Building land subsidence early warning method and system based on knowledge graph

Abstract

The invention discloses a building ground subsidence early warning method and system based on a knowledge graph, and belongs to the technical field of safety early warning. Arranging detectors at a plurality of monitoring points in the settlement monitoring area, and detecting the settlement amount of each monitoring point; acquiring laser surveying and mapping data of the settlement monitoring area to obtain topographic information of the settlement monitoring area; comparing whether each monitoring point of the monitoring point area is interfered or not according to the topographic information of the settlement monitoring area, filtering to obtain a group of laser detection data of the monitoring points without being damaged by unnatural factors, comparing the laser detection data with a preset deviation, and judging the authenticity of settlement and the comprehensive settlement amount of the sub-area; in combination with the settlement volume of each monitoring point and the comprehensive settlement volume of the sub-region, early warning and visualization are performed based on the settlement volume and the settlement rate; and judging whether the settlement rate of the settlement monitoring area accords with expectation based on the knowledge graph, and calculating the influence factor which most accords with the current settlement rate. The settlement trend of the settlement monitoring area can be presented accurately and timely, and early warning can be performed.

Description

Technical Field

[0001] The present invention relates to the field of safety warning technology, and in particular to a building ground subsidence warning method and system based on knowledge graph. Background Art

[0002] In existing construction projects, ground subsidence monitoring has always been a key monitoring indicator in various construction projects. Due to the existence of groundwater below the ground, this will lead to ground subsidence problems caused by underground cavities.

[0003] Existing technologies typically use probe technology and satellite remote sensing for ground subsidence monitoring, but each has its own limitations. Probe technology involves installing sensors on the ground or underground to monitor soil displacement and deformation in real time. Common probes include subsidence meters, inclinometers, and strain gauges. In low-rise building complexes, probe technology is often used for ground subsidence monitoring. This only measures localized deformation at the installation point and fails to fully reflect widespread ground subsidence. This can lead to significant deviations between measured values near roadsides or factories and actual values, making it difficult for the system to effectively filter them. Satellite remote sensing technology uses sensors onboard satellites to capture electromagnetic waves reflected or emitted by the ground, and then analyzes ground deformation. Common remote sensing technologies include synthetic aperture radar (SAR) and optical remote sensing. Remote sensing technology is often used for ground subsidence detection in high-rise buildings, but this results in a slow detection period. Ground subsidence is not detected until the building shifts significantly, making it difficult for the system to quickly identify it.

[0004] Therefore, both probe and satellite remote sensing technologies have their own limitations. Probe technology cannot be deployed in areas with large ground fluctuations or soft ground. Remote sensing technology generally works better on tall buildings, but not on low-rise structures. Addressing the inaccurate data and untimely monitoring of ground subsidence caused by environmental factors is a pressing technical challenge. Summary of the Invention

[0005] To overcome the above problems, the present invention provides a building ground subsidence early warning method and system based on knowledge graph, and the technical solutions adopted are as follows:

[0006] In a first aspect, the present invention provides a building ground subsidence early warning method based on a knowledge graph, comprising the following steps:

[0007] (1) Detectors are arranged at multiple monitoring points in the settlement monitoring area. Lasers are transmitted to the settlement monitoring area. Detectors in the area receive laser signals and return signals, obtaining laser detection data and calculating the settlement amount at each monitoring point.

[0008] (2) Obtain laser mapping data of the subsidence monitoring area and obtain the terrain information of the subsidence monitoring area;

[0009] (3) Compare the monitoring points in the monitoring area with the terrain information of the subsidence monitoring area to see if there is interference, and filter out a set of laser detection data of monitoring points that are not damaged by non-natural factors;

[0010] (4) Compare the laser detection data of a set of monitoring points without damage caused by non-natural factors with the pre-set deviation to determine the authenticity of the settlement and the comprehensive settlement amount of the sub-area;

[0011] (5) Combine the settlement amount of each monitoring point and the comprehensive settlement amount of the sub-region, and issue early warnings based on settlement amount and settlement rate respectively and visualize them;

[0012] (6) Based on the knowledge graph, determine whether the settlement rate in the settlement monitoring area is in line with expectations, and calculate the influencing factors that best match the current settlement rate.

[0013] Furthermore, in step (1), the laser detection data includes the emission distance of the laser and the vertical height of the detector relative to the laser emitting device.

