponding to that location in the bridge deck wind field analysis are matched with the current risk scenario jointly determined by real-time bridge deck condition monitoring data and real-time traffic flow data. The risk threshold matrix is ​​queried to obtain the basic risk level. Then, the basic risk level is dynamically corrected based on the deflection effect index corresponding to that location in the bridge deck wind field analysis results. Finally, the dynamic comprehensive risk level of each characteristic location is calculated, specifically:

[0107] After constructing the S31 real-time data input set, risk level calculations were performed for each characteristic location on the bridge deck. Specifically, firstly, based on the real-time bridge deck condition monitoring data corresponding to the characteristic location, the bridge deck condition codes and friction coefficients collected by road surface sensors were analyzed to determine the dominant bridge deck condition type at that location at the current moment. Simultaneously, based on the real-time traffic flow data corresponding to the characteristic location, the cross-sectional traffic flow, average vehicle speed, and proportion of large vehicles were analyzed to determine the dominant vehicle type at the current moment. The dominant bridge deck condition type and dominant vehicle type were combined to uniquely determine the risk scenario corresponding to that bridge deck characteristic location at the current moment.

[0108] After identifying the risk scenario, the refined wind speed and refined wind direction corresponding to the characteristic location are retrieved from the bridge deck wind field analysis results, and the angle between the refined wind direction and the normal to the bridge deck axis is calculated. Specifically, let the direction angle of the bridge deck axis in the horizontal plane at this characteristic location be denoted as . Then its normal direction angle is To refine the wind direction Subtracting the wind direction angle from the normal direction angle and taking its absolute value yields the refined angle between the wind direction and the normal direction of the bridge deck axis:

[0109] ;

[0110] in, Indicates the location of bridge deck features The angle of crosswind action at a given location is used as the input parameter for querying the risk threshold matrix.

[0111] Subsequently, the risk scenarios, refined wind speed, and wind direction-normal angle were analyzed. As a joint input, interval matching is performed within the corresponding dimension interval in the risk threshold matrix to output the basic risk level of the bridge deck feature location. After obtaining the basic risk level, the basic risk level is further dynamically corrected based on the deflection effect index corresponding to the bridge deck feature location in the bridge deck wind field analysis results. Specifically, the deflection effect index of this location is... Compared with the preset first angle threshold Compare; when satisfied When the basic risk level is raised by one level from the original level, it will be used as the dynamic comprehensive risk level for that characteristic location of the bridge deck; when At the same time, the basic risk level remains unchanged and is directly used as the dynamic comprehensive risk level of the characteristic location of the bridge deck.

[0112] Through the above processing, the dynamic comprehensive risk level of each characteristic location on the bridge deck is calculated.

[0113] S33: Based on the spatial distribution of the dynamic comprehensive risk level of all characteristic locations and its corresponding deflection effect index, combined with the preset early warning rule base and control strategy base, automatically generate graded early warning information and decision instructions that match the content and level for different audiences and control equipment, and output the graded early warning information and decision instructions to the designated early warning release platform and the control terminal of the bridge active traffic control system through the communication network, specifically:

[0114] After calculating the dynamic comprehensive risk level at all feature locations on the bridge deck, the distribution of the dynamic comprehensive risk level across the bridge deck space is first statistically analyzed. Specifically, the maximum value of the dynamic comprehensive risk level is extracted from all feature locations on the bridge deck:

[0115] ;

[0116] in, Location of bridge deck features The dynamic comprehensive risk level, Used to characterize the dominant risk level within the bridge deck area at the current moment.

[0117] After obtaining the dominant risk level, the dominant risk level is mapped to the corresponding warning color, warning text template and warning information release range according to the warning rule base. In addition, the distribution of risk levels at each characteristic location along the bridge deck is combined to determine the release section of the warning information, thereby generating hierarchical warning information that matches the current risk situation.

[0118] Meanwhile, based on the control strategy library, the spatial distribution characteristics of the dominant risk level, the risk level of the bridge surface feature location, and the corresponding deflection effect index are used as joint input conditions to match and determine the combination of decision instructions to be triggered; the decision instructions include at least the text content for the variable message sign, the target speed limit value for the variable speed limit sign, and the lane closing suggestion for the lane control light.

