Intelligent monitoring method, device and equipment for gradient of high-rise building

Abstract

The invention relates to the technical field of building engineering monitoring, and discloses an intelligent monitoring method, device and equipment for the gradient of a high-rise building. The method comprises the steps that original monitoring data of a high-rise building are collected, and multi-source standardized data of building top displacement, appearance deformation point cloud and foundation settlement are obtained through multiple data preprocessing; performing tight coupling fusion on the building top displacement data to obtain a high-precision three-dimensional displacement time sequence, and extracting building three-dimensional deformation features from the point cloud data; gradient multi-source calculation is carried out on the displacement sequence, the deformation characteristics and the settlement data, and the building inclination state is obtained through federated filtering fusion; performing time sequence frequency decomposition and environmental influence parameter elimination on the inclination state to obtain an ontology inclination feature sequence; trend extrapolation and early warning judgment are carried out based on the body inclination characteristics, and a building inclination detection result is generated. According to the method, the inclination state of the high-rise building is accurately monitored, and the accuracy of inclination trend early warning is improved.

Description

Technical Field

[0001] The present invention relates to the technical field of construction engineering monitoring, and particularly to an intelligent monitoring method, device and equipment for the inclination of high-rise buildings. Background Art

[0002] In the field of high-rise building engineering, inclination monitoring is an important part of structural safety assessment and risk warning, and the intelligent fusion of multi-source data and the prediction of inclination trends are the most critical and challenging stages in inclination monitoring. Accurately detecting the inclination state of high-rise buildings is crucial for construction units, supervision agencies and safety management departments, which directly affects structural safety assessment, service function guarantee and the overall service life of buildings. Therefore, an effective intelligent monitoring method for inclination to monitor the building inclination state in real time is crucial for ensuring the structural stability and service safety of high-rise buildings.

[0003] Among the existing high-rise building inclination monitoring methods, traditional total station measurement cannot achieve high-frequency continuous monitoring due to manual operation, and often misses the instantaneous deformation of the building; although single GNSS monitoring can provide continuous observations, it is affected by multipath effects and satellite signal blockage, and the accuracy is unstable in the urban high-density building environment; laser scanning technology can provide overall appearance deformation information, but the acquisition frequency is low and it is restricted by weather conditions. More importantly, these single monitoring means cannot effectively separate the periodic inclination caused by environmental factors (such as temperature changes, wind loads) and the abnormal deformation of the structure itself, resulting in a high false alarm rate and untimely warning. Therefore, there is an urgent need for an intelligent monitoring method that can fuse multi-source data, eliminate environmental impacts, and accurately identify the inclination characteristics of the main body. That is, the existing high-rise building inclination monitoring methods ignore some unique characteristics of multi-source data fusion and building response, resulting in low accuracy of the final inclination detection result and insufficient warning timeliness. Summary of the Invention

[0004] The main object of the present invention is to solve the problem that the existing high-rise building inclination monitoring methods ignore some unique characteristics of multi-source data fusion and building response, resulting in low accuracy of the final inclination detection result and insufficient warning timeliness.

[0005] The first aspect of the present invention provides an intelligent monitoring method for the inclination of high-rise buildings. The intelligent monitoring method for the inclination of high-rise buildings includes: collecting the original multi-source monitoring data of the high-rise building to be monitored by using a preset multi-source sensor network, and performing multiple data pre-processings on the original multi-source monitoring data to obtain multi-source standardized data, where the multi-source standardized data includes building top displacement monitoring data, appearance deformation monitoring point cloud data, and building foundation settlement monitoring data; performing a tight coupling fusion of building top displacement calculation and displacement accuracy on the building top displacement monitoring data to obtain the three-dimensional displacement time series of the top feature points of the high-rise building, and extracting various building deformation features from the appearance deformation monitoring point cloud data to obtain the three-dimensional building deformation features; performing multi-source inclination calculation on the three-dimensional displacement time series, the three-dimensional building deformation features, and the building foundation settlement monitoring data to obtain various inclination measurement values, and performing federated filtering fusion on each of the inclination measurement values to generate the inclination state of the high-rise building; performing time series frequency decomposition and elimination of various preset environmental impact parameters on the inclination state of the high-rise building to obtain the corrected body inclination feature sequence after environmental correction, and performing detection trend extrapolation and early warning judgment on the body inclination feature sequence to generate the inclination detection result of the high-rise building to be monitored.

[0006] Optionally, in the first implementation manner of the first aspect of the present invention, the original multi-source monitoring data includes satellite navigation original data, inertial measurement original data, building original point cloud data, and leveling original data. The multi-source standardized data is obtained by performing multiple data preprocessing on the original multi-source monitoring data, including: identifying the mutation of the observation sequence generated by the building swing in the satellite navigation original data to obtain the top displacement cycle skip marker data, and performing building displacement curve fitting repair on the top displacement cycle skip marker data to obtain the initial top displacement observation sequence, and performing building swing temperature response correction on the inertial measurement original data to obtain the corrected initial inertial response data, and performing building facade noise filtering on the building original point cloud data to obtain the initial monitoring point cloud data, and performing foundation settlement error compensation on the leveling original data to obtain the initial building settlement data; synchronizing the time bases of the initial top displacement observation sequence, the corrected initial inertial response data, the initial monitoring point cloud data, and the initial building settlement data to obtain a time-consistent building multi-source monitoring data stream, and performing frequency-domain decomposition noise reduction and standardization transformation on the building multi-source monitoring data stream to obtain the building top displacement observation sequence, the building inertial response data, the appearance deformation monitoring point cloud data, and the building foundation settlement monitoring data; generating multi-source standardized data based on the building top displacement observation sequence, the building inertial response data, the appearance deformation monitoring point cloud data, and the building foundation settlement monitoring data, wherein the building top displacement observation sequence and the building inertial response data are combined to generate the building top displacement monitoring data.

[0007] Optionally, in the second implementation manner of the first aspect of the present invention, the three-dimensional displacement time series of the top feature points of the high-rise building is obtained by performing building top displacement calculation and tight-coupling fusion of displacement accuracy on the building top displacement monitoring data, including: performing position calculation on the building top displacement observation sequence based on the preset site positions of the reference station and the monitoring station to obtain the building top positioning position data, and performing building swing attitude calculation and building vibration monitoring coordinate system conversion on the building inertial response data to obtain the navigation acceleration data, and performing frequency-domain second integration on the navigation acceleration data to obtain the building inertial displacement data; using the building top positioning position data and the building inertial displacement data to perform tight-coupling fusion modeling of the high-rise building displacement error to generate a building displacement error state model, and performing filtering fusion of the building swing displacement on the building displacement error state model to obtain the building displacement error estimate value; based on the building displacement error estimate value, performing drift correction on the building top positioning position data and the building inertial displacement data to obtain the fused building displacement data, and performing time serialization on the fused building displacement data to obtain the three-dimensional displacement time series of the top feature points of the high-rise building.

[0008] Optionally, in the third implementation manner of the first aspect of the present invention, the extraction of multiple building deformation features from the appearance deformation monitoring point cloud data to obtain the three-dimensional building deformation features includes: performing multi-station point cloud registration on the building facades of the appearance deformation monitoring point cloud data to obtain the complete point cloud of the building in a unified point cloud coordinate system, and calculating the geometric features of the building facades for the complete point cloud of the building to obtain the building feature enhanced point cloud; performing building structure surface area segmentation and extraction of building cross-section sequences at multiple heights on the building feature enhanced point cloud to obtain building contour lines at multiple heights, and performing anti-interference straight line fitting on the building contour lines at multiple heights to obtain the building tilt angle height distribution curve; performing feature point-to-point matching and displacement calculation for different acquisition time periods on the appearance deformation monitoring point cloud data to obtain the overall building displacement and local deformation components, and performing spatial feature fusion on the building tilt angle height distribution curve and the overall building displacement and local deformation components to obtain the three-dimensional building deformation features.

