A roof construction monitoring method combining three-dimensional laser scanning and unmanned aerial vehicle aerial surveying

Abstract

The present application relates to the technical field of building structure monitoring, and more particularly to a roof construction monitoring method based on three-dimensional laser scanning and unmanned aerial vehicle aerial survey, comprising: collecting initial state data sets; collecting roof structures of each construction stage at corresponding collection nodes of each construction stage; comparing the pre-processed monitoring data sets with the reference data; comparing the comprehensive deformation of each target in each construction stage with the preset deformation threshold of the target; adjusting the construction equipment according to the alarm data, and after adjustment, analyzing the re-collected monitoring data and the preset deformation threshold corresponding to the abnormal target until the monitoring data corresponding to the abnormal target is within the range of the preset deformation threshold corresponding to the abnormal target; and archiving the monitoring data archives corresponding to each construction stage. The present application can quantitatively compare and timely correct and adjust the deformation of the roof structure during the construction stage, thereby improving the accuracy of roof construction monitoring.

Description

Technical Field

[0001] This invention relates to the field of building structure monitoring technology, and in particular to a method for monitoring roof construction using three-dimensional laser scanning and UAV aerial surveying. Background Technology

[0002] The roof structure is a core load-bearing component of a building project. Its construction process is complex, with varying degrees of structural displacement and deformation occurring at each stage, including assembly, hoisting, and unloading. The deformation directly determines the construction quality and future safety of the roof structure. Traditional construction monitoring methods rely primarily on total stations, levels, or string-type displacement gauges to discretely sample a few characteristic points. This results in low measurement frequency and limitations on the placement of measurement points due to constraints on the work surface, making it difficult to capture the overall deformation field of the roof surface in three-dimensional space. While finite element numerical simulation can provide theoretical deformation predictions, existing calculation models are often based on idealized boundary conditions and material constitutive models, which deviate from the nonlinear effects of semi-rigid node connections, temporary constraint stiffness degradation, and foundation settlement under actual construction conditions.

[0003] Existing monitoring technologies mostly focus on structural inspection after construction is completed. The monitoring coverage for the dynamic process of construction is limited, and data information can only be collected from local discrete points. It is impossible to form a complete structural deformation data system. Therefore, the existing monitoring methods are difficult to standardize and quantify the deformation data of the entire roof construction process for comparative analysis. They cannot capture small structural deformations and deformation development trends in a timely manner. Problems such as structural offsets and deformation exceeding the standard during construction cannot be detected and corrected in real time. Small deformation errors will continue to accumulate as construction procedures progress, which can easily cause the overall deformation of the roof structure to exceed the allowable limit of the standard. This results in a significant lag in roof construction deformation control, making it impossible to achieve accurate control throughout the entire process and greatly reducing the accuracy and reliability of roof construction monitoring.

[0004] Chinese Patent Publication No. CN118548816A discloses a monitoring system, method, electronic device, and storage medium for tubular truss roof structures. The system includes a stress-strain monitoring module and a visual monitoring module. The stress-strain monitoring module monitors the stress and strain of the tubular truss roof structure and transmits the stress-strain data to a data management and application platform. The visual monitoring module monitors the displacement changes of the tubular truss roof structure and transmits the displacement change data to the data management and application platform. The data management and application platform analyzes and judges the stress-strain data and displacement change data. If the data is normal, it displays the data normally in real time; if the data is abnormal, it displays the abnormal data abnormally and determines the location of the abnormality based on the stress-strain data. This solution only monitors the operating status of the tubular truss roof using stress-strain sensors and a visual camera. It cannot quantitatively compare and promptly correct the deformation of the roof structure during the construction phase, making it difficult to control structural deformation exceeding limits in real time during construction and reducing the accuracy of roof construction monitoring. Summary of the Invention

[0005] To address this issue, the present invention provides a method for monitoring roof construction using three-dimensional laser scanning and UAV aerial surveying. This method overcomes the problem in existing technologies where it is impossible to quantitatively compare and promptly correct roof structural deformation during the construction phase, leading to difficulties in real-time control of excessive structural deformation during construction and reducing the accuracy of roof construction monitoring.

[0006] To achieve the above objectives, this invention provides a method for monitoring roof construction using three-dimensional laser scanning and UAV aerial surveying, comprising the following steps: S1. The initial state of the roof structure is collected to obtain the initial state dataset. The initial state dataset is preprocessed to obtain the preprocessed initial state dataset. The coordinate data of each target is monitored by a total station to verify the preprocessed initial state dataset. The data that passes the preset verification conditions is determined as the reference data of the roof structure. S2. At the data collection nodes corresponding to each construction stage, the roof structure of each construction stage is collected by a 3D laser scanner, a vision camera and a drone to obtain the monitoring data corresponding to each construction stage. S3. Preprocess the monitoring data corresponding to each construction stage to obtain the preprocessed monitoring dataset. Compare and analyze the monitoring data corresponding to each construction stage in the preprocessed monitoring dataset with the benchmark data to obtain the comprehensive deformation of each target in each construction stage. S4. Compare the comprehensive deformation of each target in each construction stage with the preset deformation threshold of the target, and identify the target whose comprehensive deformation exceeds the preset deformation threshold as an abnormal target, and generate alarm data based on the abnormal target. S5. Determine the location of the abnormal roof structure where the abnormal target is located based on the alarm data. Adjust the construction equipment according to the construction parameter adjustment instructions corresponding to the abnormal roof structure location. After adjustment, re-collect the monitoring data corresponding to the abnormal target and analyze the re-collected monitoring data with the preset deformation threshold corresponding to the abnormal target until the monitoring data corresponding to the abnormal target is within the range of the preset deformation threshold corresponding to the abnormal target. S6. Archive the monitoring data, alarm data, and construction parameter adjustment records corresponding to each construction stage to obtain the monitoring data archive corresponding to each construction stage.

[0007] The technical principle of this application is as follows: A 3D laser scanner is used to acquire high-precision point cloud data of the roof structure to obtain 3D spatial coordinate information; a vision camera is used to acquire surface texture and morphological information of the roof structure; and an unmanned aerial vehicle (UAV) is used for aerial surveying of the roof structure. The acquired multi-source data is fused to form monitoring data corresponding to each construction stage. The initial state of the roof structure is collected to obtain an initial state dataset. This initial state dataset is preprocessed to obtain a preprocessed initial state dataset. The coordinate data of each target is monitored using a total station to verify the preprocessed initial state dataset. Data that passes the preset verification criteria is determined as the baseline data for the roof structure. Data is collected at the acquisition nodes corresponding to each construction stage to obtain monitoring data for that stage. This monitoring data is then preprocessed to obtain a preprocessed monitoring dataset. The preprocessed monitoring dataset is used to compare and analyze the coordinates of the monitoring data corresponding to each construction stage with the baseline data, and to calculate the comprehensive deformation of each target in three-dimensional space. The comprehensive deformation of each target in each construction stage is compared with the preset deformation threshold of the target. Targets with comprehensive deformation exceeding the preset deformation threshold are identified as abnormal targets and alarm data is generated. The location of the abnormal roof structure where the abnormal target is located is determined based on the alarm data. The construction parameter adjustment command corresponding to the location is extracted and sent to the construction equipment for adjustment. After the adjustment is completed, the abnormal target data is collected again and analyzed and compared with the preset deformation threshold. The adjustment and collection comparison process is repeated until the monitoring data is within the threshold range. Finally, the monitoring data, alarm data and construction parameter adjustment records of each construction stage are archived in a unified manner to obtain the monitoring data archive corresponding to each construction stage.

