Real-time localized restoration method and system for expressway asphalt surface layer segregation based on water seepage-flatness double-index coupling

Abstract

The invention discloses a real-time localized restoration method and system for expressway asphalt surface layer segregation based on water seepage-flatness double-index coupling, and the method comprises the steps: synchronously collecting the pavement water seepage rate and multidirectional flatness data of an expressway, and converting the water seepage rate distribution into an equivalent capillary seepage path network; performing space-time alignment fusion on the equivalent capillary seepage path network and the multi-directional flatness data, and constructing a segregation cooperative criterion by analyzing a coupling relationship between the seepage path aggregation degree and a flatness mutation region; identifying a segregation core area based on segregation collaborative criteria, and generating a graded repair topological graph according to the seepage aggregation strength and flatness mutation degree of the segregation core area; and controlling the multi-nozzle repair system to execute differential non-uniform grouting operation according to the graded repair topological graph to realize repair of the segregation area. By utilizing the embodiment of the invention, the accurate, efficient and energy-saving integrated operation from identification to repair can be realized, and the repair quality and the resource utilization efficiency are improved.

Description

Technical Field

[0001] This invention belongs to the field of highway maintenance technology, and in particular, it is a method and system for real-time localized repair of segregation in asphalt pavement of highways based on the coupling of water permeability and smoothness dual indicators. Background Technology

[0002] In the construction and maintenance of asphalt pavements on highways, material segregation is one of the key factors leading to early pavement damage. Traditional segregation identification methods often rely on single indicators, such as infrared temperature differences or manual on-site observation, which are insufficient to accurately reflect the coupling degradation relationship between the internal pore structure of the material and the external pavement function, easily leading to missed or incorrect diagnoses. In the repair phase, existing technologies often employ homogenization and large-area treatment methods, lacking precise definition of the spatial distribution and severity of segregated areas, resulting in wasted repair materials, low work efficiency, and difficulty in restoring the overall uniformity of the pavement structure. Summary of the Invention

[0003] The purpose of this invention is to provide a real-time localized repair method and system for asphalt pavement segregation on highways based on the coupling of water permeability and smoothness dual indicators, so as to overcome the shortcomings of the existing technology and realize a precise, efficient and energy-saving integrated operation from identification to repair, thereby improving repair quality and resource utilization efficiency.

[0004] One embodiment of this application provides a real-time localized repair method for segregation of asphalt pavement on highways based on the coupling of water permeability and smoothness dual indicators. The method includes: Simultaneously collect data on pavement seepage rate and multi-directional smoothness of highways, and transform the seepage rate distribution into an equivalent capillary seepage path network that characterizes the interconnected pores inside the material. The equivalent capillary seepage path network is spatiotemporally aligned and fused with the multi-directional smoothness data. By analyzing the coupling relationship between the seepage path aggregation degree and the abrupt smoothness region, a separation synergy criterion is constructed. Based on the segregation synergy criterion, the segregation core region is identified, and a graded repair topology map is generated according to the seepage aggregation intensity and the degree of abrupt change in smoothness of the segregation core region. The segregation synergy criterion identifies the segregation core region by judging the coupling relationship between the aggregation of seepage paths and the abrupt change in smoothness. The graded repair topology map defines the repair priority and material penetration depth of different segregation regions. Based on the graded repair topology diagram, the multi-nozzle repair system is controlled to perform differentiated non-uniform grouting operations to achieve the repair of segregation areas.

[0005] Another embodiment of this application provides a real-time localized repair system for segregation of asphalt pavement on highways based on the coupling of water permeability and smoothness dual indicators. The system includes: The acquisition module is used to simultaneously acquire the pavement seepage rate and multi-directional smoothness data of the highway, and transform the seepage rate distribution into an equivalent capillary seepage path network that characterizes the interconnected pores inside the material. The construction module is used to perform spatiotemporal alignment and fusion of the equivalent capillary seepage path network and the multi-directional smoothness data, and to construct a separation synergy criterion by analyzing the coupling relationship between seepage path aggregation degree and smoothness abrupt change region. The identification module is used to identify the segregation core area based on the segregation synergy criterion, and generate a graded repair topology map according to the seepage aggregation intensity and the degree of abrupt change in smoothness of the segregation core area. The segregation synergy criterion identifies the segregation core area by judging the coupling relationship between the aggregation of seepage paths and the abrupt change in smoothness. The graded repair topology map defines the repair priority and material penetration depth of different segregation areas. The repair module is used to control the multi-nozzle repair system to perform differentiated non-uniform grouting operations according to the graded repair topology map, so as to achieve the repair of segregation areas.

[0006] Another embodiment of this application provides a storage medium storing a computer program, wherein the computer program is configured to execute the method described in any of the preceding claims when running.

[0007] Another embodiment of this application provides an electronic device including a memory and a processor, wherein the memory stores a computer program and the processor is configured to run the computer program to perform the method described in any of the preceding claims.

[0008] Compared with existing technologies, the present invention provides a real-time localized repair method for asphalt pavement segregation on highways based on the coupling of water permeability and smoothness dual indicators. This method can achieve precise, efficient, and energy-saving integrated operation from identification to repair, thereby improving repair quality and resource utilization efficiency. Attached Figure Description

[0009] Figure 1 The hardware structure block diagram of a computer terminal for a real-time localized repair method for segregation of asphalt pavement on highways based on the coupling of water permeability and smoothness dual indicators is provided in an embodiment of the present invention. Figure 2 A flowchart illustrating a real-time localized repair method for segregation of asphalt pavement on highways based on the coupling of water permeability and smoothness in an embodiment of the present invention. Figure 3 This is a schematic diagram of a real-time localized repair system for segregation of asphalt pavement on highways based on the coupling of water permeability and smoothness, provided in an embodiment of the present invention. Detailed Implementation

[0010] The embodiments described below with reference to the accompanying drawings are exemplary and are only used to explain the present invention, and should not be construed as limiting the present invention.

[0011] The present invention first provides a real-time localized repair method for segregation of asphalt pavement on highways based on the coupling of water permeability and smoothness dual indicators. This method can be applied to electronic devices, such as computer terminals, specifically ordinary computers.

[0012] The following detailed explanation uses a computer terminal as an example. Figure 1 This is a hardware structure diagram of a computer terminal for a real-time localized repair method for asphalt pavement segregation on highways based on the coupling of water permeability and smoothness dual indicators, provided as an embodiment of the present invention. Figure 1 As shown, the computer device includes a processor, memory, and network interface connected via a system bus, wherein the memory may include non-volatile storage media and internal memory.

[0013] See Figure 2 The embodiments of the present invention provide a real-time localized repair method for segregation of asphalt pavement on highways based on the coupling of water permeability and smoothness dual indicators, which may include the following steps: S201, synchronously collects the pavement seepage rate and multi-directional smoothness data of highways, and transforms the seepage rate distribution into an equivalent capillary seepage path network that characterizes the interconnected pores inside the material. Specifically, it can simultaneously collect data on the seepage rate and multi-directional smoothness of highway pavement to ensure the spatiotemporal synchronization of data collection and generate an original seepage smoothness dataset. The core of this step is to acquire spatiotemporally consistent seepage and flatness data through collaborative acquisition devices and time calibration mechanisms, providing a reliable foundation for subsequent analysis. The specific implementation is as follows: First, determine the data acquisition equipment and parameter configuration. Infiltration rate data are collected using a vehicle-mounted pavement permeability meter (such as the LYS-3 model). Its core component is a 150mm diameter permeability ring (sealed in contact with the pavement to ensure that only the area within the ring is measured), equipped with a high-precision flow meter (measurement range 0-500mL / min, accuracy ±2mL / min). Data points are collected every 5 meters—the 5-meter interval is chosen because asphalt pavement segregation on highways often presents as localized, continuous patches; too dense (e.g., 1 meter) would increase data redundancy, while too sparse (e.g., 10 meters) might miss small segregation areas. Multi-directional smoothness data is collected using a multi-laser array smoothness meter (such as the TRL type). The device is equipped with 8 laser sensors (arranged along the longitudinal, transverse, and 45° diagonal directions respectively), with a sampling frequency of 50Hz (i.e., when the vehicle is traveling at 100km / h, 50 data points are collected per second, with a spatial resolution of approximately 0.56 meters), and a measurement range of ±50mm (road surface undulations exceeding this range are automatically marked as abnormal). It can simultaneously output multi-directional indicators such as the International Roughness Index (IRI) and the Transverse Force Coefficient (SFC).

