Engineering deformation monitoring method and system based on big data

Abstract

The application discloses an engineering deformation monitoring method and system based on big data and relates to the technical field of intelligent monitoring, including the following steps: setting a screening standard, screening first data, analyzing to obtain a first deformation rate, executing a first operation, setting a key subspace, performing first regulation and second marking, and second analysis to obtain an influence relationship. Through secondary subdivision of the subspace and sensitivity, the standard deviation of the measured rate is reduced, after introducing a meteorological and time baseline coupling model, a data, model and decision closed loop is realized, operation and maintenance risks are reduced, the method is applicable to three kinds of data sources, namely BIM, point cloud and grid, covers various engineering types such as dams, fills, bridges, landslides and subway tunnels, the meteorological and rock-soil parameter table is loosely coupled with the algorithm module, can be directly popularized to different climate zones and geological conditions by updating the parameter library, and the core code does not need to be modified.

Description

Technical Field

[0001] This invention relates to the technical field of intelligent monitoring, and in particular to a method and system for monitoring engineering deformation based on big data. Background Technology

[0002] In recent years, engineering deformation monitoring has entered an intelligent era characterized by "sub-millimeter precision, second-level response, AI self-diagnosis, and digital twin closed-loop," evolving towards autonomous monitoring driven by "6G integrated sensing, AIGE zero-code, and quantum sensing." The collaborative use of BeiDou-3, InSAR, UAV lasers, distributed fiber optics, and 5G / 6G communication enables second-level acquisition and transmission of millimeter-level or even sub-millimeter-level displacements. Edge AI completes millisecond-level anomaly identification at base stations or airborne terminals, uploading only features to save bandwidth, and providing real-time visualization through UE5 digital twins. This technology is widely applied in typical projects such as high-speed railway bridges, subway tunnels, dams, mines, and ancient buildings, significantly reducing accident rates and maintenance costs.

[0003] Currently, Chinese invention patent CN117029708A discloses a method for monitoring bridge deformation. This method involves determining the marker points, support reference points, and multiple measuring points of the bridge under test to obtain a frontal view and a rotated image of the bridge. Using the frontal view as a reference image and the corresponding points in the overlapping area of ​​the frontal and rotated images as constraints, the rotated image is projected onto the reference image to obtain an equivalent frontal view. The SIFT algorithm is then used to perform feature matching between the reference image and the equivalent frontal view to obtain high-precision corresponding points. Based on these high-precision corresponding points... The precise relative orientation elements of the equivalent orthographic image are obtained by combining point and bundle adjustment methods. The reference image and the equivalent orthographic image are then stitched together based on the precise relative orientation elements to obtain a panoramic image of the bridge under test. The deformation values ​​of multiple measuring points of the bridge under test are obtained based on the panoramic image, marker points, and support reference points to obtain the deformation curve. However, the related technologies do not use deformation degree algorithms to screen parts of the project and conduct further analysis, which is not conducive to the simplicity of data analysis. Furthermore, the monitoring time baseline of the satellite is not adjusted according to influencing factors, which is not conducive to the accuracy and customization of the detection. Summary of the Invention

[0004] The technical problem solved by this invention is that related technologies do not use deformation degree algorithms to screen and further analyze engineering parts, which is not conducive to the simplicity of data analysis, and do not adjust the satellite monitoring time baseline according to influencing factors, which is not conducive to the accuracy and customization of detection.

[0005] To solve the above-mentioned technical problems, the present invention provides the following technical solution: Firstly, a method for monitoring engineering deformation based on big data, comprising the following steps:

[0006] Step S100: Select any engineering subspace, set a filtering standard, filter the first data according to the filtering standard, analyze the first data to obtain the first deformation rate, perform the first operation according to the first deformation rate, and mark the engineering subspace with the first mark.

[0007] Step S200: In response to the first operation, a key subspace is set, the time baseline of the key subspace is first adjusted, and the key subspace is second-marked according to the first deformation rate after the first adjustment.

[0008] Step S300: Perform a second analysis on the relevant data, the first marker, and the second marker to obtain the influence relationship between meteorological data and the time baseline, and the influence relationship between engineering subspace distribution parameters and the time baseline.

[0009] As a preferred embodiment of the big data-based engineering deformation monitoring method of the present invention, the engineering project to be monitored is first segmented before setting the screening criteria. The segmentation includes semantic segmentation, parameter segmentation, obtaining and automatically numbering the engineering subspaces, and outputting the engineering subspaces and their numbers. The number of the engineering subspace is represented as X. i where i is a natural number;

[0010] The semantic segmentation uses open-source tools and a classification manager to segment the 3D model of the project to be monitored. The semantic representation is the name of the constituent parts of the project. Open IFC, click the classification option, click create independent model, and you will get N independent objects.