[0119] After generating graded early warning information and decision instructions, the early warning information and decision instructions are processed for protocol adaptation and formatting, and then sent to the designated early warning release platform and the control terminal of the bridge active traffic control system through the communication network to realize real-time prompting and active control of bridge traffic operation risks.

[0120] like Figure 2 As shown, a risk early warning service system is used to provide early warning services for the aforementioned method of forecasting wind on the bridge deck of a cross-river bridge, and includes the following modules:

[0121] Data acquisition and preprocessing module: used to acquire numerical weather forecast data, three-dimensional terrain data, three-dimensional model data of bridge structure and real-time traffic monitoring data in parallel for the area where the target cross-river bridge is located, and to perform spatiotemporal benchmark unification and standardization processing on the above data, as well as alignment, fusion and encapsulation under a unified spatial reference system to construct the initial forecast dataset;

[0122] Bridge Deck Wind Field Simulation and Analysis Module: Connected to the data acquisition and preprocessing module, it is used to establish a computational fluid dynamics model covering the target cross-river bridge and surrounding river canyon areas based on the forecast initial dataset. By simulating the disturbance of the background wind field by the river canyon and bridge structure, it generates bridge deck wind field analysis results including wind direction, wind speed and deflection effect index at various characteristic locations on the bridge deck.

[0123] Real-time status access module: It is used to continuously access real-time bridge surface condition monitoring data from the bridge surface sensor network and real-time traffic flow data from the traffic monitoring system through the data interface, and to perform time synchronization and validity verification of the real-time data to provide real-time input for risk assessment.

[0124] Dynamic risk assessment module: Connected to the bridge deck wind field simulation analysis module and the real-time status access module, it is used to match the wind direction and wind speed at each characteristic location of the bridge deck with the preset risk threshold matrix based on different bridge deck conditions and vehicle types, calculate the basic risk level, and dynamically correct the basic risk level by combining the deflection effect index of the corresponding characteristic location to obtain the dynamic comprehensive risk level of each characteristic location of the bridge deck.

[0125] Early warning decision generation module: connected to the dynamic risk assessment module, it is used to automatically generate graded early warning information that matches the risk level and decision instructions for the bridge active traffic control system based on the spatial distribution of the dynamic comprehensive risk level of each characteristic location on the bridge surface and its corresponding deflection effect index, by calling the preset early warning rule library and control strategy library.

[0126] Early warning release and control interface module: Connected to the early warning decision generation module, it is used to release graded early warning information to the designated early warning release platform through the communication network, and send decision instructions to the control terminal of the bridge active traffic control system to realize active intervention and risk prevention and control of bridge traffic.

[0127] This invention encompasses any substitutions, modifications, equivalent methods, and solutions made within the spirit and scope of this invention. To provide the public with a thorough understanding of this invention, specific details are described in detail in the following preferred embodiments; however, those skilled in the art will fully understand the invention even without these details. Furthermore, to avoid unnecessary misunderstanding of the essence of this invention, well-known methods, processes, procedures, components, and circuits are not described in detail.

[0128] The above description is only a preferred embodiment of the present invention. It should be noted that for those skilled in the art, several improvements and modifications can be made without departing from the principle of the present invention, and these improvements and modifications should also be considered within the scope of protection of the present invention.

Claims

1. A method for forecasting wind on the deck of a cross-river bridge, characterized in that, Includes the following steps: S1: Obtain numerical weather forecast data, 3D terrain data, 3D model data of the bridge structure, and real-time traffic monitoring data for the area where the target cross-river bridge is located, and construct the initial forecast dataset; S2: Based on the forecast initial dataset, simulate the disturbance of the background wind field by river canyons and bridge structures, and generate bridge deck wind field analysis results; the bridge deck wind field analysis results include wind direction, wind speed at each characteristic location on the bridge deck, and a deflection effect index used to quantify the degree of horizontal deflection of wind direction caused by terrain. S3: Receive the bridge surface wind field analysis results, and simultaneously access real-time bridge surface condition monitoring data and real-time traffic flow data; match the wind direction and wind speed with a preset risk threshold matrix based on different bridge surface conditions and vehicle types to calculate the dynamic comprehensive risk level; Based on the dynamic comprehensive risk level and the deflection effect index, the system automatically generates and outputs corresponding graded early warning information and decision instructions for the bridge active traffic control system.

