A scene-aware multi-source deformation monitoring data fusion method and system

Abstract

This invention discloses a multi-source deformation monitoring data fusion method and system based on scene awareness, belonging to the field of deformation monitoring data fusion technology. The invention involves: S1: acquiring GNSS time-series data, InSAR raster data, LiDAR point cloud data, and ECMWF meteorological data of the dam monitoring area; S2: constructing a multi-dimensional scene feature vector containing signal quality and environmental geometry dimensions based on the acquired data; S3: calculating a first weight, a second weight, and a third weight corresponding to the GNSS time-series data, InSAR raster data, and LiDAR point cloud data respectively, based on the multi-dimensional scene feature vector and a preset physical threshold indication function; S4: calculating a first deformation value, a second deformation value, and a third deformation value of the dam monitoring point based on the acquired data; S5: fusing the first deformation value, the second deformation value, and the third deformation value according to the first weight, the second weight, and the third weight to obtain the fused deformation value of the dam monitoring point.

Description

Technical Field

[0001] This invention relates to the field of deformation monitoring data fusion technology, and in particular to a method and system for fusion of multi-source deformation monitoring data based on scene perception. Background Technology

[0002] As large-scale water conservancy projects are concentrated in complex mountainous areas such as plateaus, dam deformation monitoring faces a triple challenge of "complex terrain, harsh climate, and high monitoring accuracy requirements".

[0003] Currently, there are significant limitations to using a single monitoring technology: 1. Limitations of using GNSS technology for dam deformation monitoring: High precision (±2mm for horizontal, ±5mm for vertical), but low spatial resolution and sparse point coverage. 2. Limitations of using InSAR technology for dam deformation monitoring: Wide coverage (100-500km²), but affected by atmospheric delay in mountainous areas (error can reach 10mm), and insensitive to north-south deformation. 3. Limitations of using LiDAR technology for dam deformation monitoring: Millimeter-level three-dimensional point cloud, but lacks the ability to monitor temporal deformation.

[0004] The limitations of using a single monitoring technology for dam deformation monitoring necessitate the fusion of three monitoring technologies to achieve comprehensive dam deformation monitoring. However, existing technologies often employ fixed-weight fusion (e.g., GNSS 50%, InSAR 30%, LiDAR 20%), which cannot adapt to dynamic scenarios. For instance, during severe convective weather, InSAR coherence drops sharply, and LiDAR point cloud noise increases dramatically in steep slope areas. Fixed weights cause fusion accuracy fluctuations exceeding 5mm, all of which affect the final monitoring results of dam deformation. Summary of the Invention

[0005] In view of the above-mentioned prior art, the present invention provides a method and system for fusing multi-source deformation monitoring data based on scene perception, which mainly solves the technical problems existing in the background art.

[0006] To achieve the above objectives, the technical solution of this invention is implemented as follows: In a first aspect, the present invention provides a method for fusing multi-source deformation monitoring data based on scene awareness, the method comprising the following steps: S1: Acquire GNSS time-series data, InSAR raster data, LiDAR point cloud data, and ECMWF meteorological data for the dam monitoring area; S2: Based on the GNSS time series data, the InSAR raster data, the LiDAR point cloud data, and the ECMWF meteorological data, construct a multi-dimensional scene feature vector that includes signal quality dimension and environmental geometry dimension; S3: Based on the multi-dimensional scene feature vector, calculate the first weight, second weight, and third weight corresponding to the GNSS time series data, the InSAR raster data, and the LiDAR point cloud data respectively based on the preset physical threshold indication function; S4: Calculate the first deformation value, the second deformation value, and the third deformation value of the dam body monitoring point based on the GNSS time series data, the InSAR grid data, and the LiDAR point cloud data, respectively. S5: Based on the first weight, the second weight, and the third weight, the first deformation value, the second deformation value, and the third deformation value are fused to obtain the fused deformation value of the dam body monitoring point.

[0007] As a preferred embodiment of the present invention, the scene feature vector of the signal quality dimension in step S2 includes the GNSS signal-to-noise ratio and the InSAR coherence coefficient, and the scene feature vector of the environmental geometry dimension includes the slope of the dam monitoring area and the hourly rainfall.

[0008] As a preferred embodiment of the present invention, step S3, based on the multi-dimensional scene feature vector, calculates the first weight, second weight, and third weight corresponding to the GNSS time-series data, the InSAR raster data, and the LiDAR point cloud data respectively based on a preset physical threshold indication function, specifically including: Based on the physical threshold indication function, the indication values ​​of the GNSS signal-to-noise ratio, the InSAR coherence coefficient, the slope of the dam monitoring area, and the hourly rainfall are calculated respectively. The indication values ​​indicate whether the GNSS signal-to-noise ratio, the InSAR coherence coefficient, the slope of the dam monitoring area, or the hourly rainfall is greater than the corresponding preset threshold. The first weight is calculated based on the indicated value of the GNSS signal-to-noise ratio and the indicated value of the hourly rainfall; The second weight is calculated based on the indicative value of the InSAR coherence coefficient and the indicative value of the slope of the dam monitoring area; The third weight is calculated based on the first weight and the second weight.

[0009] As a preferred embodiment of the present invention, step S4, before calculating the first deformation value, the second deformation value, and the third deformation value of the dam monitoring points based on the GNSS time series data, the InSAR raster data, and the LiDAR point cloud data, further includes: Based on the GNSS time series data, the LiDAR point cloud data, and the ECMWF meteorological data, the atmospheric delay of the InSAR raster data is corrected. Spatial registration is performed on the GNSS time series data, the InSAR raster data, and the LiDAR point cloud data to obtain spatially registered GNSS time series data, the InSAR raster data, and the LiDAR point cloud data.