[0011] The parameter segmentation includes setting the compressible layer thickness abruptly to 2 meters as the first segmentation condition, setting the slope threshold to 10 degrees as the second segmentation condition, and setting the minimum number of points to 1000 as the third segmentation condition, thereby segmenting N independent objects to obtain M independent objects.

[0012] Based on the processing order of the segmentation software, the M independent objects are automatically numbered, and the M independent objects are recorded as a project subspace. The numbers of the M independent objects are recorded as the numbers of the project subspace.

[0013] As a preferred embodiment of the big data-based engineering deformation monitoring method of the present invention, the following is provided: after the engineering to be monitored is divided, any engineering subspace is selected and a screening criterion is set. The screening criterion is that the time baseline is the first day number. The time baseline represents the interval duration of SAR detection. The first day number is obtained by the time baseline of engineering of the same type as the engineering to be monitored. The first day number is represented as the average value of the time baseline of engineering of the same type as the engineering to be monitored.

[0014] The first data is selected according to the selection criteria. The first data is represented as each historical monitoring data corresponding to the periodicity of the first day. The historical monitoring data is represented as SAR image. The SAR image is represented as an image of each pixel point on the surface of the engineering subspace taken by a satellite with a first distance spatial baseline. The transmitted radar wavelength is the first wavelength and the time baseline is the first day. The spatial baseline is represented as the distance between the geometric center points of any two satellites used for taking pictures.

[0015] Extract the deformation phase corresponding to any SAR image, convert the deformation phase into deformation rate according to InSAR technology, traverse the deformation rate at each pixel in the engineering subspace, calculate the average value of the deformation rate at each pixel, and record the average value of the deformation rate at each pixel as the first deformation rate.

[0016] As a preferred embodiment of the big data-based engineering deformation monitoring method of the present invention, wherein: a first operation is performed according to a first deformation rate, the first operation including marking an engineering subspace and setting a key engineering subspace;

[0017] The setup method for the first operation includes:

[0018] Set the first value as the deformation rate threshold, select any engineering subspace, compare the first deformation rate corresponding to the engineering subspace with the first value, when the first deformation rate is less than or equal to the first value, set the first operation to mark the engineering subspace, when the first deformation rate is greater than the first value, set the first operation to set a critical engineering subspace.

[0019] In a preferred embodiment of the big data-based engineering deformation monitoring method of the present invention, when the first operation is to mark the first engineering subspace, the corresponding engineering subspace number is highlighted in red, and the first day number is added after the engineering subspace number. The first mark is represented as X. i On the first day, create the first folder and store the first tag in the first folder;

[0020] When the first operation is to set up a critical engineering subspace, a second folder is created, and the corresponding engineering subspace number and the corresponding second data backup are stored in the second folder. The second data represents the historical monitoring data of each critical engineering subspace. Each engineering subspace in the second folder is set as a critical engineering subspace.

[0021] As a preferred embodiment of the big data-based engineering deformation monitoring method of the present invention, the first change amount is set as the change gradient of the first day number, and the original time baseline of the key engineering subspace is adjusted according to the first change amount, that is, the first adjustment is made to the first day number, the first adjustment is expressed as continuous reduction, and the reduced first day number is continuously set as the new time baseline.

[0022] Select any new time baseline, starting from the starting time point of the historical monitoring data, and select historical monitoring data at each new time baseline, and set the selected historical monitoring data as the data to be analyzed;

[0023] By iterating through each new time baseline, the corresponding data to be analyzed is obtained.

[0024] As a preferred embodiment of the big data-based engineering deformation monitoring method of the present invention, wherein: the first deformation rate corresponding to the new time baseline is continuously acquired in descending order of the new time baseline;

[0025] The continuous reduction of the first day number is stopped until the first deformation rate corresponding to the new time baseline is less than or equal to the first value, and the adjusted time baseline at this time is obtained and recorded as the critical day number.

[0026] A second label is applied to the critical subspace, which represents the critical subspace number and the critical time baseline. The second label is denoted as X. i And the key time baseline.

[0027] As a preferred embodiment of the big data-based engineering deformation monitoring method of the present invention, the relevant data includes meteorological data and engineering subspace distribution parameters;

[0028] The meteorological data include average annual precipitation, average daily wind speed, average annual air humidity, average annual temperature range, and average annual snow depth.

[0029] The engineering subspace distribution parameters include compressible layer thickness, fill thickness, distance from water source, ground slope, average daily building load, groundwater level depth, and vegetation coverage.

