Bridge structure safety early warning method based on multi-source heterogeneous data and related device

Abstract

The embodiment of the application relates to the technical field of bridge structure safety detection and data processing, and provides a bridge structure safety early warning method based on multi-source heterogeneous data and related devices, which comprises the following steps: acquiring a bridge multi-source heterogeneous data set comprising perception layer data; performing space-time fusion on the basis of the bridge multi-source heterogeneous data set to obtain a space-time fusion multi-source heterogeneous data set; performing bridge damage feature extraction processing on the perception layer data based on multi-modal data to obtain a bridge damage feature data set; performing multi-level feature fusion processing on the basis of the bridge damage feature data set and the space-time fusion multi-source heterogeneous data set to obtain a fusion damage feature data set; performing bridge damage evolution processing on the basis of the fusion damage feature data set to obtain a bridge damage evolution feature result; and performing dynamic intelligent early warning processing based on risk coupling on the basis of the bridge damage evolution feature result and the fusion damage feature data set, so that a more accurate bridge structure safety early warning result can be obtained.

Description

Technical Field

[0001] This application relates to the field of bridge structure safety detection and data processing technology, specifically to a bridge structure safety early warning method and related device based on multi-source heterogeneous data. Background Technology

[0002] With the rapid advancement of transportation infrastructure construction, the scale of long-span, high-load bridges continues to expand. During their long-term service, bridges are susceptible to structural damage and performance degradation due to multiple factors, including structural aging, repeated traffic loads, and erosion from extreme weather conditions. Bridge structural safety management has become a core task in transportation operation and maintenance, and accurate and efficient safety early warning technology is crucial for ensuring safe bridge operation. Currently, traditional bridge safety monitoring and early warning technologies mostly have significant shortcomings. For example, relying solely on single-point monitoring data at the sensing layer for early warning analysis leads to incomplete damage feature extraction and insufficient risk assessment basis, making it difficult to fully reflect the actual service conditions of bridges. Consequently, the accuracy of safety early warning results is low, with a high rate of false alarms and missed alarms, failing to meet the actual needs of modern bridge long-term safety management. Therefore, how to achieve more accurate bridge structural safety early warning has become an urgent problem to be solved. Summary of the Invention

[0003] This application provides a bridge structure safety early warning method and related device based on multi-source heterogeneous data, which can achieve more accurate bridge structure safety early warning.

[0004] The first aspect of this application provides a bridge structure safety early warning method based on multi-source heterogeneous data. This method includes: Acquire a multi-source heterogeneous data set of bridges; the multi-source heterogeneous data set of bridges includes data from the perception layer, design layer, operation and maintenance layer, traffic layer, and geographic and meteorological layer; Spatiotemporal fusion processing is performed on each bridge multi-source heterogeneous data in the bridge multi-source heterogeneous dataset to obtain a spatiotemporal fused multi-source heterogeneous dataset. The bridge damage feature extraction process based on multimodal data is performed on the perception layer data to obtain a bridge damage feature data set; Multi-level feature fusion processing is performed on the bridge damage feature dataset and the spatiotemporally fused multi-source heterogeneous dataset to obtain the fused damage feature dataset. Bridge damage evolution characteristics are obtained by performing bridge damage evolution processing based on the fused damage feature dataset; Based on the bridge damage evolution characteristics and the fusion damage characteristic data set, dynamic intelligent early warning processing based on risk coupling is carried out to obtain bridge structural safety early warning results.

[0005] A second aspect of this application provides a bridge structure safety early warning device based on multi-source heterogeneous data, the bridge structure safety early warning device based on multi-source heterogeneous data comprising: The acquisition unit is used to acquire a multi-source heterogeneous data set of bridges; the multi-source heterogeneous data set of bridges includes data from the perception layer, design layer, operation and maintenance layer, traffic layer, and geographic and meteorological layer. The first processing unit is used to perform spatiotemporal fusion processing on each bridge multi-source heterogeneous data in the bridge multi-source heterogeneous data set to obtain a spatiotemporal fused multi-source heterogeneous data set. The second processing unit is used to perform bridge damage feature extraction processing on the perception layer data based on multimodal data to obtain a bridge damage feature data set. The third processing unit is used to perform multi-level feature fusion processing based on the bridge damage feature data set and the spatiotemporal fusion multi-source heterogeneous data set to obtain the fused damage feature data set. The fourth processing unit is used to perform bridge damage evolution processing based on the fused damage feature data set to obtain bridge damage evolution feature results. The fifth processing unit is used to perform dynamic intelligent early warning processing based on risk coupling according to the bridge damage evolution characteristics and the fused damage characteristic data set, so as to obtain the bridge structural safety early warning results.

[0006] A third aspect of this application provides a terminal including a processor, an input device, an output device, and a memory, wherein the processor, input device, output device, and memory are interconnected, wherein the memory is used to store a computer program, the computer program including program instructions, and the processor is configured to invoke the program instructions to execute the step instructions as described in the first aspect of this application.

[0007] A fourth aspect of this application provides a computer-readable storage medium storing a computer program for electronic data interchange, wherein the computer program causes a computer to perform some or all of the steps described in the first aspect of this application.

[0008] A fifth aspect of this application provides a computer program product, wherein the computer program product includes a non-transitory computer-readable storage medium storing a computer program operable to cause a computer to perform some or all of the steps described in the first aspect of this application. The computer program product may be a software installation package.

[0009] Implementing the embodiments of this application has the following beneficial effects: By acquiring a multi-source heterogeneous bridge data set including perception layer data, design layer data, operation and maintenance layer data, traffic layer data, and geographic and meteorological layer data, spatiotemporal fusion processing can be performed on each bridge multi-source heterogeneous data set to obtain a spatiotemporally fused multi-source heterogeneous data set. Furthermore, bridge damage feature extraction processing based on multimodal data can be performed on the perception layer data to obtain a bridge damage feature data set. Multi-level feature fusion processing can then be performed on the bridge damage feature data set and the spatiotemporally fused multi-source heterogeneous data set to obtain a fused damage feature data set. Bridge damage evolution processing can then be performed based on the fused damage feature data set to obtain bridge damage evolution feature results. Finally, dynamic intelligent early warning processing based on risk coupling can be performed based on the bridge damage evolution feature results and the fused damage feature data set to obtain more accurate bridge structural safety early warning results, which is beneficial for achieving a more accurate bridge structural safety early warning process. Attached Figure Description

[0010] To more clearly illustrate the technical solutions in the embodiments of this application 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 some embodiments of this application. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0011] Figure 1 This application provides a schematic diagram of a bridge structure safety early warning method based on multi-source heterogeneous data. Figure 2 This application provides a flowchart illustrating a bridge structure safety early warning method based on multi-source heterogeneous data. Figure 3 This is a schematic diagram of the structure of a terminal provided in an embodiment of this application; Figure 4 This application provides a schematic diagram of a bridge structure safety early warning device based on multi-source heterogeneous data. Detailed Implementation

[0012] The technical solutions of the embodiments of this application will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of this application, and not all embodiments. Based on the embodiments of this application, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of this application.

[0013] The terms "first," "second," etc., in the specification, claims, and accompanying drawings of this application are used to distinguish different objects, not to describe a specific order. Furthermore, the terms "comprising" and "having," and any variations thereof, are intended to cover non-exclusive inclusion. For example, a process, method, system, product, or apparatus that includes a series of steps or units is not limited to the listed steps or units, but may optionally include steps or units not listed, or may optionally include other steps or units inherent to these processes, methods, products, or apparatuses.

[0014] In this application, the reference to "embodiment" means that a specific feature, structure, or characteristic described in connection with an embodiment may be included in at least one embodiment of this application. The appearance of this phrase in various places throughout the specification does not necessarily refer to the same embodiment, nor is it a mutually exclusive, independent, or alternative embodiment. It will be explicitly and implicitly understood by those skilled in the art that the embodiments described in this application can be combined with other embodiments.

[0015] To better understand the bridge structure safety early warning method based on multi-source heterogeneous data provided in this application, a brief introduction to existing bridge structure safety early warning technologies is given below. Currently, conventional bridge structure safety early warning mainly relies on local single-item monitoring data such as crack width, structural displacement, and strain, comparing and alarming by setting fixed thresholds. This method relies on only a single parameter, reflecting only a local instantaneous state, lacking data reliability, being susceptible to noise interference, and failing to comprehensively reflect the overall structural damage status. Furthermore, the use of static threshold judgments does not consider the gradual characteristics of structural aging and damage accumulation, resulting in delayed identification of slowly developing damage, untimely warnings, and a high risk of false alarms and missed alarms. In addition, existing early warning systems mostly make single-dimensional judgments based on the structure's own parameters, failing to comprehensively consider the coupled influence of external factors such as weather and traffic loads on damage development, thus failing to accurately quantify the overall safety risk. This leads to one-sided and inaccurate early warning results, making it difficult to meet the actual needs of refined safety management.

[0016] In order to solve at least one of the above problems, this application provides a bridge structure safety early warning method based on multi-source heterogeneous data. It can achieve deep fusion of multi-dimensional heterogeneous data, and can customize the sensitivity coefficient in combination with the characteristics of the target bridge. It can also modify the damage evolution rate by coupling with meteorological risk, so as to achieve dynamic and accurate bridge structure safety early warning and improve the intelligent level of bridge safety management.

[0017] Please see Figure 1 , Figure 1 A schematic diagram of a bridge structure safety early warning system based on multi-source heterogeneous data is shown. Figure 1As shown, a bridge structural safety early warning system based on multi-source heterogeneous data may include a data acquisition module, a spatiotemporal fusion module, a bridge damage feature extraction module, a multi-level feature fusion module, a bridge damage evolution module, and a dynamic intelligent early warning module. The system comprises several modules: a data acquisition module for acquiring multi-source heterogeneous bridge data sets, providing comprehensive multi-source data support for subsequent bridge structural damage assessment and safety early warning; a spatiotemporal fusion module for processing the spatiotemporal fusion data of various bridge multi-source heterogeneous data collected by the data acquisition module, eliminating data biases, unifying spatiotemporal benchmarks, and obtaining a spatiotemporally fused multi-source heterogeneous data set; a bridge damage feature extraction module for extracting multimodal damage features from the perception layer data, filtering and extracting effective structural damage features, and obtaining a bridge damage feature data set; a multi-level feature fusion module for combining the bridge damage feature data set with the spatiotemporally fused multi-source heterogeneous data set, performing multi-level feature association fusion, eliminating redundant features, integrating core feature information, and obtaining a fused damage feature data set; a bridge damage evolution module for analyzing damage development trends and deducing damage evolution patterns based on the fused damage feature data set, and obtaining bridge damage evolution feature results; and a dynamic intelligent early warning module for combining the bridge damage evolution feature results with the fused damage feature data set to conduct multi-dimensional risk coupling calculations and dynamic assessments, ultimately obtaining accurate and reliable bridge structural safety early warning results.

[0018] Please see Figure 2 , Figure 2 This application provides a flowchart illustrating a bridge structure safety early warning method based on multi-source heterogeneous data. Figure 2 As shown, the bridge structural safety early warning method based on multi-source heterogeneous data includes: S10: Obtain a multi-source heterogeneous data set of bridges; the multi-source heterogeneous data set of bridges includes data from the perception layer, design layer, operation and maintenance layer, traffic layer, and geographic and meteorological layer.

