Asphalt concrete heat-induced healing safety regulation and control method based on fatigue damage evaluation

Abstract

The invention belongs to the technical field of automatic control, and particularly relates to an asphalt concrete heat-induced healing safety regulation and control method based on fatigue damage evaluation. Comprising the following steps: step 1, scanning an asphalt concrete pavement of a target road section through a vehicle-mounted detection device to generate a to-be-healed unit sequence; step 2, moving the phased array microwave heating device to a position above the current unit to be healed, establishing a dielectric constant three-dimensional grid model based on the dielectric constant three-dimensional distribution data body, and determining a focusing target point; the microwave energy is focused on the focusing target point to form a three-dimensional focusing hot area; monitoring the surface temperature and executing heat preservation control; and step 3, retesting the current to-be-healed unit after healing, carrying out statistics on the healing standard-reaching rate of the road section and outputting a safety regulation and control report. According to the invention, the technical problem of accurate focusing of microwave energy in a non-uniform asphalt concrete medium is solved, and the pertinence and effectiveness of heat-induced healing are improved.

Description

Technical Field

[0001] This invention relates to the field of automatic control technology, specifically to a method for safe regulation of thermally induced healing of asphalt concrete based on fatigue damage assessment, which belongs to the category of industrial control software. Background Technology

[0002] As a viscoelastic material, asphalt concrete possesses self-healing capabilities under specific temperature conditions. When the pavement temperature rises above the asphalt softening point, the viscosity of the asphalt binder decreases, and the molecular chains gain fluidity, enabling them to diffuse and entangle across the crack interface, thereby achieving crack healing and closure. This characteristic provides a new technical approach for pavement maintenance: inducing self-healing in asphalt concrete through artificial heating to actively repair fatigue cracks. Compared to traditional passive repair methods, thermally induced healing technology can intervene in the early stages of crack development, resulting in better repair effects, higher construction efficiency, and less impact on traffic.

[0003] Currently, heating methods used for thermally induced healing of asphalt concrete mainly include infrared heating, hot air heating, and microwave heating. Infrared heating transfers heat to the pavement surface through infrared radiation, and then conducts it from the surface to the interior. This heating method has high energy utilization efficiency, but heat transfer depends on the heat conduction process, resulting in the surface temperature of the pavement being much higher than the internal temperature. This can easily cause the surface asphalt to overheat and age, while deep cracks fail to reach the effective healing temperature. Hot air heating uses high-temperature airflow to heat the pavement, resulting in a temperature gradient problem similar to infrared heating, and is greatly affected by ambient wind speed, making it difficult to guarantee heating uniformity. Microwave heating utilizes the interaction between microwave electromagnetic fields and medium molecules to generate heat, enabling volumetric heating of the pavement structure and alleviating the problem of excessive temperature gradients to some extent. However, existing microwave heating equipment usually uses a single radiation source or a simple radiation source array. The distribution of microwave energy inside the pavement is mainly determined by the medium characteristics, lacking precise control over the heating area and depth, making it difficult to concentrate heat on the cracks that need healing.

[0004] Asphalt concrete pavement is a non-homogeneous medium composed of asphalt binder, aggregates, and voids, with significant differences in dielectric constants among its components. When microwaves propagate in this non-homogeneous medium, refraction, scattering, and multipath effects occur, causing the energy distribution to deviate from the expected value. While traditional phased array beamforming technology can achieve directional heating of target points in homogeneous media, its phase calculations are based on the assumption of a homogeneous medium and cannot accurately compensate for the phase distortion caused by the non-homogeneous medium, often resulting in a significantly reduced focusing effect. Therefore, how to achieve precise energy focusing at crack locations in non-homogeneous asphalt concrete media is a key technical challenge for thermally induced healing technology. Summary of the Invention

[0005] The main objective of this invention is to provide a safe control method for thermally induced healing of asphalt concrete based on fatigue damage assessment. This method achieves closed-loop control of damage identification, directional heating, and effect verification, solves the technical challenge of precise focusing of microwave energy in non-uniform asphalt concrete media, and improves the pertinence and effectiveness of thermally induced healing.

[0006] To address the aforementioned technical problems, this invention provides a method for safe regulation of thermally induced healing in asphalt concrete based on fatigue damage assessment, comprising the following steps: Step 1: Fatigue damage state identification and determination of areas to be healed: The asphalt concrete pavement of the target road section is scanned using an on-board detection device to obtain pavement images and ground-penetrating radar echo data; crack density and crack connectivity values ​​are extracted based on the pavement images, and a three-dimensional distribution data volume of dielectric constant is generated based on the ground-penetrating radar echo data, and crack depth values ​​are extracted; the pavement is divided into grid cells, and the units to be healed are determined based on the crack density, crack connectivity, and crack depth values ​​of each grid cell, generating a sequence of units to be healed; Step 2: Phased array microwave directional heating based on inverse-time focusing algorithm: Move the phased array microwave heating device above the current unit to be healed, establish a three-dimensional grid model of dielectric constant based on the three-dimensional distribution data of dielectric constant, and determine the focusing target point; perform inverse-time focusing phase calculation, set a virtual point source at the focusing target point and simulate the electromagnetic wave propagation process, obtain the receiving phase at the corresponding position of each microwave radiation unit, and perform phase conjugation processing on the receiving phase to obtain the transmitting phase; load the transmitting phase onto each microwave radiation unit, control each microwave radiation unit to simultaneously emit microwave signals, so that the microwave energy is focused on the focusing target point to form a three-dimensional focusing hot zone; monitor the surface temperature and perform heat preservation control; Step 3, Healing Effect Verification and Safety Control Output: Retest the current unit to be healed after healing, obtain the crack density value, crack connectivity value, and maximum crack depth value after healing, calculate the decrease rate of each indicator and determine the healing status; traverse all units to be healed in the sequence of units to be healed, count the road section healing compliance rate and output a safety control report.

[0007] Furthermore, in step 1, the process of extracting crack density and crack connectivity values ​​based on the road surface image includes: inputting the road surface image into the crack segmentation model and outputting a crack binary mask; performing skeleton refinement processing on the crack binary mask to obtain a crack skeleton image with a single pixel width; counting the total number of crack pixels on the crack skeleton image and dividing it by the road surface area covered by the image to obtain the crack density value; identifying all crack intersections on the crack skeleton image, counting the number of crack intersections and dividing it by the total number of crack pixels to obtain the crack connectivity value.

[0008] Furthermore, in step 1, the process of generating a three-dimensional distribution data volume of dielectric constant based on ground-penetrating radar echo data and extracting crack depth values ​​includes: performing time-depth conversion on the ground-penetrating radar echo data to generate a three-dimensional distribution data volume of dielectric constant of the pavement structure; in the three-dimensional distribution data volume of dielectric constant, areas with dielectric constant values ​​lower than the reference dielectric constant of asphalt are marked as crack void areas, and the maximum depth value of each crack void area is extracted as the crack depth value.

[0009] Furthermore, in step 1, the process of determining the units to be healed and generating a sequence of units to be healed includes: dividing the road surface into square grid units with equal side lengths; for each grid unit, calculating the crack density value, crack connectivity value, and maximum crack depth value within the grid unit; when the crack density value of any grid unit exceeds a density threshold, or the crack connectivity value exceeds a connectivity threshold, or the maximum crack depth value exceeds a depth threshold, the corresponding grid unit is marked as a unit to be healed; and sorting all units to be healed in descending order of maximum crack depth value to generate a sequence of units to be healed.

[0010] Furthermore, in step 2, the phased array microwave heating device includes multiple microwave radiation units arranged in a rectangular array. Each microwave radiation unit is equipped with an independent phase shifter and a power amplifier. The phase adjustment range of the phase shifter is from 0 degrees to 360 degrees.

[0011] Furthermore, in step 2, the process of establishing a three-dimensional dielectric constant mesh model based on the three-dimensional dielectric constant distribution data volume and determining the focusing target point includes: extracting local dielectric constant distribution data of the corresponding region of the current unit to be healed from the three-dimensional dielectric constant distribution data volume; discretizing the spatial range covered by the local dielectric constant distribution data into a cubic mesh, with each cubic mesh storing the dielectric constant value at the corresponding position, forming a three-dimensional dielectric constant mesh model; the horizontal position of the focusing target point is the geometric center of the point with the maximum crack density in the current unit to be healed, and the depth position of the focusing target point is half of the maximum crack depth value of the current unit to be healed.

[0012] Furthermore, in step 2, the process of performing the reverse-time focusing phase calculation includes: setting a virtual point source at the focusing target point position, and the virtual point source emitting spherical electromagnetic waves into the surrounding space; simulating the propagation process of the spherical electromagnetic waves in the three-dimensional grid model of the dielectric constant using the finite-difference time-domain method; recording the electric field time-domain waveform at the corresponding position directly below each microwave radiating unit when the spherical electromagnetic waves propagate to the top surface of the three-dimensional grid model of the dielectric constant; for the electric field time-domain waveform recorded at the position directly below each microwave radiating unit, selecting the complete period with the largest amplitude in the electric field time-domain waveform as the reference period, and recording the phase angle of the moment when the electric field changes from negative to positive at the zero-crossing point within the reference period relative to the start time of the reference period as the receiving phase of the corresponding microwave radiating unit.

