Bridge disease identification method based on multispectral imaging

Abstract

The present application relates to the technical field of bridge disease identification, and particularly relates to a bridge disease identification method based on multispectral imaging, which comprises the following steps: introducing multispectral imaging data and bridge structure parameters into a corrosion disease state evolution analysis strategy to perform corrosion disease state evolution analysis, so as to obtain the corrosion disease state evolution of the single cable surface of the cable-stayed bridge; introducing two-way lane traffic monitoring data and bridge structure parameters into a bearing influence evolution analysis strategy to perform bearing influence evolution analysis, so as to obtain the bearing influence evolution of the single cable surface of the cable-stayed bridge; introducing the corrosion disease state evolution and the bearing influence evolution into a bridge disease dangerous state analysis strategy, so as to obtain the disease dangerous state of the single cable surface of the cable-stayed bridge; and performing maintenance on the single cable surface of the cable-stayed bridge according to the disease dangerous state; thereby improving the optimized allocation efficiency of maintenance resources, reducing the resource waste caused by excessive maintenance or delayed treatment of key hidden dangers, and improving the targeting and timeliness of maintenance operations.

Description

Technical Field

[0001] This invention relates to the field of bridge defect identification technology, and in particular to a bridge defect identification method based on multispectral imaging. Background Technology

[0002] In modern transportation networks, cable-stayed bridges are widely used in major projects such as those spanning rivers and seas due to their advantages of strong spanning capacity and lightweight, aesthetically pleasing structures. The central single-cable-stayed bridge, as the core force-transfer structure, bears the crucial function of transferring the main beam load to the towers, and its health directly determines the overall load-bearing capacity and service life of the bridge. However, the central single-cable-stayed bridge cable often faces a dual threat of damage during long-term service. On the one hand, the erosion of the natural environment continuously undermines the integrity of the cable structure, leading to oxidation and corrosion of the steel cables. The resulting corrosion products not only cause cross-sectional loss but also cause sheath cracking due to volume expansion, accelerating the spread of damage. On the other hand, the repeated impact of traffic loads drives the continuous evolution of damage. The asymmetrical distribution of traffic flow in both directions generates periodic off-center loading, subjecting the cables to alternating additional stresses, causing cracks in the corroded areas to continuously expand. From initial localized material deterioration, this gradually develops into a decrease in cross-sectional load-bearing capacity, potentially leading to major accidents such as cable breakage and cable-stayed instability. In recent years, multispectral imaging technology has begun to be applied to bridge monitoring. By capturing the spectral reflectance characteristics of different bands, it is possible to achieve early identification of corrosion.

[0003] However, existing technologies for bridge defect identification lack a correlation model between spectral characteristics and the type and thickness of corrosion products, making it impossible to accurately quantify cross-sectional losses and distribution uniformity, thus hindering the assessment of defect severity. Furthermore, current technologies neglect the fact that cable defect evolution is a result of the combined effects of corrosion and load, and fail to incorporate the dynamic impact of asymmetrical traffic loads into the evaluation. For example, when cables show signs of corrosion, without considering the difference in axle load between upstream and downstream traffic and the duration of peak periods, it is impossible to determine whether load impacts will accelerate failure in the corroded area. Finally, existing methods do not consider the varying importance of cables within the cable plane; cables at mid-span and those near the tower have the same risk weight, making it impossible to identify core defect risks in cables at critical locations.

[0004] To address these issues, this application proposes a bridge defect identification method based on multispectral imaging. Summary of the Invention

[0005] The purpose of this invention is to provide a bridge defect identification method based on multispectral imaging. By acquiring multispectral imaging data of cables and combining bridge structural parameters with two-way traffic monitoring data, the method first extracts spectral reflectance characteristics to invert the type and thickness of corrosion products and calculates the evolution of corrosion defects. Then, a structural model is established to analyze the impact of asymmetrical traffic loads on cables and obtain the evolution of load-bearing impact. By coupling the two methods, the risk of a single cable is calculated, and the overall risk of the cable surface is obtained by combining the cable position weights. Finally, maintenance plans are formulated according to the urgency of maintenance.

[0006] This invention is implemented as follows: In a first aspect, the present invention provides a method for identifying bridge defects based on multispectral imaging, comprising the following steps: S1. Acquire multispectral imaging data of the central single cable plane of the cable-stayed bridge, bridge structural parameters, and two-way lane traffic flow monitoring data; S2. Import the multispectral imaging data and bridge structural parameters into the corrosion disease state evolution analysis strategy to conduct corrosion disease state evolution analysis and obtain the corrosion disease state evolution of the central single cable surface of the cable-stayed bridge. S3. Import the two-way traffic flow monitoring data and bridge structural parameters into the load-bearing impact evolution analysis strategy to conduct load-bearing impact evolution analysis and obtain the load-bearing impact evolution of the central single cable surface of the cable-stayed bridge. S4. Obtain the evolution of corrosion and load-bearing impact on the central single cable surface of the cable-stayed bridge, and import the bridge defect risk analysis strategy to obtain the defect risk status of the central single cable surface of the cable-stayed bridge. S5. Based on the condition of the defects and hazards of the central single cable surface of the cable-stayed bridge, maintain the central single cable surface of the cable-stayed bridge.