[0019] Among them, the multi-source heterogeneous data set of bridges can refer to a comprehensive data set that covers information from multiple sources, including the operation, design, maintenance, and environment of bridges throughout their entire life cycle, and where various types of information differ in data type, format, collection frequency, and representation dimensions.

[0020] The sensing layer data refers to real-time data that reflects the current physical state and operating conditions of the bridge, directly acquired through various sensors, monitoring equipment, and data acquisition devices on the bridge. This sensing layer data can intuitively demonstrate the bridge's damage and structural performance changes, and can be considered the core data for sensing the bridge's safety status. Optionally, sensing devices such as crack width sensors, displacement monitoring sensors, vibration acceleration sensors, and high-definition cameras can be deployed in key parts of the bridge, such as the main beams, piers, and supports, to collect real-time data on crack width changes, structural displacement, vibration frequency, and surface images of the bridge deck and components to obtain the aforementioned sensing layer data. This application does not impose any limitations on this.

[0021] Design layer data refers to the fundamental parameters and technical information generated during the planning and design phases of a bridge. This design layer data can be understood as the inherent attributes of the bridge structure, determining core indicators such as the bridge's design load-bearing capacity, structural form, and material properties. Design layer data can be seen as an important reference for judging the actual performance degradation of a bridge. Optionally, the aforementioned design layer data can be obtained by retrieving bridge design drawings, design specifications, material testing reports, structural calculation sheets, and other relevant documents. This allows for the acquisition of data such as the bridge's design load-bearing capacity, structural parameters (e.g., main beam cross-sectional dimensions, pier height, bearing type), material performance parameters (e.g., concrete strength grade, steel reinforcement yield strength), and design service life.

[0022] Operation and maintenance (O&M) layer data refers to historical records and actual working condition data generated during the long-term service, operation, and maintenance of bridges. This O&M layer data reflects the bridge's performance changes over time, damage evolution process, and the effectiveness of maintenance measures, and can be regarded as the core data for analyzing the long-term safety status of bridges. Optionally, O&M layer data can be obtained by collecting data such as daily inspection records, periodic inspection reports, maintenance and repair archives, and historical records of bridge defects, including inspection time, inspection results (such as whether damage was found and the preliminary type of damage), quantitative data of damage from periodic inspections, maintenance and repair time, type of maintenance measures, and repair effects.

[0023] Traffic layer data refers to data reflecting the actual traffic load on a bridge. This data reflects information such as the bridge's load conditions, traffic density, vehicle type, and traffic frequency, and can be considered an important reference for assessing the bridge's stress state and the dynamics of damage evolution. Optionally, traffic layer data can be obtained from channels such as traffic flow monitoring systems, license plate recognition systems, and toll station passage records to acquire data on the bridge's traffic flow, traffic density, vehicle type classification, traffic frequency, and actual load distribution for the corresponding time period.

[0024] Geographic meteorological layer data refers to the natural environmental data of the bridge's geographical location, including meteorological elements such as temperature, humidity, rainfall, wind speed, and wind direction, as well as geographical conditions such as topography, geology, and hydrology. This geographic meteorological layer data can be considered as external environmental factors affecting the bridge's structural durability and the rate of damage evolution. Optionally, data such as daily average temperature, daily maximum temperature, daily minimum temperature, relative humidity, rainfall, wind speed, wind direction, geological conditions (such as soil type and foundation bearing capacity), and hydrological conditions (such as river water level and flow velocity) of the area where the bridge is located can be obtained through meteorological stations, hydrological monitoring stations, and geographic information systems around the bridge to obtain the aforementioned geographic meteorological layer data.

[0025] S20: Perform spatiotemporal fusion processing on each bridge multi-source heterogeneous data in the bridge multi-source heterogeneous data set to obtain a spatiotemporal fused multi-source heterogeneous data set.

[0026] Spatiotemporal fusion processing refers to the process of integrating, aligning, and correcting heterogeneous bridge data from different sources and with different acquisition dimensions, based on a unified temporal and spatial reference. It can be understood that spatiotemporal fusion processing can eliminate temporal misalignment and spatial deviation between different data, forming a unified and interconnected whole, thus providing a standardized data foundation for subsequent feature extraction and fusion analysis.

[0027] A spatiotemporally fused multi-source heterogeneous dataset refers to a bridge multi-source heterogeneous dataset that has undergone spatiotemporal fusion processing to achieve time synchronization and spatial uniformity. It can be understood that a clear correlation is established between various types of data in the spatiotemporally fused multi-source heterogeneous dataset, enabling a comprehensive and systematic reflection of the overall condition of the bridge and the impact of the external environment.

[0028] Specifically, the spatial coordinates of the bridge can be used as a unified spatial reference, and a unified timestamp as a unified time reference, to align, correct, and integrate various types of data. Optionally, for the perception layer data, a unified format timestamp can be added to each data point according to the chronological order of collection time, while simultaneously marking the corresponding spatial location of the bridge (such as the specific location at the middle 1 / 3 of the main beam span, the top of the pier, etc.). For the design layer, operation and maintenance layer, traffic layer, and geographic and meteorological layer data, corresponding timestamps and spatial location information can be added. Among them, the timestamp for the design layer data can be the design completion time, and the spatial location can be the entire bridge; the timestamp for the operation and maintenance layer data can be the inspection / testing / repair time, and the spatial location can be the corresponding damage location or the entire bridge; the timestamp for the traffic layer data can be the monitoring period, and the spatial location can be each lane on the bridge deck; the timestamp for the geographic and meteorological layer data can be the monitoring time, and the spatial location can be the area within 1 kilometer of the bridge. This application does not impose any restrictions on this.

[0029] For example, the various types of data in the final spatiotemporal fusion multi-source heterogeneous data set can be arranged in an orderly manner according to the temporal and spatial correlation. For instance, at 10:00:00 on May 10, 2024, the crack width at the middle 1 / 3 of the main beam span was 0.25 mm, the traffic flow at the corresponding time was 800 vehicles / hour, the rainfall at the corresponding time was 5 mm / hour, and the average daily temperature of the corresponding area was 20℃.

[0030] Optionally, a hierarchical spatiotemporal fusion approach can be adopted during spatiotemporal fusion processing to improve the fusion accuracy and correlation effectiveness of multi-source heterogeneous data. Specifically, hierarchical alignment along the time dimension can be performed, such as dividing data into different time levels according to real-time acquisition, periodic acquisition, and static archive data. For example, high-frequency acquisition data from the perception layer and transportation layer can be aligned using minute-level timestamps, mid-frequency acquisition data from the geographic and meteorological layer can be matched using hour-level timestamps, and low-frequency updated data from the design layer and operation and maintenance layer can be correlated using date-level timestamps. This ensures that data from different acquisition frequencies form a corresponding relationship within a unified time system, thereby guaranteeing the consistency of data from different frequencies in the time dimension.

[0031] Furthermore, hierarchical matching can be performed based on spatial dimensions. For example, different spatial levels can be defined according to the overall bridge structure, key components, and monitoring points. Global geographic and meteorological data and traffic data can be matched with the overall bridge structure; design and maintenance data corresponding to key components can be matched with key parts such as main beams, piers, and bearings; and perception data collected from monitoring points can be matched with specific sensor installation locations. This ensures that data from different coverage areas form a corresponding relationship within a unified spatial system. Through this hierarchical spatiotemporal fusion processing, temporal deviations and spatial misalignments between multi-source heterogeneous data can be effectively eliminated, which is beneficial for improving the accuracy and reliability of subsequent feature extraction and multi-level feature fusion.

[0032] Optionally, for data with outliers or missing values, interpolation, mean averaging, or elimination methods can be used for processing. For example, if there are missing values ​​in the displacement data of a certain sensing layer, linear interpolation can be used to supplement them using displacement data from adjacent time periods; if there are outliers in the rainfall data of a certain geographic meteorological layer (far higher than the historical data for the same period), it can be corrected by combining data from surrounding meteorological stations.

[0033] S30: Perform bridge damage feature extraction processing based on multimodal data on the perception layer data to obtain a bridge damage feature data set.

[0034] Bridge damage feature extraction based on multimodal data refers to the process of using feature extraction techniques to identify and extract core feature information that can characterize the type, degree, and location of bridge damage for different types of multimodal data (such as image data, vibration data, displacement data, etc.) in the perception layer.

[0035] The bridge damage feature dataset can refer to a set of feature data that comprehensively reflects the bridge damage situation, obtained by extracting features from the multimodal data of the sensing layer. This bridge damage feature dataset may include quantitative and qualitative features corresponding to various types of damage; this application does not impose any limitations on this.

[0036] Optionally, when performing bridge damage feature extraction processing based on multimodal data on the perception layer data, an initial bridge image set acquired by the acquisition device can be obtained. This initial set can be combined with the acquisition device status data set to perform acquisition device error correction processing, eliminating errors such as device vibration blurring and positional offset, to obtain a reference bridge image set. Environmental interference correction processing can be performed on the reference bridge image set based on the environmental status data set, progressively eliminating environmental interference such as traffic obstruction, heavy pollution blurring, and strong reflection overexposure, to obtain a target bridge image set that accurately reflects the true state of the bridge. Furthermore, a multi-algorithm fusion strategy can be used to extract damage features from the target bridge image set. Edge detection algorithms can identify grayscale abrupt damage such as linear cracks, and semantic segmentation algorithms can achieve pixel-level segmentation of planar damage such as spalling and exposed reinforcement. Geometric parameters and damage levels can then be calculated for the identified damage types. Simultaneously, bridge structural response data can be combined to supplement and verify the damage features. Finally, information such as damage location, size, shape, and level can be integrated to obtain a structured bridge damage feature data set.

[0037] Optionally, the above-mentioned multi-algorithm fusion strategy can specifically adopt a combination of traditional image processing algorithms and / or lightweight deep learning models. For example, edge detection can use the Canny algorithm, semantic segmentation can use a U-shaped network lightweight model, and model training can be based on a bridge damage annotation dataset (including typical damage images such as cracks, spalling, and exposed rebar, and corresponding quantization parameters) and put into use after iterative training. This application does not impose any restrictions on this.

[0038] S40: Perform multi-level feature fusion processing based on the bridge damage feature data set and the spatiotemporal fusion multi-source heterogeneous data set to obtain a fused damage feature data set.

[0039] Multi-level feature fusion processing refers to the process of gradually integrating bridge damage feature data with various related data from spatiotemporally fused multi-source heterogeneous data, following a hierarchical logic of basic engineering fusion and risk factor fusion. By achieving deep correlation between data from different dimensions through hierarchical fusion (i.e., multi-level feature fusion), it is possible to comprehensively integrate information on bridge damage itself, design attributes, operation and maintenance history, traffic load, environmental impact, and other aspects, forming more comprehensive damage feature data.