[0008] Compared with existing technologies, the advantages of this application are as follows: High-precision point cloud acquisition of the roof structure is achieved through a 3D laser scanner, combined with multi-view and multi-dimensional construction stage monitoring data obtained from a visual camera and UAV aerial surveying. This provides a more comprehensive coverage of the geometric morphology information of each area of ​​the roof structure. By comparing and analyzing the monitoring data of each construction stage in the preprocessed monitoring dataset with the benchmark data calibrated by a total station, the comprehensive deformation of each target is obtained. This allows for the quantitative characterization of structural deformation at each stage of construction and precise comparison with the initial state. By comparing the comprehensive deformation with a preset deformation threshold, abnormal targets are identified and alarm data is generated. The construction equipment is adjusted according to the construction parameter adjustment instructions corresponding to the abnormal roof structure location of the abnormal target, and monitoring data is re-acquired after adjustment until the standard is met. This allows for immediate correction of structural deformation exceeding limits during construction, thereby improving the accuracy of roof construction monitoring.

[0009] Furthermore, S1 includes the following steps: S11. After the roof structure is assembled and before it is lifted, set up several targets in the preset area of ​​the roof structure and use the set targets as spatial positioning reference points for the 3D laser scanner, vision camera and UAV. S12. The roof structure is scanned in its entirety by a 3D laser scanner to obtain an initial point cloud model. Target image data is collected by a vision camera. A reference coordinate system is constructed based on the target image data, and the initial coordinates of each target are recorded. The roof structure is scanned at high altitude by a drone to obtain an initial 3D model. The initial point cloud model, the reference coordinate system, and the initial 3D model are then stitched together to obtain an initial state dataset. S13. Perform denoising filtering on the initial state dataset to obtain denoised and filtered data. Perform multi-site cloud registration on the denoised and filtered data to obtain registered data. Then, stitch the registered data together to obtain the preprocessed initial state dataset. S14. Measure each target using a total station to obtain the total station measured coordinate set. Perform deviation analysis between each measured coordinate in the total station measured coordinate set and the corresponding target in the preprocessed initial state data set. Determine the preprocessed data with deviation values ​​within the preset deviation value range as the reference data for the roof structure.

[0010] In this solution, a unified target is deployed to build a shared positioning benchmark for multiple devices, enabling synchronous acquisition and fusion of multi-source data from 3D laser scanning visual imaging drone aerial surveys. Multi-level data preprocessing is used to eliminate noise and complete accurate registration of the site cloud to optimize the quality of the original data. Qualified benchmark data is screened by comparing and verifying the target coordinates measured by a total station to ensure the spatial accuracy of the initial state data of the roof.

[0011] Furthermore, S2 includes the following steps: S21. At the data collection nodes corresponding to each construction stage, the roof structure of each construction stage is scanned by a 3D laser scanner to obtain the stage point cloud data corresponding to each construction stage. The targets of each construction stage are collected by a vision camera to obtain the stage target image data corresponding to each construction stage. The roof structure of each construction stage is scanned at high altitude by a UAV to obtain the stage aerial survey data corresponding to each construction stage. S22. Time-align the stage point cloud data, the stage target image data, and the stage aerial survey data to obtain the monitoring data corresponding to each construction stage.

[0012] In this solution, information on different dimensions of the roof structure is collected synchronously from multiple devices in stages, and the timing of data collection is uniformly calibrated to enrich the monitoring materials throughout the construction process. This eliminates the deviation caused by the timing misalignment of different data sources, fully restores the true state of the structure at each stage, and improves the completeness of construction deformation monitoring and the accuracy of data matching.

[0013] Furthermore, S3 includes the following steps: S31. Perform noise reduction and filtering on the monitoring data corresponding to each construction stage to obtain the preprocessed monitoring dataset. Extract feature points from the preprocessed monitoring dataset to obtain the feature coordinate data corresponding to each target in each construction stage. S32. Perform point-by-point difference analysis on the characteristic coordinate data corresponding to each target in each construction stage and the reference coordinate data corresponding to each target in the reference data to obtain the initial deformation components of each target in each construction stage. Perform weighted fusion on the initial deformation components of each target in each construction stage to obtain the fused deformation amount of each target in each construction stage. S33. Multiply the fusion deformation of each target in each construction stage with the preset coordinate correction coefficient to obtain the comprehensive deformation of each target in each construction stage.

[0014] In this scheme, feature coordinate data is extracted by noise reduction filtering, initial deformation components are obtained by point-by-point difference, weighted fusion is used to calculate the fused deformation, and the comprehensive deformation is obtained by multiplying the coordinate correction coefficients, thereby improving the processing accuracy of construction deformation monitoring data.

[0015] Furthermore, in S33, the mathematical expression for the comprehensive deformation of each target in each construction stage is as follows: In the formula, Indicates the first The first construction phase The total deformation of each target This indicates the total number of monitoring data sources. Indicates the first The weighting coefficients corresponding to each monitoring data source Indicates the first The first construction phase The target in the first Initial deformation components from each monitoring data source Indicates the first The first construction phase Preset coordinate correction coefficients for each target.

[0016] In this scheme, the overall deformation is calculated by integrating the initial deformation components from multiple monitoring sources and combining them with coordinate correction coefficients, thereby quantifying the weighted influence of each monitoring data source and improving the accuracy of target deformation monitoring results during the construction phase.

[0017] Furthermore, S4 includes the following steps: S41. Compare the comprehensive deformation of each target in each construction stage with the preset deformation threshold corresponding to the target to obtain the over-limit status identification data of each target. Then, classify and mark the over-limit status identification data of each target to obtain the over-limit status data after classification and marking. S42. In the over-limit status data after grading and marking, the targets whose over-limit level reaches the preset alarm level are identified as abnormal targets. The location information and deformation characteristics of the abnormal targets are correlated to obtain abnormal target correlation data. Based on the abnormal target correlation data, alarm data containing abnormal target number, deformation value and location coordinates is generated.