[0010] As a preferred embodiment of the present invention, the atmospheric delay of the InSAR raster data is corrected based on the GNSS time-series data, the LiDAR point cloud data, and the ECMWF meteorological data, including: Based on the ECMWF meteorological data, the horizontal turbulent mixing delay at the k-th monitoring point on the dam body was calculated. ; The elevation attenuation coefficient of the k-th monitoring point is calibrated based on measured data from the GNSS reference station. ; Based on the LiDAR point cloud data, the elevation of the k-th monitoring point is obtained. ; Obtain the initial value of the vertical stratification delay of the land surface and the residual delay component of the kth monitoring point Among them, the residual delay component Obtained through Kalman filtering optimization; Based on the horizontal turbulent mixing delay at the kth monitoring point of the dam body Elevation attenuation coefficient of the kth monitoring point Elevation of the kth monitoring point Initial value of vertical stratification delay on the ground surface and the residual delay component of the kth monitoring point Calculate the zenith tropospheric delay at the k-th monitoring point. ; The zenith troposphere delay at the k-th monitoring point Converted to interferometric phase delay ; Remove from the interferometric phase of the InSAR raster data The phase-corrected InSAR raster data is obtained.

[0011] As a preferred embodiment of the present invention, the horizontal turbulent mixing delay based on the k-th monitoring point of the dam body... Elevation attenuation coefficient of the kth monitoring point Elevation of the kth monitoring point Initial value of vertical stratification delay on the ground surface and the residual delay component of the kth monitoring point Calculate the zenith tropospheric delay at the k-th monitoring point. The specific formula is as follows:

[0012] The zenith troposphere delay at the k-th monitoring point Converted to interferometric phase delay The specific formula is as follows:

[0013] in, The incident angle of the LiDAR radar wave. This is the operating wavelength.

[0014] As a preferred embodiment of the present invention, spatial registration is performed on the GNSS time-series data, the InSAR raster data, and the LiDAR point cloud data to obtain spatially registered GNSS time-series data, InSAR raster data, and LiDAR point cloud data, including: A local coordinate system was established using the GNSS reference station as the origin to construct the BIM model of the dam body; The LiDAR point cloud data and the InSAR raster data are spatially partitioned using an octree index. Each node of the octree is a substructure of the dam body BIM model, and each substructure represents a different spatial region of the dam body. For InSAR raster data and LiDAR point cloud data at each node, using GNSS coordinates as the reference, coordinate transformation is performed on the InSAR raster point coordinates and LiDAR point cloud coordinates through the first coordinate transformation function and the second coordinate transformation function, respectively. The first spatial registration error between the InSAR raster point coordinates and GNSS coordinates is calculated, as well as the second spatial registration error between the LiDAR point cloud coordinates and GNSS coordinates. Based on the first spatial registration error and the second spatial registration error, the coordinates of the InSAR grid points and the coordinates of the LiDAR point cloud are corrected respectively.

[0015] As a preferred embodiment of the present invention, the specific formula for calculating the first spatial registration error between InSAR grid point coordinates and GNSS coordinates is as follows:

[0016] in, Indicates the coordinates of the GNSS reference station. Indicates the coordinates of InSAR grid points. This represents the first coordinate transformation function. Indicates the first spatial registration error; To minimize the first spatial registration error To find the first coordinate transformation function, we need to determine the objective function. Obtain the minimized first spatial registration error ; The specific formula for calculating the second spatial registration error between LiDAR point cloud coordinates and GNSS coordinates is as follows:

[0017] in, Represents LiDAR point cloud coordinates. This represents the second coordinate transformation function. Indicates the first spatial registration error; To minimize the second spatial registration error To find the objective function, solve for the second coordinate transformation function. To obtain the minimized second spatial registration error .

[0018] As a preferred embodiment of the present invention, based on atmospheric delay-corrected InSAR raster data and spatially registered GNSS time-series data, InSAR raster data and LiDAR point cloud data, the deformation values ​​of the dam body are calculated respectively, referred to as the first deformation value, the second deformation value and the third deformation value, including: Unify the GNSS station coordinates, InSAR raster pixel coordinates, and LiDAR point cloud coordinates into a local coordinate system with the GNSS reference station as the origin, and align them with the spatial structure of the dam body BIM model. In the dam body BIM model, a key point of the dam body to be monitored is selected. The coordinates of the key point of the dam body BIM model are synchronously mapped to the GNSS network, InSAR deformation raster map and LiDAR point cloud dataset through the established coordinate transformation relationship to obtain the target monitoring point at the same physical location. Based on the three-dimensional coordinates of the target monitoring point, calculate the linear rate of change of GNSS data in the elevation direction within the specified monitoring period T, and use the linear rate of change in the elevation direction as the first deformation value. Based on the three-dimensional coordinates of the target monitoring point, the vertical linear change rate of the InSAR raster data within the specified monitoring period T is calculated, and the vertical linear change rate is used as the second deformation value. Based on the three-dimensional coordinates of the target monitoring point, the cumulative elevation change of the LiDAR point cloud data within the monitoring period T is calculated, and the cumulative elevation change is used as the third deformation value.

[0019] Secondly, the present invention also provides a scene-aware multi-source deformation monitoring data fusion system, the system comprising: The data acquisition module is used to acquire GNSS time-series data, InSAR raster data, LiDAR point cloud data, and ECMWF meteorological data of the dam monitoring area; The extraction module is used to extract a multi-dimensional scene feature vector containing signal quality dimension and environmental geometry dimension based on the GNSS time series data, the InSAR raster data, the LiDAR point cloud data and the ECMWF meteorological data. The first calculation module is used to calculate, based on the multi-dimensional scene feature vector and a preset physical threshold indication function, the first weight, the second weight, and the third weight corresponding to the GNSS time series data, the InSAR raster data, and the LiDAR point cloud data, respectively. The second calculation module calculates the first deformation value, the second deformation value, and the third deformation value of the dam body monitoring point based on the GNSS time series data, the InSAR raster data, and the LiDAR point cloud data, respectively. The fusion module is used to fuse the first deformation value, the second deformation value, and the third deformation value according to the first weight, the second weight, and the third weight to obtain the fused deformation value of the dam body monitoring point.

[0020] Thirdly, the present invention also provides an electronic device, including a memory and a processor, wherein the processor is used to implement a scene-aware multi-source deformation monitoring data fusion method when executing a computer management program stored in the memory.