[0009] Optionally, in the fourth implementation manner of the first aspect of the present invention, the tilt measurement values include the building tilt distribution function, the top displacement tilt rate, and the settlement gradient tilt component. The multi-source calculation of the tilt for the three-dimensional displacement time series, the three-dimensional building deformation features, and the building foundation settlement monitoring data to obtain multiple tilt measurement values, and performing federated filtering fusion on each of the tilt measurement values to generate the high-rise building tilt state includes: performing unified conversion of the preset building engineering unified coordinates and time reference alignment on the three-dimensional displacement time series, the three-dimensional building deformation features, and the building foundation settlement monitoring data to obtain the spatio-temporally unified building tilt multi-source monitoring data, and calculating the standard deviations of the preset multiple monitoring accuracy indicators for the building tilt monitoring data set to obtain multiple monitoring measurement standard deviation values; performing reciprocal square weighted calculation on each of the monitoring measurement standard deviation values to obtain the multi-source fusion weight vector, and performing building bending axis fitting on the three-dimensional building deformation features to obtain the building tilt distribution function, and performing building top displacement tilt calculation on the three-dimensional displacement time series to obtain the top displacement tilt rate, and performing uneven settlement surface fitting on the building foundation settlement monitoring data to obtain the settlement gradient tilt component; based on the multi-source fusion weight vector, performing weighted calculation on the building tilt distribution function, the top displacement tilt rate, and the settlement gradient tilt component to obtain the overall building tilt degree value and the overall building tilt direction, and generating the high-rise building tilt state based on the overall building tilt degree value and the overall building tilt direction.

[0010] Optionally, in the fifth implementation manner of the first aspect of the present invention, the original multi-source monitoring data further includes building environment monitoring data. The time series frequency decomposition of the high-rise building inclination state and the elimination of various preset environmental impact parameters to obtain the body inclination feature sequence after environmental correction include: performing multi-scale decomposition of the building inclination time series of the high-rise building inclination state to obtain a set of building inclination decomposition components, and identifying the periodic characteristics of the building inclination for the set of building inclination decomposition components to obtain a periodic characteristic identification result; performing multiple regression calculations of inclination-environment association on the periodic characteristic identification result and the building environment monitoring data to obtain an inclination-environment response coefficient matrix, and based on the inclination-environment response coefficient matrix, performing environmental impact compensation calculation on the high-rise building inclination state to obtain the body inclination feature sequence after environmental correction.

[0011] Optionally, in the sixth implementation manner of the first aspect of the present invention, the detection trend extrapolation and early warning judgment of the body inclination feature sequence to generate the inclination detection result of the to-be-monitored high-rise building include: performing fitting calculation of multi-period trends on the body inclination feature sequence to obtain a trend mixed prediction result, performing multi-scale window statistical calculation on the body inclination feature sequence to obtain a building inclination statistical anomaly index, and performing multi-path probability prediction on the trend mixed prediction result to obtain a building inclination prediction confidence interval; based on the building inclination prediction confidence interval, performing residual sequence calculation on the body inclination feature sequence and the building inclination trend mixed prediction result to obtain a prediction anomaly index, and performing weighted combination calculation on the building inclination statistical anomaly index and the prediction anomaly index to obtain an inclination anomaly level; performing dynamic threshold judgment on the inclination anomaly level to obtain an inclination early warning level, and based on the inclination early warning level and a preset emergency response strategy library, matching the corresponding emergency warning response parameters of the to-be-monitored high-rise building to generate an inclination detection result.

[0012] In a second aspect of the present invention, an intelligent monitoring device for the inclination of high-rise buildings is provided. The intelligent monitoring device for the inclination of high-rise buildings includes: a data acquisition module, configured to collect original multi-source monitoring data of a high-rise building to be monitored by using a preset multi-source sensor network, and perform multiple data preprocessings on the original multi-source monitoring data to obtain multi-source standardized data, where the multi-source standardized data includes building top displacement monitoring data, appearance deformation monitoring point cloud data, and building foundation settlement monitoring data; a feature extraction module, configured to perform tight-coupling fusion of building top displacement calculation and displacement accuracy on the building top displacement monitoring data to obtain a three-dimensional displacement time series of feature points at the top of the high-rise building, and extract various building deformation features from the appearance deformation monitoring point cloud data to obtain three-dimensional building deformation features; a data fusion module, configured to perform multi-source calculation of the inclination on the three-dimensional displacement time series, the three-dimensional building deformation features, and the building foundation settlement monitoring data to obtain multiple inclination measurement values, and perform federated filter fusion on each of the inclination measurement values to generate the inclination state of the high-rise building; and an early warning judgment module, configured to perform time series frequency decomposition and elimination of multiple preset environmental impact parameters on the inclination state of the high-rise building to obtain an ontology inclination feature sequence after environmental correction, and perform detection trend extrapolation and early warning judgment on the ontology inclination feature sequence to generate an inclination detection result of the high-rise building to be monitored.

[0013] In a third aspect of the present invention, an intelligent monitoring device for the inclination of high-rise buildings is provided, including: a memory and at least one processor, where instructions are stored in the memory; the at least one processor calls the instructions in the memory to cause the intelligent monitoring device for the inclination of high-rise buildings to execute each step of the above-mentioned intelligent monitoring method for the inclination of high-rise buildings.

[0014] The above-mentioned intelligent monitoring method, device and equipment for the inclination of high-rise buildings. In the embodiments of the present invention, through multi-source sensor network monitoring and data preprocessing of the high-rise building to be monitored, standardized monitoring data of the displacement of the building top, appearance deformation and foundation settlement are obtained; then displacement calculation and feature extraction are performed on these data to obtain the three-dimensional displacement time series of the building top and the three-dimensional deformation features of the building; further, through multi-source inclination calculation and federated filtering fusion, the comprehensive inclination state of the high-rise building is obtained; thus, based on time series frequency decomposition and environmental impact elimination, trend extrapolation and early warning judgment are carried out, and the inclination detection result is output. Through multi-level data fusion and feature analysis, accurate identification of the inclination of high-rise buildings is realized. Especially in aspects such as the displacement of the building top, appearance deformation and foundation settlement, the dynamic characteristics and inclination characteristics of the building structure are fully considered, effectively improving the monitoring accuracy; and the multi-source data tightly coupled fusion and federated filtering strategy are adopted, which not only realizes complementary analysis between different monitoring means, but also enhances the reliability of the inclination state evaluation; in addition, through the elimination of environmental impact parameters and trend extrapolation early warning, the inclination characteristics and development trends of the main body are accurately identified, thus overall realizing the efficient and intelligent monitoring of the inclination of high-rise buildings.

[0015] Other features and advantages of the present invention will be described in the following specification, and, in part, will be obvious from the specification, or will be understood by implementing the present invention. The objectives and other advantages of the present invention are achieved and obtained by the structures specifically pointed out in the specification, claims and drawings.

[0016] To make the above-mentioned objectives, features and advantages of the present invention more obvious and understandable, the following specific preferred embodiments are given, and in conjunction with the accompanying drawings, the detailed description is as follows. BRIEF DESCRIPTION OF THE DRAWINGS

[0017] Figure 1 Schematic diagram of the first embodiment of the intelligent monitoring method for the inclination of high-rise buildings in the embodiments of the present invention; Figure 2 Schematic diagram of an embodiment of the intelligent monitoring device for the inclination of high-rise buildings in the embodiments of the present invention; Figure 3 Schematic diagram of an embodiment of the intelligent monitoring equipment for the inclination of high-rise buildings in the embodiments of the present invention. DETAILED DESCRIPTION OF THE EMBODIMENTS

[0018] To make the objectives, technical solutions and advantages of the embodiments of the present invention clearer, the technical solutions of the present invention will be clearly and completely described below in conjunction with the accompanying drawings. Obviously, the described embodiments are part of the embodiments of the present invention, rather than all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those of ordinary skill in the art without creative work fall within the scope of protection of the present invention.

[0019] As used in the embodiments of the present invention, the terms "comprising" and "having" and any variations thereof are intended to cover non-exclusive inclusion. For example, a process, method, system, product, or device that includes a series of steps or units is not limited to the listed steps or units, but may optionally further include other unlisted steps or units, or may optionally further include other steps or units inherent to these processes, methods, products, or devices.