[0018] In this solution, the over-limit status identification data is obtained by comparing the comprehensive deformation amount with the preset deformation threshold, the over-limit status is marked in a grade, the abnormal target is locked and associated with the location and deformation characteristics, and alarm data containing number and coordinates is generated, so as to realize accurate graded early warning of construction deformation and efficient anomaly identification.

[0019] Furthermore, S5 includes the following steps: S51. Determine the location of the abnormal roof structure where the abnormal target is located based on the location coordinates of the abnormal target in the alarm data, and obtain the current construction parameter data based on the location of the abnormal roof structure. S52. Match and analyze the current construction parameter data with the preset construction parameter adjustment rules to obtain the construction parameter adjustment instructions. Adjust the operating status of the construction equipment according to the construction parameter adjustment instructions to obtain the adjusted construction equipment status data. S53. Under the adjusted construction equipment status, re-collect the monitoring data corresponding to the abnormal target to obtain the re-collected monitoring data. Compare and analyze the re-collected monitoring data with the preset deformation threshold corresponding to the abnormal target until the monitoring data corresponding to the abnormal target is within the range of the preset deformation threshold corresponding to the abnormal target.

[0020] In this solution, the corresponding construction parameters are obtained by locating abnormal targets, instructions are generated to adjust the equipment operation status according to the preset construction parameter adjustment rules, and monitoring data is re-collected and compared with the preset deformation threshold to achieve automatic optimization and adjustment of construction parameters, ensuring that the deformation of the roof structure is within a safe range.

[0021] Furthermore, S6 includes the following steps: S61. Obtain monitoring data, alarm data and construction parameter adjustment records corresponding to each construction stage, perform format unification processing on the monitoring data to obtain unified monitoring data, classify and encode the alarm data and construction parameter adjustment records to obtain classified and coded record data. S62. Associate the monitoring data after format unification with the record data after classification and coding to obtain the associated construction stage dataset. Then, perform time series arrangement on the associated construction stage dataset to obtain the construction stage dataset. S63. Store the construction stage dataset according to the preset archiving rules to obtain the monitoring data archive corresponding to each construction stage.

[0022] This solution unifies the format of various construction monitoring data, classifies and codes the records, links multiple types of data to complete the chronological arrangement, and then archives them in a standardized manner. This organizes scattered construction information, ensures the integrity and traceability of data chronology, simplifies the file retrieval process, and improves the standardization of construction data organization and archiving. Attached Figure Description

[0023] Figure 1 This is a flowchart illustrating a method for monitoring roof construction using three-dimensional laser scanning and UAV aerial surveying, according to an embodiment of the present invention. Detailed Implementation

[0024] The following detailed description illustrates the specific implementation method: like Figure 1 The diagram shown is a flowchart of a roof construction monitoring method using three-dimensional laser scanning and UAV aerial surveying, including the following steps: S1. The initial state of the roof structure is collected to obtain the initial state dataset. The initial state dataset is preprocessed to obtain the preprocessed initial state dataset. The coordinate data of each target is monitored by a total station to verify the preprocessed initial state dataset. The data that passes the preset verification conditions is determined as the reference data of the roof structure. S2. At the data collection nodes corresponding to each construction stage, the roof structure of each construction stage is collected by a 3D laser scanner, a vision camera and a drone to obtain the monitoring data corresponding to each construction stage. S3. Preprocess the monitoring data corresponding to each construction stage to obtain the preprocessed monitoring dataset. Compare and analyze the monitoring data corresponding to each construction stage in the preprocessed monitoring dataset with the benchmark data to obtain the comprehensive deformation of each target in each construction stage. S4. Compare the comprehensive deformation of each target in each construction stage with the preset deformation threshold of the target, and identify the target whose comprehensive deformation exceeds the preset deformation threshold as an abnormal target, and generate alarm data based on the abnormal target. S5. Determine the location of the abnormal roof structure where the abnormal target is located based on the alarm data. Adjust the construction equipment according to the construction parameter adjustment instructions corresponding to the abnormal roof structure location. After adjustment, re-collect the monitoring data corresponding to the abnormal target and analyze the re-collected monitoring data with the preset deformation threshold corresponding to the abnormal target until the monitoring data corresponding to the abnormal target is within the range of the preset deformation threshold corresponding to the abnormal target. S6. Archive the monitoring data, alarm data, and construction parameter adjustment records corresponding to each construction stage to obtain the monitoring data archive corresponding to each construction stage.

[0025] Specifically, S1 includes the following steps: S11. After the roof structure is assembled and before it is lifted, set up several targets in the preset area of ​​the roof structure and use the set targets as spatial positioning reference points for the 3D laser scanner, vision camera and UAV. S12. The roof structure is scanned in its entirety by a 3D laser scanner to obtain an initial point cloud model. Target image data is collected by a vision camera. A reference coordinate system is constructed based on the target image data, and the initial coordinates of each target are recorded. The roof structure is scanned at high altitude by a drone to obtain an initial 3D model. The initial point cloud model, the reference coordinate system, and the initial 3D model are then stitched together to obtain an initial state dataset. S13. Perform denoising filtering on the initial state dataset to obtain denoised and filtered data. Perform multi-site cloud registration on the denoised and filtered data to obtain registered data. Then, stitch the registered data together to obtain the preprocessed initial state dataset. S14. Measure each target using a total station to obtain the total station measured coordinate set. Perform deviation analysis between each measured coordinate in the total station measured coordinate set and the corresponding target in the preprocessed initial state data set. Determine the preprocessed data with deviation values ​​within the preset deviation value range as the reference data for the roof structure.

[0026] Furthermore, in step S11, after the roof structure is assembled but not yet lifted, several targets are set in a preset area of ​​the roof structure. Active targets are placed at key nodes, namely mid-span, cantilever ends, and lifting points. These active targets are powered by a centralized power supply. Passive targets are placed in general monitoring areas. The surfaces of the passive targets are clean and non-reflective, and they are firmly adhered to prevent detachment. The targets are placed away from stress concentration areas such as welds and bolt connections. The targets are perpendicular to the roof structure surface with a deviation of less than or equal to 5 degrees. The set targets are used as spatial positioning reference points for the 3D laser scanner, vision camera, and UAV. The number of targets depends on the roof structure. The structural area is determined to be no less than 4 per 100 square meters. For example, the target is set to 12. The 3D laser scanner station is set up in a location with a wide field of view and good stability and is fixed with a tripod. The leveling accuracy is less than or equal to 0.1 degrees. The overlap area between stations is greater than or equal to 20%. The vision camera is fixed with a bracket and installed at a height greater than or equal to 10 meters. The lens is aimed at the monitoring area so that the target accounts for more than or equal to 5%. The drone flies along the planned route and the flight altitude avoids the operation range of construction machinery. Outdoor equipment is covered with rain shelters and electrical equipment is grounded. Data transmission cables are laid overhead to avoid being crushed.