[0021] Fourthly, the present invention also provides a computer-readable storage medium storing a computer management program thereon, wherein the computer management program, when executed by a processor, implements the steps of a scene-aware multi-source deformation monitoring data fusion method.

[0022] The beneficial effects of this invention are as follows: This invention provides a multi-source deformation monitoring data fusion method and system based on scene awareness. Based on the GNSS signal-to-noise ratio, InSAR coherence coefficient, slope of the dam monitoring area, and hourly rainfall, it calculates the weights corresponding to GNSS time-series data, InSAR raster data, and LiDAR point cloud data respectively based on a preset physical threshold indicator function. Based on these weights, the dam deformation values ​​calculated from the GNSS time-series data, InSAR raster data, and LiDAR point cloud data are fused to obtain the fused deformation value of the dam monitoring points. This invention uses four-dimensional operating parameters—GNSS signal quality, InSAR coherence coefficient, terrain slope, and meteorological conditions—and dynamically calculates the fusion weights of the three data sources in real time based on a physical threshold indicator function. It achieves high fusion accuracy in complex scenarios such as severe convective weather and steep terrain, effectively meeting the high-precision, real-time monitoring needs of dams in complex mountainous areas. Attached Figure Description

[0023] Figure 1 A flowchart of a multi-source deformation monitoring data fusion method based on scene awareness provided by the present invention; Figure 2 A schematic diagram of the structure of a scene-aware multi-source deformation monitoring data fusion system provided by the present invention; Figure 3 A schematic diagram of the hardware structure of an electronic device provided by the present invention; Figure 4 This is a schematic diagram of the hardware structure of a computer-readable storage medium provided by the present invention. Detailed Implementation

[0024] The technical solution of the present invention will be further described in detail below with reference to the accompanying drawings and specific embodiments. Unless otherwise defined, all technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the art to which this invention pertains. The terminology used in this specification is for the purpose of describing particular embodiments only and is not intended to limit the invention. In the following description, the expression "some embodiments" refers to a subset of all possible embodiments; however, it should be understood that "some embodiments" can be the same subset or different subsets of all possible embodiments and can be combined with each other without conflict.

[0025] In the following description, numerous specific details are set forth in order to provide a more thorough understanding of the invention. However, it will be apparent to those skilled in the art that the invention can be practiced without one or more of these details. In other instances, certain technical features well-known in the art have not been described in order to avoid obscuring the invention.

[0026] It should be understood that the present invention can be embodied in various forms and should not be construed as being limited to the embodiments set forth herein. Rather, providing these embodiments will make the disclosure thorough and complete, and will fully convey the scope of the invention to those skilled in the art. Furthermore, the terminology used herein is intended only to describe particular embodiments and is not intended to limit the invention. When used herein, the singular forms “a,” “an,” and “the” are also intended to include the plural forms unless the context clearly indicates otherwise. It should also be understood that the terms “compose” and / or “comprising,” when used in this specification, identify the presence of the stated features, integers, steps, operations, elements, and / or components, but do not exclude the presence or addition of one or more other features, integers, steps, operations, elements, components, and / or groups. When used herein, the term “and / or” includes any and all combinations of the associated listed items.

[0027] It should also be noted that when an element is referred to as being "fixed to" another element, it can be directly attached to the other element or there may be an intervening element. When an element is referred to as being "connected to" another element, it can be directly connected to the other element or there may be an intervening element. The terms "vertical," "horizontal," "inner," "outer," "left," "right," and similar expressions used herein are for illustrative purposes only and do not represent the only possible implementation.

[0028] To fully understand this invention, a detailed structure will be presented in the following description to illustrate the technical solution proposed by this invention. Optional embodiments of the invention are described in detail below; however, in addition to these detailed descriptions, the invention may have other embodiments.

[0029] Firstly, this invention provides a method for fusing multi-source deformation monitoring data based on scene awareness, please refer to the appendix. Figure 1 The method includes the following steps: S1: Acquire GNSS time-series data, InSAR raster data, LiDAR point cloud data, and ECMWF meteorological data for the dam monitoring area.

[0030] In this embodiment, monitoring equipment, including a GNSS reference station, InSAR equipment, LiDAR radar, and a meteorological station, is deployed on the hydropower station dam to acquire GNSS time-series data, InSAR raster data, LiDAR point cloud data, and ECMWF meteorological data of the dam monitoring area.

[0031] S2: Based on the GNSS time series data, the InSAR raster data, the LiDAR point cloud data, and the ECMWF meteorological data, construct a multi-dimensional scene feature vector that includes signal quality dimension and environmental geometry dimension.

[0032] As a preferred embodiment of the present invention, the scene feature vector of the signal quality dimension in step S2 includes the GNSS signal-to-noise ratio and the InSAR coherence coefficient, and the scene feature vector of the environmental geometry dimension includes the slope of the dam monitoring area and the hourly rainfall.

[0033] Understandably, based on the four-dimensional operating parameters of GNSS signal-to-noise ratio, InSAR coherence coefficient, slope of the dam monitoring area, and hourly rainfall, a scene-sensitive segmented weight model is used to assign corresponding weights to the three source data: GNSS time series data, InSAR raster data, and LiDAR point cloud data.

[0034] Specifically, multi-dimensional scene feature vectors, including signal quality and environmental geometry dimensions, were extracted from GNSS time-series data, InSAR raster data, LiDAR point cloud data, and ECMWF meteorological data. These primarily consist of four-dimensional scene feature vectors. The scene feature vectors containing the signal quality dimension include the GNSS signal-to-noise ratio and InSAR coherence coefficient, while the scene feature vectors containing the environmental geometry dimension include the slope of the dam monitoring area and hourly rainfall.

[0035] Specifically, the signal-to-noise ratio (SNR) of Global Navigation Satellite System (GNSS) signals is a key physical quantity characterizing the strength of the received signal relative to the background noise level. A higher SNR means that the effective signal power is significantly higher than the noise floor. A high SNR reduces the random error of carrier phase observations, making phase ambiguity resolution more robust, thereby directly improving absolute and relative positioning accuracy. In complex environments with significant multipath effects or radio frequency interference, a high SNR helps the receiver signal processing module more effectively distinguish between direct signals and reflected / interference signals, improving the system's robustness. A high SNR reduces the risk of data interruption due to transient signal loss, ensuring the continuity and stability of deformation time series. Therefore, this invention uses the GNSS signal SNR as the core indicator of signal quality.

