Steel-concrete composite beam bridge deck slab automatic construction system and method based on formwork platform circulation

Abstract

The invention provides a steel-concrete composite beam bridge deck slab automatic construction system and method based on formwork platform circulation, and is suitable for the technical field of prefabricated bridge deck slab construction. The method comprises the steps of automatic cleaning of a formwork platform, automatic oil injection of the formwork platform, installation of a formwork, pouring and vibration of a bridge deck slab, flattening of the bridge deck slab and maintenance and formwork removal, and full-process automatic construction of the steel-concrete composite beam bridge deck slab is achieved through automatic cleaning of the formwork platform, a defect recognition model and a gray gradient quantitative evaluation system. In the step of cleaning the mold table, a linear array camera and LED diffused light sources on the two sides are combined with an improved Scharr operator, a machine vision gray gradient analysis system is established, and the problem that residues are not cleaned thoroughly is solved through Gaussian weighted fusion of local roughness and feedback adjustment of the inclination angle and shear force of a scraper; in the bridge deck slab pouring and flattening links, on the basis of YOLOv7-tin lightweight model construction, a dynamic threshold value judgment method of a sub-region gray variance and gradient amplitude defect area proportion is adopted, and subjective errors of manual visual judgment are avoided.

Description

Technical Field

[0001] The present invention relates to the technical field of prefabricated bridge deck construction, and in particular to an automated construction system and method for steel-concrete composite beam bridge decks based on formwork circulation. Background Art

[0002] In recent years, steel-concrete composite beam bridges have become a mainstream bridge type due to their lightweight, high load-bearing capacity, and ease of construction. Industry demands for construction efficiency, quality control, and cost optimization continue to rise, driving the transition of bridge construction toward automation and intelligentization. However, traditional bridge deck construction suffers from low formwork recycling efficiency and high reliance on manual labor. This is particularly true in critical areas such as formwork cleanliness control, concrete defect detection, and surface treatment. Existing technologies struggle to meet the high-precision, high-stability requirements of industrialized production.

[0003] Existing technologies often rely on manual or mechanical cleaning with fixed parameters to clean residual material from the formwork surface. These methods lack a real-time feedback adjustment mechanism based on surface roughness, which can easily lead to excessive wear or incomplete cleaning, impacting subsequent demolding quality and formwork life. Concrete vibration quality testing primarily relies on visual judgment based on worker experience, which is highly subjective and makes it difficult to achieve traceability of quality data. Furthermore, surface leveling effectiveness is often evaluated through contact measurement or qualitative observation after the concrete has solidified. A quantitative evaluation system based on image grayscale gradient analysis is lacking, resulting in insufficient construction consistency and a high rate of post-process repairs.

[0004] These shortcomings have led to extended construction periods and increased material losses, restricting the standardization process of industrialized bridge construction. There is an urgent need to achieve technological innovation through intelligent technology. Summary of the Invention

[0005] The purpose of the present invention is to solve the problems of low recycling efficiency of existing bridge deck prefabrication construction formwork, high dependence on manpower, and easy lead to extended construction period and increased material loss. An automated construction system and method for steel-concrete composite beam bridge deck based on formwork circulation is proposed, which can be widely used in the field of prefabricated bridge deck construction technology.

[0006] To achieve the above object, the present invention provides the following technical solutions:

[0007] The automated construction system for steel-concrete composite beam bridge decks based on formwork circulation includes a formwork cleaning module, a formwork oiling module, a material distribution module, a leveling module, a maintenance module, a visual inspection and analysis module, a motion control module, a data integration module, and an exception handling module.

[0008] The die table cleaning module is used to adjust the scraper shear force and angle according to the surface roughness data fed back by the visual inspection and analysis module, and integrates the roller brush speed control system and the residue recovery device;

[0009] The die table oil spraying module is used to spray oil on the die table surface according to the pressure feedback of the oil spraying station and the oil film thickness parameters;

[0010] The distribution module is used to control the concrete conveyor to transport concrete from the mixing plant to the production line. The distribution machine's mixing shaft and worm extrusion and discharging method are controlled by a wireless portable remote controller. The built-in PLC control chip automatically sets the travel trajectory and distribution speed according to the component's external dimensions and thickness parameters. It is also used to adjust the excitation force and amplitude of multiple independent vibrating tables.