[0007] In a preferred embodiment of the present invention, the strategy for analyzing the evolution of rust disease status in step S2 specifically includes: S21. Extract the spectral reflectance characteristics of the cable sheath material under different bands from the multispectral imaging data of the central single cable plane of the cable-stayed bridge, and calculate the corrosion characteristic spectral index of each cable; simultaneously obtain the nominal diameter of each cable from the bridge structural parameters. S22. Based on the spectral index of the corrosion characteristics of each cable and the wave transmission characteristics of the sheath material, a quantitative identification model of corrosion products on the inner wall of the sheath is established, and the type and equivalent thickness of corrosion products of each cable are obtained by inversion calculation. S23. Correlate the equivalent thickness of the corrosion products of each cable obtained by inversion with the nominal diameter of the cable obtained through bridge structural parameters, and use the ratio of the equivalent thickness of the corrosion products to the nominal diameter of the cable as the basic parameter of the cross-sectional loss rate of each cable; at the same time, calculate the non-uniform corrosion risk index of each cable based on the equivalent thickness of the corrosion products. S24. The basic parameters of the cross-sectional loss rate of each cable are weighted and summed with the non-uniform corrosion risk index to obtain the evolution of corrosion disease status of each cable in the central single cable plane of the cable-stayed bridge.

[0008] In a preferred embodiment of the present invention, the strategy for analyzing the evolution of the bearing influence in step S3 specifically includes: S31. Extract the average daily traffic volume, average axle load, and duration of peak traffic periods for the upstream and downstream lanes of the bridge during the monitoring period from the two-way lane traffic flow monitoring data. S32. Calculate the traffic flow asymmetry coefficient, average axle load difference coefficient, and peak load duration difference coefficient in the upstream and downstream directions. S33. Obtain the spatial arrangement coordinates, cable surface inclination angle, and stiffness distribution data of each cable from the bridge structural parameters, establish a planar truss model, and calculate the cable force influence coefficient of each cable under a unit asymmetric load. S34. Correlate the traffic flow asymmetry coefficient, average axle load difference coefficient, and peak load duration difference coefficient with the corresponding cable force influence coefficient to calculate the periodic asymmetric load strength index of each cable. S35. Divide the periodic asymmetric load strength index of each cable by the cumulative amount of the periodic asymmetric load strength index within the monitoring period to obtain the evolution of the load-bearing influence of each cable.

[0009] In a preferred embodiment of the present invention, the bridge defect risk analysis strategy in step S4 specifically includes: S41. Obtain the evolution of the corrosion disease status of each cable calculated in step S24, and the evolution of the load-bearing impact of each cable calculated in step S35. S42. The weighted sum of the evolution of corrosion and load-bearing impact of each cable is used to obtain the bridge defect risk for each cable. S43. The bridge defect risk of each cable is weighted and combined according to the position weight of each cable in the entire central single cable plane to calculate the defect risk status of the central single cable plane of the cable-stayed bridge.

[0010] In a preferred embodiment of the present invention, step S5 involves maintaining the central single cable face of the cable-stayed bridge according to the defect and danger status of the cable face, specifically including: Extract the disease risk status of the central single cable surface of the cable-stayed bridge calculated in step S43, and preset the bridge disease threat threshold; when the disease risk status of the central single cable surface of the cable-stayed bridge is less than the bridge disease threat threshold, continue to monitor the central single cable surface of the cable-stayed bridge; when the disease risk status of the central single cable surface of the cable-stayed bridge is greater than or equal to the bridge disease threat threshold, perform maintenance on the central single cable surface of the cable-stayed bridge.

[0011] In a preferred embodiment of the present invention, when the defect risk state of the central single cable plane of a cable-stayed bridge is greater than or equal to the bridge defect threat threshold, maintenance is carried out on the central single cable plane of the cable-stayed bridge, specifically including: S51. Obtain the bridge defect risk of each cable and the corresponding position weight of the cable in the entire central single cable plane, and calculate the defect risk status of each cable. S52. The defect and danger status of each cable is taken as the maintenance urgency of each cable. All cables in the central single cable plane are sorted in descending order according to the maintenance urgency index of each cable to obtain the order of maintenance operations for the central single cable plane of the cable-stayed bridge. All cables in the central single cable plane are maintained according to the order of maintenance operations.

[0012] In a second aspect, the present invention provides an electronic device, comprising: a processor and a memory, wherein the memory stores a computer program that can be called by the processor, and the processor executes a bridge defect identification method based on multispectral imaging by calling the computer program stored in the memory.

[0013] Compared with the prior art, the present invention has the following advantages and beneficial effects: 1. This invention integrates cross-sectional loss rate and uneven corrosion risk index to achieve quantitative evolution analysis of corrosion disease state, improves the quantitative accuracy of corrosion disease, reduces the risk of misjudging the severity of disease caused by qualitative judgment, and enhances the ability to accurately identify early corrosion hazards. 2. This invention extracts the asymmetric features of traffic flow, combines cable parameters to model and calculate the cable force influence coefficient, couples load asymmetry with cable sensitivity, quantifies load intensity and evolution trend, and reduces the evaluation bias caused by isolated load analysis. 3. This invention introduces location weights from structural sensitivity analysis to synthesize the overall risk of the cable surface, which reflects both the hidden dangers of individual cables and highlights the impact of key cables, improving the systematic comprehensiveness of risk assessment, reducing the risk of missing hidden dangers in key locations due to average assessment, and enhancing the ability to accurately judge the risks of the overall cable surface and individual cables. Attached Figure Description