[0014] Secondly, a spatiotemporal synchronization mechanism is established. Both types of equipment integrate a GPS module (positioning accuracy ±0.5 meters) and a BeiDou time synchronization module (time accuracy ±10 ns). During data acquisition, they are linked by a trigger signal—when the percolation meter starts measuring, it sends a synchronization pulse to the flatness meter. Both use this time as a timestamp (format YYYY-MM-DD HH:MM:SS.ssssss) and are associated with GPS coordinates (e.g., 30°15'22.34" N, 120°05'11.67" E). For example, at a certain data collection point at 10:23:45.123456, the percolation meter measures a percolation rate of 180 mL / min, while the flatness meter simultaneously measures an IRI of 2.3 m / km and a lateral flatness deviation of 3.5 mm at the same location. Both are bound by the same timestamp and coordinates to ensure data correspondence at the "same location, same time".

[0015] Finally, the original dataset is generated. The data is stored in real time in binary format on an onboard industrial computer (storage rate ≥10MB / s to avoid data loss). Each record contains fields such as "timestamp, GPS coordinates, seepage rate (mL / min), longitudinal flatness (mm / m), lateral flatness (mm / m), 45° diagonal flatness (mm / m), and equipment status code (0 = normal, 1 = abnormal)". For example, a complete record is "2025-10-20 10:23:45.123456, 30.256205°N, 120.086575°E, 180, 2.1, 3.5, 2.8, 0", where the equipment status code is used to mark abnormal data (e.g., status code 1 when the seepage meter is not sealed properly). This ultimately forms the original seepage flatness dataset covering the collected road sections.

[0016] The original seepage rate and flatness dataset was preprocessed by using a filtering algorithm to remove noise interference and interpolation to fill in missing data points, generating seepage rate and flatness distribution data after cleaning. This step aims to eliminate interference during data acquisition (such as equipment vibration and road impurities), fill data gaps (such as missed data collection caused by vehicles avoiding obstacles), and ensure data continuity and accuracy. The specific implementation is as follows: First, noise filtering is performed. For the seepage rate data, a Gaussian filtering algorithm is used to remove random noise. Because tiny particles on the road surface may clog the seepage holes during seepage measurement, causing instantaneous data jumps (e.g., a normal 180 mL / min suddenly changing to 50 mL / min), Gaussian filtering smooths out these sudden changes by calculating the weighted average of a data point and its eight neighboring points (the weight decreases with distance according to a Gaussian function, σ=1.2, meaning the closer the distance, the greater the weight). For example, if the original data for a point is 50 mL / min and the average of the surrounding eight points is 170 mL / min, after filtering, it is corrected to 165 mL / min, preserving the overall trend while eliminating abnormal jumps. For multi-directional flatness data, a median filtering algorithm is used to remove impulse noise. Vehicle bumps may cause instantaneous misreading of the laser sensor (such as a sudden change in lateral flatness to -20mm). The median filter takes the median of a certain point and the five data points before and after it as the new value (window size 11, corresponding to a spatial range of about 6 meters), which can effectively remove extreme values. For example, the median of the original sequence [3.5, 4.2, -20, 3.8, 4.0] is 3.8, and the sequence is smoother after replacing -20.

[0017] Next, missing data is filled in. Approximately 3%-5% of the points in the original data are missing due to GPS signal loss or temporary equipment malfunction. Kriging interpolation is used to fill in these missing points. This method is based on spatial autocorrelation (data from adjacent points is similar). It calculates the spatial weights between the missing points and known points using a variogram (e.g., a spherical model with a range of 50 meters, meaning data within 50 meters is correlated), and then sums these weights to obtain the filled value. For example, if seepage data is missing at kilometer marker K10+200 on a road section, and three surrounding points K10+195 (170 mL / min), K10+205 (190 mL / min), and K10+190 (165 mL / min) are known, Kriging interpolation calculates the filled value to be 178 mL / min, with an error of ≤2% compared to the actual subsequent verification value (175 mL / min).

[0018] Finally, distribution data is generated. The filtered and completed data is mapped to a 1m × 1m grid coordinate system (with the road starting point as the origin, the X-axis along the longitudinal direction of the road, and the Y-axis along the transverse direction). Each grid cell is assigned the average value of all data points within that area. For example, if a 1m × 1m grid contains 3 infiltration data points (170, 180, 175 mL / min), then the infiltration rate of that grid is 175 mL / min. This results in an infiltration rate distribution (one infiltration value per grid) and a smoothness distribution (each grid contains three smoothness values: longitudinal, transverse, and diagonal) covering the entire road section. The data format is GeoTIFF (supporting spatial coordinate association), providing structured input for subsequent modeling.

[0019] Based on the seepage rate distribution data after cleaning, a pore network modeling algorithm is applied to map the seepage rate values ​​to equivalent capillary diameter and connectivity parameters, generating an initial capillary seepage path network. This step transforms macroscopic seepage rates into microscopic pore characteristics, constructs a quantifiable seepage path model, and reveals the distribution patterns of interconnected pores within the material. The specific implementation is as follows: First, let's clarify the core logic of pore network modeling: the seepage rate of asphalt pavement is determined by the size (diameter) and connectivity (connectivity) of the internal interconnected pores. A higher seepage rate indicates the presence of more and larger interconnected pores. The pore network modeling algorithm treats the pavement as a network composed of countless "pore nodes" and "connecting throats." Nodes represent pores, and throats represent the connecting channels between pores. The seepage rate is used to infer the throat diameter and the node connectivity probability.

[0020] Next, parameter mapping is performed. The seepage rate (q, in mL / min) is mapped to the equivalent capillary diameter (d, in mm) using the empirical formula d = 0.002q + 0.05 (this formula was calibrated through indoor experiments: when q = 100 mL / min, the measured average throat diameter is 0.25 mm, substituting into the formula yields 0.002 × 100 + 0.05 = 0.25 mm, with an error ≤ 3%). For example, for a certain grid with a seepage rate of 180 mL / min, the corresponding equivalent capillary diameter is d = 0.002 × 180 + 0.05 = 0.41 mm. The infiltration rate is mapped to a connectivity parameter (C, unitless, range 0-1), using the exponential function C=1-e^(-0.005q) (C=0 when q=0, no connectivity; C≈0.91 when q=500, highly connected). For example, when q=180mL / min, C=1-e^(-0.005×180)=1-e^(-0.9)≈0.59, indicating that the pore connectivity probability in this region is approximately 59%.

