A road surface three-dimensional texture uniformity evaluation method based on grid distribution characteristics

Abstract

The application relates to a road surface three-dimensional texture uniformity evaluation method based on grid distribution characteristics, which comprises the following steps: obtaining road surface three-dimensional texture data, pre-processing the data to obtain samples, carrying out grid division on the samples, calculating three-dimensional texture feature parameters and statistical indexes of each grid unit of a single sample, evaluating the uniformity of the single sample based on the statistical indexes and normal distribution test, if the single sample is determined to be uniform, carrying out homodistribution test on multiple samples, if the multiple samples all obey the same distribution, determining that the overall distribution of the road surface is uniform, if the distribution of the multiple samples is inconsistent, obtaining coarse and fine aggregate distribution ratios based on two-dimensional wavelet decomposition, judging whether coarse and fine aggregate distribution is uneven among the multiple samples, if the single sample is determined to be uneven, adopting a spatial autocorrelation analysis method to describe and depict local uneven forms. Compared with the prior art, the application can establish a fine uniformity evaluation system of a surface domain, and can evaluate various uneven phenomena.

Description

Technical Field

[0001] This invention relates to the field of road engineering quality assessment technology, and in particular to a method for evaluating the three-dimensional texture uniformity of road surfaces based on grid distribution characteristics. Background Technology

[0002] Asphalt pavement is formed by paving and compacting an asphalt mixture composed of asphalt, aggregates, and additives. The aggregates, composed of coarse and fine aggregates of different sizes in a specified proportion, directly affect the pavement performance of the asphalt mixture. The uniformity of the pavement mixture mainly refers to the spatial uniformity of the coarse and fine aggregates. In practice, due to the influence of many uncertain factors at different stages of road construction (mixing, paving, and compaction), the distribution of the pavement mixture may be uneven, causing changes in parameters such as the density and porosity of the asphalt mixture, leading to a decline in pavement performance and premature damage. Therefore, evaluating pavement uniformity is of great significance for the monitoring, assessment, and control of construction quality.

[0003] The uniformity of asphalt mixtures is mainly divided into internal structural uniformity and surface uniformity. Testing the internal structural uniformity of asphalt mixtures is further divided into non-destructive testing (NDT) and destructive testing (DT). NDT uses equipment such as densitometers and ground-penetrating radar to indirectly obtain the internal structural state of the pavement, used to assess the compaction degree and longitudinal uniformity of the pavement structure. However, the accuracy of NDT equipment is limited; therefore, DT is still considered the only current method to obtain the true internal condition of asphalt mixtures. Traditionally, core samples are drilled from the pavement, and multi-directional mechanical property tests are performed on the samples to assess their homogeneity in different directions. Furthermore, with the development of digital technology, computed tomography (CT) technology can obtain cross-sectional image data of core samples or laboratory-prepared Marshall specimens, serving as an important data source for uniformity analysis. Although internal uniformity can reflect construction quality and pavement performance to some extent, the amount of sample data is limited due to the destructive nature of core sampling. Current methods and indicators for analyzing horizontal uniformity are relatively simple and cannot be used to detect surface uniformity over a larger area.

[0004] While evaluating the uniformity of asphalt mixture surface layers cannot reveal the internal state of the mixture, it can assess the overall construction quality and promptly identify areas of uneven distribution. Surface unevenness in pavement mixtures mainly manifests as segregation and localized damage. Traditional evaluation methods primarily involve visual inspection and the sand-paving method. Visual inspection is typically used for determining asphalt mixture segregation on-site, but it is subjective and difficult to standardize. The sand-paving method reflects the uniformity of aggregate distribution by assessing differences in the macroscopic structure of the pavement surface; however, its testing method has poor accuracy and is time-consuming and labor-intensive. Currently, to quickly acquire surface data for uniformity evaluation, various emerging sensing methods have been introduced, including image acquisition, thermal infrared imaging, and laser 3D scanning. Image-based identification of pavement segregation has become a popular research direction; however, current methods merely classify pavement segregation using machine learning models, lacking specific criteria and indicators. Although overall uniformity evaluation parameters have been proposed, descriptions of spatial distribution patterns and various unevenness conditions are lacking. Summary of the Invention

[0005] The purpose of this invention is to overcome the shortcomings of the existing technology and provide a method for evaluating the uniformity of three-dimensional road surface texture based on grid distribution characteristics. This method can establish a refined uniformity evaluation system for the surface region and can evaluate various non-uniform phenomena accordingly.