[0040] A fused damage feature dataset refers to a comprehensive set of feature data that integrates bridge damage characteristics, design attributes, operation and maintenance history, traffic load, and geographical and meteorological influences after multi-level feature fusion processing. This fused damage feature dataset can comprehensively and systematically reflect the actual safety status of bridges, providing complete data support for subsequent damage evolution analysis and early warning.

[0041] S50: Perform bridge damage evolution processing based on the fused damage feature data set to obtain bridge damage evolution feature results.

[0042] Bridge damage evolution processing refers to the process of analyzing the current damage state and rate of damage development of a bridge based on a fusion of damage feature data sets, and incorporating external environmental influencing factors to predict the trend of bridge damage over time. Bridge damage evolution processing enables the prediction of dynamic changes in bridge damage and is a crucial link connecting the current state of a bridge with future safety risks.

[0043] Bridge damage evolution characteristics refer to the comprehensive characteristic results obtained after bridge damage evolution processing, which reflect the current damage state, damage development rate, and future evolution trend of the bridge. These bridge damage evolution characteristics may include the quantitative status of the damage, the dynamic evolution rate, and trend prediction information.

[0044] S60: Based on the bridge damage evolution characteristics and the fused damage characteristic data set, perform dynamic intelligent early warning processing based on risk coupling to obtain bridge structural safety early warning results.

[0045] Risk-coupled dynamic intelligent early warning processing refers to a process that combines the results of bridge damage evolution characteristics with core data from a fusion damage feature dataset. It comprehensively considers multiple dimensions of factors, including the current degree of bridge damage, dynamic evolution trends, and meteorological risks, to conduct multi-dimensional weighted risk coupling calculations and dynamically generate early warning information. This risk-coupled dynamic intelligent early warning processing enables accurate and dynamic assessment of bridge safety risks.

[0046] Bridge structural safety early warning results refer to comprehensive early warning results obtained after dynamic intelligent early warning processing based on risk coupling, which reflect the bridge's safety risk level, early warning information, and disposal suggestions. These results can be considered the core output guiding bridge safety management and operation and maintenance decisions.

[0047] In this embodiment, by acquiring a multi-source heterogeneous bridge data set including perception layer data, design layer data, operation and maintenance layer data, traffic layer data, and geographic and meteorological layer data, spatiotemporal fusion processing can be performed on each bridge multi-source heterogeneous data set to obtain a spatiotemporal fused multi-source heterogeneous data set. Furthermore, bridge damage feature extraction processing based on multimodal data can be performed on the perception layer data to obtain a bridge damage feature data set. Multi-level feature fusion processing is then performed on the bridge damage feature data set and the spatiotemporal fused multi-source heterogeneous data set to obtain a fused damage feature data set. Bridge damage evolution processing can then be performed based on the fused damage feature data set to obtain bridge damage evolution feature results. Finally, dynamic intelligent early warning processing based on risk coupling can be performed based on the bridge damage evolution feature results and the fused damage feature data set to obtain more accurate bridge structural safety early warning results, which is beneficial for achieving a more accurate bridge structural safety early warning process.

[0048] In one possible implementation, during multi-level feature fusion processing, the associated features of the design layer, operation and maintenance layer, transportation layer, and geographic and meteorological layer can be extracted from the spatiotemporally fused multi-source heterogeneous data set to form feature data sets corresponding to each level, thereby achieving accurate decomposition and extraction of multi-dimensional engineering and environmental features. Further, basic engineering fusion is performed, fusing the bridge damage feature data set with three types of feature data sets: structural mechanics at the design layer, historical damage evolution at the operation and maintenance layer, and load at the transportation layer, to endow the bridge damage features with basic engineering attributes, resulting in a bridge damage feature data set with these attributes. Then, risk trigger fusion is carried out based on this data set, further integrating the risk trigger feature data set at the geographic and meteorological layer and the historical damage evolution feature data set at the operation and maintenance layer, achieving a deep correlation between bridge damage features and external environmental risk triggers, thus obtaining a fused damage feature data set integrating bridge damage itself, basic engineering attributes, and external risk triggers. Specifically, a method for obtaining a fused damage feature data set by performing multi-level feature fusion processing based on the bridge damage feature data set and the spatiotemporally fused multi-source heterogeneous data set may include: A1. Extract the design layer-related features from the spatiotemporal fusion multi-source heterogeneous data set to obtain the design layer structural mechanical feature data set; A2. Extract the features associated with the operation and maintenance layer from the spatiotemporal fusion multi-source heterogeneous data set to obtain the historical damage evolution feature data set of the operation and maintenance layer. A3. Extract traffic layer-related features from the spatiotemporal fusion multi-source heterogeneous data set to obtain a traffic layer load feature data set; A4. Extract features associated with the geographic meteorological layer from the spatiotemporal fusion multi-source heterogeneous data set to obtain a geographic meteorological layer risk inducement feature data set. A5. Perform basic engineering fusion processing on the bridge damage feature data set, the design layer structural mechanics feature data set, the operation and maintenance layer historical damage evolution feature data set, and the traffic layer load feature data set to obtain a bridge damage feature data set with engineering basic attributes. A6. Perform risk fusion processing on the bridge damage feature data set with engineering basic attributes, the risk induced feature data set of the geographic and meteorological layer, and the historical damage evolution feature data set of the operation and maintenance layer to obtain a fused damage feature data set.

[0049] The design-level structural mechanics characteristic data set refers to a collection of characteristic data extracted from a spatiotemporally fused multi-source heterogeneous data set, reflecting the structural mechanics properties of the bridge during the design phase. This design-level structural mechanics characteristic data set may include structural parameters, material mechanical performance parameters, design bearing capacity parameters, and structural stress calculation parameters for the bridge design, and can be regarded as the core data set reflecting the inherent mechanical characteristics of the bridge structure.

[0050] Optionally, structural mechanics characteristic parameters with high correlation to bridge damage characteristics can be extracted from the design layer data of the spatiotemporally fused multi-source heterogeneous dataset, based on the needs of bridge structural analysis, to form a design layer structural mechanics characteristic dataset. For example, characteristic data such as the cross-sectional dimensions of the bridge main girder, the design strength grade of concrete, the reinforcement ratio, the design stiffness of the piers, the overall design bearing capacity of the bridge, and the design shear strength of the supports can be extracted; for a reinforced concrete box girder bridge, data such as the main girder cross-sectional height of 2.5 meters, the web thickness of 0.25 meters, the design strength of C50 concrete, and the highway-I grade design load can be extracted.

[0051] The historical damage evolution characteristic dataset at the maintenance layer refers to a collection of characteristic data extracted from a spatiotemporally fused multi-source heterogeneous dataset, reflecting the generation, development, and evolution patterns of historical damage during bridge service. This dataset may include characteristic data such as the type, location, extent, and development rate of historical bridge damage, as well as corresponding maintenance measures and their effects. It can be considered an important historical basis for analyzing the development trend of bridge damage.

[0052] Optionally, relevant feature data such as damage monitoring records and maintenance archives of bridges over the years can be extracted from the operation and maintenance layer data of the spatiotemporally fused multi-source heterogeneous data set. This data can be structured according to damage type and monitoring time to form a historical damage evolution feature data set for the operation and maintenance layer. For example, historical development data of cracks in the main beam of a bridge over the past three years can be extracted. This data could include the first detection of a crack width of 0.1 mm in 2020, its expansion to 0.15 mm in 2021, and the absence of significant expansion after sealing treatment in 2022, as well as annual changes in pier displacement and the degree of bearing corrosion. Corresponding maintenance measures and their effects can also be included; this application does not impose any restrictions on this.

[0053] The traffic layer load characteristic dataset refers to a dataset extracted from a spatiotemporally fused multi-source heterogeneous dataset that reflects the actual traffic load characteristics borne by the bridge. This traffic layer load characteristic dataset may include characteristic data such as traffic flow density, vehicle type distribution, peak load, load frequency, and spatiotemporal distribution patterns of the load. It is the core dataset reflecting the actual service load conditions of the bridge.

[0054] Optionally, characteristic data reflecting the actual load on bridges can be extracted from the traffic layer data of a spatiotemporally fused multi-source heterogeneous dataset. Statistical analysis can then be performed in conjunction with the time dimension to form a traffic layer load characteristic data set. For example, data such as the average daily traffic volume, peak-hour traffic density, the proportion and frequency of large trucks and overloaded vehicles, load distribution characteristics of different lanes, and monthly average peak load can be extracted for a specific bridge. For example, for a main urban road bridge, characteristic data such as an average daily traffic volume of 12,000 vehicles, a large truck proportion of 15%, and a morning peak traffic density of 800 vehicles per hour can be extracted.

[0055] The geographic meteorological layer risk inducing factor characteristic dataset refers to a dataset extracted from a spatiotemporally fused multi-source heterogeneous dataset that reflects the characteristics of environmental factors in the natural environment in which a bridge is located that may induce or accelerate bridge damage. This geographic meteorological layer risk inducing factor characteristic dataset may include the variation characteristics of meteorological elements such as rainfall, wind speed, temperature, and humidity, as well as the basic characteristics of geographic elements such as topography and geology. It can be regarded as a key dataset for analyzing the impact of environmental factors on bridge damage.

[0056] Optionally, environmental factor characteristic data affecting bridge structural damage can be extracted from spatiotemporally fused multi-source heterogeneous geographic meteorological layer data. This data can be categorized and organized by meteorological and geographic elements to form a geographic meteorological layer risk factor characteristic data set. For example, meteorological characteristic data such as monthly average rainfall, maximum wind speed, extreme temperature, and relative humidity of the area where the bridge is located can be extracted, as well as geographic characteristic data such as soil type, foundation bearing capacity, and river water level changes of the bridge foundation. For example, for coastal bridges, characteristic data such as annual average rainfall of 1500 mm, maximum winter wind speed of 25 m / s, and annual average relative humidity of 75% can be extracted.

[0057] Basic engineering fusion processing refers to the process of associating, integrating, and analyzing bridge damage characteristic data sets with characteristic data sets from the design, operation and maintenance, and traffic layers. This basic engineering fusion processing process can deeply integrate the current damage characteristics of the bridge with its inherent design attributes, historical damage evolution, and actual load conditions, giving the bridge damage characteristics fundamental attributes at the engineering level, making the representation of damage characteristics more consistent with the actual engineering context of the bridge.

[0058] A bridge damage feature dataset with engineering-based attributes refers to a comprehensive damage feature dataset obtained after basic engineering fusion processing. This dataset integrates current bridge damage characteristics, structural mechanical characteristics at the design level, historical damage evolution characteristics at the operation and maintenance level, and traffic load characteristics. In essence, the damage features in this dataset are no longer singular state representations, but rather multi-dimensional damage features incorporating the engineering context, thus providing a more comprehensive reflection of the engineering causes of bridge damage.

[0059] Specifically, the aforementioned basic engineering fusion processing can be achieved using feature association matching. For example, each damage feature in the bridge damage feature dataset can be spatiotemporally associated and fused with corresponding feature data from the design layer structural mechanics feature dataset, the maintenance layer historical damage evolution feature dataset, and the traffic layer load feature dataset, thereby obtaining a bridge damage feature dataset with basic engineering attributes. For instance, for the damage feature of a 0.3mm wide crack at the mid-span of the bridge main girder, it can be associated with the design parameters of the main girder section and the design strength of C50 concrete in the design layer, with the historical expansion data and past maintenance measures of the crack at that location in the maintenance layer, and with the peak load at the mid-span location and the frequency of large truck traffic in the traffic layer. This allows the crack damage feature to possess corresponding engineering design, historical evolution, and actual load attributes, thus achieving the aforementioned basic engineering fusion.

