A method for evaluating the spalling and loosening tendency of asphalt pavement aggregates

Abstract

The present invention proposes a method for evaluating asphalt pavement aggregate spalling and measuring loosening trends, comprising: 1) acquiring three-dimensional pavement data and performing preprocessing to obtain three-dimensional precise pavement data; 2) identifying and segmenting aggregate areas and non-aggregate areas in the three-dimensional precise pavement data to determine the aggregate spalling condition; 3) establishing an asphalt pavement looseness evaluation index based on the aggregate spalling condition, and evaluating the asphalt pavement aggregate loosening condition; 4) establishing a three-dimensional analysis model of the asphalt pavement based on the three-dimensional precise pavement data, and simulating the aggregate loosening and spalling condition of the three-dimensional analysis model of the asphalt pavement under vehicle load to estimate the deterioration development trend of the asphalt pavement; through the method of the present invention, the accuracy of spalling assessment and the measurement trend of pavement looseness can be effectively improved, which is suitable for current asphalt pavement aggregate spalling evaluation and loosening trend measurement applications, has higher objectivity and accuracy, and can thus more accurately obtain the asphalt pavement performance condition.

Description

Technical Field

[0001] The present invention relates to the technical field of preventive maintenance of asphalt pavements, and in particular to a method for evaluating the spalling of asphalt pavement aggregates and measuring their loosening tendency. Background Art

[0002] Aggregate spalling and loosening in asphalt pavements are among the earliest signs of asphalt pavement failure and are a key concern during preventive maintenance. Pavement loosening typically manifests as the spalling of aggregate within the asphalt mixture, resulting in surface roughness, exposed surfaces, peeling, missing aggregate, and even small potholes. This condition is most common in drainage pavements using porous asphalt mixtures and can manifest in two primary forms: adhesion failure due to a decrease in the interaction between the asphalt and aggregate interface, and cohesive failure due to aging or reduced fatigue durability of the asphalt mortar bonding the aggregates.

[0003] However, existing research on pavement looseness assessment primarily uses the area of ​​the loosened area as a measure of looseness. The degree of looseness is categorized based on the size of the missing aggregates: the absence of fine aggregate on the surface is considered mild looseness, while the absence of coarse aggregate on the surface is considered severe looseness. This evaluation method, due to its ambiguity and subjectivity, presents difficulties in evaluating actual looseness damage. On the one hand, in early-stage pavement damage, looseness manifests itself simply as the spalling and loss of individual aggregates, with no clear clustering between the missing aggregates. This damage is a localized phenomenon on the asphalt pavement surface, making it difficult to define its clear boundaries, which complicates the measurement of the loosened area. On the other hand, both coarse and fine aggregates may be missing simultaneously within the same loosened area, and it is difficult to discern the aggregate size of the missing areas before the loss. This presents challenges in evaluating the degree of pavement looseness. Furthermore, no research has demonstrated clear differences in the mechanisms and difficulty of spalling between coarse and fine aggregates.

[0004] Therefore, the traditional research on the evaluation of aggregate spalling and pavement loosening of asphalt pavements is highly subjective, the meaning of indicators is vague, and there is a lack of sufficient understanding of pavement loosening and aggregate spalling, which increases the difficulty of evaluating pavement aggregate spalling and measuring loosening trends. Summary of the Invention

[0005] The purpose of the present invention is to overcome the defects of the above-mentioned prior art and provide a method for evaluating the spalling and loosening tendency of asphalt pavement aggregate with higher objectivity and accuracy.