[0014] Other features, objects, and advantages of the invention will become more apparent from the following detailed description of non-limiting embodiments with reference to the accompanying drawings: Figure 1This is a schematic diagram of the overall process of the bridge defect identification method based on multispectral imaging of the present invention; Figure 2 This is a schematic diagram of the central single cable plane of a cable-stayed bridge in the bridge defect identification method based on multispectral imaging of the present invention; Figure 3 This is a flowchart of the data acquisition process for step S21 of the bridge defect identification method based on multispectral imaging of the present invention. Figure 4 This is a data analysis flowchart for step S22 of the bridge defect identification method based on multispectral imaging of the present invention; Figure 5 This is a data analysis flowchart for step S24 of the bridge defect identification method based on multispectral imaging of the present invention. Detailed Implementation

[0015] The technical solution of the present invention will be described in detail below with reference to the accompanying drawings and specific embodiments. It should be understood that the embodiments of the present invention and the specific features in the embodiments are detailed descriptions of the technical solution of the present invention, rather than limitations thereof. In the absence of conflict, the embodiments of the present invention and the technical features in the embodiments can be combined with each other.

[0016] Example 1 like Figure 1 As shown, this embodiment provides a bridge defect identification method based on multispectral imaging, which specifically includes the following steps: S1. Acquire multispectral imaging data of the central single cable plane of the cable-stayed bridge, bridge structural parameters, and two-way lane traffic flow monitoring data; Step S1 specifically includes: S11, such as Figure 2 As shown, multispectral imaging data of the central single cable plane of the cable-stayed bridge was acquired. A UAV equipped with a high-resolution multispectral camera was used as the data acquisition device. The camera's wavelength coverage included the visible light (400-700nm), near-infrared (700-1100nm), and short-wave infrared (1100-2500nm) regions, with a resolution of no less than 0.1m. Before acquisition, the flight path was planned according to the span and cable plane height of the cable-stayed bridge, ensuring the path was parallel to the central single cable plane. The flight altitude was controlled within the range that allowed for clear imaging of the complete length of each cable, and at least three images with an overlap of no less than 60% were captured for each cable. After acquisition, the raw images were imported into multispectral image processing software, where radiometric calibration (to eliminate sensor errors), atmospheric correction (to remove atmospheric effects based on the dark target method), and geometric correction (to correct image distortion using the bridge tower embedded parts as control points) were performed sequentially. Finally, a standardized multispectral imaging dataset for each cable was output.

[0017] S12. Obtain the bridge structural parameters of the central single cable plane of the cable-stayed bridge. Prioritize retrieving parameters from bridge design drawings or the bridge management information system. Core parameters include the nominal diameter of each cable, spatial layout coordinates (three-dimensional coordinates of the connection nodes between the cable ends and the towers and main beams), cable plane inclination angle, and main beam stiffness distribution data. Furthermore, in this embodiment, if the design drawings are missing or the parameters are deviated, on-site measurements are used to supplement them: a laser diameter gauge is used to measure the diameter at the top, middle, and bottom positions of each cable, and the average value is taken as the nominal diameter; an ultrasonic thickness gauge is used to select 5 measuring points evenly on the sheath surface, and the thickness is calculated by the propagation speed and reflection time of ultrasonic waves in the sheath material, and the average value is taken as the thickness; a total station is used to perform three-dimensional coordinate measurements on the cable connection nodes to obtain the spatial arrangement coordinates; an inclinometer is used to measure the angle between the cable surface and the vertical direction, and the average value of the three measurements is taken as the cable surface inclination angle; a dynamic loading test is used to apply standard loads at different positions of the main beam, measure the deflection at the corresponding positions, and calculate the stiffness distribution data of the main beam using material mechanics formulas.

[0018] S13. Acquire traffic flow monitoring data for both directions of the bridge. Piezoelectric axle load meters, microwave traffic flow counters, and high-definition video surveillance equipment are deployed on both sides of the upstream and downstream lanes of the bridge, respectively. The monitoring period is no less than 30 days, and the sampling frequency is set to 1Hz. The piezoelectric axle load meters are used to collect axle load data for each passing vehicle, the microwave traffic flow counter records the traffic flow per unit time, and the high-definition video surveillance equipment is used to identify peak traffic periods (in this embodiment, a peak traffic period is defined as a period where the traffic flow exceeds 1.5 times the daily average traffic flow within 15 consecutive minutes). After data collection, the data is preprocessed: abnormal data caused by axle load meter zero-point drift (data with axle load less than 0.5t) and falsely triggered data from the traffic flow counter (data with a single trigger duration less than 0.1 seconds) are removed. The start and end times of peak periods are calibrated based on the video surveillance data. Finally, the daily average traffic flow, average axle load, and peak period duration data for both upstream and downstream lanes are output.

[0019] It should be noted that step S1 achieves comprehensive coverage of the basic data required for disease identification by acquiring multi-source data through UAV remote sensing, instrument measurement, and system retrieval. Among them, multispectral imaging data reflects the material state of the cables, bridge structural parameters provide a mechanical analysis benchmark, and traffic flow monitoring data characterizes the external load characteristics. The combination of these three lays a reliable data foundation for subsequent corrosion disease analysis and load-bearing impact analysis, effectively avoiding the one-sidedness of analysis caused by a single data source.