[0006] The objective of this invention can be achieved through the following technical solution: a method for evaluating the uniformity of three-dimensional road surface texture based on grid distribution characteristics, comprising the following steps:

[0007] S1. Obtain three-dimensional texture data of the road surface through three-dimensional laser detection technology, and perform data preprocessing to obtain sample data;

[0008] S2. Divide the sample into grids and calculate the three-dimensional texture feature parameters of each grid cell in a single sample;

[0009] S3. Calculate the statistical index of the three-dimensional texture feature parameters of a single sample mesh. Based on the statistical index and the normal distribution test, evaluate the uniformity of the single sample. If the single sample is determined to be uniform, proceed to step S4; otherwise, proceed to step S6.

[0010] S4. Perform an isodistribution test on multiple samples to compare the uniformity among the samples. If all samples follow the same distribution, the road surface is determined to be uniform overall. If the distributions of multiple samples are inconsistent, proceed to step S5.

[0011] S5. Based on the coarse and fine aggregate distribution ratio obtained by two-dimensional wavelet decomposition, determine whether there is uneven distribution of coarse and fine aggregates among multiple samples, and obtain the global uniformity refinement evaluation result.

[0012] S6. Spatial autocorrelation analysis is used to describe and characterize the local non-uniformity, and the results of the refined evaluation of local uniformity are obtained.

[0013] Furthermore, step S1 specifically involves using a three-dimensional laser scanning device to acquire three-dimensional texture data of the road surface and performing data preprocessing, including but not limited to meshing, interpolation, and filtering noise reduction operations, so as to cut the three-dimensional texture data of the road surface into multiple samples of the same size.

[0014] Furthermore, the three-dimensional texture feature parameters of each mesh unit in step S2 include contour parameters, height parameters, functional parameters, volume parameters, and composite parameters. The contour parameters include arithmetic mean height Ra, root mean square height Rq, arithmetic mean slope Da, root mean square slope Dq, arithmetic mean wavelength La, root mean square wavelength Lq, and mean section depth MPD.

[0015] The height parameters include skewness Ssk, kurtosis Sku, maximum peak height Sp, and maximum valley depth Sv;

[0016] The functional parameters include the horizontal difference Sk at the center, the height of the prominent peak Spk, the height of the prominent valley Svk, and the height of the pole Sxp;

[0017] The volume parameters include the void volume of the valley Vvv, the void volume of the center Vvc, the solid volume of the peak Vmc, and the solid volume of the center Vmp.

[0018] The composite parameter includes the interface expansion ratio Sdr.

[0019] Furthermore, step S2 specifically involves dividing the sample into grids based on the sliding window concept, so as to divide a road surface sample into several grids on an average basis. Then, by calculating the parameters of the road surface in each grid, the distribution characteristics of the texture parameters in the region are obtained, which are the three-dimensional texture feature parameters of each grid unit of a single sample.

[0020] Furthermore, the statistical indicators of the three-dimensional texture feature parameters of the sample mesh in step S3 include the sample mean, sample standard deviation, coefficient of variation, coefficient of deviation, and kurtosis coefficient. Specifically, the sample mean is:

[0021]

[0022] The sample standard deviation is specifically:

[0023]

[0024] The coefficient of variation is specifically:

[0025]

[0026] The deviation coefficient is specifically:

[0027]

[0028] The kurtosis coefficient is specifically:

[0029]

[0030] in, Here, m is the sample mean, n is the number of samples, Y is the texture calculation parameter, S is the sample standard deviation, and C is the sample mean. v C is the coefficient of variation. s C is the deviation coefficient. K This is the kurtosis coefficient.

[0031] Furthermore, step S3 specifically includes the following steps:

[0032] S31. Calculate the statistical indicators of the three-dimensional texture feature parameters of a single sample mesh;

[0033] S32. Perform a normality test on the sample parameters;

[0034] S33. Combining the statistical indicators calculated in step S31 and the normal distribution test results obtained in step S32, the uniformity of a single road surface sample is comprehensively evaluated. If the single sample is determined to be uniform, then step S4 is executed; otherwise, step S6 is executed.