[0013] Furthermore, the process of obtaining the transmission phase by performing phase conjugation processing on the received phase includes: for the received phase of each microwave radiation unit, calculating the difference between 360 degrees and the received phase as the transmission phase of the corresponding microwave radiation unit; if the difference exceeds 360 degrees, subtracting 360 degrees; if the difference is negative, adding 360 degrees to ensure that the transmission phase value is within the range of 0 degrees to 360 degrees; loading the transmission phase of each microwave radiation unit onto the corresponding phase shifter, controlling all microwave radiation units to transmit microwave signals simultaneously, so that the microwave signals transmitted by each microwave radiation unit are superimposed in phase at the focusing target point, forming a three-dimensional focusing hot zone.

[0014] Furthermore, in step 2, the process of monitoring surface temperature and performing heat preservation control includes: continuously acquiring surface temperature distribution images of the current unit to be healed using an infrared thermal imager, and extracting the highest temperature value and average temperature value from each frame of surface temperature distribution image; when the highest temperature value reaches the preset upper limit threshold of the highest temperature, reducing the output power of the power amplifiers of all microwave radiation units and entering the heat preservation stage; during the heat preservation stage, periodically detecting the average temperature value, increasing the output power of the power amplifiers when the average temperature value is lower than the preset lower limit threshold of the average temperature, and reducing the output power of the power amplifiers when the average temperature value is higher than the preset upper limit threshold of the average temperature; and turning off all microwave radiation units after the heat preservation stage ends.

[0015] Furthermore, in step 3, the process of calculating the decline rate of each indicator and determining the healing status includes: calculating the difference between the crack density value and the crack density value after healing, and dividing the difference by the crack density value to obtain the crack density decline rate; calculating the difference between the crack connectivity value and the crack connectivity value after healing, and dividing the difference by the crack connectivity value to obtain the crack connectivity decline rate; calculating the difference between the maximum crack depth value and the maximum crack depth value after healing, and dividing the difference by the maximum crack depth value to obtain the crack depth decline rate; when the crack density decline rate, crack connectivity decline rate, and crack depth decline rate are all greater than the preset decline rate threshold, the current unit to be healed is marked as having reached the healing standard; the ratio of the number of all units to be healed marked as having reached the healing standard to the number of all units to be healed that have been processed is used as the road segment healing standard compliance rate; the road segment safety status label is determined based on the road segment healing standard compliance rate, and a safety control report is output.

[0016] The method for controlling the safety of thermally induced healing in asphalt concrete based on fatigue damage assessment, as described in this invention, has the following beneficial effects: This invention not only acquires the planar distribution characteristics of cracks on the road surface but also detects the depth information of cracks extending into the road surface, achieving a comprehensive understanding of the three-dimensional distribution of cracks. Through the joint evaluation of three indicators—crack density, crack connectivity, and maximum crack depth—the severity of fatigue damage can be accurately assessed from three dimensions: crack area, network topology, and longitudinal propagation, providing a scientific basis for subsequent heat healing treatment. Prioritizing the units to be healed based on damage severity ensures that maintenance resources are preferentially invested in the most severely damaged areas, improving overall maintenance efficiency.

[0017] This invention effectively solves the technical challenge of achieving precise energy focusing in non-uniform asphalt concrete media. Traditional beamforming methods calculate phase delay based on the assumption of a uniform medium, which cannot accurately compensate for phase distortion caused by differences in the dielectric constants of different components such as asphalt binder, aggregates, and voids. This invention simulates the propagation process of electromagnetic waves emitted by a virtual point source in a three-dimensional grid model of dielectric constants, obtains the received phase including the influence of medium inhomogeneity, and then performs phase conjugation processing on the received phase to obtain the emitted phase. Utilizing the time-reversal symmetry of electromagnetic wave propagation, the microwave signals emitted by each microwave radiation unit automatically converge in the opposite direction along the original propagation path to the focusing target point. This method does not require precise measurement of the electromagnetic parameter distribution of the medium; the optimal phase configuration can be obtained solely through numerical simulation. It achieves directional heating in the crack depth direction, avoiding the problem of surface overheating and insufficient deep heating in traditional heating methods, significantly improving heat utilization efficiency and healing effect. Attached Figure Description

[0018] Figure 1 A schematic diagram of the characteristic curve of road surface temperature changing with time during microwave heating provided in an embodiment of the present invention; Figure 2 A schematic diagram of the electromagnetic wave propagation process in the forward propagation stage of the reverse-time focusing algorithm provided in an embodiment of the present invention; Figure 3 A schematic diagram of the electromagnetic wave focusing process during the backward propagation stage in the reverse-time focusing algorithm provided in an embodiment of the present invention; Figure 4 This is a schematic diagram of the adaptive adjustment curve of the power amplifier output power changing with time during microwave heating, provided in an embodiment of the present invention. Detailed Implementation

[0019] The method of the present invention will be further described in detail below with reference to the accompanying drawings and embodiments.

[0020] A method for controlling the safety of thermally induced healing in asphalt concrete based on fatigue damage assessment includes the following steps: Step 1: Fatigue damage state identification and determination of areas to be healed: The asphalt concrete pavement of the target road section is scanned using an on-board detection device to obtain pavement images and ground-penetrating radar echo data; crack density and crack connectivity values ​​are extracted based on the pavement images, and a three-dimensional distribution data volume of dielectric constant is generated based on the ground-penetrating radar echo data, and crack depth values ​​are extracted; the pavement is divided into grid cells, and the units to be healed are determined based on the crack density, crack connectivity, and crack depth values ​​of each grid cell, generating a sequence of units to be healed; Step 2: Phased array microwave directional heating based on inverse-time focusing algorithm: Move the phased array microwave heating device above the current unit to be healed, establish a three-dimensional grid model of dielectric constant based on the three-dimensional distribution data of dielectric constant, and determine the focusing target point; perform inverse-time focusing phase calculation, set a virtual point source at the focusing target point and simulate the electromagnetic wave propagation process, obtain the receiving phase at the corresponding position of each microwave radiation unit, and perform phase conjugation processing on the receiving phase to obtain the transmitting phase; load the transmitting phase onto each microwave radiation unit, control each microwave radiation unit to simultaneously emit microwave signals, so that the microwave energy is focused on the focusing target point to form a three-dimensional focusing hot zone; monitor the surface temperature and perform heat preservation control; Step 3, Healing Effect Verification and Safety Control Output: Retest the current unit to be healed after healing, obtain the crack density value, crack connectivity value, and maximum crack depth value after healing, calculate the decrease rate of each indicator and determine the healing status; traverse all units to be healed in the sequence of units to be healed, count the road section healing compliance rate and output a safety control report.

[0021] The process of fatigue damage state identification and determination of the area to be healed involves multiple steps, including road image acquisition, ground penetrating radar detection, crack feature extraction, and division of the area to be healed. The specific implementation methods of each step are explained in detail below.

[0022] The vehicle-mounted detection device is installed under the chassis of the detection vehicle, maintaining a fixed distance from the road surface. In this embodiment, the vehicle-mounted detection device includes two types of sensors: a line-scan camera and a ground-penetrating radar. The line-scan camera is mounted perpendicular to the driving direction, and its photosensitive element is a single-row pixel array with 4096 pixels, each pixel being 7 micrometers in size. When the detection vehicle travels at a constant speed, the line-scan camera continuously acquires road surface images at a fixed frequency. The distance the vehicle travels between two adjacent acquisitions determines the image resolution in the driving direction. For example, when the vehicle speed is 20 kilometers per hour and the frequency is 2000 lines per second, the resolution in the driving direction is approximately 2.78 millimeters. The focal length of the line-scan camera lens and its mounting height together determine the imaging width perpendicular to the driving direction. In this embodiment, the imaging width covers a single lane, approximately 3.75 meters. By stitching together the continuously acquired multi-row pixel data in chronological order, a road surface image covering the entire scanned section can be formed.

[0023] Ground-penetrating radar (GPR) uses an air-coupled antenna, operates at a frequency of 2 GHz, and is positioned 300 mm above the road surface. The GPR emits electromagnetic pulse (EMP) signals, which are reflected at different interfaces within the road structure. These reflected signals are received and recorded by the receiving antenna. Due to differences in the dielectric constants of media such as asphalt concrete, base materials, cracks, and moisture, the reflection intensity and propagation speed of the EMP signals vary at each interface. This provides a physical basis for identifying the internal structure and damage state of the road surface. The GPR continuously collects data along the driving direction at 50 mm intervals, forming a series of echo data arranged along the depth direction. Arranging these echo data according to their acquisition locations constitutes the GPR echo data.

[0024] The first step in processing road surface images is to extract crack information. The road surface image is input into a pre-trained crack segmentation model. This model employs a convolutional neural network with an encoder-decoder structure. The input is a grayscale road surface image, and the output is a binary crack mask of the same size as the input image. In the crack binary mask, crack pixels have a value of 1, and non-crack pixels have a value of 0. The training data for the crack segmentation model comes from manually annotated road surface crack image samples. Annotators mark crack regions pixel-by-pixel on the road surface image, and the model learns from these annotated samples to establish a mapping relationship from the road surface image to the crack mask. In an alternative implementation, the crack segmentation model can also be implemented using traditional image processing methods, such as first performing Gaussian filtering on the road surface image to remove noise, then using adaptive thresholding to extract dark crack regions, and finally using morphological opening operations to remove isolated noise points.