[0020] S2. Import the multispectral imaging data and bridge structural parameters into the corrosion disease state evolution analysis strategy to conduct corrosion disease state evolution analysis and obtain the corrosion disease state evolution of the central single cable surface of the cable-stayed bridge. S3. Import the two-way traffic flow monitoring data and bridge structural parameters into the load-bearing impact evolution analysis strategy to conduct load-bearing impact evolution analysis and obtain the load-bearing impact evolution of the central single cable surface of the cable-stayed bridge. S4. Obtain the evolution of corrosion and load-bearing impact on the central single cable surface of the cable-stayed bridge, and import the bridge defect risk analysis strategy to obtain the defect risk status of the central single cable surface of the cable-stayed bridge. S5. Based on the condition of the defects and hazards of the central single cable surface of the cable-stayed bridge, maintain the central single cable surface of the cable-stayed bridge.

[0021] In this embodiment, the strategy for analyzing the evolution of rust disease status in step S2 specifically includes: S21, such as Figure 3 As shown, from the multispectral imaging data of the central single cable surface of the cable-stayed bridge, the spectral reflectance characteristics of the cable sheath material under different bands are extracted. Specifically, in this embodiment, from the multispectral imaging data processed in S11, for the sheath area of ​​each cable, the region of interest (ROI) extraction method is used. A 10*10 pixel area without obvious stains is selected on the sheath surface, and the reflectance values ​​of each preset band (e.g., visible light red band 650-680nm, near-infrared band 900-950nm) are extracted respectively. The average reflectance of each band is calculated as the spectral reflectance characteristic of the cable sheath in the corresponding band. The corrosion characteristic spectral index of each cable is calculated. Specifically, the calculation strategy of the corrosion characteristic spectral index is: subtract the red band spectral reflectance characteristic from the near-infrared band spectral reflectance characteristic of each cable, and then divide by the sum of the two to obtain the corrosion characteristic spectral index of each cable. Its value increases with the deepening of the corrosion of the inner wall of the sheath. The nominal diameter of each cable is obtained simultaneously from the bridge structural parameters. It should be noted that the red light band is sensitive to corrosion products, while the near-infrared band is stable in response to the sheath substrate. The corrosion characteristic spectral index calculated by the two can effectively remove the interference of stains on the sheath surface and accurately reflect the corrosion state of the inner wall. The nominal diameter of the cable, as a structural benchmark parameter, provides a scale reference for the subsequent quantitative correlation of corrosion degree.

[0022] S22, such as Figure 4As shown, based on the spectral index of corrosion characteristics of each cable and the transmittance characteristics of the sheath material, a quantitative identification model for corrosion products on the inner wall of the sheath is established, and the type and equivalent thickness of corrosion products of each cable are calculated by inversion. Specifically, this embodiment establishes a quantitative identification model and inverts the parameters of corrosion products. Standard samples of cable sheaths with different degrees of corrosion are collected (in this embodiment, the degree of corrosion covers four levels: no corrosion, light corrosion, moderate corrosion, and heavy corrosion). In a laboratory environment, the spectral reflectance characteristics of the samples are measured using the same multispectral camera as S11. The spectral index of corrosion characteristics of the samples in the laboratory environment is calculated. At the same time, the type of corrosion products in the samples (e.g., iron oxide, iron hydroxide, etc.) is observed and determined using a scanning electron microscope. The actual thickness of the corrosion products is measured using a micrometer screw gauge and converted into an equivalent thickness (in this embodiment, the equivalent thickness is the average thickness of the corrosion products at different locations). The transmission characteristics of the sheath material are obtained (in this embodiment, the transmission characteristics of the sheath material are obtained by placing a sample of the sheath material in a spectrometer and measuring the transmittance of light in each wavelength band; that is, the transmission characteristics of the sheath material are the average transmittance of light in all wavelength bands). A linear fitting model is established with the rust characteristic spectral index and transmission characteristics as inputs and the rust product type and equivalent thickness as outputs. The model fitting process is achieved using the least squares method. The rust characteristic spectral index of each cable and the corresponding transmission characteristics of the sheath material obtained in S21 are substituted into this model, and the rust product type and equivalent thickness of each cable are calculated by inversion. It should be noted that in this embodiment, step S22 establishes a quantitative relationship through standard sample calibration, transforming the qualitative characteristics of multispectral data into quantitative parameters of rust products. This overcomes the deficiency of traditional manual visual inspection in being unable to quantify the degree of rust, and provides accurate physical parameter support for subsequent analysis of rust disease evolution.

[0023] S23. The equivalent thickness of the corrosion products of each cable obtained through inversion is correlated with the nominal diameter of the cable obtained through bridge structural parameters. The ratio of the equivalent thickness of the corrosion products to the nominal diameter of the cable is used as the basic parameter for the cross-sectional loss rate of each cable, thereby reflecting the potential for cable cross-sectional reduction caused by corrosion. At the same time, the non-uniform corrosion risk index of each cable is calculated based on the equivalent thickness of the corrosion products. Specifically, in this embodiment, the calculation strategy for the non-uniform corrosion risk index is as follows: 10 monitoring points are uniformly selected along the length of the cable. The equivalent thickness of the corrosion products at each point is obtained through inversion using the quantitative identification model of the corrosion products on the inner wall of the sheath in S22. The standard deviation of the equivalent thickness at these 10 points is calculated. The standard deviation is divided by the total length of the cable to obtain the non-uniform corrosion risk index of each cable. The larger the value, the more uneven the distribution of corrosion along the length of the cable, and the higher the risk of local stress concentration. It should be noted that the cross-sectional loss rate, a fundamental parameter, is directly related to the degree of load-bearing capacity attenuation in cables and is a core indicator of the impact of corrosion. The non-uniform corrosion risk index accurately reflects the spatial distribution characteristics of corrosion, and its value is positively correlated with the risk of local stress concentration. A larger value indicates more significant differences in local corrosion thickness within the cable, making it prone to forming weak areas such as corrosion pits and bands. Under load, these areas will bear stress far exceeding the average level, thus increasing the risk of sudden fracture. Combining these two factors allows for a comprehensive characterization of the morphology and extent of corrosion from two dimensions: overall load-bearing capacity deterioration and local sudden fracture risk. Furthermore, the standard deviation quantifies the dispersion of the equivalent corrosion thickness along the cable length. A larger standard deviation indicates a more uneven corrosion distribution, making it easier to form local stress concentration points, which aligns with the theory of stress concentration risk in engineering. This embodiment uses division by the total cable length for standardization, eliminating the influence of cable dimensions and ensuring comparability between cables of different lengths. After standardization, the index reflects the non-uniformity per unit length, which is more consistent with the fairness of risk assessment.