[0035] Furthermore, the methods for performing the normality test in step S32 include, but are not limited to, the Lilliefers test and the Kolmogorov-Smirnov test. The null hypothesis is h. If h = 0, it is assumed that the distribution is normal. If h = 1, it is assumed that the distribution is not normal. The significance level alpha is between 0.01 and 0.2. p is the probability value of accepting the hypothesis. If p is less than alpha, the null hypothesis that the distribution is normal can be rejected.

[0036] Furthermore, the specific process of performing the same distribution test on multiple samples in step S4 is as follows: compare the samples pairwise, test whether the variances of the two pairs of samples are equal by using the homogeneity of variance test, test whether the means of the two pairs of samples are equal by using the independent two-sample T test, and if the samples of the road surface have similar distributions in pairs, then the overall distribution of the road surface is determined to be uniform.

[0037] Furthermore, the specific process of step S5 is as follows:

[0038] Using multi-scale texture information obtained from two-dimensional wavelet decomposition, an evaluation index for the coarse-to-fine aggregate distribution ratio ξ is determined, which is used to represent fine and coarse aggregates respectively. The fine aggregate is the sum of the relative energies from level 1 to level I, and the coarse aggregate is the sum of the relative energies from level J to the highest level. The specific calculation formula is as follows:

[0039]

[0040]

[0041]

[0042] Where RE is the relative energy, E is the energy value, x and y are the x-axis and y-axis coordinates of each point in the three-dimensional data, respectively, and z... c,xy Let (x, y) be the height coordinates of the point (x, y) on the two-dimensional level Lc, where c is the decomposition level of the two-dimensional wavelet;

[0043] By plotting the distribution of the coarse and fine aggregate ratio ξ, if the ξ value is higher than the set first threshold, it indicates that there is more fine aggregate; if the ξ value is lower than the set second threshold, it indicates that there is more coarse aggregate. Thus, the distribution of ξ values ​​among multiple samples can be used to determine whether there is an uneven distribution of coarse and fine aggregates among multiple samples.

[0044] Furthermore, step S6 specifically involves using the global Moran index I to observe the overall clustering effect of the grid parameters, in order to measure the correlation of the spatial distribution of the overall sample. The formula for calculating the global Moran index I is as follows:

[0045]

[0046] Among them, w ij This represents the spatial weight value;

[0047] Step S6 specifically involves using the distribution of the local Moran index to characterize specific spatial aggregation phenomena in detail, thereby describing the spatial morphology and severity of local damage and surface skewness. The formula for calculating the local Moran index is as follows:

[0048]

[0049] Among them, I i This is the local Moran index.

[0050] Compared with existing technologies, this invention uses high-precision laser three-dimensional scanning technology to acquire road surface texture data, and employs mathematical statistics, two-dimensional wavelet transform, spatial autocorrelation analysis and other methods. Based on grid division, it proposes a method for evaluating the uniformity of asphalt pavement, and establishes a refined uniformity evaluation system for surface regions. This system can assess the uniformity of the pavement and describe and characterize various non-uniform phenomena such as segregation, caking, and peeling.

[0051] This invention, after acquiring three-dimensional road surface texture data, divides the samples into grids, calculates the three-dimensional texture feature parameters of each grid cell in a single sample, and further calculates corresponding statistical indicators. Combined with a normal distribution test, the uniformity of a single sample is comprehensively evaluated. If a single sample is determined to be uniform, a multi-sample uniformity evaluation is then performed; otherwise, spatial autocorrelation analysis is used to refine the evaluation of local non-uniformities. This ensures both the richness of the sample data and the ability to describe spatial distribution patterns and various non-uniformities.

[0052] When evaluating the uniformity of multiple samples, this invention uses multi-scale texture information obtained from two-dimensional wavelet decomposition to propose an evaluation index for the coarse-fine aggregate distribution ratio ξ, which represents the sum of the relative energies of fine aggregates (from grade 1 to grade I) and coarse aggregates (from grade J to the highest grade). This allows the invention to determine whether there is uneven distribution of coarse and fine aggregates among multiple samples based on the distribution of ξ values, thus enabling the detection of surface uniformity over a wider range.

