A landscape paving structure performance evaluation method suitable for differentiated sites

Abstract

The application discloses a kind of landscape paving structure performance evaluation methods suitable for differentiating site, and the application relates to engineering detection technical field, and three-dimensional point cloud data of paving surface is obtained by synchronously carrying camera and three-dimensional laser scanner of unmanned aerial vehicle, reference state vector is constructed in combination with design drawing and completion data, point cloud data is fitted and refined using frequency domain decoupling idea, respectively macroscopic deformation field and residual elevation field are constructed, macroscopic geometric state parameter and microstructure state parameter are extracted, and double-scale quantitative characterization of geometric deformation and microstructure damage of paving surface is realized;Composite state vector is constructed by feature level fusion, and performance degradation evaluation function based on singular value decomposition difference is established, the deviation of reference performance matrix and actual performance matrix is calculated, and finally the structure comprehensive performance index and the evaluation report containing performance level and maintenance suggestion are output.

Description

Technical Field

[0001] This invention belongs to the field of engineering testing technology, specifically, it relates to a method for evaluating the performance of landscape paving structures applicable to different sites. Background Technology

[0002] The rapid development of UAV mapping, 3D laser scanning and point cloud data processing technologies has provided new technical pathways for the refined monitoring of landscape paving structures.

[0003] Existing technologies for evaluating the performance of landscape paving structures in differentiated sites typically rely on manual inspections, local sampling measurements, or simple two-dimensional flatness tests. These methods are inherently inefficient, subjective, have limited data coverage, and cannot fully reflect the overall and microscopic synergistic state of the paving structure. Specifically, traditional methods, which rely on visual observation combined with tools such as rulers and levels to measure local points, are not only time-consuming and labor-intensive, but also yield discrete and partial data. This fails to construct a continuous and complete deformation field information for the pavement, especially for large areas or complex terrain. This point-to-surface assessment method easily misses critical structural defect areas. Furthermore, current technologies lack the ability to quantitatively capture the microscopic texture and local abrupt changes of pavement surfaces, focusing only on macroscopic, obvious settlement or bulging, while neglecting minute surface wear, localized material loosening, or early-stage defects. Uneven deformation is difficult to detect and warn of, resulting in delayed assessment results and failing to provide an effective basis for preventive maintenance. In terms of data integration and assessment, existing methods usually use single indicators such as geometric flatness as the main evaluation criteria, failing to effectively decouple and integrate macroscopic geometric deformation with microscopic structural characteristics. The assessment models are relatively simple and cannot comprehensively reflect the multidimensional degradation state of structural performance. The generated reports are mostly qualitative descriptions and lack a unified, quantifiable comprehensive performance index, which is not conducive to accurate comparison of performance status and long-term tracking management between different periods or sites.

[0004] To address the aforementioned issues, this invention proposes a method for evaluating the performance of landscape paving structures applicable to diverse sites. Summary of the Invention

[0005] To address the shortcomings of existing technologies, this invention provides a method for evaluating the performance of landscape paving structures suitable for differentiated sites, solving the problems of incomplete and inaccurate evaluations and the inability to quantify and provide early warnings of the degradation of the microscopic and macroscopic synergistic performance of paving structures.

[0006] The objective of this invention can be achieved through the following technical solutions: A method for evaluating the performance of landscape paving structures applicable to differentiated sites, the method comprising: Step 1: Determine the target differentiated site and construct the baseline state vector of the pavement surface of the target differentiated site; Step 2: Map a 3D point cloud map of the target's differentiated site paving surface in relation to the current time. Based on the 3D point cloud data, fitting and refinement processes are performed to construct the macroscopic deformation field and residual elevation field of the target differentiated site paving surface, and the macroscopic geometric state parameters and microscopic structural state parameters of frequency domain decoupling are calculated. Step 3: Perform feature-level fusion based on macroscopic geometric state parameters and microscopic structural state parameters to construct a composite state vector of the target differentiated site pavement surface; Structural performance evaluation is performed by combining the baseline state vector and the composite state vector, and the comprehensive structural performance index associated with the target differentiated site pavement is output, generating a comprehensive structural performance report.

[0007] As a further aspect of the present invention, the specific method for constructing the reference state vector of the target differentiated site pavement in step one is as follows: Identify target differentiated sites; Based on the design drawings and as-built data of the target differentiated site, extract the baseline values ​​of all parameter indicators associated with the target differentiated site, and calculate the baseline state vector B=(b1,b2,...,bj), where bi is the baseline value corresponding to the i-th parameter indicator, j is the total number of indicators, and i is the counting index, with a value range from 1 to j.