[0030] As a preferred embodiment of the big data-based engineering deformation monitoring method of the present invention, the method for performing a second analysis on relevant data, a first marker, and a second marker to obtain the influence relationship of meteorological data on the time baseline and the influence relationship of engineering subspace distribution parameters on the time baseline includes:

[0031] Calculate the average of the first day number corresponding to the first mark and the critical day number corresponding to the second mark, and denote it as the first average. Using the first average as the dependent variable and meteorological data as the independent variable, perform multiple regression analysis to obtain the first multiple regression equation. The first multiple regression equation is denoteed as the influence relationship between meteorological data and the time baseline.

[0032] Calculate the first difference between the first day number corresponding to the first mark and the critical day number corresponding to the second mark. Using the first difference as the dependent variable and the engineering subspace distribution parameter as the independent variable, perform a multiple regression analysis to obtain the second multiple regression equation. The second multiple regression equation is denoted as the influence relationship between the engineering subspace distribution parameter and the time baseline.

[0033] By inputting meteorological data into the influence relationship of meteorological data on the time baseline, the first average value of a new project with the same type of project to be monitored is obtained. Based on the first day and the first average value of the new project, the corresponding critical days are calculated.

[0034] Calculate the first difference between the first day and the critical day corresponding to the new project. By inputting the corresponding first difference into the influence relationship between the project subspace distribution parameters and the time baseline, the time baseline of each project subspace corresponding to the new project is obtained.

[0035] Secondly, the engineering deformation monitoring system based on big data includes a marking module, an analysis module, and a construction module;

[0036] The marking module selects any engineering subspace, sets a filtering standard, filters the first data according to the filtering standard, analyzes the first data to obtain the first deformation rate, performs the first operation according to the first deformation rate, and marks the engineering subspace with the first mark.

[0037] The analysis module responds to the first operation by setting a key subspace, performing a first adjustment on the time baseline of the key subspace, and performing a second marking on the key subspace according to the first deformation rate after the first adjustment.

[0038] The construction module performs a second analysis on the relevant data, the first label, and the second label to obtain the influence relationship of meteorological data on the time baseline and the influence relationship of engineering subspace distribution parameters on the time baseline.

[0039] The beneficial effects of this invention are as follows: By subspace and sensitivity-based secondary partitioning, the deformation gradient within each computational unit is ensured to be less than 3 mm per day, reducing the standard deviation of the measured rate. After introducing a coupled meteorological and temporal baseline model, the atmospheric phase residual is reduced, and the annual rate error is compressed. The output meteorological and temporal baselines, along with the dual models of distribution parameters and temporal baselines, can be directly embedded into a digital twin platform to achieve a data, model, and decision-making closed loop. This provides interpretable sensitivity factor ranking for operation and maintenance units, assisting in the development of targeted reinforcement, drainage, or loading schemes, reducing operation and maintenance risks. The method is applicable to three data sources: BIM, point cloud, and grid, covering various engineering types such as dams, embankments, bridges, landslides, and subway tunnels. The meteorological and geotechnical parameter tables are loosely coupled with the algorithm module, allowing direct extension to different climate zones and geological conditions through parameter library updates without modifying the core code. Attached Figure Description

[0040] Figure 1 This is a schematic diagram of the basic process of an engineering deformation monitoring method based on big data, provided as an embodiment of the present invention. Detailed Implementation

[0041] To make the above-mentioned objects, features and advantages of the present invention more apparent and understandable, the specific embodiments of the present invention will be described in detail below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments.

[0042] Example, refer to Figure 1 As an embodiment of the present invention, an engineering deformation monitoring method based on big data is provided, comprising the following steps:

[0043] Step S100: Select any engineering subspace, set a filtering standard, filter the first data according to the filtering standard, analyze the first data to obtain the first deformation rate, perform the first operation according to the first deformation rate, and mark the engineering subspace with the first mark.

[0044] Step S200: In response to the first operation, a key subspace is set, the time baseline of the key subspace is first adjusted, and the key subspace is second-marked according to the first deformation rate after the first adjustment.

[0045] Step S300: Perform a second analysis on the relevant data, the first marker, and the second marker to obtain the influence relationship between meteorological data and the time baseline, and the influence relationship between engineering subspace distribution parameters and the time baseline.