[0060] Optionally, the process of fusing the bridge damage feature data set with the design layer structural mechanics feature data set, the operation and maintenance layer historical damage evolution feature data set, and the traffic layer load feature data set to obtain a bridge damage feature data set with engineering foundation attributes can be seen in the following formula: in, A dataset representing bridge damage characteristics with basic engineering attributes; This represents the original quantized value of the bridge damage features extracted from the bridge damage feature dataset; This represents the influence coefficient of structural mechanical characteristics extracted from the set of structural mechanical characteristics data at the design level. This influence coefficient is obtained through inverse quantization; that is, the better the design parameters, the better the influence coefficient. The smaller the value, the weaker the amplification effect on damage; the worse the design parameters, the worse the damage amplification effect. The larger the value, the stronger the amplification effect on the damage; This represents the influence coefficient of historical damage evolution characteristics extracted from the historical damage evolution characteristic data set of the operations and maintenance layer. This influence coefficient is obtained through positive vectorization; the better the operations and maintenance status, the higher the influence coefficient. The smaller the value, the weaker the amplification effect on damage; the worse the operational status, the weaker the amplification effect. The larger the value, the stronger the amplification effect on the damage; This represents the traffic layer load characteristic influence coefficient extracted from the traffic layer load characteristic data set. This coefficient is obtained through positive vectorization; the better the load conditions, the better the influence coefficient. The smaller the value, the weaker the amplification effect on damage; the more unfavorable the load conditions, the weaker the amplification effect. The larger the value, the stronger the amplification effect on the damage; , and It can represent the engineering integration weight coefficient, which can be adapted and set according to the bridge structure type, damage type, and engineering operation and maintenance needs, and always satisfies α+β+γ=1 to ensure the rationality of weight allocation.

[0061] Optionally, before substituting into the above formula for calculation, All have undergone dimensional normalization. For example, for different physical quantities such as crack width, displacement, and corrosion area, dimensional normalization has been performed. The 0~1 dimension normalization process can be completed according to the grading and evaluation standards for bridge component damage in the bridge maintenance specifications. Alternatively, the parameters can be converted to the 0~1 range using the above quantization method to ensure that all parameters involved in the calculation have the same dimensions and comparable values, thus avoiding the problem of dimension confusion in subsequent weighted coupling calculations.

[0062] It should be noted that design parameters directly related to the current damage type can be extracted from the structural mechanical characteristic data set of the design layer. For example, concrete design strength and main beam section stiffness can be extracted from crack damage, and pier design lateral stiffness and overall bridge design bearing capacity can be extracted from displacement damage. Furthermore, the inverse quantization method can be used to convert these parameters into 0~1 design layer structural mechanical characteristic influence coefficients. Understandably, the better the design parameters (i.e., the stronger the constraint on damage), the better. The smaller the value, the worse the design parameters (i.e., the weaker the constraint on damage). The larger the value, the better. Historical damage evolution characteristic data sets at the maintenance layer can be used to extract parameters such as the annual crack propagation rate, the stabilization time of cracks after maintenance, the annual displacement variation, and the displacement rebound after maintenance. These parameters can then be converted into a 0-1 influence coefficient representing the historical damage evolution characteristics of the maintenance layer. The slower the rate of damage development, the better the maintenance effect. The smaller the value, the faster the damage history develops / the worse the maintenance effect. The larger the value, the more accurate it is to extract actual load parameters directly related to the current damage location from the traffic layer load characteristic data set. For example, for crack damage, the actual peak load at the mid-span location and the frequency of large truck passage can be extracted; for displacement damage, the actual eccentric load coefficient and traffic density at the pier can be extracted. These parameters are then converted into traffic layer load characteristic influence coefficients of 0 to 1 using a positive vectorization method. The closer the actual load is to the bridge's design load, the better. The larger the value, the smaller the actual load is compared to the bridge's design load. The smaller.

[0063] Risk factor fusion processing refers to the process of deeply fusing bridge damage characteristic data sets with engineering-based attributes, risk factor characteristic data sets from the geographic and meteorological layers, and historical damage evolution characteristic data sets from the operation and maintenance layer. This risk factor fusion processing can uncover the correlation between geographic and meteorological environmental risk factors and bridge damage, clarify the influence patterns of environmental factors on bridge damage development, and ensure that damage characteristics possess both engineering-based attributes and environmental risk factor correlation attributes.

[0064] Specifically, the aforementioned risk factor fusion processing can be achieved using feature correlation analysis. For example, various damage features in the bridge damage feature dataset with engineering foundation attributes can be correlated with meteorological and geographical feature data in the geographical and meteorological risk factor feature dataset. Simultaneously, the correlation between damage development and environmental factors in the historical damage evolution feature dataset of the operation and maintenance layer can be combined to assign environmental risk factor attributes to the bridge damage features, thus obtaining a fused damage feature dataset. For instance, regarding the aforementioned mid-span crack damage in the main beam, the correlation between its crack propagation rate and rainfall and air humidity can be analyzed. If the data shows that the crack propagation rate increases by 50% when humidity is above 80%, then this humidity risk factor feature can be deeply fused with the crack damage feature. Furthermore, by combining the development pattern of the crack during the rainy season in historical operation and maintenance data, the correlation between environmental factors and damage can be further strengthened, completing the risk factor fusion.

[0065] It should be noted that the extraction methods for various features in this step can be adjusted according to actual engineering needs. For example, a feature importance ranking algorithm can be used to select the feature parameters with the highest correlation to bridge damage features for extraction. Optionally, different algorithms such as weighted fusion and feature-level fusion can be used for feature fusion processing, which can be flexibly adapted according to the bridge's structural type and service environment to further improve the accuracy and relevance of the fusion processing.

[0066] In this embodiment, a hierarchical extraction strategy is employed to accurately separate the design, operation and maintenance, transportation, and geographic / meteorological layer-related features from a spatiotemporally fused multi-source heterogeneous dataset. Core data is then selected as needed, avoiding fusion biases caused by mixed multi-source data and laying a standardized data foundation for subsequent fusion. Furthermore, through hierarchical extraction and a two-level progressive fusion logic, the bridge damage characteristics are endowed with basic engineering attributes through basic engineering fusion. This integrates design, operation and maintenance, and transportation layer information, upgrading the damage characteristics from single-state data to comprehensive data encompassing inherent design, historical evolution, and actual loads. This allows for precise mining of the engineering causes of damage, and further... By deeply integrating risk-induced factors with damage characteristics and geographical and meteorological risk factors, the correlation between environment and damage is explored. The integrated characteristics possess both engineering foundation and environmental risk attributes, comprehensively reflecting the true damage state of bridges under the combined effects of engineering, load, and environmental factors. This achieves deep integration of bridge damage characteristics with multi-dimensional heterogeneous data, breaking down data barriers. The resulting fused damage characteristic dataset can systematically and comprehensively characterize the bridge damage state and influencing factors, providing high-quality multi-dimensional data support for subsequent damage evolution and dynamic early warning, and significantly improving the scientificity and accuracy of damage analysis and early warning assessment.

[0067] In one possible implementation, when performing risk factor fusion processing, a bridge damage feature data set with engineering foundation attributes can be used as a basis, combined with a geographic and meteorological layer risk factor feature data set and a historical damage evolution feature data set from the operation and maintenance layer. The risk factor fusion processing is then performed according to a process of risk quantification, coefficient adaptation, coupled calculation, and incremental fusion, thereby obtaining a fused damage feature data set that combines engineering foundation and meteorological risk coupling attributes. Specifically, a method for performing risk factor fusion processing on the bridge damage feature data set with engineering foundation attributes, the geographic and meteorological layer risk factor feature data set, and the historical damage evolution feature data set from the operation and maintenance layer to obtain a fused damage feature data set may include: B1. Perform risk quantification processing on each geographical meteorological layer risk inducement feature data in the geographical meteorological layer risk inducement feature data set to obtain a geographical meteorological layer risk inducement quantification feature data set. B2. Based on the bridge damage type, the initial sensitivity coefficient is adjusted to match the target bridge according to the historical damage evolution characteristic data set of the operation and maintenance layer and the risk inducement characteristic data set of the geographical and meteorological layer, so as to obtain the target bridge adaptation sensitivity coefficient. B3. Based on the target bridge adaptation sensitivity coefficient and the set of quantitative characteristic data of risk inducements of the geographic meteorological layer, risk coupling increment calculation is performed to obtain meteorological risk coupling increment data; B4. Perform meteorological risk fusion processing on the meteorological risk coupled incremental data and the bridge damage feature data set with engineering basic attributes to obtain a fused damage feature data set.

[0068] Risk quantification refers to the standardization and quantification of risk-inducing characteristic data in the geographic meteorological layer. Optionally, linear extreme value normalization can be used to achieve the above-mentioned risk quantification. For example, raw meteorological and geographic data such as rainfall, wind speed, relative humidity, and extreme temperatures can be converted into risk quantification values ​​in the range of 0 to 1. The set of geographic meteorological layer risk-inducing characteristic data formed by integrating all quantification values ​​can accurately characterize the degree of risk of various environmental factors inducing or accelerating bridge damage.

[0069] The set of quantitative characteristic data of risk inducement in the geo-meteorological layer refers to the set obtained after risk quantification processing of the characteristic data of risk inducement in the geo-meteorological layer. Each data in the set is a quantitative value of the corresponding meteorological and geographical risk inducement, which can intuitively and standardizedly reflect the strength of various environmental risks and is the basic data for carrying out risk coupling incremental calculation.

[0070] The initial sensitivity coefficient refers to a pre-set basic coefficient, based on the bridge structure type and damage type, characterizing the sensitivity of bridge damage to various geographical and meteorological risk factors. Its value ranges from 0 to 1. For example, for crack damage in concrete bridges, the initial sensitivity coefficient to humidity changes can be set to 0.3, and the initial sensitivity coefficient to temperature changes can be set to 0.2. These coefficients can be obtained by consulting bridge design specifications, material property manuals, or historical data statistics of similar bridges.

[0071] The target bridge adaptation sensitivity coefficient refers to a coefficient obtained by adjusting the initial sensitivity coefficient based on the specific bridge's damage type, combined with the historical damage evolution characteristic data set of the operation and maintenance layer and the risk inducement characteristic data set of the geographical and meteorological layer. It can accurately match the actual damage evolution law and environmental risk response characteristics of the target bridge, and its value ranges from 0 to 1.