[0014] Furthermore, the laser emission distance is calculated based on the laser return time, and the detector position is calculated in combination with the laser emission angle. The detector position obtained by the previous detection is compared to calculate whether the detector position has settled and the amount of settlement.

[0015] Furthermore, step (2) includes:

[0016] (2-1) The laser pulse is emitted vertically downward by the laser. The laser beam hits the surface of the earth and then returns. The distance between each surveying point and the laser is obtained. The three-dimensional coordinates of a large number of point data on the surface of the earth in the subsidence monitoring area are obtained by multiple measurements.

[0017] (2-2) De-noising and filtering the collected original point data to remove noise points and abnormal points;

[0018] (2-3) Classify the processed point data into different categories, including ground points, building points, and vegetation points;

[0019] (2-4) Extract ground points from point data, use interpolation algorithm to generate regular grids based on the elevation data of ground points to obtain digital elevation model, combine with the classification results of point data, identify and mark building points and vegetation points in the area, and obtain terrain information of the subsidence monitoring area.

[0020] Furthermore, a positioning system and an inertial measurement unit are integrated on the laser to obtain position and attitude information respectively.

[0021] Furthermore, step (3) includes:

[0022] (3-1) Obtain the location of each monitoring point in the settlement monitoring area;

[0023] (3-2) Based on the current topographic information and the previous topographic information of the subsidence monitoring area, compare the fluctuation amplitude of each monitoring point to see if it is greater than a threshold. If so, the monitoring point is judged to be disturbed, and the interference is removed and re-surveyed. If the interference cannot be removed, the data of the monitoring point is directly eliminated to obtain a set of laser detection data of the monitoring point without non-natural factors.

[0024] If not, directly obtain a set of laser detection data of monitoring points without damage caused by natural factors.

[0025] Furthermore, step (4) includes:

[0026] (4-1) Perform local spatial autocorrelation analysis on the monitoring points and the surrounding monitoring points. If the settlement trend of a monitoring point is significantly negatively correlated with that of the surrounding monitoring points, it is determined to be a suspected outlier. Alternatively, compare the settlement of two adjacent monitoring points. If the difference in settlement between a monitoring point and more than three adjacent monitoring points exceeds 2 times the standard deviation, it is determined to be a suspected outlier.

[0027] (4-2) For suspected abnormal points, the historical settlement curve of the monitoring point is combined to determine whether the current data of the monitoring point is a mutation point. If so, the monitoring point is marked as a fault; otherwise, it is regarded as a normal monitoring point and continues to be processed;

[0028] (4-3) Dynamic spatial clustering is performed on normal monitoring points. The cumulative settlement, average settlement rate, and second-order derivative of the settlement curve of each monitoring point are extracted as clustering features. Through the density clustering algorithm, sub-areas with consistent settlement patterns are divided based on the similarity of spatiotemporal characteristics, and the comprehensive settlement of each sub-area is calculated.

[0029] Furthermore, step (5) includes:

[0030] When the settlement rate of a monitoring point is greater than the preset settlement rate, or the total settlement of a monitoring point is greater than the preset maximum settlement, an early warning of a dangerous monitoring point will be issued;

[0031] When the average settlement rate of a sub-area is greater than the preset settlement rate, or when the comprehensive settlement amount of a sub-area is greater than the preset maximum settlement amount, a danger range warning is issued;

[0032] Visualize the sedimentation rate, average sedimentation rate, dangerous monitoring points and dangerous range.

[0033] Furthermore, step (6) includes:

[0034] (6-1) Constructing a knowledge graph: Divide the factors affecting sedimentation rate into natural factors and human factors. Both natural factors and human factors contain several secondary factors.

[0035] (6-2) The relationship between each factor and sedimentation rate is fitted by historical data to obtain the preset weight of each factor;

[0036] (6-3) Survey the influencing factors in the settlement monitoring area, calculate the expected settlement rate based on the preset weights, and determine whether the actual settlement rate in the settlement monitoring area meets the expected settlement rate;

[0037] If it is consistent with expectations, the influencing factors obtained by the survey are considered to be actual factors;

[0038] If it does not meet expectations, execute (6-4);

[0039] (6-4) Calculate the current settlement rate obtained in step (5) and the influencing factors obtained from the survey, arrange and combine the various influencing factors obtained from the survey, and obtain the combination closest to the current settlement rate as the actual factor.

[0040] In the second aspect, the present invention proposes a building ground subsidence early warning system based on knowledge graph, which is used to implement the above-mentioned building ground subsidence early warning method based on knowledge graph.