[0021] Then, an initial network is constructed. Based on a 1m × 1m grid, each grid cell serves as a pore node, with the node coordinates being the grid center (e.g., X = 100.5m, Y = 2.5m). Node attributes include the equivalent capillary diameter *d* and connectivity *C*. The connection rule between nodes is: if the connectivity parameter of adjacent grid cells (up, down, left, right, and diagonally, a total of 8 directions) is ≥ 0.3 (critical connectivity threshold; values ​​below this are considered disconnected), a connecting throat is generated. The throat diameter is the minimum of the two node diameters (ensuring fluid preferentially flows through the smaller channel, conforming to actual seepage patterns). For example, if node A (d = 0.41mm, C = 0.59) and the right-hand node B (d = 0.38mm, C = 0.55) both satisfy *C* ≥ 0.3, a throat AB with a diameter of 0.38mm is generated. By traversing all grid nodes, an initial capillary seepage path network is formed, which includes nodes (the number is consistent with the number of grids) and throats (the number varies with connectivity). The network is stored in the form of a topology graph (node ​​ID, coordinates, d, C; throat ID, end node IDs, diameter).

[0022] The initial capillary seepage path network is topologically optimized by calculating the connectivity and aggregation coefficient of the paths, extracting key seepage path features, and finally generating an equivalent capillary seepage path network.

[0023] This step simplifies the network structure while preserving the core principles of seepage by eliminating redundant paths and retaining key features. The specific implementation is as follows: First, the topology parameters are calculated. Connectivity (K) describes the connectivity of a node and is defined as the number of throats a node has (e.g., if node A connects to 5 neighboring nodes, then K=5). Higher connectivity indicates that the node is a key hub in the seepage path. Clustering coefficient (C) describes the degree of clustering in the local network and is defined as the ratio of the actual number of throats a node has to the maximum possible number of throats (e.g., if node A has 5 neighboring nodes, the theoretical maximum number of throats is 10 (each of the 5 nodes is connected to the next), and there are actually 3, then C=3 / 10=0.3). Higher clustering coefficient indicates a denser pore distribution in the area, making it easier to form seepage clusters. For example, the core node in a segregated region has a connectivity K=8 (connecting all 8 neighboring nodes) and a clustering coefficient C=0.6, which is significantly higher than in normal areas (K=2-3, C=0.1-0.2).

[0024] Next, path selection is performed. A critical path threshold is set: nodes with connectivity K≥5 and clustering coefficient C≥0.4 are defined as core nodes, and throats connecting core nodes are defined as critical throats; other nodes and throats are considered redundant (e.g., in a normal region, nodes K=2 and C=0.1 have minimal impact on overall seepage flow). Using a depth-first search algorithm, starting from the core nodes, all paths forming critical throats are traced, and redundant nodes and throats are eliminated. For example, an initial network containing 1000 nodes, after selection, retains 200 core nodes and 350 critical throats, reducing the network size by 65% ​​but retaining over 90% of the seepage flow (verified through flow simulation: the total seepage flow of the optimized network deviates from the initial network by ≤5%).

[0025] Finally, an equivalent network is generated. The selected critical paths are simplified and represented: continuous core nodes and critical throats are merged into "main seepage channels," and the average pipe diameter (average of all throat diameters) and length (total channel length) of the channels are labeled. For example, a main channel consists of 5 core nodes, with an average pipe diameter of 0.4 mm and a length of 5 meters, representing a continuous high seepage path in that area. The equivalent capillary seepage path network is finally presented in vector graphic form, including the distribution, direction, pipe diameter, and length of the main seepage channels, clearly reflecting the spatial distribution characteristics of the interconnected pores inside the asphalt pavement, providing a structured seepage feature carrier for subsequent fusion with smoothness data.

[0026] S202, the equivalent capillary seepage path network and the multi-directional smoothness data are spatiotemporally aligned and fused, and the separation synergistic criterion is constructed by analyzing the coupling relationship between the seepage path aggregation degree and the smoothness abrupt change region. Specifically, the equivalent capillary flow path network and multi-directional flatness data can be aligned in time and space. A gridding method can be used to map the two types of data to a unified spatial coordinate system to generate a time-space aligned fused data grid. First, a unified spatial coordinate system needs to be established, with the starting point of the data collection section (e.g., K0+000 milestone) as the origin. The X-axis runs longitudinally along the road (in the direction of travel, the direction of mileage increase), with units in meters; the Y-axis runs transversely along the road (perpendicular to the direction of travel), with the inner side of the left curb as the Y=0 baseline, also with units in meters, forming a Cartesian plane rectangular coordinate system. This coordinate system needs to be calibrated with the GPS system used during the initial data collection. Coordinate transformation is performed using three known control points (e.g., GPS coordinates at milestones K0+000, K1+000, and K2+000) to ensure a transformation error ≤ 0.1 meters, avoiding data misalignment due to coordinate system deviation.

[0027] Data mapping was achieved using a 1m×1m gridding method. This size was consistent with the node grid (1m×1m) of the previous equivalent capillary seepage path network and the flatness data preprocessing grid, ensuring uniform data granularity. The specific mapping process was as follows: For the equivalent capillary seepage path network, the coordinates of each seepage node (e.g., X=100.5 meters, Y=2.5 meters, corresponding to the grid center) were directly assigned to the corresponding 1m×1m grid along with the node's equivalent capillary diameter (e.g., 0.41mm) and the length of the main seepage channel (e.g., 0.8 meters). For multi-directional flatness data, the arithmetic mean of the longitudinal, lateral, and 45° oblique flatness values ​​of all original sampling points within each grid (sampling by a laser sensor at 50Hz, approximately 2 sampling points per grid at 100km / h speed) was calculated and used as the flatness index for that grid. For example, within a certain grid (X range 100-101 meters, Y range 2-3 meters), the original longitudinal flatness values ​​are 2.1 mm / m and 2.3 mm / m, averaging to 2.2 mm / m; the original transverse values ​​are 3.5 mm / m and 3.7 mm / m, averaging to 3.6 mm / m. This grid ultimately includes both seepage parameters (pipe diameter 0.41 mm, main channel length 0.8 meters) and flatness parameters (longitudinal 2.2 mm / m, transverse 3.6 mm / m). After mapping all grids, a spatiotemporally aligned fused data grid is generated, with each grid cell achieving a one-to-one correspondence between "seepage" and "flatness" data, without any spatiotemporal bias.

[0028] Seepage path aggregation characteristics and smoothness abrupt change characteristics are extracted from the spatiotemporally aligned fused data grid, and seepage aggregation index and smoothness abrupt change index are obtained through feature calculation. The density of seepage paths needs to be quantified within a unit space. The calculation is based on the total length of the main seepage channels within a 3m × 3m neighborhood – this neighborhood covers the local concentration areas of seepage paths, avoiding the limitations of single-grid calculations. The formula for the seepage density index (C_seep) is C_seep = L / S, where L is the total length (in meters) of all main seepage channels within a 3m × 3m neighborhood, and S is the neighborhood area (9m²). 2 The unit of measurement is m / m. 2 A larger value indicates a denser seepage path. For example, in a 3m × 3m neighborhood of a merged grid, there are 3 main seepage channels with lengths of 0.6 meters, 0.8 meters, and 0.7 meters, respectively, with a total length L = 2.1 meters. Therefore, C_seep = 2.1 / 9 ≈ 0.23 m / m 2 If the total length of the main channel within the neighborhood of a certain segregated region is L = 7.2 meters, then C_seep = 7.2 / 9 = 0.8 m / m 2 It is significantly higher than that in the normal area.