[0008] As a further aspect of the present invention, the specific method for mapping the three-dimensional point cloud diagram of the differentiated site paving surface associated with the current time in step two is as follows: Using a pre-prepared drone carrying a camera and a 3D laser scanner, the target differentiated site paving surface is photographed from above. The camera's captured image is larger than or equal to the target differentiated site paving surface, and the center O1 of the captured image coincides with the geometric center O2 of the target differentiated site. The pixel ratio of the captured image to the actual object is a preset fixed ratio. The scanning range of the 3D laser scanner is greater than or equal to that of the target differentiated site, and the scanning center O3 of the 3D laser scanner coincides with the geometric center O2 of the pavement surface of the target differentiated site. The captured images are correlated and mapped with the 3D point cloud data measured by the 3D laser scanner to obtain the 3D point cloud map associated with the target differentiated site pavement at the current moment, denoted as PM.

[0009] As a further aspect of the present invention, the specific method for associating and mapping the captured image with the three-dimensional point cloud data measured by the three-dimensional laser scanner in step two is as follows: Extract actual dimensions from design drawings of site paving surfaces based on target differentiation; Extract the camera's captured image, combine the geometric center O2 with the actual size to crop the captured image, retain the pixel portion of the target differentiated site paving surface, and record it as the first captured image A1; Based on center O1 and scanning center O3, the first shooting screen A1 is overlapped with the three-dimensional point cloud data measured by the three-dimensional laser scanner in the first shooting screen A1 to determine the three-dimensional point cloud data of each pixel in the shooting screen and complete the association mapping.

[0010] As a further aspect of the present invention, in step two, the macroscopic deformation field of the target differentiated site pavement is constructed by fitting based on three-dimensional point cloud data, and the specific method for calculating the macroscopic geometric state parameters is as follows: Extract the three-dimensional point cloud map PM, construct a two-dimensional coordinate system on the target differentiated site paving surface with the geometric center O2 as the origin, and let the Z coordinate value of the three-dimensional point cloud data of each point in the three-dimensional point cloud map PM be used as the elevation value of each coordinate in the two-dimensional coordinate system to form the original deformation field C1. The original deformation field C1 is divided into regions using a preset grid. The resulting region grids are sorted in the order from the negative X-axis to the positive X-axis and from the positive Y-axis to the negative Y-axis in the original deformation field C1, and are denoted as region grid sequence Q1, Q2, ..., Qm. The macroscopic deformation field C2 is composed of region grid sequence Q1, Q2, ..., Qm, where m is the total number of region grids. For any region grid Qn, extract all elevation values ​​within it and calculate the mean, denoted as the average relative elevation Un_avg of the region grid Qn, where n is the counting index, with a value ranging from 1 to m; Similarly, determine the average relative elevation of all regional grids: U1_avg, U2_avg, ..., Um_avg; Calculate the initial elevation U_fir of the target differentiated site pavement based on as-built data; Calculate the absolute value of the difference between each average relative elevation and the initial elevation U_fir, and represent them as follows: U1_ad, U2_ad, ..., Um_ad; Calculate the standard deviations of U1_ad, U2_ad, ..., Um_ad, and label them as the macroscopic geometric state parameters (MGP) of the target differentiated site pavement.

[0011] As a further aspect of the present invention, in step two, the residual elevation field of the target differentiated site pavement is constructed by refining the three-dimensional point cloud data, and the specific method for calculating the microstructure state parameters is as follows: For any region grid Qn in the macroscopic deformation field C2, extract all elevation values ​​inside it and denote them as the elevation value set {U1,U2,...,Uk}, where k is the total number of coordinates in the region grid Qn; Calculate the residual elevation set {R1,R2,...,Rk} corresponding to the elevation set {U1,U2,...,Uk}, where Ro=Uo-Un_avg, and o is the counting index, with a value range from 1 to k; Similarly, determine the set of residual elevation values ​​for all regional grids and perform a summary to generate a residual elevation field C3 covering the target differentiated site paving surface; Using MI=(1 / m×k)∑ e=1 m×k |Re| calculates the micro-irregularity index MI, where e is the counting index, ranging from 1 to m×k, and m×k represents the total number of coordinates in the residual elevation field C3; Extract the preset residual change threshold δ, traverse the residual elevation value of each coordinate in the residual elevation field C3, determine the absolute value of the difference between the residual elevation value of each coordinate and the residual elevation value of the adjacent coordinate, compare it with the residual change threshold δ, and count the number of coordinate pairs that exceed the residual change threshold δ, denoted as cout; The local mutation density LD is calculated using LD=cout / ( / m×k), and the microstructure state parameters M=(MI,LD) of the target differentiated site pavement are formed by combining the micro-irregularity index MI.