[0011] The leveling module is used to control the leveling machine to level and smooth the entire top surface of the bridge deck, and to level and smooth local surfaces according to the relative coordinate positions of the bridge deck;

[0012] The curing module is used for automatic sensing of the entry and exit of the formwork platform, as well as temperature and humidity control of the curing kiln during the curing of the bridge deck;

[0013] The visual inspection and analysis module is used to perform image recognition through an improved Scharr operator gradient analysis algorithm, a YOLOv7-tiny defect detection model, and a grayscale variance dynamic evaluation program;

[0014] The motion control module is used to control the movement and transfer of the mold platform on the assembly line;

[0015] The data integration module is used for data collection and storage during the construction process;

[0016] The exception handling module is used to record faults during the construction process and alert maintenance personnel.

[0017] The automated construction method for steel-concrete composite beam bridge deck based on formwork cycle is controlled by the automated construction system for steel-concrete composite beam bridge deck based on formwork cycle, and specifically includes the following steps:

[0018] S1. Automatic formwork cleaning: A tiltable scraper is used to scrape off the concrete residue solidified on the formwork during demoulding. A roller brush is then used to further clean the formwork surface. After cleaning, the formwork surface roughness is measured using machine vision grayscale gradient analysis. When the surface roughness of the measured area is greater than z, the formwork is retracted, the scraper tilt angle is adjusted to make the scraper fit more closely to the formwork surface, and the scraper shear force is increased to scrape off the concrete residue in the measured area.

[0019] S2, automatic oil spraying of the mold table, move the mold table to the bottom side mold arrangement station, and use the automatic oil sprayer to evenly spray a layer of oil film on the surface of the mold table during the movement;

[0020] S3, formwork installation, the formwork is assembled in blocks, the formwork is assembled on the ground, the release agent is sprayed and the whole is installed on the formwork platform, the steel cage, shear groove and embedded parts are placed, and the formwork is connected with bolts. After the installation is completed, the formwork platform is moved to the pouring and vibrating station;

[0021] S4: The bridge deck is poured and vibrated. The concrete placing boom and the bottom vibration table work together to complete the forming, pouring and vibration of the bridge deck. Then, an industrial camera is used to capture images of the bridge deck surface. Image recognition is used to determine whether there is obvious laitance or air holes on the bridge deck surface. If not, the formwork table is moved to the leveling station. Otherwise, vibration continues.

[0022] S5: Bridge deck leveling: Lower the leveling machine to the top of the alignment template, turn on the vibrator and rubbing plate, and then slowly and evenly move the vibration leveling machine to level and smooth the top surface of the bridge deck. Use image recognition to determine whether the bridge deck surface has a smooth mirror effect. If so, move the formwork table to a static position. Otherwise, the defects caused by leveling will be filled with a spreading machine and then leveled.

[0023] S6, maintenance and demoulding. After the bridge deck is left to stand and reaches the loose mold and roughened state, use a roughening machine to roughen the bridge deck, and then enter the curing cellar. Through automatic temperature and humidity control, the bridge deck reaches the demoulding state in 10 hours. After the bridge deck is cured, the mold platform is transferred to the lifting station and transported to the cleaning station.

[0024] As a preferred technical solution of the present invention, in step S1, the machine vision grayscale gradient analysis measures the surface roughness of the mold platform, including using a line array camera and a double-sided LED diffuse light source installed above the mold platform, with the angle θ between the camera axis and the normal line of the mold platform plane being no less than 15°; establishing a mapping relationship between the pixel coordinate system and the physical coordinate system,

[0025]

[0026] Where, (u,v)-pixel coordinates, (x,y)-physical coordinates, k x ,k y - The calibration coefficients are obtained through a standard checkerboard calibration plate, with the direction of the template movement set as the x-axis and the template width direction as the y-axis. The acquisition frequency is determined according to the image width of the linear array camera and the template movement speed. The collected template image is preprocessed using the non-local mean denoising algorithm and limited contrast adaptive histogram equalization. The image matrix I obtained after preprocessing is used to calculate the gradient amplitude using the improved Scharr operator.