[0006] To achieve the above object, the present invention proposes a method for evaluating the spalling of asphalt pavement aggregate and measuring its loosening tendency, comprising the following steps:

[0007] S1: Acquire 3D road surface data and perform preprocessing to obtain precise 3D road surface data;

[0008] S2: Identify and segment aggregate and non-aggregate areas in the 3D precise pavement data and determine the aggregate spalling condition;

[0009] S3: Based on the aggregate spalling condition, establish an evaluation index for the looseness of asphalt pavement and evaluate the looseness of asphalt pavement aggregates;

[0010] S4: Based on the three-dimensional precise data of the pavement, a three-dimensional analysis model of the asphalt pavement is established, and the loosening and peeling condition of the aggregate of the three-dimensional analysis model of the asphalt pavement under the action of vehicle load is simulated to estimate the deterioration trend of the asphalt pavement.

[0011] Furthermore, in step S1, three-dimensional road surface data is acquired by scanning, and the three-dimensional road surface data includes x-coordinates, y-coordinates, z-coordinates of road surface points and scanned reflection intensity.

[0012] Furthermore, in step S1, the preprocessing method is: based on the 3σ rule of statistical distribution, for the z coordinate and reflection intensity, the data exceeding 3σ are regarded as abnormal points and eliminated, and linear interpolation is used to replace them to obtain three-dimensional precise road surface data.

[0013] Furthermore, in step S2, a point cloud segmentation algorithm is used to identify and segment the aggregate area and non-aggregate area in the three-dimensional precise data of the road surface. The point cloud segmentation algorithm includes a watershed algorithm, an edge segmentation algorithm, a region growing algorithm, a model fitting algorithm, a point cloud clustering algorithm or a deep learning method.

[0014] Furthermore, in step S2, the non-aggregate area is divided into: aggregate spalling area, aggregate missing area and algorithm missed identification area.

[0015] Aggregate spalling areas refer to areas of non-aggregate material formed when asphalt is peeled off from the aggregate surface under the action of rain and load, causing the aggregate to become loose or even carried away by the rolling wheels; aggregate missing areas refer to areas of non-aggregate material where pavement aggregate itself does not exist due to uneven construction; and algorithm missed identification areas refer to areas where aggregate exists but is identified as non-aggregate material by the model due to the limited accuracy of the algorithm.

[0016] Furthermore, in step S2, the method for determining the aggregate spalling condition is to identify the aggregate spalling area from the non-aggregate area. The identification method is to use SVM support vector machine for identification. The input is the z coordinate and reflection intensity index of the three-dimensional precise data of the pavement, and the output is 0 or 1 or 2; among which 0 indicates that the non-aggregate area is the aggregate spalling area; 1 indicates that the non-aggregate area is the aggregate missing area; 2 indicates that the non-aggregate area is the area that the algorithm misses to identify.

[0017] Furthermore, the z-coordinates and reflection intensity of the 3D precise pavement data are encoded and fed into a support vector machine (SVM). The encoding method uses frequency distribution. For example, the z-coordinates of the 3D precise data of the non-aggregate area are obtained, and the frequency distribution of the z-coordinates is calculated. The number of distribution groups should be no fewer than 10 and no more than 100. The frequency of each z-coordinate group is used as the z-coordinate code and fed into the SVM.

[0018] Furthermore, the z coordinate and reflection intensity index are used for splicing because the z coordinate distribution and reflection intensity index distribution in the aggregate spalling area are significantly different from those in other non-aggregate areas. Specifically, the average value of the z coordinate in the aggregate spalling area is low, and the bottom surface of the local missing area formed by aggregate spalling is relatively flat, and the peak distribution of the z coordinate is not prominent, presenting a flat peak shape; the reflection intensity index distribution in the aggregate spalling area shows obvious polarization characteristics. In the aggregate spalling area where cohesive failure occurs, the surface of the spalling area is covered with an asphalt film, and the reflection intensity is low. In the aggregate spalling area where adhesion failure occurs, the surface of the spalling area is exposed to the aggregate surface, and the reflection intensity is high at this time, close to the reflection intensity of the rock itself.

