A method for detecting fireproof and heat insulation performance of a coating of a main cable of a suspension bridge

Abstract

This invention belongs to the field of performance analysis technology and discloses a method for detecting the fire resistance and heat insulation performance of the coating of the main cable of a suspension bridge. The method includes: thermally exciting the surface of the main cable coating, collecting the temperature sequence of the main cable coating surface, analyzing and obtaining uniformity characteristics, and dividing suspected areas according to the uniformity characteristics; performing secondary thermal excitation and synchronous monitoring in the suspected areas, obtaining temperature sequences and extracting and identifying them to obtain thermal anomaly zones and corresponding channel parameters; performing three-level thermal excitation tests on the thermal anomaly zones, recording the corresponding temperature sequences and corresponding temperature decay curves, and analyzing and obtaining channel depths; combining the channel parameters and channel depths of the thermal anomaly zones for analysis to obtain the fire resistance performance degradation rate; and making a comprehensive judgment based on the channel parameters, channel depth, and fire resistance performance degradation rate to obtain the detection result. This invention realizes integrated detection and early warning of channel effects from early sign detection and geometric feature extraction to quantification of the impact on fire resistance performance.

Description

Technical Field

[0001] This invention relates to the field of performance analysis technology, and more specifically, to a method for testing the fireproof and heat-insulating performance of a coating used in the main cable of a suspension bridge. Background Technology

[0002] As a primary structural form of long-span bridges, suspension bridges rely heavily on their main cables, which are the core load-bearing components. Damage to these cables in a fire could lead to catastrophic structural failure. To ensure the safety of the main cables under fire conditions, multiple layers of fire-resistant and heat-insulating coatings are typically applied to their surface to slow temperature rise and protect the steel wires from softening at high temperatures. However, in actual engineering environments, factors such as construction techniques, coating aging, mechanical damage, and marine salt spray corrosion can cause microscopic weak areas or directional degradation paths to form within the coating. This localized degradation induced by geometric defects or gradient interfaces significantly reduces the overall fire-resistant and heat-insulating performance of the coating, allowing heat or corrosive media to rapidly penetrate into the main cable along specific channels, creating the so-called channel effect.

[0003] Existing fire resistance and thermal insulation performance testing methods mostly focus on coating surface quality inspection or overall thermal performance testing, such as back temperature testing and thermal conductivity measurement. These methods are insufficient to effectively identify directional and continuous hidden degradation areas within the coating. Especially in the early stages of the channel effect, the coating surface may show no obvious abnormalities, but its fire resistance performance has already begun to deteriorate locally. Traditional methods are unable to provide early warning and quantitative assessment, leading to delayed maintenance decisions and increased safety hazards.

[0004] In view of this, the present invention proposes a method for testing the fireproof and heat insulation performance of the coating of the main cable of a suspension bridge to solve the above problems. Summary of the Invention

[0005] To overcome the aforementioned deficiencies of the prior art and to achieve the above objectives, the present invention provides the following technical solution: a method for testing the fireproof and heat-insulating performance of a coating for the main cable of a suspension bridge, comprising:

[0006] Thermal excitation was applied to the surface of the main cable coating, and the first temperature sequence of the main cable coating surface was collected. The uniformity characteristics were analyzed and the suspected areas were divided according to the uniformity characteristics.

[0007] Secondary thermal excitation and synchronous monitoring are performed in the suspected area to obtain a second temperature sequence. The second temperature sequence is then extracted, identified, and analyzed to obtain the thermal anomaly zone and its corresponding channel parameters.

[0008] Three-level thermal excitation tests were performed on the thermal anomaly zone, the corresponding third temperature sequence was recorded, and the corresponding temperature decay curve was statistically analyzed; the channel depth was obtained by analyzing the third temperature sequence and the temperature decay curve.

[0009] By combining the channel parameters and channel depth of the thermal anomaly zone, the rate of decline in fire resistance performance can be obtained;

[0010] The test results are obtained by comprehensively judging based on the channel parameters, channel depth, and the rate of decline in fire resistance performance.

[0011] Furthermore, methods for obtaining thermal anomaly zones include:

[0012] N marker points in the suspected area are selected for pulsed thermal excitation again. A polar coordinate system is established within a circular region of radius R around marker point L, using the propagation vector direction of marker point L as the reference zero. This circular region is divided into n sectors. The local heat flow vectors corresponding to all pixels in the k-th sector are averaged to obtain the corresponding average propagation vector. The magnitude of the average propagation vector is recorded as the average diffusion intensity. The average, maximum, and minimum values ​​of the average diffusion intensity in all sectors are obtained, and the thermal anisotropy index is calculated. The centerline of the sector corresponding to the maximum value is taken as the corresponding orientation angle. All marker points with thermal anisotropy indices greater than the anisotropy threshold are selected to form an initial set of anomalous marker points. Spatial adjacency detection is performed on each marker point in the initial set of anomalous marker points to obtain M groups of anomalous marker points, where the marker points within each group are spatially adjacent. A directional consistency check is performed on each group of anomalous marker points to obtain Q groups of consistent marker points. Principal component analysis and region determination are performed on the consistent marker point groups to obtain the thermal anomaly zone.

[0013] Furthermore, during the spatial adjacency detection process, if the Euclidean distance between marker point i and marker point j is not higher than the maximum adjacent distance, then marker point i and marker point j are determined to be adjacent.

[0014] Furthermore, the method for verifying directional consistency includes:

[0015] Double the orientation angles of the markers in the abnormal marker group G to obtain a unit vector composed of the cosine and sine values ​​of the doubled orientation angles. Sum the unit vectors of all markers in each abnormal marker group to obtain the average vector. Calculate the average orientation angle and average vector length based on the average vector, and statistically analyze the standard deviation of the angle based on the average orientation angle. If the number of markers in the abnormal marker group is not less than the minimum number threshold, the standard deviation of the angle is less than the angle threshold, and the average vector length is greater than the length threshold, then the orientation consistency test is passed.

[0016] Furthermore, the method for obtaining thermal anomaly zones by performing principal component analysis and region determination on the consistency marker point group includes:

[0017] Extract the coordinates of the marker points in the consistency marker group to form a coordinate matrix. Calculate the covariance matrix based on the coordinate matrix, and then calculate the first two eigenvalues ​​and their corresponding two eigenvectors. Calculate the square root of the ratio of the first eigenvalue to the second eigenvalue to obtain the aspect ratio; the first eigenvalue must not be less than the second eigenvalue. Mark the direction corresponding to the first eigenvalue as the primary direction and the direction corresponding to the second eigenvalue as the secondary direction. Calculate the coverage area and perimeter of the marker point coordinates in the consistency marker group using the convex hull method. Calculate the compactness based on the coverage area and perimeter. Calculate the channel length based on the maximum and minimum coordinate values ​​along the primary direction within the consistency marker group. Calculate the channel width based on the maximum and minimum coordinate values ​​along the secondary direction.