[0027] In step S12, a 3D laser scanner performs a full-area scan of the roof structure. The scanner scans the roof structure from multiple angles at various stations to obtain an initial point cloud model. This initial point cloud model contains all 3D coordinate information of the roof structure surface. A vision camera acquires target image data. The vision camera captures images containing the targets in real time and constructs a reference coordinate system based on the target image data. The reference coordinate system establishes a 3D rectangular coordinate system with the target center as the origin, and records the initial coordinates of each target. The initial coordinates of each target are obtained by inverse calculation using the intrinsic parameters of the vision camera combined with the pixel coordinates of the target image. The drone performs a high-altitude scan of the roof structure. The drone takes aerial photos of the roof structure from above along a preset flight path to obtain an initial 3D model. The initial 3D model contains spatial morphological data of the top and sides of the roof structure. The initial point cloud model, the reference coordinate system, and the initial 3D model are stitched and fused together. The stitching and fusion uses the reference coordinate system as a unified reference frame to align and merge the coordinates of the initial point cloud model and the initial 3D model, thereby obtaining the initial state dataset. The fusion weight coefficient is determined according to the accuracy allocation of each data source. For example, the weight of the 3D laser scanner data is set to 0.6 and the weight of the drone data is set to 0.4.

[0028] In step S13, the initial state dataset is denoised using a statistical outlier removal method. This method removes noise points in the initial state dataset that deviate from the main point cloud by a set distance. The set distance is determined based on twice the average spacing of the point clouds, for example, 15 millimeters. Simultaneously, isolated noise points caused by reflections or occlusions are removed from the initial state dataset, resulting in denoised and filtered data. Multi-site cloud registration is then performed on the denoised and filtered data. Multi-site cloud registration uses an iterative nearest-point algorithm with target coordinates as constraints to register different... The point clouds acquired by the stations are aligned using rigid body transformation. The iterative convergence threshold is determined based on the registration accuracy requirements. For example, the iterative convergence threshold is set to 0.5 mm to obtain the registered data. In the registered data, the point clouds of each station coincide in a unified coordinate system. The registered data is then stitched together. The stitching process averages and merges the repeated points in the overlapping areas of the registered data to eliminate seam traces, thus obtaining the preprocessed initial state dataset. The stitching overlap rate is determined based on the proportion of overlapping areas of adjacent point clouds. For example, the stitching overlap rate is set to be no less than 30%.

[0029] In step S14, a total station is used to measure each target. The total station is set up sequentially at a stable station and accurately aimed at the center of each target to observe the angle and distance, obtaining a total station measured coordinate set. The total station measured coordinate set contains the three-dimensional coordinate values ​​of each target in the total station coordinate system. Deviation analysis is performed on each measured coordinate in the total station measured coordinate set and the corresponding target in the preprocessed initial state data set. The deviation analysis calculates the Euclidean distance between each measured coordinate and the corresponding target coordinate in the preprocessed initial state data set as the deviation value. The deviation value reflects the degree of deviation of the fusion data of the 3D laser scanner and the UAV from the high-precision measurement of the total station. The preset verification qualified range is determined according to the engineering measurement accuracy level requirements. For example, the preset verification qualified range is set to a deviation value less than or equal to 3 mm, and the preprocessed data with deviation values ​​within the preset verification qualified range is determined as the reference data of the roof structure. If the deviation value of a target exceeds the preset verification qualified range, the data of the surrounding area of ​​the target is re-registered and corrected, and the deviation analysis is performed again until the requirements are met.

[0030] Specifically, S2 includes the following steps: S21. At the data collection nodes corresponding to each construction stage, the roof structure of each construction stage is scanned by a 3D laser scanner to obtain the stage point cloud data corresponding to each construction stage. The targets of each construction stage are collected by a vision camera to obtain the stage target image data corresponding to each construction stage. The roof structure of each construction stage is scanned at high altitude by a UAV to obtain the stage aerial survey data corresponding to each construction stage. S22. Time-align the stage point cloud data, the stage target image data, and the stage aerial survey data to obtain the monitoring data corresponding to each construction stage.

[0031] Furthermore, in step S21, at the corresponding data acquisition nodes for each construction stage, covering four key nodes—before lifting, after lifting 30cm, after lifting to the design elevation, and after sliding into position—a 3D laser scanner is used to scan the roof structure at each construction stage. Initial baseline data is acquired before lifting; during the 30cm lift, a simultaneous scan is conducted to obtain the initial deformation data of the structure after it leaves the ground; at the design elevation, the aerial posture of the structure is scanned, with a focus on acquiring data from the lifting points and the mid-span area; and after sliding into position, the final position data of the structure is scanned, resulting in stage point cloud data corresponding to each construction stage. A vision camera is then used to acquire data from the targets at each construction stage, continuously acquiring data throughout the lifting and sliding process. The sampling rate is determined based on the construction speed. For example, the sampling rate for the lifting stage is set to 0.1Hz and the sampling rate for the sliding stage is set to 0.5Hz. This obtains the target image data corresponding to each construction stage. The roof structure at each construction stage is scanned from above by a drone, and the aerial image acquisition is completed synchronously at each acquisition node. This yields the aerial survey data corresponding to each construction stage. The stage point cloud data, stage target image data, and stage aerial survey data are simultaneously stored on the local hard drive and simultaneously uploaded to the server for dual backup. The triggering time of the acquisition node is determined based on the completion time of the key construction procedures of the roof structure. For example, the acquisition node is set to start acquisition immediately after each roof displacement adjustment action is completed.

[0032] In step S22, the stage point cloud data, stage target image data, and stage aerial survey data are time-aligned. The allowable time deviation threshold for time alignment is determined based on the dynamic deformation rate of construction. For example, the allowable time deviation threshold is set to 2 seconds. Using the unified time synchronization as a reference, the acquisition timestamps of the stage point cloud data, stage target image data, and stage aerial survey data are matched and calibrated one by one. Invalid data with time deviations exceeding the allowable time deviation threshold are removed, and valid data with perfectly matched timestamps are retained to obtain the monitoring data corresponding to each construction stage.

[0033] Specifically, S3 includes the following steps: S31. Perform noise reduction and filtering on the monitoring data corresponding to each construction stage to obtain the preprocessed monitoring dataset. Extract feature points from the preprocessed monitoring dataset to obtain the feature coordinate data corresponding to each target in each construction stage. S32. Perform point-by-point difference analysis on the characteristic coordinate data corresponding to each target in each construction stage and the reference coordinate data corresponding to each target in the reference data to obtain the initial deformation components of each target in each construction stage. Perform weighted fusion on the initial deformation components of each target in each construction stage to obtain the fused deformation amount of each target in each construction stage. S33. Multiply the fusion deformation of each target in each construction stage with the preset coordinate correction coefficient to obtain the comprehensive deformation of each target in each construction stage.