[0024] S24, such as Figure 5As shown, the basic parameter of cross-sectional loss rate of each cable is weighted and summed with the non-uniform corrosion risk index to obtain the evolution of corrosion disease status of each cable in the central single cable plane of the cable-stayed bridge. Specifically, this embodiment can determine the weight of the basic parameter of cross-sectional loss rate and the weight of the non-uniform corrosion risk index based on the statistical analysis of historical data in the bridge corrosion disease database (the sum of the weights of the two is 1, and the weight value is determined by comparing the degree of consistency between the evaluation results under different weight combinations and the actual corrosion evolution; the combination with the highest degree of consistency is the optimal weight; for example, in this embodiment, the weight of the basic parameter of cross-sectional loss rate is 0.6 by default, and the weight of the non-uniform corrosion risk index is 0.4 by default; among them, the weight of the basic parameter of cross-sectional loss rate is higher because it directly reflects the core attenuation degree of the cable bearing capacity, while the non-uniform corrosion risk index focuses on local hidden dangers, which need to be taken into account but have a slightly lower priority).

[0025] In this embodiment, the specific strategy for calculating the evolution of corrosion disease status is as follows: multiply the basic parameter of cross-sectional loss rate by 0.6, and add the uneven corrosion risk index multiplied by 0.4 to obtain the evolution of corrosion disease status for each cable. The larger the value, the more significant the development of corrosion disease.

[0026] It should be noted that step S24 in this embodiment achieves a comprehensive assessment of the overall deterioration of the load-bearing capacity due to corrosion (basic parameter of cross-sectional loss rate) and the level of local sudden risk (non-uniform corrosion risk index) through weighted summation: the basic parameter of cross-sectional loss rate quantifies the decrease in the overall load-bearing capacity of the cable due to corrosion, while the non-uniform corrosion risk index captures the hidden dangers of local stress concentration. The combination of the two can effectively avoid the limitations of a single indicator that only considers the overall amount of corrosion while ignoring the risk of local fracture or only focuses on local distribution while ignoring the overall deterioration. The output of the evolution of the corrosion disease state can dynamically reflect the trend of the superposition of the dual risks of the disease, providing a time-series dimension for subsequent dangerous state analysis that combines overall strength and local stability.

[0027] In this embodiment, the strategy for analyzing the evolution of the bearing influence in step S3 specifically includes: S31. From the bidirectional lane traffic flow monitoring data, extract the daily average traffic volume, average axle load, and peak traffic duration for the upstream and downstream lanes of the bridge within the monitoring period. Specifically, in this embodiment, from the bidirectional lane traffic flow monitoring data preprocessed in S13, the traffic volume of the upstream and downstream lanes is statistically analyzed on a daily basis, and the average value within the monitoring period is taken as the daily average traffic volume. The axle load data of all vehicles passing through each lane is statistically analyzed, and the average value after removing the maximum and minimum values ​​is taken as the average axle load. Based on the peak period records calibrated by video surveillance, the duration of the daily peak period for each lane is statistically analyzed, and the average value within the monitoring period is taken as the peak period duration. It should be noted that the daily average traffic volume, average axle load, and peak period duration characterize traffic flow characteristics from three dimensions: flow rate, load intensity, and duration of action. These three parameters are the core external factors affecting the cable bearing capacity and can provide basic data input for subsequent asymmetric load analysis.

[0028] S32. Calculate the traffic flow asymmetry coefficient, average axle load difference coefficient, and peak load duration difference coefficient in the upstream and downstream directions. Specifically, in this embodiment, the calculation strategy for the traffic flow asymmetry coefficient is as follows: the absolute difference between the average daily traffic flow of the upstream lane and the average daily traffic flow of the downstream lane is divided by the maximum value of the two to obtain the traffic flow asymmetry coefficient in the upstream and downstream directions. The larger the traffic flow asymmetry coefficient, the stronger the traffic flow asymmetry. Specifically, in this embodiment, the calculation strategy for the average axle load difference coefficient is as follows: the absolute difference between the average axle load of the upstream lane and the average axle load of the downstream lane is divided by the maximum value of the two to obtain the average axle load difference coefficient in the upstream and downstream directions, which is used to reflect the degree of difference in load intensity between the upstream and downstream. Specifically, in this embodiment, the calculation strategy for the peak load duration difference coefficient is as follows: the absolute difference between the peak duration of the upstream lane and the peak duration of the downstream lane is divided by the maximum of the two to obtain the peak load duration difference coefficient in the upstream and downstream directions, which is used to characterize the difference in the duration of asymmetrical load action.