[0046] This invention employs subspace and sensitivity-based secondary partitioning to ensure that the deformation gradient within each computational unit is less than 3 mm per day, reducing the standard deviation of measured rates. By introducing a coupled meteorological and temporal baseline model, atmospheric phase residuals are reduced, and annual rate errors are compressed. The output meteorological and temporal baselines, along with the distributed parameters and temporal baseline dual model, can be directly embedded into a digital twin platform, achieving a data, model, and decision-making closed loop. This provides operation and maintenance units with interpretable sensitivity factor ranking, assisting in the development of targeted reinforcement, drainage, or loading schemes, reducing operation and maintenance risks. The method is applicable to three data sources: BIM, point cloud, and grid, covering various engineering types such as dams, embankments, bridges, landslides, and subway tunnels. The meteorological and geotechnical parameter tables are loosely coupled with the algorithm module, allowing for direct extension to different climate zones and geological conditions through parameter library updates without modifying the core code.

[0047] Before setting the screening criteria, the project to be monitored is first segmented. Segmentation includes semantic segmentation, parametric segmentation, obtaining project subspaces and automatically numbering them, and outputting the project subspaces and their numbers. The project subspace number is represented as X. i where i is a natural number;

[0048] Semantic segmentation uses open-source tools and a classification manager to segment the 3D model of the project to be monitored. The semantic representation is the name of the components of the project. Open IFC, click the classification option, click Create Independent Model, and you will get N independent objects.

[0049] The parameter segmentation includes setting the compressible layer thickness abruptly to 2 meters as the first segmentation condition, setting the slope threshold to 10 degrees as the second segmentation condition, and setting the minimum number of points to 1000 as the third segmentation condition, thereby segmenting N independent objects to obtain M independent objects.

[0050] Based on the processing order of the segmentation software, the M independent objects are automatically numbered, and the M independent objects are recorded as a project subspace. The numbers of the M independent objects are recorded as the numbers of the project subspace.

[0051] In practical implementation, using open-source BlenderBIM and a classification manager, semantic segmentation is automatically completed in 3 minutes, completely eliminating the need for traditional manual CAD drawing and improving segmentation efficiency by more than 10 times. Parameter segmentation, such as a 2-meter thickness change, a 10-degree slope, and a minimum of 1000 points, is bound to the numbering process, allowing for the output of M project subspaces at once, avoiding subsequent manual matching errors and ensuring 100% traceability of the data link. A compressible layer thickness change of ≥2 meters is used as the first cutting plane, ensuring that the settlement gradient within the same subspace is less than 3 mm per day, reducing the measured rate standard deviation by 35%. The 10-degree slope threshold automatically separates the slope's leading and trailing edges, allowing independent calculation of rate differences between different landslide blocks, avoiding a 5-10 mm error per day caused by mixed calculations. The minimum of 1000 points ensures that each Xi still has greater than or equal to 50 high coherence values ​​after SAR multi-view. The pixels meet the Level 1 accuracy requirements of the "Geological Disaster InSAR Monitoring Specification". The numbering order is consistent with the software processing time. It naturally records semantics, geometry, and parameter versions, which can be directly used as panel data keys for subsequent big data regression. Xi encoding can be seamlessly embedded into blockchain or digital twin platforms to achieve tamper-proof and lifecycle management of subspace-level deformation data. Homogeneous units with pre-cut deformation sensitivity are selected so that subsequent screening criteria can locate anomaly areas in only one run. The amount of computation is reduced from the whole image level to the subspace level, saving 60% of CPU time. Because the thickness, slope, and number of points are controlled, the overall image rejection rate is reduced from 25% to 8%, effectively improving the utilization rate of SAR data. The segmentation conditions (2 meters / 10 degrees / 1000 points) are based on national standards and can be adjusted parametrically with one click. It is suitable for various projects such as dams, embankments, bridges, landslides, and subway foundation pits.

[0052] After the monitoring project is divided, select any project subspace and set the screening criteria. The screening criteria is that the time baseline is the first day. The time baseline represents the interval of SAR detection. The first day is obtained by the time baseline of projects of the same type as the project to be monitored. The first day is represented as the average value of the time baseline of projects of the same type as the project to be monitored.

[0053] The first data is selected according to the selection criteria. The first data is represented as each historical monitoring data corresponding to the periodicity of the first day. The historical monitoring data is represented as SAR images. The SAR images are images taken by the satellite of the first distance spatial baseline of each pixel point on the surface of the engineering subspace. The transmitted radar wavelength is the first wavelength, the time baseline is the first day, and the spatial baseline is represented as the distance between the geometric center points of any two satellites used for imaging.

[0054] Extract the deformation phase corresponding to any SAR image, convert the deformation phase into deformation rate according to InSAR technology, traverse the deformation rate at each pixel in the engineering subspace, calculate the average value of the deformation rate at each pixel, and record the average value of the deformation rate at each pixel as the first deformation rate.