[0072] Optionally, based on the bridge damage type, the initial sensitivity coefficient is adjusted to match the target bridge according to the historical damage evolution characteristic data set of the operation and maintenance layer and the risk inducement characteristic data set of the geographic and meteorological layer. The process of obtaining the target bridge adaptation sensitivity coefficient can be seen in the following formula: in, The sensitivity coefficient for the target bridge corresponding to the risk factors of the i-th type of geographical and meteorological layer; denoted as , where is the initial sensitivity coefficient corresponding to the risk factor of the i-th type of geographic meteorological layer; k is the adaptation adjustment coefficient, ranging from 0 to 0.5. This adaptation adjustment coefficient can be determined by the actual correlation between the bridge's historical damage and the geographic meteorological environment, as reflected in the historical damage evolution characteristic data set of the target bridge's operation and maintenance layer. If the correlation between the bridge damage development and changes in geographic meteorological factors in the historical data is higher, k can take a high value of 0.4 to 0.5; if the correlation is moderate, k can take a medium value of 0.2 to 0.3; if there is no obvious correlation, k can take a low value of 0 to 0.1. By limiting the range of k, the adjustment range of geographic meteorological factors on the initial sensitivity coefficient can be effectively controlled, avoiding excessive deviation of the coefficient adjustment from the actual engineering situation. Let be the correlation coefficient between the i-th type of geo-meteorological risk factor and the development of bridge damage. This correlation coefficient can be calculated through joint analysis of the historical damage evolution characteristic data set of the maintenance layer and the risk factor characteristic data set of the geo-meteorological layer. Specifically, the historical development rate time series data of a certain type of bridge damage can be extracted from the historical damage evolution characteristic data set of the maintenance layer, and the monitoring time series data of the i-th type of geo-meteorological risk factor corresponding to the same time dimension can be extracted from the risk factor characteristic data set of the geo-meteorological layer. For example, the Pearson correlation coefficient method can be used to calculate the correlation between the two sets of time series data. The calculation result can range from -1 to 1. Since only when geo-meteorological factors have a positive driving effect on the development of bridge damage is it meaningful for engineering analysis, only the positive correlation calculation results are retained. If the calculation result is negative or has no correlation, it is directly... Take 0.

[0073] It should be noted that the above formula is used to calculate the adaptation sensitivity coefficient of the target bridge. During the process, the calculation result may fall between 1 and 1.5. To address this, the following methods can be used to ensure that the final target bridge adaptation sensitivity coefficient meets the required values: for example, by completing all... After calculation, normalization is performed to adjust all results to the 0-1 range; alternatively, a threshold truncation can be applied directly, meaning that if the result is greater than 1, it can be forcibly reduced to the 0-1 range. The value is set to 1 to ensure that it is always within the defined range of 0 to 1. This application does not impose any restrictions on this.

[0074] It should be noted that the correlation coefficient used in calculating the sensitivity coefficient adjustment... The historical damage evolution characteristic data set of the operation and maintenance layer can be the complete original monitoring data and historical records of the operation and maintenance layer, and can include multi-dimensional content such as damage development time series, full information on maintenance measures, and original data on the correlation between damage and environmental factors; while the influence coefficient of historical damage evolution characteristics of the operation and maintenance layer in the aforementioned formula This can be viewed as a single-dimensional impact coefficient extracted and quantified from the full set of raw data for the current damage type. The two represent the relationship between the full set of raw data and the single-dimensional quantified feature, respectively serving the environmental correlation analysis of damage correction and sensitivity coefficient adjustment in basic engineering integration, providing data support for two different technical aspects.

[0075] It should also be noted that the sensitivity coefficient adjustment in this application is not a generalized and fixed coefficient setting method used in the prior art. Instead, it is based on the historical damage evolution characteristic data of the operation and maintenance layer (the historical damage pattern of the bridge itself) and the risk inducement characteristic data of the geographical and meteorological layer (the correlation pattern between the environment and damage). The initial sensitivity coefficient is adapted to the target bridge. By using the correlation coefficient and adaptation adjustment coefficient calculated from historical data, the sensitivity coefficient can accurately match the structural characteristics, damage type and actual service environment of the target bridge. This solves the technical defect in the prior art where the general coefficient does not match the actual situation of the bridge, resulting in a large deviation in the environmental risk coupling calculation.

[0076] Risk coupling increment calculation refers to the process of combining the target bridge's adaptability sensitivity coefficient with the quantitative characteristic data of geographical and meteorological risk factors, and then calculating the meteorological risk coupling increment through a quantitative formula. This risk coupling increment calculation process can achieve the coupled calculation of bridge environmental sensitivity and actual environmental risk level, obtaining a comprehensive acceleration increment that can characterize the damage caused by environmental factors to the bridge.

[0077] Meteorological risk coupling increment data refers to the quantitative data obtained by calculating the risk coupling increment, which can accurately characterize the comprehensive accelerated impact of various risk factors in the geographic meteorological layer on the development of bridge damage.

[0078] Specifically, a weighted summation method can be used to calculate the above-mentioned risk coupling increment. This involves multiplying the target bridge adaptation sensitivity coefficient corresponding to each geographic meteorological risk factor by its quantitative characteristic value, and then summing all the results to obtain the meteorological risk coupling increment data. Optionally, the process of calculating the risk coupling increment based on the target bridge adaptation sensitivity coefficient and the set of quantitative characteristic data of geographic meteorological risk factors to obtain the meteorological risk coupling increment data can be found in the following formula: in, The meteorological risk coupled incremental data is a larger value, which indicates a stronger overall accelerating effect of meteorological risk on bridge damage, and a smaller value indicates a weaker overall accelerating effect of meteorological risk on bridge damage. The sensitivity coefficient for the target bridge to be adapted to the risk factors of the i-th type of geographical and meteorological layer; is the quantitative characteristic value of the i-th type of geometeorological layer risk inducement, with a value of 0~1; n is the number of geometeorological layer risk inducement types participating in the coupled calculation.

[0079] It should be noted that the above formula is used to calculate the coupled incremental data of meteorological risk. During the process, because the number of geographic and meteorological risk factors n involved in the coupled calculation is greater than 1, the calculation result may be greater than 1. To address this situation, this application can employ a normalization method, such as after completing all... After weighted summation, all calculation results are normalized to uniformly adjust them to the range of 0 to 1, thereby ensuring that the final value range of the meteorological risk coupled incremental data is 0 to 1.

[0080] Meteorological risk fusion processing refers to the process of deeply integrating incremental meteorological risk data with bridge damage characteristic data sets that include engineering fundamental attributes. This meteorological risk fusion processing can attribute the comprehensive and accelerating impact of environmental risks to bridge damage characteristics, making the fused damage characteristics possess both engineering fundamental attributes and meteorological risk coupling attributes, thus more closely reflecting the actual damage state of bridges under real environmental conditions.

[0081] Specifically, the above-mentioned meteorological risk fusion processing can be achieved using an incremental overlay fusion method. For example, the quantitative values ​​of bridge damage features with engineering basic attributes can be overlaid with the incremental meteorological risk coupling data to obtain the quantitative values ​​of bridge damage features with fused meteorological risk coupling attributes. After all the quantitative values ​​are bound to the corresponding damage features, they are integrated to form a fused damage feature data set. Optionally, the process of performing meteorological risk fusion processing based on the meteorological risk coupling incremental data and the bridge damage feature data set with engineering basic attributes to obtain the fused damage feature data set can be seen in the following formula: in, The data set is a fused damage feature dataset, which contains bridge damage feature quantification data that incorporates meteorological risk coupling attributes. Quantitative values ​​of bridge damage characteristics with engineering-based attributes; Incremental data is coupled to meteorological risk.

[0082] It should be noted that the above formula is used to calculate the fusion damage characteristic data. During the process, due to the coupling of meteorological risks to incremental The value range is 0 1. When When the value is greater than 0, the calculation result may be greater than 1. To address this, this application can use normalization to adjust the calculation result to the range of 0 to 1, ensuring that the final fused damage feature data has a value range of 0 to 1.

[0083] In this embodiment, by standardizing and quantifying the geographical and meteorological risk factors, environmental data can be directly involved in the coupled calculation, and the values ​​are consistent with those of the preceding data, ensuring computational continuity. The method abandons the generalized coefficient setting and adjusts the sensitivity coefficient based on the bridge's historical damage and actual environmental characteristics, significantly improving the accuracy and adaptability of the coupled calculation. Through weighted summation, a deep coupled calculation of multiple factors is completed, upgrading the comprehensive accelerating impact of environmental risks on bridge damage from a qualitative description to a quantitative characterization for the first time, reflecting the engineering innovation logic of multi-factor coupling. Incremental superposition and fusion yield a fused damage feature data set that combines engineering fundamentals and meteorological risk coupling attributes, comprehensively reflecting the actual damage state of the bridge under the combined effect of multiple factors, breaking through the limitations of traditional fusion methods that only consider engineering attributes. The quantitative fusion process constructed in this step achieves deep quantitative fusion of bridge damage characteristics and environmental risk factors, providing high-quality multi-dimensional data support for subsequent analysis, significantly improving the scientific rigor and accuracy of bridge damage evolution prediction and intelligent early warning. The overall method possesses outstanding engineering innovation and practical value.

[0084] It should be noted that this application adopts a two-step layered fusion mechanism of basic engineering fusion and risk-induced factor fusion. First, through basic engineering fusion, the bridge damage characteristics are deeply bound with the core engineering data of the design, operation and maintenance, and traffic layers, giving the damage characteristics basic engineering attributes. This solves the problem that the damage characteristics of existing technologies lack engineering background support and are out of touch with actual service conditions. Then, through risk-induced factor fusion, the geographical and meteorological environmental data are specifically integrated to achieve a progressive fusion of engineering attributes and environmental risk attributes, rather than the indiscriminate data fusion of existing technologies. This allows the fused damage characteristics to accurately reflect the real damage state of the bridge under the combined effects of multiple factors such as engineering, load, and environment. This step-by-step layered fusion logic lays the core data foundation for subsequent accurate damage evolution analysis and risk early warning.

[0085] In one possible implementation, when processing bridge damage evolution, the current damage development state and the original damage development rate can be extracted first. Then, the original rate is specifically corrected by combining meteorological risk coupled incremental data from the fused data. Furthermore, based on the corrected damage development rate and the current damage state, damage evolution inference is completed, thereby obtaining bridge damage evolution characteristic results that reflect the future development trend of bridge damage. Specifically, a method for processing bridge damage evolution based on the fused damage characteristic data set to obtain bridge damage evolution characteristic results may include: C1. Perform damage state extraction processing on each fused damage feature data in the fused damage feature data set to obtain the current damage development state of the bridge; C2. Perform damage development rate extraction processing on each fusion damage feature data in the fusion damage feature data set to obtain the damage development change rate; C3. Based on the meteorological risk coupling incremental data in the fused damage feature data set, the damage development rate is corrected according to the risk coupling to obtain the damage development correction rate. C4. Based on the damage development correction rate and the current damage development state of the bridge, perform bridge damage evolution processing to obtain bridge damage evolution characteristic results.

[0086] Damage state extraction processing refers to the process of extracting key information that characterizes the current development stage, severity, and distribution characteristics of bridge damage from each fused damage feature data set in the fused damage feature dataset, forming a standardized damage state description. The current damage development state of the bridge refers to the specific development representation of various types of bridge damage at the current time point obtained after damage state extraction processing, which may include core information such as damage level, morphology, location, and quantification degree.