[0029] It should be noted that, under actual working conditions, the cable stress on the central single cable plane of a cable-stayed bridge is extremely sensitive to the asymmetry of two-way traffic flow. The above three coefficients, by quantifying the asymmetric characteristics of two-way traffic flow, can provide a direct indicator for analyzing the unevenness of cable stress.

[0030] S33. Obtain the spatial arrangement coordinates, cable surface inclination angle, and main beam stiffness distribution data of each cable from the bridge structural parameters, establish a planar truss model, and calculate the cable force influence coefficient of each cable under a unit asymmetric load. Specifically, in this embodiment, the spatial arrangement coordinates, cable surface inclination angle, and main beam stiffness distribution data of each cable are obtained from S12. These parameters are imported into professional structural mechanics analysis software (in this embodiment, the professional structural mechanics analysis software can be Midas-Civil software) to establish a planar truss model of the central single cable surface of the cable-stayed bridge. The cables are simplified to tension-only rod elements, the main beam is simplified to beam elements, and the towers are simplified to fixed supports. The model boundary conditions are set according to the actual support conditions of the bridge. A unit asymmetric load is applied to the model (in this embodiment, the unit asymmetric load can be defined as a 1kN concentrated load applied at the mid-span of the upstream lane, with no load applied to the downstream lane, thus the unit asymmetric load is 1kN). The change in cable force for each cable under the unit asymmetric load is calculated. The ratio of this change in cable force to the unit asymmetric load is used as the cable force influence coefficient for each cable under the unit asymmetric load. The larger the value, the more sensitive the cable is to the asymmetric load. It should be noted that in this embodiment, step S33 correlates the asymmetric characteristics of traffic flow with the mechanical response of the cables through structural mechanics modeling. The cable force influence coefficient quantifies the differences in sensitivity of different cables to asymmetric loads, laying a mechanical foundation for the subsequent accurate assessment of the load-bearing impact of each cable.

[0031] S34. The traffic flow asymmetry coefficient, average axle load difference coefficient, and peak load duration difference coefficient are associated with the cable force influence coefficient of the corresponding cable. Specifically, in this embodiment, the association strategy is as follows: the traffic flow asymmetry coefficient, average axle load difference coefficient, and peak load duration difference coefficient are arithmetically averaged, and the arithmetic average is multiplied by the cable force influence coefficient of each cable under a unit asymmetric load. The multiplication is used as the periodic asymmetric load strength index of each cable. S35. Divide the periodic asymmetric load strength index of each cable by the cumulative amount of the periodic asymmetric load strength index within the monitoring period to obtain the evolution of the load-bearing impact of each cable. Specifically, in this embodiment, the periodic asymmetric load strength index of each cable is calculated daily within the monitoring period, and the sum of the periodic asymmetric load strength indices of each cable for all days up to the previous day within the monitoring period is calculated as the cumulative amount of the periodic asymmetric load strength index of that cable up to the previous day. This avoids including the index of the current day in the cumulative amount in advance and eliminates the logical contradiction of repeated statistics. The calculation strategy for the evolution of the load-bearing impact of each cable is as follows: First, divide the cumulative amount of the periodic asymmetric load strength index up to the previous day by the number of monitoring days up to the previous day to obtain the average strength index up to the previous day; then, perform a division operation using the difference between the periodic asymmetric load strength index of the current day and the average strength index up to the previous day, and add the value 1 to the division operation result to obtain the evolution of the load-bearing impact of each cable. Specifically, in this embodiment: When the load-bearing impact evolution of each cable is greater than 1, it indicates that the load-bearing impact of each cable on that day is higher than the average level up to the previous day in the monitoring period, and the load-bearing impact shows an intensifying evolution trend. When the load-bearing impact evolution of each cable is equal to 1, it indicates that the load-bearing impact of each cable on that day is the same as the historical average level, and the load-bearing impact shows a stable evolution trend. When the load-bearing impact evolution of each cable is less than 1, it indicates that the load-bearing impact of each cable on that day is lower than the historical average level, and the load-bearing impact shows a slowing evolution trend.

[0032] It should be noted that step S35 of this embodiment uses the historical average level as a benchmark to compare the data of the day with the previous trend, which truly realizes the time-series dynamic analysis of the evolution. This makes the analysis results of this embodiment not only reflect the absolute size of the load-bearing impact on the day, but also accurately capture its rising and falling direction relative to the historical trend, providing key basis for the dynamic trend of load impact for subsequent dangerous state analysis.

[0033] It should be noted that step S3 in this embodiment deeply integrates traffic flow monitoring data with the structural mechanics model, constructing a technical path from traffic flow feature extraction, asymmetry quantification, to mechanical response coupling, and finally reflecting the evolution trend analysis. This breaks through the limitation of traditional load analysis relying solely on static structural parameters, and realizes dynamic and accurate assessment of the impact of external loads on cables, providing a basis for load evolution for subsequent coupled analysis of disease and danger states.