[0041] The beneficial effects of the present invention are:

[0042] The present invention adopts laser detection technology in conjunction with laser surveying and mapping technology to monitor ground subsidence. It can timely obtain the settlement amount of each monitoring point and the terrain information of the settlement monitoring area through the accumulated data changes in a short period of time, identify the interference points, avoid the influence of the interference points, and then further judge the authenticity of the settlement and the comprehensive settlement amount of the sub-area, thereby obtaining a set of relatively accurate, intuitive and visual monitoring data, which can timely present the ground subsidence phenomenon based on the settlement amount and settlement rate warning respectively; judge whether the settlement rate of the settlement monitoring area meets the expectation based on the knowledge graph, calculate the influencing factors that best meet the current settlement rate, and provide guidance for settlement control. BRIEF DESCRIPTION OF THE DRAWINGS

[0043] Figure 1 It is a schematic diagram of the framework of the knowledge graph-based building ground subsidence early warning method shown in an embodiment of the present invention.

[0044] Figure 2 It is a flow chart of a knowledge graph-based building ground subsidence early warning method shown in an embodiment of the present invention.

[0045] Figure 3 It is a schematic diagram of a knowledge graph shown in an embodiment of the present invention. DETAILED DESCRIPTION

[0046] The present invention will be further described and illustrated below in conjunction with specific embodiments. The embodiments are merely illustrative of the present disclosure and do not limit its scope. The technical features of the various embodiments of the present invention may be combined accordingly, provided that there is no conflict between them.

[0047] The accompanying drawings are merely schematic illustrations of the present invention and are not necessarily drawn to scale. Some of the blocks shown in the accompanying drawings are functional entities that do not necessarily correspond to physically or logically separate entities. These functional entities may be implemented in software, in one or more hardware modules or integrated circuits, or in different networks and / or processor devices and / or microcontroller devices.

[0048] The flowcharts shown in the accompanying drawings are merely illustrative and do not necessarily include all steps. For example, some steps may be decomposed, while some steps may be combined or partially combined, so the actual execution order may change according to actual circumstances.

[0049] Figure 1 The figure shows a framework diagram of a knowledge graph-based building ground subsidence early warning method of the present invention, which introduces laser ranging, laser mapping, positioning system and settlement factors to realize the visualization and early warning of settlement speed, settlement range, etc., as well as to judge whether the treatment plan is effective and analyze the most appropriate influencing factors.

[0050] like Figure 2 As shown, the specific implementation process mainly includes the following steps:

[0051] S1, multiple detectors are arranged in the settlement monitoring area, and lasers are emitted to the settlement monitoring area. The detectors in the area receive the laser signals and return the signals, obtain the laser detection data and calculate the settlement amount of each monitoring point.

[0052] The laser detection data includes the emission distance of the laser and the vertical height of the detector relative to the laser emission device.

[0053] The laser emission distance is calculated based on the laser return time, combined with the laser emission angle , the detector position can be calculated, and by comparing the detector position obtained in the last detection, it can be determined whether the detector position has shifted and the up and down offset angles; here, the detector is fixed on the monitoring point on the ground in the settlement monitoring area and will settle with the monitoring point on the ground.

[0054] In one embodiment of the present invention, a laser transmitter transmits a laser signal to a detector located in a subsidence monitoring area. The detector receives the laser signal and returns a signal. The time t from the laser signal's transmission to its reception is measured. The laser's transmission distance can be calculated based on the signal speed (generally considered to be the speed of light c) and the time difference. :

[0055]

[0056] Here, the division by 2 is because the laser signal needs to make one round trip; to improve accuracy, multiple measurements can be taken and the average value can be obtained.

[0057] Using trigonometric relationships, calculate the vertical height of the detector relative to the laser emitting device:

[0058]

[0059] The distance data of each measurement is stored in the data storage unit, including the current measured data and historical data. In order to analyze the settlement of the detector position, the last height is read from the historical data of the data storage unit. The amount of subsidence at the detector position is equal to the difference in the vertical height of the detector in the two measurements. .

[0060] Measurement results can be corrected for the effects of the laser's angular measurement error, atmospheric refraction during laser propagation, and detector positional deviation. Atmospheric refraction can be corrected by calculating a refraction correction factor based on current meteorological data (such as temperature, pressure, and humidity) using an atmospheric optical model, and then correcting the measured distance. The laser's angular measurement error and detector positional deviation can be minimized through regular equipment calibration, increasing the number of measurement points, and performing data averaging.