[0087] Damage development rate extraction processing refers to the process of extracting the rate of change of various bridge damages over time from each fused damage feature data set, combined with historical evolution correlation information. The damage development rate can be defined as the quantitative value of the development and change of various bridge damages per unit time, obtained after damage development rate extraction processing, such as the monthly crack expansion rate and the annual growth rate of corrosion area. This damage development rate reflects the original development trend of the damage without additional risk intervention.

[0088] Risk-coupled rate correction processing refers to the process of dynamically adjusting the rate of damage development by combining quantified incremental meteorological risk coupling data from a fusion damage feature dataset. This risk-coupled rate correction processing incorporates the comprehensive accelerating effect of meteorological risk on damage into the damage rate consideration, making the corrected rate more consistent with the damage development patterns under actual environmental conditions. The corrected damage development rate can refer to the quantified damage development rate obtained after risk-coupled rate correction processing, which incorporates the accelerating effect of meteorological risk coupling, and can reflect the actual rate of change of bridge damage under the combined effects of engineering, load, and environmental factors.

[0089] Optionally, a risk increment weighted correction method can be used to weight the meteorological risk coupling increment data in the fused damage feature dataset with the damage development change rate, thereby realizing dynamic correction of the original rate (i.e., the damage development change rate) and thus obtaining the damage development correction rate.

[0090] Optionally, linear evolutionary extrapolation calculations can be performed based on the damage development correction rate and the current damage development status of the bridge, combined with a preset evolution time period (such as 3 months, 6 months, or 12 months), to obtain the predicted damage development values ​​for various types of damage within the preset period. These values ​​are then combined with bridge maintenance standards to determine the risk level of the predicted damage. By integrating information such as damage development trends, predicted severity, and risk evolution characteristics, a bridge damage evolution characteristic result is formed. For example, if the current state of a main beam crack on a bridge is 0.3 mm, the damage development correction rate is 0.05 mm / month, and the preset evolution period is 6 months, then the predicted crack width after 6 months can be extrapolated to be 0.3 + 0.05 × 6 = 0.6 mm. Further, based on the standards, it can be determined that the crack will enter a highly developed state after 6 months, posing a structural risk. The evolution trend of the crack, the predicted level, and the evolution information of all damage constitute the aforementioned bridge damage evolution characteristic result.

[0091] It should be noted that the preset evolution time period in the embodiments of this application can be flexibly adjusted according to the actual needs of bridge safety management. The damage evolution simulation method can also be selected according to the bridge structure type and damage characteristics, such as linear simulation or nonlinear simulation, which can improve the pertinence and practicality of damage evolution treatment.

[0092] The above uses linear extrapolation as an example for explanation. In practical applications, nonlinear evolution extrapolation can be carried out based on historical damage evolution data of the maintenance layer, structural mechanical parameters of the design layer, and the nonlinear development characteristics of bridge damage and the impact of sudden risks. The relevant rate correction coefficient and sudden risk multiplier can be pre-calibrated according to the actual project. This application does not limit this.

[0093] In this embodiment, by separating and extracting the current damage state and the original development rate of the bridge, a clear and independent basis is provided for subsequent analysis. By introducing meteorological risk coupled incremental data to dynamically correct the damage development rate, the actual evolution law of damage under the combined action of multiple factors such as engineering, load, and environment is realistically restored, effectively solving the technical limitation of traditional evolution analysis that ignores environmental risks. Based on the corrected rate and the current state, the evolutionary deduction is completed, and the damage evolution characteristic results obtained can directly support subsequent dynamic early warning, realizing a technical upgrade from current status judgment to trend prediction. This significantly improves the scientificity, accuracy, and foresight of damage evolution deduction, and provides key and reliable trend basis for full-process early warning.

[0094] It should be noted that the damage evolution rate correction in this application incorporates the aforementioned meteorological risk coupled incremental data, adapted and adjusted based on historical data, into the damage development rate correction process, thus achieving dynamic risk coupling correction of the damage development rate. This correction mechanism is not a simple rate superposition, but rather based on the correlation between the bridge's historical damage and the environment. This allows the corrected damage development rate to truly reflect the actual evolution law under the combined effects of multiple factors, including engineering, load, and environment. It solves the problems of large deviations between existing technology damage evolution projections and actual working conditions, and inaccurate predictions of future damage trends, significantly improving the scientific rigor and foresight of the damage evolution characteristic results.

[0095] In one possible implementation, when performing risk-coupled dynamic intelligent early warning processing, the process can be based on bridge damage evolution characteristics and a fused damage feature dataset. Following a procedure of damage severity determination, evolution trend quantification, meteorological risk normalization, multi-dimensional weighted coupling, and early warning result determination, the process outputs a bridge structural safety early warning result. Specifically, a method for performing risk-coupled dynamic intelligent early warning processing based on the bridge damage evolution characteristics and the fused damage feature dataset to obtain a bridge structural safety early warning result may include: D1. Based on the bridge damage evolution characteristics, determine the degree of bridge damage to obtain the current degree of bridge damage. D2. Based on the bridge damage evolution characteristics, perform evolution trend analysis and processing to obtain dynamic evolution trend quantitative data; D3. Normalize the meteorological risk coupled incremental data in the fused damage feature dataset to obtain the current meteorological risk quantification data. D4. Perform multi-dimensional weighted risk coupling calculation on the current bridge damage level, the dynamic evolution trend quantification data, and the current meteorological risk quantification data to obtain dynamic comprehensive risk quantification data; D5. Based on the dynamic comprehensive risk quantification data, determine the bridge structure safety early warning result.

[0096] The bridge damage severity assessment process refers to the process of classifying and determining the severity of current bridge damage based on the bridge damage evolution characteristics. The current bridge damage severity quantification data refers to the 0-1 range value obtained after classifying and quantifying the severity of current bridge damage based on the bridge damage evolution characteristics; this value can be used to characterize the degree of current bridge damage.

[0097] Optionally, the damage quantification value at the current moment can be extracted from the bridge damage evolution characteristic results, and determined according to a preset damage level range to obtain the current bridge damage degree. For example, if the damage quantification value at the current moment is 0.85 extracted from the bridge damage evolution characteristic results, and the preset range is 0–0.3 (mild, assigned a value of 0.2), 0.3–0.7 (moderate, assigned a value of 0.6), and 0.7–1.0 (severe, assigned a value of 0.9), then 0.85 can be mapped to severe damage, and the damage quantification value of 0.85 at the current moment can be assigned a value of 0.9 as the quantification value of the current bridge damage degree. The above assignment can be adaptively set and adjusted according to bridge maintenance specifications, bridge structure type, and actual engineering experience, aiming to uniformly quantify the damage level into a value within the range of 0–1, facilitating subsequent risk coupling calculations.

[0098] Evolutionary trend analysis refers to the trend analysis of bridge damage evolution characteristics to obtain quantitative data on the future speed of damage development, whether it is accelerating, and whether it is stable. Dynamic evolutionary trend quantitative data refers to quantitative indicators that characterize the future development trend of damage, reflecting the speed and danger of damage development.

[0099] Optionally, damage development and change data over future time periods can be extracted from the bridge damage evolution characteristics results to calculate a quantitative index of the damage development rate, serving as the dynamic evolution trend quantification data. For example, assuming the preset evolution trend quantification range can be: 0–0.02 mm / month corresponds to a value of 0–0.3, 0.02–0.05 mm / month corresponds to a value of 0.3–0.7, and greater than 0.05 mm / month corresponds to a value of 0.7–1.0, if the damage change over the next 6 months is extracted to be 0.30 mm, and the calculated monthly damage development change is 0.05 mm / month, then according to the preset evolution trend quantification range, the dynamic evolution trend quantification data can be determined to be 0.7. The aforementioned preset evolution trend quantification range can be adjusted based on bridge type and engineering experience.

[0100] Normalization refers to the process of uniformly converting incremental meteorological risk data to the 0-1 range to facilitate coupling calculations with data from other dimensions. Current meteorological risk quantification data refers to the normalized quantified value that reflects the overall risk level of the current meteorological environment to the bridge structure.

[0101] Multi-dimensional weighted risk coupling calculation refers to the process of comprehensively coupling the current bridge damage level, dynamic evolution trend quantification data, and current meteorological risk quantification data according to weights to obtain an overall risk value. Dynamic comprehensive risk quantification data refers to the total risk quantification value obtained by combining the current damage status, development trend, and meteorological risk, which can be used to directly determine the warning level and further obtain bridge structural safety warning results.

[0102] Optionally, the process of performing multi-dimensional weighted risk coupling calculations on the current bridge damage level, the dynamic evolution trend quantification data, and the current meteorological risk quantification data to obtain dynamic comprehensive risk quantification data can be found in the following formula: in, This is dynamic, comprehensive risk quantification data, with a value range of 0. 1; This data quantifies the current degree of bridge damage, with values ​​ranging from 0. 1; Quantify data for dynamic evolution trends, with values ​​ranging from 0. 1; This is the current meteorological risk quantification data, with a value range of 0. 1; , and For the weighting coefficients, satisfying + + =1, the weighting coefficient can be adjusted according to the bridge type and engineering experience.

[0103] Furthermore, the bridge structural safety warning result can be determined based on dynamic comprehensive risk quantification data and a preset bridge safety warning level range. This preset range can be pre-set according to bridge maintenance specifications, structural type, and engineering experience. For example, the preset bridge safety warning level ranges could be: 0–0.3 corresponding to Level I (safe state), 0.3–0.6 corresponding to Level II (attention state), 0.6–0.8 corresponding to Level III (warning state), and 0.8–1.0 corresponding to Level IV (emergency warning state). Different control and response measures can be set for different warning levels. If the calculated dynamic comprehensive risk quantification data is 0.87, falling within the range of 0.8–1.0, the current bridge structural safety warning result is determined to be a Level IV emergency warning state, and corresponding control and response measures can be implemented. These measures may include initiating on-site inspections, intensified structural monitoring, abnormal data verification, and safety warning reporting. This application does not impose any restrictions on these measures.

[0104] In this embodiment, by grading and quantifying the degree of damage and analyzing the evolution trend based on the bridge damage evolution characteristics, and by combining meteorological risk data to carry out multi-dimensional weighted coupling calculations, a dynamic, comprehensive, and quantitative assessment of the safety status of the bridge structure can be achieved. This not only outputs standardized early warning levels but also matches corresponding response measures, effectively improving the accuracy, timeliness, and operability of bridge structure safety monitoring, and providing reliable quantitative basis and decision support for bridge safety management.