[0029] The flatness abrupt change characteristic needs to reflect the flatness difference between adjacent grids. The "average flatness difference of 4 neighboring grids" is chosen as the calculation method (4 neighboring grids refer to the four directly adjacent grids in the top, bottom, left, and right directions, avoiding insignificant abrupt changes caused by diagonal adjacency). Taking the longitudinal flatness abrupt change index (C_rough_long) as an example, the calculation formula is C_rough_long = (|Δ1| +|Δ2| + |Δ3| + |Δ4|) / 4, where Δ1-Δ4 are the longitudinal flatness differences (unit: mm / m) between this grid and its four neighboring grids, respectively. A larger index value indicates a more significant flatness abrupt change. For example, if the longitudinal smoothness of a fused grid is 2.2 mm / m, the upper neighboring grid has 2.8 mm / m (Δ1=0.6), the lower one has 1.9 mm / m (Δ2=0.3), the left one has 2.3 mm / m (Δ3=0.1), and the right one has 2.7 mm / m (Δ4=0.5), then C_rough_long=(0.6+0.3+0.1+0.5) / 4=0.375 mm / m. If the longitudinal smoothness of a segregated grid is 3.5 mm / m, the neighborhood average is 2.2 mm / m, and the average difference is 1.3 mm / m, then C_rough_long=1.3 mm / m, showing a significant abrupt change. Through the above calculations, each fused grid obtains quantified seepage aggregation indices and smoothness abrupt change indices, providing basic data for subsequent coupled analysis.

[0030] The coupling coefficient between the seepage aggregation index and the smoothness mutation index was calculated using a correlation analysis algorithm, and highly coupled regions were identified by clustering methods to generate a coupling relationship map. Correlation analysis used the Pearson correlation coefficient (r), an algorithm suitable for measuring the degree of linear correlation between two continuous variables. The value range is [-1, 1], where |r| ≥ 0.8 indicates a strong correlation, 0.5 ≤ |r| < 0.8 indicates a moderate correlation, and |r| < 0.5 indicates a weak correlation. The calculation used a fused data grid as the sample, with variable X representing the seepage aggregation index (C_seep) and variable Y representing the longitudinal smoothness abrupt change index (C_rough_long). The formula r = [nΣxy - (Σx)(Σy)] / √{[nΣx... 2 - (Σx) 2 ][nΣy 2 - (Σy) 2 The calculation is performed, where n is the number of samples (i.e., the total number of fused grids). For example, if 10 fused grids of a certain road section are selected as samples, and C_seep is 0.5, 0.6, 0.7, 0.8, 0.8, 0.9, 0.9, 1.0, 1.0, 1.1 respectively, and C_rough_long is 0.4, 0.5, 0.6, 0.7, 0.7, 0.8, 0.8, 0.9, 0.9, 1.0 respectively, substituting them into the formula yields r=0.99, indicating a strong positive correlation between the two, verifying the coupling relationship between seepage accumulation and abrupt changes in smoothness.

[0031] The K-means clustering method was used to divide the samples into different regions according to their coupling degree. First, the number of clusters, K=3, was determined using the "elbow rule": the sum of squared errors (SSE) within clusters was calculated from K=1 to K=5. At K=1, SSE=2.5; at K=2, SSE=1.2; at K=3, SSE=0.3; and at K=4, SSE=0.25. The largest decrease in SSE was observed at K=3, so K=3 was chosen (high coupling, medium coupling, low coupling). The initial cluster centers were set as follows: high coupling centers (C_seep=1.0, C_rough_long=0.9), medium coupling centers (C_seep=0.7, C_rough_long=0.6), and low coupling centers (C_seep=0.4, C_rough_long=0.3). During the iteration process, the Euclidean distance from each sample to the three centers is calculated, and the sample is assigned to the nearest cluster. The cluster center is then updated to the mean of the samples in that cluster, until the change in center value is ≤0.01. The final clustering results are: high-coupling regions (C_seep≥0.8, C_rough_long≥0.7, r≥0.85), medium-coupling regions (0.6≤C_seep<0.8, 0.5≤C_rough_long<0.7, 0.6≤r<0.85), and low-coupling regions (C_seep<0.6, C_rough_long<0.5, r<0.6). The coupling relationship map is drawn using ArcGIS, based on a fused grid. High-coupling regions are marked in red, medium-coupling regions in yellow, and low-coupling regions in blue. Road mileage and lane markings are overlaid. For example, the continuous red area in the right lane from K10+200 to K10+300 is a high-coupling region.

[0032] Based on the coupling relationship graph, a separation synergy criterion is constructed, and the critical conditions for separation are determined by setting a threshold, thus generating a separation synergy criterion model.

[0033] The core logic of the segregation synergistic criterion is that "segregation must simultaneously satisfy seepage aggregation (internal pore connectivity) and abrupt changes in smoothness (uneven surface structure)" to avoid misjudgment based on a single indicator (e.g., high seepage alone may indicate localized water seepage, and abrupt changes in smoothness alone may indicate uneven compaction). The threshold is determined through indoor tests and field verification: Asphalt specimens with mild segregation (porosity 6-8%), moderate segregation (8-10%), severe segregation (>10%), and normal segregation (<4%) are prepared indoors, and their seepage aggregation degree and smoothness abrupt change values ​​are measured; combined with field core sampling (core samples from segregated areas with a porosity ≥6%), the critical threshold is finally determined: seepage aggregation degree index T1 = 0.75 m / m 2 (Lower limit of high coupling region), longitudinal flatness abrupt change index T2=0.65mm / m (lower limit of high coupling region), continuous area threshold S0=3m2 (Avoid misjudgment of a single grid).

[0034] The logical expression for the criterion is as follows: if a merged mesh satisfies (C_seep≥T1) and (C_rough_long≥T2), it is a "potentially segregating mesh"; if three or more consecutive potentially segregating meshes form a contiguous region (area ≥ S0), it is determined to be a "segregated associated region". For example, a mesh has C_seep=0.82 and C_rough_long=0.71, satisfying the threshold condition; the meshes to its right and below also satisfy the condition, forming a 3m... 2 A contiguous area is classified as a segregated region. The segregation synergy criterion model is expressed as: Segregation judgment result = 1 (segregated region), if (C_seep≥0.75)∩(C_rough_long≥0.65)∩(contiguous area≥3m²) 2 Otherwise = 0 (non-separated region). Model verification: 100 grids each of known separated road segments and normal road segments were selected. The accuracy rate of separated road segment judgment was 98%, and the misjudgment rate of normal road segments was 2%, which met the engineering accuracy requirements. Finally, a separation collaborative criterion model that can be directly applied was generated.

[0035] S203, Identify the segregation core region based on the segregation synergy criterion, and generate a graded repair topology map based on the seepage aggregation intensity and the degree of abrupt change in smoothness of the segregation core region. The segregation synergy criterion identifies the segregation core region by judging the coupling relationship between the aggregation of seepage paths and the abrupt change in smoothness. The graded repair topology map defines the repair priority and material penetration depth of different segregation regions. Specifically, the segregation synergy criterion model can be applied to the entire road surface data to calculate the segregation index of each grid cell, identify potential segregation areas based on preset thresholds, and generate a segregation index distribution map. The core of this step is to transform the segregation synergy criterion into a calculable value through quantitative indicators, thereby achieving preliminary screening of segregation regions. The specific implementation is as follows: The segregation index is calculated to comprehensively reflect the coupling degree between seepage aggregation and abrupt changes in smoothness, using a weighted summation formula: Segregation Index I = 0.6 × C_seep_norm + 0.4 × C_rough_norm. Here, C_seep_norm is the standardized value of the seepage aggregation index (the original C_seep is normalized to the [0,1] interval, the formula is C_seep_norm = (C_seep - C_seep_min) / (C_seep_max - C_seep_min), where C_seep_min is the minimum seepage aggregation degree for the entire road section, and C_seep_max is the maximum). The weight of 0.6 is because seepage aggregation degree more directly reflects the pore connectivity of internal segregation; C_rough_norm is the standardized value of the smoothness abrupt change index (calculated in the same way as C_seep_norm), with a weight of 0.4 to balance the influence of abrupt changes in surface structure. For example, for a certain grid, the original C_seep = 0.8 m / m 2 (For the entire road section, C_seep_min=0.2, C_seep_max=1.0), then C_seep_norm=(0.8-0.2) / (1.0-0.2)=0.75; the original C_rough=0.7mm / m (for the entire road section, C_rough_min=0.1, C_rough_max=1.2), then C_rough_norm=(0.7-0.1) / (1.2-0.1)≈0.545, and the final segregation index I=0.6×0.75 + 0.4×0.545≈0.45 + 0.218=0.668.