[0020] To facilitate the understanding of this embodiment, the specific process of the embodiments of the present invention will be described below. Please refer to Figure 1 , the first embodiment of the intelligent monitoring method for the inclination of high-rise buildings in the embodiments of the present invention includes: 101. Use a preset multi-source sensor network to collect the original multi-source monitoring data of the high-rise building to be monitored, and perform multiple data preprocessing on the original multi-source monitoring data to obtain multi-source standardized data, which includes building top displacement monitoring data, appearance deformation monitoring point cloud data, and building foundation settlement monitoring data; The embodiments of the present application can acquire and process relevant data based on artificial intelligence technology. Among them, artificial intelligence (AI) is a theory, method, technology, and application system that uses digital computers or machines controlled by digital computers to simulate, extend, and expand human intelligence, perceive the environment, acquire knowledge, and use knowledge to obtain the best results.

[0021] Artificial intelligence basic technologies generally include technologies such as sensors, dedicated artificial intelligence chips, cloud computing, distributed storage, big data processing technology, operation / interaction systems, and mechatronics. Artificial intelligence software technologies mainly include several major directions such as computer vision technology, robotics technology, biometric technology, speech processing technology, natural language processing technology, and machine learning / deep learning.

[0022] In this embodiment, the original multi-source monitoring data includes satellite navigation original data, inertial measurement original data, building original point cloud data, and leveling original data. By identifying the mutation of the observation sequence generated by the building swing in the satellite navigation original data, the top displacement cycle slip marking data is obtained, and the building displacement curve fitting repair is performed on the top displacement cycle slip marking data to obtain the initial top displacement observation sequence. In addition, the building swing temperature response correction is performed on the inertial measurement original data to obtain the corrected initial inertial response data. The building elevation noise filtering is performed on the building original point cloud data to obtain the initial monitoring point cloud data. The foundation settlement error compensation is performed on the leveling original data to obtain the initial building settlement data. The time base synchronization is performed on the initial top displacement observation sequence, the corrected initial inertial response data, the initial monitoring point cloud data, and the initial building settlement data to obtain the building multi-source monitoring data stream with consistent time. The frequency domain decomposition noise reduction and standardization transformation of the building tilt characteristics are performed on the building multi-source monitoring data stream to obtain the building top displacement observation sequence, the building inertial response data, the appearance deformation monitoring point cloud data, and the building foundation settlement monitoring data. Based on the building top displacement observation sequence, the building inertial response data, the appearance deformation monitoring point cloud data, and the building foundation settlement monitoring data, multi-source standardized data is generated. Among them, the building top displacement observation sequence and the building inertial response data are combined to generate the building top displacement monitoring data.

[0023] In practical applications, a preset multi-source sensor network (including GNSS (Global Navigation Satellite System) receivers distributed on the top of the building, IMU (Inertial Measurement Unit) sensors, laser scanning equipment around the building, and leveling points arranged at the foundation of the building, thus forming an "air-sky-ground" three-dimensional monitoring system. For example, in the monitoring of a 300-meter-high super-high-rise building, 3-4 GNSS receivers may be arranged on the top, 10-15 IMU sensors may be installed on key floors, 4-6 laser scanning stations may be set up around the building, and 20-30 precise leveling points may be arranged on the foundation to form a full-range monitoring network) is used to collect the original multi-source monitoring data of the high-rise building to be monitored. These sensors work together to fully capture the overall displacement of the building. , local deformation and foundation settlement conditions, and obtain complete monitoring information including satellite navigation raw data, inertial measurement raw data, building raw point cloud data and leveling measurement raw data; then for the satellite navigation raw data, identify the observation sequence mutation caused by the building swing, that is, by calculating the carrier phase ionospheric combination and the geometrically independent combination, and analyzing the continuity of its time series. When the change value at a certain moment exceeds the preset threshold, the point is marked as a cycle slip point, thereby generating top displacement cycle slip marker data, and these cycle slip marker data are repaired by building displacement curve fitting, that is, using the continuous observation data before and after the cycle slip, through cubic spline function or polynomial function fitting, Reconstruct the data of the cycle slip segment to ensure the continuity of the observation sequence, and finally obtain the initial top displacement observation sequence; and for the original inertial measurement data (since high-rise buildings will experience obvious day and night temperature changes, these temperature changes will cause the IMU sensor zero bias and proportional coefficient to drift, affecting the measurement accuracy), by establishing a polynomial mapping relationship between temperature and sensor output, eliminate the influence of temperature on the accelerometer and gyroscope, and obtain the corrected initial inertial response data to more accurately reflect the actual motion state of the building; and since the original point cloud data of the building usually contains a large number of noise points (these noises come from glass curtain wall reflections, building surface ancillary facilities, hanging objects, etc.), for the building For the original point cloud data of the building, by calculating the statistical characteristics of the distance between each point and its neighboring points, the local density distribution is determined, and the abnormal points that obviously deviate from the local average distance are eliminated, especially those points that may be caused by glass reflection and interference from hanging objects, and finally the initial monitoring point cloud data that can accurately express the appearance of the building is obtained; and for the original data of leveling measurement (due to the existence of systematic errors and random errors in the measurement process, foundation settlement error compensation is required), by applying the adjustment calculation method based on the observation values of known stable points, adjusting the elevation values in the observation network, eliminating the influence of systematic errors, and thus obtaining high-precision initial building settlement data to accurately reflect the settlement state of the building foundation;Furthermore, perform time-base synchronization on the initial top displacement observation sequence, the corrected initial inertial response data, the initial monitoring point cloud data, and the initial building settlement data. Reconstruct the data at a unified time node through the control point interpolation method to ensure that all monitoring data has a consistent time reference, obtaining a building multi-source monitoring data stream with consistent time. Then, perform frequency-domain decomposition noise reduction and standardization transformation on the building multi-source monitoring data stream for building tilt characteristics, that is, convert the time series data in the building multi-source monitoring data stream to the frequency domain through Fourier transform, identify and filter out high-frequency noise and interference signals, retain the frequency band reflecting the true tilt characteristics of the building, and then obtain the noise-reduced time series through inverse transformation. Furthermore, by eliminating the differences in dimension and order of magnitude between different types of data, make the data comparable, and finally obtain high-quality building top displacement observation sequences, building inertial response data, appearance deformation monitoring point cloud data, and building foundation settlement monitoring data. Furthermore, based on the processed data above, generate multi-source standardized data, in which the building top displacement observation sequence and the building inertial response data are combined to generate building top displacement monitoring data after fusion processing. This fusion utilizes the long-term stability of GNSS and the short-term high-precision characteristics of IMU to make up for the deficiencies of a single sensor, while the appearance deformation monitoring point cloud data provides deformation information on the overall appearance of the building, and the building foundation settlement monitoring data reflects the tilt caused by uneven settlement of the foundation.

[0024] 102. Perform tight-coupling fusion of building top displacement calculation and displacement accuracy on the building top displacement monitoring data to obtain the three-dimensional displacement time series of the characteristic points at the top of the high-rise building, and extract various building deformation characteristics from the appearance deformation monitoring point cloud data to obtain the three-dimensional building deformation characteristics. In this embodiment, based on the preset site positions of the reference station and the monitoring station, the position of the displacement observation sequence at the top of the building is solved to obtain the positioning position data of the top of the building, and the inertial response data of the building is subjected to building swing attitude solution and building vibration monitoring coordinate system conversion to obtain navigation acceleration data, and the navigation acceleration data is subjected to frequency-domain double integration to obtain building inertial displacement data; the positioning position data of the top of the building and the building inertial displacement data are used to perform tight-coupling fusion modeling of the displacement error of the high-rise building, generate a building displacement error state model, and perform filtering fusion of the building swing displacement on the building displacement error state model to obtain a building displacement error estimate; based on the building displacement error estimate, drift correction is performed on the positioning position data of the top of the building and the building inertial displacement data to obtain fused building displacement data, and the fused building displacement data is time-serialized to obtain the three-dimensional displacement time series of the characteristic points at the top of the high-rise building; the multi-site cloud registration of the building facade is performed on the appearance deformation monitoring point cloud data to obtain the complete point cloud of the building in the unified point cloud coordinate system, and the geometric features of the building facade are calculated for the complete point cloud of the building to obtain a building feature-enhanced point cloud; the building structure surface area segmentation and the extraction of the building cross-section sequence at multiple heights are performed on the building feature-enhanced point cloud to obtain building contour lines at multiple heights, and the anti-interference straight line fitting of the building inclination is performed on the building contour lines at multiple heights to obtain a building inclination angle height distribution curve; the feature point-to-point matching and displacement calculation of different acquisition periods are performed on the appearance deformation monitoring point cloud data to obtain the overall displacement and local deformation components of the building, and the spatial feature fusion of the building inclination angle height distribution curve and the overall displacement and local deformation components of the building is performed to obtain the three-dimensional deformation characteristics of the building.

