Automatic repairing process for road surface crack punching and pressure grouting

Abstract

The application discloses a kind of road surface crack punching pressurized grouting automatic repair processes, comprising: crack automatic detection step, by mobile platform is carried image acquisition device, utilize AI to identify crack position, length, width and depth;Precise positioning and punching step, based on detection result control arm automatic drilling, hole depth is according to crack depth self-adapting adjustment;Pressurized grouting step, real-time monitoring pressure flow using closed-loop pressure feedback system, according to crack depth self-adapting adjustment grouting pressure, filling rate reaches 95% when automatic stop;Solidification and surface treatment step, heating solidification, leveling and spraying sealing layer;Quality detection and data recording step, nondestructive testing repair quality, unqualified automatic repair, data upload cloud platform generates repair report.The application realizes full-process automatic repair, improves precision and efficiency, reduces artificial cost and security risk, and is suitable for high-speed and municipal road.

Description

Technical Field

[0001] This invention belongs to the field of road maintenance and repair technology, specifically relating to an automatic repair process for road cracks by drilling and pressurizing grouting. Background Technology

[0002] Currently, road crack repair technology mainly relies on manual or semi-mechanized operations. The conventional repair process includes: visually inspecting and identifying cracks, marking their locations, drilling holes at the cracks using handheld drilling equipment, injecting grout into the holes using a manual grouting machine or a simple grouting jug, and finally manually smoothing the repaired surface. Some improved technologies use vehicle-mounted grouting equipment, but manual assistance is still required for positioning and operating the grouting gun. In these traditional methods, crack detection relies entirely on human experience, drilling locations are determined visually, grouting pressure is manually adjusted by the operator, and repair quality control depends on post-repair sampling inspections.

[0003] The above-mentioned traditional repair process has the following technical defects: (1) Crack detection relies on manual visual inspection, which easily misses small cracks with a width of less than 2mm. Moreover, the detection accuracy drops significantly under adverse weather conditions such as night, rain, and fog, making it impossible to achieve all-weather operation; (2) Drilling position relies on manual marking and hand-held drilling, resulting in poor positioning accuracy. The drilling depth cannot be adaptively adjusted according to the crack depth, which leads to the grout not being able to effectively fill the deep cracks during grouting; (3) The pressure is manually controlled during the grouting process, which cannot maintain a constant grouting pressure. This easily leads to incomplete grouting or grout overflow and waste. Furthermore, the grouting volume cannot be accurately measured, resulting in large fluctuations in repair quality; (4) The entire repair process involves many manual interventions, resulting in high labor intensity. There are serious safety hazards for personnel working on busy roads, making it impossible to achieve large-scale, high-efficiency automated repair operations. These problems lead to high road repair costs, long repair cycles, and difficulty in guaranteeing the service life of the repaired road surface. Secondary cracking usually occurs within 1-2 years after repair. Summary of the Invention

[0004] The purpose of this invention is to provide an automated repair process for road surface cracks by drilling and pressurizing grouting, in order to solve the problems of low repair efficiency, insufficient precision and high reliance on manual labor in the existing technology.

[0005] To achieve the above objectives, the present invention adopts the following technical solution: 1. An automatic repair process for road surface cracks by drilling and pressurizing grouting, characterized by comprising the following steps: S1. Automatic crack detection steps: The image acquisition device mounted on the mobile platform acquires road surface images, and artificial intelligence algorithms are used to identify the location, length, width and depth of cracks; S2. Precise positioning and drilling steps: Based on the crack location information obtained in the automatic crack detection step, control the robotic arm to drive the drilling device to automatically locate and drill at the crack. S3. Pressurized grouting step: Grout is injected into the borehole through a high-pressure grouting device, and a closed-loop pressure feedback system is used to monitor and control the grouting pressure and grouting flow rate in real time. S4. Curing and Surface Treatment Steps: Accelerate curing and smooth the surface of the repaired area where the injected slurry is applied. S5. Quality Inspection and Data Recording Steps: Perform non-destructive testing on the repaired area to determine the repair quality, and upload the process data to the cloud platform.

[0006] Preferably, in the automatic crack detection step, the artificial intelligence algorithm is an image recognition model based on a convolutional neural network (CNN). This model is pre-trained using an image dataset containing various crack types, different lighting conditions, different road surface backgrounds, and different levels of pollution. The crack types include at least transverse cracks, longitudinal cracks, mesh cracks, and block cracks.