[0036] The preset threshold is determined based on the critical conditions of the segregation synergy criterion model. Through statistical analysis of 100 known segregated samples, the segregation identification accuracy reaches over 95% when I ≥ 0.7, therefore the threshold T = 0.7 is set. If a grid has I ≥ 0.7, it is marked as a "potential segregated grid"; otherwise, it is a non-segregated grid. For example, a grid with I = 0.72 satisfies I ≥ 0.7 and is marked as a potential segregated grid; another grid with I = 0.65 does not satisfy this condition and is marked as a non-segregated grid.

[0037] The segregation index distribution map was generated using GIS mapping technology. Based on a 1m×1m grid, gradient colors were used to represent the segregation index magnitude: blue (I<0.5) indicated no segregation risk, yellow (0.5≤I<0.7) indicated low risk, and red (I≥0.7) indicated potential segregation areas. Geographic information such as road lane lines and mileage markers were overlaid on the map. For example, in the right lane from K10+250 to K10+300, 20 consecutive red grids clearly showed the spatial distribution of potential segregation.

[0038] The segregation index distribution map is processed by a region growing algorithm to merge adjacent high segregation index units to form contiguous segregation core regions, and a boundary map of the segregation core regions is generated. This step uses an algorithm to merge scattered potential segregated meshes into contiguous regions, eliminates false positives for isolated meshes, and clarifies the spatial extent of segregation. The specific implementation is as follows: The core of the region growing algorithm is "seed point initiation - neighborhood expansion - conditional termination". First, seed points are selected: in the segregation index distribution map, the grid with the highest segregation index (I≥0.8) and no marking is selected as the initial seed point (the seed point must satisfy I≥0.8 to ensure that it is a segregation core). For example, if a grid has I=0.85 and there are multiple grids around it with I≥0.7, it is selected as the seed point.

[0039] Neighborhood expansion rule: Centered on the seed point, check its 8 neighboring grids (top, bottom, left, right, and diagonal). If a neighboring grid satisfies "separation index I ≥ 0.7 and has not been merged", it is included in the current region, and the 8 neighboring grids are checked again using this neighboring grid as the new expansion point. During the expansion process, the grid coordinates of each included region are recorded. For example, if the seed point coordinates (X=100.5, Y=2.5), its right grid (X=101.5, Y=2.5) has I=0.73 and is included in the region; the grid above this right grid (X=101.5, Y=3.5) has I=0.71 and is included again, until the expansion reaches a neighboring grid I < 0.7.

[0040] Growth termination condition: If, during a certain expansion round, all neighboring meshes fail to satisfy I≥0.7 or have been merged with other regions, then growth of that region stops. A minimum region area threshold of 3m is also set. 2 (i.e., at least 3 1m×1m grids), if the area of ​​the grown region is < 3m 2 If a region contains only 2 grids and has an area of ​​2m², it is considered an isolated misjudgment region and is removed. 2 One area was removed; another area contained 5 grids and had an area of ​​5m². 2 This area is retained as an effective separation region.

[0041] Generation of the boundary map of the segregated core region: For each valid segregated region, a polygon boundary is formed by using a boundary extraction algorithm (connecting the edge points of the outermost grid of the region), and the region number (e.g., R1, R2) and area (e.g., 5m²) are labeled. 2 ), and the average segregation index (e.g., 0.78). For example, the boundary of region R1 is a polygon (100,2)-(102,2)-(102,4)-(100,4), with an area of ​​6m². 2With an average I=0.76, the spatial morphology and extent of the core segregated area are clearly presented.

[0042] The intensity of seepage accumulation and the degree of abrupt change in smoothness in the segregation core area were analyzed. A multi-index weighted scoring method was used to calculate the urgency of remediation for each area, and a remediation priority scoring table was generated. This step involves quantitatively assessing the severity of the segregation areas, determining the remediation sequence, and ensuring that resources are prioritized for the areas most in need of remediation. The specific implementation is as follows: The intensity of seepage accumulation is quantified using the maximum value of the seepage accumulation index (C_seep_max) within the region. A higher value indicates a denser internal network of interconnected pores and a greater risk of water damage. For example, the maximum C_seep value in region R1 is 0.9 m / m. 2 The R2 region has a diameter of 0.7 m / m. 2 This indicates that R1 has stronger seepage accumulation.

[0043] The degree of surface roughness abrupt change is quantified by the maximum value of the surface roughness abrupt change index (C_rough_max) within the region. The higher the value, the more severe the surface undulation, and the greater the impact on driving safety and comfort. For example, C_rough_max = 0.8 mm / m in region R1 and 0.6 mm / m in region R2, indicating that the surface roughness abrupt change is more significant in region R1.

[0044] The formula for calculating the multi-index weighted score is: Repair urgency S = 0.5 × (C_seep_max / C_seep_global_max) + 0.5 × (C_rough_max / C_rough_global_max). Where C_seep_global_max is the maximum value of C_seep_max across all segregated areas in the entire road segment (e.g., 1.0 m / m). 2 C_rough_global_max is the maximum C_rough_max value for the entire road segment (e.g., 1.0 mm / m), which is normalized to eliminate the influence of dimensions; the weights are all 0.5 to balance the influence of internal structure and surface properties. For example, in region R1, C_seep_max=0.9, C_rough_max=0.8, C_seep_global_max=1.0, and C_rough_global_max=1.0, then S=0.5×(0.9 / 1.0) + 0.5×(0.8 / 1.0)=0.45+0.4=0.85; in region R2, C_seep_max=0.7, C_rough_max=0.6, then S=0.5×0.7 + 0.5×0.6=0.35+0.3=0.65, indicating that the urgency of repair in R1 is higher than that in R2.

[0045] The repair priority scoring table includes region number and area (m²). 2 ), C_seep_max (m / m 2 The table includes C_rough_max (mm / m), repair urgency S, and priority ranking (sorted in descending order of S). For example, in the table, R1 has S=0.85 and ranks 1st; R2 has S=0.65 and ranks 2nd; R3 has S=0.50 and ranks 3rd, clearly reflecting the repair urgency of each region.

[0046] Based on the repair priority scoring table and the material penetration depth requirements, the topological structure relationship of the repair area is constructed, and finally a graded repair topology map is generated.

[0047] This step integrates repair priorities and technical parameters to construct a spatial topology model that guides the repair work, ensuring that the repair process is orderly and precise. The specific implementation is as follows: The material penetration depth is determined based on the average segregation index (I_avg) of the segregated core region. The more severe the segregation (the higher the I_avg), the larger the internal pores and the stronger the connectivity, requiring a deeper penetration depth to ensure sufficient filling of the grouting material. A corresponding relationship was established through laboratory tests: I_avg ≥ 0.8 (severe segregation) → penetration depth 5cm; 0.7 ≤ I_avg < 0.8 (moderate segregation) → penetration depth 3cm; I_avg < 0.7 (slight segregation) → penetration depth 2cm. For example, in region R1, I_avg = 0.78 (moderate), penetration depth 3cm; in region R3, I_avg = 0.68 (slight), penetration depth 2cm.