[0025] In practical applications, first, based on the preset site positions of the reference station and the monitoring station (referring to the spatial coordinate positions of two types of key stations used for differential positioning in the GNSS satellite positioning system), the position of the displacement observation sequence at the top of the building is solved. This process usually uses a double-difference positioning model, that is, by establishing a carrier phase double-difference observation equation between the monitoring station and the reference benchmark station to eliminate common errors such as satellite clock errors and receiver clock errors, and improve the positioning accuracy. The expression of the double-difference observation equation is: ; Among them, represents the carrier phase double-difference observation value formed by the monitoring station r relative to the reference station between satellites s and t, is the geometric distance double-difference, is the wavelength, is the integer ambiguity double-difference, is the observation noise. For example, in the monitoring of a 380-meter high-rise office building, the reference station is set in a stable area 1 kilometer away from the building, and the monitoring station is installed on the top of the building. The horizontal position accuracy at the top obtained by solving this equation can reach ±3 mm, and the vertical accuracy is about ±5 mm. And for the building inertial response data, through the building swing attitude solution, using the quaternion algorithm or the direction cosine matrix method, processing the data from the accelerometer and gyroscope, determining the three-dimensional attitude parameters of the building (these parameters reflect the inclination and rotation states of the building in space), and after obtaining the attitude parameters, performing the conversion of the building vibration monitoring coordinate system, converting the acceleration data in the body coordinate system to the navigation coordinate system, that is, using the rotation matrix to convert the acceleration measured by the inertial measurement unit from the sensor local coordinate system to the global navigation coordinate system compatible with GNSS, and at the same time eliminating the influence of the gravitational acceleration, finally obtaining the navigation acceleration data representing the pure motion acceleration of the building. Then, for the navigation acceleration data, the acceleration data is converted to the frequency domain through Fourier transform, and then a band-pass filter is applied to retain the frequency band reflecting the true motion characteristics of the building. Then, a second integral calculation is performed in the frequency domain, and finally, the displacement data in the time domain, that is, the building inertial displacement data, is obtained through the inverse Fourier transform (these data have high-frequency characteristics (usually 100 Hz) and can capture the instantaneous vibration response of the building). Then, using the building top positioning position data and the building inertial displacement data to perform a tightly coupled fusion modeling of the high-rise building displacement error, that is, by constructing an error state vector including position error, velocity error, attitude error, and accelerometer and gyroscope zero bias error, forming a building displacement error state model, where this model describes the evolution law of the error state through the system equation, establishes the relationship between the GNSS position and the IMU integrated displacement through the observation equation, and based on the established building displacement error state model, by adopting the extended Kalman filter algorithm, including the prediction step and the update step. In the prediction step, according to the system dynamics equation, the error state and the error covariance matrix are predicted. In the update step, using the difference between the GNSS observation value and the IMU predicted value, the optimal Kalman gain is calculated, and the state estimate and the covariance matrix are updated, and finally the building displacement error estimate value is obtained. Then, based on the building displacement error estimate value, using the estimated error parameters to correct the drift generated in the IMU integration process, and at the same time, the GNSS position data may also be fine-tuned, integrating the advantages of the two data sources, obtaining the fused building displacement data. Then, the fused building displacement data is time-seriesized, organizing the data according to the unified time tag, forming the three-dimensional displacement time series of the characteristic points at the top of the high-rise building. These data not only have the high time resolution characteristics of the IMU but also maintain the long-term stability of the GNSS.

[0026] Secondly, the appearance deformation monitoring point cloud data is subjected to multi-site cloud registration of the building facade, that is, first relying on the target balls or feature control points arranged around the building (these control points provide a common reference benchmark between different scanning stations), the target ball position in each point cloud is automatically detected by the feature recognition algorithm, and the precise coordinates of each target ball are calculated. Then, the rigid body transformation relationship between the target ball coordinates is established, and the coarse registration parameters between different scanning stations are initially obtained. Then, the iterative closest point (ICP) algorithm is used for fine registration, that is, by iteratively optimizing the transformation matrix between the two sets of point clouds, the distance and normal vector difference between the corresponding points are minimized, and finally a high-precision image is obtained. The registration result of each site cloud is converted into a unified coordinate system to form a complete 3D model of the building appearance, that is, the complete point cloud of the building in the unified point cloud coordinate system. Then, the geometric features of the building facade are calculated for the complete point cloud of the building. That is, for each point in the point cloud, the distribution characteristics of its neighborhood points are analyzed, and the principal component analysis is used to determine the direction of the normal vector of the local surface. At the same time, the relative size of the eigenvalue is calculated to obtain the curvature information (where the normal vector reflects the direction of the surface where the point is located, and the curvature represents the degree of curvature of the surface. These two features are crucial for distinguishing different structural elements of the building, such as the plane, edge, corner point, etc.). Through the calculation of these features, The original point cloud is transformed into a building feature enhanced point cloud containing rich geometric information, which significantly improves the accuracy of subsequent structural surface recognition and deformation analysis; then based on the building feature enhanced point cloud, a regional growing algorithm based on normal vector and spatial continuity is adopted. First, points with smaller curvature are selected as seed points, and then gradually expanded to points with similar normal vectors and adjacent spatial positions, and finally continuous plane areas are formed. These areas usually correspond to the main structural elements of the building, such as walls and columns. After the structural surface segmentation is completed, cross sections are extracted at fixed intervals (usually 3-5 meters) along the height direction of the building, that is, by creating a series of horizontal section planes perpendicular to the main axis of the building, these planes are calculated. The intersection line with the point cloud is used to obtain the building contour lines at different heights. These contour lines are combined to form multi-height building contour lines. Then, for the obtained multi-height building contour lines, a robust fitting method such as M-estimation or RANSAC is used. First, a suitable objective function is defined, and the influence weight of the outliers is gradually reduced through iteration. The fitting parameters are optimized until the best fitting straight line is obtained. Then, the angle between each fitting straight line and the plumb line is calculated. This angle is the building inclination angle at that height. The inclination angles at different heights are connected to form a building inclination angle height distribution curve. This curve intuitively shows the law of the building's inclination change along the height direction.Furthermore, feature point-to-point matching and displacement calculation are performed on the appearance deformation monitoring point cloud data for different acquisition periods. The feature point-to-point matching is based on local feature descriptors (such as point feature histograms). By calculating the similarity of feature points in the point clouds of different periods, a one-to-one correspondence is established. After determining the corresponding point pairs, the displacement vectors between each pair of feature points are calculated. These vectors constitute the deformation vector field of the building. Thus, by analyzing the spatial distribution characteristics of the deformation vector field, the overall displacement of the building (such as overall tilt, translation, etc.) and local deformations (such as local bulging, depression, etc.) can be distinguished, and the overall displacement and local deformation components of the building are obtained. Furthermore, for the building tilt angle height distribution curve and the overall displacement and local deformation components of the building, weighted combination or hierarchical fusion methods are adopted, considering the global nature of the tilt angle distribution curve and the locality of the displacement and deformation components. Appropriate weights are assigned to different regions and features, and finally, the comprehensive three-dimensional deformation characteristics of the building are obtained (where this characteristic includes multi-dimensional information such as the overall tilt trend of the building, axis bending deformation, local abnormal deformation regions, etc. For example, in the monitoring of a certain super high-rise building F, a complex "S"-shaped deformation pattern was reconstructed through feature fusion, capturing the differential tilt caused by the combined action of wind load and temperature gradient, and the deformation field reconstruction accuracy reached ±0.3 mm).;