[0053] This invention, when further analyzing single-sample heterogeneity, utilizes the global Moran's index to observe the overall clustering effect of grid parameters and measure the correlation of the spatial distribution of the overall samples. It also uses the distribution of the local Moran's index to characterize specific spatial clustering phenomena in detail, describing the spatial morphology and severity of local damage and surface skewness. This fully ensures the reliability of the refined evaluation results of local homogeneity. Attached Figure Description

[0054] Figure 1 This is a schematic diagram of the method flow of the present invention;

[0055] Figure 2 This is a diagram showing the distribution of computational parameters for three different grid sizes in the embodiment;

[0056] Figure 3 The uniform distribution of road surface parameters and statistical index values ​​are shown in the examples.

[0057] Figure 4 The uneven distribution of road surface parameters and statistical index values ​​in the example;

[0058] Figure 5A schematic diagram illustrating the effectiveness of the two-sample identical distribution test method in the embodiment;

[0059] Figure 6 This is a schematic diagram showing the distribution of ξ values ​​for different road surface types in the embodiment;

[0060] Figure 7 Example diagram illustrating the spatial morphology of partial damage in the embodiment;

[0061] Figure 8 An example diagram illustrating the spatial morphological description of the surface region skewness in the embodiment. Detailed Implementation

[0062] The present invention will now be described in detail with reference to the accompanying drawings and specific embodiments.

[0063] Example

[0064] like Figure 1 As shown, a method for evaluating the uniformity of three-dimensional road surface texture based on grid distribution characteristics includes the following steps:

[0065] S1. Obtain three-dimensional texture data of the road surface through three-dimensional laser detection technology, and perform data preprocessing to obtain sample data;

[0066] S2. Divide the sample into grids and calculate the three-dimensional texture feature parameters of each grid cell in a single sample;

[0067] S3. Calculate the statistical index of the three-dimensional texture feature parameters of a single sample mesh. Based on the statistical index and the normal distribution test, evaluate the uniformity of the single sample. If the single sample is determined to be uniform, proceed to step S4; otherwise, proceed to step S6.

[0068] S4. Perform an isodistribution test on multiple samples to compare the uniformity among the samples. If all samples follow the same distribution, the road surface is determined to be uniform overall. If the distributions of multiple samples are inconsistent, proceed to step S5.

[0069] S5. Based on the coarse and fine aggregate distribution ratio obtained by two-dimensional wavelet decomposition, determine whether there is uneven distribution of coarse and fine aggregates among multiple samples, and obtain the global uniformity refinement evaluation result.

[0070] S6. Spatial autocorrelation analysis is used to describe and characterize the local non-uniformity, and the results of the refined evaluation of local uniformity are obtained.

[0071] This embodiment applies the above technical solution, and the specific process includes:

[0072] 1. Obtain three-dimensional texture data of the road surface using three-dimensional laser detection technology, and divide the samples into meshes;

[0073] 2. Calculate the 3D texture feature parameters of each grid cell in a single sample;

[0074] 3. Calculate the statistical index of the three-dimensional texture feature parameters of a single sample mesh, and comprehensively evaluate the uniformity of the single sample based on the statistical index and the normal distribution test;

[0075] A. If a single sample is determined to be uniform:

[0076] Then, perform an isodistribution test on multiple samples to compare the uniformity among the samples. If all samples follow the same distribution, then the road surface is judged to be uniformly distributed overall.

[0077] If the distribution of multiple samples is inconsistent, the coarse and fine aggregate distribution ratio obtained based on two-dimensional wavelet decomposition is used to determine whether there is uneven distribution of coarse and fine aggregates among multiple samples.

[0078] B. If a single sample is determined to be uneven:

[0079] Spatial autocorrelation analysis is used to describe and characterize local non-uniform morphology.

[0080] In step one, the three-dimensional texture data of the road surface is obtained using a three-dimensional laser scanning device, and the data is preprocessed, including operations such as meshing, interpolation, filtering and noise reduction, and the data is cut into samples of the same size.

[0081] In step two, the sample is divided into grids. Based on the sliding window concept, a road surface sample is divided into several grids on average. Parameters are calculated for the road surface within each grid to obtain the distribution characteristics of texture parameters in that region. For example... Figure 2 As shown, considering that the size of the panes affects the parameter distribution, if the pane size is too large, it will be difficult to reflect the characteristics of the grid distribution; if the pane size is too small, it will be difficult to distinguish the difference between the shape of the aggregate itself and the distribution of the aggregate. Therefore, in practical applications, the pane size includes, but is not limited to, 500mm×500mm, 250mm×250mm and 100mm×100mm.