[0018] When the aspect ratio is not lower than the aspect ratio threshold, the channel length is not lower than the channel length threshold, and the compactness is not higher than the compactness threshold, a thermal anomaly zone is determined to exist, and channel parameters are output. The channel parameters include channel length, channel width, orientation angle, standard deviation of thermal anisotropy index, average orientation angle, standard deviation of angle, and coverage area.

[0019] Furthermore, the uniformity characteristics include thermal diffusivity shape factor and average axial diffusion velocity; the method for obtaining the uniformity characteristics includes:

[0020] A two-dimensional coordinate system is established in each detection area, with the axial direction along the length of the main cable and the circumferential direction along the circumference of the main cable, and the origin as the starting point of the detection section. The surface of the main cable coating is discretized into a regular grid of marker points, with each grid corresponding to a marker point coordinate. The initial surface temperature and ambient temperature of each marker point are recorded. A preset thermal excitation is applied to each marker point. For marker point P, the temperature rise at each moment is calculated to generate the initial temperature rise. The ambient temperature rise at each moment is calculated, and the difference between the initial temperature rise and the ambient temperature rise is calculated to obtain the temperature rise sequence. The highest temperature value of marker point P in the temperature rise sequence is obtained. The largest value is marked as the peak value; the moment when the temperature rise at marker point P reaches half the peak value after the thermal excitation ends is marked as t1; a fixed delay is superimposed on t1 to obtain the diffusion observation time, marked as t2; at time t2, the temperature distribution is extracted along the axial direction with marker point P as the center, and the axial diffusion distance is calculated; the temperature distribution is extracted along the circumferential direction with marker point P as the center, and the circumferential diffusion distance is calculated; the ratio of the axial diffusion distance to the circumferential diffusion distance is calculated to obtain the thermal diffusion shape factor; the ratio of the axial diffusion distance to the fixed delay is calculated to obtain the average axial diffusion velocity.

[0021] Furthermore, methods for obtaining suspected areas include:

[0022] When the average axial diffusion velocity is greater than the axial diffusion velocity threshold and the average thermal diffusion shape factor is greater than the shape ratio threshold, it is marked as an anomaly. All marked anomalies are sorted by axial position to find continuous anomaly segments. Only continuous anomaly segments with more than the minimum threshold are retained. For each continuous anomaly segment, a safe distance is extended to both ends to form a suspected zone.

[0023] Furthermore, methods for obtaining channel depth include:

[0024] The peak time ratio is calculated based on the peak time of each test level; the temperature decay curve of each test level is extracted and analyzed to obtain the temperature decay slope ratio; the temperature decay curve of each test level is integrated to obtain the corresponding thermal response area, and the normalized thermal response area is calculated; the peak time ratio, temperature decay slope ratio, and normalized thermal response area are concatenated to obtain the depth vector; the Euclidean distance between the depth vector and the depth vector in the standard database is calculated; the top X depth vectors with the highest similarity are selected to obtain the channel depths corresponding to the X depth vectors.

[0025] Furthermore, methods for obtaining the rate of decline in fire resistance include:

[0026] The channel length, channel width, channel depth, orientation angle, standard deviation of thermal anisotropy index, mean orientation angle, standard deviation of angle, and coverage area of ​​each channel sample in the standard database are concatenated to obtain the channel vector of the thermal anomaly zone and the sample vector of each channel sample in the standard database.

[0027] Calculate the Euclidean distance between the channel vector and each sample vector in the standard database. The smaller the Euclidean distance, the higher the similarity. Select the top Y channel samples with the highest similarity, obtain the fire performance degradation rate corresponding to the Y channel samples, calculate the weighted average, and obtain the fire performance degradation rate of the thermal anomaly zone.

[0028] Furthermore, methods for obtaining test results include:

[0029] A Level 4 warning is triggered when any of the triggering conditions for a Level 4 warning are met.

[0030] If any of the triggering conditions for a Level 4 warning is not met, but any of the triggering conditions for a Level 3 warning are met, a Level 3 warning will be triggered.

[0031] If any of the triggering conditions for a Level 3 warning is not met, but any of the triggering conditions for a Level 2 warning are met, a Level 2 warning is triggered.

[0032] If any of the triggering conditions for a Level II warning is not met, but any of the triggering conditions for a Level I warning are met, then a Level I warning is triggered.

[0033] Otherwise, it is judged to be in the normal stage.

[0034] The technical effects and advantages of the fireproof and heat-insulating performance testing method for the coating of the main cable of a suspension bridge according to the present invention are as follows:

[0035] This invention rapidly identifies suspected areas through preliminary thermal excitation and uniformity analysis, and then accurately identifies thermal anomaly zones with continuity and directional consistency using secondary excitation and directionality analysis. Depth vectors are extracted through three levels of thermal excitation testing (surface, middle, and deep layers). Combined with a pre-constructed standard database, the channel depth and fire resistance degradation rate are quantitatively assessed, ultimately achieving graded early warning based on multi-parameter thresholds. This effectively overcomes the shortcomings of traditional methods in failing to identify and quantitatively assess the directional degradation process within the coating at an early stage. It achieves integrated detection and early warning of channel effects, from early sign detection and geometric feature extraction to quantification of fire resistance impact. This provides objective and repeatable scientific evidence for preventative maintenance of suspension bridge main cable coatings, significantly improving the timeliness and accuracy of safety warnings. Attached Figure Description

[0036] Figure 1 This is a schematic diagram of a method for testing the fireproof and heat-insulating performance of a coating for the main cable of a suspension bridge according to the present invention.

[0037] Figure 2 This is a schematic diagram of the method for obtaining uniformity features according to the present invention;

[0038] Figure 3 This is a schematic flowchart of the method for obtaining thermal anomaly zones and corresponding channel parameters according to the present invention.

[0039] Figure 4 This is a schematic diagram of the method for obtaining channel depth according to the present invention. Detailed Implementation

[0040] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0041] Please see Figure 1 As shown in the figure, this embodiment provides a method for testing the fire resistance and heat insulation performance of the coating on the main cable of a suspension bridge, including:

[0042] Thermal excitation was applied to the surface of the main cable coating, and the first temperature sequence of the main cable coating surface was collected. The uniformity characteristics were analyzed and the suspected areas were divided according to the uniformity characteristics.

[0043] Uniformity characteristics include thermal diffusivity shape factor and average axial diffusion velocity;

[0044] Reference Figure 2 Methods for obtaining uniformity features include:

[0045] A two-dimensional coordinate system was established in each detection area, with the axial direction along the length of the main cable and the circumferential direction along the circumference of the main cable, and the origin at the starting point of the detection section. The surface of the main cable coating was discretized into a regular grid of marker points, with each grid corresponding to a marker point coordinate. The discretized marker point grid is the spatial basis for all subsequent quantitative analyses, ensuring that each location has a corresponding data point and facilitating independent axial and circumferential analyses. The initial surface temperature and ambient temperature of each marker point were recorded to provide a spatial reference and environmental compensation basis for subsequent dynamic response analysis. The initial temperature distribution of the coating surface is affected by environmental factors such as sunlight and wind speed. Recording the initial values ​​and monitoring environmental changes are prerequisites for extracting the pure thermal excitation response, ensuring that subsequent characteristic parameters only reflect the coating's own properties.