[0027]

[0028] Where, The gradient amplitude at the coordinate, *-convolution operation; the image is divided into N small windows along the width of the template, and the gradient amplitude of the i-th pixel in the small window is Calculate the gradient statistic R for each small window local ,

[0029]

[0030] Where n is the number of pixels in the small window, and the local roughness R is fused by Gaussian weighted fusion. area ,

[0031]

[0032] Where w k - Small window weight, y k -The center position of the kth small window in the width direction of the mold platform, μ-the geometric center position in the width direction of the mold platform, σ-the standard deviation of the Gaussian distribution, controls the weight decay speed, takes W / 5, W is the width of the mold platform; establish the mold platform surface roughness R according to the standard roughness sample a and local roughness R area Mapping relationship,

[0033] R a =aR area +b(8)

[0034] The coefficients a and b are fitted by the least squares method.

[0035] As a preferred technical solution of the present invention, in step S4, the image recognition adopts a deep learning-based method, including shooting images of the surface of the concrete bridge deck just after vibration from a fixed angle on the formwork under different lighting conditions: lights on at night, lights on on cloudy days, and lights off during the day, marking irregular flaky areas with whitish color as floating slurry defects, and marking circular / elliptical holes as air pore defects; unifying the image size to 640×640 pixels and normalizing the pixel values ​​to the range of [0,1]; applying a random combination enhancement strategy to perform data enhancement through geometric transformation, color perturbation, and noise injection; using a lightweight target detection model YOLOv7-tiny, pre-training on a general data set to learn general object features; freezing the parameters of the first 50% layers of the backbone network, fine-tuning the parameters of the last 50% layers and the detection head, using the AdamW optimizer as the optimizer, and using a composite loss function, wherein the positioning loss uses CIoULoss and the classification loss uses FocalLoss; finally, using 5-fold cross validation to ensure the generalization ability of the model.

[0036] As a preferred technical solution of the present invention, in step S5, the image recognition is based on image data processing, including quantitative evaluation through grayscale variance and gradient mutation. An industrial-grade RGB camera is used with a ring-shaped LED uniform light source to capture the surface image of the bridge deck after leveling. The RGB image is converted into a grayscale image, and the image is divided into sub-regions with a resolution of w×h. The grayscale value standard deviation σ is calculated for each sub-region. g ,

[0037]

[0038] Where, N′ is the total number of pixels in the sub-region, g i - Gray value of a single pixel in the sub-region, μ g -The average gray value of the sub-region, the Sobel operator is used to calculate the gradient amplitude in the sub-region,

[0039]

[0040] Where R(x,y)-the gradient amplitude at the (x,y) coordinate in the sub-region, *-convolution operation, I′-the image matrix of the sub-region; the gradient amplitude is binarized,

[0041]

[0042] Defect area ratio R t for

[0043]

[0044] Set the judgment threshold σ max and R max , when σ g Greater than σ max or R t Greater than R max When , it means that the bridge deck surface does not achieve a smooth mirror effect; among them, σ max Determined based on multiple groups of bridge deck specimens, R max Based on the ratio of the acceptable defect area to the actual block area of ​​the bridge deck represented by the sub-area.