[0082] In step two, the three-dimensional texture feature parameters of each grid cell in a single sample are calculated. The three-dimensional texture feature parameters include contour parameters, height parameters, functional parameters, volume parameters, and composite parameters. The contour parameters include arithmetic mean height Ra, root mean square height Rq, arithmetic mean slope Da, root mean square slope Dq, arithmetic mean wavelength La, root mean square wavelength Lq, and mean cross-sectional depth MPD. The height parameters include skewness Ssk, kurtosis Sku, maximum peak height Sp, and maximum valley depth Sv. The functional parameters include the horizontal difference Sk at the center, the height of the prominent peak Spk, the height of the prominent valley Svk, and the height of the pole Sxp. The volume parameters include the void volume Vvv in the valley, the void volume Vvc in the center, the solid volume Vmc in the peak, and the solid volume Vmp in the center. The composite parameter includes the interface expansion ratio Sdr.

[0083] In step three, the statistical indices of the three-dimensional texture feature parameters of a single sample grid are first calculated, including the sample mean, sample standard deviation, coefficient of variation, coefficient of deviation, and kurtosis coefficient. Then, the normality test is performed on the sample parameters, and the uniformity of the sample is determined by combining the statistical indices and the distribution test results.

[0084] Among them, the sample mean of the grid division calculation parameters. The calculation is as follows:

[0085]

[0086] In the formula, m represents the number of samples, n represents the number of grids, and Y refers to the texture calculation parameters.

[0087] The sample standard deviation S is calculated as follows:

[0088]

[0089] Coefficient of variation C v The calculation is as follows:

[0090]

[0091] Deviation coefficient C s The calculation is as follows:

[0092]

[0093] Kurtosis coefficient C K The calculation is as follows:

[0094]

[0095] When testing the normality of sample parameters, the testing methods include, but are not limited to, the Lilliefers test and the Kolmogorov-Smirnov test. The null hypothesis is h; if h = 0, it assumes a normal distribution; if h = 1, it assumes a non-normal distribution. The significance level alpha is between 0.01 and 0.2, and p is the probability value of accepting the hypothesis. If p is less than alpha, the null hypothesis that the distribution is normal can be rejected.

[0096] like Figure 3 and Figure 4 As shown, in practical applications, the uniformity of road surface samples is evaluated through comprehensive analysis, based on the statistical indicators obtained from calculations and combined with existing standards and specifications.

[0097] In step four, the same distribution test is performed on multiple samples. The samples are compared pairwise. The homogeneity of variance test is used to test whether the variances of the two pairs of samples are equal. The independent two-sample t-test is also used to test whether the means of the two pairs of samples are equal. If the samples of the road surface have similar distributions in each pair, the road surface is determined to be uniformly distributed overall.

[0098] Among these, the same distribution test methods include, but are not limited to, Hartley's test, Bartlett's test, and Leyene's test (e.g., Figure 5 (As shown).

[0099] When using an independent two-sample t-test to test whether the means of two pairs of samples are equal, the test statistic t is calculated as follows:

[0100]

[0101]

[0102] In step five, if the distribution of multiple samples is inconsistent, multi-scale texture information obtained based on two-dimensional wavelet decomposition is used. Continuous wavelet transform is applied to decompose the samples in the x and y directions respectively. The decomposition level can be set to 9, with each level having the same size. For the wavelength of the decomposition, the 9 levels are level 1 (0.1–0.2 mm), level 2 (0.2–0.4 mm), level 3 (0.4–0.8 mm), ..., level 9 (>25.6 mm), and the two-dimensional level L... c The calculation is as follows:

[0103]

[0104] In the formula, a represents the decomposition level in the x-direction, and b represents the decomposition level in the y-direction. Each two-dimensional level is the sum of all signals containing that level, not exceeding its level scale.

[0105] Step five also proposes an evaluation index for the coarse-fine aggregate distribution ratio ξ, which represents the sum of the relative energy of fine aggregate (grade 1 to grade I) and coarse aggregate (grade J to the highest grade). Its calculation formula is as follows:

[0106]

[0107]

[0108]

[0109] In the formula: RE represents relative energy, E represents energy value, x and y are the x-axis and y-axis coordinates of each point in the three-dimensional data, respectively; z c,xy Let (x, y) be the height coordinates of the point (x, y) on the two-dimensional level Lc, where c is the decomposition level of the two-dimensional wavelet.