[0036] Therefore, when assigning weights, the first weight corresponding to the GNSS time-series data is positively correlated with the SNR. The higher the SNR, the better the signal quality of the GNSS monitoring point in the current observation period, and the first deformation value calculated from it should be given a higher confidence weight in the fusion process.

[0037] Specifically, meteorological conditions, especially short-duration heavy rainfall, significantly impact GNSS signal propagation. A sudden increase in water vapor content in the ionosphere and troposphere causes signal path delay, while rain attenuation can lead to signal power reduction, equivalent to a decrease in SNR. Incorporating hourly rainfall into the scene feature vector aims to achieve environmentally-aware correction of GNSS data weights. When hourly rainfall is high, even if the original output SNR of the GNSS receiver has not decreased significantly, its observations may already contain systematic meteorological delay errors, and the risk of signal loss increases.

[0038] When calculating the weights for GNSS data, both the signal-to-noise ratio (SNR) and hourly rainfall are considered. High rainfall will trigger a weight decay function, dynamically reducing the weight to reflect the potential decrease in the reliability of GNSS observations under adverse weather conditions. Therefore, this invention uses hourly rainfall as a correction factor for the environmental geometry.

[0039] Specifically, the coherence coefficient of Synthetic Aperture Radar Interferometry (InSAR) is a fundamental indicator for evaluating the reliability of interferometric phase information. Its physical meaning is the temporal and spatial correlation of the backscattered signals of corresponding pixels in two SAR images. A high coherence coefficient (close to 1) indicates that the scattering characteristics of the target pixel remain stable between two imaging sessions. Its interferometric phase mainly includes deformation, elevation, and controllable atmospheric phase information, which can be used for high-precision deformation inversion. A low coherence coefficient (close to 0) means that the phase information is dominated by noise, making it impossible to reliably extract deformation signals. The coherence coefficient is directly related to the phase noise variance. In areas with high coherence, deformation monitoring accuracy can reach the millimeter level; in areas with excessively low coherence, effective deformation data cannot be generated. In time-series InSAR raster data analysis, high-coherence points are the foundation for constructing stable deformation sequences, and the coherence threshold directly determines the spatial density of effective monitoring points. Therefore, the second weight corresponding to the InSAR raster data is strongly correlated with the coherence coefficient. A higher coherence coefficient indicates better phase quality of the pixel at that location, and the weight of the calculated second deformation value in data fusion should be increased accordingly. Therefore, this invention uses the InSAR coherence coefficient as a direct measure of signal quality.

[0040] S3: Based on the multi-dimensional scene feature vector, calculate the first weight, second weight, and third weight corresponding to the GNSS time series data, the InSAR raster data, and the LiDAR point cloud data respectively based on the preset physical threshold indication function.

[0041] In this embodiment, the slope of the dam monitoring area extracted by the LiDAR is used to calculate the third weight of the LiDAR point cloud data and serve as the geometric constraint condition for the InSAR raster data.

[0042] Specifically, slope is a key geometric constraint affecting InSAR coherence and LiDAR accuracy; terrain slope is a core feature in environmental geometry and has a decisive impact on the quality of both InSAR and LiDAR data. The slope significantly alters the geometry of radar side-looking imaging, impacting InSAR coherence. Steep slopes tend to cause severe geometric decoherence, specifically including: ① inducing overlay or shadowing phenomena, leading to pixel decorrelation; ② increasing sensitivity to spatial baselines, where even small vertical baselines can generate large phase gradients on slopes; and ③ altering the surface scattering mechanism, introducing unstable volume scattering components. Therefore, the steeper the slope, the lower the InSAR coherence coefficient typically is, and the worse its applicability for deformation monitoring.

[0043] This invention uses slope information extracted from LiDAR point cloud data as a priori geometric constraint for evaluating the reliability of InSAR data. Before calculating the weights of the InSAR data, it first determines whether the region is affected by severe geometric distortion based on the slope. For regions exceeding a preset slope threshold, their InSAR data weights can be directly set to zero or significantly attenuated, thereby avoiding the introduction of low-reliability data into the fusion process.

[0044] The slope is also a key parameter for evaluating the accuracy of LiDAR point cloud data, affecting its own weighting. On steep slopes, the distribution of laser footprints in LiDAR point cloud data may be sparse and uneven, and increasing the scanning angle introduces additional uncertainty in elevation measurement. Therefore, when calculating the third weight of LiDAR point cloud data, slope is used as a direct input parameter. In areas with gentle slopes, the LiDAR point cloud data has high accuracy and a high weight; in areas with steep slopes, the measurement uncertainty model indicates that the weight should be reduced.

[0045] As a preferred embodiment of the present invention, step S3, based on the multi-dimensional scene feature vector, calculates the first weight, second weight, and third weight corresponding to the GNSS time-series data, the InSAR raster data, and the LiDAR point cloud data respectively based on a preset physical threshold indication function, specifically including: Based on the physical threshold indication function, the indication values ​​of the GNSS signal-to-noise ratio, the InSAR coherence coefficient, the slope of the dam monitoring area, and the hourly rainfall are calculated respectively. The indication values ​​indicate whether the GNSS signal-to-noise ratio, the InSAR coherence coefficient, the slope of the dam monitoring area, or the hourly rainfall is greater than the corresponding preset threshold. Based on the indicated value of the GNSS signal-to-noise ratio and the indicated value of the hourly rainfall, the first weight is calculated, and the specific calculation formula is as follows:

[0046] The second weight is calculated based on the InSAR coherence coefficient indication value and the slope indication value of the dam monitoring area. The specific calculation formula is as follows:

[0047] The third weight is calculated based on the first weight and the second weight, using the following formula:

[0048] in, As the first weight, As the second weight, As the third weight, This represents the physical threshold indicator function, which takes a value of 1 when the condition is true, and a value of 0 otherwise. This represents the signal-to-noise ratio of GNSS signals, in dB. Under normal operating conditions, the value range is 30-40 dB. This represents the InSAR coherence coefficient, which typically ranges from 0.6 to 0.8 under normal operating conditions. Indicates the slope of the monitored area, in degrees (°). This represents hourly rainfall, measured in mm / h; for severe convective weather, rainfall is >10 mm / h. , , and These are the first preset threshold, the second preset threshold, the third preset threshold, and the fourth preset threshold, respectively.