[0055] In practice, the average historical time baseline of similar projects is used as the first day number to avoid manual trial and error. The process is 100% data-driven, with deviations between different teams' results being less than one day. The average value is automatically updated quarterly to form an industry baseline database, enabling the monitoring scheme to adaptively optimize with regional climate and satellite formation evolution. The selection criteria lock the first day number plus or minus one day in the cycle pool. The utilization rate of historical SAR images has increased from 65% to 88%, significantly increasing the number of effective interferometric pairs. The spatial baseline (first distance) and wavelength (first wavelength) selection are unified to eliminate additional phase noise caused by frequency mixing and baseline mixing. The measured phase standard deviation has decreased by 0.3 radians. The rate of each pixel in the subspace is calculated individually and then averaged, which is equivalent to the arithmetic average of 50–200 virtual PSs. Random errors have decreased, and the rate uncertainty has been reduced from plus or minus 3 millimeters per second. The number of days per day is reduced to plus or minus 0.7 mm. The averaging process suppresses local scatterer anomalies, making the results insensitive to transient targets such as temporary metal sheet construction barriers on rooftops. This improves the stability of engineering-level interpretation. Batch matching of historical images requires only one SQL query (time baseline = number of days per day plus or minus 1 day). The data filtering time for 1000 scenes in a single project is less than 30 seconds. The pixel-average rate algorithm has been encapsulated into a single pygmtsar call. The entire S100 process takes less than 10 minutes and requires no manual intervention. After the phase noise is reduced, the annual rate error at the subspace level is reduced by 40%. Accelerated deformations greater than 10 mm per day per day can be detected 15 to 30 days in advance. The output rate format is automatically bound to the IFC subspace Xi number, which is convenient for writing into the BIM and FEM coupled model, realizing standardization of the entire life cycle of surveying, monitoring, and operation and maintenance.

[0056] Perform a first operation according to a first deformation rate, the first operation including marking a first engineering subspace and setting a critical engineering subspace;

[0057] The setup methods for the first operation include:

[0058] Set the first value as the deformation rate threshold, select any engineering subspace, compare the first deformation rate corresponding to the engineering subspace with the first value, when the first deformation rate is less than or equal to the first value, set the first operation to mark the engineering subspace, when the first deformation rate is greater than the first value, set the first operation to set the critical engineering subspace.

[0059] When the first operation is to mark the first subspace of a project, the corresponding subspace number is highlighted in red, and the first day number is added after the subspace number. The first mark is represented as X. i On the first day, create the first folder and store the first tag in the first folder;

[0060] When the first operation is to set up a critical engineering subspace, a second folder is created, and the corresponding engineering subspace number and the corresponding second data backup are stored in the second folder. The second data represents the historical monitoring data of each critical engineering subspace. Each engineering subspace in the second folder is set as a critical engineering subspace.

[0061] In practice, the first value is used as the sole threshold. The system automatically divides the data into two levels: red-highlighted archives and critical backups. It is 100% scripted, requires zero manual judgment, avoids human error in standard shifting, and the data division time is less than 1 second per subspace. For projects with tens of thousands of subspaces, the first round of risk identification can be completed within 1 minute. The subspaces are highlighted in red and the first day number is appended to the number. Engineers can easily identify low-risk and latest cycle subspaces. On-site inspection forms are automatically generated, reducing internal processing time by 30%. The numbering rules are linked with subsequent reports and BIM color cards, and one-click location of the corresponding components in the 3D model is supported.

[0062] The first change is set as the change gradient of the first day number. The original time baseline of the key engineering subspace is adjusted according to the first change, that is, the first adjustment is made to the first day number. The first adjustment is represented by continuous reduction, and the reduced first day number is continuously set as the new time baseline.

[0063] Select any new time baseline, starting from the starting time point of the historical monitoring data, and select historical monitoring data at each new time baseline, and set the selected historical monitoring data as the data to be analyzed;

[0064] By iterating through each new time baseline, the corresponding data to be analyzed is obtained.

[0065] According to the new time baseline, the first deformation rate corresponding to the new time baseline is continuously acquired in descending order;

[0066] The continuous reduction of the first day number is stopped until the first deformation rate corresponding to the new time baseline is less than or equal to the first value, and the adjusted time baseline at this time is obtained and recorded as the critical day number.

[0067] The critical subspaces are then labeled a second time, which consists of the critical subspace number and the critical time baseline. The second time label is denoted as X. i And the key time baseline.