[0034] In this embodiment, the bridge defect risk state analysis strategy in step S4 specifically includes: S41. Obtain the evolution of the corrosion disease status of each cable calculated in step S24, and the evolution of the load-bearing impact of each cable calculated in step S35. S42. The weighted sum of the evolution of corrosion and load-bearing impact of each cable is used to obtain the bridge defect risk for each cable. S43. The bridge defect risk of each cable is weighted and synthesized according to the position weight of each cable in the entire central single cable plane to calculate the defect risk status of the cable-stayed bridge's central single cable plane. Specifically, in this embodiment, the position weight is determined by combining structural sensitivity analysis with actual engineering logic. The specific process is as follows: Based on the previously obtained data such as the spatial arrangement position of the cables, the inclination angle of the cable plane, and the stiffness distribution of the main beam, a planar rod system model of the central single cable plane of the cable-stayed bridge is built in professional structural mechanics analysis software. In the model, the cables are simplified into rod-like units that only bear tensile force, the main beam is simplified into beam-like units that bear bending and shear, and the towers are simplified into fixed supports according to the actual support situation. The stiffness of the target cable is reduced to simulate the extreme case of cable performance weakening, while maintaining the stiffness of other cables. With the cable and structural parameters unchanged, a standard uniformly distributed load is applied to the model according to the bridge design specifications. The deflection variation at the mid-span of the main girder is monitored using displacement measurement equipment. The deflection variation at the mid-span of the main girder corresponding to each cable is statistically analyzed. The deflection variation of a single cable is compared with the largest deflection variation among all cables, and the ratio obtained is the position weight of that cable. Furthermore, in this embodiment, the larger the deflection variation of a cable, the more critical its supporting role in the overall stiffness of the cable surface. Once its performance weakens, the impact on the overall deformation of the bridge is more significant, and therefore, the higher its position weight. For example, the deflection variation of cables in the mid-span area of ​​a cable-stayed bridge is usually the largest, and the position weight is mostly between 0.9 and 1.0; while the deflection variation of cables near the tower area is relatively smaller, and the position weight is mostly between 0.5 and 0.7. The strategy for calculating the bridge defect risk status is as follows: multiply the bridge defect risk of each cable by the position weight of the corresponding cable in the entire central single cable plane to obtain the defect risk status of each cable; sum the defect risk statuses of all cables in the central single cable plane of the cable-stayed bridge, divide the sum by the sum of the position weights of all cables in the central single cable plane of the cable-stayed bridge, and use the result of the division as the defect risk status of the central single cable plane of the cable-stayed bridge.

[0035] In this embodiment, step S5 involves maintaining the central single cable plane of the cable-stayed bridge according to its defect and danger status, specifically including: Extract the disease risk status of the central single cable surface of the cable-stayed bridge calculated in step S43, and preset the bridge disease threat threshold; when the disease risk status of the central single cable surface of the cable-stayed bridge is less than the bridge disease threat threshold, continue to monitor the central single cable surface of the cable-stayed bridge; when the disease risk status of the central single cable surface of the cable-stayed bridge is greater than or equal to the bridge disease threat threshold, perform maintenance on the central single cable surface of the cable-stayed bridge.

[0036] In this embodiment, when the defect risk status of the central single cable plane of the cable-stayed bridge is greater than or equal to the bridge defect threat threshold, maintenance is carried out on the central single cable plane of the cable-stayed bridge, specifically including: S51. Obtain the bridge defect risk of each cable and the corresponding position weight of the cable in the entire central single cable plane, and calculate the defect risk status of each cable. S52. The defect risk status of each cable is used as the maintenance urgency of each cable. All cables in the central single-cable plane are sorted in descending order according to the maintenance urgency index of each cable to obtain the order of maintenance operations for the central single-cable plane of the cable-stayed bridge. Maintenance is then performed on all cables in the central single-cable plane according to this order. It should be noted that the parameters set in this embodiment (such as weights and thresholds) need to be set by those skilled in the art based on relevant experiments. Specifically, the experimental method is as follows: Multispectral imaging data of the central single-cable plane of the cable-stayed bridge, bridge structural parameters, and two-way lane traffic flow monitoring data are obtained and substituted into the steps of this embodiment to calculate the defect risk status of the central single-cable plane. The calculated defect risk status of the central single-cable plane is obtained. The defect risk status of the central single-cable plane and the actual defect risk judgment results of the central single-cable plane are imported into fitting software for continuous fitting. The output values ​​of the set parameters (such as weights and thresholds) that match the defect risk status with the actual defect risk judgment results are then obtained.

[0037] Example 2 An electronic device according to an embodiment of the present invention includes a processor and a memory, wherein the memory stores a computer program that can be called by the processor, and the processor executes a bridge defect identification method based on multispectral imaging by calling the computer program stored in the memory. It should be noted that all computer programs for the bridge defect identification method based on multispectral imaging are implemented using the C language.

[0038] In the description of this specification, references to terms such as "an embodiment," "example," "specific example," etc., indicate that a specific feature, structure, material, or characteristic described in connection with that embodiment or example is included in at least one embodiment or example of the invention. In this specification, illustrative expressions of the above terms do not necessarily refer to the same embodiment or example. Furthermore, the specific features, structures, materials, or characteristics described may be combined in any suitable manner in one or more embodiments or examples.

[0039] The preferred embodiments of the present invention disclosed above are merely illustrative of the invention. These preferred embodiments do not exhaustively describe all details, nor do they limit the invention to the specific implementations described. Clearly, many modifications and variations can be made based on the content of this specification. This specification selects and specifically describes these embodiments to better explain the principles and practical applications of the invention, thereby enabling those skilled in the art to better understand and utilize the invention. The invention is limited only by the claims and their full scope and equivalents.