[0110] Therefore, by plotting the distribution of the coarse and fine aggregate ratio ξ (e.g., Figure 6 As shown in the figure, if the ξ value is high, it indicates that there is more fine aggregate; if the ξ value is low, it indicates that there is more coarse aggregate. The distribution of ξ values ​​among multiple samples can be used to determine whether there is an uneven distribution of coarse and fine aggregates among multiple samples.

[0111] In step six, if a single sample is determined to be heterogeneous, spatial autocorrelation analysis is used to analyze the heterogeneity of the single sample. Specifically, the global Moran index I is used to observe the overall clustering effect of the grid parameters and to measure the degree of correlation of the spatial distribution of the overall samples. The calculation of the global Moran index I is as follows:

[0112]

[0113] In the formula, w ij This represents the spatial weight value.

[0114] Furthermore, the distribution of local Moran's indices is used to characterize specific spatial aggregation phenomena in detail, in order to describe the spatial morphology and severity of local damage and surface skewness (e.g., Figure 7 and Figure 8 As shown, the local Moran index is calculated as follows:

[0115]

[0116] In summary, this technical solution employs high-precision laser 3D scanning technology to acquire pavement texture data. It utilizes mathematical statistics, two-dimensional wavelet transform, and spatial autocorrelation analysis, based on a grid-based approach, to propose a method and system for evaluating the uniformity of asphalt pavements. This enables the assessment of pavement uniformity, describing and characterizing non-uniform phenomena such as segregation, caking, and spalling, thus enriching the evaluation methods for asphalt pavements during on-site construction and operation and maintenance. This technical solution leverages high-precision 3D data to establish a refined uniformity evaluation system and method for surface regions, providing new insights for pavement evaluation in road engineering quality control, construction acceptance, and operational testing stages.

Claims

1. A method for evaluating the uniformity of three-dimensional road surface texture based on grid distribution characteristics, characterized in that, Includes the following steps: S1. Obtain three-dimensional texture data of the road surface through three-dimensional laser detection technology, and perform data preprocessing to obtain sample data; S2. Divide the sample into grids and calculate the three-dimensional texture feature parameters of each grid cell in a single sample; S3. Calculate the statistical index of the three-dimensional texture feature parameters of a single sample mesh. Based on the statistical index and the normal distribution test, evaluate the uniformity of the single sample. If the single sample is determined to be uniform, proceed to step S4; otherwise, proceed to step S6. S4. Perform an isodistribution test on multiple samples to compare the uniformity among the samples. If all samples follow the same distribution, the road surface is determined to be uniform overall. If the distributions of multiple samples are inconsistent, proceed to step S5. S5. Based on the coarse and fine aggregate distribution ratio obtained by two-dimensional wavelet decomposition, determine whether there is uneven distribution of coarse and fine aggregates among multiple samples, and obtain the global uniformity refinement evaluation result. S6. Spatial autocorrelation analysis is used to describe and characterize the local non-uniformity, and the refined evaluation results of local uniformity are obtained. The specific process of performing the same distribution test on multiple samples in step S4 is as follows: compare the samples pairwise, test whether the variances of the two pairs of samples are equal by using the homogeneity of variance test, test whether the means of the two pairs of samples are equal by using the independent two-sample T test, and if the samples of the road surface have similar distributions in pairs, then the overall distribution of the road surface is determined to be uniform. The specific process of step S5 is as follows: The ratio of coarse to fine aggregate distribution is determined by using multi-scale texture information obtained from two-dimensional wavelet decomposition. ξ The evaluation index is used to represent fine aggregate and coarse aggregate, where fine aggregate is the sum of the relative energy of grades 1 to 1, and coarse aggregate is the sum of the relative energy of grades J to the highest grade. The specific calculation formula is as follows: ， Where RE is the relative energy, E is the energy value, x and y are the x-axis and y-axis coordinates of each point in the three-dimensional data, respectively, and z... c,xy Let (x, y) be the height coordinates of the point (x, y) on the two-dimensional level Lc, where c is the decomposition level of the two-dimensional wavelet; By drawing the distribution ratio of coarse and fine aggregates ξ The distribution, if ξ A value higher than the set first threshold indicates a high amount of fine aggregate. ξ If the value is lower than the set second threshold, it indicates that there is a lot of coarse aggregate. Therefore, based on the data from multiple samples... ξ The value distribution is used to determine whether there is uneven distribution of coarse and fine aggregates among multiple samples.