[0105] For examples consistent with the above embodiments, please refer to... Figure 3 , Figure 3 This is a schematic diagram of the structure of a terminal provided in an embodiment of this application, such as... Figure 3 As shown, it includes a processor, an input device, an output device, and a memory, which are interconnected. The memory is used to store a computer program, which includes program instructions. The processor is configured to call the program instructions. The program includes instructions for performing the following steps. Acquire a multi-source heterogeneous data set of bridges; the multi-source heterogeneous data set of bridges includes data from the perception layer, design layer, operation and maintenance layer, traffic layer, and geographic and meteorological layer. Spatiotemporal fusion processing is performed on each bridge multi-source heterogeneous data in the bridge multi-source heterogeneous data set to obtain a spatiotemporal fused multi-source heterogeneous data set. The perception layer data is processed by extracting bridge damage features based on multimodal data to obtain a bridge damage feature data set. Multi-level feature fusion processing is performed on the bridge damage feature data set and the spatiotemporal fusion multi-source heterogeneous data set to obtain a fused damage feature data set. Bridge damage evolution processing is performed based on the fused damage feature data set to obtain bridge damage evolution feature results; Based on the bridge damage evolution characteristics and the fused damage characteristic data set, dynamic intelligent early warning processing based on risk coupling is performed to obtain bridge structural safety early warning results.

[0106] The above mainly describes the solutions of the embodiments of this application from the perspective of the method execution process. It is understood that, in order to achieve the above functions, the terminal includes the corresponding hardware structure and / or software modules for executing each function. Those skilled in the art should readily recognize that, in conjunction with the units and algorithm steps of the various examples described in the embodiments provided herein, this application can be implemented in hardware or a combination of hardware and computer software. Whether a function is executed in hardware or by computer software driving hardware depends on the specific application and design constraints of the technical solution. Those skilled in the art can use different methods to implement the described functions for each specific application, but such implementation should not be considered beyond the scope of this application.

[0107] This application embodiment can divide the terminal into functional units according to the above method example. For example, each function can be divided into a separate functional unit, or two or more functions can be integrated into one processing unit. The integrated unit can be implemented in hardware or as a software functional unit. It should be noted that the unit division in this application embodiment is illustrative and only represents one logical functional division. In actual implementation, there may be other division methods.

[0108] For those consistent with the above, please refer to Figure 4 , Figure 4 This application provides a schematic diagram of a bridge structure safety early warning device based on multi-source heterogeneous data. For example... Figure 4 As shown, the device includes: The acquisition unit 101 is used to acquire a multi-source heterogeneous data set of bridges; the multi-source heterogeneous data set of bridges includes sensing layer data, design layer data, operation and maintenance layer data, traffic layer data, and geographic and meteorological layer data; The first processing unit 102 is used to perform spatiotemporal fusion processing on each bridge multi-source heterogeneous data in the bridge multi-source heterogeneous data set to obtain a spatiotemporal fused multi-source heterogeneous data set. The second processing unit 103 is used to perform bridge damage feature extraction processing based on multimodal data on the perception layer data to obtain a bridge damage feature data set. The third processing unit 104 is used to perform multi-level feature fusion processing based on the bridge damage feature data set and the spatiotemporal fusion multi-source heterogeneous data set to obtain a fused damage feature data set. The fourth processing unit 105 is used to perform bridge damage evolution processing based on the fused damage feature data set to obtain bridge damage evolution feature results. The fifth processing unit 106 is used to perform dynamic intelligent early warning processing based on risk coupling according to the bridge damage evolution characteristic results and the fused damage characteristic data set, so as to obtain the bridge structure safety early warning result.

[0109] In one possible implementation, the third processing unit 104 is configured to perform multi-level feature fusion processing based on the bridge damage feature data set and the spatiotemporal fusion multi-source heterogeneous data set to obtain a fused damage feature data set, specifically for: The design layer-related features are extracted from the spatiotemporal fusion multi-source heterogeneous data set to obtain the design layer structural mechanical feature data set. From the spatiotemporal fusion multi-source heterogeneous data set, features associated with the operation and maintenance layer are extracted to obtain a historical damage evolution feature data set of the operation and maintenance layer. The traffic layer-related features are extracted from the spatiotemporally fused multi-source heterogeneous data set to obtain the traffic layer load feature data set. From the spatiotemporal fusion multi-source heterogeneous data set, features related to the geographic meteorological layer are extracted to obtain a geographic meteorological layer risk inducement feature data set. The bridge damage feature data set is fused with the design layer structural mechanics feature data set, the operation and maintenance layer historical damage evolution feature data set, and the traffic layer load feature data set through basic engineering fusion processing to obtain a bridge damage feature data set with engineering basic attributes. The bridge damage feature data set with engineering basic attributes is fused with the risk inducement feature data set of the geographic and meteorological layer and the historical damage evolution feature data set of the operation and maintenance layer to obtain a fused damage feature data set.

[0110] In one possible implementation, the third processing unit 104 is used to perform risk factor fusion processing on the bridge damage feature data set with engineering basic attributes, the geographical and meteorological layer risk inducement feature data set, and the operation and maintenance layer historical damage evolution feature data set to obtain a fused damage feature data set, specifically for: Each risk-inducing feature data of the geographic meteorological layer risk inducement feature data set in the geographic meteorological layer risk inducement feature data set is subjected to risk quantification processing to obtain the geographic meteorological layer risk inducement quantification feature data set. Based on the bridge damage type, the initial sensitivity coefficient is adjusted to match the target bridge according to the historical damage evolution characteristic data set of the operation and maintenance layer and the risk inducement characteristic data set of the geographic and meteorological layer, so as to obtain the target bridge adaptation sensitivity coefficient. Based on the target bridge adaptation sensitivity coefficient and the set of quantitative characteristic data of risk inducements in the geographic meteorological layer, risk coupling increment calculation is performed to obtain meteorological risk coupling increment data; Meteorological risk fusion processing is performed on the meteorological risk coupled incremental data and the bridge damage feature data set with engineering basic attributes to obtain a fused damage feature data set.

[0111] In one possible implementation, the fourth processing unit 105 is used to perform bridge damage evolution processing based on the fused damage feature data set to obtain bridge damage evolution feature results, specifically for: Damage state extraction processing is performed on each fused damage feature data in the fused damage feature data set to obtain the current damage development state of the bridge; Each fusion damage feature data in the fusion damage feature data set is processed to extract the damage development rate, thereby obtaining the damage development change rate. Based on the meteorological risk coupling incremental data in the fused damage feature dataset, the damage development rate is corrected by risk coupling to obtain the damage development correction rate. Based on the damage development correction rate and the current damage development state of the bridge, bridge damage evolution processing is performed to obtain bridge damage evolution characteristic results.

[0112] In one possible implementation, the fifth processing unit 106 is used to perform dynamic intelligent early warning processing based on risk coupling according to the bridge damage evolution characteristic results and the fused damage characteristic data set, to obtain bridge structural safety early warning results, specifically for: Based on the bridge damage evolution characteristics, the degree of bridge damage is determined to obtain quantitative data on the current degree of bridge damage. Based on the bridge damage evolution characteristics, an evolution trend analysis was performed to obtain dynamic evolution trend quantitative data. The meteorological risk coupled incremental data in the fused damage feature dataset are normalized to obtain the current meteorological risk quantification data; The current bridge damage level quantification data, the dynamic evolution trend quantification data, and the current meteorological risk quantification data are combined with multi-dimensional weighted risk coupling calculation to obtain dynamic comprehensive risk quantification data; Based on the dynamic comprehensive risk quantification data, the bridge structure safety early warning results are determined.

[0113] This application also provides a computer storage medium storing a computer program for electronic data interchange, which causes a computer to perform some or all of the steps of any of the bridge structure safety early warning methods based on multi-source heterogeneous data as described in the above method embodiments.

[0114] This application also provides a computer program product, which includes a non-transitory computer-readable storage medium storing a computer program that causes a computer to perform some or all of the steps of any of the bridge structure safety early warning methods based on multi-source heterogeneous data as described in the above method embodiments.

[0115] It should be noted that, for the sake of simplicity, the foregoing method embodiments are all described as a series of actions. However, those skilled in the art should understand that this application is not limited to the described order of actions, as some steps may be performed in other orders or simultaneously according to this application. Furthermore, those skilled in the art should also understand that the embodiments described in the specification are preferred embodiments, and the actions and modules involved are not necessarily essential to this application.

[0116] In the above embodiments, the descriptions of each embodiment have different focuses. For parts not described in detail in a certain embodiment, please refer to the relevant descriptions in other embodiments.

[0117] In the several embodiments provided in this application, it should be understood that the disclosed apparatus can be implemented in other ways. For example, the apparatus embodiments described above are merely illustrative; for instance, the division of units is only a logical functional division, and in actual implementation, there may be other division methods. For example, multiple units or components may be combined or integrated into another system, or some features may be ignored or not executed. Furthermore, the coupling or direct coupling or communication connection shown or discussed may be through some interfaces; the indirect coupling or communication connection between devices or units may be electrical or other forms.

[0118] The units described as separate components may or may not be physically separate. The components shown as units may or may not be physical units; that is, they may be located in one place or distributed across multiple network units. Some or all of the units can be selected to achieve the purpose of this embodiment according to actual needs.

[0119] Furthermore, the functional units in the various embodiments of the application can be integrated into one processing unit, or each unit can exist physically separately, or two or more units can be integrated into one unit. The integrated unit can be implemented in hardware or as a software program module.

[0120] If the integrated unit is implemented as a software program module and sold or used as an independent product, it can be stored in a computer-readable storage device (CMD). Based on this understanding, the technical solution of this application, in essence, or the part that contributes to the prior art, or all or part of the technical solution, can be embodied in the form of a software product. This computer software product is stored in a memory and includes several instructions to cause a computer device (which may be a personal computer, server, or network device, etc.) to execute all or part of the steps of the methods described in the various embodiments of this application. The aforementioned memory includes various media capable of storing program code, such as USB flash drives, read-only memory (ROM), random access memory (RAM), portable hard drives, magnetic disks, or optical disks.

[0121] Those skilled in the art will understand that all or part of the steps in the various methods of the above embodiments can be implemented by a program instructing related hardware. The program can be stored in a computer-readable storage device, which may include: a flash drive, a read-only memory, a random access memory, a magnetic disk, or an optical disk, etc.

[0122] The embodiments of this application have been described in detail above. Specific examples have been used to illustrate the principles and implementation methods of this application. The description of the above embodiments is only for the purpose of helping to understand the method and core ideas of this application. At the same time, for those skilled in the art, there will be changes in the specific implementation methods and application scope based on the ideas of this application. Therefore, the content of this specification should not be construed as a limitation of this application.