2. The method for forecasting wind on the bridge deck of a cross-river bridge according to claim 1, characterized in that, S1 includes: S11: Parallel acquisition of numerical weather forecast data, 3D terrain data, 3D model data of bridge structure, and real-time traffic monitoring data of the target cross-river bridge area to form initial, unstandardized numerical weather forecast data, 3D terrain data, 3D model data of bridge structure, and real-time traffic monitoring data; S12: Perform spatiotemporal benchmark unification and standardization processing on the initial, unstandardized numerical weather forecast data, three-dimensional terrain data, bridge structure three-dimensional model data and real-time traffic monitoring data, and output standardized numerical weather forecast data, three-dimensional terrain data, bridge structure three-dimensional model data and real-time traffic monitoring data. S13: Align, fuse, and encapsulate the standardized numerical weather forecast data, three-dimensional terrain data, three-dimensional bridge structure model data, and real-time traffic monitoring data within a unified spatiotemporal framework to construct an initial forecast dataset containing multi-dimensional data layers with consistent internal relationships.

3. The method for forecasting wind on the bridge deck of a cross-river bridge according to claim 2, characterized in that, The numerical weather forecast data is derived from the fusion of a mesoscale weather forecast model and a regional high-resolution numerical model; the three-dimensional terrain data is acquired through airborne lidar scanning; the three-dimensional model data of the bridge structure is obtained by lightweight conversion of the bridge design BIM model; and the real-time traffic monitoring data is collected by video vehicle detectors and microwave vehicle detectors deployed at the bridgehead and on the bridge deck.

4. The method for forecasting wind on the bridge deck of a cross-river bridge according to claim 2, characterized in that, The alignment, fusion, and encapsulation within a unified spatiotemporal framework refers to: establishing a quasi-linear spatial reference system extending along the bridge deck, using the centerline of the target cross-river bridge as the spatial reference baseline; mapping the standardized numerical weather forecast data, 3D terrain data, 3D bridge structure model data, and real-time traffic monitoring data to the corresponding segments of this reference system according to their spatial locations; and finally encapsulating and generating the initial forecast dataset with a unified index, capable of quickly retrieving and extracting all related data based on bridge deck mileage location.

5. The method for forecasting wind on the bridge deck of a cross-river bridge according to claim 2, characterized in that, S2 includes: S21: Based on the initial forecast dataset, establish a computational fluid dynamics model covering the target cross-river bridge and the surrounding river canyon area. The model uses the three-dimensional terrain data and the three-dimensional model data of the bridge structure to construct the physical geometry and computational grid, and sets the numerical weather forecast data interpolation as the model inlet boundary condition, thereby generating a customized computational fluid dynamics model. S22: Run the computational fluid dynamics model to simulate the flow around, downdraft and septum effects generated when airflow passes through the space defined by the three-dimensional terrain data and the three-dimensional bridge structure model data, and solve for the three-dimensional wind field including the bridge deck space in the future specified forecast period. S23: Extract airflow information of the height layer of each characteristic location on the bridge deck from the three-dimensional wind field, directly obtain the wind direction and wind speed at that location, and simultaneously calculate the absolute value of the horizontal angle between the wind direction at that location and the background wind direction provided by the numerical weather forecast data. Define the absolute value of the horizontal angle as the deflection effect index at that location. Integrate the wind direction, wind speed and deflection effect index of all characteristic locations to finally generate a complete bridge deck wind field analysis result.

6. The method for forecasting wind on the bridge deck of a cross-river bridge according to claim 5, characterized in that, The step of setting the numerical weather forecast data interpolation as the model inlet boundary condition specifically includes: spatially interpolating the horizontal wind field and temperature field data of the upwind boundary of the target area in the numerical weather forecast data to the corresponding inlet boundary grid points of the customized high-resolution computational fluid dynamics model, and assigning the surface roughness parameter represented by the three-dimensional terrain data to the bottom boundary of the model.