[0034] Specifically, in S33, the mathematical expression for the comprehensive deformation of each target in each construction stage is as follows: In the formula, Indicates the first The first construction phase The total deformation of each target This indicates the total number of monitoring data sources. Indicates the first The weighting coefficients corresponding to each monitoring data source Indicates the first The first construction phase The target in the first Initial deformation components from each monitoring data source Indicates the first The first construction phase Preset coordinate correction coefficients for each target.

[0035] Further, in step S31, the monitoring data corresponding to each construction stage is subjected to denoising and filtering. First, a three-dimensional point cloud processing method is used to denoise the three-dimensional point cloud data in the monitoring data corresponding to each construction stage using point cloud processing software, removing redundant data and outliers. At the same time, registration and stitching operations are used to spatially align and integrate the monitoring data corresponding to different construction stages to obtain a preprocessed monitoring dataset. The denoising and filtering process uses a Gaussian filtering algorithm to smooth and suppress noise signals in the monitoring data corresponding to each construction stage. The convolution kernel size of the Gaussian filtering algorithm is determined according to the spatial resolution of the monitoring data. For example, the convolution kernel size of the Gaussian filtering algorithm is set to 5x5. Then, feature points are extracted from the preprocessed monitoring dataset. The KLT algorithm is used to track feature points in the visual image data corresponding to each target in the preprocessed monitoring dataset, calculate the pixel displacement of the feature points corresponding to each target, and convert the pixel displacement into the real coordinate displacement through the conversion coefficient between pixel displacement and real displacement, finally obtaining the feature coordinate data corresponding to each target in each construction stage. The core principle of the KLT (Kanade-Lucas-Tomasi Tracking) algorithm is to examine the motion of pixels in a small neighborhood of an image and use the optical flow constraint equation to estimate the motion vector of the feature points in the image sequence between adjacent frames, thereby realizing frame-by-frame tracking of the target feature points and calculation of pixel displacement.

[0036] In step S32, a point-by-point difference analysis is performed between the feature coordinate data corresponding to each target in each construction stage and the reference coordinate data corresponding to each target in the reference data. Specifically, the difference calculation is performed on the three-dimensional coordinate values ​​of the feature coordinate data corresponding to each target in each construction stage and the three-dimensional coordinate values ​​of the reference coordinate data corresponding to each target in the reference data along each coordinate axis to obtain the initial deformation components of each target in each construction stage. Simultaneously, combined with UAV data processing methods, dense point clouds are generated from the UAV image data of each construction stage using the SFM and MVS algorithms. The dense point cloud data generated by the UAV is then fused with ground scan data to supplement the high-altitude field-of-view data. The fused data is used as the first... Multiple monitoring data sources are used to calculate the initial deformation components. The initial deformation components of each target in each construction stage are weighted and fused. A weighted average method is used to fuse the initial deformation components of each target in each construction stage from multiple monitoring data sources. The weighting coefficients are determined according to the accuracy level of each monitoring data source. For example, the weighting coefficient for total station monitoring data is set to 0.5, the weighting coefficient for laser scanning monitoring data is set to 0.3, and the weighting coefficient for visual monitoring data is set to 0.2. Finally, the fused deformation amount of each target in each construction stage is obtained. The SFM (Structure from Motion) algorithm is a photogrammetric technique that recovers the 3D scene structure and camera pose by analyzing disordered image sequences. This photogrammetric technique relies on a pinhole camera model and achieves 3D reconstruction through steps such as feature extraction, feature matching, camera parameter estimation, and bundle adjustment. MVS (Multi-View Stereo) is a technique that infers the depth information of each pixel from multiple images acquired from different perspectives to construct a highly detailed 3D model. Based on the camera parameters output by the SFM algorithm, it reconstructs dense point clouds through depth assumptions and cost calculations.

[0037] In step S33, the fused deformation of each target in each construction stage is multiplied by a preset coordinate correction coefficient to obtain the comprehensive deformation of each target in each construction stage. The preset coordinate correction coefficient is determined based on the degree of influence of environmental temperature and humidity changes at the location of each target in each construction stage on the monitoring coordinates. The rule for determining the preset coordinate correction coefficient is to statistically analyze the coordinate drift of each target under different temperature and humidity conditions based on historical monitoring data and take the weighted average as the correction benchmark. For example, the first... The first construction phase The preset coordinate correction coefficient for each target is set to 1.0. The comprehensive deformation of each target in each construction stage is obtained by multiplying the fused deformation of each target in each construction stage by the corresponding preset coordinate correction coefficient. At the same time, the fused deformation result of laser scanning monitoring data and visual monitoring data is corrected and verified based on the total station measured data to ensure the reliability and accuracy of the comprehensive deformation of each target in each construction stage. The comprehensive deformation is finally used to generate deformation cloud maps and deformation curves for each construction stage to reflect the deformation trend during the construction process.

[0038] Specifically, S4 includes the following steps: S41. Compare the comprehensive deformation of each target in each construction stage with the preset deformation threshold corresponding to the target to obtain the over-limit status identification data of each target. Then, classify and mark the over-limit status identification data of each target to obtain the over-limit status data after classification and marking. S42. In the over-limit status data after grading and marking, the targets whose over-limit level reaches the preset alarm level are identified as abnormal targets. The location information and deformation characteristics of the abnormal targets are correlated to obtain abnormal target correlation data. Based on the abnormal target correlation data, alarm data containing abnormal target number, deformation value and location coordinates is generated.