[0068] In practical implementation, the time baseline is automatically reduced using the first change (gradient reduction). The system can converge to the critical number of days in 3 to 5 steps, requiring no expert experience. The result deviation between different operators is less than 0.5 days. Compared with a fixed empirical value (such as a uniform 12 days), the image utilization rate is increased by 20%, and the effective interferogram increases by 1.3 to 1.8 times. The process of cyclically reducing, extracting, and comparing continues until the rate is less than or equal to the first value, ensuring that the critical subspace maintains the lowest detectable rate under the shortest time baseline. The theoretical monitoring sensitivity is increased by 40% (from 5 mm per day to 3 mm per day). The critical number of days is the optimal observation period for that subspace and can be directly written into the observation task book, avoiding oversampling or undersampling. During the reduction process, only the critical subspace is locally recalculated, resulting in low... The risk zone still uses the original long cycle, saving 35% of the overall CPU time. The critical days usually fall between 6 and 24 days. Compared with the traditional 60-day scheme, the time for detecting deformation and mutation is on average 18 days earlier, winning a golden window for emergency response. The second marker Xi and the critical time baseline are written into the scheduling system in real time. Satellite mission lists can be automatically scheduled according to the number of critical days to achieve a closed loop of monitoring and decision-making. The gradient reduction can be set with a lower limit (greater than or equal to 6 days) to prevent excessively low coherence due to infinite shortening. The algorithm has an embedded coherence check. If γ is less than 0.3, it automatically rolls back to the previous cycle to ensure the reliability of the results. γ is a coherence index, which is used to measure the stability of the calculation results in the current cycle. The value of γ is between 0 and 1. The larger the value of γ, the more stable the calculation results in the current cycle.

[0069] The relevant data includes meteorological data and engineering subspace distribution parameters;

[0070] Meteorological data include average annual precipitation, average daily wind speed, average annual air humidity, average annual temperature range, and average annual snow depth;

[0071] The engineering subspace distribution parameters include compressible layer thickness, fill thickness, distance from water source, ground slope, average daily building load, groundwater level depth, and vegetation coverage.

[0072] The methods for conducting a second analysis on relevant data, the first marker, and the second marker to obtain the influence relationship between meteorological data and the time baseline, and the influence relationship between engineering subspace distribution parameters and the time baseline include:

[0073] Calculate the average of the first day number corresponding to the first mark and the critical day number corresponding to the second mark, and denote it as the first average. Using the first average as the dependent variable and meteorological data as the independent variable, perform multiple regression analysis to obtain the first multiple regression equation. The first multiple regression equation is denoteed as the influence relationship between meteorological data and the time baseline.

[0074] Calculate the first difference between the first day number corresponding to the first mark and the critical day number corresponding to the second mark. Use the first difference as the dependent variable and the engineering subspace distribution parameter as the independent variable to perform a multiple regression analysis to obtain the second multiple regression equation. The second multiple regression equation is recorded as the influence relationship between the engineering subspace distribution parameter and the time baseline.

[0075] By inputting meteorological data into the influence relationship of meteorological data on the time baseline, the first average value of a new project with the same type of project to be monitored is obtained. Based on the first day and the first average value of the new project, the corresponding critical days are calculated.

[0076] Calculate the first difference between the first day and the critical day corresponding to the new project. By inputting the corresponding first difference into the influence relationship between the project subspace distribution parameters and the time baseline, the time baseline of each project subspace corresponding to the new project is obtained.

[0077] In practice, only five meteorological parameters and seven types of distribution parameters need to be input, and the first average value and critical number of days can be output within 1 second without having to rerun the entire InSAR process. The efficiency of scheme design is improved by 20 times. For new projects, the satellite imaging plan can be determined 3-6 months in advance, avoiding the waste of resources caused by shooting first and then calculating. The first multiple regression equation gives explicit coefficients such as the shortening of the first average value by 1.2 days for every 100 mm increase in annual precipitation, providing a scientific basis for emergency intensive observation in extreme climate years. The second multiple regression equation reveals the law such as the expansion of the first difference by 0.7 days for every 10 meters increase in compressible layer thickness, which can be directly used for rapid recalculation under sudden geological conditions.

[0078] This invention employs subspace and sensitivity-based secondary partitioning to ensure that the deformation gradient within each computational unit is less than 3 mm per day, reducing the standard deviation of measured rates. By introducing a coupled meteorological and temporal baseline model, atmospheric phase residuals are reduced, and annual rate errors are compressed. The output meteorological and temporal baselines, along with the distributed parameters and temporal baseline dual model, can be directly embedded into a digital twin platform, achieving a data, model, and decision-making closed loop. This provides operation and maintenance units with interpretable sensitivity factor ranking, assisting in the development of targeted reinforcement, drainage, or loading schemes, reducing operation and maintenance risks. The method is applicable to three data sources: BIM, point cloud, and grid, covering various engineering types such as dams, embankments, bridges, landslides, and subway tunnels. The meteorological and geotechnical parameter tables are loosely coupled with the algorithm module, allowing for direct extension to different climate zones and geological conditions through parameter library updates without modifying the core code.