[0049] In this embodiment of the invention, the first preset threshold is 25dB, the second preset threshold is 5mm / h, the third preset threshold is 0.5, and the fourth preset threshold is 20°.

[0050] The signal-to-noise ratio of the GNSS signal and the hourly rainfall extracted from the collected data are substituted into the first weighting formula to calculate the first weight. Similarly, the extracted InSAR coherence coefficient and the slope of the monitoring area are substituted into the second weighting formula to calculate the second weight.

[0051] The third weight is calculated based on the first and second weights.

[0052] In this embodiment of the invention, weights are assigned according to specific application scenarios, for example: Under normal operating conditions ( >30dB, >0.6, <20°, <5mm / h): =0.40, =0.35, =0.25 Under severe convective weather ( >10mm / h, <0.5): =0.50, =0.15, =0.35, reducing the weight of InSAR raster data and increasing the contribution of GNSS time series data and LiDAR point cloud data.

[0053] Below the steep slope area ( >30°): =0.40, =0.40, =0.20, reducing the LiDAR weights to avoid point cloud noise interference.

[0054] S4: Calculate the first deformation value, the second deformation value, and the third deformation value of the dam monitoring points based on the GNSS time series data, the InSAR raster data, and the LiDAR point cloud data.

[0055] As a preferred embodiment of the present invention, due to the complex atmospheric errors in mountainous areas, considering the differentiated impacts of terrain and meteorology on different data sources, and in order to improve data accuracy, atmospheric delay correction needs to be applied to the InSAR raster data for the acquired raw GNSS time-series data, InSAR raster data, and LiDAR point cloud data. Therefore, in step S4, before calculating the first deformation value, second deformation value, and third deformation value of the dam monitoring point based on the GNSS time-series data, the InSAR raster data, and the LiDAR point cloud data, the following steps are included: Based on the GNSS time series data, the LiDAR point cloud data, and the ECMWF meteorological data, the atmospheric delay of the InSAR raster data is corrected. Spatial registration is performed on the GNSS time series data, the InSAR raster data, and the LiDAR point cloud data to obtain spatially registered GNSS time series data, the InSAR raster data, and the LiDAR point cloud data.

[0056] As a preferred embodiment of the present invention, the atmospheric delay of the InSAR raster data is corrected based on the GNSS time-series data, the LiDAR point cloud data, and the ECMWF meteorological data, including: Based on the ECMWF meteorological data, the horizontal turbulent mixing delay at the k-th monitoring point on the dam body was calculated. ; The elevation attenuation coefficient of the k-th monitoring point is calibrated based on measured data from the GNSS reference station. ; Based on the LiDAR point cloud data, the elevation of the k-th monitoring point is obtained. ; Obtain the initial value of the vertical stratification delay of the land surface and the residual delay component of the kth monitoring point Among them, the residual delay component Obtained through Kalman filtering optimization; Based on the horizontal turbulent mixing delay at the kth monitoring point of the dam body Elevation attenuation coefficient of the kth monitoring point Elevation of the kth monitoring point Initial value of vertical stratification delay on the ground surface and the residual delay component of the kth monitoring point Calculate the zenith tropospheric delay at the k-th monitoring point. ; The zenith troposphere delay at the k-th monitoring point Converted to interferometric phase delay ; Remove from the interferometric phase of the InSAR raster data The phase-corrected InSAR raster data is obtained.

[0057] It should be noted that, in this embodiment of the invention, the horizontal turbulent mixing delay at the k-th monitoring point is calculated. Elevation attenuation coefficient of the kth monitoring point Elevation of the kth monitoring point Initial value of vertical stratification delay on the ground surface and the residual delay component of the kth monitoring point All calculations are performed using existing technology, so they will not be repeated here.

[0058] As a preferred embodiment of the present invention, the horizontal turbulent mixing delay based on the k-th monitoring point of the dam body... Elevation attenuation coefficient of the kth monitoring point Elevation of the kth monitoring point Initial value of vertical stratification delay on the ground surface and the residual delay component of the kth monitoring point Calculate the zenith tropospheric delay at the k-th monitoring point. The specific formula is as follows:

[0059] The zenith troposphere delay at the k-th monitoring point Converted to interferometric phase delay The specific formula is as follows:

[0060] in, The incident angle of the LiDAR radar wave. This is the operating wavelength.

[0061] Specifically, the three types of data—GNSS time-series data, InSAR raster data, and LiDAR point cloud data—have large differences in time sampling frequencies and inconsistent spatial references, resulting in large direct fusion errors. Therefore, after obtaining the phase-corrected InSAR raster data, spatial registration is performed on the GNSS time-series data, InSAR raster data, and LiDAR point cloud data to obtain spatially registered GNSS time-series data, InSAR raster data, and LiDAR point cloud data.

[0062] As a preferred embodiment of the present invention, spatial registration is performed on the GNSS time-series data, the InSAR raster data, and the LiDAR point cloud data to obtain spatially registered GNSS time-series data, InSAR raster data, and LiDAR point cloud data, including: A local coordinate system was established using the GNSS reference station as the origin to construct the BIM model of the dam body; The LiDAR point cloud data and the InSAR raster data are spatially partitioned using an octree index. Each node of the octree is a substructure of the dam body BIM model, and each substructure represents a different spatial region of the dam body. For InSAR raster data and LiDAR point cloud data at each node, using GNSS coordinates as the reference, coordinate transformation is performed on the InSAR raster point coordinates and LiDAR point cloud coordinates through the first coordinate transformation function and the second coordinate transformation function, respectively. The first spatial registration error between the InSAR raster point coordinates and GNSS coordinates is calculated, as well as the second spatial registration error between the LiDAR point cloud coordinates and GNSS coordinates. Based on the first spatial registration error and the second spatial registration error, the coordinates of the InSAR grid points and the coordinates of the LiDAR point cloud are corrected respectively.