[0046] Apply a preset thermal excitation to each marker point; for example, apply thermal excitation with consistent energy, density, and spot diameter using a non-contact flash lamp, low-power laser, or low-temperature hot air gun, such as an energy density of 0.5 J / cm². 2 Thermal excitation with a pulse width of 50ms and a spot diameter of 15mm was used; the dynamic change sequence of surface temperature was acquired simultaneously to obtain the first temperature sequence. Standardized excitation ensured that all marker points received the same thermal input, so that the response differences between marker points only stemmed from the differences in the internal state of the coating, which is the experimental control basis for comparative analysis.

[0047] For each marker point P, the temperature rise at each time step is calculated to generate the initial temperature rise; the ambient temperature rise at each time step is also calculated, and the difference between the initial temperature rise and the ambient temperature rise is calculated to obtain the temperature rise sequence. This process removes the influence of ambient temperature drift, resulting in a clean signal for subsequent feature extraction. This step significantly improves the signal-to-noise ratio, allowing weak coating anomaly signals to be revealed.

[0048] Obtain the maximum value of marker point P in the temperature rise sequence and mark it as the peak value;

[0049] After the thermal excitation is completed, the temperature rise at marker point P reaches half-peak value, i.e., the time corresponding to half of the peak value is marked as t1. A fixed delay is superimposed on t1 to obtain the diffusion observation time, marked as t2. Time t1 corresponds to the state where heat has entered the coating from the surface but has not yet diffused far, and is sensitive to the initial thermal inertia of the coating. The fixed delay is calculated using the target detection depth and the thermal diffusion rate of the coating, specifically the ratio of the square of the target detection depth to four times the thermal diffusion rate of the coating. For preliminary screening, the target detection depth can be set to 50% of the nominal coating thickness. The thermal diffusion rate of the coating can be pre-measured through calibration experiments in an intact coating area or by referring to typical values ​​provided by the material supplier. The fixed delay ensures that the heat wave has propagated a certain distance and its spatial distribution pattern tends to be stable, facilitating the observation of anisotropy. t1 reflects the rate at which heat enters the coating, and t2 reflects the spatial distribution of heat within the coating; the combination of these two factors characterizes the thermal diffusion process from both temporal and spatial dimensions.

[0050] At time t2, with the marker point P as the center, the temperature distribution is extracted along the axial direction to obtain an axial temperature profile. Two locations on the left and right sides of the axial temperature profile where the temperature value drops to 1 / e of the peak value are identified and labeled P1 and P2, respectively, where e is a mathematical constant. Labeling P1 and P2 is a common method for defining the heat wave front in heat conduction analysis. The distances between the axial coordinates of P1 and P2 and the axial coordinates of the marker point P are calculated to obtain the diffusion distances on the left and right sides. The sum of these diffusion distances is then calculated to obtain the axial diffusion distance. Similarly, with the marker point P as the center, the temperature distribution is extracted along the circumferential direction, and the circumferential diffusion distance is calculated. Axial and circumferential diffusion distances are directly measurable physical quantities. They intuitively reflect the range of heat arrival in different directions. The channel effect manifests as a significant increase in the axial diffusion distance while the circumferential diffusion distance remains relatively unchanged.

[0051] Calculate the ratio of axial diffusion distance to circumferential diffusion distance to obtain the thermal diffusion shape factor;

[0052] The ratio of axial diffusion distance to fixed delay is calculated to obtain the average axial diffusion velocity. The thermal diffusion shape factor and average axial diffusion velocity quantify thermal diffusion anomalies from two dimensions: relative proportion and absolute rate, respectively. These two parameters quantify anomalies from two complementary perspectives: the thermal diffusion shape factor focuses on shape distortion, while the average axial diffusion velocity focuses on conduction acceleration. Joint judgment using these two parameters can effectively distinguish channels from other local defects.

[0053] Methods for obtaining suspected areas include:

[0054] The average axial diffusion rate, the average thermal diffusion shape factor, the standard deviation of velocity, and the standard deviation of shape ratio are statistically analyzed throughout the entire testing period to ensure adaptability to different coating materials and environmental conditions.

[0055] An anomaly is identified when the average axial diffusion velocity exceeds the axial diffusion velocity threshold and the average thermal diffusion shape factor exceeds the shape ratio threshold. The axial diffusion velocity threshold is set based on the average axial diffusion velocity and velocity standard deviation, using a multiple principle (e.g., 2-3 times the standard deviation). Similarly, the shape ratio threshold is set based on the average thermal diffusion shape factor and shape ratio standard deviation, using a multiple standard deviation principle. Both the axial diffusion velocity threshold and the shape ratio threshold are applicable to main cables of different materials and environments, requiring no pre-calibration. Furthermore, the logic requires both anomaly characteristics to appear simultaneously, significantly reducing false alarms caused by random noise or single interference. All marked anomalies are sorted by axial position to identify continuous anomaly segments. Only continuous anomaly segments with more than a minimum threshold are retained, excluding isolated anomalies. For each continuous anomaly segment, a safe distance is extended to both ends to form a suspected zone, providing a clear target for subsequent fine-tuning. Channel effects are essentially spatially continuous degradation paths. This step integrates point-like anomaly signals into linear or strip-like regions through spatial clustering, which is precisely the geometric characteristic of a channel. Size filtering and boundary extension enhance the robustness and engineering practicality of the results.

[0056] The filtered abnormal segments are then extended with safety boundaries to form the final suspected area. Through these steps, the abnormal thermal diffusion areas can be objectively and quantitatively identified from a large area of ​​the main cable surface, solving the technical problem that traditional methods cannot effectively detect early signs of channel effects. This approach does not rely on human experience; all judgments are based on quantitative parameters and statistical criteria, ensuring high repeatability and objectivity.

[0057] Secondary thermal excitation and synchronous monitoring are performed in the suspected area to obtain a second temperature sequence. The second temperature sequence is then extracted, identified, and analyzed to obtain the thermal anomaly zone and its corresponding channel parameters.

[0058] Reference Figure 3 Methods for obtaining thermal anomaly zones and corresponding channel parameters include:

[0059] N marker points in the suspected area are selected for further pulsed thermal excitation. At each marker point, at least two infrared thermal imagers with non-parallel viewing angles are used to acquire a second temperature sequence corresponding to each marker point. For example, one imager is positioned directly facing the main cable surface, and the other is at a 45° angle to the axis, to capture images of heat wave propagation from two significantly different perspectives. Using a calibration board of known size, such as a checkerboard, the intrinsic and extrinsic parameters of each thermal imager are calibrated to obtain the intrinsic and extrinsic parameter matrices, establishing a mapping from image pixel coordinates to a unified world coordinate system. A synchronous triggering device ensures that all thermal imagers acquire data strictly synchronously with the thermal excitation pulse, using the same standardized excitation parameters as the initial screening.