[0027] 103. Perform multi-source calculations of the tilt on the three-dimensional displacement time series, the three-dimensional deformation characteristics of the building, and the building foundation settlement monitoring data to obtain various tilt measurement values, and perform federated filtering fusion on each tilt measurement value to generate the tilt state of the high-rise building; In this embodiment, unified transformation of the preset building engineering unified coordinates and time reference alignment are performed on the three-dimensional displacement time series, the three-dimensional deformation characteristics of the building, and the building foundation settlement monitoring data to obtain the spatio-temporally unified building tilt multi-source monitoring data. The standard deviations of various preset monitoring accuracy indicators are calculated for the building tilt monitoring data set to obtain various monitoring measurement standard deviation values. Reciprocal square weighted calculation is performed on each monitoring measurement standard deviation value to obtain the multi-source fusion weight vector. The building bending axis is fitted to the three-dimensional deformation characteristics of the building to obtain the building tilt distribution function, and the top displacement tilt rate is calculated for the three-dimensional displacement time series, and the uneven settlement surface is fitted to the building foundation settlement monitoring data to obtain the settlement gradient tilt component. Based on the multi-source fusion weight vector, weighted calculation is performed on the building tilt distribution function, the top displacement tilt rate, and the settlement gradient tilt component to obtain the overall tilt degree value and the overall tilt direction of the building, and based on the overall tilt degree value and the overall tilt direction of the building, the tilt state of the high-rise building is generated.

[0028] In practical applications, the process of unified transformation of the preset building engineering unified coordinates and time reference alignment for three-dimensional displacement time series, three-dimensional building deformation characteristics, and building foundation settlement monitoring data first determines a standard coordinate system, usually using the local survey coordinate system or the national coordinate system. That is, a control network is laid out around the building using a total station, and coordinate transformation parameters, including translation, rotation, and scale factor, are established through the known control point coordinates. For GNSS data, it is transformed from the WGS84 coordinate system to the engineering coordinate system through the seven-parameter Helmert transformation model. The transformation parameters include three translation amounts (ΔX, ΔY, ΔZ), three rotation angles (ωX, ωY, ωZ), and a scale factor (μ), which are accurately determined through the registration of control point coordinates; for point cloud data, the local scanning coordinate system is transformed to the control network coordinate system; for settlement data, the elevation of the benchmark point is incorporated into the three-dimensional coordinate system, and in terms of time reference alignment, a unified timestamp standard is adopted, and the data collected at different times is transformed to consistent time nodes through time interpolation methods. For example: for an office building with a height of 100 meters, GNSS data may be collected once per second, while point cloud data may be collected once per week, and settlement measurement data may be collected once per month. It is necessary to unify these data with different time resolutions to the same observation time to form a complete monitoring data set. Subsequently, the standard deviations of various monitoring accuracy indicators are calculated for the building tilt monitoring data set, including GNSS position accuracy, point cloud registration accuracy, settlement measurement accuracy, etc., to ensure that all data is analyzed in a unified spatio-temporal reference system and eliminate systematic errors caused by different coordinate systems and inconsistent times; furthermore, the reciprocal square weighted calculation of the standard deviations of each monitoring measurement utilizes the principle that the monitoring accuracy is inversely proportional to the weight. The higher the accuracy of the data, the greater the weight is assigned. That is, the reciprocal square of the standard deviation of each monitoring data is calculated, and then normalized to obtain the weight vector. For example: assuming that the GNSS position standard deviation is 5mm, the point cloud analysis standard deviation is 3mm, and the settlement measurement standard deviation is 2mm, then their weights are 1 / 25, 1 / 9, and 1 / 4 respectively, and the weight vector [0.13, 0.36, 0.51] is obtained after normalization. Then, the building bending axis is fitted for the three-dimensional building deformation characteristics by fitting a polynomial curve to the tilt angle data at different height levels to form a function describing the bending degree of the building axis.Meanwhile, the displacement inclination rate at the top of the building is calculated using the three-dimensional displacement time series, that is, the horizontal displacement of the top of the building relative to the bottom is measured and divided by the total height of the building to obtain the inclination rate. For the building foundation settlement monitoring data, the least squares plane fitting method is used to fit the foundation settlement surface, extract the maximum inclination gradient and its direction, and obtain the settlement gradient inclination component, so as to characterize the inclination state of the building from different angles, verify each other, and enhance the reliability of the monitoring results. For example, in the monitoring of a certain super high-rise building, when using the federated filter to fuse multi-source inclination data, the weight is set to 0.13 when the GNSS position standard deviation is 5 mm, the weight is set to 0.36 when the point cloud analysis standard deviation is 3 mm, and the weight is set to 0.51 when the settlement measurement standard deviation is 2 mm. The inclination warning thresholds are set as follows: observation level (0.5‰ - 1‰), attention level (1‰ - 2‰), warning level (2‰ - 3‰), and emergency level (> 3‰).; Furthermore, based on the multi-source fusion weight vector, the weighted calculation of the building inclination distribution function, the displacement inclination rate at the top, and the settlement gradient inclination component is to comprehensively combine the inclination measurement values obtained by various monitoring methods and make full use of their respective advantages. For example, the point cloud analysis can provide a continuous inclination distribution along the height, the GNSS provides long-term stable displacement monitoring at the top, and the settlement measurement reflects the inclination caused by uneven foundation settlement. By weighted averaging these three inclination measurement values through the weight vector, a more reliable overall building inclination value can be obtained. At the same time, through the method of vector synthesis, the spatial direction of the inclination is determined to form the overall building inclination direction. For example, for a certain high-rise residential building, the inclination rate obtained by the point cloud analysis is 2‰, the top inclination rate measured by the GNSS is 2.2‰, and the gradient inclination obtained by the settlement measurement is 1.8‰. Through weighted calculation with the weight [0.3, 0.4, 0.3], the final overall building inclination is 2.04‰, and the inclination direction is 15° east of north, so as to fully combine the advantages and disadvantages of various monitoring means and obtain a more comprehensive and accurate description of the inclination state.

[0029] 104. Perform time-series frequency decomposition and elimination of various preset environmental impact parameters on the inclination state of high-rise buildings to obtain the body inclination feature sequence after environmental correction, and perform detection trend extrapolation and warning judgment on the body inclination feature sequence to generate the inclination detection result of the high-rise building to be monitored.

[0030] In this embodiment, the multi-scale decomposition of the building tilt time series is carried out on the tilt state of the high-rise building to obtain the building tilt decomposition component set, and the building tilt cycle characteristics are identified for the building tilt decomposition component set to obtain the cycle characteristic recognition result; the multivariate regression calculation of tilt-environment correlation is carried out on the cycle characteristic recognition result and the building environment monitoring data to obtain the tilt-environment response coefficient matrix, and based on the tilt-environment response coefficient matrix, the environmental impact compensation calculation is carried out on the tilt state of the high-rise building to obtain the corrected body tilt characteristic sequence after environmental correction; the fitting calculation of the multi-period trend is carried out on the body tilt characteristic sequence to obtain the trend mixed prediction result, and the multi-scale window statistical calculation is carried out on the body tilt characteristic sequence to obtain the building tilt statistical anomaly index, and the multi-path probability prediction is carried out on the trend mixed prediction result to obtain the building tilt prediction confidence interval; based on the building tilt prediction confidence interval, the residual sequence calculation is carried out on the body tilt characteristic sequence and the building tilt trend mixed prediction result to obtain the prediction anomaly index, and the weighted combination calculation is carried out on the building tilt statistical anomaly index and the prediction anomaly index to obtain the tilt anomaly level; the dynamic threshold judgment is carried out on the tilt anomaly level to obtain the tilt warning level, and based on the tilt warning level and the preset emergency response strategy library, the corresponding emergency warning response parameters of the high-rise building to be monitored are matched to generate the tilt detection result.