[0025] Crack regions in a binary crack mask typically have a certain width. To accurately calculate crack length and topology, the crack binary mask needs to undergo skeleton refinement. Skeleton refinement employs an iterative erosion algorithm. In each iteration, this algorithm examines each pixel at the edge of the crack region, determining whether deleting the pixel would change the connectivity of the crack region. If connectivity is not changed, the pixel is deleted; otherwise, it is retained. After multiple iterations, the crack region is refined into a crack skeleton map with a single pixel width. This skeleton map preserves the original crack's topology and orientation information while eliminating the influence of crack width on subsequent statistical analysis.

[0026] Crack density is calculated based on the ratio of the total number of crack pixels in the crack skeleton image to the road surface area covered by the image. Specifically, the total crack length is obtained by multiplying the number of pixels with a value of 1 in the crack skeleton image by the actual length corresponding to a single pixel. The road surface area covered by the image is calculated based on the number of rows and columns of the image and the corresponding actual resolution. The crack density value equals the total crack length divided by the road surface area, and its physical meaning is the total crack length per unit area of ​​the road surface, expressed in meters per square meter. A higher crack density value indicates a more severe degree of road surface crack damage.

[0027] Calculating crack connectivity requires first identifying crack intersections in the crack skeleton map. A crack intersection is defined as a pixel in the crack skeleton map that connects three or more crack branches. The method for identifying crack intersections is as follows: for each pixel with a value of 1 in the crack skeleton map, count the number of pixels with a value of 1 in its 8-neighborhood. If this number is greater than or equal to 3, the pixel is marked as a crack intersection. The crack connectivity value equals the number of crack intersections divided by the total number of crack pixels; its physical meaning is the degree of branching in the crack network. A higher crack connectivity value indicates a more complex crack network, a higher degree of interconnection between cracks, and more severe overall damage to the pavement structure. When the crack connectivity value exceeds a certain threshold, it usually means that the pavement has developed severe damage patterns such as network cracks or alligator cracks.

[0028] The purpose of processing ground-penetrating radar (GPR) echo data is to obtain the three-dimensional distribution data of the dielectric constant of the road surface structure. First, time-depth conversion is performed on the GPR echo data, converting the two-way propagation time of the echo signal into the corresponding depth value. The basic principle of time-depth conversion is that the propagation speed of electromagnetic waves in a medium is related to the dielectric constant of the medium; the propagation speed... It can be represented as ,in The speed of light in a vacuum is taken as 300 million meters per second. Let be the relative permittivity of the medium. For asphalt concrete pavement, the typical range of relative permittivity is 4 to 8, corresponding to electromagnetic wave propagation speeds of approximately 100 million to 150 million meters per second. This is based on the two-way propagation time of the echo signal. and propagation speed The depth of the reflecting interface can then be calculated. The calculation method is as follows The division by 2 is because the electromagnetic wave undergoes a two-way propagation from the emission point to the reflection interface and back to the receiving antenna.

[0029] The generation of the three-dimensional dielectric constant distribution data volume employs a full-waveform inversion method. This method treats ground-penetrating radar (GPR) echo data as a function of the dielectric constant distribution of the road structure, and iteratively optimizes the data to find the dielectric constant distribution that best fits the measured echo data. Specifically, an initial dielectric constant model of the road structure is first established, assuming a uniform dielectric constant distribution across all layers. Then, a forward electromagnetic wave propagation simulation is performed based on this initial model to calculate theoretical echo data. Next, the residual between the theoretical and measured echo data is calculated, and the dielectric constant model is updated using backpropagation based on the residual. This process is repeated until the residual converges to below a preset threshold. The resulting three-dimensional dielectric constant distribution data volume is stored as a three-dimensional array, with the three dimensions corresponding to the lateral, longitudinal, and depth positions, respectively. Each element in the array represents the dielectric constant at its corresponding spatial location. In an alternative implementation, a simpler tomographic imaging method can be used to directly generate the three-dimensional dielectric constant distribution data volume by offsetting the echo data. This method is more computationally efficient but slightly less accurate.

[0030] The extraction of crack depth values ​​is based on the identification of abnormal regions in the three-dimensional dielectric constant distribution data volume. Intact asphalt concrete pavements have a relatively uniform dielectric constant distribution, while cracked voids, due to being filled with air or moisture, exhibit a significant difference in dielectric constant compared to the surrounding asphalt concrete. The relative dielectric constant of air is approximately 1, much lower than that of asphalt concrete; therefore, cracked voids appear as low-dielectric-constant regions in the three-dimensional dielectric constant distribution data volume. A baseline dielectric constant for asphalt is set, which in this embodiment is 4. Connected regions in the three-dimensional dielectric constant distribution data volume with dielectric constant values ​​lower than the baseline dielectric constant are marked as cracked voids. For each cracked void, its maximum extension value in the depth direction is calculated; this maximum extension value is the crack depth value. The crack depth value reflects the extent to which the crack extends into the pavement structure; a larger depth value indicates deeper damage to the pavement structure, and a correspondingly greater difficulty in healing and repair.

[0031] Determining the area to be healed first requires dividing the road surface into grid cells that are easy to manage and process. In this embodiment, the road surface is divided into square grid cells with equal side lengths, each 500 mm. The consideration for choosing 500 mm as the grid cell side length is that this size can cover the influence range of typical cracks and matches the effective heating area of ​​the phased array microwave heating device, facilitating subsequent cell-by-cell heating and healing operations. In an optional implementation, the side length of the grid cells can also be adjusted according to the actual road conditions and heating equipment parameters. For example, a smaller grid side length can be used in areas with dense cracks to improve processing accuracy, while a larger grid side length can be used in areas with sparse cracks to improve processing efficiency.

[0032] For each grid cell, the crack density, crack connectivity, and maximum crack depth are calculated. The crack density and crack connectivity values ​​within the grid cell are derived from the pixels falling within that grid cell in the crack skeleton map, using the same calculation method as the overall calculation described above. The maximum crack depth of the grid cell is the maximum crack depth value among all crack void regions falling within the horizontal range of that grid cell in the dielectric constant 3D distribution data volume.

[0033] The determination of units requiring healing employs a multi-index joint criterion. Density threshold, connectivity threshold, and depth threshold are set. In this embodiment, the density threshold is 0.5 meters per square meter, the connectivity threshold is 0.1, and the depth threshold is 10 millimeters. When the crack density value of any grid unit exceeds the density threshold, or the crack connectivity value exceeds the connectivity threshold, or the maximum crack depth value exceeds the depth threshold, the corresponding grid unit is marked as a unit requiring healing. The disjunctive criterion is used instead of the conjunctive criterion because the three damage indices reflect different aspects of crack damage. Exceeding any one index indicates damage requiring healing in that area; waiting for all indices to exceed simultaneously may result in some damaged areas being missed. The density threshold is based on research findings in road maintenance specifications regarding the relationship between crack density and pavement service performance; the connectivity threshold aims to identify networked crack damage patterns; and the depth threshold is based on engineering experience regarding the impact of asphalt concrete pavement thickness and crack propagation on structural performance. In an optional embodiment, the above thresholds can also be differentiated according to the maintenance standards for different road grades.

[0034] The generation of the sequence of units to be healed is based on the severity of the damage. All units to be healed are sorted from largest to smallest according to their maximum crack depth, forming the sequence. The reason for choosing the maximum crack depth as the sorting criterion is that deeper cracks pose a greater threat to the pavement structure and require more timely healing. If not treated promptly, the cracks may extend further downwards until they penetrate the entire surface layer, causing irreversible structural damage. In addition, the thermally induced healing of deep cracks requires longer heating time and more precise energy control; prioritizing the treatment of deep cracks helps to rationally allocate construction resources and time. In an optional implementation, a comprehensive scoring method can also be used for sorting. Different scoring coefficients are assigned to the crack density value, crack connectivity value, and maximum crack depth value, respectively, and a comprehensive damage score is calculated for each unit to be healed. The sequence of units to be healed is then generated by sorting them from highest to lowest comprehensive damage score.

[0035] The phased array microwave directional heating process based on the inverse-time focusing algorithm involves multiple steps, including heating device positioning, dielectric constant modeling, focusing target point determination, inverse-time focusing phase calculation, microwave transmission control, and temperature monitoring. The specific implementation methods of each step are described in detail below.

[0036] The phased array microwave heating device is installed on a movable working platform, which can move longitudinally and laterally along the road surface with a positioning accuracy of ±5 mm. The core component of the phased array microwave heating device is a microwave radiating element array. In this embodiment, the microwave radiating element array consists of 64 microwave radiating elements, evenly distributed in an 8x8 rectangular arrangement. The center-to-center distance between adjacent microwave radiating elements is 60 mm, and the entire array covers an area of ​​420 mm x 420 mm. Each microwave radiating element adopts a rectangular waveguide open antenna structure with an inner cross-sectional dimension of 86 mm x 43 mm, corresponding to the main mode transmission conditions at a working frequency of 2.45 GHz. The working height of the microwave radiating element array above the road surface is 100 mm. This distance ensures that microwaves can effectively radiate to the road surface while avoiding heat transfer between the device and the road surface during the heating process, which could affect the temperature control accuracy.