[0061] Through this step, the settlement amount of each monitoring point in the monitoring area can be preliminarily obtained.

[0062] S2, obtaining laser mapping data of the subsidence monitoring area to obtain terrain information of the subsidence monitoring area.

[0063] Laser mapping technology uses laser pulses to determine ground distances and shapes by emitting them and measuring their return time. It is widely used in fields such as topographic mapping, architectural modeling, forestry surveys, urban planning, and disaster monitoring. Laser pulses are emitted vertically downward from a laser. The laser beam strikes the Earth's surface and then returns. Based on principles similar to laser detection, the distance between each surveying point and the laser is determined. Multiple measurements are performed to obtain data on a large number of points on the Earth's surface in the subsidence monitoring area. Information such as the time, angle, and reflection intensity from laser emission to reception is recorded. Combined with the laser's position and attitude information obtained from the Beidou positioning system and an inertial measurement unit, the three-dimensional coordinate data for each laser point is generated. The collected raw laser data is processed through denoising and filtering to remove noise and outliers caused by equipment errors, environmental interference, and other factors, thereby improving data quality. Subsequently, the laser point cloud data is classified into different categories, such as ground points, building points, and vegetation points, through machine learning and geometric feature analysis. This technology is well known in the field and will not be further elaborated here.

[0064] Ground points are extracted from the laser point cloud data, and interpolation algorithms such as Kriging interpolation and spline interpolation are used to generate a regular grid based on the elevation data of the ground points to obtain a digital elevation model, which intuitively reflects the terrain undulations of the area. Combined with the classification results of the laser point cloud data, the main ground object information such as buildings and vegetation points in the area is identified and extracted, and they are classified and labeled to obtain the terrain information of the subsidence monitoring area.

[0065] Here, the laser can choose airborne, vehicle-mounted or ground laser scanners, with airborne or vehicle-mounted equipment being preferred. The flight or driving route should be planned in advance. The airborne equipment can be carried by a drone to fly along the planned route, and the vehicle-mounted equipment can be carried by a vehicle to drive along the planned route. Both airborne and vehicle-mounted equipment can combine the Beidou positioning system and the inertial measurement unit to obtain positioning and attitude information.

[0066] S3, using the Beidou positioning system to obtain the location of each monitoring point in the subsidence monitoring area, and combining it with the terrain information of the subsidence monitoring area to compare whether there is a large fluctuation in the monitoring point area, and determine whether the monitoring point is interfered with.

[0067] This step determines whether a monitoring point has been damaged by unnatural factors, such as human intervention or damage by plants and animals. If the current mapping data for a monitoring point differs significantly from the previously mapped data, the laser detection data for that monitoring point is considered invalid. By calculating these two data points, a set of laser detection data indicating that the monitoring point has not been damaged by unnatural factors can be obtained.

[0068] The disturbed monitoring points are reported to the staff for processing. After the processing is completed, the surveying and mapping process can be restarted for secondary surveying and mapping. If the staff does not process it, the data of the erroneous monitoring point will be set as an interference item and the monitoring of the erroneous monitoring point will be suspended.

[0069] S4, comparing the laser detection data of the monitoring point after eliminating interference with the pre-set deviation to determine the authenticity of the settlement and the comprehensive settlement amount.

[0070] Verification: A local spatial autocorrelation analysis is performed on the monitoring point and its surrounding monitoring points, and the Moran's I index is calculated. If the settlement trend of a monitoring point is significantly negatively correlated with that of surrounding monitoring points (p < 0.05), it is identified as a suspected outlier. The true settlement curve should conform to the laws of geomechanics. Furthermore, the historical vertical height curve (settlement curve) of the monitoring point is combined to determine whether the current data of the monitoring point is a mutation point. This embodiment uses Bayesian change point analysis to identify mutation points caused by non-geological factors (such as equipment failure). If so, the monitoring point is marked as a fault and reported to the staff for processing; otherwise, it is treated as a normal monitoring point and processed further.

[0071] In addition to the above solution, those skilled in the art can also compare the settlement of two adjacent monitoring points. If the settlement difference between a monitoring point and three or more adjacent nodes exceeds 2 standard deviations, the monitoring point is identified as a suspected outlier. Similarly, the historical vertical height curve (settlement curve) of the monitoring point can be combined to determine whether the monitoring point is faulty.