[0019] Furthermore, in step S3, the calculation formula of the asphalt pavement looseness evaluation index (RD, Rave Light Degree) is:

[0020]

[0021] Among them, S l Refers to the asphalt pavement spalling area caused by aggregate loss and spalling, which can be calculated by the number of pixels in the aggregate spalling area in the aggregate spalling identification result; S o Refers to the overall area of ​​the road surface, which can be calculated by counting the number of pixels in the area.

[0022] Furthermore, in step S4, the simulation includes an asphalt pavement simulation model, a vehicle load simulation model, and an interaction simulation. The asphalt pavement simulation model is a pavement simulation model with physical properties established based on precise three-dimensional pavement data. The vehicle load simulation model is a standard vehicle axle load set according to standards and specifications, including static axle loads and dynamic loads. The interaction simulation is a force simulation that connects the asphalt pavement simulation model and the vehicle load simulation model in series.

[0023] Furthermore, in step S4, the specific method of estimating the deterioration trend of the asphalt pavement is: using simulation to obtain the change of the looseness RD of the asphalt pavement under the vehicle load over time.

[0024] Compared with the prior art, the advantages of the present invention are:

[0025] The present invention uses high-precision laser scanning equipment to obtain precise three-dimensional data of the road surface, and combines it with a point cloud segmentation algorithm to automatically identify the spalling condition of the road aggregate, thereby achieving the purpose of evaluating the spalling of the asphalt pavement aggregate. On this basis, an evaluation index for the looseness of the asphalt pavement is established to evaluate the looseness and spalling condition of the aggregate. Simulation software is then used to simulate the decline in the service level of the asphalt pavement, effectively measuring the development trend of the road looseness. This method is suitable for the current asphalt pavement aggregate spalling evaluation and looseness trend measurement applications, and has higher objectivity and accuracy, can effectively improve the accuracy of the spalling evaluation, and more accurately obtain the performance status of the asphalt pavement. BRIEF DESCRIPTION OF THE DRAWINGS

[0026] Figure 1 This is a flow chart of the method for evaluating the spalling and loosening tendency of asphalt pavement aggregates according to the present invention;

[0027] Figure 2 This is a graph showing the aggregate identification and segmentation results in an embodiment of the present invention;

[0028] Figure 3 Schematic diagram of the non-aggregate area in an embodiment of the present invention;

[0029] Figure 4 Schematic diagram of aggregate loss distribution of loose damaged pavement in an embodiment of the present invention;

[0030] Figure 5 This is a diagram of an aggregate spalling area identification algorithm based on a support vector machine (SVM) in an embodiment of the present invention;

[0031] Figure 6 This is a diagram showing the result of identifying the aggregate spalling area in an embodiment of the present invention;

[0032] Figure 7 This is a diagram of a discrete element simulation model of an asphalt pavement in an embodiment of the present invention;

[0033] Figure 8 A diagram of a vehicle load simulation model in an embodiment of the present invention;

[0034] Figure 9 This is a simulation result diagram of the loose damage and deterioration trend of asphalt pavement in an embodiment of the present invention. DETAILED DESCRIPTION

[0035] In order to make the purpose, technical solution and advantages of the present invention clearer, the technical solution of the present invention will be further described below.

[0036] like Figure 1As shown, the present invention proposes a method for evaluating the spalling of asphalt pavement aggregates and measuring their loosening tendency. The method mainly includes data collection and preprocessing, point cloud segmentation and aggregate identification, spalling judgment and loosening evaluation, and simulation and loosening tendency measurement. The specific steps are as follows:

[0037] Step 1: Data collection and preprocessing

[0038] 1.1: First, select landmark asphalt pavement points for data collection (preferably loose areas of the pavement where aggregate spalling is clearly visible to the naked eye); scan the pavement using high-precision 3D scanning equipment. The vertical resolution of the scanning equipment should be no less than 0.01mm, the horizontal and vertical acquisition spacing should be no more than 0.05mm, and the area scanned at one time should be no less than 10cm×10cm. The data obtained from the scanning should at least include the x-coordinate, y-coordinate, and z-coordinate of the pavement point, as well as the scanned reflection intensity.