[0007] Preferably, in the automatic crack detection step, the crack parameters output by the artificial intelligence algorithm include at least the crack centerline coordinates, length, average width, and estimated depth; and when the identified crack width is ≥1mm and the depth is ≥5mm, the subsequent precise positioning and drilling steps are automatically triggered.

[0008] Preferably, in the precise positioning and drilling step, a GPS / RTK differential positioning system and an inertial navigation IMU module are used for data fusion to achieve centimeter-level precise positioning during the dynamic movement of the mobile platform; the robotic arm is a six-degree-of-freedom robotic arm with an end-effector positioning accuracy of less than or equal to ±1mm, and its motion trajectory is planned in real time by the controller based on the coordinates of the crack centerline.

[0009] Preferably, in the precise positioning and drilling step, the drilling depth is automatically calculated based on the crack depth, and is 1.5-2 times the crack depth and does not exceed the road surface thickness; during drilling, atomized coolant is injected synchronously through the coolant channel built into the drill bit; the number of holes is automatically calculated based on the crack length, and the spacing between adjacent holes is controlled within the range of 10-20cm.

[0010] Preferably, in the pressurized grouting step, the closed-loop pressure feedback system includes a pressure sensor, a flow meter, and a programmable logic controller (PLC) installed in the grouting pipeline; the PLC receives feedback signals from the pressure sensor and the flow meter in real time and compares them with a preset pressure curve; when the PLC determines that the grout filling rate has reached a preset threshold based on the ratio of the cumulative flow of the flow meter to the theoretical grouting volume, it automatically stops grouting.

[0011] Preferably, in the pressurized grouting step, the grouting pressure is adaptively adjusted according to the crack depth: for shallow cracks with a depth <10mm, the grouting pressure is controlled at 0.5-1MPa; for deep cracks with a depth ≥10mm, the grouting pressure is controlled at 2-5MPa.

[0012] Preferably, in the curing and surface treatment steps, an infrared heating lamp or a hot air gun is used to heat the repair area, and the curing time is controlled within 10-30 minutes; after curing, a scraper is used to smooth the repair surface, and a layer of silane-based or acrylic-based waterproof sealant is sprayed onto the surface of the repair area using a spraying device.

[0013] Preferably, in the quality inspection and data recording step, an ultrasonic detector or an infrared thermal imager is used to perform non-destructive testing on the repair area; if the test result is unqualified, the system automatically marks the position and returns to execute the precise positioning and drilling step and subsequent steps until the test is qualified.

[0014] Preferably, in the quality inspection and data recording step, the process data includes at least the original image of the crack, identification parameters, drilling coordinates, grouting volume, grouting pressure curve, curing temperature curve, and inspection results; the above data is uploaded to the cloud server via 5G or 4G network to automatically generate a digital repair report containing time, location, and repair process parameters.

[0015] The present invention has the following beneficial effects: 1. This invention achieves high-precision automatic detection of pavement cracks by introducing an artificial intelligence image recognition algorithm based on convolutional neural networks. Simultaneously, the quantitative parameters such as crack depth and width output by the AI ​​model provide precise input conditions for adaptive drilling and grouting, enabling the entire repair process to truly achieve data-driven intelligent operation.

[0016] 2. This invention achieves centimeter-level precise positioning during the dynamic movement of a mobile platform by fusing GPS / RTK differential positioning with an inertial navigation IMU module and combining it with a six-degree-of-freedom high-precision robotic arm. The drilling device can automatically calculate the drilling depth based on the crack depth detected by AI, maintaining an optimal hole depth ratio of 1.5-2 times, and prevents road surface thermal damage through built-in coolant channels.

[0017] 3. This invention employs a closed-loop pressure feedback system, which, through the coordinated operation of pressure sensors, flow meters, and programmable logic controllers, achieves real-time monitoring and precise control of the grouting process. Simultaneously, an adaptive mechanism that automatically switches the grouting pressure based on crack depth ensures that shallow cracks will not damage the pavement structure due to excessive pressure, and deep cracks will not result in incomplete filling due to insufficient pressure.

[0018] 4. This invention constructs a complete closed-loop control system from detection and repair to quality inspection. Immediately after repair, non-destructive testing is performed using ultrasonic or infrared thermal imagers. Non-conforming areas automatically trigger a rework process, ensuring that every repaired area meets quality standards. Simultaneously, all process data is uploaded to a cloud platform in real time via a 5G network, generating a digital repair report. This enables full traceability and querying of the repair process, providing reliable data support for road maintenance decisions and significantly improving the digitalization level of road maintenance management. Attached Figure Description

[0019] Figure 1 This is a process flow diagram of the present invention. Detailed Implementation

[0020] The technical solutions of the present invention will be clearly and completely described below with reference to the accompanying drawings of the embodiments of the present invention. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments.