[0063] Specifically, the formula for calculating the first spatial registration error between InSAR grid point coordinates and GNSS coordinates is as follows:

[0064] in, Indicates the coordinates of the GNSS reference station. Indicates the coordinates of InSAR grid points. This represents the first coordinate transformation function, which includes translation, rotation, and scaling parameters. Indicates the first spatial registration error; To minimize the first spatial registration error To find the first coordinate transformation function, we need to determine the objective function. Obtain the minimized first spatial registration error ; Specifically, similarly, using GNSS coordinates as the reference, the formula for calculating the second spatial registration error between LiDAR point cloud coordinates and GNSS coordinates is as follows:

[0065] in, Represents LiDAR point cloud coordinates. This represents the second coordinate transformation function, which includes translation, rotation, and scaling parameters. Indicates the first spatial registration error; To minimize the second spatial registration error To find the objective function, solve for the second coordinate transformation function. To obtain the minimized second spatial registration error The original coordinates of the LiDAR point cloud are corrected based on the second spatial registration error. In this embodiment of the invention, the second spatial registration error is required to be ≤0.5mm to meet the millimeter-level fusion requirement.

[0066] As a preferred embodiment of the present invention, based on atmospheric delay-corrected InSAR raster data and spatially registered GNSS time-series data, InSAR raster data and LiDAR point cloud data, the deformation values ​​of the dam body are calculated respectively, referred to as the first deformation value, the second deformation value and the third deformation value, including: Unify the GNSS station coordinates, InSAR raster pixel coordinates, and LiDAR point cloud coordinates into a local coordinate system with the GNSS reference station as the origin, and align them with the spatial structure of the dam body BIM model. In the dam body BIM model, a key point of the dam body to be monitored is selected. The coordinates of the key point of the dam body BIM model are synchronously mapped to the GNSS network, InSAR deformation raster map and LiDAR point cloud dataset through the established coordinate transformation relationship to obtain the target monitoring point at the same physical location. Based on the three-dimensional coordinates of the target monitoring point, calculate the linear rate of change of GNSS data in the elevation direction within the specified monitoring period T, and use the linear rate of change in the elevation direction as the first deformation value. Based on the three-dimensional coordinates of the target monitoring point, the vertical linear change rate of the InSAR raster data within the specified monitoring period T is calculated, and the vertical linear change rate is used as the second deformation value. Based on the three-dimensional coordinates of the target monitoring point, the cumulative elevation change of the LiDAR point cloud data within the monitoring period T is calculated, and the cumulative elevation change is used as the third deformation value.

[0067] S5: Based on the first weight, the second weight, and the third weight, the first deformation value, the second deformation value, and the third deformation value are fused to obtain the fused deformation value of the dam monitoring points. The specific fusion formula is as follows:

[0068] in, The deformation value is expressed in mm. , and These represent the first deformation value, the second deformation value, and the third deformation value, respectively. , and These represent the first weight, the second weight, and the third weight, respectively.

[0069] To illustrate the technical effectiveness of the fusion method of this invention, a specific example is used below to explain the multi-source deformation monitoring data fusion method based on scene awareness provided in the embodiments of this invention.

[0070] Monitoring object: A hydropower station dam in a plateau canyon area The basic parameters of the dam include: dam height: 150m; dam length: 500m; slope range: 25-40°; average number of severe convective weather days per year: 30; altitude: 3200-3400m.

[0071] The monitoring equipment is arranged as follows: GNSS: 3 reference stations (left shoulder of the dam, right shoulder of the dam, and crest of the dam), sampling interval 30 seconds; InSAR: Sentinel-1A C-band data, revisit period 12 days, resolution 5m×5m; LiDAR: Scanning frequency 200kHz, point cloud density 100 points / m²; Meteorological data: ECMWF hourly data.

[0072] The data collected by the monitoring equipment is shown in Table 1 below: Table 1 Data collected

[0073] Based on the collected data, four-dimensional working condition parameters were extracted, as shown in Table 2 below.

[0074] Table 2 Four-dimensional working condition parameters

[0075] (1) Adaptive weight allocation is performed based on the four-dimensional working condition parameters. The calculation formula is as follows: = I(28>25)×[0.3+0.1×I(12<5)] = 1×[0.3+0.1×0]= 0.30; = I(0.45>0.5)×[0.25+0.1×I(35<20)] = 0×[...]= 0; because = 0 (coherence coefficient too low), therefore reallocation: = 0.15 (Retain minimum weight to avoid complete information loss); = 1 - 0.30 - 0.15 = 0.55; However, considering the impact of steep slopes, the adjustment is as follows: = 0.50 (Increase the weight of stable sources); = 0.15 (reduce the weight of failure sources); = 0.35 (Appropriately reduce the weight of noise sources).

[0076] Compared to normal operating conditions: =35dB, =0.70, =25°, =2mm / h, =0.40, =0.35, =0.25.

[0077] (2) Atmospheric delay correction for InSAR raster data: For example, the correction process for a certain monitoring point is as follows: Input parameters: = 3.35 km (elevation); = 12 mm (ECMWF meteorological data interpolation); = 75 mm (value taken in the altitude range of 3000-4000m); = 0.13 km -1 (GNSS reference station calibration); = -1.5 mm (Kalman filter optimization).

[0078] = 12 + 75×e(-0.13×3.35) + (-1.5) = 12 + 75 × e -0.4355 + (-1.5) = 12 + 75 × 0.647 + (-1.5) = 12 + 48.5 - 1.5 = 59.0 mm.

[0079] Phase delay conversion: =( / 0.056)×59.0×cos(39 °) = 224.3×59.0×0.777= 10289 degrees≈28.6 ° (phase correction).

[0080] The correction resulted in an InSAR phase error of 15mm before correction, which was reduced to 6mm after correction, improving accuracy by 60%.