[0048] The topological structure of the repair area includes spatial location relationships and repair sequence relationships. Spatial location relationships are described by coordinates to indicate the relative positions of each area, such as "R1 is located in the right lane at K10+250, and R2 is located 5 meters ahead of R1, in the same lane". The repair sequence relationship is determined by priority ranking, with the area ranked 1 being repaired first, followed by the areas ranked 2 and 3 in sequence. If areas are adjacent (distance < 2 meters), they are merged into a continuous work section to improve efficiency.

[0049] The generation of the graded repair topology map uses vector drawing technology, with the road plane as the base map. Different colored polygons represent segregated areas (red = priority 1, yellow = priority 2, blue = priority 3). The penetration depth (e.g., "3cm") is marked inside the polygons, and arrows connect the areas to indicate the repair order (arrows point from higher priority areas to lower priority areas), along with the area number and area. For example, in the map, R1 (red polygon) is marked "3cm", and the arrow points to R2 (yellow polygon, marked "3cm") on the right, clearly showing the repair order of "R1 first, then R2" and their respective technical requirements, providing intuitive guidance for subsequent grouting operations.

[0050] S204, according to the graded repair topology diagram, control the multi-nozzle repair system to perform differentiated non-uniform grouting operations to achieve the repair of segregation areas.

[0051] Specifically, it can parse the hierarchical remediation topology map, extract the coordinate range, remediation priority and required penetration depth parameters of each remediation area, and generate a remediation operation parameter table; The core of parsing the hierarchical repair topology map is to transform the graphical topology information into structured parameter data, providing a precise basis for subsequent repair operations. The specific implementation is as follows: First, ensure that the analysis tool and coordinate system are compatible. Use professional geographic information analysis software (such as ArcGIS Engine secondary development module) to read the vector data of the topology map (including polygon boundaries, attribute labels, etc.) and ensure that its coordinate system is completely consistent with the Cartesian coordinate system (X is the vertical mileage, Y is the horizontal distance) of the previous data collection. Calibrate through 3 key control points (such as the coordinates at K10+000 and K10+500) to ensure that the analysis error is ≤0.05 meters.

[0052] When extracting the coordinate range, for each segregated region's polygon boundary, the coordinate values ​​of its outermost vertex are read to determine the region's minimum bounding rectangle. The coordinate range is represented as "X start - X end, Y start - Y end". For example, the polygon vertex coordinates of segregated region R1 are (100.0, 2.0), (102.0, 2.0), (102.0, 4.0), and (100.0, 4.0). Therefore, its coordinate range is X: 100.0 meters - 102.0 meters, Y: 2.0 meters - 4.0 meters, covering a 2-meter wide and 2-meter long area in the right lane.

[0053] When extracting repair priorities, they are directly mapped according to the color labels of the regions in the topology map (red = priority 1, yellow = priority 2, blue = priority 3), and at the same time associated with the repair urgency score (e.g., R1 priority 1 corresponds to urgency 0.85, R2 priority 2 corresponds to 0.65), to ensure that the priority and urgency correspond one-to-one.

[0054] When extracting the penetration depth parameter, read the marked values ​​(such as "3cm" and "2cm") inside the region in the topology map. This value is determined based on the previous segregation index (5cm for severe segregation, 3cm for moderate segregation, and 2cm for mild segregation). For example, the R1 region is marked "3cm", indicating that the grouting material needs to penetrate to a depth of 3cm below the asphalt surface layer.

[0055] When generating the remediation operation parameter table, the extracted information above is integrated. Each record includes "region number, coordinate range (starting from X / ending from X, starting from Y / ending from Y), remediation priority (1 / 2 / 3), penetration depth (cm), and region area (m²)". 2 Fields such as "Estimated operation time (min)" are included. For example, the parameter table record for R1 is "R1, 100.0-102.0 meters, 2.0-4.0 meters, 1, 3cm, 4m". 2 "15min", where the estimated operation time is based on the area and nozzle efficiency (e.g., 1m²). 2 The estimate ( / 3min) provides a reference for subsequent construction scheduling.

[0056] Based on the repair operation parameter table, the control strategy of the multi-nozzle repair system is determined. The strategy includes the nozzle movement path, grouting pressure and grouting volume settings, and a nozzle control command sequence is generated. This step requires developing a precise control strategy based on the characteristics of the repair area and the performance of the equipment to ensure that the grouting operation is efficient and meets design requirements. The specific implementation is as follows: The multi-nozzle repair system consists of 6 grouting nozzles (arranged in a 2×3 matrix with a lateral spacing of 0.5 meters, covering a 2-meter-wide lane), mounted on an autonomous repair vehicle (positioning accuracy ±0.03 meters). Each nozzle is independently controllable (pressure and flow can be adjusted individually).

[0057] When determining the nozzle movement path, a "serpentine path + area coverage" strategy is adopted: For a rectangular area (such as 100.0-102.0 meters × 2.0-4.0 meters in R1), with the X-axis as the main direction of movement, starting from X=100.0 meters and Y=2.0 meters, move along the positive X-axis to X=102.0 meters to complete the first row of coverage; then move 0.5 meters (nozzle lateral spacing) in the positive Y-axis direction, and move in the opposite direction along the X-axis to X=100.0 meters to complete the second row of coverage, until the entire area is traversed. Path planning should avoid non-segregated areas outside the designated area, and the overlap rate should be controlled within 5% (to avoid repeated grouting and material waste). For example, area R1 requires 4 rows of paths to be completely covered.

[0058] When determining the grouting pressure, an empirical formula based on the penetration depth and degree of segregation is used: P (MPa) = 0.1 × H (cm) + 0.2 × S, where H is the penetration depth (cm) and S is the urgency of repair (0-1). This formula is calibrated through indoor tests: for every 1cm increase in depth, the pressure needs to be increased by 0.1MPa to overcome material resistance; the higher the urgency (the more severe the segregation, the larger the pores), the pressure can be appropriately increased to accelerate penetration. For example, in region R1, H = 3cm, S = 0.85, then P = 0.1 × 3 + 0.2 × 0.85 = 0.3 + 0.17 = 0.47MPa, rounded to 0.5MPa (rounded to 0.05MPa accuracy in actual control); in a region with slight segregation, H = 2cm, S = 0.5, then P = 0.1 × 2 + 0.2 × 0.5 = 0.3MPa.

[0059] When determining the grouting volume, calculate using the "volume filling method": Q(L) = V × φ × 1.2, where V is the volume of the repaired area (V = area × penetration depth, unit L, 1m). 2 ×1cm=10L), φ is the porosity of the segregated region (severe segregation φ=10%, moderate segregation 8%, mild segregation 5%), and 1.2 is the safety factor (to ensure sufficient filling). For example, the area of ​​region R1 is 4m². 2 , H=3cm (0.03m), V=4×0.03×1000=120L (1m 3 =1000L), with moderate segregation φ=8%, then Q=120×8%×1.2=11.52L, which is distributed to 6 nozzles, with each nozzle injecting approximately 1.92L of grout.

[0060] When generating the nozzle control command sequence, each command includes "timestamp, nozzle number, X coordinate, Y coordinate, grouting pressure (MPa), grouting flow rate (L / min), and execution duration (s)". For example, the command for the first path in area R1 is "10:30:00.000, No.1, 100.0, 2.0, 0.5, 1.2, 10", which means that at 10:30:00, nozzle No.1 is at the coordinate (100.0, 2.0) and grouts at a pressure of 0.5MPa and a flow rate of 1.2L / min for 10 seconds to ensure uniform grouting along the path.