[0079] Those skilled in the art will understand that embodiments of the present invention can be provided as methods, systems, or computer program products. Therefore, the present invention can take the form of a completely hardware embodiment, a completely software embodiment, or an embodiment combining software and hardware aspects. Furthermore, the present invention can take the form of a computer program product implemented on one or more computer-usable storage media containing computer-usable program code. The storage medium can be implemented by any type of volatile or non-volatile storage device or a combination thereof, such as Static Random Access Memory (SRAM), Electrically Erasable Programmable Read-Only Memory (EEPROM), Erasable Programmable Read Only Memory (EPROM), Programmable Red-Only Memory (PROM), Read-Only Memory (ROM), magnetic storage, flash memory, magnetic disk, or optical disk. These computer program instructions may also be stored in a computer-readable storage medium that can direct a computer or other programmable data processing device to function in a particular manner, such that the instructions stored in the computer-readable storage medium produce an article of manufacture including instruction means, which are implemented in a process Figure 1 One or more processes and / or boxes Figure 1 The function specified in one or more boxes.

[0080] It should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention and are not intended to limit it. Although the present invention has been described in detail with reference to preferred embodiments, those skilled in the art should understand that modifications or equivalent substitutions can be made to the technical solutions of the present invention without departing from the spirit and scope of the technical solutions of the present invention, and all such modifications or substitutions should be covered within the scope of the claims of the present invention.

Claims

1. A big data-based engineering deformation monitoring method, characterized in that, Includes the following steps: Step S100: Select any engineering subspace, set a filtering standard, filter the first data according to the filtering standard, analyze the first data to obtain the first deformation rate, perform the first operation according to the first deformation rate, and mark the engineering subspace with the first mark. Step S200: In response to the first operation, a key subspace is set, the time baseline of the key subspace is first adjusted, and the key subspace is second-marked according to the first deformation rate after the first adjustment. Step S300: Perform a second analysis on the relevant data, the first marker, and the second marker to obtain the influence relationship of meteorological data on the time baseline and the influence relationship of engineering subspace distribution parameters on the time baseline; The methods for conducting a second analysis on relevant data, the first marker, and the second marker to obtain the influence relationship between meteorological data and the time baseline, and the influence relationship between engineering subspace distribution parameters and the time baseline include: Calculate the average of the first day number corresponding to the first mark and the critical day number corresponding to the second mark, and denote it as the first average. Using the first average as the dependent variable and meteorological data as the independent variable, perform multiple regression analysis to obtain the first multiple regression equation. The first multiple regression equation is denoteed as the influence relationship between meteorological data and the time baseline. Calculate the first difference between the first day number corresponding to the first mark and the critical day number corresponding to the second mark. Using the first difference as the dependent variable and the engineering subspace distribution parameter as the independent variable, perform a multiple regression analysis to obtain the second multiple regression equation. The second multiple regression equation is denoted as the influence relationship between the engineering subspace distribution parameter and the time baseline. By inputting meteorological data into the influence relationship of meteorological data on the time baseline, the first average value of a new project with the same type of project to be monitored is obtained. Based on the first day and the first average value of the new project, the corresponding critical days are calculated. Calculate the first difference between the first day and the critical day corresponding to the new project. By inputting the corresponding first difference into the influence relationship between the project subspace distribution parameters and the time baseline, the time baseline of each project subspace corresponding to the new project is obtained.

2. The engineering deformation monitoring method based on big data as described in claim 1, characterized in that: Before setting the screening criteria, the project to be monitored is first segmented. This segmentation includes semantic segmentation, parametric segmentation, obtaining project subspaces and automatically numbering them, and outputting the project subspaces and their numbers. The project subspace number is represented as X. i where i is a natural number; The semantic segmentation uses open-source tools and a classification manager to segment the 3D model of the project to be monitored. The semantic representation is the name of the constituent parts of the project. Open IFC, click the classification option, click create independent model, and you will get N independent objects. The parameter segmentation includes setting the compressible layer thickness abruptly to 2 meters as the first segmentation condition, setting the slope threshold to 10 degrees as the second segmentation condition, and setting the minimum number of points to 1000 as the third segmentation condition, thereby segmenting N independent objects to obtain M independent objects. Based on the processing order of the segmentation software, the M independent objects are automatically numbered, and the M independent objects are recorded as a project subspace. The numbers of the M independent objects are recorded as the numbers of the project subspace.