[0060] The moment when the temperature rise of the marker point of the secondary thermal excitation reaches half the peak value is marked as t3; the diffusion observation time is obtained by superimposing a fixed delay on t3 and marked as t4.

[0061] In the heat wave propagation image at time t4, corresponding feature points of the heat wave front region centered on marker point L are identified within the field of view of two infrared thermal imagers. Using the camera's projection matrix, the three-dimensional spatial coordinates of each feature point are calculated using the forward intersection principle. Specifically, in each heat wave propagation image, a square search area with a side length of 2U is defined centered on the projected position of marker point L, where U is set according to the estimated radius of heat diffusion. The temperature rise ΔT of each pixel within the square search area relative to time t3 is calculated. Pixels whose ΔT is greater than the peak temperature rise of the front, i.e., 30% of the temperature rise of marker point L at time t4, are initially defined as the heat wave front region.

[0062] Within the heat wave front region, a batch of salient feature points are extracted using feature point detection algorithms such as SIFT, SURF, or ORB. For each feature point, a feature descriptor vector is generated to uniquely characterize the local temperature rise or texture pattern around that point.

[0063] For any feature point pA extracted from the thermal propagation image of camera A, the corresponding feature descriptor DA is compared with the descriptors of all feature points in the thermal propagation image of camera B. Similarity calculations are performed, such as Euclidean distance or Hamming distance. If a feature point pB exists, and its feature descriptor DB has the best similarity to DA among all matches, and the similarity is higher than a preset matching threshold (which can be set empirically, such as 0.7), then pA and pB are determined to be a pair of corresponding feature points, representing the same physical point on the surface of the main cable coating. After completing this step, a pair of corresponding image points pA and pB that have successfully matched in the two views are obtained, providing input for subsequent 3D reconstruction.

[0064] For each successfully matched pair of image points pA and pB, using the projection matrices of camera A and camera B obtained from calibration, and based on the principle of stereo vision forward intersection, a linear equation is constructed to solve for the three-dimensional spatial coordinates of the image point pair pA and pB in the world coordinate system. Repeating this process yields the set of three-dimensional coordinates of discrete physical points within the heat wave front region.

[0065] Using the coordinates of marker point L at time t3 as the starting point and the coordinates of feature points obtained at time t4 as the ending point group, the displacement vector is obtained by fitting using the least squares method and marked as the propagation vector of marker point L. The direction of the propagation vector is the key directional feature.

[0066] With the propagation vector direction of the marker point L as the reference zero point, a polar coordinate system is established in a circular region with a radius of R around it. The circular region is divided into n sectors. The local heat flow vectors corresponding to all pixels in the kth sector are averaged to obtain the corresponding average propagation vector. The magnitude of the average propagation vector is denoted as the average diffusion intensity.

[0067] The average, maximum, and minimum values ​​of the average diffusion intensity in all sectors are obtained. The difference between the maximum and minimum values ​​of the average diffusion intensity is calculated to obtain the velocity difference, which characterizes the maximum directional difference. The ratio of the velocity difference to the average diffusion intensity of each sector is calculated to obtain the thermal anisotropy index of each sector. The thermal anisotropy index reflects the degree of directional concentration of thermal diffusion. Simultaneously, the centerline of the sector corresponding to the maximum value is used as the corresponding directional angle. Specifically, the sector containing the maximum value is recorded, and the directional angle of the average propagation vector corresponding to the sector containing the maximum value, i.e., the sector centerline, is used as the directional angle of the marker point L.

[0068] All markers with thermal anisotropy indices greater than the anisotropy threshold are selected to form an initial set of anomaly markers. The anisotropy threshold can be set based on the average thermal anisotropy index and the standard deviation of the thermal anisotropy index, and can be set by multiples, such as 2-3 times the standard deviation.

[0069] Spatial adjacency detection is performed on each marker in the initial set of anomaly markers. This yields M groups of anomaly markers; markers within each group are spatially adjacent. Marker i and marker j are considered adjacent if the Euclidean distance between them is not greater than the maximum adjacency distance. The maximum adjacency distance is typically set to 1.5-2 times the spacing between markers. This setting is based on the physical characteristics of the channel; the actual distance between adjacent anomaly points should not be significantly greater than the grid spacing of regular markers. Spatial adjacency detection distinguishes between truly continuous channels and isolated random anomalies, which is crucial for identifying linear channels from point anomalies.

[0070] For each abnormal marker group G, a directional consistency test is performed to obtain Q consistent marker groups; verifying the directional coordination of the abnormal point group is the core feature that distinguishes the channel from other defects.

[0071] Methods for verifying directional consistency include:

[0072] The orientation angle θ of each marker point in the abnormal marker point group G is doubled to 2θ and converted into a unit vector (cos2θ, sin2θ). The average vector is obtained by summing all unit vectors. The average orientation angle and average vector length are calculated based on the average vector. Specifically, the arctangent value is calculated based on the average vector. Half of the arctangent value is calculated. Since the angle was doubled before, half of the arctangent value is taken to obtain the true orientation angle. The average orientation radian is obtained. The average orientation radian is converted into angle and normalized to the range of 0° to 180° to obtain the average orientation angle. The magnitude of the average vector is calculated to obtain the average vector length. The standard deviation of the angle is calculated based on the average orientation angle. Specifically, the absolute value of the difference between the angle of each marker point and the average orientation angle is calculated to obtain the first angle difference. Then, the difference between 180° and the first angle difference is calculated to obtain the second angle difference. The smaller value between the first angle difference and the second angle difference is selected as the angle difference. The angle difference is calculated to obtain the standard deviation of the angle.

[0073] If the number of markers in an abnormal marker group is not less than the minimum number threshold, the angle standard deviation is less than the angle threshold, and the average vector length is greater than the length threshold, then the directional consistency test is passed. The minimum number threshold is generally set to 3-5 based on experience to ensure statistical significance; the angle threshold is set to 15° based on engineering experience, with a directional deviation less than 15° indicating strong directional consistency; the length threshold corresponds to at least 0.9 times the average vector length, indicating high consistency. Only a group of points that simultaneously passes the spatial continuity and directional consistency tests can correspond to the actual physical channel. This step significantly reduces the false alarm rate.

[0074] Principal component analysis and region determination were performed on the consistency marker point group to obtain the thermal anomaly zone; the point group that passed the test was transformed into channel parameters with clear engineering significance, providing input for subsequent performance evaluation. Details are as follows:

[0075] Extract the coordinates of the marker points from the consistency marker point group to construct a coordinate matrix. Calculate the covariance matrix based on the coordinate matrix, and then calculate the first two eigenvalues ​​and their corresponding two eigenvectors. The method for calculating eigenvalues ​​and eigenvectors based on the covariance matrix is ​​existing technology and will not be elaborated further. Calculate the square root of the ratio of the first eigenvalue to the second eigenvalue to obtain the aspect ratio. The first eigenvalue should not be less than the second eigenvalue. Mark the direction corresponding to the first eigenvalue as the primary direction and the direction corresponding to the second eigenvalue as the secondary direction. The aspect ratio characterizes the elongation or narrowness of the point group distribution. For channels, the aspect ratio should be significantly greater than 1.