[0072] Dynamic spatial clustering was performed on the monitoring points judged to be real, and the cumulative settlement, average settlement rate, and second-order derivative (acceleration) of the curve of each monitoring point were extracted as clustering features. The density clustering (OPTICS) algorithm was used to divide the sub-regions with consistent settlement patterns based on the similarity of spatiotemporal characteristics, and the statistical indicators of each sub-region were calculated.

[0073] In this embodiment, weights are assigned according to the density of monitoring points and geological sensitivity (such as soft soil thickness) in the sub-region, and the weighted average cumulative settlement is calculated to represent the comprehensive settlement of the sub-region.

[0074] S5, settlement warning and visualization.

[0075] By calculating the settlement data detected in the recent period, the settlement rate of the monitoring point is calculated;

[0076] When the settlement rate of a monitoring point is greater than the preset settlement rate, or the total settlement of a monitoring point is greater than the preset maximum settlement, an early warning of a dangerous monitoring point will be issued;

[0077] When the average settlement rate of the sub-area is greater than the preset settlement rate, or when the comprehensive settlement amount of the sub-area is greater than the preset maximum settlement amount, a danger range warning is issued.

[0078] Visualize the above-mentioned sedimentation rate, dangerous monitoring points and dangerous range.

[0079] In one implementation of the present invention, a high-resolution satellite image or vector map (generated using a GIS tool such as QGIS or ArcGIS) is used as a basemap. Monitoring points within the area are marked, with circles representing normal monitoring points and triangles representing dangerous monitoring points. Sub-area boundaries are identified using polygons. Sub-areas within the dangerous range are marked in a semi-transparent red color, and a color gradient is used to represent the average sedimentation rate. Through visualization technology, users can intuitively identify high-risk areas for sedimentation, trace temporal and spatial evolution patterns, and determine the effectiveness of sedimentation treatment measures within the monitoring area.

[0080] S6, based on the knowledge graph, determines whether the settlement rate in the settlement monitoring area meets expectations and calculates the influencing factors that best match the current settlement rate.

[0081] like Figure 3 As shown, a knowledge graph is constructed: the factors affecting the sedimentation rate are divided into natural factors and human factors. Natural factors include primary factors (underground cave collapse, groundwater changes, earthquakes, animal excavation, plant rooting) and secondary factors (reduction of groundwater flow, groundwater diversion, small animal excavation, large animal excavation, large perennial plants, large annual plants); human factors include primary factors (ground engineering, underground engineering) and secondary factors (surface excavation, deep ground excavation, ground reinforcement, subway construction, well digging, basement digging, underground reinforcement).

[0082] By fitting the relationship between each factor and the sedimentation rate through historical data, the preset weight of each factor is obtained; by manually detecting the influencing factors in the area and calculating the expected sedimentation rate, it can be judged whether the sedimentation rate in the monitored area meets expectations and whether the treatment plan is effective; if it does not meet expectations, continue to calculate the sedimentation rate and the determined sedimentation factors obtained through S5, and arrange and combine multiple preset influencing factors to obtain the combination most similar to the sedimentation rate as the influencing factor that best meets the current sedimentation rate, providing guidance for formulating sedimentation treatment measures.

[0083] Based on the same inventive concept, this embodiment also provides a knowledge graph-based building ground subsidence early warning system, which is used to implement the above-mentioned embodiment. The terms "module", "unit", etc. used below can be a combination of software and / or hardware that implements the predetermined function. Although the system described in the following embodiments is preferably implemented in software, it is also possible to implement it in hardware, or a combination of software and hardware.

[0084] In this embodiment, a building ground subsidence early warning system based on a knowledge graph includes:

[0085] A laser detection module is used to transmit laser light to a settlement monitoring area. Detectors are arranged at multiple monitoring points in the settlement monitoring area. The detectors receive laser signals and return signals, acquire laser detection data, and calculate the settlement amount at each monitoring point.

[0086] Laser mapping module, which is used to obtain laser mapping data of the subsidence monitoring area and obtain terrain information of the subsidence monitoring area;

[0087] A filtering module is used to compare the various monitoring points in the monitoring area with the terrain information of the subsidence monitoring area to determine whether interference has occurred, and filter out a set of laser detection data of monitoring points that are not damaged by non-natural factors;

[0088] The data analysis and visualization module is used to compare the laser detection data of a group of monitoring points without non-natural factors with the pre-set deviation to determine the authenticity of the settlement and the comprehensive settlement amount of the sub-area. It combines the settlement amount of each monitoring point and the comprehensive settlement amount of the sub-area to issue early warnings based on settlement amount and settlement rate, and visualize them.