3. The engineering deformation monitoring method based on big data as described in claim 2, characterized in that: After the monitoring project is divided, select any project subspace and set a screening criterion. The screening criterion is that the time baseline is the first day. The time baseline represents the interval of SAR detection. The first day is obtained by the time baseline of projects of the same type as the project to be monitored. The first day is represented as the average value of the time baseline of projects of the same type as the project to be monitored. The first data is selected according to the selection criteria. The first data is represented as each historical monitoring data corresponding to the periodicity of the first day. The historical monitoring data is represented as SAR image. The SAR image is represented as an image of each pixel point on the surface of the engineering subspace taken by a satellite with a first distance spatial baseline. The transmitted radar wavelength is the first wavelength and the time baseline is the first day. The spatial baseline is represented as the distance between the geometric center points of any two satellites used for taking pictures. Extract the deformation phase corresponding to any SAR image, convert the deformation phase into deformation rate according to InSAR technology, traverse the deformation rate at each pixel in the engineering subspace, calculate the average value of the deformation rate at each pixel, and record the average value of the deformation rate at each pixel as the first deformation rate.

4. The engineering deformation monitoring method based on big data as described in claim 3, characterized in that: Perform a first operation according to a first deformation rate, the first operation including marking a first engineering subspace and setting a critical engineering subspace; The setup method for the first operation includes: Set the first value as the deformation rate threshold, select any engineering subspace, compare the first deformation rate corresponding to the engineering subspace with the first value, when the first deformation rate is less than or equal to the first value, set the first operation to mark the engineering subspace, when the first deformation rate is greater than the first value, set the first operation to set a critical engineering subspace.

5. The engineering deformation monitoring method based on big data as described in claim 4, characterized in that: When the first operation is to mark the first subspace of a project, the corresponding subspace number is highlighted in red, and the first day number is added after the subspace number. The first mark is represented as X. i On the first day, create the first folder and store the first tag in the first folder; When the first operation is to set up a critical engineering subspace, a second folder is created, and the corresponding engineering subspace number and the corresponding second data backup are stored in the second folder. The second data represents the historical monitoring data of each critical engineering subspace. Each engineering subspace in the second folder is set as a critical engineering subspace.

6. The engineering deformation monitoring method based on big data as described in claim 5, characterized in that: The first change is set as the change gradient of the first day number. The original time baseline of the key engineering subspace is adjusted according to the first change, that is, the first adjustment is made to the first day number. The first adjustment is represented by continuous reduction, and the reduced first day number is continuously set as the new time baseline. Select any new time baseline, starting from the starting time point of the historical monitoring data, and select historical monitoring data at each new time baseline, and set the selected historical monitoring data as the data to be analyzed; By iterating through each new time baseline, the corresponding data to be analyzed is obtained.

7. The engineering deformation monitoring method based on big data as described in claim 6, characterized in that: According to the new time baseline, the first deformation rate corresponding to the new time baseline is continuously acquired in descending order; The continuous reduction of the first day number is stopped until the first deformation rate corresponding to the new time baseline is less than or equal to the first value, and the adjusted time baseline at this time is obtained and recorded as the critical day number. A second label is applied to the critical subspace, which represents the critical subspace number and the critical time baseline. The second label is denoted as X. i And the key time baseline.

8. The engineering deformation monitoring method based on big data as described in claim 1, characterized in that: The relevant data includes meteorological data and engineering subspace distribution parameters; The meteorological data include average annual precipitation, average daily wind speed, average annual air humidity, average annual temperature range, and average annual snow depth. The engineering subspace distribution parameters include compressible layer thickness, fill thickness, distance from water source, ground slope, average daily building load, groundwater level depth, and vegetation coverage.

9. A big data-based engineering deformation monitoring system, wherein the system is used to execute the big data-based engineering deformation monitoring method as described in claim 1, characterized in that, It includes a tagging module, an analysis module, and a building module; The marking module selects any engineering subspace, sets a filtering standard, filters the first data according to the filtering standard, analyzes the first data to obtain the first deformation rate, performs the first operation according to the first deformation rate, and marks the engineering subspace with the first mark. The analysis module responds to the first operation by setting a key subspace, performing a first adjustment on the time baseline of the key subspace, and performing a second marking on the key subspace according to the first deformation rate after the first adjustment. The construction module performs a second analysis on the relevant data, the first label, and the second label to obtain the influence relationship of meteorological data on the time baseline and the influence relationship of engineering subspace distribution parameters on the time baseline.

Images