[0045] The beneficial effects of the present invention are as follows: through closed-loop control of automatic formwork cleaning, multi-illumination scene defect recognition model and grayscale gradient quantitative evaluation system, the full-process automated construction of steel-concrete composite beam bridge deck is realized; in the formwork cleaning link, a linear array camera and a double-sided LED diffuse light source are combined with an improved Scharr operator to establish a machine vision grayscale gradient analysis system, and the local roughness is fused by Gaussian weighted, and the scraper inclination angle and shear force are adjusted by feedback, thereby solving the problem of incomplete residue cleaning; in the bridge deck pouring and leveling link, a floating slurry and pore detection network across illumination scenes is constructed based on the YOLOv7-tiny lightweight model, and an innovative dynamic threshold judgment method of the sub-region grayscale variance and gradient amplitude defect area ratio is adopted to avoid the subjective error of manual visual judgment. BRIEF DESCRIPTION OF THE DRAWINGS

[0046] Figure 1 This is a flow chart of the automated construction method for steel-concrete composite beam bridge deck based on formwork circulation of the present invention. DETAILED DESCRIPTION

[0047] The following describes in detail specific embodiments of the present invention in conjunction with the accompanying drawings. It should be understood that the specific embodiments provided herein are intended only to illustrate and explain the present invention and are not intended to limit the present invention. It should be noted that many specific details are set forth in the following description to facilitate a full understanding of the present invention. However, the present invention may also have other embodiments and variations thereof. Therefore, the scope of protection of the present invention is not limited by the specific embodiments disclosed below.

[0048] In the first embodiment, the automated construction method of the steel-concrete composite beam bridge deck based on the formwork cycle is as follows:

[0049] S1. Automatic cleaning of the formwork platform. The formwork platform adopts a standardized steel formwork with sufficient rigidity. The standard formwork platform size is 4m×6m. The side formwork adopts a standardized steel formwork to prevent deformation of the formwork. A tiltable scraper is used to scrape off the concrete residue solidified on the formwork platform during demoulding. Then, a roller brush is used to further clean the formwork platform surface. After cleaning, the formwork platform surface roughness is measured by machine vision grayscale gradient analysis. The machine vision grayscale gradient analysis measures the formwork platform surface roughness, including using a linear array camera and a double-sided LED diffuse light source installed above the formwork platform. The angle θ between the camera axis and the normal line of the formwork platform plane is not less than 15°. A mapping relationship between the pixel coordinate system and the physical coordinate system is established.

[0050]

[0051] Where, (u,v)-pixel coordinates, (x,y)-physical coordinates, k x ,k y- The calibration coefficients are obtained through a standard checkerboard calibration plate, with the direction of the template movement set as the x-axis and the template width direction as the y-axis. The acquisition frequency is determined according to the image width of the linear array camera and the template movement speed. The collected template image is preprocessed using the non-local mean denoising algorithm and limited contrast adaptive histogram equalization. The image matrix I obtained after preprocessing is used to calculate the gradient amplitude using the improved Scharr operator.

[0052]

[0053] Where, The gradient amplitude at the coordinate, *-convolution operation; the image is divided into N small windows along the width of the template, and the gradient amplitude of the i-th pixel in the small window is Calculate the gradient statistic R for each small window local ,

[0054]

[0055] Where n is the number of pixels in the small window, and the local roughness R is fused by Gaussian weighted fusion. area ,

[0056]

[0057] Where w k - Small window weight, y k -The center position of the kth small window in the width direction of the mold platform, μ-the geometric center position in the width direction of the mold platform, σ-the standard deviation of the Gaussian distribution, controls the weight decay speed, takes W / 5, W is the width of the mold platform; establish the mold platform surface roughness R according to the standard roughness sample a and local roughness R area Mapping relationship,

[0058] R a =aR area +b(8)

[0059] The coefficients a and b are fitted by the least square method; the surface roughness R of the measurement area is obtained by formula (8) a When the surface roughness of the measurement area is greater than the roughness threshold z, the formwork is retracted and the scraper tilt angle is adjusted to make the scraper fit more closely to the formwork surface. At the same time, the scraper shear force is increased to scrape off the concrete residue in the measurement area.

[0060] S2, automatic oil spraying of the mold table: the mold table is moved to the bottom side mold arrangement station. During the movement, an automatic oil sprayer sprays a layer of oil film evenly on the mold table surface. The automatic oil sprayer has a spraying width of 4m. The mold table movement and the start and stop of the oil sprayer are linked and controlled. Multiple groups of nozzles are configured to achieve spraying of the entire mold table.