[0037] Each microwave radiating unit is equipped with an independent phase shifter and power amplifier. The phase shifter is a digitally controlled ferrite phase shifter with a phase adjustment range of 0 to 360 degrees, an adjustment step of 1 degree, and a phase switching time of less than 10 microseconds. The power amplifier is a solid-state power amplifier with a maximum output power of 200 watts per amplifier. The power adjustment range is 10% to 100% of the maximum output power, with an adjustment step of 1% of the maximum output power. The total maximum output power of the 64 microwave radiating units is 12.8 kilowatts. All phase shifters and power amplifiers are centrally controlled and managed by a central control unit. The central control unit issues phase setting commands to each phase shifter based on the counter-time focusing phase calculation results and power adjustment commands to each power amplifier based on temperature monitoring feedback.

[0038] Once the work platform moves directly above the currently healing cell in the sequence of cells to be healed, it first needs to extract local data of the corresponding region from the 3D dielectric constant distribution data volume to establish a 3D dielectric constant mesh model. The local dielectric constant distribution data covers a cuboid region with a length of 480 mm, a width of 480 mm, and a depth of 100 mm. This spatial range is slightly larger than the coverage area of ​​the microwave radiation cell array, with a 30 mm margin reserved at the edge to accommodate boundary effects during electromagnetic wave propagation. The depth of 100 mm is chosen because the typical thickness of asphalt concrete surface layers is 40 mm to 80 mm, and a depth of 100 mm is sufficient to cover all possible cracks within the surface layer, while avoiding an excessively large computational domain that would reduce simulation efficiency.

[0039] The process of establishing the three-dimensional grid model of the dielectric constant is as follows: the aforementioned cuboid space is uniformly discretized into cubic grids with a side length of 2 mm. The entire computational domain contains 240 x 240 x 50 cubic grids, totaling 2.88 million grid nodes. Each cubic grid stores the dielectric constant value at the corresponding location, which is obtained from the three-dimensional distribution data volume of the dielectric constant using trilinear interpolation. The basis for choosing 2 mm as the grid side length is: the free space wavelength corresponding to the operating frequency of 2.45 GHz is approximately 122 mm, and the wavelength in asphalt concrete medium is approximately 50 to 60 mm. According to the sampling theorem of the finite-difference time-domain method, the grid side length should be less than one-tenth of the wavelength in the medium to ensure computational accuracy. The 2 mm grid side length meets this requirement, and the computational load is within an acceptable range.

[0040] Determining the location of the focus point requires comprehensive consideration of both the horizontal and depth distribution of the cracks. The horizontal position of the focus point is set as the geometric center of the point with the highest crack density within the current unit to be healed. Specifically, within the crack skeleton map area corresponding to the current unit to be healed, a sliding scan is performed using a circular window with a radius of 50 mm. The number of crack pixels within each window position is counted, and the center of the window with the highest number of crack pixels is taken as the point of maximum crack density. The reason for choosing the point of maximum crack density as the horizontal focus position is that this location is usually the convergence point of the crack network or the core area of ​​the main crack; heating and healing at this location can restore the structural integrity of the pavement to the greatest extent. The depth position of the focus point is set at half the maximum crack depth value of the current unit to be healed. The reason for choosing the midpoint of the depth rather than the bottom of the crack as the focus position is that microwave energy forms a high-energy-density area around the focus point. Setting the focus position at the midpoint of the depth allows this high-energy-density area to cover the entire depth range from the surface to the bottom of the crack, achieving uniform heating of the entire crack depth. In an optional implementation, for cracks with greater depth, a multi-layer focusing strategy can also be adopted, sequentially setting the focus point at different depth positions for layered heating.

[0041] Phase calculation for reverse-time focusing is a core step in achieving directional focusing of microwave energy. Traditional phased array beamforming methods assume that electromagnetic waves propagate in a homogeneous medium, and the phase delay from each array element to the target point can be directly calculated based on the geometric distance. However, asphalt concrete pavement is a non-homogeneous medium composed of asphalt binder, aggregate, and voids, with significant differences in dielectric constant at different locations. The propagation path and velocity of electromagnetic waves within this medium are affected by its inhomogeneity. If the homogeneous medium assumption is used to calculate the phase, the electromagnetic waves emitted by each microwave radiating element will not be able to achieve in-phase superposition when they reach the target point, resulting in a significant reduction in focusing effect or even complete failure. The reverse-time focusing algorithm solves this problem by utilizing the time-reversal symmetry of electromagnetic wave propagation: if an electromagnetic wave emitted from a point source propagates through a non-homogeneous medium and is recorded by each receiving point, and the signals recorded by each receiving point are time-reversed before being re-emitted, the emitted electromagnetic wave will propagate in reverse along the original propagation path and automatically converge at the original point source position. Time reversal is equivalent to phase conjugation in the frequency domain. Therefore, by simply obtaining the received phase at the location of each microwave radiation unit and taking its conjugate value as the transmitted phase, precise focusing on the target point in a non-uniform medium can be achieved.

[0042] The specific implementation process of reverse-time focusing phase calculation is as follows: First, a virtual point source is set at the focusing target point. The virtual point source emits spherical electromagnetic waves into the surrounding space at a working frequency of 2.45 GHz. The mathematical form of the spherical electromagnetic wave is a radially propagating sine wave, with the initial phase set to 0 degrees. The propagation process of the spherical electromagnetic wave is simulated in a three-dimensional grid model of dielectric constant using the finite-difference time-domain method. The finite-difference time-domain method discretizes Maxwell's curl equation in time and space, and obtains the electric and magnetic field components at each grid node after each time step through iterative calculation. Time step The setting must satisfy the Courant-Friedrich-Levy stability condition, which requires... ,in The side length of the spatial grid is 2 millimeters. The speed of light in a vacuum is taken as 300 million meters per second. Substituting this value into the calculation, the time step should be less than 3.85 picoseconds. In this embodiment, the time step is taken as 3 picoseconds.

[0043] The total duration of the electromagnetic wave propagation simulation was set to 10 nanoseconds, which was sufficient for the spherical electromagnetic wave to propagate from the focusing target point to the top boundary of the dielectric constant 3D mesh model. During the simulation, virtual probe points were set at positions directly below the 64 microwave radiating elements on the top surface of the dielectric constant 3D mesh model. The electric field time-domain waveform at each virtual probe point was continuously recorded. After the simulation, 64 sets of electric field time-domain waveform data were obtained, each set containing approximately 3333 sampling points.

[0044] refer to Figure 2In this schematic diagram, the asphalt concrete pavement structure, from top to bottom along the depth direction, includes an asphalt surface layer and an asphalt intermediate layer. The thickness of the asphalt surface layer is 40 mm, and the thickness of the asphalt intermediate layer is 60 mm. The pavement structure contains multiple irregularly shaped mineral aggregate particles. The dielectric constant of these aggregate particles differs from that of the asphalt binder, resulting in a non-homogeneous dielectric characteristic of the entire pavement structure. A microwave radiating element array is arranged above the pavement structure. In the diagram, eight square symbols represent the microwave radiating elements arranged laterally, with adjacent microwave radiating elements maintaining equal spacing.

[0045] In the forward propagation phase, the location of the focusing target point is first determined based on the crack detection results. The focusing target point is located at the center of the crack depth within the pavement structure, marked with a pentagram symbol in the diagram. A virtual point source is set up at the focusing target point, emitting spherical electromagnetic waves into the surrounding space at a working frequency of 2.45 GHz. The wavefront of the spherical electromagnetic wave is represented by multiple concentric arcs in the diagram, with the spacing between the arcs corresponding to the wavelength of the electromagnetic wave. Because the pavement structure is a non-uniform medium, the spherical electromagnetic wave undergoes refraction and scattering when it encounters aggregate particles during propagation, causing the wavefront shape to distort and no longer maintain an ideal circular arc shape. The diagram illustrates the impact of this medium non-uniformity on electromagnetic wave propagation through the local bending and irregular deformation of the arcs.

[0046] Spherical electromagnetic waves propagate outward from the focal point, eventually reaching the road surface and being received by virtual detection points at the locations of each microwave radiating element. Because the geometric distances between each microwave radiating element and the focal point differ, and the medium distribution along the propagation path also varies, the timing and phase of the received spherical electromagnetic waves differ at each microwave radiating element location. The diagram uses dashed lines to represent the propagation paths from the focal point to each microwave radiating element location. These paths are not straight lines but exhibit a degree of curvature, reflecting the actual propagation trajectory of the electromagnetic waves in a non-uniform medium. The received phase value is marked below each microwave radiating element. to These phase values ​​are obtained by analyzing the time-domain waveforms of the electric field recorded at the virtual detection points. The specific value of the received phase depends on the combined effects of the propagation distance and the medium properties; different microwave radiating elements at different locations correspond to different received phase values.

[0047] The core objective of the forward propagation phase is to acquire the received phase, which includes information about the inhomogeneity of the medium. Traditional phased array beamforming methods assume that electromagnetic waves propagate in a straight line in a homogeneous medium, calculating the phase delay solely based on geometric distance. This simplification fails to accurately reflect the complex propagation process in actual road structures. By simulating the propagation of spherical electromagnetic waves emitted from a virtual point source within a three-dimensional grid model of dielectric constant, the effects of medium refraction, scattering, and multipath effects on the phase can be precisely captured, providing accurate input data for subsequent phase conjugation processing.