[0012] As a further aspect of the present invention, in step three, the specific method for constructing the composite state vector of the target differentiated site pavement surface by performing feature-level fusion based on macroscopic geometric state parameters and microscopic structural state parameters is as follows: Extract the macroscopic geometric state parameter MGP and the microscopic structural state parameter M=(MI,LD); Combine MGP, MI, and LD in sequence and denote them as the composite state vector V=(MGP,MI,LD) of the target differentiated site pavement.

[0013] As a further aspect of the present invention, in step three, the specific method for generating a structural comprehensive performance report by outputting the structural comprehensive performance index associated with the target differentiated site pavement surface is as follows: Extract the reference state vector B and the composite state vector V; A benchmark performance matrix B_mat is established based on the benchmark state vector B; The actual performance matrix V_mat is established based on the composite state vector V; Establish a performance degradation evaluation function f(B_mat, V_mat) and calculate the singular value decomposition difference between B_mat and V_mat: in, The first digit of the baseline performance matrix B_mat There are t singular values, where σ_Vt is the t-th singular value of the actual performance matrix V_mat, r is the total number of singular values, and t is the counting index, which ranges from 1 to r. Re-adopt Calculate the structural comprehensive performance index SPI, where ℮ is the base of the natural logarithm function, and ℮≠e. The preset performance degradation sensitivity coefficient is used to adjust the sensitivity to performance degradation. The larger the value of θ, the higher the sensitivity, and vice versa. The Structural Performance Index (SPI) is used to generate a structural performance report, which includes basic information about the target differentiated site, the assessment time, the Structural Performance Index (SPI), the performance level assessment conclusions based on the numerical preset division of SPI, and maintenance recommendations.

[0014] The beneficial effects of this invention are: This invention achieves multi-dimensional and precise diagnosis of differentiated landscape paving structures from macro to micro dimensions using three-dimensional point cloud technology. Its advantage lies in the frequency domain decoupling and fusion analysis of the geometric deformation and microstructural features of the paving surface, thereby breaking through the limitations of traditional single detection methods. It can efficiently and accurately quantify the overall deformation and local damage of the paving structure. By comparing the composite state vector with the baseline state, it achieves a comprehensive quantitative assessment of structural performance, thereby improving the systematicness and scientific nature of the assessment. This is conducive to the early detection of potential risks and the guidance of precise maintenance. While ensuring the safety and durability of the landscape, it provides data support and decision-making basis for long-term management. This invention ensures the objectivity, accuracy, and authority of the baseline state vector by explicitly using design drawings and as-built data as the data source for extracting benchmark values, thus providing a data origin for subsequent differential analysis and state assessment. Secondly, by systematically integrating multi-dimensional parameter indicators through vectorization, complex site conditions can be accurately quantified and characterized, providing an analytical basis for subsequent real-time monitoring data comparison, performance degradation identification, and preventive maintenance decisions. This invention utilizes a drone carrying both a camera and a 3D laser scanner for top-down data acquisition, ensuring that the center of the image capture coincides with the center of the scan. This achieves high-precision correlation mapping between visual images and 3D point clouds, improving data consistency. Based on this, a macroscopic deformation field is constructed through grid segmentation, and the overall geometric deformation is quantified using standard deviation to comprehensively assess the settlement or undulation trend of the site pavement. Simultaneously, the microscopic unevenness index and local mutation density are analyzed using residual elevation field analysis to precisely capture local surface defects and mutation characteristics. Based on a multi-level analysis method combining macroscopic and microscopic perspectives, comprehensive and efficient monitoring of the deformation and structural status of the site pavement is achieved. Automated calculations reduce human error, providing accurate data support for engineering maintenance and risk warning. This invention achieves a comprehensive description of the pavement state by fusing macroscopic geometric parameters and microscopic structural parameters at the feature level to form a multidimensional composite state vector. Its advantages lie in introducing a matrix difference evaluation function based on singular value decomposition to deeply quantify the degree of structural performance degradation, and combining it with an exponential sensitivity coefficient to flexibly adjust the evaluation sensitivity, making the calculation results more accurate and reliable. The final generated comprehensive structural performance report quantifies the comprehensive structural performance index and provides intuitive performance levels and maintenance recommendations based on numerical classification. Attached Figure Description

[0015] The invention will now be further described with reference to the accompanying drawings.

[0016] Figure 1 This is a flowchart illustrating the method described in this invention; Figure 2 This is a flowchart illustrating the content described in Embodiment 2 of the present invention. Detailed Implementation

[0017] 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.