7. The method for forecasting wind on the bridge deck of a cross-river bridge according to claim 5, characterized in that, The three-dimensional wind field has a grid resolution of no less than 10 meters in the horizontal direction and no less than 5 layers in the vertical direction near the bridge deck.

8. The method for forecasting wind on the bridge deck of a cross-river bridge according to claim 5, characterized in that, S3 includes: S31: Receive the bridge deck wind field analysis results transmitted by S2, and continuously access real-time bridge deck condition monitoring data from the bridge deck sensor network and real-time traffic flow data from the traffic monitoring system through the data interface to provide real-time data input for risk assessment; S32: For each characteristic location on the bridge deck, the wind direction and wind speed corresponding to it in the bridge deck wind field analysis results are matched with the current risk scenario jointly determined by the real-time bridge deck condition monitoring data and real-time traffic flow data. The risk threshold matrix is ​​queried to obtain the basic risk level. Then, the basic risk level is dynamically corrected based on the deflection effect index corresponding to the location in the bridge deck wind field analysis results. Finally, the dynamic comprehensive risk level of each characteristic location is calculated. S33: Based on the spatial distribution of the dynamic comprehensive risk level of all characteristic locations and its corresponding deflection effect index, combined with the preset early warning rule base and control strategy base, automatically generate hierarchical early warning information and decision instructions that match the content and level for different audiences and control equipment, and output the hierarchical early warning information and decision instructions to the designated early warning release platform and the control terminal of the bridge active traffic control system through the communication network.

9. A method for forecasting wind on the bridge deck of a cross-river bridge according to claim 8, characterized in that, The real-time bridge surface condition monitoring data includes dry / wet / ice condition codes and friction coefficients collected by road surface sensors, and the real-time traffic flow data includes cross-sectional traffic flow, average vehicle speed, and the proportion of large vehicles obtained by video vehicle detectors.

10. A risk early warning service system, used to provide early warning services for the forecasting method as described in any one of claims 1-9, characterized in that, include: Data acquisition and preprocessing module: used to acquire numerical weather forecast data, three-dimensional terrain data, three-dimensional model data of bridge structure and real-time traffic monitoring data in parallel for the area where the target cross-river bridge is located, and to perform spatiotemporal benchmark unification and standardization processing on the above data, as well as alignment, fusion and encapsulation under a unified spatial reference system to construct the initial forecast dataset; Bridge deck wind field simulation and analysis module: connected to the data acquisition and preprocessing module, used to establish a computational fluid dynamics model covering the target cross-river bridge and surrounding river canyon areas based on the forecast initial dataset, and generate bridge deck wind field analysis results including wind direction, wind speed and deflection effect index at various characteristic locations on the bridge deck by simulating the disturbance of the background wind field by the river canyon and bridge structure. Real-time status access module: It is used to continuously access real-time bridge surface condition monitoring data from the bridge surface sensor network and real-time traffic flow data from the traffic monitoring system through the data interface, and to perform time synchronization and validity verification of the real-time data to provide real-time input for risk assessment. Dynamic risk assessment module: connected to the bridge deck wind field simulation analysis module and the real-time status access module, used to match the wind direction and wind speed at each characteristic location of the bridge deck with a preset risk threshold matrix based on different bridge deck conditions and vehicle types, calculate the basic risk level, and dynamically correct the basic risk level by combining the deflection effect index of the corresponding characteristic location to obtain the dynamic comprehensive risk level of each characteristic location of the bridge deck. Early warning decision generation module: connected to the dynamic risk assessment module, used to automatically generate graded early warning information matching the risk level and decision instructions for the bridge active traffic control system based on the spatial distribution of the dynamic comprehensive risk level of each characteristic location on the bridge surface and its corresponding deflection effect index, by calling the preset early warning rule library and control strategy library. Early warning release and control interface module: connected to the early warning decision generation module, used to release the graded early warning information to the designated early warning release platform through the communication network, and send the decision instructions to the control terminal of the bridge active traffic control system, so as to realize active intervention and risk prevention and control of bridge traffic.

Images