[0048] refer to Figure 3 The diagram is consistent with Figure 2 They have the same road surface structure and microwave radiation unit array arrangement, but the propagation direction of electromagnetic waves is different. Figure 2 Conversely, the microwave radiation unit array propagates into the interior of the road structure and eventually converges at the focal target point.

[0049] In the backpropagation phase, firstly... Figure 2 The received phases at the locations of each microwave radiating element are obtained and subjected to phase conjugation. The mathematical operation of phase conjugation is to invert the received phase, i.e., the transmitted phase. Equals 360 degrees minus the corresponding receiving phase The diagram shows the emitted phase value after phase conjugation below each microwave radiating element. to These transmitted phases are applied to the phase shifters of the corresponding microwave radiating units, so that the microwave signals emitted by each microwave radiating unit carry a specific initial phase.

[0050] When all microwave radiating units simultaneously emit continuous microwave signals at an operating frequency of 2.45 GHz, the microwave signals emitted by each unit enter the road surface structure and propagate towards depth. The figure shows multiple arcs representing the wavefronts of the microwave signals emitted by each unit; these wavefronts are also distorted during propagation due to the influence of the non-uniform medium. Solid arrows in the figure represent the propagation paths of the microwave signals from each radiating unit to the focusing target point. These propagation paths are related to… Figure 2 The forward propagation paths in the two vectors completely overlap but in opposite directions.

[0051] According to the principle of time-reversal symmetry in electromagnetic wave propagation, when the transmitted phase is the conjugate of the received phase, the microwave signals emitted by each microwave radiating element will propagate backward along the original path of forward propagation. During propagation, the phase effect of medium inhomogeneity on the microwave signals is exactly the same as during forward propagation. However, since the transmitted phase has already compensated for these effects, the microwave signals will achieve in-phase superposition when they reach the focusing target point. In-phase superposition means that the electric field vectors of the microwave signals are in the same direction, and the superimposed electric field strength is the sum of the electric field strengths of each signal, which is much greater than the electric field strength of a single signal.

[0052] The image shows multiple concentric ellipses at the focal point, with the color of the ellipses gradually lightening from the inside out, indicating a gradual decrease in energy density from the center outwards. These ellipses represent the projection of the three-dimensional focused thermal zone onto the cross-section. The major axis of the ellipses is along the depth direction, and the minor axis is along the horizontal direction, reflecting the greater extension range of the focused thermal zone in the depth direction. The formation of the three-dimensional focused thermal zone concentrates microwave energy at the depth of the crack, achieving directional heating of the asphalt binder inside the crack. Outside the focused thermal zone, the road surface area has a lower energy density and limited temperature rise because the microwave signals cannot be superimposed in phase, thus avoiding unnecessary heating and potential thermal damage to the intact road surface area.

[0053] The received phase is extracted from the electric field time-domain waveform corresponding to each microwave radiating unit. Taking the first microwave radiating unit as an example, its electric field time-domain waveform exhibits a sinusoidal oscillation pattern that first increases and then tends to stabilize. The leading part of the waveform corresponds to the transition process when the spherical electromagnetic wave just arrives at that position, and the middle and later parts correspond to the steady-state oscillation. The complete period with the largest amplitude in the electric field time-domain waveform is selected as the reference period. This selection method ensures that the extracted phase information comes from the steady-state propagation stage rather than the transition process. Within the reference period, the zero-crossing moment when the electric field value changes from negative to positive is identified, and the phase angle of this zero-crossing moment relative to the start of the reference period is calculated. This phase angle is the received phase of the first microwave radiating unit. The phase angle is calculated as follows: Let the zero-crossing time be... The reference period starts at the time when The reference period duration is Then receive phase Degree. The electric field time-domain waveforms corresponding to the 2nd to 64th microwave radiating units are processed sequentially in the same manner to obtain the 2nd receiving phase. Up to the 64th receiving phase .

[0054] The transmitted phase is obtained by performing phase conjugation on the received phase. The mathematical operation of phase conjugation is to invert the received phase; that is, the transmitted phase equals 0 degrees minus the received phase, or equivalently, 360 degrees minus the received phase. Taking the first microwave radiating unit as an example, its transmitted phase... Degrees. If the calculated result exceeds 360 degrees, subtract 360 degrees; if the calculated result is negative, add 360 degrees to ensure that the value of the transmission phase is between 0 and 360 degrees. This value range matches the phase adjustment range of the phase shifter. Calculate the second transmission phase in the same way. Up to the 64th launch phase .

[0055] The calculated 64 transmission phases are applied to the phase shifters of the corresponding microwave radiating units. The central control unit sends phase setting commands sequentially to the 1st through 64th phase shifters, with the command content being the corresponding transmission phase value. Upon receiving the command, the phase shifter adjusts the magnetization state of its internal ferrite elements, causing a corresponding phase shift in the microwave signal passing through the phase shifter. After phase loading is complete, the output power of the power amplifiers in all microwave radiating units is set to the same value. In this embodiment, the initial output power is set to 80% of the maximum output power, i.e., each power amplifier outputs 160 watts.

[0056] The central control unit sends a synchronization trigger signal, controlling 64 microwave radiating units to simultaneously transmit continuous microwave signals at an operating frequency of 2.45 GHz. Since the phase carried by the microwave signal emitted by each radiating unit is the phase conjugate of the corresponding received phase, according to the time-reversal symmetry of electromagnetic wave propagation, the microwave signals emitted by each radiating unit will propagate backward along the propagation path of the spherical electromagnetic waves in the simulation stage. During propagation, the microwave signals experience the same medium refraction, scattering, and multipath effects as in the simulation stage. These effects cancel each other out during forward and backward propagation, ultimately resulting in all microwave signals arriving in phase at the focusing target point and undergoing coherent superposition. This coherent superposition makes the electric field intensity at the focusing target point much higher than the surrounding area, forming a three-dimensional focusing thermal zone with a significantly higher energy density than the surrounding area.

[0057] The spatial distribution of the three-dimensional focusing thermal zone is ellipsoidal, with its major axis along the depth direction and its minor axis along the horizontal direction. Under the parameter configuration of this embodiment, the half-power beamwidth of the three-dimensional focusing thermal zone is approximately 40 mm in the horizontal direction and approximately 60 mm in the depth direction. The half-power beamwidth is defined as the spatial range corresponding to when the energy density drops to half of its peak value. The three-dimensional focusing thermal zone covers most of the crack in the depth direction, achieving directional heating of the asphalt binder inside the crack. After absorbing microwave energy, the asphalt binder within the focusing area experiences a temperature increase. When the temperature reaches above the asphalt softening point, the viscosity of the asphalt binder decreases, and the molecular chains gain fluidity, enabling them to diffuse and entangle across the crack interface, thereby achieving crack healing and closure. Outside the three-dimensional focusing thermal zone, the road surface area experiences a limited temperature increase due to the lower microwave energy density, avoiding unnecessary energy waste and thermal damage to the intact road surface.

[0058] Temperature monitoring during the microwave heating process is performed by an infrared thermal imager installed at the center of the phased array microwave heating device. The infrared thermal imager's detector has a resolution of 640 pixels by 480 pixels, operates in the 8-14 micrometer wavelength range, measures temperatures from -20 degrees Celsius to 200 degrees Celsius, and has a temperature resolution of 0.1 degrees Celsius. The infrared thermal imager's field of view is set to cover the current unit to be healed and its surrounding area, with an imaging range of approximately 600 mm by 450 mm. The infrared thermal imager continuously acquires surface temperature distribution images of the current unit to be healed at a rate of 10 frames per second, or one frame of surface temperature distribution image every 100 milliseconds.

[0059] The central control unit receives surface temperature distribution images from the infrared thermal imager and extracts two characteristic temperature values ​​from each frame: the maximum temperature value and the average temperature value. The maximum temperature value is the highest value among all pixels in the image, reflecting the temperature of the hottest spot on the road surface. The average temperature value is the arithmetic mean of the temperature values ​​of all pixels within the image area corresponding to the current unit to be healed, reflecting the overall temperature level of the area to be healed. Extracting the maximum temperature value aims to prevent localized overheating, as asphalt concrete will age and deteriorate at excessively high temperatures; if the maximum temperature value exceeds a safe threshold, the heating power needs to be reduced. Extracting the average temperature value aims to assess the overall healing effect; the asphalt binder needs to be maintained within a certain temperature range for a sufficient time to complete the molecular diffusion and entanglement process.

[0060] refer to Figure 1 The horizontal axis represents the heating time. The unit is seconds, and the vertical axis represents temperature. The unit is degrees Celsius. The graph contains two temperature curves, one for the highest temperature and the other for the lowest temperature. and average temperature curve It also marks three temperature threshold reference lines and the division of three working stages.