Claims

1. A bridge structural safety early warning method based on multi-source heterogeneous data, characterized in that, The bridge structural safety early warning method based on multi-source heterogeneous data includes: Acquire a multi-source heterogeneous data set of bridges; the multi-source heterogeneous data set of bridges includes data from the perception layer, design layer, operation and maintenance layer, traffic layer, and geographic and meteorological layer. Spatiotemporal fusion processing is performed on each bridge multi-source heterogeneous data in the bridge multi-source heterogeneous data set to obtain a spatiotemporal fused multi-source heterogeneous data set. The perception layer data is processed by extracting bridge damage features based on multimodal data to obtain a bridge damage feature data set. Multi-level feature fusion processing is performed on the bridge damage feature data set and the spatiotemporal fusion multi-source heterogeneous data set to obtain a fused damage feature data set. Bridge damage evolution processing is performed based on the fused damage feature data set to obtain bridge damage evolution feature results; Based on the bridge damage evolution characteristics and the fused damage characteristic data set, dynamic intelligent early warning processing based on risk coupling is performed to obtain bridge structural safety early warning results. The process of performing multi-level feature fusion processing based on the bridge damage feature data set and the spatiotemporal fused multi-source heterogeneous data set to obtain a fused damage feature data set includes: The design layer-related features are extracted from the spatiotemporal fusion multi-source heterogeneous data set to obtain the design layer structural mechanical feature data set. From the spatiotemporal fusion multi-source heterogeneous data set, features associated with the operation and maintenance layer are extracted to obtain a historical damage evolution feature data set of the operation and maintenance layer. The traffic layer-related features are extracted from the spatiotemporally fused multi-source heterogeneous data set to obtain the traffic layer load feature data set. From the spatiotemporal fusion multi-source heterogeneous data set, features related to the geographic meteorological layer are extracted to obtain a geographic meteorological layer risk inducement feature data set. The bridge damage feature data set is fused with the design layer structural mechanics feature data set, the operation and maintenance layer historical damage evolution feature data set, and the traffic layer load feature data set through basic engineering fusion processing to obtain a bridge damage feature data set with engineering basic attributes. The process of fusing the bridge damage feature data set with the design-level structural mechanics feature data set, the operation and maintenance-level historical damage evolution feature data set, and the traffic-level load feature data set to obtain a bridge damage feature data set with engineering foundation attributes is illustrated in the following formula: in, A dataset representing bridge damage characteristics with basic engineering attributes; This represents the original quantized value of the bridge damage features extracted from the bridge damage feature dataset; This represents the influence coefficient of structural mechanical characteristics extracted from the set of structural mechanical characteristics data at the design level; that is, the better the design parameters, the better the influence coefficient. The smaller the value, the weaker the amplification effect on damage; the worse the design parameters, the worse the damage amplification effect. The larger the value, the stronger the amplification effect on the damage; This represents the influence coefficient of historical damage evolution characteristics extracted from the historical damage evolution characteristic data set of the operations and maintenance layer. The better the operations and maintenance status, the higher the coefficient. The smaller the value, the weaker the amplification effect on damage; the worse the operational status, the weaker the amplification effect. The larger the value, the stronger the amplification effect on the damage; This represents the traffic layer load characteristic influence coefficient extracted from the traffic layer load characteristic data set. The more favorable the load conditions, the greater the influence coefficient. The smaller the value, the weaker the amplification effect on damage; the more unfavorable the load conditions, the weaker the amplification effect. The larger the value, the stronger the amplification effect on the damage; , and It can represent the engineering integration weight coefficient, which can be adapted and set according to the bridge structure type, damage type, and engineering operation and maintenance needs, and always satisfies α+β+γ=1 to ensure the rationality of weight allocation; The bridge damage feature data set with engineering basic attributes is fused with the risk inducement feature data set of the geographic and meteorological layer and the historical damage evolution feature data set of the operation and maintenance layer to obtain a fused damage feature data set. The process involves fusing the bridge damage feature data set with engineering fundamental attributes, the geographical and meteorological risk factor feature data set, and the historical damage evolution feature data set from the operation and maintenance layer to obtain a fused damage feature data set, including: Each risk-inducing feature data of the geographic meteorological layer risk inducement feature data set in the geographic meteorological layer risk inducement feature data set is subjected to risk quantification processing to obtain the geographic meteorological layer risk inducement quantification feature data set. Based on the bridge damage type, the initial sensitivity coefficient is adjusted to match the target bridge according to the historical damage evolution characteristic data set of the operation and maintenance layer and the risk inducement characteristic data set of the geographic and meteorological layer, so as to obtain the target bridge adaptation sensitivity coefficient. Based on the target bridge adaptation sensitivity coefficient and the set of quantitative characteristic data of risk inducements in the geographic meteorological layer, risk coupling increment calculation is performed to obtain meteorological risk coupling increment data; Meteorological risk fusion processing is performed on the meteorological risk coupled incremental data and the bridge damage feature data set with engineering basic attributes to obtain a fused damage feature data set.

2. The bridge structure safety early warning method based on multi-source heterogeneous data according to claim 1, characterized in that, The process of performing bridge damage evolution processing based on the fused damage feature data set to obtain bridge damage evolution feature results includes: Damage state extraction processing is performed on each fused damage feature data in the fused damage feature data set to obtain the current damage development state of the bridge; Each fusion damage feature data in the fusion damage feature data set is processed to extract the damage development rate, thereby obtaining the damage development change rate. Based on the meteorological risk coupling incremental data in the fused damage feature dataset, the damage development rate is corrected by risk coupling to obtain the damage development correction rate. Based on the damage development correction rate and the current damage development state of the bridge, bridge damage evolution processing is performed to obtain bridge damage evolution characteristic results.

3. The bridge structure safety early warning method based on multi-source heterogeneous data according to claim 2, characterized in that, The dynamic intelligent early warning processing based on risk coupling, performed according to the bridge damage evolution characteristics and the fused damage characteristic data set, yields a bridge structural safety early warning result, including: Based on the bridge damage evolution characteristics, the degree of bridge damage is determined to obtain quantitative data on the current degree of bridge damage. Based on the bridge damage evolution characteristics, an evolution trend analysis was performed to obtain dynamic evolution trend quantitative data. The meteorological risk coupled incremental data in the fused damage feature dataset are normalized to obtain the current meteorological risk quantification data; The current bridge damage level quantification data, the dynamic evolution trend quantification data, and the current meteorological risk quantification data are combined with multi-dimensional weighted risk coupling calculation to obtain dynamic comprehensive risk quantification data; Based on the dynamic comprehensive risk quantification data, the bridge structure safety early warning results are determined.

4. A bridge structure safety early warning device based on multi-source heterogeneous data, characterized in that, The device includes: The acquisition unit is used to acquire a multi-source heterogeneous data set of bridges; the multi-source heterogeneous data set of bridges includes data from the perception layer, design layer, operation and maintenance layer, traffic layer, and geographic and meteorological layer. The first processing unit is used to perform spatiotemporal fusion processing on each bridge multi-source heterogeneous data in the bridge multi-source heterogeneous data set to obtain a spatiotemporal fused multi-source heterogeneous data set. The second processing unit is used to perform bridge damage feature extraction processing on the perception layer data based on multimodal data to obtain a bridge damage feature data set. The third processing unit is used to perform multi-level feature fusion processing based on the bridge damage feature data set and the spatiotemporal fusion multi-source heterogeneous data set to obtain a fused damage feature data set. The fourth processing unit is used to perform bridge damage evolution processing based on the fused damage feature data set to obtain bridge damage evolution feature results. The fifth processing unit is used to perform dynamic intelligent early warning processing based on risk coupling according to the bridge damage evolution characteristic results and the fused damage characteristic data set, so as to obtain the bridge structure safety early warning result. The third processing unit is used to perform multi-level feature fusion processing based on the bridge damage feature data set and the spatiotemporal fused multi-source heterogeneous data set to obtain a fused damage feature data set, specifically for: The design layer-related features are extracted from the spatiotemporal fusion multi-source heterogeneous data set to obtain the design layer structural mechanical feature data set. From the spatiotemporal fusion multi-source heterogeneous data set, features associated with the operation and maintenance layer are extracted to obtain a historical damage evolution feature data set of the operation and maintenance layer. The traffic layer-related features are extracted from the spatiotemporally fused multi-source heterogeneous data set to obtain the traffic layer load feature data set. From the spatiotemporal fusion multi-source heterogeneous data set, features related to the geographic meteorological layer are extracted to obtain a geographic meteorological layer risk inducement feature data set. The bridge damage feature data set is fused with the design layer structural mechanics feature data set, the operation and maintenance layer historical damage evolution feature data set, and the traffic layer load feature data set through basic engineering fusion processing to obtain a bridge damage feature data set with engineering basic attributes. The bridge damage feature data set with engineering basic attributes is fused with the risk inducement feature data set of the geographic and meteorological layer and the historical damage evolution feature data set of the operation and maintenance layer to obtain a fused damage feature data set. The process of fusing the bridge damage feature data set with the design-level structural mechanics feature data set, the operation and maintenance-level historical damage evolution feature data set, and the traffic-level load feature data set to obtain a bridge damage feature data set with engineering foundation attributes is illustrated in the following formula: in, A dataset representing bridge damage characteristics with basic engineering attributes; This represents the original quantized value of the bridge damage features extracted from the bridge damage feature dataset; This represents the influence coefficient of structural mechanical characteristics extracted from the set of structural mechanical characteristics data at the design level; that is, the better the design parameters, the better the influence coefficient. The smaller the value, the weaker the amplification effect on damage; the worse the design parameters, the worse the damage amplification effect. The larger the value, the stronger the amplification effect on the damage; This represents the influence coefficient of historical damage evolution characteristics extracted from the historical damage evolution characteristic data set of the operations and maintenance layer. The better the operations and maintenance status, the higher the coefficient. The smaller the value, the weaker the amplification effect on damage; the worse the operational status, the weaker the amplification effect. The larger the value, the stronger the amplification effect on the damage; This represents the traffic layer load characteristic influence coefficient extracted from the traffic layer load characteristic data set. The more favorable the load conditions, the greater the influence coefficient. The smaller the value, the weaker the amplification effect on damage; the more unfavorable the load conditions, the weaker the amplification effect. The larger the value, the stronger the amplification effect on the damage; , and It can represent the engineering integration weight coefficient, which can be adapted and set according to the bridge structure type, damage type, and engineering operation and maintenance needs, and always satisfies α+β+γ=1 to ensure the rationality of weight allocation; The third processing unit is used to perform risk factor fusion processing on the bridge damage feature data set with engineering basic attributes, the geographical and meteorological layer risk inducement feature data set, and the operation and maintenance layer historical damage evolution feature data set to obtain a fused damage feature data set, specifically used for: Each risk-inducing feature data of the geographic meteorological layer risk inducement feature data set in the geographic meteorological layer risk inducement feature data set is subjected to risk quantification processing to obtain the geographic meteorological layer risk inducement quantification feature data set. Based on the bridge damage type, the initial sensitivity coefficient is adjusted to match the target bridge according to the historical damage evolution characteristic data set of the operation and maintenance layer and the risk inducement characteristic data set of the geographic and meteorological layer, so as to obtain the target bridge adaptation sensitivity coefficient. Based on the target bridge adaptation sensitivity coefficient and the set of quantitative characteristic data of risk inducements in the geographic meteorological layer, risk coupling increment calculation is performed to obtain meteorological risk coupling increment data; Meteorological risk fusion processing is performed on the meteorological risk coupled incremental data and the bridge damage feature data set with engineering basic attributes to obtain a fused damage feature data set.

5. A terminal, characterized in that, The system includes a processor, an input device, an output device, and a memory, which are interconnected. The memory stores a computer program, which includes program instructions. The processor is configured to invoke the program instructions to execute the bridge structure safety early warning method based on multi-source heterogeneous data as described in any one of claims 1-3.

6. A computer-readable storage medium, characterized in that, The computer-readable storage medium stores a computer program, the computer program including program instructions, which, when executed by a processor, cause the processor to perform the bridge structure safety early warning method based on multi-source heterogeneous data as described in any one of claims 1-3.

Images