[0021] Example 1: Repair of deep longitudinal cracks in highways Please see Figure 1 This embodiment provides an automated repair process for deep longitudinal cracks in highways, specifically including the following steps: Automatic crack detection steps: A tracked robot is used as the mobile platform. The platform is 2.5m long, 1.8m wide, and 1.2m high, with a maximum travel speed of 15km / h. The front of the platform is equipped with a Sony IMX series high-definition visible light camera (1920×1080 resolution, 30fps) and a FLIR infrared thermal imager (640×512 resolution, thermal sensitivity <0.05℃). The mobile platform travels on the highway at a speed of 8km / h, continuously acquiring road surface images. The image data is transmitted in real time to an onboard industrial control computer, which has a built-in image recognition model based on a convolutional neural network (CNN). This model was trained using 120,000 road surface images, including 25,000 images of transverse cracks, 30,000 images of longitudinal cracks, 40,000 images of network cracks, and 25,000 images of block cracks. The training set covers various scenarios such as daytime, nighttime, backlighting, rain, and oily road surfaces. During the inspection, a longitudinal crack measuring 50cm in length, 3mm in width, and 15mm in depth was detected. The AI ​​model output the coordinates of the crack's centerline (starting point: X=125.36m, Y=3.45m; ending point: X=125.86m, Y=3.45m), a crack width of 3.2mm, and an estimated depth of 15mm. The identification results triggered subsequent steps.

[0022] Precise positioning and drilling steps: A GPS / RTK antenna is installed on the top of the mobile platform to receive differential signals from the BeiDou satellite, achieving a positioning accuracy of ±2cm. Simultaneously, an inertial navigation IMU module is installed at the bottom of the platform, which is fused with GPS data using Kalman filtering to achieve a dynamic positioning error ≤±3cm. A six-degree-of-freedom robotic arm (model: UR10e, repeatability ±0.1mm) plans its motion trajectory based on the coordinates of the crack centerline. The controller calculates the crack length to be 50cm and, according to a standard hole spacing of 17cm, determines to drill three holes at positions 12cm, 29cm, and 46cm from the crack (distance from the starting point). The robotic arm's end carries a high-pressure pneumatic drill bit (model: AtlasCopcoLBB16, adjustable speed 0-25000rpm, adjustable impact frequency 0-50Hz), with a drill diameter of 8mm. The control system automatically calculates the drilling depth to be 23mm based on a crack depth of 15mm (1.5 times the crack depth and less than the road surface layer thickness of 40mm). During drilling, atomized water-based coolant is simultaneously injected into the drill bit's built-in coolant channel at a flow rate of 0.1 L / min, keeping the drill bit temperature below 60°C. Each hole takes approximately 8 seconds to drill, and all three holes are completed in a total of 24 seconds.

[0023] Pressurized grouting procedure: A cement-epoxy mixed grout is pre-prepared with the following proportions: 100 parts P.O42.5 cement, 25 parts bisphenol A epoxy resin, 12.5 parts polyamide curing agent, 35 parts water, 1.5 parts carbon nanofibers (diameter 10-20nm, length 10-30μm), and 2 parts silicone waterproofing agent. The grout viscosity, measured by a rotational viscometer, is 105 Pa·s (25℃). The high-pressure grouting device includes a stainless steel storage tank (50L capacity), an electric high-pressure plunger pump (model: GracoXtremeMix, pressure range 0-8MPa, flow rate 0-5L / min), a grouting gun, and a grouting pipe. The grouting gun is equipped with a 10mm diameter rubber sealing head at the front end, which fits tightly with the drilled hole. During grouting, the robotic arm inserts the grouting gun into three drilled holes sequentially. The control system automatically selects the high-pressure mode based on a crack depth of 15mm (≥10mm), setting the grouting pressure to 2.5MPa. The closed-loop pressure feedback system includes a pressure sensor (Siemens 7MF1565, range 0-6MPa, accuracy 0.5%), a turbine flow meter (OMEGAFTB-4605, range 0.1-5L / min, accuracy 1%), and a programmable logic controller (PLC) (Siemens S7-1200) installed on the grouting pipeline. The PLC collects pressure and flow data in real time at a sampling frequency of 100Hz. The theoretical grouting volume is calculated based on the crack volume: Crack volume = length × width × depth × filling coefficient (taken as 1.2) = 0.5m × 0.0032m × 0.015m × 1.2 = 2.88 × 10⁻ 5m³ = 28.8 ml. Considering the drilling volume (3 holes, 8 mm diameter, 23 mm depth, total drilling volume = 3 × π × 0.004² × 0.023 = 3.47 × 10⁻ 6 The theoretical grouting volume is approximately 32.3 ml (m³ = 3.47 ml). During actual grouting, the PLC monitors the cumulative flow rate. When the cumulative flow rate reaches 95% of the theoretical grouting volume (i.e., 30.7 ml), the PLC issues a command to stop grouting. In the actual grouting process, 11.2 ml, 10.8 ml, and 10.5 ml were injected into the three holes respectively, totaling 32.5 ml, which matches the theoretical value. The grouting pressure curve shows that the pressure remained stable within the range of 2.45-2.55 MPa throughout the entire grouting process.