Claims

1. A method for identifying bridge defects based on multispectral imaging, characterized in that, Includes the following steps: S1. Acquire multispectral imaging data of the central single cable plane of the cable-stayed bridge, bridge structural parameters, and two-way lane traffic flow monitoring data; S2. Import the multispectral imaging data and bridge structural parameters into the corrosion disease state evolution analysis strategy to conduct corrosion disease state evolution analysis and obtain the corrosion disease state evolution of the central single cable surface of the cable-stayed bridge. S3. Import the two-way traffic flow monitoring data and bridge structural parameters into the load-bearing impact evolution analysis strategy to conduct load-bearing impact evolution analysis, and obtain the load-bearing impact evolution of the central single cable plane of the cable-stayed bridge; specifically including: S33. Obtain the spatial arrangement coordinates, cable surface inclination angle, and stiffness distribution data of each cable from the bridge structural parameters, establish a planar truss model, and calculate the cable force influence coefficient of each cable under a unit asymmetric load. S34. Correlate the traffic flow asymmetry coefficient, average axle load difference coefficient, and peak load duration difference coefficient with the corresponding cable force influence coefficient to calculate the periodic asymmetric load strength index of each cable. S35. Divide the periodic asymmetric load strength index of each cable by the cumulative amount of the periodic asymmetric load strength index within the monitoring period to obtain the evolution of the load-bearing influence of each cable. S4. Obtain the evolution of corrosion and load-bearing impact on the central single cable face of the cable-stayed bridge, and import the bridge defect risk analysis strategy to obtain the defect risk status of the central single cable face; specifically including: S42. The weighted sum of the evolution of corrosion and load-bearing impact of each cable is used to obtain the bridge defect risk for each cable. S43. The bridge defect risk of each cable is weighted and combined according to the position weight of each cable in the entire central single cable plane to calculate the defect risk status of the central single cable plane of the cable-stayed bridge. S5. Based on the condition of the defects and hazards of the central single cable surface of the cable-stayed bridge, maintain the central single cable surface of the cable-stayed bridge. The specific strategies for analyzing the evolution of rust disease status in step S2 include: S21. Extract the spectral reflectance characteristics of the cable sheath material under different bands from the multispectral imaging data of the central single cable plane of the cable-stayed bridge, and calculate the corrosion characteristic spectral index of each cable; simultaneously obtain the nominal diameter of each cable from the bridge structural parameters. S22. Based on the spectral index of the corrosion characteristics of each cable and the wave transmission characteristics of the sheath material, a quantitative identification model of corrosion products on the inner wall of the sheath is established, and the type and equivalent thickness of corrosion products of each cable are obtained by inversion calculation. S23. Correlate the equivalent thickness of the corrosion products of each cable obtained by inversion with the nominal diameter of the cable obtained through bridge structural parameters, and use the ratio of the equivalent thickness of the corrosion products to the nominal diameter of the cable as the basic parameter of the cross-sectional loss rate of each cable; at the same time, calculate the non-uniform corrosion risk index of each cable based on the equivalent thickness of the corrosion products. S24. The basic parameters of the cross-sectional loss rate of each cable are weighted and summed with the non-uniform corrosion risk index to obtain the evolution of corrosion disease status of each cable in the central single cable plane of the cable-stayed bridge.

2. The bridge defect identification method based on multispectral imaging according to claim 1, characterized in that, The specific strategies for analyzing the evolution of bearing capacity impact in step S3 include: S31. Extract the average daily traffic volume, average axle load, and duration of peak traffic periods for the upstream and downstream lanes of the bridge during the monitoring period from the two-way lane traffic flow monitoring data. S32. Calculate the traffic flow asymmetry coefficient, average axle load difference coefficient, and peak load duration difference coefficient in the upstream and downstream directions.

3. The bridge defect identification method based on multispectral imaging according to claim 2, characterized in that, The bridge defect risk status analysis strategy in step S4 specifically includes: S41. Obtain the evolution of the corrosion disease status of each cable calculated in step S24, and the evolution of the load-bearing impact of each cable calculated in step S35.

4. The bridge defect identification method based on multispectral imaging according to claim 3, characterized in that, Step S5 involves maintaining the central single cable face of the cable-stayed bridge according to its defect and danger status. Specifically, this includes: Extract the disease risk status of the central single cable surface of the cable-stayed bridge calculated in step S43, and preset the bridge disease threat threshold; when the disease risk status of the central single cable surface of the cable-stayed bridge is less than the bridge disease threat threshold, continue to monitor the central single cable surface of the cable-stayed bridge; when the disease risk status of the central single cable surface of the cable-stayed bridge is greater than or equal to the bridge disease threat threshold, perform maintenance on the central single cable surface of the cable-stayed bridge.

5. The bridge defect identification method based on multispectral imaging according to claim 4, characterized in that, When the defect risk status of the central single cable plane of a cable-stayed bridge is greater than or equal to the bridge defect threat threshold, maintenance is carried out on the central single cable plane of the cable-stayed bridge, specifically including: S51. Obtain the bridge defect risk of each cable and the corresponding position weight of the cable in the entire central single cable plane, and calculate the defect risk status of each cable. S52. The defect and danger status of each cable is taken as the maintenance urgency of each cable. All cables in the central single cable plane are sorted in descending order according to the maintenance urgency index of each cable to obtain the order of maintenance operations for the central single cable plane of the cable-stayed bridge. All cables in the central single cable plane are maintained according to the order of maintenance operations.

6. An electronic device, comprising: A processor and a memory, wherein the memory stores a computer program that can be called by the processor; characterized in that the processor executes the bridge defect identification method based on multispectral imaging as described in any one of claims 1-5 by calling the computer program stored in the memory.

Images