[0081] Specifically, after phase correction of InSAR raster data, the phase error of InSAR raster data is reduced by 60%, and the accuracy after correction is ≤3mm.

[0082] (3) Spatial registration of the three source data: Registration reference: GNSS dam crest reference station coordinates: = (500250.123, 2785630.456, 3215.789) mm.

[0083] LiDAR point cloud data registration: 1. Construct a BIM model of the dam, dividing it into 20 dam sections; 2. Octree node size: 1m × 1m × 1m; 3. Initial conversion parameters This includes translations (Δx, Δy, Δz) and rotations (α, β, γ). 4. After 100 iterations of optimization, the final registration error is:

[0084] To minimize the first spatial registration error To find the first coordinate transformation function, we need to determine the objective function. Obtain the minimized first spatial registration error .

[0085] InSAR raster data registration: With a grid resolution of 5m, registration to the GNSS coordinate system was performed using bilinear interpolation. 1. Construct a BIM model of the dam, dividing it into 20 dam sections; 2. Octree node size: 1m × 1m × 1m; 3. Initial conversion parameters This includes translations (Δx, Δy, Δz) and rotations (α, β, γ). 4. After 100 iterations of optimization, the final registration error is:

[0086] To minimize the second spatial registration error To find the objective function, solve for the second coordinate transformation function. To obtain the minimized second spatial registration error .

[0087] Dynamic weighted fusion: Three-source deformation data (vertical direction ΔZ) at a certain monitoring point: = 0.82 mm (sinking); = 0.68 mm (sinking); = 0.91 mm (sinking).

[0088] = 0.50×0.82+0.15×0.68+0.35×0.91=0.41+0.10+0.32=0.83 mm.

[0089] Verification was performed using a total station (accuracy ±0.5mm): Measured value: 0.85 mm; Fusion deviation: |0.83 - 0.85| = 0.02 mm ≤ 0.15 mm.

[0090] In this embodiment of the invention, the weights of the three data sources are allocated according to the specific application scenario, and the deformation values ​​of the dam body from the three data sources are fused according to the allocated weights to obtain the fused deformation value. The deformation monitoring error of the dam body is ≤0.15mm.

[0091] Secondly, this invention also provides a scene-aware multi-source deformation monitoring data fusion system, please refer to [link to relevant documentation]. Figure 2 The system includes: Data acquisition module 201 is used to acquire GNSS time-series data, InSAR raster data, LiDAR point cloud data and ECMWF meteorological data of the dam monitoring area; Extraction module 202 is used to extract a multi-dimensional scene feature vector containing signal quality dimension and environmental geometry dimension based on the GNSS time series data, the InSAR raster data, the LiDAR point cloud data and the ECMWF meteorological data; The first calculation module 203 is used to calculate, based on the multi-dimensional scene feature vector and a preset physical threshold indication function, the first weight, the second weight, and the third weight corresponding to the GNSS time series data, the InSAR raster data, and the LiDAR point cloud data, respectively. The second calculation module 204 calculates the first deformation value, the second deformation value, and the third deformation value of the dam body monitoring point based on the GNSS time series data, the InSAR grid data, and the LiDAR point cloud data, respectively. The fusion module 205 is used to fuse the first deformation value, the second deformation value and the third deformation value according to the first weight, the second weight and the third weight to obtain the fused deformation value of the dam body monitoring point.

[0092] It is understood that the scene-aware multi-source deformation monitoring data fusion system provided by the present invention corresponds to the scene-aware multi-source deformation monitoring data fusion method provided in the foregoing embodiments. The relevant technical features of the scene-aware multi-source deformation monitoring data fusion system can be referred to the relevant technical features of the scene-aware multi-source deformation monitoring data fusion method, and will not be repeated here.

[0093] As can be seen, this invention uses four-dimensional working condition parameters—GNSS signal quality, InSAR coherence coefficient, terrain slope, and meteorological conditions—and constructs a segmented weight model based on a physical threshold indicator function to dynamically calculate the fusion weight of three-source data in real time. It achieves high fusion accuracy in complex scenarios such as severe convective weather and steep terrain, and can effectively meet the high-precision and real-time monitoring needs of dams in complex mountainous areas.

[0094] In addition to collecting GNSS time-series data and InSAR raster data, LiDAR point cloud data was also collected. LiDAR point cloud data can be used not only as a deformation monitoring source but also as a scene perception source (to extract slope information). The weights of InSAR raster data and LiDAR point cloud data are then adjusted based on the slope information, making the final weight adjustment of the three data sources more flexible and accurate.

[0095] Thirdly, the present invention also provides an electronic device 300, please refer to [link to relevant documentation]. Figure 3 The system includes a memory 310, a processor 320, and a computer program 311 stored on the memory 310 and capable of running on the processor 320. When the processor 320 executes the computer program 311, it implements the steps of a scene-aware multi-source deformation monitoring data fusion method.

[0096] Fourthly, the present invention also provides a computer-readable storage medium 400, see below. Figure 4 It stores a computer program 411, which, when executed by a processor, implements the steps of a scene-aware multi-source deformation monitoring data fusion method.

[0097] It should be noted that the descriptions of each embodiment in the above embodiments have different focuses. For parts that are not described in detail in a certain embodiment, please refer to the relevant descriptions in other embodiments.

[0098] Those skilled in the art will understand that embodiments of the present invention can be provided as methods, systems, or computer program products. Therefore, the present invention can take the form of a completely hardware embodiment, a completely software embodiment, or an embodiment combining software and hardware aspects. Furthermore, the present invention can take the form of a computer program product embodied on one or more computer-usable storage media (including, but not limited to, disk storage, CD-ROM, optical storage, etc.) containing computer-usable program code.

[0099] This invention is described with reference to flowchart illustrations and / or block diagrams of methods, apparatus (systems), and computer program products according to embodiments of the invention. It will be understood that each block of the flowchart illustrations and / or block diagrams, and combinations of blocks in the flowchart illustrations and / or block diagrams, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, special-purpose computer, embedded computer, or other programmable data processing apparatus to produce a machine, such that the instructions, which execute via the processor of the computer or other programmable data processing apparatus, generate instructions for implementing the flowchart illustrations. Figure 1 One or more processes and / or boxes Figure 1 A device that provides the functions specified in one or more boxes.