[0089] The knowledge graph module is used to determine whether the settlement rate in the settlement monitoring area meets expectations based on the knowledge graph, and calculate the influencing factors that best match the current settlement rate.

[0090] As for the system embodiment, since it basically corresponds to the method embodiment, the relevant parts can be referred to the partial description of the method embodiment, and the implementation methods of the remaining modules will not be repeated here. The system embodiment described above is only illustrative, wherein the units described as separate components may or may not be physically separated, and the components displayed as units may or may not be physical units, that is, they may be located in one place, or they may be distributed on multiple network units. Some or all of the modules can be selected according to actual needs to achieve the purpose of the solution of the present invention. Ordinary technicians in this field can understand and implement it without paying any creative work.

[0091] Embodiments of the system of the present invention can be applied to any device with data processing capabilities, such as a computer or other device. System embodiments can be implemented through software, hardware, or a combination of software and hardware. For example, a software implementation, as a logical device, is implemented by a processor of any device with data processing capabilities, reading corresponding computer program instructions from non-volatile memory into internal memory and executing them.

[0092] The above examples are merely specific embodiments of the present invention. Obviously, the present invention is not limited to the above examples, and many variations are possible. All variations that can be directly derived or imagined by a person skilled in the art from the disclosure of the present invention should be considered to be within the scope of protection of the present invention.

Claims

1. A building ground subsidence early warning method based on knowledge graph, characterized in that: The following steps are involved: (1) Detectors are arranged at multiple monitoring points in the settlement monitoring area. Lasers are transmitted to the settlement monitoring area. Detectors in the area receive laser signals and return signals, obtaining laser detection data and calculating the settlement amount at each monitoring point. (2) Obtain laser mapping data of the subsidence monitoring area and obtain the terrain information of the subsidence monitoring area; (3) Compare the monitoring points in the monitoring area with the terrain information of the subsidence monitoring area to see if there is interference, and filter out a set of laser detection data of monitoring points that are not damaged by non-natural factors; (4) Compare the laser detection data of a set of monitoring points without damage caused by non-natural factors with the pre-set deviation to determine the authenticity of the settlement and the comprehensive settlement amount of the sub-area; (5) Combine the settlement amount of each monitoring point and the comprehensive settlement amount of the sub-region, and issue early warnings based on settlement amount and settlement rate respectively and visualize them; (6) Based on the knowledge graph, determine whether the settlement rate in the settlement monitoring area is in line with expectations, and calculate the influencing factors that best match the current settlement rate.

2. The building ground subsidence early warning method based on knowledge graph according to claim 1 is characterized in that: In step (1), the laser detection data includes the emission distance of the laser and the vertical height of the detector relative to the laser emission device.

3. The building ground subsidence early warning method based on knowledge graph according to claim 2 is characterized in that: The laser emission distance is calculated based on the laser return time, and the detector position is calculated in combination with the laser emission angle. The detector position obtained by the previous detection is compared to calculate whether the detector position has settled and the amount of settlement.

4. The building ground subsidence early warning method based on knowledge graph according to claim 1 is characterized in that: Step (2) includes: (2-1) The laser pulse is emitted vertically downward by the laser. The laser beam hits the surface of the earth and then returns. The distance between each surveying point and the laser is obtained. The three-dimensional coordinates of a large number of point data on the surface of the earth in the subsidence monitoring area are obtained by multiple measurements. (2-2) De-noising and filtering the collected original point data to remove noise points and abnormal points; (2-3) Classify the processed point data into different categories, including ground points, building points, and vegetation points; (2-4) Extract ground points from point data, use interpolation algorithm to generate regular grids based on the elevation data of ground points to obtain digital elevation model, combine with the classification results of point data, identify and mark building points and vegetation points in the area, and obtain terrain information of the subsidence monitoring area.

5. The building ground subsidence early warning method based on knowledge graph according to claim 4 is characterized in that: The laser is integrated with a positioning system and an inertial measurement unit to obtain position and attitude information respectively.