[0031] In practical applications, first, a multi-scale decomposition of the building inclination time series is performed on the inclination state of high-rise buildings. That is, the high-rise building inclination time series is input into the EMD (Empirical Mode Decomposition) algorithm, and each IMF component is extracted through an iterative "screening" process (where each IMF component represents the variation characteristics of the building inclination state at different time scales), arranged in order from high frequency to low frequency, usually manifested as a high-frequency component of daily variation (24-hour cycle), a medium-frequency component of weekly variation (7-day cycle), a low-frequency component of seasonal variation (90 - 120-day cycle), and a long-term trend term. For example: After the inclination monitoring data of a super high-rise building is decomposed, 4 IMF components and 1 trend term are obtained, forming a set of building inclination decomposition components. Then, the periodic characteristics of these components are identified. The power spectral density is calculated using the Fast Fourier Transform (FFT) method to determine the main frequency, amplitude, and phase parameters of each component. And the first IMF component shows an obvious 24-hour periodicity, with an amplitude of about 0.05‰, which is highly correlated with the temperature change caused by solar radiation; the second IMF component shows a periodicity of about 7 days, with an amplitude of about 0.08‰, which is related to the change in personnel load on weekends and weekdays; the third IMF component shows seasonal variation, with an amplitude of about 0.2‰, which is related to the seasonal temperature difference change, thus obtaining the periodic characteristic identification result. This multi-scale decomposition and periodic characteristic identification can effectively distinguish the periodic fluctuations in building inclination from the true structural change trend, avoiding the interference of short-term fluctuations caused by environmental factors on the judgment of the long-term trend; then, a multiple regression calculation of the inclination-environment correlation is performed on the periodic characteristic identification result and the building environment monitoring data. That is, by aligning the building environment monitoring data (such as temperature, humidity, wind speed, wind direction, etc.) with each IMF component in time, and then calculating the response coefficient corresponding to each environmental factor through the multiple linear regression method. Taking temperature as an example: By calculating the regression relationship between the temperature change and the daily cycle IMF component, the temperature influence coefficient is obtained; similarly, by calculating the relationship between the wind speed and the high-frequency component, the wind load influence coefficient is obtained. These coefficients form an inclination-environment response coefficient matrix, and each element in the matrix represents the influence intensity of a specific environmental factor on the inclination component at a specific time scale. For example: In the monitoring of an office building, the change in inclination caused by a 1℃ increase in temperature is about 0.03‰, and the change in inclination caused by a 1m / s increase in wind speed is about 0.01‰; then, based on this response coefficient matrix, an environmental impact compensation calculation is performed on the original high-rise building inclination state, that is, subtracting the inclination components caused by each environmental factor from the original inclination to obtain the body inclination characteristic sequence after environmental correction. This environmental impact compensation can significantly improve the accuracy of inclination monitoring, especially in areas with significant environmental changes, such as areas with large day-night temperature differences, obvious seasonal changes, or frequent wind loads. The compensated inclination characteristic sequence can better reflect the change state of the building structure itself.

[0032] Secondly, the fitting calculation of the multi-period trend of the body tilt feature sequence adopts a combined prediction method, including short-term prediction (24 - 72 hours) and medium- and long-term prediction (7 - 30 days). Among them, the short-term prediction uses the autoregressive integrated moving average (ARIMA) model, which considers the autocorrelation and moving average characteristics of the data and is suitable for capturing short-term fluctuations and trends; the medium- and long-term prediction uses the support vector regression (SVR) method, which performs well in dealing with non-linear trends, and adaptively weights and combines the two prediction results according to the time span to form a trend hybrid prediction result. At the same time, multi-scale window statistical calculations are performed on the body tilt feature sequence, setting three time windows of 24 hours (short-term), 7 days (medium-term), and 30 days (long-term), and calculating the mean, standard deviation, and change rate within each window respectively. By comparing the statistical feature differences between the short-term window and the medium- and long-term windows, abnormal acceleration changes are identified to obtain the building tilt statistical anomaly index. Furthermore, multi-path probability prediction is carried out on the trend hybrid prediction result. By adding random perturbations to the prediction model parameters, various possible development paths under different environmental conditions and loads are simulated to generate multiple prediction trajectories, thereby calculating the upper and lower limit ranges at the 95% confidence level to form the building tilt prediction confidence interval. This multi-period trend analysis and statistical method can identify the abnormal development trend of building tilt in advance and provide a time buffer for early warning decision-making. Furthermore, based on the building tilt prediction confidence interval, a residual sequence calculation is performed on the body tilt feature sequence and the building tilt trend hybrid prediction result, that is, calculating the difference between the measured value and the predicted value, and judging whether the measured value exceeds the prediction confidence interval. When multiple consecutive data points (such as more than 3) exceed the confidence interval boundary, it is considered that a prediction anomaly occurs, and a prediction anomaly index is generated. Furthermore, weighted combination calculation is carried out on the building tilt statistical anomaly index and the prediction anomaly index to comprehensively evaluate the abnormal degree of tilt development. By using weighted average, the weight of the prediction anomaly index is 0.6 and the weight of the statistical anomaly index is 0.4 in the weight assignment. Thus, the abnormal score obtained after combination is divided into the range of 0 - 100, and further classified into four levels: normal (0 - 40), attention (41 - 60), warning (61 - 80), and danger (81 - 100) to form the tilt anomaly level, which can balance the influence of short-term fluctuations and long-term trends and improve the accuracy and robustness of anomaly detection. Furthermore, the dynamic threshold judgment of the tilt anomaly level is to set a differentiated early warning threshold according to the characteristics and safety level of the building itself. Among them, the dynamic threshold is not a fixed value, but a discriminant standard that dynamically adjusts with the building's service life, structural form, and foundation conditions. For example: for a newly built building, the threshold is set more strictly, while for a building that has been operating stably for many years, the threshold is set relatively loosely. Similarly, for important public buildings, the threshold is also set more strictly than for general buildings.Thus, through dynamic threshold judgment, the tilt anomaly level is converted into a specific tilt warning level, which is usually divided into five levels: normal, observation, attention, warning, and emergency. Furthermore, based on the determined tilt warning level, corresponding emergency warning response parameters are matched from the preset emergency response strategy library. For example, for the "observation" level, the response parameters include increasing the monitoring frequency and notifying relevant technical personnel; for the "attention" level, the response parameters include on-site inspection and expert evaluation; for the "warning" level, the response parameters include temporary reinforcement and restricted use of some areas; for the "emergency" level, the response parameters include evacuating personnel and comprehensive reinforcement measures. Thus, a closed-loop control from monitoring to management is achieved, ensuring that corresponding measures can be taken in a timely manner when tilt anomalies are detected, effectively reducing safety risks.

[0033] In the embodiment of the present invention, through multi-source sensor network monitoring and data preprocessing of the high-rise building to be monitored, standardized monitoring data of the displacement of the building top, appearance deformation, and foundation settlement are obtained. Furthermore, displacement calculation and feature extraction are performed on these data to obtain the three-dimensional displacement time series of the building top and the three-dimensional deformation features of the building. Furthermore, through multi-source tilt calculation and federated filter fusion, the comprehensive tilt state of the high-rise building is obtained. Thus, based on time series frequency decomposition and environmental impact elimination, trend extrapolation and early warning judgment are performed to output the tilt detection result. Through multi-level data fusion and feature analysis, accurate identification of the tilt of the high-rise building is achieved. Especially in aspects such as the displacement of the building top, appearance deformation, and foundation settlement, the dynamic characteristics and tilt characteristics of the building structure are fully considered, effectively improving the monitoring accuracy. And the multi-source data tightly coupled fusion and federated filter strategy are adopted, which not only realizes complementary analysis between different monitoring means but also enhances the reliability of tilt state evaluation. In addition, through environmental impact parameter elimination and trend extrapolation early warning, the tilt characteristics and development trends of the entity are accurately identified, thus realizing the efficient and intelligent monitoring of the tilt of the high-rise building as a whole.