[0039] 1.2: Preprocess the 3D road surface data obtained through scanning to remove abnormal data. Specifically, based on the 3σ rule of statistical distribution, data with an error exceeding 3σ for both the z-coordinate and reflection intensity are considered outliers and removed. Linear interpolation is then used to replace the data to obtain precise 3D road surface data.

[0040] Step 2: Point cloud segmentation and aggregate identification

[0041] Based on the obtained three-dimensional precise data of the road surface, the point cloud data belonging to the aggregate area and the point cloud data belonging to the non-aggregate area are identified; the watershed algorithm is used to segment the three-dimensional precise data of the road surface (the watershed algorithm is to obtain the contours of different objects by imitating the terrain such as mountains, gullies, and basins in the geographical structure. It is a segmentation algorithm based on geographical morphology. It compares the input three-dimensional data to the geographical space, and regards the local minimum value and its affected area as a watershed. The boundaries of different watersheds, that is, the contours of the target, constitute the watershed). In the segmentation process, the lowest point of the algorithm is the starting point of the threshold, and the watershed area is gradually covered. As the threshold gradually increases, the two adjacent but isolated watersheds tend to be connected. At this time, the pixels of the convergence point are connected to form a watershed. After the depth of the three-dimensional precise data of the road surface is inverted, the aggregate area is turned into a valley, and the aggregate boundary area is turned into a peak, thereby completing the watershed segmentation. After the segmentation is completed, the aggregate area is identified, such as Figure 2 As shown in Figure 1, (a) is the aggregate area identification mask map, and (b) is the aggregate area identification overlay map. On this basis, the remaining part of the aggregate area is taken as the non-aggregate area, as shown in Figure 1. Figure 3 As shown, (a) is the non-aggregate area identification map, and (b) is the non-aggregate area identification map after superimposed display.

[0042] Step 3: Peeling Judgment and Looseness Assessment

[0043] Schematic diagram of loose road damage Figure 4 As shown in the figure, it can be seen that pavement looseness is mainly manifested in the spalling and missing of aggregate. Therefore, the identification of aggregate spalling is the basis for judging and assessing the degree of pavement looseness. Therefore, the problem can be described as identifying the spalling area from the non-aggregate area. However, from the results of aggregate segmentation and identification, there are mainly three types of non-aggregate areas: 1) Aggregate spalling area: Aggregate spalling causes the localized aggregate missing and unidentified; 2) Aggregate missing area: Aggregate uneven distribution during road construction results in the absence of aggregate at this location and is therefore unidentified; 3) Algorithm missed identification area: The performance of the segmentation model is limited, and some aggregates are missed.

[0044] For the three aforementioned situations, through analysis of their causes and combined with on-site observations, we summarized their differentiated characteristics in the laser point cloud data. The results revealed significant differences in the z-coordinate distribution and reflection intensity intensities of aggregate spalling areas compared to those in other non-aggregate areas. Specifically, the z-coordinate average value in the spalling areas is low, and the bottom surface of the localized loss areas caused by aggregate spalling is relatively flat, resulting in a flat peak. Furthermore, the reflection intensity intensities in the spalling areas exhibit distinct polarization compared to other non-aggregate areas. In areas of aggregate spalling that have undergone cohesive failure, the spalling surface is covered with an asphalt film, resulting in low reflection intensity. In areas of aggregate spalling that have undergone adhesive failure, the aggregate surface is exposed, resulting in high reflection intensity, approaching that of the rock itself.