[0018] like Figure 1 As shown, this application provides a method for evaluating the performance of landscape paving structures applicable to differentiated sites; As an embodiment 1 of this application, it specifically includes: Step 1: Determine the target differentiated site and construct the baseline state vector of the pavement surface of the target differentiated site; Step 2: Map a 3D point cloud map of the target's differentiated site paving surface in relation to the current time. Based on the 3D point cloud data, fitting and refinement processes are performed to construct the macroscopic deformation field and residual elevation field of the target differentiated site paving surface, and the macroscopic geometric state parameters and microscopic structural state parameters of frequency domain decoupling are calculated. Step 3: Perform feature-level fusion based on macroscopic geometric state parameters and microscopic structural state parameters to construct a composite state vector of the target differentiated site pavement surface; Structural performance evaluation is performed by combining the baseline state vector and the composite state vector, and the comprehensive structural performance index associated with the target differentiated site pavement is output, generating a comprehensive structural performance report. Example 2

[0019] This embodiment continues to explain and illustrate steps one and two described in embodiment one, based on embodiment one. Figure 2 As shown, it specifically includes the following: First, the target differentiated site, which has been pre-marked by the operators and has already been paved with paving materials, is obtained. The design drawings and as-built data involved in the paving process of the target differentiated site are extracted. The baseline values ​​of all parameter indicators associated with the target differentiated site are extracted from the design drawings and as-built data. The parameter indicators include, but are not limited to, paving material type, paving thickness, paving surface flatness, paving surface slope, elevation, surface texture depth, joint width, slab size, and material reflectivity. Based on the baseline values ​​of all parameter indicators associated with the target differentiated site, construct a baseline state vector B=(b1,b2,...,bj), where bi is the baseline value corresponding to the i-th parameter indicator, j is the total number of indicators, and i is the counting index, with a value range from 1 to j. It should be noted that the baseline state vector B is the benchmark for all subsequent state assessments. It represents the ideal state of the pavement at the time of completion and acceptance. Any state parameters detected later will be compared with this benchmark to quantify the degree of deviation.

[0020] After determining the baseline state vector B associated with the target differentiated site, a drone equipped with both a visible light camera and a 3D laser scanner is used to collect data on the target differentiated site.

[0021] The camera and 3D laser scanner are in a top-down view of the target differentiated site paving surface, and the camera's captured image is greater than or equal to the target differentiated site paving surface. The center O1 of the captured image coincides with the geometric center O2 of the target differentiated site, and the pixel ratio of the captured image to the actual object is a fixed ratio preset by the operator. The camera is used to acquire high-resolution images, providing texture and visual information, which can be used to help identify non-geometric features such as materials, cracks, stains, and vegetation erosion after being associated with 3D point cloud data, or for color rendering of point cloud data to facilitate human interpretation. 3D laser scanners are used to directly acquire high-precision 3D point cloud data as the basis for geometric deformation analysis; The scanning range of the three-dimensional laser scanner is greater than or equal to that of the target differentiated site, and the scanning center O3 of the three-dimensional laser scanner coincides with the geometric center O2 of the pavement surface of the target differentiated site. By aligning O1, O2, and O3, the registration error between the two data sources is minimized, ensuring that each pixel in the image can accurately match the corresponding point in the 3D point cloud. In addition, the directly acquired data also needs to undergo distortion correction processing, which can be achieved directly using existing technology, so this solution will not elaborate further.

[0022] Finally, the captured images are correlated and mapped with the 3D point cloud data measured by the 3D laser scanner, specifically as follows: Extract the actual dimensions of the target differentiated site paving surface from the design drawings; Then, the camera's captured image is extracted, and the image is cropped by combining the geometric center O2 with the actual size, retaining the pixel portion of the target differentiated site paving surface, which is recorded as the first captured image A1; Based on center O1 and scanning center O3, the first captured image A1 is overlapped with the 3D point cloud data measured by the 3D laser scanner within the first captured image A1 to determine the 3D point cloud data of each pixel in the captured image. Combined with orthorectification and texture mapping algorithms, the association mapping is automatically completed in the computer. It should be noted that the cropping ratio here needs to take into account the drone's flight altitude and the point density of the laser scan. That is to say, it is necessary to pre-build a unified spatial coordinate system for the drone, camera, and 3D laser scanner. This step is performed by the operator beforehand to calibrate the drone, camera, and 3D laser scanner. In addition, during the actual acquisition process, it is also necessary to pre-set high-visibility ground control points for image geometric correction and point cloud-image registration of the captured images and 3D point cloud data.

[0023] Finally, after completing the association mapping, a three-dimensional point cloud map of the target differentiated site paving surface at the current moment is obtained and labeled as PM.