[0061] Differentiated non-uniform grouting operations are performed according to the nozzle control command sequence. Grouting parameters are adjusted through real-time sensor feedback to ensure that the grouting material fills the segregation area according to the designed penetration depth and to generate real-time monitoring data of the grouting process. This step is the core execution phase of the repair operation. Closed-loop control ensures that the grouting effect meets the design requirements, and the specific implementation is as follows: When performing differentiated non-uniform grouting operations, the repair vehicle travels along a preset path (speed 5 m / min, matched to the nozzle flow rate). The six nozzles operate independently according to the command sequence: the R1 area (priority 1) uses a pressure of 0.5 MPa and a flow rate of 1.2 L / min; the R2 area (priority 2, penetration depth 3 cm, urgency 0.65) uses a pressure of 0.4 MPa and a flow rate of 1.0 L / min; and the R3 area (priority 3, penetration depth 2 cm) uses a pressure of 0.3 MPa and a flow rate of 0.8 L / min, achieving differentiated operations with different parameters for different areas. The grouting material is modified epoxy resin (viscosity 500-800 mPa·s, curing time 30 min at 25℃, good compatibility with asphalt), which maintains its fluidity through an insulated storage tank (temperature controlled at 25±2℃).

[0062] The real-time sensor feedback system includes three types of sensors: a pressure sensor (accuracy ±0.01MPa) installed in each nozzle to monitor the grouting pressure in real time; a radar depth sounder (resolution 0.1cm) installed on the bottom of the repair vehicle to monitor the actual penetration depth of the material; and a vehicle GPS (update frequency 10Hz) to record the real-time location. The feedback mechanism is as follows: when the deviation between the pressure sensor detection value and the command value exceeds ±0.05MPa (e.g., command 0.5MPa, actual 0.43MPa), the control system automatically adjusts the nozzle valve opening to increase the flow rate and raise the pressure to 0.5MPa; when the radar depth sounder shows that the actual depth is lower than the design value (e.g., design 3cm, actual 2.5cm), the grouting time in that area is extended by 10% (from 10 seconds to 11 seconds) to ensure that the depth meets the standard.

[0063] When generating real-time monitoring data for the grouting process, data is recorded every 0.5 seconds, including "time, area number, real-time coordinates (X, Y), pressure of each nozzle, actual penetration depth, cumulative grouting volume, and equipment status (normal / abnormal)". For example, a monitoring data entry might read "10:30:05.500, R1, 100.5, 2.0, No.1: 0.5MPa, No.2: 0.49MPa, 2.8cm, 5.2L, normal". This complete record of key parameters during the operation provides the original basis for subsequent quality assessment.

[0064] Based on real-time monitoring data of the grouting process, the repair effect is verified and compared with the graded repair topology map to complete the repair quality assessment and finally generate a report on the completion of segregation area repair.

[0065] This step, through effect verification and evaluation, ensures that the repair achieves the expected goals and generates a traceable report, specifically implemented as follows: When verifying the repair effect, a combination of "in-situ re-testing + sampling inspection" was adopted: the in-situ re-testing used the same permeability meter and flatness meter as the previous one, and sampling was conducted on 20% of the repaired area (e.g., 4m² in area R1). 2 Sampling 0.8m 2 The process involves detecting the seepage rate (target ≤ 50 mL / min) and multi-directional flatness (longitudinal IRI ≤ 2.0 m / km); sampling is conducted by drilling core samples (10 cm in diameter) from the repaired area to observe the penetration depth of the grouting material (≥ 90% of the design value, e.g., design 3 cm, actual ≥ 2.7 cm) and pore filling rate (microscopic observation, target ≥ 90%). For example, in area R1, the in-situ re-measurement of the seepage rate was 35 mL / min (≤ 50), and the longitudinal IRI was 1.8 m / km (≤ 2.0); the core sample test showed a penetration depth of 2.8 cm (≥ 2.7), and a filling rate of 92% (≥ 90%), indicating that the effect met the standard.

[0066] When comparing with the hierarchical repair topology map, check the overlap between the actual range of the repair area and the range marked on the topology map (target ≥95%). For example, the range of the R1 topology map is 100.0-102.0 meters × 2.0-4.0 meters, and the actual repair range is 100.1-101.9 meters × 2.1-3.9 meters, with an overlap of 96%, which meets the requirements. At the same time, check whether the repair order of each area is executed according to priority (R1 first, then R2, then R3) to ensure reasonable resource allocation.

[0067] When completing the repair quality assessment, the following assessment indicators are set: Effectiveness compliance rate (number of compliant areas / total number of areas, target 100%), parameter compliance rate (percentage of actual pressure / flow rate deviating from commanded values ​​by ≤10%, target ≥90%), and material utilization rate (actual grouting volume / theoretical calculation volume, target 90%-110%). For example, in this repair of 3 areas, the effectiveness compliance rate was 100%, the parameter compliance rate was 95%, and the material utilization rate was 105% (within a reasonable range), resulting in a comprehensive assessment of "qualified".

[0068] The generated report for the remediation of segregated areas includes sections such as "Project Information, Details of the Remediated Areas, Summary of Process Parameters, Effectiveness Verification Data, Quality Assessment Conclusions, and Attached Figures (Before and After Remediation Comparison Images, Core Sample Photos)". For example, the "Details of the Remediated Areas" section lists the coordinates, priorities, and actual penetration depths of areas R1-R3; the "Effectiveness Verification Data" section includes in-situ testing records and core sample testing photos; and the "Conclusion" section clearly states that "all segregated areas have been remediated to the required standard, and a follow-up inspection is recommended after 3 months". The report is ultimately archived in PDF format and simultaneously uploaded to the road maintenance management system, completing the entire remediation process.

[0069] Another embodiment of the present invention provides a real-time localized repair system for asphalt pavement segregation on highways based on the coupling of water permeability and smoothness dual indicators, see [link to relevant documentation]. Figure 3 The system may include: The acquisition module 301 is used to simultaneously acquire the pavement seepage rate and multi-directional smoothness data of the highway, and convert the seepage rate distribution into an equivalent capillary seepage path network that characterizes the interconnected pores inside the material. Construction module 302 is used to perform spatiotemporal alignment and fusion of the equivalent capillary seepage path network and the multi-directional smoothness data, and to construct a separation synergy criterion by analyzing the coupling relationship between seepage path aggregation degree and smoothness abrupt change region; The identification module 303 is used to identify the segregation core area based on the segregation synergy criterion, and generate a graded repair topology map according to the seepage aggregation intensity and the degree of abrupt change in smoothness of the segregation core area. The segregation synergy criterion identifies the segregation core area by judging the coupling relationship between the aggregation of seepage paths and the abrupt change in smoothness. The graded repair topology map defines the repair priority and material penetration depth of different segregation areas. Repair module 304 is used to control the multi-nozzle repair system to perform differentiated non-uniform grouting operations according to the graded repair topology map, so as to achieve the repair of segregation areas.

[0070] The above description, based on the embodiments shown in the figures, details the structure, features, and effects of the present invention. The above description is only a preferred embodiment of the present invention, but the present invention is not limited to the scope of implementation shown in the figures. Any changes made in accordance with the concept of the present invention, or equivalent embodiments modified to have equivalent changes, that do not exceed the spirit covered by the specification and figures, should be within the protection scope of the present invention.