[0024] Curing and Surface Treatment Steps: After grouting, the infrared heating lamp array (3kW power, 2-10μm wavelength) at the rear of the moving platform is activated to heat the repair area. The ambient temperature monitoring module (temperature sensor model: PT100, accuracy ±0.2℃) detects an ambient temperature of 18℃, and the PLC automatically sets the heating time to 18 minutes according to the preset algorithm (curing time = reference time × temperature compensation coefficient). During the heating process, the surface temperature of the repair area is maintained between 45-55℃. After curing, the end of the robotic arm is replaced with a leveling tool to scrape away excess grout above the road surface. The leveling pressure is controlled at 10N to ensure that the surface flatness error is ≤±0.5mm. Subsequently, the spraying device sprays a layer of silane-based waterproof sealant (model: BASF MasterSeal345) onto the surface of the repair area, with a spray thickness of 0.2mm and a spray width covering 5cm on both sides of the crack.

[0025] Quality inspection and data recording steps: After curing, the ultrasonic detector (model: PunditPL-200, frequency 50 kHz) at the rear of the mobile platform detects the repaired area. The ultrasonic probe moves on the surface of the repaired area to collect sound velocity data. The average sound velocity of the repaired area is 4120 m / s, and the average sound velocity of the surrounding intact road surface is 4180 m / s. The sound velocity ratio is 98.6%, which is greater than the qualified threshold of 95%, so it is determined to be qualified. At the same time, the infrared thermal imager performs thermal imaging detection on the repaired area, and the heat diffusion is uniform without abnormal cold spots. After all the detection data is determined to be qualified, the vehicle-mounted industrial computer packages all the data of this repair, including: original crack images (JPEG format, 12 pieces), AI recognition parameters (JSON format, including crack coordinates and dimensions), drilling coordinates (CSV format), grouting volume record (32.5 ml), grouting pressure curve (5000 sampling points), curing temperature curve (1000 sampling points), and ultrasonic detection data (waveform diagram and sound velocity value). The data is uploaded to the cloud server in real time through the 5G module (model: Huawei MH5000). The server automatically generates a digital repair report with the report number: HW-20230516-003, including the repair time (14:23:15 on May 16, 2023), location coordinates (GPS coordinates: N32°12'36.5", E118°48'22.3"), repair process parameters, etc. The whole process takes 5 minutes and 48 seconds without manual intervention.

[0026] Example 2: Repair of shallow mesh cracks on municipal roads Please refer to Figure 1 , this example provides an automatic repair process for shallow mesh cracks on municipal roads, which specifically includes the following steps: Automatic crack detection steps: An unmanned vehicle platform was used. This platform, based on a modified electric vehicle, is 3.2m long, 1.6m wide, and 1.5m high, with a maximum speed of 30km / h. The platform's front end is equipped with a high-definition camera (model: Baslerac A1920-40gc, resolution 1920×1200, frame rate 40fps) and a line laser scanner (model: SICKLMS511, scanning angle 190°, angular resolution 0.25°) to acquire the three-dimensional contours of the road surface. The mobile platform travels at 5km / h on a municipal road, scanning a section of road surface with a network of cracks. The AI ​​image recognition model detected a network of cracks approximately 0.8m × 0.6m in area, containing 5 main cracks. The crack width ranged from 1.2-2.5mm, and the laser scanner measured the crack depth to be between 4-7mm. The AI ​​model outputs the centerline coordinates and parameters of five cracks: Crack 1 (length 45cm, average width 1.5mm, depth 5mm), Crack 2 (length 38cm, average width 1.8mm, depth 6mm), Crack 3 (length 52cm, average width 2.2mm, depth 7mm), Crack 4 (length 30cm, average width 1.2mm, depth 4mm), and Crack 5 (length 25cm, average width 1.3mm, depth 4mm). All cracks are ≥1mm wide, and their depths are ≥5mm, except for Cracks 4 and 5, which are 4mm deep (slightly below the 5mm threshold). The control system performs secondary analysis on Cracks 4 and 5, finding them located at the edge of the mesh area and connected to adjacent cracks. Therefore, it decides to include them in the repair scope, triggering subsequent steps.