[0045] On this basis, a classification and recognition model for non-aggregate areas is established to identify aggregate spalling areas in non-aggregate areas. The algorithm model is as follows: Figure 5 As shown below: 1) First, the z coordinates and reflection intensity intensities of the non-aggregate area in all the three-dimensional precise data of the pavement are extracted; 2) The distribution of the z coordinates and reflection intensity are counted respectively. For the z coordinates, they are divided into 50 groups from 0.25 to 0.5, and the frequency of each group is counted respectively as the z coordinate code of the non-aggregate area; for the reflection intensity intensities, they are divided into 50 groups from 0 to 3000, and the frequency of each group is counted as the reflection intensity intensities code of the non-aggregate area; 3) After splicing the two groups of codes, a 1×100 vector is formed, which is input into the SVM support vector machine. The support vector machine SVM outputs a 1×3 vector, which represents that the non-aggregate area belongs to 0 - aggregate spalling area, 1 - aggregate missing area, 2 - algorithm missed area. The recognition results are shown as follows: Figure 6As shown, (a) is a schematic diagram of the three-dimensional accuracy data of the pavement, (b) is the identified non-aggregate area, and (c) is the identified aggregate spalling area.

[0046] Finally, the area S of the aggregate spalling area is calculated separately. l and the area of ​​the scan area S o The ratio of the two is used as the evaluation index RD of the looseness of asphalt pavement. The performance formula is:

[0047]

[0048] Where S l Refers to the asphalt pavement spalling area caused by aggregate loss and spalling, which can be calculated by the number of pixels in the aggregate spalling area in the aggregate spalling identification result; S o Refers to the total area of ​​the road surface, which can be calculated by the total number of pixels in the scanned area.

[0049] Step 4: Simulate the loose trend measure

[0050] After identifying the aggregate spalling area and calculating the evaluation index of the pavement looseness, the looseness condition of the current asphalt pavement can be understood. By further simulating the aging process of the asphalt pavement, the deterioration trend of the loose damage of the asphalt pavement can be obtained, providing a theoretical basis for road maintenance and management. The details are as follows:

[0051] 4.1: First, based on the 3D precise data of the pavement and the identified aggregate area, a discrete element simulation model of the asphalt pavement is established, such as Figure 7 As shown, at the same time, the vehicle load simulation model is established as Figure 8 As shown, in addition, when establishing the interaction model, the main considerations are 1) dynamically attenuating the bonding strength of the asphalt elements while applying the moving load to simulate the impact of the dynamic water pressure on the asphalt film and the aging effect of the asphalt; 2) simplifying the definition conditions of the spalling of the surface aggregate elements, and simplifying all the contact fractures of the original elements and the substantial displacement from the road surface to only judging whether the contact relationship is broken. Because under the action of a simple load, the asphalt and aggregate elements are less subject to upward forces, even if the elements have broken in contact, they may not necessarily have substantial displacement behavior from the road surface. Therefore, this simplification of the definition can more realistically measure the proportion of aggregate spalling. On this basis, the discrete element software was used for simulation and the degree of looseness RD of the asphalt pavement was evaluated, and the results were obtained as shown below. Figure 9 shown.

[0052] The looseness rate variation chart shows that early loosening damage in asphalt pavements primarily progresses through three stages: early loosening formation, mid-stage loosening exacerbation, and late-stage loosening stability. In the early loosening formation stage, the pavement's looseness is low and its deterioration trend is slow. During this stage, the aggregate spalling rate is generally low, but fluctuates significantly. When the aggregate spalling ratio in the asphalt mixture reaches a certain level, the pavement enters the mid-stage loosening exacerbation stage. During this stage, the overall contact force network of surface elements undergoes significant structural changes, losing its previously stable state. During this stage, the loosening deterioration trend and aggregate spalling rate accelerate, leading to rapid loosening. When the pavement looseness reaches a higher level, the pavement enters the late-stage loosening stability stage. During this stage, the local contact force network structure is destroyed, leaving only localized areas capable of independently resisting vehicle loads and hydrodynamic pressure. However, the remaining unstripped areas are typically areas with strong cohesion and enhanced spalling resistance. At this stage, the aggregate spalling rate decreases significantly, marking the entry into the loosening stability stage.