[0061] The entire heating process is divided into three continuous working stages: the heating stage, the holding stage, and the cooling stage. The heating stage starts from time... Start to time The process ends in seconds; during this phase, the phased array microwave heating device continuously radiates microwave energy to the road surface at its initial set power. The road surface temperature rises from the ambient temperature. As the temperature begins to rise, both the maximum temperature curve and the average temperature curve exhibit an exponential growth trend. The pattern of change in the maximum temperature can be approximated as follows: ,in Celsius is the upper limit threshold for the maximum temperature. This represents the heating time constant. The average temperature changes similarly to the maximum temperature, but because heat diffusion within the road surface area takes time, the average temperature rises at a slightly lower rate than the maximum temperature. When the maximum temperature first reaches... When the temperature reaches a certain degree Celsius, the system determines that the heating phase has ended and automatically switches to the heat preservation phase.

[0062] The heat preservation stage starts from time to time Seconds start to time Ends in seconds, duration is Seconds. This stage is the critical period for asphalt binder to complete molecular diffusion and interfacial healing. The ideal healing temperature range is marked by a light blue area in the figure, with the lower limit of this range being the lower threshold of the average temperature. Celsius, with the upper limit being the upper threshold of the average temperature. Temperature in degrees Celsius. During the heat preservation phase, the control system adjusts the microwave radiation power to maintain the average temperature. Maintain at to Within the ideal temperature range of 100 degrees Celsius. As can be observed from the graph, the highest temperature curve during the heat preservation stage is... to The temperature fluctuates slightly within a range of degrees Celsius, consistently remaining below the critical temperature that could lead to asphalt aging; the average temperature curve is... to The temperature fluctuates periodically within a certain range of degrees Celsius, with the fluctuation period corresponding to the power adjustment period. This temperature fluctuation is a normal phenomenon in power feedback control, and the fluctuation amplitude is controlled within an acceptable range and does not affect the healing effect.

[0063] Cooling phase from time Starting at a certain second, all microwave radiation units are switched off, and the road surface gradually cools down through natural convection and heat conduction. Both the maximum and average temperatures exhibit an exponential decay trend, which can be expressed as follows: ,in This is the temperature value at the end of the heat preservation stage. The second marks the end of the heat preservation phase. This is the cooling time constant. As time progresses, the road surface temperature gradually approaches the ambient temperature, and the asphalt binder re-solidifies during the cooling process, completing the crack healing.

[0064] The image also shows three horizontal dashed lines, each corresponding to a maximum temperature threshold. Celsius, upper limit of average temperature threshold Celsius and average lower temperature threshold Celsius. These three thresholds are the core parameters of the thermal insulation control algorithm. Their values ​​are set by comprehensively considering the softening point temperature of asphalt concrete, the optimal healing temperature range, and the safety margin to prevent thermal aging.

[0065] The heat preservation control process is divided into two stages: a heating stage and a heat preservation stage. During the heating stage, all microwave radiation units continuously emit microwave signals at their initial output power, gradually increasing the road surface temperature. The central control unit monitors the maximum temperature value in real time. When the maximum temperature value first reaches 85 degrees Celsius, the heating stage is considered complete, and the heat preservation stage begins. The 85-degree Celsius setting is based on the fact that the softening point of asphalt concrete is typically between 45 and 55 degrees Celsius, and the healing effect is ideal within the range of 60 to 80 degrees Celsius. 85 degrees Celsius, as the upper limit threshold for the maximum temperature, ensures that the healing area reaches the effective healing temperature while also providing a safety margin to prevent further temperature increases that could lead to asphalt aging.

[0066] After entering the heat preservation stage, the central control unit reduces the output power of all microwave radiation unit power amplifiers to 50% of the initial setting, i.e., each power amplifier outputs 80 watts. The purpose of reducing the power is to slow the rate of temperature rise and stabilize the road surface temperature within the target range. During the heat preservation stage, the central control unit checks the average temperature value every 6 seconds and adjusts the power accordingly: if the average temperature is below 65 degrees Celsius, it indicates that the heating power is insufficient to maintain the healing temperature, and the output power of all power amplifiers is increased by 10%; if the average temperature is above 75 degrees Celsius, it indicates that excessive heating power may lead to overheating, and the output power of all power amplifiers is reduced by 10%; if the average temperature is between 65 and 75 degrees Celsius, it indicates that the temperature is within the ideal healing range, and the current power is maintained. 65 degrees Celsius and 75 degrees Celsius are designated as the lower and upper thresholds of the average temperature, respectively, and this temperature range corresponds to the optimal healing temperature range of the asphalt binder.

[0067] The duration of the insulation phase is set at 300 seconds. This duration is based on the fact that asphalt binder requires a certain amount of time to complete molecular diffusion and interface healing at the healing temperature. Experimental studies have shown that a 300-second insulation time can effectively heal most cracks. Further extending the insulation time has limited effect on improving the healing effect but increases energy consumption and construction time. In an optional implementation, the insulation time can also be adaptively adjusted according to the crack depth, with a longer insulation time for deep cracks and a shorter insulation time for shallow cracks.

[0068] After the 300-second heat preservation phase ends, the central control unit sends a shutdown command to all power amplifiers to stop microwave signal transmission, completing the thermally induced healing process for the current unit to be healed. After the microwave transmission is shut down, the road surface temperature gradually decreases under the action of natural convection and heat conduction. During the cooling process, the asphalt binder re-solidifies, and a continuous asphalt film layer is formed at the healed crack interface, restoring the integrity and mechanical properties of the road surface structure.

[0069] refer to Figure 4 The horizontal axis represents the heating time. The unit is seconds, and the vertical axis represents the output power of the power amplifier of a single microwave radiating unit. The unit is watts. The light green filled area below the curve in the graph visually represents the time integral of power output, i.e., total energy consumption.

[0070] During the warming phase, that is, from time... At that time Within a 1-second interval, the power amplifiers of all microwave radiating units were set to their initial power. The power output is continuous. This initial power value is the unit's maximum output power. tile The calculation relationship is as follows .choose Instead The consideration of maximum power is to reserve a certain upward adjustment margin for power regulation, while avoiding the device from operating at its limit for extended periods. The initial power level is indicated by a horizontal dashed line in the diagram. The position of the tile indicates that the power curve during the heating phase completely coincides with the dashed line, indicating that the power output is stable and without fluctuations.

[0071] When the system detects that the highest temperature has reached At a temperature of [temperature value missing] degrees Celsius, the control system reduces the output power of all power amplifiers to the initial set value. That is, the insulation reference power Watts, the calculation relationship is as follows The figure shows the time... At the second, the power curve can be observed to change from... Wa step jump down to Near the tile, this step drop signifies the transition from the heating phase to the heat preservation phase. The purpose of the significant power reduction is to slow the rate of temperature rise and prevent the road surface temperature from continuing to rise beyond a safe range.

[0072] During the heat preservation stage, that is, from time to time... Seconds to the moment Within a time interval of seconds, the control system periodically adjusts the power output based on the average temperature value fed back from the infrared thermal imager. The adjustment strategy follows these rules: every [number] seconds... Measure the average temperature value once per second ;like The temperature in degrees Celsius indicates that the heating power is insufficient to maintain the healing temperature. At this point, the output power of all power amplifiers should be increased. ;like The temperature in degrees Celsius indicates that excessive heating power may lead to overheating. In this case, the output power of all power amplifiers should be reduced. ;like The temperature in degrees Celsius indicates that the temperature is within the ideal healing range, maintaining the current power constant. As can be observed from the graph, the power curve during the heat preservation phase... The area around the tile exhibits periodic fluctuations, with a range of approximately... tile to Watts, corresponding to the reference power to The fluctuation of the power curve and Figure 1 The fluctuations in the average temperature curve show a negative correlation, that is, when the temperature is too high, the power decreases and when the temperature is too low, the power increases. This is a typical characteristic of closed-loop feedback control.

[0073] During the heat preservation stage After the duration of the specified number of seconds ends, the central control unit sends a shutdown command to all power amplifiers. (See diagram at time...) At approximately 10 seconds, the power curve can be observed to change from approximately 100%. Wa step jump down to The power output remains zero until the heating process is completely finished. After microwave radiation stops, the road surface enters a natural cooling phase, during which the asphalt binder re-solidifies, and a continuous asphalt film layer forms at the healed crack interface. The baseline insulation power is indicated by a horizontal dotted line in the diagram. The position of the tile facilitates observation of the center value and amplitude range of power fluctuations during the heat preservation stage.

[0074] The process of verifying healing effect and controlling safety involves multiple steps, including post-healing retesting, index comparison calculation, healing status determination, sequence traversal control, and safety control report generation. The specific implementation methods of each step are explained in detail below.

[0075] After the heat-induced healing treatment is completed, the pavement temperature gradually decreases under natural cooling. The asphalt binder re-solidifies during the cooling process, and the healed crack interface requires sufficient curing time to restore stable mechanical properties. In this embodiment, the cooling waiting time for the current healing unit is set to 30 minutes, ensuring that the pavement surface temperature drops to within ±5 degrees Celsius of the ambient temperature. The cooling waiting time is set considering that: if a retest is performed too early, before the asphalt binder is fully cured, the retest results may overestimate the healing effect; if the waiting time is too long, it will prolong the overall construction cycle and reduce work efficiency. In an optional embodiment, the cooling waiting time can also be achieved by monitoring the pavement temperature in real time using an infrared thermal imager. When the surface temperature drops below a preset retest temperature threshold, the retest procedure is automatically triggered. This method can adaptively adjust the waiting time according to the actual cooling rate.