[0061] S3. Formwork installation. The formwork is assembled in blocks and is divided into upper side formwork, lower side formwork, left side formwork, right side formwork, and center formwork. The side formwork adopts a standard modular combination design of customized steel components and bridge decks. By modularizing component sizes and interface parameters, a reusable industrialized component system is formed to adapt to the standardized and large-scale production of various types of bridge decks. The side formwork adopts an adaptive stepless adjustable system to meet the production of different stepless side panel sizes. The formwork is assembled on the ground, sprayed with a release agent, and installed as a whole on the formwork platform. The steel cage, shear grooves, and embedded parts are placed. The formwork is connected with bolts. After installation, the formwork platform is moved to the pouring and vibrating station. The formwork installation must meet the requirements of Table 1.

[0062] Table 1 Template installation measured items

[0063]

[0064] S4. Bridge deck pouring and vibration. Concrete is centrally mixed at a mixing station using a roller feeder with an additional lifting function. The feeder has a feeding speed of 0 to 60 m / min and can be frequency-controlled. Concrete is transported to the production workshop and poured by a concrete placing machine. The concrete placing machine consists of a placing machine frame, a placing hopper, a traveling trolley, a cleaning platform, a traveling car, a hydraulic system, safety protection devices, and an electronic control system. The traveling speed of the trolley is 0 to 30 m / min, the traveling speed of the car is 0 to 10 m / min, and the unloading speed is 0.5 to 1.5 m. 3 / min; the concrete placing boom and the bottom vibrating table cooperate to complete the forming, pouring and vibrating of the bridge deck. Then, an industrial camera is used to capture images of the bridge deck surface, and image recognition is used to determine whether there is obvious laitance or air holes on the bridge deck surface. If not, the formwork table is moved to the leveling station, otherwise, vibration is continued. The concrete mix ratio meets the requirements of Table 2.

[0065] Table 2 Concrete mix ratio

[0066]

[0067] The image recognition adopts a deep learning-based approach, including taking images of the surface of a concrete bridge deck just after vibration from a fixed angle on a formwork platform under different lighting conditions: at night with lights on, on cloudy days with lights on, and during the day with lights off. Irregular flaky areas with whitish colors are marked as slurry defects, and circular / elliptical holes are marked as air pore defects. The images are resized to 640×640 pixels and the pixel values ​​are normalized to the range of [0,1]. A random combination enhancement strategy is applied to perform data enhancement through geometric transformation, color perturbation, and noise injection. A lightweight target detection model YOLOv7-tiny is used and pre-trained on a general dataset to learn general object features. The parameters of the first 50% of the backbone network layers are frozen, and the parameters of the last 50% of the layers and the detection head are fine-tuned. The AdamW optimizer is used as the optimizer, and a composite loss function is used, in which the localization loss is CIoULoss and the classification loss is FocalLoss. Finally, 5-fold cross-validation is used to ensure the generalization ability of the model.

[0068] S5: Bridge deck leveling: Lower the leveling machine to the top of the alignment template, turn on the vibrator and rubbing plate, and then slowly and evenly move the vibration leveling machine to level and smooth the top surface of the bridge deck. Use image recognition to determine whether the bridge deck surface has a smooth mirror effect. If so, move the formwork table to a static position. Otherwise, the defects caused by leveling will be filled with a spreading machine and then leveled.

[0069] The image recognition is based on image data processing, including quantitative evaluation through grayscale variance and gradient mutation. An industrial-grade RGB camera is used with a ring-shaped LED uniform light source to capture the surface image of the bridge deck after leveling. The RGB image is converted into a grayscale image and the image is divided into sub-areas with a resolution of w×h. The grayscale value standard deviation σ is calculated for each sub-area. g ,

[0070]

[0071] Where, N′ is the total number of pixels in the sub-region, g i - Gray value of a single pixel in the sub-region, μ g -The average gray value of the sub-region, the Sobel operator is used to calculate the gradient amplitude in the sub-region,

[0072]

[0073] Where R(x,y)-the gradient amplitude at the (x,y) coordinate in the sub-region, *-convolution operation, I′-the image matrix of the sub-region; the gradient amplitude is binarized,