2. The method for evaluating the uniformity of three-dimensional road surface texture based on grid distribution characteristics according to claim 1, characterized in that, Step S1 specifically involves using a three-dimensional laser scanning device to acquire three-dimensional texture data of the road surface and performing data preprocessing, including but not limited to meshing, interpolation, and filtering noise reduction operations, to extract the three-dimensional texture data of the road surface into multiple samples of the same size.

3. The method for evaluating the uniformity of three-dimensional road surface texture based on grid distribution characteristics according to claim 1, characterized in that, In step S2, the three-dimensional texture feature parameters of each mesh unit include contour parameters, height parameters, functional parameters, volume parameters, and composite parameters. The contour parameters include arithmetic mean height Ra, root mean square height Rq, arithmetic mean slope Da, root mean square slope Dq, arithmetic mean wavelength La, root mean square wavelength Lq, and mean section depth MPD. The height parameters include skewness Ssk, kurtosis Sku, maximum peak height Sp, and maximum valley depth Sv; The functional parameters include the horizontal difference Sk at the center, the height of the prominent peak Spk, the height of the prominent valley Svk, and the height of the pole Sxp; The volume parameters include the void volume of the valley Vvv, the void volume of the center Vvc, the solid volume of the peak Vmc, and the solid volume of the center Vmp. The composite parameter includes the interface expansion ratio Sdr.

4. A method for evaluating the uniformity of three-dimensional road surface texture based on grid distribution characteristics according to any one of claims 1 to 3, characterized in that, Step S2 specifically involves dividing the sample into grids based on the sliding window concept, so that a road surface sample is divided into several grids on average. Then, by calculating the parameters of the road surface in each grid, the distribution characteristics of the texture parameters in the grid area are obtained, which are the three-dimensional texture feature parameters of each grid unit of a single sample.

5. The method for evaluating the uniformity of three-dimensional road surface texture based on grid distribution characteristics according to claim 3, characterized in that, The statistical indicators of the three-dimensional texture feature parameters of the sample mesh in step S3 include the sample mean, sample standard deviation, coefficient of variation, coefficient of deviation, and kurtosis coefficient. The sample mean is specifically: The sample standard deviation is specifically: The coefficient of variation is specifically: The deviation coefficient is specifically: The kurtosis coefficient is specifically: in, The sample mean. m For the number of samples, n The number of grids, Y For texture calculation parameters, S The standard deviation of the sample is 1. C v The coefficient of variation is 1. C s The deviation coefficient, C K This is the kurtosis coefficient.

6. The method for evaluating the uniformity of three-dimensional road surface texture based on grid distribution characteristics according to claim 1, characterized in that, Step S3 specifically includes the following steps: S31. Calculate the statistical indicators of the three-dimensional texture feature parameters of a single sample mesh; S32. Perform a normality test on the sample parameters; S33. Combining the statistical indicators calculated in step S31 and the normal distribution test results obtained in step S32, the uniformity of a single road surface sample is comprehensively evaluated. If the single sample is determined to be uniform, then step S4 is executed; otherwise, step S6 is executed.

7. The method for evaluating the uniformity of three-dimensional road surface texture based on grid distribution characteristics according to claim 6, characterized in that, The methods for testing the normality of the distribution in step S32 include, but are not limited to, the Lilliefers test and the Kolmogorov-Smirnov test. The null hypothesis is h. If h=0, it is assumed that the distribution is normal. If h=1, it is assumed that the distribution is not normal. The significance level alpha is between 0.01 and 0.

2. p is the probability value of accepting the hypothesis. If p is less than alpha, the null hypothesis that the distribution is normal can be rejected.

8. The method for evaluating the uniformity of three-dimensional road surface texture based on grid distribution characteristics according to claim 1, characterized in that, Step S6 specifically involves using the global Moran index I to observe the overall clustering effect of the grid parameters, in order to measure the spatial distribution correlation of the overall sample. The formula for calculating the global Moran index I is as follows: in, w ij This represents the spatial weight value; Step S6 specifically involves using the distribution of the local Moran index to characterize specific spatial aggregation phenomena in detail, thereby describing the spatial morphology and severity of local damage and surface skewness. The formula for calculating the local Moran index is as follows: in, This is the local Moran index.

Images