[0076] The retesting process uses the same vehicle-mounted detection device and data processing method as in step 1. A linear scan camera acquires images of the current unit to be healed, obtaining the healed road surface image. The healed road surface image is input into the crack segmentation model, outputting a binary mask image of the healed crack. The binary mask image of the healed crack is then refined into a skeleton image. The total number of crack pixels on the healed crack skeleton image is counted and divided by the road surface area covered by the image to obtain the healed crack density value. All crack intersections are identified on the healed crack skeleton image, and the number of intersections is counted and divided by the total number of crack pixels to obtain the healed crack connectivity value. Using the same crack segmentation model and processing flow for both pre- and post-healing detection eliminates systematic errors caused by differences in detection methods, ensuring the accuracy and comparability of the healing effect evaluation.

[0077] Ground-penetrating radar (GPR) acquires echo data of the current unit to be healed, obtaining GPR echo data after healing. The healed GPR echo data undergoes time-depth conversion to generate a three-dimensional dielectric constant distribution data volume. Within this volume, regions with dielectric constant values ​​lower than the reference dielectric constant of asphalt are marked as healed crack void regions. The maximum depth value of each healed crack void region is extracted, and the maximum value is taken as the maximum healed crack depth. Thermally induced healing causes the asphalt binder at the crack interface to flow and re-bond, reducing the volume of the healed crack voids. This is reflected in the three-dimensional dielectric constant distribution data volume as a reduction in the range and depth of low dielectric constant regions.

[0078] The calculation of the decline rate of each indicator is based on the comparison of values ​​before and after healing. The calculation method for the crack density decline rate is as follows: subtract the crack density value after healing from the crack density value obtained in step 1 to obtain the crack density reduction; divide the crack density reduction by the crack density value obtained in step 1 to obtain the crack density decline rate. Described in mathematical terms ,in The crack density value obtained in step 1. This represents the density value of the crack after healing. This refers to the crack density reduction rate. The crack density reduction rate reflects the degree of repair effect of the healing treatment on visible cracks on the pavement surface; a higher reduction rate indicates a better crack closure effect.

[0079] The crack connectivity reduction rate is calculated as follows: subtract the healed crack connectivity value from the crack connectivity value obtained in step 1 to obtain the crack connectivity reduction amount; divide the crack connectivity reduction amount by the crack connectivity value obtained in step 1 to obtain the crack connectivity reduction rate. Described in mathematical terms ,in The crack connectivity value obtained in step 1. This represents the connectivity value of the healed crack. The crack connectivity reduction rate reflects the degree of damage to the crack network structure caused by the healing treatment. A higher reduction rate indicates that the originally interconnected crack network has been effectively isolated, and the integrity of the pavement structure has been restored.

[0080] The crack depth reduction rate is calculated as follows: subtract the maximum crack depth after healing from the maximum crack depth value obtained in step 1 to obtain the crack depth reduction; divide the crack depth reduction by the maximum crack depth value obtained in step 1 to obtain the crack depth reduction rate. Described in mathematical terms ,in The maximum crack depth value obtained in step 1. This represents the maximum crack depth after healing. This represents the crack depth reduction rate. The crack depth reduction rate reflects the degree of repair effect of the healing treatment on the longitudinal propagation of the crack. The higher the reduction rate, the better the healing effect at the bottom of the crack and the more sufficient the rebonding of the deep asphalt binder.

[0081] The three reduction rates evaluate the healing effect from different dimensions: the crack density reduction rate evaluates the area repair of surface cracks, the crack connectivity reduction rate evaluates the topological repair of the crack network, and the crack depth reduction rate evaluates the volumetric repair of the crack depth. These three rates are independent yet complementary, together constituting a comprehensive assessment of the healing effect.

[0082] The healing status is determined using a multi-indicator joint criterion. A reduction rate threshold is set, which is uniformly set to 60% in this embodiment. The determination logic is as follows: when all three indicators—crack density reduction rate, crack connectivity reduction rate, and crack depth reduction rate—are greater than the reduction rate threshold, the current unit to be healed is marked as having achieved the healing target and removed from the unit to be healed sequence; when any one of the three reduction rates is less than or equal to the reduction rate threshold, the current unit to be healed is marked as having failed to achieve the healing target and is retained in the unit to be healed sequence for further processing.

[0083] The 60% threshold for the reduction rate is chosen based on the engineering evaluation standards for the healing effect of asphalt concrete. When the crack index decreases by more than 60%, the pavement's water resistance and structural bearing capacity can be restored to more than 70% of their pre-damage levels, meeting the safety requirements for normal traffic. The setting of the reduction rate threshold needs to strike a balance between the required healing effect and construction efficiency: a threshold that is too high will lead to repeated treatment of a large number of units to be healed, prolonging the construction period; a threshold that is too low may result in poorly healed sections being mistakenly judged as meeting the standards, leaving safety hazards. In one optional implementation, the reduction rate threshold can also be differentiated according to the road grade. Higher reduction rate thresholds are used for high-grade roads such as highways to ensure higher healing quality, while lower reduction rate thresholds are used for ordinary roads to balance construction efficiency.

[0084] The reason for using the conjunction criterion instead of the disjunctive criterion to determine the healing status is that the three indicators reflect different aspects of the healing effect, and meeting only some indicators does not guarantee the restoration of the overall pavement performance. For example, if the surface crack density decreases significantly but the depth of deep cracks does not decrease significantly, it may indicate that the heat mainly acts on the surface layer and does not effectively penetrate to the deeper layers. In this case, although the pavement surface may appear intact, internal damage still exists, and surface cracks may reappear after continued service. Therefore, all three indicators must be met simultaneously to determine the healing status as satisfactory.

[0085] For units that have not yet reached the healing standard, retaining them in the unit sequence indicates that they will undergo reheating and healing. During reheating, the heating parameters can be adjusted based on the decrease in the indicators from the initial healing: if the crack depth decrease rate is significantly lower than the other two indicators, it indicates insufficient deep heating, and the depth position of the focus target point can be appropriately shifted downwards during the reverse-time focusing phase calculation; if the crack density decrease rate is significantly lower than the other two indicators, it indicates insufficient surface heating range, and a scanning focusing method can be used to expand the heating coverage area; if the crack connectivity decrease rate is significantly lower than the other two indicators, it indicates poor healing at crack intersections, and the focus target point can be adjusted to the crack intersection position. In an optional implementation, units that have not yet reached the healing standard can also be classified according to the lowest decrease rate, and different secondary heating strategies can be adopted for different categories.

[0086] The traversal of the sequence of units to be healed is performed sequentially. After verifying the healing effect and determining the status of the current unit, the next unit to be healed is taken from the sequence and repeated step 2 (heat healing treatment) and step 3 (effect verification and determination). The traversal process continues until all units in the sequence have been processed. For units marked as not healing up to standard in the first traversal, a second traversal is performed after the first traversal. The second traversal only applies to units that are not healing up to standard. If units are still not healing up to standard after the second traversal, a third traversal can be performed, with a maximum of 3 traversals. Units that are still not healing up to standard after more than 3 traversals are marked as intractable damage and require further maintenance.

[0087] After the traversal is complete, the road segment healing compliance rate is used as a quantitative evaluation index of the overall healing effect. The road segment healing compliance rate is calculated as follows: count the number of all unhealed units marked as having met the healing criteria, divide this number by the total number of all processed unhealed units to obtain the road segment healing compliance rate. Described in mathematical terms ,in The number of units to be healed that are marked as having reached the healing target status. For the total number of units to be healed that have been processed, This refers to the road section healing compliance rate. The road section healing compliance rate ranges from 0 to 1 and is usually expressed as a percentage.

[0088] The safety status of a road segment is determined based on its healing compliance rate. In this embodiment, two threshold levels are set: when the healing compliance rate is greater than 90%, the target road segment is marked as safe for passage, indicating that the cracks and damage have been effectively repaired, the pavement structure performance has been well restored, and traffic can resume normally; when the healing compliance rate is between 70% and 90%, the target road segment is marked as speed-limited, indicating that some incompletely healed areas still exist, requiring reduced vehicle speeds to minimize dynamic load impact, and subsequent re-inspections should be arranged; when the healing compliance rate is less than 70%, the target road segment is marked as requiring close monitoring, indicating that the overall healing effect is poor, potentially due to severe material aging, excessive crack penetration depth, etc., requiring further analysis of the causes and the development of a specific maintenance plan.

[0089] The grading thresholds for safety status markings are set based on risk assessments of road safety operations. A 90% compliance rate means that the vast majority of damaged areas have returned to normal, and the few non-compliant areas are scattered and will not significantly affect overall traffic safety. A 70% compliance rate means that approximately one-third of the damaged areas have not been effectively repaired; these areas may continue to deteriorate under vehicle loads and require control measures. In an alternative implementation, the grading thresholds can also be adjusted according to road grade and traffic volume, with stricter thresholds used for main roads with high traffic volume and relatively more lenient thresholds used for secondary roads with low traffic volume.