[0027] Precise positioning and drilling steps: After fusion of GPS / RTK positioning system and IMU data, the dynamic positioning accuracy reaches ±2.5cm. A six-degree-of-freedom robotic arm (model: KUKAKR6R700, repeatability ±0.05mm) plans the drilling positions according to the distribution of 5 cracks. Considering the characteristics of the network cracks, a grid-based drilling strategy is adopted: within a 0.8m×0.6m area, 20 holes are arranged in a 15cm×15cm grid. The drilling device uses an electric pulse drill bit (model: Bosch GSB18V-150, maximum speed 1900rpm, impact energy 1.7J), with a drill bit diameter of 6mm. The control system automatically calculates the drilling depth based on the crack depth below each drilling position: for areas with crack depth ≥5mm, the drilling depth is 1.8 times the crack depth; for areas with crack depth 4mm (cracks 4 and 5), to ensure grouting effect, the drilling depth is 8mm (i.e., twice the crack depth, but not exceeding 50mm of road surface thickness). Specific calculations: Drilling depth in crack 1 area was 9mm, crack 2 area was 11mm, crack 3 area was 13mm, crack 4 and 5 areas were 8mm, and the drilling depth in crack-free areas was based on the average depth of the surrounding cracks, 10mm. During drilling, atomized coolant was injected through the drill bit's built-in coolant channel at a flow rate of 0.08L / min. The total drilling time for 20 holes was approximately 2 minutes and 40 seconds.

[0028] Pressurized grouting steps: Prepare epoxy resin-based grout with the following proportions: 100 parts bisphenol F type epoxy resin, 32 parts polyetheramine curing agent, 3 parts nano silica (particle size 20-30nm), 1.5 parts organosilicon waterproofing agent, and 5 parts diluent (acetone). The grout viscosity was measured to be 82 Pa·s (25℃). The high-pressure grouting device includes a storage tank (30L capacity), a pneumatic double diaphragm pump (model: Wilden P025, pressure range 0-8 bar, maximum flow rate 15L / min), a grouting gun, and a grouting pipe. The control system sets the grouting pressure according to the crack depth at each borehole location: for boreholes with a depth ≥5mm (crack areas 1-3), a high-pressure mode of 2.0MPa is used; for boreholes with a depth <5mm (crack areas 4-5 and surrounding areas), a low-pressure mode of 0.8MPa is used. During grouting, the robotic arm sequentially inserts the grouting gun into 20 boreholes. The PLC adjusts the air intake pressure of the pneumatic pump in real time according to the pressure set value corresponding to each borehole, ensuring that the grouting pressure is stable within ±5% of the set value. The closed-loop pressure feedback system includes a pressure sensor (model: Danfoss MBS3000, range 0-4MPa, accuracy 1%), an electromagnetic flowmeter (model: Yokogawa AXF005G, range 0-5L / min, accuracy 0.5%), and a PLC (model: Mitsubishi FX5U). Theoretical grouting volume calculation: The theoretical grouting volume for each borehole is calculated based on the volume of the corresponding crack, with a total theoretical grouting volume of approximately 145ml. The PLC monitors the cumulative flow rate in real time, and automatically stops grouting when the cumulative flow rate reaches 95% (138ml) of the theoretical grouting volume. The actual grouting volume is 152ml, which is in line with expectations considering the additional filling of micro-cracks around the boreholes.

[0029] Curing and Surface Treatment Steps: After grouting, an infrared heating lamp array (2kW power, 2-4μm wavelength) heats the repaired area. Ambient temperature monitoring shows 25℃, humidity 65%, and the heating time is set to 12 minutes. During curing, the surface temperature of the repaired area is maintained at 50-60℃. After curing, a leveling machine smooths the repaired surface and removes excess grout. A layer of acrylic-based waterproof sealant (model: Sika1KFlex) is sprayed onto the surface with a thickness of 0.15mm.