[0074]

[0075] Defect area ratio R t for

[0076]

[0077] Set the judgment threshold σ max and R max , when σ g Greater than σ max or R t Greater than R max When , it means that the bridge deck surface does not achieve a smooth mirror effect; among them, σ max Determined based on multiple groups of bridge deck specimens, R max Based on the ratio of the acceptable defect area to the actual block area of ​​the bridge deck represented by the sub-area;

[0078] S6, curing and demoulding. After the bridge deck is left to stand and reaches the loose mold and roughening state, the bridge deck is roughened with a roughening machine. After the roughening of the bridge deck surface is completed, it enters the curing pit for curing. The curing station is arranged in two rows of 20 stations. The side insulation board is made of polyurethane material with a thickness of 50mm. The front and rear lifting doors can automatically sense the mold platform in and out, control the opening and closing of the door, fully reduce the heat loss in the kiln, and save energy efficiently. The bridge deck reaches the demoulding state in 10 hours through automatic temperature and humidity control. After demoulding, the bridge deck is transported by a flatbed transport vehicle. Transport to the storage area, sprinkle water to keep moist for no less than 7 days, and the storage time (including storage time in the civil engineering yard) shall not be less than 180 days; 4 0.5×0.2×0.1m wooden strips are used between the bridge panels, and the wooden strips must be treated with anti-corrosion, and the upper and lower wooden strips must be kept in the same vertical line. When the storage height exceeds 8 layers, it must be subject to design verification; the bridge panels are sprayed for maintenance in the storage area, and the maintenance time is no less than 28 days. The bridge panels put into storage should be printed with the production date and plate number as required; after the bridge panels are demoulded, the mold platform is transferred to the lifting station and then transported to the cleaning station.

[0079] Example 2, an automated construction system for steel-concrete composite beam bridge decks based on formwork circulation, including a formwork cleaning module, a formwork oiling module, a material distribution module, a leveling module, a maintenance module, a visual inspection and analysis module, a motion control module, a data integration module, and an exception handling module;

[0080] The die table cleaning module is used to adjust the scraper shear force and angle according to the surface roughness data fed back by the visual inspection and analysis module, and integrates the roller brush speed control system and the residue recovery device;

[0081] The die table oil spraying module is used to spray oil on the die table surface according to the pressure feedback of the oil spraying station and the oil film thickness parameters;

[0082] The distribution module is used to control the concrete conveyor to transport concrete from the mixing plant to the production line. The distribution machine's mixing shaft and worm extrusion and discharging method are controlled by a wireless portable remote controller. The built-in PLC control chip automatically sets the travel trajectory and distribution speed according to the component's external dimensions and thickness parameters. It is also used to adjust the excitation force and amplitude of multiple independent vibrating tables.

[0083] The leveling module is used to control the leveling machine to level and smooth the entire top surface of the bridge deck, and to level and smooth local surfaces according to the relative coordinate positions of the bridge deck;

[0084] The curing module is used for automatic sensing of the entry and exit of the formwork platform, as well as temperature and humidity control of the curing kiln during the curing of the bridge deck;

[0085] The visual inspection and analysis module is used to perform image recognition through an improved Scharr operator gradient analysis algorithm, a YOLOv7-tiny defect detection model, and a grayscale variance dynamic evaluation program;

[0086] The motion control module is used to control the movement and transfer of the mold platform on the assembly line;

[0087] The data integration module is used for data collection and storage during the construction process;

[0088] The exception handling module is used to record faults during the construction process and alert maintenance personnel.

[0089] In summary, the automated construction system and method of steel-concrete composite beam bridge panels based on formwork circulation of the present invention has the characteristics of significantly improving the stability of demolding quality, the generalization ability of defect identification and the standardization level of construction technology in the field of prefabricated bridge panel construction technology.

[0090] It should be understood that the above embodiments are one or more embodiments of the present invention, and there are many other embodiments and variations thereof based on the present invention; the variations and modifications made by ordinary technicians in this industry through the present invention without making groundbreaking innovations all fall within the scope of protection of the present invention.