[0034] The intelligent monitoring method for the tilt of a high-rise building in the embodiment of the present invention has been described above. Next, the intelligent monitoring device for the tilt of a high-rise building in the embodiment of the present invention will be described. Please refer to Figure 2 , an embodiment of the intelligent monitoring device for the tilt of a high-rise building in the embodiment of the present invention includes: A data acquisition module 201, configured to collect the original multi-source monitoring data of the high-rise building to be monitored by using a preset multi-source sensor network, and perform multiple data preprocessing on the original multi-source monitoring data to obtain multi-source standardized data, where the multi-source standardized data includes building top displacement monitoring data, appearance deformation monitoring point cloud data, and building foundation settlement monitoring data; The feature extraction module 202 is configured to perform tightly coupled fusion of the displacement calculation and displacement accuracy of the building top displacement monitoring data to obtain the three-dimensional displacement time series of the feature points at the top of the high-rise building, and extract various building deformation features from the appearance deformation monitoring point cloud data to obtain the three-dimensional building deformation features; The data fusion module 203 is configured to perform multi-source calculation of the inclination on the three-dimensional displacement time series, the three-dimensional building deformation features, and the building foundation settlement monitoring data to obtain various inclination measurement values, and perform federated filter fusion on each of the inclination measurement values to generate the inclination state of the high-rise building; The warning judgment module 204 is configured to perform time series frequency decomposition and elimination of various preset environmental impact parameters on the inclination state of the high-rise building to obtain the body inclination feature sequence after environmental correction, and perform detection trend extrapolation and warning judgment on the body inclination feature sequence to generate the inclination detection result of the high-rise building to be monitored.

[0035] In the embodiment of the present invention, through multi-source sensor network monitoring and data preprocessing of the high-rise building to be monitored, standardized building top displacement, appearance deformation, and foundation settlement monitoring data are obtained; then, displacement calculation and feature extraction are performed on these data to obtain the three-dimensional displacement time series at the top of the building and the three-dimensional building deformation features; then, through multi-source inclination calculation and federated filter fusion, the comprehensive inclination state of the high-rise building is obtained; thus, based on time series frequency decomposition and environmental impact elimination, trend extrapolation and warning judgment are performed to output the inclination detection result. Through multi-level data fusion and feature analysis, accurate identification of the inclination of the high-rise building is realized. Especially in aspects such as building top displacement, appearance deformation, and foundation settlement, the dynamic characteristics and inclination characteristics of the building structure are fully considered, effectively improving the monitoring accuracy; and the multi-source data tightly coupled fusion and federated filter strategy are adopted, which not only realizes the complementary analysis between different monitoring means but also enhances the reliability of the inclination state evaluation; in addition, through the elimination of environmental impact parameters and trend extrapolation warning, the body inclination features and development trends are accurately identified, thus overall realizing the efficient and intelligent monitoring of the inclination of the high-rise building.

[0036] Above Figure 2 The high-rise building inclination intelligent monitoring device in the embodiment of the present invention is described in detail from the perspective of modular functional entities. Next, the high-rise building inclination intelligent monitoring device in the embodiment of the present invention is described in detail from the perspective of hardware processing.

[0037] Figure 3It is a schematic structural diagram of an intelligent monitoring device for the inclination of high-rise buildings provided by an embodiment of the present invention. The intelligent monitoring device 300 for the inclination of high-rise buildings may vary greatly due to different configurations or performances, and may include one or more processors (central processing units, CPU) 310 (for example, one or more processors) and a memory 320, and one or more storage media 330 for storing application programs 333 or data 332 (for example, one or more mass storage devices). Among them, the memory 320 and the storage medium 330 may be transient storage or persistent storage. The program stored in the storage medium 330 may include one or more modules (not shown in the figure), and each module may include a series of instruction operations on the intelligent monitoring device 300 for the inclination of high-rise buildings. Further, the processor 310 may be configured to communicate with the storage medium 330 and execute a series of instruction operations in the storage medium 330 on the intelligent monitoring device 300 for the inclination of high-rise buildings.

[0038] The intelligent monitoring device 300 for the inclination of high-rise buildings may further include one or more power supplies 340, one or more wired or wireless network interfaces 350, one or more input / output interfaces 360, and / or one or more operating systems 331, such as Windows Serve, Mac OS X, Unix, Linux, FreeBSD, and so on. Those skilled in the art can understand that Figure 3 The shown structural diagram of the intelligent monitoring device for the inclination of high-rise buildings does not limit the intelligent monitoring device for the inclination of high-rise buildings, and may include more or fewer components than shown, or combine certain components, or have different component arrangements.

[0039] The present invention also provides an intelligent monitoring device for the inclination of high-rise buildings. The computer device includes a memory and a processor. When the computer-readable instructions stored in the memory are executed by the processor, the processor executes each step of the method for intelligent monitoring of the inclination of high-rise buildings in the above embodiments.

[0040] Those skilled in the art can clearly understand that for the convenience and brevity of description, the specific working processes of the above-described systems, devices, and units can refer to the corresponding processes in the foregoing method embodiments and will not be repeated here.

[0041] This application can be used in numerous general-purpose or special-purpose computer system environments or configurations. For example: personal computers, server computers, handheld or portable devices, tablet devices, multiprocessor systems, microprocessor-based systems, set-top boxes, programmable consumer electronics, network PCs, minicomputers, mainframe computers, distributed computing environments including any of the above systems or devices, and so on. This application can be described in the general context of computer-executable instructions executed by a computer, such as program modules. Generally, program modules include routines, programs, objects, components, data structures, etc. that perform specific tasks or implement specific abstract data types. This application can also be practiced in a distributed computing environment where tasks are performed by remote processing devices connected through a communication network. In a distributed computing environment, program modules can be located in local and remote computer storage media including storage devices.

[0042] As described above, the above embodiments are only used to illustrate the technical solutions of the present invention and are not intended to limit them; although the present invention has been described in detail with reference to the foregoing embodiments, those of ordinary skill in the art should understand that they can still modify the technical solutions recorded in the foregoing embodiments or perform equivalent replacements for some of the technical features; and these modifications or replacements do not cause the essence of the corresponding technical solutions to deviate from the spirit and scope of the technical solutions of the embodiments of the present invention.

Claims

1. An intelligent monitoring method for the inclination of high-rise buildings, characterized in that, The intelligent monitoring method for the inclination of high-rise buildings includes: Collecting the original multi-source monitoring data of the high-rise building to be monitored by using a preset multi-source sensor network, and performing multiple data preprocessing on the original multi-source monitoring data to obtain multi-source standardized data, where the multi-source standardized data includes building top displacement monitoring data, appearance deformation monitoring point cloud data, and building foundation settlement monitoring data; Performing building top displacement calculation and tightly coupled fusion of displacement accuracy on the building top displacement monitoring data to obtain the three-dimensional displacement time series of the top characteristic points of the high-rise building, and extracting various building deformation characteristics from the appearance deformation monitoring point cloud data to obtain the building three-dimensional deformation characteristics; Performing multi-source calculation of inclination on the three-dimensional displacement time series, the building three-dimensional deformation characteristics, and the building foundation settlement monitoring data to obtain various inclination measurement values, and performing federated filtering fusion on each of the inclination measurement values to generate the inclination state of the high-rise building; Performing time series frequency decomposition and elimination of various preset environmental impact parameters on the inclination state of the high-rise building to obtain the body inclination characteristic sequence after environmental correction, and performing detection trend extrapolation and early warning judgment on the body inclination characteristic sequence to generate the inclination detection result of the high-rise building to be monitored.

2. The intelligent monitoring method for the inclination of high-rise buildings according to claim 1, characterized in that The original multi-source monitoring data includes satellite navigation original data, inertial measurement original data, building original point cloud data, and leveling original data. The multiple data preprocessing of the original multi-source monitoring data to obtain multi-source standardized data includes: Identifying the mutation of the observation sequence generated by the building swing in the satellite navigation original data to obtain the top displacement cycle skip marker data, and performing building displacement curve fitting repair on the top displacement cycle skip marker data to obtain the initial top displacement observation sequence, and performing building swing temperature response correction on the inertial measurement original data to obtain the corrected initial inertial response data, and performing building elevation noise filtering on the building original point cloud data to obtain the initial monitoring point cloud data, and performing foundation settlement error compensation on the leveling original data to obtain the initial building settlement data; Performing time base synchronization on the initial top displacement observation sequence, the corrected initial inertial response data, the initial monitoring point cloud data, and the initial building settlement data to obtain a building multi-source monitoring data stream with consistent time, and performing frequency domain decomposition noise reduction and standardization transformation on the building multi-source monitoring data stream to obtain the building top displacement observation sequence, building inertial response data, appearance deformation monitoring point cloud data, and building foundation settlement monitoring data; Generating multi-source standardized data based on the building top displacement observation sequence, the building inertial response data, appearance deformation monitoring point cloud data, and building foundation settlement monitoring data, where the building top displacement observation sequence and the building inertial response data are combined to generate the building top displacement monitoring data.