[0039] Further, in step S41, for each target, a preset deformation threshold is determined in advance based on structural design specifications, material properties, and construction conditions. This preset deformation threshold includes a primary warning threshold and a secondary warning threshold. For example, for steel structure lifting construction, the primary warning threshold is determined based on 70% of the design allowable deformation value, and is set to 7 mm. The secondary warning threshold is determined based on 90% of the design allowable deformation value, and is set to 9 mm. The comprehensive deformation of each target is compared with the preset deformation threshold corresponding to that target, i.e., the comprehensive deformation value is read for each target according to the time series and compared with the primary and secondary warning thresholds respectively to obtain the over-limit status identification data of each target. This over-limit status identification data records the over-limit level of the current deformation of each target, including not exceeding the limit, primary over-limit, and secondary over-limit. When the overall deformation is less than the first-level warning threshold, the over-limit status identification data is marked as not exceeding the limit. When the overall deformation is greater than or equal to the first-level warning threshold but less than the second-level warning threshold, it is marked as first-level exceeding the limit. At this time, a prompt message is sent to the construction management personnel in real time to notify them to pay attention to the deformation development trend. When the overall deformation is greater than or equal to the second-level warning threshold, it is marked as second-level exceeding the limit. At this time, the audible and visual alarm device is automatically triggered and the construction operation is suspended. Subsequently, the over-limit status identification data of each target is graded and marked, and a corresponding grade code is assigned according to the over-limit level. For example, not exceeding the limit is coded as 0, first-level exceeding the limit is coded as 1, and second-level exceeding the limit is coded as 2, forming the graded and marked over-limit status data. During the lifting stage, if the height difference deviation of the lifting points exceeds 10 mm, the flow rate of the hydraulic lifting device is adjusted to correct the synchronization of each lifting point. During the sliding stage, if the lateral displacement deviation exceeds 5 mm, the thrust of the hydraulic crawler is adjusted and the sliding trajectory is corrected through the limit device. If the structural deformation approaches the design limit, i.e. the total deformation reaches the level two warning threshold, in addition to suspending construction, temporary support reinforcement and measures to reduce the lifting or sliding speed will be taken. Construction will resume after the deformation drops below the level one warning threshold and is continuously monitored for at least 10 minutes.

[0040] In step S42, based on the over-limit status data after grading and marking, targets whose over-limit levels reach the preset alarm level are identified as abnormal targets. The preset alarm level is determined according to engineering safety control requirements. For example, a level 2 over-limit is set as the preset alarm level, meaning only targets with a level 2 over-limit are considered abnormal targets; level 1 over-limit only serves as a warning and does not trigger an alarm. After identifying the abnormal target, its position information and deformation characteristics are correlated. The position information includes the three-dimensional coordinates of the abnormal target in the overall coordinate system, and the deformation characteristics include the current comprehensive deformation value, deformation direction, and deformation rate of the abnormal target. The correlation process combines the abnormal target's number, current comprehensive deformation value, three-dimensional coordinates, deformation direction, and deformation rate into a structured data record to obtain abnormal target correlation data. Alarm data is generated based on the abnormal target correlation data. This alarm data includes the abnormal target number, deformation value, and position coordinates. The abnormal target number is a unique character sequence identifying the target, the deformation value is the specific millimeter value of the current comprehensive deformation, and the position coordinates are the X, Y, and Z direction values ​​in three-dimensional space. After generating alarm data, the alarm data is pushed to the construction management platform and the on-site monitoring terminal, and the audible and visual alarm is triggered to continue working until the operator confirms. During the lifting phase, if the height difference deviation of the lifting point reflected by the abnormal target exceeds 10 mm, the hydraulic lifter is controlled to adjust the flow rate of each lifting point to restore synchronous lifting. During the sliding phase, if the lateral displacement deviation reflected by the abnormal target exceeds 5 mm, the thrust of the hydraulic crawler is adjusted and mechanical correction is performed using the limit device. If the structural deformation approaches the design limit, that is, the deformation value of the abnormal target reaches the value corresponding to the second-level warning threshold, temporary support columns or diagonal braces are added, the operating speed of the hydraulic lifter or hydraulic crawler is reduced to 50% of the original speed, and the comprehensive deformation of the abnormal target is re-collected every 2 minutes. After the deformation values ​​collected for 3 consecutive times show a downward trend and stabilize below the preset alarm level, the normal construction speed is gradually restored.

[0041] Specifically, S5 includes the following steps: S51. Determine the location of the abnormal roof structure where the abnormal target is located based on the location coordinates of the abnormal target in the alarm data, and obtain the current construction parameter data based on the location of the abnormal roof structure. S52. Match and analyze the current construction parameter data with the preset construction parameter adjustment rules to obtain the construction parameter adjustment instructions. Adjust the operating status of the construction equipment according to the construction parameter adjustment instructions to obtain the adjusted construction equipment status data. S53. Under the adjusted construction equipment status, re-collect the monitoring data corresponding to the abnormal target to obtain the re-collected monitoring data. Compare and analyze the re-collected monitoring data with the preset deformation threshold corresponding to the abnormal target until the monitoring data corresponding to the abnormal target is within the range of the preset deformation threshold corresponding to the abnormal target.

[0042] Further, in step S51, the alarm data includes the location coordinates of the abnormal target, which are the X-axis, Y-axis, and Z-axis values ​​in three-dimensional space. The location coordinates of the abnormal target are matched with the three-dimensional spatial model of the roof structure. By querying the corresponding grid number or member number in the three-dimensional model of the roof structure using the coordinates, the location of the abnormal roof structure is determined. Specifically, the location of the abnormal roof structure is a truss node or a section of the space frame within the roof structure. After determining the location of the abnormal roof structure, the current construction parameters corresponding to that location are acquired. These current construction parameters include the flow rate setting of the hydraulic lifter currently operating at that location, the thrust setting of the hydraulic crawler, the lifting or sliding speed setting, and the jacking force value of the temporary support. The acquisition method involves reading the values ​​in the operating parameter register of the aforementioned equipment at the current moment through the parameter acquisition module in the construction control system. This current construction parameter data is organized in the form of equipment number plus parameter value data pairs.

[0043] In step S52, the preset construction parameter adjustment rules are determined based on the over-limit level of the abnormal target and the construction stage corresponding to the abnormal roof structure location. These preset rules contain multiple parameter adjustment entries, each recording the correspondence between the over-limit level, construction stage, equipment type, adjustment direction, and adjustment range. For example, the preset construction parameter adjustment rules can be set to include the following entries: for the lifting stage with a level 2 over-limit, the hydraulic lifter flow rate setting is reduced by 5%; for the sliding stage with a level 2 over-limit, the hydraulic crawler thrust setting is reduced by 8%; for the lifting stage with a level 1 over-limit, the hydraulic lifter flow rate setting is reduced by 2%. During the matching analysis, the construction stage identifier, the type of equipment currently running, and the over-limit level of the abnormal target are extracted from the current construction parameter data. Parameter adjustment entries matching these three conditions are searched in the preset construction parameter adjustment rules. Based on the adjustment direction and adjustment range in the matched parameter adjustment entries, a construction parameter adjustment instruction for the specific equipment is generated. This instruction includes the target equipment number, the name of the parameter to be adjusted, and the adjusted target value. The operating status of the construction equipment is adjusted according to the construction parameter adjustment instructions. The adjustment method is to send the target value instruction to the controller of the corresponding equipment through the construction control system. The controller executes the instruction to adjust the operating parameters of the equipment to the target value, and obtains the adjusted construction equipment status data. The adjusted construction equipment status data records the equipment number, parameter name and the actual operating value after adjustment.