[0053] Implementation Cases

[0054] This embodiment proposes a method for evaluating the spalling of asphalt pavement aggregate and measuring its loosening tendency, which is as follows:

[0055] (1) Scan the road surface to obtain three-dimensional road surface information and preprocess it

[0056] In this embodiment, a road surface texture laser scanner is used to scan the road surface. The vertical resolution of the scanning instrument is 0.01 mm, the horizontal acquisition spacing is 0.0415 mm, the longitudinal acquisition spacing is 0.0496 mm, and the area scanned at one time is 10 cm × 10 cm. The scanning obtains three-dimensional information, including the x-coordinate, y-coordinate, and z-coordinate of the road surface point and the scanned reflection intensity information.

[0057] The preprocessing adopts the 3σ rule based on statistical distribution. The data with z-coordinate and reflection intensity information exceeding 3σ are regarded as abnormal points and removed. Linear interpolation is then used to replace them to obtain the three-dimensional precise data of the road surface.

[0058] (2) Aggregate identification and spalling judgment:

[0059] Based on the pre-processed 3D pavement data, a watershed algorithm was used to identify the aggregate areas within the pavement, and the remaining areas within the aggregate areas were taken as the non-aggregate areas. The z-coordinate distribution and reflection intensity distribution of each aggregate area were statistically analyzed, and a support vector machine (SVM) was used to identify the areas within the non-aggregate areas that were considered to be spalled aggregate areas.

[0060] In this example, 29 10cm x 10cm asphalt pavement points were collected, resulting in 369 non-aggregate areas. Manual annotation confirmed that 196 of these non-aggregate areas were spalled aggregate areas, 160 were missing aggregate areas, and 40 were areas missed by the algorithm. The results of using SVM to identify spalled aggregate areas in non-aggregate areas are shown below:

[0061] Table 1

[0062] Number of samples The true value is 0 The true value is 1 The true value is 2 The predicted value is 0 183 1 0 184 The predicted value is 1 5 157 2 164 The predicted value is 2 8 2 38 48 196 160 40 396

[0063] Table 1 above shows the results of the support vector machine (SVM) in identifying the non-aggregate area. As can be seen from Table 1, the recognition accuracy of the support vector machine (SVM) model is 95.45%.

[0064] (3) Looseness assessment

[0065] For each scanned 10 cm × 10 cm area, the size of the aggregate spalling area is counted and the looseness degree RD is calculated using the following formula:

[0066]

[0067] From the above calculation results, it can be seen that the looseness RD of the scanned road surface area is 9.75%.

[0068] (4) Simulation of loosening trend measurement: Using discrete element simulation software, a three-dimensional analysis model of asphalt pavement is established, and the loosening and peeling of aggregates under vehicle load is simulated to estimate the deterioration trend of the asphalt pavement.

[0069] The three-dimensional analysis model of asphalt pavement established by discrete element simulation software is as follows: Figure 7 As shown in Figure 2, the established vehicle load simulation model is as follows: Figure 8 shown.

[0070] The corresponding material parameters of the established pavement model are given according to the measured data, as shown in Table 2 below:

[0071] Table 2

[0072] Structural layer Macro modulus (MPa) Contact stiffness (N / m) Asphalt stabilized gravel 1200 <![CDATA[2.4×10 11 ]]> Graded gravel 400 <![CDATA[8×10 6 ]]> lime soil 400 <![CDATA[8×10 6 ]]> soil base 50 <![CDATA[2×10 6 ]]>

[0073] In addition, in Figure 8 In the simulation model of the vehicle load shown, the design axle load and wheel parameters are shown in Tables 3 and 4 below:

[0074] Table 3

[0075]

[0076] Table 4

[0077] Tire weight (kg) Tread width (m) Tire diameter (m) Center distance between two wheels (m) 50 0.28 1.085 0.3195

[0078] In this embodiment, the loose damage deterioration trend of the asphalt pavement obtained by simulation is as follows: Figure 9 shown.