[0030] Quality inspection and data recording steps: Use an infrared thermal imager (model: FLIRT540, resolution 464×348) to detect the repaired area. Immediately take a thermal image after heating and analyze the uniformity of heat diffusion. The thermal image shows that the heat diffusion in the entire repaired area is uniform without obvious cold spots, and it is determined to be qualified. All data is packaged and uploaded, including crack images (20 pieces), AI recognition parameters (JSON), drilling coordinates (20 points, CSV), grouting records (detailed grouting volume per hole, a total of 152 ml), grouting pressure curves (20 curves), curing temperature curves, and infrared thermal images (10 pieces). The data is uploaded to the cloud via 4G network to generate a repair report. The entire process takes 8 minutes and 15 seconds.

[0031] Example 3: Repair adaptability in rainy weather Please refer to Figure 1 , this example provides an automatic repair process for road cracks in rainy weather, highlighting the adaptability of this process under恶劣 weather conditions, specifically including the following steps: Automatic crack detection step: The mobile platform uses a tracked robot to operate in the rainy environment with a humidity of 85%. There is water accumulation on the road surface and the visibility is low. The platform simultaneously turns on a high-definition visible light camera and an infrared camera. The AI image recognition model uses a multi-modal fusion algorithm to perform pixel-level fusion of the visible light image and the infrared image. The infrared image shows that the temperature in the crack area is slightly higher than the surrounding area (due to the difference in evaporation cooling of the water accumulation), and the visible light image is greatly affected by the reflection of the water accumulation. The fused image clearly shows the crack contour. A horizontal crack with a length of 65 cm, a width of 4 mm, and a depth of 12 mm is detected. The AI model outputs the crack parameters with a confidence level of 96.5%, triggering the subsequent steps. This example specifically explains that the AI model training set contains approximately 15,000 rainy road surface images, so it has good adaptability to the rainy environment.

[0032] Precise positioning and drilling steps: GPS / RTK positioning has stable signals in rainy weather. After fusing with IMU, the positioning accuracy is ±3 cm. The robotic arm is positioned at the center of the crack. According to the crack length of 65 cm, 4 holes are determined to be drilled at intervals of 16 cm, and the hole positions are respectively at 10 cm, 26 cm, 42 cm, and 58 cm of the crack. The drilling depth is calculated as 20 mm (1.67 times) according to the crack depth of 12 mm. Before drilling, the robotic arm first executes a hole cleaning procedure: Through the compressed air channel inside the drill bit, high-pressure air (pressure 0.6 MPa, lasting for 2 seconds) is sprayed at each predetermined drilling position to dry the accumulated water on the surface of the hole position. When drilling, the coolant channel normally injects atomized coolant, and at the same time, the rotation speed of the drill bit is reduced by 10% (to prevent slipping under wet conditions) to ensure the drilling quality. The total drilling time for 4 holes is 35 seconds.

[0033] Pressurized grouting steps: The slurry formula is adjusted according to the high-humidity environment: 3 parts of water absorbent (diatomaceous earth) are added to the cement-epoxy mixed slurry, and the water consumption is reduced by 5 parts. The ratio is: 100 parts of P.O42.5 cement, 20 parts of bisphenol A epoxy resin, 10 parts of polyamide curing agent, 30 parts of water, 3 parts of diatomaceous earth, 1 part of carbon nanofiber, and 2 parts of silicone waterproofing agent. The slurry viscosity is measured as 135 Pa·s (25 °C). The grouting pressure is automatically increased according to the humidity compensation algorithm: The basic pressure should have selected the high-pressure mode of 2.5 MPa at a depth of 12 mm. The control system automatically increases the compensation coefficient by 1.2 according to the humidity of 85%. The actual grouting pressure is set at 3.0 MPa. The closed-loop pressure feedback system works normally, and the PLC monitors the pressure and flow rate in real time. Theoretical grouting volume calculation: Crack volume = 0.65 m × 0.004 m × 0.012 m × 1.2 = 3.74×10⁻ 5 m³ = 37.4 ml. Adding the drilling volume (4 holes, hole diameter 8 mm, hole depth 20 mm, total volume = 4×π×0.004²×0.02 = 4.02×10⁻ 6 m³ = 4.02 ml), the total theoretical grouting volume is about 41.4 ml. The PLC sets the stop threshold at 95% (39.3 ml). During the actual grouting process, due to the influence of the high-humidity environment, the grouting pressure fluctuates slightly larger in the initial stage, but the PID adjustment algorithm quickly stabilizes the pressure. After 3 seconds, the pressure stabilizes at 2.95 - 3.05 MPa. When the cumulative flow rate reaches 39.5 ml, it automatically stops. The actual total grouting volume is 42.1 ml.