[0024] Thus, a three-dimensional point cloud map PM representing the three-dimensional point cloud data and the captured images is obtained, as follows. Based on the three-dimensional point cloud data in the three-dimensional point cloud map PM, a fitting operation is performed to construct the macroscopic deformation field of the target differentiated site pavement, thereby calculating the macroscopic geometric state parameters of the target differentiated site pavement. First, extract the 3D point cloud map PM and determine the geometric center O2 in PM. Using this geometric center O2 as the origin, construct a 2D coordinate system on the target differentiated site paving surface. The 2D coordinate system is on the same plane as the ideal state of the target differentiated site paving surface. It should be noted that the ideal state of the target differentiated site paving surface may not be a perfect plane (but operators will undergo leveling work during the actual construction process). Therefore, the 2D coordinate system needs to fit the target differentiated site paving surface to a plane as much as possible during the actual construction process to facilitate subsequent analysis. The plane mentioned does not refer to a horizontal plane, but to the plane on which the target differentiated site paving surface is located.

[0025] Extract the 3D point cloud data of each point in the 3D point cloud map PM, and use the Z coordinate value as the elevation value of the corresponding point in the 2D coordinate system. In this way, determine the elevation values ​​of the left and right coordinates in the 2D coordinate system, and form the original deformation field C1. Next, the original deformation field C1 is divided into regions using the grid preset by the operator. This is used to perform dimensionality reduction and fitting on the three-dimensional point cloud map PM, extracting the trend deformation. After the region division, several region grids are obtained. The total number of region grids is counted and denoted as m. It is important to note that the grid preset by the operator should be determined based on the actual size of the site and the wavelength of the deformation characteristics of interest. If the grid is too small, it will lose its macroscopic significance and result in a large amount of calculation. If the grid is too large, it will smooth out important deformation information. Therefore, the operator needs to determine the grid based on experience and the block size in the as-built data.

[0026] Sort the m region grids according to the order from the negative X-axis to the positive X-axis and from the positive Y-axis to the negative Y-axis in the original deformation field C1. Record the sorting result as the region grid sequence Q1, Q2, ..., Qm. Thus, the macroscopic deformation field C2 is composed of the region grid sequence Q1, Q2, ..., Qm, where m is the total number of region grids.

[0027] Extract any one region grid Qn from the region grid sequence Q1, Q2, ..., Qm, obtain the elevation values ​​corresponding to all coordinates within region grid Qn, and perform mean processing. The calculated mean value is denoted as the average relative elevation Un_avg of region grid Qn, where n is the counting index, and the value range is from 1 to m. Repeat the above steps to process all regional grids in the same synchronous manner, and determine the average relative elevation of all regional grids: U1_avg, U2_avg, ..., Um_avg; Next, the design theoretical elevation of the target differentiated site pavement is extracted from the as-built data and marked as the initial elevation U_fir of the target differentiated site pavement; Next, calculate the absolute value of the difference between the average relative elevations U1_avg, U2_avg, ..., Um_avg and the initial elevation U_fir, and sort them in the order of U1_avg, U2_avg, ..., Um_avg, represented as: U1_ad, U2_ad, ..., Um_ad; Next, the standard deviations of U1_ad, U2_ad, ..., Um_ad are calculated. The standard deviation quantifies the dispersion of the deviation of each grid from the initial design elevation. A small standard deviation indicates that the overall macroscopic condition of the site is good. Conversely, a large standard deviation indicates that the site has experienced uneven settlement or warping, which may be due to problems with the pavement structure and requires early attention. Finally, the calculated standard deviation is labeled as the macro-geometric state parameter (MGP) of the target differentiated site pavement.

[0028] After determining the macroscopic geometric state parameters (MGP), the 3D point cloud data is further refined to construct the residual elevation field of the target differentiated site pavement, and the microstructural state parameters are calculated, as follows: First, extract any region grid Qn in the macroscopic deformation field C2, extract the elevation values ​​of all coordinates inside the region grid Qn, and sort them according to the order from the negative X-axis to the positive X-axis and from the positive Y-axis to the negative Y-axis in the original deformation field C1. This set of elevation values ​​is denoted as {U1, U2, ..., Uk}, where k is the total number of coordinates in the region grid Qn, which is the total number of elevation values.