[0076] The coverage area and perimeter of the marker coordinates in the consistency marker group are calculated using the convex hull method. The method of calculating the coverage area and perimeter based on the marker coordinates using the convex hull method is existing technology and will not be described in detail here. The product of the coverage area and 4π is calculated to obtain process quantity one; the square of the perimeter is calculated to obtain process quantity two; and the ratio of process quantity one to process quantity two is calculated to obtain the compactness. For a strip-shaped channel, the compactness is relatively small, close to 0; for a circular distribution, the compactness is close to 1.

[0077] Obtain the maximum and minimum coordinate values ​​along the main direction within the consistency flag group, calculate the coordinate difference between the maximum and minimum coordinate values, and obtain the channel length; similarly, calculate the coordinate difference in the secondary direction to obtain the channel width.

[0078] When the aspect ratio is not lower than the aspect ratio threshold, the channel length is not lower than the channel length threshold, and the compactness is not higher than the compactness threshold, a thermal anomaly zone is determined to exist, and channel parameters are output. These parameters include channel length, channel width, azimuth angle, standard deviation of thermal anisotropy index, average azimuth angle, standard deviation of angle, and coverage area. The azimuth angle is obtained by calculating the arctangent value from the eigenvector corresponding to the first eigenvalue and converting it to an angle value. The standard deviation of the thermal anisotropy index is obtained by statistically analyzing the thermal anisotropy index. The aspect ratio threshold is generally set to 2.0 based on experience to ensure a clearly banded shape. The channel length threshold is generally set based on the main cable coating thickness. The compactness threshold is generally set to 0.8 based on experience to confirm a banded channel distribution. A true channel generally simultaneously meets the following conditions: clearly elongated, sufficient length, and low compactness; these three conditions together ensure the reliability of the determination result.

[0079] Three-level thermal excitation tests were performed on the thermal anomaly zone at the surface, middle and deep layers. The corresponding third temperature sequence was recorded and the corresponding temperature decay curve was statistically analyzed. The channel depth was obtained by analyzing the third temperature sequence and the temperature decay curve.

[0080] The three-level thermal excitation tests—surface, middle, and deep—correspond to different pulse widths and energies. For surface testing, the pulse width is typically set to 0.05 s, and the energy density is generally set to 0.3 J / cm³. 2 It is mainly used to excite the surface layer of the main cable coating, i.e., the response of the coating thickness of 0-30%; the pulse width for the middle layer test is generally set to 0.10s, and the energy density is generally set to 0.5J / cm². 2 It is mainly used to excite the response of the middle layer of the main cable coating, i.e., 31%-70% of the coating thickness; the pulse width for deep layer testing is generally set to 0.20s, and the energy density is generally set to 0.8J / cm². 2 It is mainly used to stimulate the response of the deep layer of the main cable coating, namely 71%-100% coating thickness and near the substrate.

[0081] Three key points were selected within the thermal anomaly zone: the start point, the midpoint, and the end point. For each key point, surface, mid-layer, and deep layer tests were performed sequentially. Adjacent tests were spaced 30 seconds apart to ensure the coating surface temperature recovered to within ±0.5℃ of ambient temperature. During each test, an infrared thermal imager simultaneously recorded the third temperature sequence within 0-3 seconds after heating, with a sampling frequency ≥50Hz. Defects at different depths within the coating exhibit different response characteristics to thermal excitation. By designing three levels of thermal excitation with different pulse widths and energies for the surface, mid-layer, and deep layers, responses in the surface, mid-layer, and deep regions can be excited separately, thereby obtaining information on thermal response differences along the depth direction. This step provides depth-resolved thermal response data, laying the foundation for subsequent depth feature extraction. The excitation parameters for different depths were designed based on the coating's thermal diffusion characteristics to ensure that the responses of each layer can be effectively excited and separated.

[0082] Methods for obtaining temperature decay curves include:

[0083] Centered on the key point, a circular area with a radius of 5 mm is selected as the analysis region. The average key temperature rise at each moment within the analysis region is calculated. A 5-point moving average filter can be used to smooth the curve and reduce the influence of noise. The peak value of the key point is statistically obtained and marked as the peak temperature rise. The average key temperature rise is divided by the peak temperature rise to obtain the normalized temperature decay curve, which facilitates comparison between different tests. The temperature decay process reflects the heat dissipation dynamics in the coating, and its shape is significantly affected by the defect depth. Extracting the standardized temperature decay curve is a prerequisite for calculating depth feature parameters. The above steps can obtain a clean and comparable temperature decay curve, eliminating the influence of spatial inhomogeneity and random noise, making the subsequent feature parameter calculations more stable and reliable.

[0084] Reference Figure 4 Methods for obtaining channel depth include:

[0085] The peak time ratio is obtained by calculating the difference between the peak time of the surface layer test and the peak time of the deep layer test, and then comparing this difference with the peak time of the intermediate layer test. The peak time of the surface layer test is the time required for the average temperature rise of the heated area on the coating surface to reach its maximum value from the start of heating in the surface layer test; the peak time of the intermediate layer test is the time required for the average temperature rise of the heated area on the coating surface to reach its peak value from the start of heating in the intermediate layer test; and the peak time of the deep layer test is the time required for the average temperature rise of the heated area on the coating surface to reach its peak value from the start of heating in the deep layer test. The peak time ratio quantifies the difference in the delay of the surface temperature rise reaching its peak value under different depths of thermal excitation. Deep defects lead to an extended heat conduction path, significantly delaying the peak time in the deep layer test.

[0086] For the temperature decay curve of each test level, the first temperature rise and first time point when the temperature drops from the peak to 50% of the peak, and the second temperature rise and second time point when the temperature drops from the peak to 20% of the peak are extracted. The temperature rise difference between the first and second temperature rises is calculated, as is the time point difference between the first and second time points. The ratio of the temperature rise difference to the time point difference is calculated to obtain the decay slope. The ratio of the decay slope of the deep test to that of the surface test is calculated to obtain the temperature decay slope ratio. The temperature decay slope ratio quantifies the difference in surface temperature decay rate under thermal excitation at different depths. Deep defects may provide rapid heat conduction paths, accelerating the surface temperature decrease.

[0087] Integrating the temperature decay curve yields the thermal response area. The sum of the surface and middle layer thermal response areas is calculated, and the ratio of the deep layer thermal response area to this sum is calculated to obtain the normalized thermal response area. The normalized thermal response area quantifies the difference in the cumulative effect of surface temperature rise under thermal excitation at different depths. Deep defects may lead to rapid heat dissipation, reducing the accumulation of surface temperature rise.