[0079] The above description is merely a preferred embodiment of the present invention and does not limit the present invention in any way. Any person skilled in the art who, without departing from the scope of the present invention, makes any equivalent substitution, modification, or other changes to the technical solution and technical content disclosed in the present invention shall be deemed to be within the scope of the present invention and still fall within the scope of protection of the present invention.

Claims

1. A method for evaluating the spalling and loosening tendency of asphalt pavement aggregate, characterized in that: The process includes the following steps: S1: Acquire 3D road surface data and perform preprocessing to obtain precise 3D road surface data; S2: Identify and segment aggregate and non-aggregate areas in the 3D precise pavement data and determine the aggregate spalling condition; In step S2, a point cloud segmentation algorithm is used to identify and segment aggregate areas and non-aggregate areas in the three-dimensional precise road surface data. The point cloud segmentation algorithm includes a watershed algorithm, an edge segmentation algorithm, a region growing algorithm, a model fitting algorithm, a point cloud clustering algorithm, or a deep learning method. In step S2, the method for determining the aggregate spalling condition is to identify the aggregate spalling area from the non-aggregate area. The identification method is to use an SVM support vector machine for identification. The input is the z coordinate and reflection intensity of the three-dimensional precise data of the road surface, and the output is 0, 1, or 2. Among them, 0 indicates that the non-aggregate area is an aggregate spalling area; 1 indicates that the non-aggregate area is an aggregate missing area; 2 indicates that the non-aggregate area is an area that is missed by the algorithm. S3: Based on the aggregate spalling condition, establish an evaluation index for the looseness of asphalt pavement and evaluate the looseness of asphalt pavement aggregates; The calculation formula for the evaluation index of looseness of asphalt pavement is: ； in, S l Refers to the asphalt pavement spalling area caused by aggregate loss and spalling, which is calculated by the number of pixels in the aggregate spalling area in the aggregate spalling identification result; S o Refers to the overall road surface area, calculated by counting the number of pixels in the area; S4: Based on the three-dimensional precise data of the pavement, a three-dimensional analysis model of the asphalt pavement is established, and the loosening and spalling condition of the aggregate of the three-dimensional analysis model of the asphalt pavement under the action of vehicle load is simulated to estimate the deterioration trend of the asphalt pavement; In step S4, the specific method of estimating the deterioration trend of the asphalt pavement is to use simulation to obtain the change of the looseness RD of the asphalt pavement under the vehicle load over time.

2. The method for evaluating the spalling and loosening tendency of asphalt pavement aggregate according to claim 1 is characterized in that: In step S1, three-dimensional road surface data is acquired by scanning, wherein the three-dimensional road surface data includes x-coordinates, y-coordinates, z-coordinates of road surface points and scanned reflection intensity.

3. The method for evaluating the spalling and loosening tendency of asphalt pavement aggregate according to claim 2 is characterized in that: In step S1, the preprocessing method is: based on the 3σ rule of statistical distribution, for the z coordinate and reflection intensity, the data exceeding 3σ are regarded as abnormal points and eliminated, and linear interpolation is used to replace them to obtain three-dimensional precise road surface data.

4. The method for evaluating asphalt pavement aggregate spalling and measuring loosening tendency according to claim 1, characterized in that: In step S2, the non-aggregate area is divided into: aggregate spalling area, aggregate missing area and algorithm missed identification area.

5. The method for evaluating asphalt pavement aggregate spalling and measuring loosening tendency according to claim 1, characterized in that: The z coordinate and reflection intensity of the three-dimensional road surface precision data are encoded and uniformly input into the SVM support vector machine, and the encoding method is to encode through frequency distribution.

6. The method for evaluating asphalt pavement aggregate spalling and measuring loosening tendency according to claim 1, characterized in that: In step S4, the simulation includes an asphalt pavement simulation model, a vehicle load simulation model, and an interaction simulation.

Images