[0029] Next, calculate the residual elevation set {R1,R2,...,Rk} corresponding to the elevation set {U1,U2,...,Uk}, where Ro=Uo-Un_avg, representing the difference between the corresponding coordinate elevation value and the average relative elevation of the grid in its region, where o is the counting index, with a value range from 1 to k; Repeat the above steps to determine the set of residual elevation values ​​for all regional grids and perform a summary. This generates a residual elevation field covering the target differentiated site pavement, denoted as C3. Then, by using MI=(1 / m×k)∑ e=1 m×k|Re| calculates the micro-irregularity index MI associated with the residual elevation field C3, where e is the counting index, ranging from 1 to m×k, and m×k represents the total number of coordinates in the residual elevation field C3; The Micro-unevenness Index (MI) measures the average roughness of the entire paved surface. The higher the MI value, the rougher the surface is overall, which may indicate severe wear or uniformly distributed fine damage.

[0030] Next, the residual change threshold δ preset by the operator based on practical experience is extracted. The residual elevation value of each coordinate in the residual elevation field C3 is traversed in combination with the residual change threshold δ. The absolute value of the difference between the residual elevation value of the corresponding coordinate and the residual elevation value between the adjacent coordinates is determined. It should be noted that the adjacent coordinates refer to the 8-neighbor coordinates of the corresponding coordinate. The determined absolute value of the difference is compared with the residual change threshold δ. If the absolute value of the difference is less than or equal to the residual change threshold δ, it is skipped. If the absolute value of the difference exceeds the residual change threshold δ, it is marked. Finally, the number of coordinate pairs that exceed the residual change threshold δ is counted and the number of coordinate pairs is recorded as cout. The local mutation density LD is then calculated using LD=cout / ( / m×k). The local mutation density LD represents the change in residual elevation between adjacent detection points. A high LD value indicates that there may be a large number of concentrated and sharp defects on the pavement surface, such as dense crack networks and spalling. Ultimately, the combination of local mutation density LD and micro-irregularity index MI constitutes the microstructural state parameter M=(MI,LD) of the target differentiated site pavement. Example 3

[0031] This embodiment, based on Embodiment 2, further discloses a method for constructing a composite state vector of the target differentiated site pavement, outputting the structural comprehensive performance index associated with the target differentiated site pavement, and generating a structural comprehensive performance report, specifically including the following: Based on the content described in Example 2, the macroscopic geometric state parameter MGP and the microscopic structural state parameter M=(MI,LD) are extracted. MGP, MI and LD are combined in sequence to construct the composite state vector V=(MGP,MI,LD) of the target differentiated site pavement. Based on the analysis of the composite state vector V and the baseline state vector B, the structural comprehensive performance index associated with the target differentiated site pavement is calculated, and a structural comprehensive performance report is generated, as follows: First, a benchmark performance matrix B_mat is established based on the benchmark state vector B to extract the performance structure features of the ideal state of the target differentiated site pavement. The benchmark performance matrix B_mat represents the ideal state of each coordinate in the target differentiated site pavement. Next, based on the composite state vector V, an actual performance matrix V_mat is established to extract the performance structure features of the current state of the target differentiated site pavement. The benchmark performance matrix V_mat represents the current state of each coordinate in the target differentiated site pavement. It should be noted that the actual performance matrix V_mat and the benchmark performance matrix B_mat have the same dimensions; pass A performance degradation evaluation function f(B_mat, V_mat) is established to calculate the singular value decomposition difference between B_mat and V_mat. The calculation is based on the relative attenuation rate of the intensity of the t-th main mode. As a normalization process, the influence of the magnitude of different modes is eliminated.

[0032] right Performing a squaring operation means squaring the relative attenuation rate, amplifying the degraded signal while ensuring the value is positive; (1 / r) represents the average of the squared attenuation rates of the r main modes, used to assess the degree of overall performance degradation; f(B_mat,V_mat) essentially characterizes the overall state of the current target differentiated site pavement relative to the baseline state across all important feature modes, and the larger the value of f(B_mat,V_mat) (not exceeding 1, because it has been normalized), the more severe the degradation. The first digit of the baseline performance matrix B_mat There are t singular values, where σ_Vt is the t-th singular value of the actual performance matrix V_mat, r is the total number of singular values, and t is the counting index, which ranges from 1 to r. Then, by adopting Calculate the structural comprehensive performance index SPI, where ℮ is the base of the natural logarithm function (℮≠e). When the target differentiated site paving surface has no degradation, f(B_mat,V_mat)=0, SPI=1. As f(B_mat,V_mat) increases, SPI decreases from 1 to 0. The performance degradation sensitivity coefficient preset for operators is used to adjust the sensitivity to performance degradation. The larger the value of θ, the faster the SPI will drop for the same f(B_mat,V_mat), thus showing a higher sensitivity, and vice versa. Finally, a structural comprehensive performance report is generated based on the Structural Comprehensive Performance Index (SPI), including basic information about the target differentiated site (images taken by the camera carried by the UAV and 3D point cloud data measured by the 3D scanner), evaluation time (calculation time of the Structural Comprehensive Performance Index (SPI)), performance level evaluation conclusions and maintenance recommendations based on the Structural Comprehensive Performance Index (SPI) and the numerical preset division of the SPI. For example, performance levels include Grade A (Excellent, SPI ≥ 0.90): preventive monitoring; Grade B (Good, 0.75 ≤ SPI < 0.90): planned preventive maintenance; Grade C (Acceptable, 0.60 ≤ SPI < 0.75): repair work required; Grade D (Poor, SPI < 0.60): immediate structural repair. The current Structural Performance Index (SPI) is 0.70, which indicates that the performance level assessment of the target differentiated site pavement is Grade C, and the maintenance recommendation is to arrange repair works.