Claims

1. The automated construction system for steel-concrete composite beam bridge decks based on formwork circulation is characterized by: It includes die table cleaning module, die table oil injection module, material distribution module, leveling module, maintenance module, visual inspection and analysis module, motion control module, data integration module and exception handling module; The die table cleaning module is used to adjust the scraper shear force and angle according to the surface roughness data fed back by the visual inspection and analysis module, and integrates the roller brush speed control system and the residue recovery device; The die table oil spraying module is used to spray oil on the die table surface according to the pressure feedback of the oil spraying station and the oil film thickness parameters; The distribution module is used to control the concrete conveyor to transport concrete from the mixing plant to the production line. The distribution machine's mixing shaft and worm extrusion and discharging method are controlled by a wireless portable remote controller. The built-in PLC control chip automatically sets the travel trajectory and distribution speed according to the component's external dimensions and thickness parameters. It is also used to adjust the excitation force and amplitude of multiple independent vibrating tables. The leveling module is used to control the leveling machine to level and smooth the entire top surface of the bridge deck, and to level and smooth local surfaces according to the relative coordinate positions of the bridge deck; The curing module is used for automatic sensing of the entry and exit of the formwork platform, as well as temperature and humidity control of the curing kiln during the curing of the bridge deck; The visual inspection and analysis module is used to perform image recognition through an improved Scharr operator gradient analysis algorithm, a YOLOv7-tiny defect detection model, and a grayscale variance dynamic evaluation program; The motion control module is used to control the movement and transfer of the mold platform on the assembly line; The data integration module is used for data collection and storage during the construction process; The exception handling module is used to record faults during the construction process and alert maintenance personnel.

2. The automated construction method for steel-concrete composite beam bridge deck based on formwork cycle is controlled by the automated construction system for steel-concrete composite beam bridge deck based on formwork cycle according to claim 1, characterized in that: The specific steps include: S1. Automatic formwork cleaning: A tiltable scraper is used to scrape off the concrete residue solidified on the formwork during demoulding. A roller brush is then used to further clean the formwork surface. After cleaning, the formwork surface roughness is measured using machine vision grayscale gradient analysis. When the surface roughness of the measured area is greater than z, the formwork is retracted, the scraper tilt angle is adjusted to make the scraper fit more closely to the formwork surface, and the scraper shear force is increased to scrape off the concrete residue in the measured area. S2, automatic oil spraying of the mold table, move the mold table to the bottom side mold arrangement station, and use the automatic oil sprayer to evenly spray a layer of oil film on the surface of the mold table during the movement; S3, formwork installation, the formwork is assembled in blocks, the formwork is assembled on the ground, the release agent is sprayed and the whole is installed on the formwork platform, the steel cage, shear groove and embedded parts are placed, and the formwork is connected with bolts. After the installation is completed, the formwork platform is moved to the pouring and vibrating station; S4: The bridge deck is poured and vibrated. The concrete placing boom and the bottom vibration table work together to complete the forming, pouring and vibration of the bridge deck. Then, an industrial camera is used to capture images of the bridge deck surface. Image recognition is used to determine whether there is obvious laitance or air holes on the bridge deck surface. If not, the formwork table is moved to the leveling station. Otherwise, vibration continues. S5: Bridge deck leveling: Lower the leveling machine to the top of the alignment template, turn on the vibrator and rubbing plate, and then slowly and evenly move the vibration leveling machine to level and smooth the top surface of the bridge deck. Use image recognition to determine whether the bridge deck surface has a smooth mirror effect. If so, move the formwork table to a static position. Otherwise, the defects caused by leveling will be filled with a spreading machine and then leveled. S6, maintenance and demoulding. After the bridge deck is left to stand and reaches the loose mold and roughened state, use a roughening machine to roughen the bridge deck, and then enter the curing cellar. Through automatic temperature and humidity control, the bridge deck reaches the demoulding state in 10 hours. After the bridge deck is cured, the mold platform is transferred to the lifting station and transported to the cleaning station.