[0088] These three parameters characterize the relationship between thermal response and depth from three complementary perspectives: time delay, decay rate, and cumulative effect, thus forming a three-dimensional feature space for depth recognition.

[0089] The peak time ratio, temperature decay slope ratio, and normalized thermal response area are concatenated to obtain a depth vector. The Euclidean distance between this depth vector and depth vectors in the standard database is calculated. The top X depth vectors with the highest similarity are selected to obtain the channel depths corresponding to these X depth vectors. The depth vectors in the standard database are concatenated from the peak time ratio, temperature decay slope ratio, and normalized thermal response area of ​​artificial defect samples. These parameters are obtained through experimental calibration, such as preparing artificial defect samples with different channel depths and measuring the peak time ratio, temperature decay slope ratio, and normalized thermal response area of ​​each artificial defect sample. At least three sets of parallel samples are set for each artificial defect sample. The standard database serves as the benchmark for this method, establishing a correlation between the abstract feature parameter space and specific physical depth. The completeness and accuracy of the database directly determine the reliability of the depth assessment. The feature parameters obtained from field testing are matched with the standard database for similarity to find the most similar known depth samples, thereby estimating the depth of unknown channels. This step enables the conversion from feature parameters to specific depth values ​​and provides quantification of the uncertainty of the estimate, making the results more scientific and reliable.

[0090] By combining the channel parameters and channel depth of the thermal anomaly zone, the rate of decline in fire resistance performance can be obtained;

[0091] Methods for obtaining the rate of decline in fire resistance include:

[0092] The channel length, channel width, channel depth, orientation angle, standard deviation of thermal anisotropy index, mean orientation angle, standard deviation of angle, and coverage area of ​​each channel sample in the thermal anomaly zone and the standard database are normalized and then spliced ​​together to eliminate the influence of differences in the dimensions and orders of magnitude of different parameters on distance calculation, ensuring that each parameter contributes equally in similarity matching, and obtaining the channel vector of the thermal anomaly zone and the sample vector of each channel sample in the standard database; the multi-dimensional channel parameters obtained from field detection are organized into a standardized mathematical representation to facilitate quantitative comparison with the database.

[0093] Calculate the Euclidean distance between the channel vector and each sample vector in the standard database. The smaller the Euclidean distance, the higher the similarity. Select the top Y channel samples with the highest similarity and obtain the fire performance degradation rate corresponding to the Y channel samples. Calculate the weighted average to obtain the fire performance degradation rate of the thermal anomaly zone. The weight is the reciprocal of the corresponding Euclidean distance. Based on the physical assumption that similar channels have similar fire performance impact, use known samples in the database to estimate the fire performance degradation rate of unknown channels through the nearest neighbor matching method.

[0094] Methods for building standard databases include:

[0095] The standard channel sample is exactly the same as the coating system of the main cable of the suspension bridge, including primer, intermediate paint, topcoat and coating process;

[0096] The design includes F channel samples with multi-dimensional parameters, including channel length, channel width, channel depth, orientation angle, and channel shape; each set of parameters has at least three parallel samples; and artificial channel samples with precise and controllable parameters are created to cover various situations that may occur in the field.

[0097] Measure the channel length, channel width, standard deviation of thermal anisotropy index, mean orientation angle, standard deviation of angle, and coverage area for each channel sample; establish a complete record of characteristic parameters for each channel sample to ensure complete correspondence with the on-site detection parameters.

[0098] Fire resistance performance tests were conducted on each channel sample, and the rate of decline in fire resistance performance was calculated based on the test data. The fire resistance performance test quantitatively measures the loss of fire resistance performance for each sample due to channel defects. Details are as follows:

[0099] Fire resistance tests were conducted according to the temperature rise curve specified in ISO 834 standard.

[0100] The channel sample was placed vertically with the back, i.e. the uncoated side, facing the fire source to simulate the fire condition of the main cable.

[0101] Thermocouples are arranged at the center and multiple locations on the back of the channel sample. For example, a K-type thermocouple is set at the center, four corners and the midpoint of the four sides. The thermocouples are buried at a depth of 0.5 mm from the surface of the substrate, reflecting the actual temperature of the steel wire, and are used to measure the temperature on the back of the channel sample.

[0102] The first critical time is measured, which is the time when the average temperature on the back reaches the preset critical value. The preset critical value is generally set to 200℃, which is the temperature at which the strength of common steel wires begins to decrease significantly.

[0103] The critical time of the test specimen is measured, the time difference between the critical time and the first critical time of the test specimen is calculated, and the ratio of the time difference to the critical time of the test specimen is calculated to obtain the fire performance degradation rate. The fire performance degradation rate represents the percentage reduction in fire protection time due to channel defects.

[0104] Construct a database containing N channel samples. Each sample records the sample ID, channel length, channel width, standard deviation of thermal anisotropy index, mean orientation angle, standard deviation of angle, and coverage area, as well as the corresponding rate of decline in fire resistance performance.

[0105] The test results are obtained by comprehensively judging based on the channel parameters, channel depth, and the rate of decline in fire resistance performance.

[0106] Methods for obtaining test results include:

[0107] A Level 4 warning is triggered when any of the triggering conditions is met. The triggering conditions for a Level 4 warning are: the channel depth reaches Level 4, indicating deep penetration, which means the coating has essentially lost its isolation function, allowing corrosive media and heat to directly act on the main cable wires, posing an immediate threat; the fire resistance degradation rate is not lower than the first degradation rate threshold, meaning the back temperature arrival time of the main cable coating under standard fire conditions no longer meets the minimum safety redundancy required by the design; the channel length is not lower than the first length threshold and the fire resistance degradation rate is not lower than the second degradation rate threshold, as excessively long and severe channels will form significant thermal bridges and seepage paths, greatly weakening the overall protective performance of local sections; at least W thermal anomaly zones are detected within the pre-detection length range, and the fire resistance degradation rate of at least one thermal anomaly zone is not lower than the third degradation rate threshold. This condition applies to areas where channel effects may occur densely, indicating severe coating degradation and systemic risk.

[0108] If any triggering condition of a Level 4 warning is not met, but any triggering condition of a Level 3 warning is met, a Level 3 warning is triggered. The triggering conditions for a Level 3 warning are: the rate of fire resistance degradation is lower than the first degradation rate threshold but not lower than the second degradation rate threshold; at this point, performance has severely degraded, and although it has not reached an immediate danger level, the risk has significantly increased, requiring immediate intervention; the channel length is not lower than the second length threshold, and the rate of fire resistance degradation is not lower than the third degradation rate threshold; the channel depth reaches Level 3, and the rate of fire resistance degradation is not lower than the third degradation rate threshold; at this point, deep defects combined with significant performance loss indicate that degradation is approaching the base material and is developing rapidly; the thermal anomaly zone is located in a critical stress area, and the rate of fire resistance degradation is not lower than the third degradation rate threshold. Critical stress areas include the predefined saddle area within 0.5m, cable clamps, and the main cable anchorage area. Defects in critical locations have a greater impact on the overall safety of the main cable and require a higher level of attention.