[0033] All data in the formulas described above have been calculated with dimensions removed. Furthermore, any content not described in detail in this specification is existing technology known to those skilled in the art.

[0034] The above description is merely an example and illustration of the present invention. Those skilled in the art can make various modifications or additions to the specific embodiments described, or use similar methods to replace them, as long as they do not deviate from the invention or exceed the scope defined in the claims, all of which should fall within the protection scope of the present invention.

[0035] It should be stated that all user data collected in this application was collected with the user's consent and authorization. Furthermore, the uses of user data are legal and compliant, and the use and processing of user data comply with the relevant laws, regulations, and standards of the relevant regions.

Claims

1. A method for evaluating the performance of landscape paving structures suitable for differentiated sites, characterized in that, The method includes: Step 1: Determine the target differentiated site and construct the baseline state vector of the pavement surface of the target differentiated site; Step 2: Map a 3D point cloud map of the target's differentiated site paving surface in relation to the current time. Based on the 3D point cloud data, fitting and refinement processes are performed to construct the macroscopic deformation field and residual elevation field of the target differentiated site paving surface, and the macroscopic geometric state parameters and microscopic structural state parameters of frequency domain decoupling are calculated. Step 3: Perform feature-level fusion based on macroscopic geometric state parameters and microscopic structural state parameters to construct a composite state vector of the target differentiated site pavement surface; Structural performance evaluation is performed by combining the baseline state vector and the composite state vector, and the comprehensive structural performance index associated with the target differentiated site pavement is output, generating a comprehensive structural performance report.

2. The method according to claim 1, characterized in that, In step one, the specific method for constructing the reference state vector of the target differentiated site pavement is as follows: Identify target differentiated sites; Based on the design drawings and as-built data of the target differentiated site, extract the baseline values ​​of all parameter indicators associated with the target differentiated site, and calculate the baseline state vector B=(b1,b2,...,bj), where bi is the baseline value corresponding to the i-th parameter indicator, j is the total number of indicators, and i is the counting index, with a value range from 1 to j.

3. The method according to claim 1, characterized in that, In step two, the specific method for mapping the three-dimensional point cloud image of the differentiated site paving surface associated with the current time is as follows: Using a pre-prepared drone carrying a camera and a 3D laser scanner, the target differentiated site paving surface is photographed from above. The camera's captured image is larger than or equal to the target differentiated site paving surface, and the center O1 of the captured image coincides with the geometric center O2 of the target differentiated site. The pixel ratio of the captured image to the actual object is a preset fixed ratio. The scanning range of the 3D laser scanner is greater than or equal to that of the target differentiated site, and the scanning center O3 of the 3D laser scanner coincides with the geometric center O2 of the pavement surface of the target differentiated site. The captured images are correlated and mapped with the 3D point cloud data measured by the 3D laser scanner to obtain the 3D point cloud map associated with the target differentiated site pavement at the current moment, denoted as PM.

4. The method according to claim 3, characterized in that, In step two, the specific method for associating and mapping the captured images with the 3D point cloud data measured by the 3D laser scanner is as follows: Extract actual dimensions from design drawings of site paving surfaces based on target differentiation; Extract the camera's captured image, combine the geometric center O2 with the actual size to crop the captured image, retain the pixel portion of the target differentiated site paving surface, and record it as the first captured image A1; Based on center O1 and scanning center O3, the first shooting screen A1 is overlapped with the three-dimensional point cloud data measured by the three-dimensional laser scanner in the first shooting screen A1 to determine the three-dimensional point cloud data of each pixel in the shooting screen and complete the association mapping.