3. The automated construction method for steel-concrete composite beam bridge deck based on formwork circulation according to claim 2 is characterized in that: In step S1, the machine vision grayscale gradient analysis measures the surface roughness of the mold platform, including using a line array camera and a double-sided LED diffuse light source installed above the mold platform, with the angle θ between the camera axis and the normal line of the mold platform plane being no less than 15°; establishing a mapping relationship between the pixel coordinate system and the physical coordinate system, Where, (u,v)-pixel coordinates, (x,y)-physical coordinates, k x ,k y - Calibration coefficients are obtained using a standard checkerboard calibration plate, with the x-axis along the direction of the stage's motion and the y-axis along the stage's width. The acquisition frequency is determined based on the image width captured by the linear array camera and the stage's movement speed. The captured stage image is preprocessed using a non-local means denoising algorithm and contrast-constrained adaptive histogram equalization. The image matrix I obtained after preprocessing is used to calculate the gradient amplitude using the improved Scharr operator. Where, The gradient amplitude at the coordinate, *-convolution operation; the image is divided into N small windows along the width of the template, and the gradient amplitude of the i-th pixel in the small window is Calculate the gradient statistic R for each small window local , Where n is the number of pixels in the small window, and the local roughness R is fused by Gaussian weighted fusion. area , Where w k - Small window weight, y k -The center position of the kth small window in the width direction of the mold platform, μ-the geometric center position in the width direction of the mold platform, σ-the standard deviation of the Gaussian distribution, controls the weight decay speed, takes W / 5, W is the width of the mold platform; establish the mold platform surface roughness R according to the standard roughness sample a and local roughness R area Mapping relationship, R a =aR area +b (8) The coefficients a and b are fitted by the least squares method.

4. The automated construction method for steel-concrete composite beam bridge deck based on formwork circulation according to claim 2 is characterized in that: In step S4, the image recognition adopts a deep learning-based approach, including taking images of the surface of the concrete bridge deck just after vibration from a fixed angle on the formwork under different lighting conditions: lights on at night, lights on on cloudy days, and lights off during the day, marking irregular flaky areas with whitish colors as laitance defects, and marking circular / elliptical holes as air pore defects; unifying the image size to 640×640 pixels and normalizing the pixel values ​​to the range of [0,1]; applying a random combination enhancement strategy to perform data enhancement through geometric transformation, color perturbation, and noise injection; using a lightweight target detection model YOLOv7-tiny, pre-training on a general dataset to learn general object features; freezing the parameters of the first 50% layers of the backbone network, and fine-tuning the parameters of the 50% layers and the detection head; using the AdamW optimizer as the optimizer, and using a composite loss function, where the positioning loss uses CIoU Loss and the classification loss uses Focal Loss; finally, using 5-fold cross-validation to ensure the generalization ability of the model.

5. The automated construction method for steel-concrete composite beam bridge deck based on formwork circulation according to claim 2 is characterized in that: In step S5, the image recognition is based on image data processing, including quantitative evaluation through grayscale variance and gradient mutation. An industrial-grade RGB camera is used with a ring-shaped LED uniform light source to capture the surface image of the bridge deck after leveling. The RGB image is converted into a grayscale image and the image is divided into sub-areas with a resolution of w×h. The grayscale value standard deviation σ is calculated for each sub-area. g , Where, N′ is the total number of pixels in the sub-region, g i - Gray value of a single pixel in the sub-region, μ g -The average gray value of the sub-region, the Sobel operator is used to calculate the gradient amplitude in the sub-region, Where R(x,y)-the gradient amplitude at the (x,y) coordinate in the sub-region, *-convolution operation, I′-the image matrix of the sub-region; the gradient amplitude is binarized, Defect area ratio R t for Set the judgment threshold σ max and R max , when σ g Greater than σ max or R t Greater than R max When , it means that the bridge deck surface does not achieve a smooth mirror effect; among them, σ max Determined based on multiple groups of bridge deck specimens, R max Based on the ratio of the acceptable defect area to the actual block area of ​​the bridge deck represented by the sub-area.

Images