[0034] Curing and surface treatment steps: In a high-humidity environment, the curing time needs to be extended. The power of the heating device is increased to 4 kW, and the heating time is set at 28 minutes (40% longer than normal). During the heating process, the surface temperature of the repair area is maintained at 50 - 60 °C, and at the same time, a hot air gun blows hot air at a low speed towards the repair area to accelerate water evaporation. After curing, a leveling machine is used for leveling. When spraying the sealing layer, a two-component moisture-curing polyurethane sealing material is used. This material cures faster in a high-humidity environment and the surface dries within 15 minutes after spraying.

[0035] Quality inspection and data recording steps: An infrared thermal imager is used for inspection. Due to the high environmental humidity, a special algorithm is used in the thermal image analysis to exclude the interference of surface moisture. The thermal image shows that the heat diffusion in the repair area is uniform, and it is judged to be qualified. At the same time, an ultrasonic detector is used for auxiliary inspection. The sound velocity value is 4020 m / s, the surrounding is 4060 m / s, and the sound velocity ratio is 99%, which is qualified. All data is packed and uploaded, with special markings for environmental parameters (humidity 85%, temperature 22 °C) and compensation parameters (pressure compensation coefficient 1.2, curing time extension coefficient 1.4). The data is uploaded to the cloud to generate a special repair report for rainy days. The entire process takes about 32 minutes (including the extended curing time).

[0036] Example 4: Nighttime Low Temperature Environment Remediation Please see Figure 1 This embodiment provides an automatic repair process for road surface cracks in low-temperature nighttime environments, specifically including the following steps: Automatic crack detection steps: The mobile platform operates in an environment with a nighttime temperature of -5℃. A high-definition visible light camera with supplementary lighting is activated, while an infrared camera operates simultaneously. Due to the more significant temperature difference between the crack and the surrounding road surface in low-temperature conditions (higher thermal resistance at the crack, resulting in slower cooling), the infrared image contrast is high. The AI ​​model identifies multiple cracks with a detection accuracy of 99.2%, showing no significant difference compared to conditions at normal temperatures.

[0037] Precise positioning and drilling procedures: Road surface materials become brittle in low-temperature environments, requiring adjustment of drilling parameters. The control system automatically reduces the drill bit speed by 15% and the impact frequency by 20% to prevent road surface chipping. The drilling depth is calculated as 1.5 times the crack depth. The coolant channel is closed for liquid cooling, switching to pulsed compressed air cooling to prevent the coolant from freezing at low temperatures. The drilling process proceeded normally.

[0038] Pressurized grouting procedure: The viscosity of the grout increases at low temperatures, requiring formula adjustment. The grout uses a low-temperature epoxy resin (glass transition temperature -20℃), with the following proportions: 100 parts low-temperature epoxy resin, 35 parts low-temperature curing agent, 1 part nanofiber, and 5 parts antifreeze (ethylene glycol). The grout viscosity at -5℃ is 185 Pa·s, still within a controllable range. After calculating the grouting pressure based on depth, a low-temperature compensation coefficient of 1.15 is added. The grouting process proceeds normally.

[0039] Curing and Surface Treatment Steps: In low-temperature environments, conventional heating methods are insufficient to reach the curing temperature. A combination of high-power infrared heating lamps (5kW) and high-frequency induction heating technology is used to rapidly raise the surface temperature of the repaired area to 60℃ and maintain it for 25 minutes. An insulation cover is installed on the outside of the heating device to reduce heat loss. Surface treatment is performed immediately after curing.

[0040] Quality inspection and data recording steps: The inspection process was normal. All data was packaged and uploaded, with low-temperature environmental parameters labeled. This embodiment demonstrates that the process of this invention can operate normally in a -5℃ low-temperature environment, breaking through the temperature limitations of traditional repair processes.