5. The method according to claim 3, characterized in that, In step two, the macroscopic deformation field of the target differentiated site pavement is constructed by fitting the three-dimensional point cloud data. The specific method for calculating the macroscopic geometric state parameters is as follows: Extract the three-dimensional point cloud map PM, construct a two-dimensional coordinate system on the target differentiated site paving surface with the geometric center O2 as the origin, and let the Z coordinate value of the three-dimensional point cloud data of each point in the three-dimensional point cloud map PM be used as the elevation value of each coordinate in the two-dimensional coordinate system to form the original deformation field C1. The original deformation field C1 is divided into regions using a preset grid. The resulting region grids are sorted in the order from the negative X-axis to the positive X-axis and from the positive Y-axis to the negative Y-axis in the original deformation field C1, and are denoted as region grid sequence Q1, Q2, ..., Qm. The macroscopic deformation field C2 is composed of region grid sequence Q1, Q2, ..., Qm, where m is the total number of region grids. For any region grid Qn, extract all elevation values ​​within it and calculate the mean, denoted as the average relative elevation Un_avg of the region grid Qn, where n is the counting index, with a value ranging from 1 to m; Similarly, determine the average relative elevation of all regional grids: U1_avg, U2_avg, ..., Um_avg; Calculate the initial elevation U_fir of the target differentiated site pavement based on as-built data; Calculate the absolute value of the difference between each average relative elevation and the initial elevation U_fir, and represent them as follows: U1_ad, U2_ad, ..., Um_ad; Calculate the standard deviations of U1_ad, U2_ad, ..., Um_ad, and label them as the macroscopic geometric state parameters (MGP) of the target differentiated site pavement.

6. The method according to claim 5, characterized in that, In step two, the residual elevation field of the target differentiated site pavement is constructed by refining the three-dimensional point cloud data, and the specific method for calculating the microstructure state parameters is as follows: For any region grid Qn in the macroscopic deformation field C2, extract all elevation values ​​inside it and denote them as the elevation value set {U1,U2,...,Uk}, where k is the total number of coordinates in the region grid Qn; Calculate the residual elevation set {R1,R2,...,Rk} corresponding to the elevation set {U1,U2,...,Uk}, where Ro=Uo-Un_avg, and o is the counting index, with a value range from 1 to k; Similarly, determine the set of residual elevation values ​​for all regional grids and perform a summary to generate a residual elevation field C3 covering the target differentiated site paving surface; Using MI=(1 / m×k)∑ e=1 m×k |Re| calculates the micro-irregularity index MI, where e is the counting index, ranging from 1 to m×k, and m×k represents the total number of coordinates in the residual elevation field C3; Extract the preset residual change threshold δ, traverse the residual elevation value of each coordinate in the residual elevation field C3, determine the absolute value of the difference between the residual elevation value of each coordinate and the residual elevation value of the adjacent coordinate, compare it with the residual change threshold δ, and count the number of coordinate pairs that exceed the residual change threshold δ, denoted as cout; The local mutation density LD is calculated using LD=cout / ( / m×k), and the microstructure state parameters M=(MI,LD) of the target differentiated site pavement are formed by combining the micro-irregularity index MI.

7. The method according to claim 6, characterized in that, In step three, the specific method for constructing the composite state vector of the target differentiated site pavement surface by performing feature-level fusion based on macroscopic geometric state parameters and microscopic structural state parameters is as follows: Extract the macroscopic geometric state parameter MGP and the microscopic structural state parameter M=(MI,LD); The MGP, MI, and LD are combined in sequence and denoted as the composite state vector V=(MGP,MI,LD) of the target differentiated site pavement.

8. The method according to claim 7, characterized in that, In step three, the specific method for generating a structural comprehensive performance report by outputting the structural comprehensive performance index associated with the target differentiated site pavement is as follows: Extract the reference state vector B and the composite state vector V; A benchmark performance matrix B_mat is established based on the benchmark state vector B; The actual performance matrix V_mat is established based on the composite state vector V; Establish a performance degradation evaluation function f(B_mat, V_mat) and calculate the singular value decomposition difference between B_mat and V_mat: in, The first digit of the baseline performance matrix B_mat There are t singular values, where σ_Vt is the t-th singular value of the actual performance matrix V_mat, r is the total number of singular values, and t is the counting index, which ranges from 1 to r. Re-adopt Calculate the structural comprehensive performance index SPI, where ℮ is the base of the natural logarithm function, and ℮≠e. The preset performance degradation sensitivity coefficient is used to adjust the sensitivity to performance degradation. The larger the value of θ, the higher the sensitivity, and vice versa. The Structural Performance Index (SPI) is used to generate a structural performance report, which includes basic information about the target differentiated site, the assessment time, the Structural Performance Index (SPI), the performance level assessment conclusions based on the numerical preset division of SPI, and maintenance recommendations.

Images