[0109] If any of the triggering conditions for a Level 3 warning is not met, but any of the triggering conditions for a Level 2 warning are met, a Level 2 warning is triggered. The triggering conditions for a Level 2 warning are: the rate of decline in fire resistance is lower than the second decline rate threshold but not lower than the fourth decline rate threshold, indicating a significant performance decline requiring planned repair; the channel length is lower than the second length threshold but not lower than the third length threshold, and the rate of decline in fire resistance is not lower than the fourth decline rate threshold; the channel depth reaches Level 2, and the rate of decline in fire resistance is not lower than the fourth decline rate threshold, indicating that mid-layer defects have a quantifiable impact on performance; the angle standard deviation is not higher than the standard deviation threshold, the channel length is not lower than the second length threshold, and the mean thermal anisotropy index is not lower than the index threshold. This combination is a strong indicator of a typical channel effect, meaning it has a clear directionality, a certain scale, and significant thermal anisotropy. Even if the current rate of decline in fire resistance has not reached a higher threshold, it indicates that a clear deterioration channel has formed, posing a significant risk.

[0110] If any of the triggering conditions for a Level II warning is not met, but any of the triggering conditions for a Level I warning are met, a Level I warning is triggered. Level I warning triggering conditions are: 1) The fire performance degradation rate is lower than the fourth degradation rate threshold and not lower than the fifth degradation rate threshold, indicating initial signs of performance degradation, requiring monitoring; 2) The passage length is not lower than the third length threshold, and the fire performance degradation rate is lower than the fourth degradation rate threshold, indicating a certain scale of defects, but currently having a minor impact on fire performance, requiring attention to its development; 3) The passage depth reaches Level I, and the fire performance degradation rate is not lower than the fourth degradation rate threshold, indicating that surface defects have begun to have a slight impact on performance.

[0111] Otherwise, it is judged to be in the normal stage.

[0112] The channel depth levels are set according to different proportions of the coating thickness. When it is not less than the first preset proportion, it is divided into four levels; when it is less than the first preset proportion but not less than the second preset proportion, it is divided into three levels; when it is less than the second preset proportion but not less than the third preset proportion, it is divided into two levels; and when it is less than the third preset proportion but not less than the fourth preset proportion, it is divided into one level. The first to fourth preset proportions are generally set based on empirical values: the first preset proportion is set to 90%; the second preset proportion is set to 60%-89%; the third preset proportion is set to 30%-59%; and the fourth preset proportion is set to 10%-29%.

[0113] The first to fifth rate-of-failure thresholds are based on laboratory standard fire resistance tests, such as those calibrated according to ISO 834. This is achieved by preparing a series of specimens with known defects, conducting fire resistance tests, analyzing the difference in back-temperature arrival time compared to intact specimens, and establishing a corresponding database. The first rate-of-failure threshold is generally set at 35%, corresponding to the critical point of basic coating failure; the second rate-of-failure threshold is generally set at 25%, corresponding to the imminent exhaustion of safety redundancy; the third rate-of-failure threshold is generally set at 20%, and the fourth rate-of-failure threshold is generally set at 15%, with the third and fourth rate-of-failure thresholds corresponding to significant performance degradation; the fifth rate-of-failure threshold is generally set at 5%, corresponding to performance changes just exceeding the identifiable threshold of measurement uncertainty.

[0114] The first to third length thresholds are obtained based on engineering experience and thermal conduction simulation analysis; for example, the impact of different length channels on the local temperature and stress fields is simulated through finite element analysis. The first length threshold is generally set to 3.0m, corresponding to an influence range that exceeds the local maintenance scale; the second length threshold is generally set to 1.5m, corresponding to the minimum significant engineering length for forming an effective continuous thermal bridge / seepage path; the third length threshold is generally set to 0.5m, which is the smallest defect size worth recording and paying attention to in engineering.

[0115] W and the pre-inspection length are obtained based on statistical analysis and engineering judgment; for example, the pre-inspection length is generally taken as 1.0m, which is a compromise between inspection efficiency and risk identification capability. W is generally set to 3, which means that multiple defects appear within this length, indicating that the degradation is not an isolated event. This value can be optimized through cluster analysis of historical inspection data.

[0116] The standard deviation threshold and exponential threshold are obtained statistically based on measured data from intact and defective coating samples. For example, by measuring a large number of intact areas and channel areas, the mean and standard deviation of the intact coating data are statistically obtained. The standard deviation threshold is generally the mean of the intact coating data plus twice the standard deviation to filter out highly consistent orientations; the exponential threshold is generally the lower quartile of the channel area data to ensure the selection of areas with significant thermal anisotropy.

[0117] The above description is merely a specific embodiment of the present invention, but the scope of protection of the present invention is not limited thereto. Any variations or substitutions that can be easily conceived by those skilled in the art within the technical scope disclosed in the present invention should be included within the scope of protection of the present invention. Therefore, the scope of protection of the present invention should be determined by the scope of the claims.

[0118] In conclusion, the above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of the present invention should be included within the protection scope of the present invention.