[0041] The above description is only a preferred embodiment of the present invention, but the scope of protection of the present invention is not limited thereto. Any equivalent substitutions or modifications made by those skilled in the art within the scope of the technology disclosed in the present invention, based on the technical solution and inventive concept of the present invention, should be covered within the scope of protection of the present invention.

Claims

1. An automatic repair process for road surface cracks by drilling and pressurizing grouting, characterized in that, Includes the following steps: S1. Automatic crack detection steps: The image acquisition device mounted on the mobile platform acquires road surface images, and artificial intelligence algorithms are used to identify the location, length, width and depth of cracks; S2. Precise positioning and drilling steps: Based on the crack location information obtained in the automatic crack detection step, control the robotic arm to drive the drilling device to automatically locate and drill at the crack. S3. Pressurized grouting step: Grout is injected into the borehole through a high-pressure grouting device, and a closed-loop pressure feedback system is used to monitor and control the grouting pressure and grouting flow rate in real time. S4. Curing and Surface Treatment Steps: Accelerate curing and smooth the surface of the repaired area where the injected slurry is applied. S5. Quality Inspection and Data Recording Steps: Perform non-destructive testing on the repaired area to determine the repair quality, and upload the process data to the cloud platform.

2. The process according to claim 1, characterized in that, In the automatic crack detection step, the artificial intelligence algorithm is an image recognition model based on a convolutional neural network (CNN). This model is pre-trained using an image dataset containing various crack types, different lighting conditions, different road surface backgrounds, and different levels of pollution. The crack types include at least transverse cracks, longitudinal cracks, network cracks, and block cracks.

3. The process according to claim 1, characterized in that, In the automatic crack detection step, the crack parameters output by the artificial intelligence algorithm include at least the crack centerline coordinates, length, average width, and estimated depth; and when the identified crack width is ≥1mm and the depth is ≥5mm, the subsequent precise positioning and drilling steps are automatically triggered.

4. The process according to claim 1, characterized in that, In the precise positioning and drilling steps, a GPS / RTK differential positioning system and an inertial navigation IMU module are used for data fusion to achieve centimeter-level precise positioning during the dynamic movement of the mobile platform; the robotic arm is a six-degree-of-freedom robotic arm with an end-effector positioning accuracy of less than or equal to ±1mm, and its motion trajectory is planned in real time by the controller based on the coordinates of the crack centerline.

5. The process according to claim 1, characterized in that, In the precise positioning and drilling steps, the drilling depth is automatically calculated based on the crack depth, and is 1.5-2 times the crack depth and does not exceed the road surface thickness; during drilling, atomized coolant is injected synchronously through the coolant channel built into the drill bit; the number of holes is automatically calculated based on the crack length, and the spacing between adjacent holes is controlled within the range of 10-20cm.

6. The process according to claim 1, characterized in that, In the pressurized grouting step, the closed-loop pressure feedback system includes a pressure sensor, a flow meter, and a programmable logic controller (PLC) installed in the grouting pipeline. The PLC receives feedback signals from the pressure sensor and the flow meter in real time and compares them with a preset pressure curve. When the PLC determines that the grout filling rate has reached a preset threshold based on the ratio of the cumulative flow of the flow meter to the theoretical grouting volume, it automatically stops grouting.

7. The process according to claim 6, characterized in that, In the pressurized grouting step, the grouting pressure is adaptively adjusted according to the crack depth: for shallow cracks with a depth <10mm, the grouting pressure is controlled at 0.5-1MPa; for deep cracks with a depth ≥10mm, the grouting pressure is controlled at 2-5MPa.

8. The process according to claim 1, characterized in that, In the curing and surface treatment steps, an infrared heating lamp or hot air gun is used to heat the repair area, and the curing time is controlled between 10 and 30 minutes. After curing, a scraper is used to smooth the repair surface, and a layer of silane-based or acrylic waterproof sealant is sprayed onto the repair area surface using a spraying device.

9. The process according to claim 1, characterized in that, In the quality inspection and data recording steps, an ultrasonic detector or an infrared thermal imager is used to perform non-destructive testing on the repair area; if the test result is unqualified, the system automatically marks the position and returns to execute the precise positioning and drilling steps and subsequent steps until the test is qualified.

10. The process according to claim 1, characterized in that, In the quality inspection and data recording steps, the process data includes at least the original image of the crack, identification parameters, drilling coordinates, grouting volume, grouting pressure curve, curing temperature curve, and inspection results; the above data is uploaded to the cloud server via 5G or 4G network to automatically generate a digital repair report containing time, location, and repair process parameters.

Images