3. The intelligent monitoring method for the inclination of high-rise buildings according to claim 2, wherein, The performing building top displacement calculation and tightly coupled fusion of displacement accuracy on the building top displacement monitoring data to obtain the three-dimensional displacement time series of the top characteristic points of the high-rise building includes: Based on the preset site positions of the reference station and the monitoring station, perform position calculation on the displacement observation sequence at the top of the building to obtain the positioning position data of the building top, and perform building swing attitude calculation and conversion of the building vibration monitoring coordinate system on the building inertial response data to obtain the navigation acceleration data, and perform frequency-domain double integration on the navigation acceleration data to obtain the building inertial displacement data; Use the positioning position data of the building top and the building inertial displacement data to perform tightly coupled fusion modeling of the displacement error of the high-rise building, generate a building displacement error state model, and perform filtering fusion of the building swing displacement on the building displacement error state model to obtain an estimated value of the building displacement error; Based on the estimated value of the building displacement error, perform drift correction on the positioning position data of the building top and the building inertial displacement data to obtain fused building displacement data, and perform time serialization on the fused building displacement data to obtain the three-dimensional displacement time series of the characteristic points at the top of the high-rise building.

4. The intelligent monitoring method for the inclination of high-rise buildings according to claim 2, wherein The extraction of various building deformation features from the appearance deformation monitoring point cloud data to obtain three-dimensional building deformation features includes: Perform multi-site cloud registration of the building exterior on the appearance deformation monitoring point cloud data to obtain the complete point cloud of the building in a unified point cloud coordinate system, and perform geometric feature calculation of the building exterior on the complete point cloud of the building to obtain a building feature enhanced point cloud; Perform regional segmentation of the building structure surface and extraction of building cross-section sequences at multiple heights on the building feature enhanced point cloud to obtain building contour lines at multiple heights, and perform anti-interference straight line fitting of the building tilt on the building contour lines at multiple heights to obtain a building tilt angle height distribution curve; Perform feature point-to-point matching and displacement calculation at different acquisition time periods on the appearance deformation monitoring point cloud data to obtain the overall displacement and local deformation components of the building, and perform spatial feature fusion on the building tilt angle height distribution curve and the overall displacement and local deformation components of the building to obtain three-dimensional building deformation features.

5. The intelligent monitoring method for the inclination of high-rise buildings according to claim 1, characterized in that The tilt measurement values include the building tilt distribution function, the top displacement tilt rate, and the settlement gradient tilt component. The multi-source calculation of the tilt on the three-dimensional displacement time series, the three-dimensional building deformation features, and the building foundation settlement monitoring data to obtain various tilt measurement values, and perform federated filtering fusion on each of the tilt measurement values to generate the tilt state of the high-rise building, including: Perform unified conversion of the preset building engineering unified coordinates and time reference alignment on the three-dimensional displacement time series, the three-dimensional building deformation features, and the building foundation settlement monitoring data to obtain spatio-temporally unified multi-source monitoring data of the building tilt, and perform standard deviation calculation of preset multiple monitoring accuracy indicators on the building tilt monitoring data set to obtain multiple monitoring measurement standard deviation values; The reciprocal square weighting calculation is performed on each of the monitored measurement standard deviation values to obtain a multi-source fusion weight vector, and the building three-dimensional deformation characteristics are subjected to building bending axis fitting to obtain a building inclination distribution function, and the three-dimensional displacement time series is subjected to building top displacement inclination calculation to obtain a top displacement inclination rate, and the uneven settlement surface fitting is performed on the building foundation settlement monitoring data to obtain a settlement gradient inclination component; Based on the multi-source fusion weight vector, a weighted calculation is performed on the building inclination distribution function, the top displacement inclination rate, and the settlement gradient inclination component to obtain a building overall inclination value and a building overall inclination direction, and based on the building overall inclination value and the building overall inclination direction, a high-rise building inclination state is generated.

6. The intelligent monitoring method for the inclination of high-rise buildings according to claim 1, characterized in that, The original multi-source monitoring data further includes building environment monitoring data, and the time series frequency decomposition and elimination of multiple preset environmental impact parameters are performed on the high-rise building inclination state to obtain an environmentally corrected body inclination feature sequence, including: Performing multi-scale decomposition of the building inclination time series on the high-rise building inclination state to obtain a building inclination decomposition component set, and performing identification of building inclination period characteristics on the building inclination decomposition component set to obtain a period characteristic identification result; Performing multiple regression calculation of inclination-environment association on the period characteristic identification result and the building environment monitoring data to obtain an inclination-environment response coefficient matrix, and based on the inclination-environment response coefficient matrix, performing environmental impact compensation calculation on the high-rise building inclination state to obtain an environmentally corrected body inclination feature sequence.

7. The intelligent monitoring method for the inclination of high-rise buildings according to claim 1, wherein Performing detection trend extrapolation and early warning judgment on the body inclination feature sequence to generate an inclination detection result of the to-be-monitored high-rise building, including: Performing fitting calculation of multi-period trends on the body inclination feature sequence to obtain a trend mixed prediction result, performing multi-scale window statistical calculation on the body inclination feature sequence to obtain a building inclination statistical anomaly index, and performing multi-path probability prediction on the trend mixed prediction result to obtain a building inclination prediction confidence interval; Based on the building inclination prediction confidence interval, performing residual sequence calculation on the body inclination feature sequence and the building inclination trend mixed prediction result to obtain a prediction anomaly index, and performing weighted combination calculation on the building inclination statistical anomaly index and the prediction anomaly index to obtain an inclination anomaly level; Performing dynamic threshold judgment on the inclination anomaly level to obtain an inclination early warning level, and based on the inclination early warning level and a preset emergency response strategy library, matching emergency warning response parameters corresponding to the to-be-monitored high-rise building to generate an inclination detection result.

8. An intelligent monitoring device for the inclination of high-rise buildings, characterized in that, The high-rise building inclination intelligent monitoring device includes: A data acquisition module, configured to collect original multi-source monitoring data of a to-be-monitored high-rise building by using a preset multi-source sensor network, and perform multiple data preprocessings on the original multi-source monitoring data to obtain multi-source standardized data, where the multi-source standardized data includes building top displacement monitoring data, appearance deformation monitoring point cloud data, and building foundation settlement monitoring data; A feature extraction module, which is used to perform a tight coupling fusion of the displacement calculation and displacement accuracy of the building roof on the displacement monitoring data of the building roof to obtain the three-dimensional displacement time series of the feature points at the top of the high-rise building, and to extract various building deformation features from the appearance deformation monitoring point cloud data to obtain the three-dimensional building deformation features; A data fusion module, which is used to perform multi-source calculations of the inclination on the three-dimensional displacement time series, the three-dimensional building deformation features, and the building foundation settlement monitoring data to obtain various inclination measurement values, and to perform federated filtering fusion on each of the inclination measurement values to generate the inclination state of the high-rise building; An early warning judgment module, which is used to perform time series frequency decomposition and elimination of various preset environmental impact parameters on the inclination state of the high-rise building to obtain the body inclination feature sequence after environmental correction, and to perform detection trend extrapolation and early warning judgment on the body inclination feature sequence to generate the inclination detection result of the high-rise building to be monitored; 9. An intelligent monitoring device for the inclination of high-rise buildings, characterized in that, The high-rise building inclination intelligent monitoring device includes: a memory and at least one processor, and instructions are stored in the memory; The at least one processor calls the instructions in the memory so that the high-rise building inclination intelligent monitoring device executes each step of the high-rise building inclination intelligent monitoring method according to any one of claims 1-7.

Images