Claims

1. A method for detecting fireproof and thermal insulation performance of a coating for a main cable of a suspension bridge, characterized in that, include: Thermal excitation is applied to the surface of the main cable coating, and the first temperature sequence of the main cable coating surface is collected. The uniformity characteristics are analyzed to obtain the uniformity characteristics, and suspected areas are divided according to the uniformity characteristics. The uniformity characteristics include thermal diffusivity shape factor and average axial diffusion velocity. Secondary thermal excitation and synchronous monitoring are performed in the suspected area to obtain a second temperature sequence. The second temperature sequence is then extracted, identified, and analyzed to obtain the thermal anomaly zone and its corresponding channel parameters. Methods for obtaining thermal anomaly zones include: N marker points in the suspected area are selected for pulsed thermal excitation again. A polar coordinate system is established within a circular region of radius R around marker point L, using the propagation vector direction of marker point L as the reference zero. This circular region is divided into n sectors. The local heat flow vectors corresponding to all pixels in the k-th sector are averaged to obtain the average propagation vector. The magnitude of the average propagation vector is recorded as the average diffusion intensity. The average, maximum, and minimum values ​​of the average diffusion intensity in all sectors are obtained, and the thermal anisotropy index is calculated. The centerline of the sector corresponding to the maximum value is taken as the corresponding orientation angle. All marker points with thermal anisotropy indices greater than the anisotropy threshold are selected to form an initial set of anomalous marker points. Spatial adjacency detection is performed on each marker point in the initial set of anomalous marker points to obtain M groups of anomalous marker points, where marker points within each group are spatially adjacent. A directional consistency check is performed on each group of anomalous marker points to obtain Q groups of consistent marker points. Principal component analysis and region determination are performed on the consistent marker point groups to obtain the thermal anomaly zone. Three-level thermal excitation tests were performed on the thermal anomaly zone, and the corresponding third temperature sequence was recorded. The corresponding temperature decay curves were then statistically analyzed. The channel depth was obtained by analyzing the third temperature sequence and the temperature decay curves. Methods for obtaining the channel depth included: The peak time ratio is calculated based on the peak time of each test level; the temperature decay curve of each test level is extracted and analyzed to obtain the temperature decay slope ratio; the temperature decay curve of each test level is integrated to obtain the corresponding thermal response area, and the normalized thermal response area is calculated; the peak time ratio, temperature decay slope ratio, and normalized thermal response area are concatenated to obtain the depth vector; the Euclidean distance between the depth vector and the depth vector in the standard database is calculated; the top X depth vectors with the highest similarity are selected to obtain the channel depths corresponding to the X depth vectors. The rate of fire resistance degradation is obtained by analyzing the channel parameters and channel depth of the thermal anomaly zone. Methods for obtaining the rate of fire resistance degradation include: The channel length, channel width, channel depth, orientation angle, standard deviation of thermal anisotropy index, mean orientation angle, standard deviation of angle, and coverage area of ​​each channel sample in the standard database are concatenated to obtain the channel vector of the thermal anomaly zone and the sample vector of each channel sample in the standard database. Calculate the Euclidean distance between the channel vector and each sample vector in the standard database. The smaller the Euclidean distance, the higher the similarity. Select the top Y channel samples with the highest similarity, obtain the fire performance degradation rate corresponding to the Y channel samples, calculate the weighted average, and obtain the fire performance degradation rate of the thermal anomaly zone. The test results are obtained by comprehensively judging based on the channel parameters, channel depth, and the rate of decline in fire resistance performance.

2. The method for detecting fireproof and thermal insulation performance of coating for main cable of suspension bridge according to claim 1, characterized in that, During the spatial adjacency detection process, if the Euclidean distance between marker i and marker j is not higher than the maximum adjacency distance, then marker i and marker j are determined to be adjacent.

3. The method for detecting fireproof and thermal insulation performance of coating for main cable of suspension bridge according to claim 1, characterized in that, The method for verifying directional consistency includes: Double the orientation angles of the markers in the abnormal marker group G to obtain a unit vector composed of the cosine and sine values ​​of the doubled orientation angles. Sum the unit vectors of all markers in each abnormal marker group to obtain the average vector. Calculate the average orientation angle and average vector length based on the average vector, and statistically analyze the standard deviation of the angle based on the average orientation angle. If the number of markers in the abnormal marker group is not less than the minimum number threshold, the standard deviation of the angle is less than the angle threshold, and the average vector length is greater than the length threshold, then the orientation consistency test is passed.

4. The method for detecting fireproof and thermal insulation performance of coating for main cable of suspension bridge according to claim 3, characterized in that, The method for obtaining thermal anomaly zones by performing principal component analysis and region determination on a set of consistency marker points includes: Extract the coordinates of the marker points in the consistency marker group to form a coordinate matrix. Calculate the covariance matrix based on the coordinate matrix, and then calculate the first two eigenvalues ​​and their corresponding two eigenvectors. Calculate the square root of the ratio of the first eigenvalue to the second eigenvalue to obtain the aspect ratio; the first eigenvalue must not be less than the second eigenvalue. Mark the direction corresponding to the first eigenvalue as the primary direction and the direction corresponding to the second eigenvalue as the secondary direction. Calculate the coverage area and perimeter of the marker point coordinates in the consistency marker group using the convex hull method. Calculate the compactness based on the coverage area and perimeter. Calculate the channel length based on the maximum and minimum coordinate values ​​along the primary direction within the consistency marker group. Calculate the channel width based on the maximum and minimum coordinate values ​​along the secondary direction. When the aspect ratio is not lower than the aspect ratio threshold, the channel length is not lower than the channel length threshold, and the compactness is not higher than the compactness threshold, a thermal anomaly zone is determined to exist, and channel parameters are output. The channel parameters include channel length, channel width, orientation angle, standard deviation of thermal anisotropy index, average orientation angle, standard deviation of angle, and coverage area.

5. The method for testing the fire resistance and heat insulation performance of a coating for a suspension bridge main cable according to claim 1, characterized in that, Methods for obtaining uniformity characteristics include: A two-dimensional coordinate system is established in each detection area, with the axial direction along the length of the main cable and the circumferential direction along the circumference of the main cable, and the origin as the starting point of the detection section. The surface of the main cable coating is discretized into a regular grid of marker points, with each grid corresponding to a marker point coordinate. The initial surface temperature and ambient temperature of each marker point are recorded. A preset thermal excitation is applied to each marker point. For marker point P, the temperature rise at each moment is calculated to generate the initial temperature rise. The ambient temperature rise at each moment is calculated, and the difference between the initial temperature rise and the ambient temperature rise is calculated to obtain the temperature rise sequence. The highest temperature value of marker point P in the temperature rise sequence is obtained. The largest value is marked as the peak value; the moment when the temperature rise at marker point P reaches half the peak value after the thermal excitation ends is marked as t1; a fixed delay is superimposed on t1 to obtain the diffusion observation time, marked as t2; at time t2, the temperature distribution is extracted along the axial direction with marker point P as the center, and the axial diffusion distance is calculated; the temperature distribution is extracted along the circumferential direction with marker point P as the center, and the circumferential diffusion distance is calculated; the ratio of the axial diffusion distance to the circumferential diffusion distance is calculated to obtain the thermal diffusion shape factor; the ratio of the axial diffusion distance to the fixed delay is calculated to obtain the average axial diffusion velocity.

6. The method for testing the fire resistance and heat insulation performance of a coating for the main cable of a suspension bridge according to claim 5, characterized in that, Methods for obtaining suspected areas include: When the average axial diffusion velocity is greater than the axial diffusion velocity threshold and the average thermal diffusion shape factor is greater than the shape ratio threshold, it is marked as an anomaly. All marked anomalies are sorted by axial position to find continuous anomaly segments. Only continuous anomaly segments with more than the minimum threshold are retained. For each continuous anomaly segment, a safe distance is extended to both ends to form a suspected zone.

7. The method for detecting fireproof and thermal insulation performance of coating for main cable of suspension bridge according to claim 1, characterized in that, Methods for obtaining test results include: A Level 4 warning is triggered when any of the triggering conditions for a Level 4 warning are met. If any of the triggering conditions for a Level 4 warning is not met, but any of the triggering conditions for a Level 3 warning are met, a Level 3 warning will be triggered. If any of the triggering conditions for a Level 3 warning is not met, but any of the triggering conditions for a Level 2 warning are met, a Level 2 warning is triggered. If any of the triggering conditions for a Level II warning is not met, but any of the triggering conditions for a Level I warning are met, then a Level I warning is triggered. Otherwise, it is judged to be in the normal stage.

Images