[0044] In step S53, a new round of data acquisition is performed using the displacement sensor or total station originally positioned at the abnormal target location. The acquisition interval is determined based on the stabilization time after the adjustment action is completed; for example, the acquisition interval is set to wait 30 seconds after the adjustment action is completed before acquiring data. The reacquired monitoring data is compared and analyzed with the preset deformation threshold corresponding to the abnormal target. This preset deformation threshold is the same as the first-level and second-level warning thresholds. During the comparison analysis, it is first determined whether the comprehensive deformation value in the reacquired monitoring data is less than the first-level warning threshold. If the comprehensive deformation value is less than the first-level warning threshold, it is determined that the monitoring data corresponding to the abnormal target is within the range of the preset deformation threshold corresponding to the abnormal target, the adjustment process is stopped, and the current operating state of the construction equipment is maintained. If the comprehensive deformation value is still greater than or equal to the first-level warning threshold, the process returns to step S52 to perform matching analysis again and generate a new construction parameter adjustment command. The operating state of the construction equipment continues to be adjusted until the monitoring data corresponding to the abnormal target is within the range of the preset deformation threshold corresponding to the abnormal target. After each re-collection, if the overall deformation reaches the secondary warning threshold, an audible and visual alarm will be automatically triggered and construction will be suspended. If the overall deformation falls below the primary warning threshold, the construction speed will be gradually restored to 100% of the original speed. During the lifting phase, if the height difference deviation of the lifting points exceeds 10 mm, the flow rate of the hydraulic lifter will be continuously adjusted until the height difference deviation is less than 10 mm. During the sliding phase, if the lateral displacement deviation exceeds 5 mm, the thrust of the hydraulic crawler will be continuously adjusted and the position will be corrected in conjunction with the limit device until the deviation is less than 5 mm. If the structural deformation approaches the design limit, i.e., the overall deformation reaches the secondary warning threshold, temporary support measures will be added, and the operating speed of the hydraulic lifter or hydraulic crawler will be adjusted to 50% of the original speed. After three consecutive re-collected monitoring data are all within the preset deformation threshold range, the operating speed will be gradually restored to the original speed.

[0045] Specifically, S6 includes the following steps: S61. Obtain monitoring data, alarm data and construction parameter adjustment records corresponding to each construction stage, perform format unification processing on the monitoring data to obtain unified monitoring data, classify and encode the alarm data and construction parameter adjustment records to obtain classified and coded record data. S62. Associate the monitoring data after format unification with the record data after classification and coding to obtain the associated construction stage dataset. Then, perform time series arrangement on the associated construction stage dataset to obtain the construction stage dataset. S63. Store the construction stage dataset according to the preset archiving rules to obtain the monitoring data archive corresponding to each construction stage.

[0046] Further, in step S61, when acquiring monitoring data, alarm data, and construction parameter adjustment records corresponding to each construction stage, the original monitoring data, alarm data, and construction parameter adjustment records for each stage are read through the sensor acquisition devices and data recording terminals deployed at each construction stage. When standardizing the monitoring data format, the monitoring data collected by different sensors are mapped and aligned according to a unified data template. The timestamps in the monitoring data are standardized to year, month, day, hour, minute, and second format, and the numerical units in the monitoring data are uniformly converted to international standard units, resulting in standardized monitoring data. When classifying and coding alarm data and construction parameter adjustment records, labels are applied to the alarm data and construction parameter adjustment records according to preset classification and coding rules. The preset classification and coding rules are determined based on the data type and source stage of the alarm data and construction parameter adjustment records. For example, alarm data is divided into displacement exceeding limits, stress anomaly, and seepage alarm categories, and construction parameter adjustment records are divided into support parameter adjustment, excavation rate adjustment, and grouting parameter adjustment categories. Each category is assigned a unique code identifier, resulting in classified and coded record data.

[0047] In step S62, when associating the standardized monitoring data with the categorized and coded record data, the time nodes of each construction stage are used as the matching benchmark. The time information in the standardized monitoring data is bound one-to-one with the time information in the categorized and coded record data, ensuring a correspondence between the standardized monitoring data and the categorized and coded record data within the same construction stage, resulting in the associated construction stage dataset. When arranging the associated construction stage dataset in a time series, the data entries are arranged sequentially from earliest to latest according to the start and end times of each construction stage. The preset time series arrangement rules are determined based on the actual start and completion dates of each construction stage. For example, the construction stages are arranged sequentially as foundation construction stage, main structure construction stage, and decoration construction stage, with an interval of one data node taken every hour, resulting in the construction stage dataset.

[0048] In step S63, when storing the construction phase datasets according to preset archiving rules, the construction phase datasets are categorized and stored in corresponding storage areas based on the number and data type hierarchy of each construction phase. The preset archiving rules are determined based on the data volume and importance level of the construction phase datasets. For example, construction phase datasets with an average daily data volume exceeding 5000 records are stored in a high-performance storage module, and datasets marked as key monitoring nodes are set to dual-backup storage. When generating the monitoring data archives corresponding to each construction phase, an independent archive directory index is established for each construction phase. The archive directory index includes the construction phase name, data start and end time, number of data entries, and a list of associated alarm data and construction parameter adjustment record codes. For example, the monitoring data archive directory for the basic construction phase is set to include all monitoring data from March 1, 2024 to June 30, 2024, a total of 120 days, ultimately resulting in the monitoring data archives corresponding to each construction phase.

[0049] The above are merely embodiments of the present invention. Commonly known structures and characteristics are not described in detail here. Those skilled in the art are aware of all common technical knowledge in the field prior to the application date or priority date, are aware of all existing technologies in that field, and have the ability to apply conventional experimental methods prior to that date. Those skilled in the art can, under the guidance of this application, improve and implement this solution in combination with their own capabilities. Some typical known structures or methods should not be obstacles for those skilled in the art to implement this application. It should be noted that those skilled in the art can make several modifications and improvements without departing from the structure of the present invention. These should also be considered within the scope of protection of the present invention, and will not affect the effectiveness of the implementation of the present invention or the practicality of the patent. The scope of protection claimed in this application should be determined by the content of its claims, and the specific embodiments described in the specification can be used to interpret the content of the claims.