6. The building ground subsidence early warning method based on knowledge graph according to claim 1 is characterized in that: Step (3) includes: (3-1) Obtain the location of each monitoring point in the settlement monitoring area; (3-2) Based on the current topographic information and the previous topographic information of the subsidence monitoring area, compare the fluctuation amplitude of each monitoring point to see if it is greater than a threshold. If so, the monitoring point is judged to be disturbed, and the interference is removed and re-surveyed. If the interference cannot be removed, the data of the monitoring point is directly eliminated to obtain a set of laser detection data of the monitoring point without non-natural factors. If not, directly obtain a set of laser detection data of monitoring points without damage caused by natural factors.

7. The building ground subsidence early warning method based on knowledge graph according to claim 1 is characterized in that: Step (4) includes: (4-1) Perform local spatial autocorrelation analysis on the monitoring points and the surrounding monitoring points. If the settlement trend of a monitoring point is significantly negatively correlated with that of the surrounding monitoring points, it is determined to be a suspected outlier. Alternatively, compare the settlement of two adjacent monitoring points. If the difference in settlement between a monitoring point and more than three adjacent monitoring points exceeds 2 times the standard deviation, it is determined to be a suspected outlier. (4-2) For suspected abnormal points, the historical settlement curve of the monitoring point is combined to determine whether the current data of the monitoring point is a mutation point. If so, the monitoring point is marked as a fault; otherwise, it is regarded as a normal monitoring point and continues to be processed; (4-3) Dynamic spatial clustering is performed on normal monitoring points. The cumulative settlement, average settlement rate, and second-order derivative of the settlement curve of each monitoring point are extracted as clustering features. Through the density clustering algorithm, sub-areas with consistent settlement patterns are divided based on the similarity of spatiotemporal characteristics, and the comprehensive settlement of each sub-area is calculated.

8. The building ground subsidence early warning method based on knowledge graph according to claim 7 is characterized in that: Step (5) includes: When the settlement rate of a monitoring point is greater than the preset settlement rate, or the total settlement of a monitoring point is greater than the preset maximum settlement, an early warning of a dangerous monitoring point will be issued; When the average settlement rate of a sub-area is greater than the preset settlement rate, or when the comprehensive settlement amount of a sub-area is greater than the preset maximum settlement amount, a danger range warning is issued; Visualize the sedimentation rate, average sedimentation rate, dangerous monitoring points and dangerous range.

9. The building ground subsidence early warning method based on knowledge graph according to claim 7 is characterized in that: Step (6) includes: (6-1) Constructing a knowledge graph: Divide the factors affecting sedimentation rate into natural factors and human factors. Both natural factors and human factors contain several secondary factors. (6-2) The relationship between each factor and sedimentation rate is fitted by historical data to obtain the preset weight of each factor; (6-3) Survey the influencing factors in the settlement monitoring area, calculate the expected settlement rate based on the preset weights, and determine whether the actual settlement rate in the settlement monitoring area meets the expected settlement rate; If it is consistent with expectations, the influencing factors obtained by the survey are considered to be actual factors; If it does not meet expectations, execute (6-4); (6-4) Calculate the current settlement rate obtained in step (5) and the influencing factors obtained from the survey, arrange and combine the various influencing factors obtained from the survey, and obtain the combination closest to the current settlement rate as the actual factor.

10. A building ground subsidence early warning system based on knowledge graph, used to implement the method described in claim 1; characterized in that: The system comprises: A laser detection module is used to transmit laser light to a settlement monitoring area. Detectors are arranged at multiple monitoring points in the settlement monitoring area. The detectors receive laser signals and return signals, acquire laser detection data, and calculate the settlement amount at each monitoring point. Laser mapping module, which is used to obtain laser mapping data of the subsidence monitoring area and obtain terrain information of the subsidence monitoring area; A filtering module is used to compare the various monitoring points in the monitoring area with the terrain information of the subsidence monitoring area to determine whether interference has occurred, and filter out a set of laser detection data of monitoring points that are not damaged by non-natural factors; The data analysis and visualization module is used to compare the laser detection data of a group of monitoring points without non-natural factors with the pre-set deviation to determine the authenticity of the settlement and the comprehensive settlement amount of the sub-area. It combines the settlement amount of each monitoring point and the comprehensive settlement amount of the sub-area to issue early warnings based on settlement amount and settlement rate, and visualize them. The knowledge graph module is used to determine whether the settlement rate in the settlement monitoring area meets expectations based on the knowledge graph, and calculate the influencing factors that best match the current settlement rate.

Images