Claims

1. A real-time localized repair method for segregation of asphalt pavement on highways based on the coupling of permeability and smoothness dual indicators, characterized in that, The method includes: Simultaneously collect data on pavement seepage rate and multi-directional smoothness of highways, and transform the seepage rate distribution into an equivalent capillary seepage path network that characterizes the interconnected pores inside the material. The equivalent capillary seepage path network is spatiotemporally aligned and fused with the multi-directional smoothness data. By analyzing the coupling relationship between the seepage path aggregation degree and the abrupt smoothness region, a separation synergy criterion is constructed. Based on the segregation synergy criterion, the segregation core region is identified, and a graded repair topology map is generated according to the seepage aggregation intensity and the degree of abrupt change in smoothness of the segregation core region. The segregation synergy criterion identifies the segregation core region by judging the coupling relationship between the aggregation of seepage paths and the abrupt change in smoothness. The graded repair topology map defines the repair priority and material penetration depth of different segregation regions. Based on the graded repair topology diagram, the multi-nozzle repair system is controlled to perform differentiated non-uniform grouting operations to achieve the repair of segregation areas.

2. The method according to claim 1, characterized in that, The synchronous acquisition of highway pavement seepage rate and multi-directional smoothness data, and the transformation of seepage rate distribution into an equivalent capillary seepage path network characterizing the interconnected pores within the material, includes: Simultaneously collect data on the seepage rate and multi-directional smoothness of the highway pavement to ensure the spatiotemporal synchronization of data collection and generate the original seepage smoothness dataset. The original seepage rate and flatness dataset was preprocessed by using a filtering algorithm to remove noise interference and interpolation to fill in missing data points, generating seepage rate and flatness distribution data after cleaning. Based on the seepage rate distribution data after cleaning, a pore network modeling algorithm is applied to map the seepage rate values ​​to equivalent capillary diameter and connectivity parameters, generating an initial capillary seepage path network. The initial capillary seepage path network is topologically optimized by calculating the connectivity and aggregation coefficient of the paths, extracting key seepage path features, and finally generating an equivalent capillary seepage path network.

3. The method according to claim 2, characterized in that, The process of spatiotemporally aligning and fusing the equivalent capillary seepage path network with the multi-directional smoothness data, and constructing a separation-coordinated criterion by analyzing the coupling relationship between seepage path aggregation and smoothness abrupt change regions, includes: The equivalent capillary flow path network and multi-directional flatness data are aligned in time and space. A gridding method is used to map the two types of data to a unified spatial coordinate system, generating a time-space aligned fused data grid. Seepage path aggregation characteristics and smoothness abrupt change characteristics are extracted from the spatiotemporally aligned fused data grid, and seepage aggregation index and smoothness abrupt change index are obtained through feature calculation. The coupling coefficient between the seepage aggregation index and the smoothness mutation index was calculated using a correlation analysis algorithm, and highly coupled regions were identified by clustering methods to generate a coupling relationship map. Based on the coupling relationship graph, a separation synergy criterion is constructed, and the critical conditions for separation are determined by setting a threshold, thus generating a separation synergy criterion model.

4. The method according to claim 3, characterized in that, The segregation core region is identified based on the segregation synergy criterion, and a graded repair topology map is generated based on the seepage aggregation intensity and the degree of abrupt change in smoothness of the segregation core region. The segregation synergy criterion identifies the segregation core region by judging the coupling relationship between the aggregation of seepage paths and the abrupt change in smoothness. The graded repair topology map defines the repair priority and material penetration depth for different segregation regions, including: The segregation synergy criterion model is applied to the entire road surface data to calculate the segregation index of each grid cell, and potential segregation areas are identified according to a preset threshold to generate a segregation index distribution map. The segregation index distribution map is processed by a region growing algorithm to merge adjacent high segregation index units to form contiguous segregation core regions, and a boundary map of the segregation core regions is generated. The intensity of seepage accumulation and the degree of abrupt change in smoothness in the segregation core area were analyzed. A multi-index weighted scoring method was used to calculate the urgency of remediation for each area, and a remediation priority scoring table was generated. Based on the repair priority scoring table and the material penetration depth requirements, the topological structure relationship of the repair area is constructed, and finally a graded repair topology map is generated.

5. The method according to claim 4, characterized in that, The step of controlling the multi-nozzle repair system to perform differentiated non-uniform grouting operations according to the graded repair topology map to achieve the repair of segregation areas includes: Analyze the hierarchical remediation topology map, extract the coordinate range, remediation priority and required penetration depth parameters of each remediation area, and generate a remediation operation parameter table; Based on the repair operation parameter table, the control strategy of the multi-nozzle repair system is determined. The strategy includes the nozzle movement path, grouting pressure and grouting volume settings, and a nozzle control command sequence is generated. Differentiated non-uniform grouting operations are performed according to the nozzle control command sequence. Grouting parameters are adjusted through real-time sensor feedback to ensure that the grouting material fills the segregation area according to the designed penetration depth and to generate real-time monitoring data of the grouting process. Based on real-time monitoring data of the grouting process, the repair effect is verified and compared with the graded repair topology map to complete the repair quality assessment and finally generate a report on the completion of segregation area repair.

6. A real-time localized repair system for segregation of asphalt pavement on highways based on the coupling of permeability and smoothness dual indicators, characterized in that, The system includes: The acquisition module is used to simultaneously acquire the pavement seepage rate and multi-directional smoothness data of the highway, and transform the seepage rate distribution into an equivalent capillary seepage path network that characterizes the interconnected pores inside the material. The construction module is used to perform spatiotemporal alignment and fusion of the equivalent capillary seepage path network and the multi-directional smoothness data, and to construct a separation synergy criterion by analyzing the coupling relationship between seepage path aggregation degree and smoothness abrupt change region. The identification module is used to identify the segregation core area based on the segregation synergy criterion, and generate a graded repair topology map according to the seepage aggregation intensity and the degree of abrupt change in smoothness of the segregation core area. The segregation synergy criterion identifies the segregation core area by judging the coupling relationship between the aggregation of seepage paths and the abrupt change in smoothness. The graded repair topology map defines the repair priority and material penetration depth of different segregation areas. The repair module is used to control the multi-nozzle repair system to perform differentiated non-uniform grouting operations according to the graded repair topology map, so as to achieve the repair of segregation areas.

7. The system according to claim 6, characterized in that, The acquisition module is specifically used for: Simultaneously collect data on the seepage rate and multi-directional smoothness of the highway pavement to ensure the spatiotemporal synchronization of data collection and generate the original seepage smoothness dataset. The original seepage rate and flatness dataset was preprocessed by using a filtering algorithm to remove noise interference and interpolation to fill in missing data points, generating seepage rate and flatness distribution data after cleaning. Based on the seepage rate distribution data after cleaning, a pore network modeling algorithm is applied to map the seepage rate values ​​to equivalent capillary diameter and connectivity parameters, generating an initial capillary seepage path network. The initial capillary seepage path network is topologically optimized by calculating the connectivity and aggregation coefficient of the paths, extracting key seepage path features, and finally generating an equivalent capillary seepage path network.

8. The system according to claim 7, characterized in that, The building module is specifically used for: The equivalent capillary flow path network and multi-directional flatness data are aligned in time and space. A gridding method is used to map the two types of data to a unified spatial coordinate system, generating a time-space aligned fused data grid. Seepage path aggregation characteristics and smoothness abrupt change characteristics are extracted from the spatiotemporally aligned fused data grid, and seepage aggregation index and smoothness abrupt change index are obtained through feature calculation. The coupling coefficient between the seepage aggregation index and the smoothness mutation index was calculated using a correlation analysis algorithm, and highly coupled regions were identified by clustering methods to generate a coupling relationship map. Based on the coupling relationship graph, a separation synergy criterion is constructed, and the critical conditions for separation are determined by setting a threshold, thus generating a separation synergy criterion model.

9. A storage medium, characterized in that, The storage medium stores a computer program, wherein the computer program is configured to execute the method of any one of claims 1-5 when it is run.

10. An electronic device comprising a memory and a processor, characterized in that, The memory stores a computer program, and the processor is configured to run the computer program to perform the method of any one of claims 1-5.

Images