[0090] The safety control report summarizes key information from the entire healing control process. The report includes the following: basic information about the target road segment, including segment number, start and end chainage, lane number, and inspection date; damage status statistics before healing, including the total number of units to be healed, the distribution range of crack density values, the distribution range of crack connectivity values, and the distribution range of maximum crack depth values; execution records of the healing treatment, including heating time, insulation time, and power consumption for each unit to be healed; post-healing effect evaluation, including the crack density reduction rate, crack connectivity reduction rate, and crack depth reduction rate for each unit to be healed; determination of the healing status, including a list of units that have met the healing standards and a list of units that have not met the standards; the road segment healing compliance rate and safety status markers; and a list of coordinates of units that have not met the healing standards, used to guide subsequent review and supplementary maintenance.

[0091] The safety control report is output in two forms: structured data files and visual charts. The structured data files are stored in Extensible Markup Language (XML) format, facilitating integration and data exchange with road management information systems. The visual charts include a road segment plan view, a damage distribution heat map, and a healing effect comparison chart. The road segment plan view marks the location and healing status of each unit to be healed. The damage distribution heat map uses color gradations to represent the spatial distribution of crack density values ​​before and after healing. The healing effect comparison chart displays the three reduction rate values ​​of each unit to be healed in the form of a bar chart.

[0092] While specific embodiments of the present invention have been described above, those skilled in the art should understand that these specific embodiments are merely illustrative. Those skilled in the art can omit, substitute, and modify the details of the above methods and systems in various ways without departing from the principles and essence of the present invention. For example, combining the above method steps to perform substantially the same function and achieve substantially the same result according to substantially the same method falls within the scope of the present invention. Therefore, the scope of the present invention is defined only by the appended claims.

Claims

1. A method for controlling the safety of thermally induced healing in asphalt concrete based on fatigue damage assessment, characterized in that, Includes the following steps: Step 1: Fatigue damage state identification and determination of areas to be healed: The asphalt concrete pavement of the target road section is scanned using an on-board detection device to obtain pavement images and ground-penetrating radar echo data; crack density and crack connectivity values ​​are extracted based on the pavement images, and a three-dimensional distribution data volume of dielectric constant is generated based on the ground-penetrating radar echo data, and crack depth values ​​are extracted; the pavement is divided into grid cells, and the units to be healed are determined based on the crack density, crack connectivity, and crack depth values ​​of each grid cell, generating a sequence of units to be healed; Step 2: Phased array microwave directional heating based on inverse-time focusing algorithm: Move the phased array microwave heating device above the current unit to be healed, establish a three-dimensional grid model of dielectric constant based on the three-dimensional distribution data of dielectric constant, and determine the focusing target point; perform inverse-time focusing phase calculation, set a virtual point source at the focusing target point and simulate the electromagnetic wave propagation process, obtain the receiving phase at the corresponding position of each microwave radiation unit, and perform phase conjugation processing on the receiving phase to obtain the transmitting phase; load the transmitting phase onto each microwave radiation unit, control each microwave radiation unit to simultaneously emit microwave signals, so that the microwave energy is focused on the focusing target point to form a three-dimensional focusing hot zone; monitor the surface temperature and perform heat preservation control; Step 3, Healing Effect Verification and Safety Control Output: Retest the current unit to be healed after healing, obtain the crack density value, crack connectivity value, and maximum crack depth value after healing, calculate the decrease rate of each indicator and determine the healing status; traverse all units to be healed in the sequence of units to be healed, count the road section healing compliance rate and output a safety control report.

2. The method according to claim 1, characterized in that, In step 1, the process of extracting crack density and crack connectivity values ​​based on the road surface image includes: inputting the road surface image into the crack segmentation model and outputting a crack binary mask; performing skeleton refinement on the crack binary mask to obtain a crack skeleton image with a single pixel width; counting the total number of crack pixels on the crack skeleton image and dividing it by the road surface area covered by the image to obtain the crack density value; identifying all crack intersections on the crack skeleton image, counting the number of crack intersections and dividing it by the total number of crack pixels to obtain the crack connectivity value.

3. The method according to claim 2, characterized in that, In step 1, the process of generating a three-dimensional distribution data volume of dielectric constant based on ground-penetrating radar echo data and extracting crack depth values ​​includes: performing time-depth conversion on the ground-penetrating radar echo data to generate a three-dimensional distribution data volume of dielectric constant of the pavement structure; in the three-dimensional distribution data volume of dielectric constant, areas with dielectric constant values ​​lower than the reference dielectric constant of asphalt are marked as crack void areas, and the maximum depth value of each crack void area is extracted as the crack depth value.

4. The method according to claim 3, characterized in that, Step 1, the process of determining the units to be healed and generating a sequence of units to be healed, includes: dividing the road surface into square grid units with equal side lengths; for each grid unit, calculating the crack density value, crack connectivity value, and maximum crack depth value within the grid unit; when the crack density value of any grid unit exceeds the density threshold, or the crack connectivity value exceeds the connectivity threshold, or the maximum crack depth value exceeds the depth threshold, the corresponding grid unit is marked as a unit to be healed; and sorting all units to be healed in descending order of maximum crack depth value to generate a sequence of units to be healed.

5. The method according to claim 4, characterized in that, In step 2, the phased array microwave heating device includes multiple microwave radiation units arranged in a rectangular array. Each microwave radiation unit is equipped with an independent phase shifter and a power amplifier. The phase adjustment range of the phase shifter is from 0 degrees to 360 degrees.

6. The method according to claim 5, characterized in that, In step 2, the process of establishing a three-dimensional dielectric constant mesh model and determining the focus target point based on the three-dimensional dielectric constant distribution data volume includes: extracting local dielectric constant distribution data of the corresponding region of the current unit to be healed from the three-dimensional dielectric constant distribution data volume; discretizing the spatial range covered by the local dielectric constant distribution data into a cubic mesh, with each cubic mesh storing the dielectric constant value at the corresponding location, forming a three-dimensional dielectric constant mesh model; the horizontal position of the focus target point is the geometric center of the location with the maximum crack density in the current unit to be healed, and the depth position of the focus target point is half of the maximum crack depth value of the current unit to be healed.

7. The method according to claim 6, characterized in that, Step 2, the process of performing the reverse-time focusing phase calculation includes: setting a virtual point source at the focusing target point, and the virtual point source emitting spherical electromagnetic waves into the surrounding space; simulating the propagation process of the spherical electromagnetic waves in the three-dimensional grid model of the dielectric constant using the finite-difference time-domain method; recording the electric field time-domain waveform at the corresponding position directly below each microwave radiating element when the spherical electromagnetic waves propagate to the top surface of the three-dimensional grid model of the dielectric constant; for the electric field time-domain waveform recorded at the position directly below each microwave radiating element, selecting the complete period with the largest amplitude in the electric field time-domain waveform as the reference period, and recording the phase angle of the moment when the electric field changes from negative to positive at the zero-crossing point within the reference period relative to the start time of the reference period as the receiving phase of the corresponding microwave radiating element.

8. The method according to claim 7, characterized in that, The process of obtaining the transmission phase by performing phase conjugation processing on the received phase includes: for the received phase of each microwave radiation unit, the difference between 360 degrees and the received phase is calculated as the transmission phase of the corresponding microwave radiation unit. If the difference exceeds 360 degrees, 360 degrees is subtracted; if the difference is negative, 360 degrees is added to ensure that the transmission phase value is within the range of 0 degrees to 360 degrees. The transmission phase of each microwave radiation unit is loaded onto the corresponding phase shifter, and all microwave radiation units are controlled to transmit microwave signals simultaneously, so that the microwave signals transmitted by each microwave radiation unit are superimposed in phase at the focusing target point, forming a three-dimensional focusing hot zone.

9. The method according to claim 8, characterized in that, Step 2, the process of monitoring surface temperature and performing heat preservation control includes: continuously acquiring surface temperature distribution images of the current unit to be healed using an infrared thermal imager, and extracting the highest temperature value and average temperature value from each frame of surface temperature distribution image; when the highest temperature value reaches the preset upper limit threshold of the highest temperature, reducing the output power of the power amplifiers of all microwave radiation units and entering the heat preservation stage; during the heat preservation stage, periodically detecting the average temperature value, increasing the output power of the power amplifiers when the average temperature value is lower than the preset lower limit threshold of the average temperature, and reducing the output power of the power amplifiers when the average temperature value is higher than the preset upper limit threshold of the average temperature; and turning off all microwave radiation units after the heat preservation stage ends.

10. The method according to claim 9, characterized in that, Step 3, the process of calculating the decline rate of each indicator and determining the healing status, includes: calculating the difference between the crack density value and the crack density value after healing, and dividing the difference by the crack density value to obtain the crack density decline rate; calculating the difference between the crack connectivity value and the crack connectivity value after healing, and dividing the difference by the crack connectivity value to obtain the crack connectivity decline rate; calculating the difference between the maximum crack depth value and the maximum crack depth value after healing, and dividing the difference by the maximum crack depth value to obtain the crack depth decline rate; when the crack density decline rate, crack connectivity decline rate, and crack depth decline rate are all greater than the preset decline rate threshold, the current unit to be healed is marked as having reached the healing standard; the ratio of the number of all units to be healed marked as having reached the healing standard to the number of all units to be healed that have been processed is used as the road segment healing standard compliance rate; the road segment safety status label is determined based on the road segment healing standard compliance rate, and a safety control report is output.

Images