Claims

1. A method for monitoring roof construction using three-dimensional laser scanning and UAV aerial surveying, characterized in that: Includes the following steps: S1. The initial state of the roof structure is collected to obtain the initial state dataset. The initial state dataset is preprocessed to obtain the preprocessed initial state dataset. The coordinate data of each target is monitored by a total station to verify the preprocessed initial state dataset. The data that passes the preset verification conditions is determined as the reference data of the roof structure. S2. At the data collection nodes corresponding to each construction stage, the roof structure of each construction stage is collected by a 3D laser scanner, a vision camera and a drone to obtain the monitoring data corresponding to each construction stage. S3. Preprocess the monitoring data corresponding to each construction stage to obtain the preprocessed monitoring dataset. Compare and analyze the monitoring data corresponding to each construction stage in the preprocessed monitoring dataset with the benchmark data to obtain the comprehensive deformation of each target in each construction stage. S4. Compare the comprehensive deformation of each target in each construction stage with the preset deformation threshold of the target, and identify the target whose comprehensive deformation exceeds the preset deformation threshold as an abnormal target, and generate alarm data based on the abnormal target. S5. Determine the location of the abnormal roof structure where the abnormal target is located based on the alarm data. Adjust the construction equipment according to the construction parameter adjustment instructions corresponding to the abnormal roof structure location. After adjustment, re-collect the monitoring data corresponding to the abnormal target and analyze the re-collected monitoring data with the preset deformation threshold corresponding to the abnormal target until the monitoring data corresponding to the abnormal target is within the range of the preset deformation threshold corresponding to the abnormal target. S6. Archive the monitoring data, alarm data, and construction parameter adjustment records corresponding to each construction stage to obtain the monitoring data archive corresponding to each construction stage.

2. The method for monitoring roof construction using three-dimensional laser scanning and UAV aerial surveying according to claim 1, characterized in that: S1 includes the following steps: S11. After the roof structure is assembled and before it is lifted, set up several targets in the preset area of ​​the roof structure and use the set targets as spatial positioning reference points for the 3D laser scanner, vision camera and UAV. S12. The roof structure is scanned in its entirety by a 3D laser scanner to obtain an initial point cloud model. Target image data is collected by a vision camera. A reference coordinate system is constructed based on the target image data, and the initial coordinates of each target are recorded. The roof structure is scanned at high altitude by a drone to obtain an initial 3D model. The initial point cloud model, the reference coordinate system, and the initial 3D model are then stitched together to obtain an initial state dataset. S13. Perform denoising filtering on the initial state dataset to obtain denoised and filtered data. Perform multi-site cloud registration on the denoised and filtered data to obtain registered data. Then, stitch the registered data together to obtain the preprocessed initial state dataset. S14. Measure each target using a total station to obtain the total station measured coordinate set. Perform deviation analysis between each measured coordinate in the total station measured coordinate set and the corresponding target in the preprocessed initial state data set. Determine the preprocessed data with deviation values ​​within the preset deviation value range as the reference data for the roof structure.

3. The roof construction monitoring method based on three-dimensional laser scanning and UAV aerial surveying according to claim 2, characterized in that: S2 includes the following steps: S21. At the data collection nodes corresponding to each construction stage, the roof structure of each construction stage is scanned by a 3D laser scanner to obtain the stage point cloud data corresponding to each construction stage. The targets of each construction stage are collected by a vision camera to obtain the stage target image data corresponding to each construction stage. The roof structure of each construction stage is scanned at high altitude by a UAV to obtain the stage aerial survey data corresponding to each construction stage. S22. Time-align the stage point cloud data, the stage target image data, and the stage aerial survey data to obtain the monitoring data corresponding to each construction stage.

4. The roof construction monitoring method based on three-dimensional laser scanning and UAV aerial surveying according to claim 1, characterized in that: S3 includes the following steps: S31. Perform noise reduction and filtering on the monitoring data corresponding to each construction stage to obtain the preprocessed monitoring dataset. Extract feature points from the preprocessed monitoring dataset to obtain the feature coordinate data corresponding to each target in each construction stage. S32. Perform point-by-point difference analysis on the characteristic coordinate data corresponding to each target in each construction stage and the reference coordinate data corresponding to each target in the reference data to obtain the initial deformation components of each target in each construction stage. Perform weighted fusion on the initial deformation components of each target in each construction stage to obtain the fused deformation amount of each target in each construction stage. S33. Multiply the fusion deformation of each target in each construction stage with the preset coordinate correction coefficient to obtain the comprehensive deformation of each target in each construction stage.

5. The roof construction monitoring method based on three-dimensional laser scanning and UAV aerial surveying according to claim 4, characterized in that: In S33, the mathematical expression for the comprehensive deformation of each target in each construction stage is: In the formula, Indicates the first The first construction phase The total deformation of each target This indicates the total number of monitoring data sources. Indicates the first The weighting coefficients corresponding to each monitoring data source Indicates the first The first construction phase The target in the first Initial deformation components from each monitoring data source Indicates the first The first construction phase Preset coordinate correction coefficients for each target.

6. The roof construction monitoring method based on three-dimensional laser scanning and UAV aerial surveying according to claim 1, characterized in that: S4 includes the following steps: S41. Compare the comprehensive deformation of each target in each construction stage with the preset deformation threshold corresponding to the target to obtain the over-limit status identification data of each target. Then, classify and mark the over-limit status identification data of each target to obtain the over-limit status data after classification and marking. S42. In the over-limit status data after grading and marking, the targets whose over-limit level reaches the preset alarm level are identified as abnormal targets. The location information and deformation characteristics of the abnormal targets are correlated to obtain abnormal target correlation data. Based on the abnormal target correlation data, alarm data containing abnormal target number, deformation value and location coordinates is generated.

7. The roof construction monitoring method based on three-dimensional laser scanning and UAV aerial surveying according to claim 1, characterized in that: S5 includes the following steps: S51. Determine the location of the abnormal roof structure where the abnormal target is located based on the location coordinates of the abnormal target in the alarm data, and obtain the current construction parameter data based on the location of the abnormal roof structure. S52. Match and analyze the current construction parameter data with the preset construction parameter adjustment rules to obtain the construction parameter adjustment instructions. Adjust the operating status of the construction equipment according to the construction parameter adjustment instructions to obtain the adjusted construction equipment status data. S53. Under the adjusted construction equipment status, re-collect the monitoring data corresponding to the abnormal target to obtain the re-collected monitoring data. Compare and analyze the re-collected monitoring data with the preset deformation threshold corresponding to the abnormal target until the monitoring data corresponding to the abnormal target is within the range of the preset deformation threshold corresponding to the abnormal target.

8. The roof construction monitoring method based on three-dimensional laser scanning and UAV aerial surveying according to claim 1, characterized in that: S6 includes the following steps: S61. Obtain monitoring data, alarm data and construction parameter adjustment records corresponding to each construction stage, perform format unification processing on the monitoring data to obtain unified monitoring data, classify and encode the alarm data and construction parameter adjustment records to obtain classified and coded record data. S62. Associate the monitoring data after format unification with the record data after classification and coding to obtain the associated construction stage dataset. Then, perform time series arrangement on the associated construction stage dataset to obtain the construction stage dataset. S63. Store the construction stage dataset according to the preset archiving rules to obtain the monitoring data archive corresponding to each construction stage.

Images