[0100] These computer program instructions may also be stored in a computer-readable storage medium that can direct a computer or other programmable data processing device to function in a particular manner, such that the instructions stored in the computer-readable storage medium produce an article of manufacture including instruction means, which are implemented in a process Figure 1 One or more processes and / or boxes Figure 1 The function specified in one or more boxes.

[0101] These computer program instructions may also be loaded onto a computer or other programmable data processing equipment to cause a series of operational steps to be performed on the computer or other programmable equipment to produce a computer-implemented process, thereby providing instructions that execute on the computer or other programmable equipment for implementing the process. Figure 1 One or more processes and / or boxesFigure 1 The steps of the function specified in one or more boxes.

[0102] Although preferred embodiments of the invention have been described, those skilled in the art, upon learning the basic inventive concept, can make other changes and modifications to these embodiments. Therefore, the appended claims are intended to be interpreted as including both the preferred embodiments and all changes and modifications falling within the scope of the invention.

[0103] The above are merely specific embodiments of the present invention, but the scope of protection of the present invention is not limited thereto. Any variations or substitutions that can be easily conceived by those skilled in the art within the technical scope disclosed in the present invention should be included within the scope of protection of the present invention. The scope of protection of the present invention should be determined by the scope of the claims.

Claims

1. A method for fusing multi-source deformation monitoring data based on scene awareness, characterized in that, The method includes the following steps: S1: Acquire GNSS time-series data, InSAR raster data, LiDAR point cloud data, and ECMWF meteorological data for the dam monitoring area; S2: Based on the GNSS time series data, the InSAR raster data, the LiDAR point cloud data, and the ECMWF meteorological data, construct a multi-dimensional scene feature vector that includes signal quality dimension and environmental geometry dimension; S3: Based on the multi-dimensional scene feature vector, calculate the first weight, second weight, and third weight corresponding to the GNSS time series data, the InSAR raster data, and the LiDAR point cloud data respectively based on the preset physical threshold indication function; S4: Calculate the first deformation value, the second deformation value, and the third deformation value of the dam body monitoring point based on the GNSS time series data, the InSAR grid data, and the LiDAR point cloud data, respectively. S5: Based on the first weight, the second weight, and the third weight, the first deformation value, the second deformation value, and the third deformation value are fused to obtain the fused deformation value of the dam body monitoring point.

2. The multi-source deformation monitoring data fusion method based on scene awareness according to claim 1, characterized in that, In step S2, the scene feature vector for the signal quality dimension includes the GNSS signal-to-noise ratio and the InSAR coherence coefficient, while the scene feature vector for the environmental geometry dimension includes the slope of the dam monitoring area and the hourly rainfall.

3. The multi-source deformation monitoring data fusion method based on scene awareness according to claim 2, characterized in that, In step S3, based on the multi-dimensional scene feature vector, the first weight, second weight, and third weight corresponding to the GNSS time-series data, the InSAR raster data, and the LiDAR point cloud data are calculated according to a preset physical threshold indication function. Specifically, this includes: Based on the physical threshold indication function, the indication values ​​of the GNSS signal-to-noise ratio, the InSAR coherence coefficient, the slope of the dam monitoring area, and the hourly rainfall are calculated respectively. The indication values ​​indicate whether the GNSS signal-to-noise ratio, the InSAR coherence coefficient, the slope of the dam monitoring area, or the hourly rainfall is greater than the corresponding preset threshold. The first weight is calculated based on the indicated value of the GNSS signal-to-noise ratio and the indicated value of the hourly rainfall; The second weight is calculated based on the indicative value of the InSAR coherence coefficient and the indicative value of the slope of the dam monitoring area; The third weight is calculated based on the first weight and the second weight.

4. The multi-source deformation monitoring data fusion method based on scene awareness according to claim 3, characterized in that, In step S4, before calculating the first deformation value, second deformation value, and third deformation value of the dam monitoring points based on the GNSS time-series data, the InSAR raster data, and the LiDAR point cloud data, the following steps are included: Based on the GNSS time series data, the LiDAR point cloud data, and the ECMWF meteorological data, the atmospheric delay of the InSAR raster data is corrected. Spatial registration is performed on the GNSS time series data, the InSAR raster data, and the LiDAR point cloud data to obtain spatially registered GNSS time series data, the InSAR raster data, and the LiDAR point cloud data.

5. The multi-source deformation monitoring data fusion method based on scene awareness according to claim 4, characterized in that, Based on the GNSS time-series data, the LiDAR point cloud data, and the ECMWF meteorological data, atmospheric delay correction is performed on the InSAR raster data, including: Based on the ECMWF meteorological data, the horizontal turbulent mixing delay at the k-th monitoring point on the dam body was calculated. ; The elevation attenuation coefficient of the k-th monitoring point is calibrated based on measured data from the GNSS reference station. ; Based on the LiDAR point cloud data, the elevation of the k-th monitoring point is obtained. ; Obtain the initial value of the vertical stratification delay of the land surface and the residual delay component of the kth monitoring point Among them, the residual delay component Obtained through Kalman filtering optimization; Based on the horizontal turbulent mixing delay at the kth monitoring point of the dam body Elevation attenuation coefficient of the kth monitoring point Elevation of the kth monitoring point Initial value of vertical stratification delay on the ground surface and the residual delay component of the kth monitoring point Calculate the zenith tropospheric delay at the k-th monitoring point. ; The zenith troposphere delay at the k-th monitoring point Converted to interferometric phase delay ; Remove from the interferometric phase of the InSAR raster data The phase-corrected InSAR raster data is obtained.

6. The multi-source deformation monitoring data fusion method based on scene awareness according to claim 5, characterized in that, The horizontal turbulent mixing delay based on the k-th monitoring point of the dam body Elevation attenuation coefficient of the kth monitoring point Elevation of the kth monitoring point Initial value of vertical stratification delay on the ground surface and the residual delay component of the kth monitoring point Calculate the zenith tropospheric delay at the k-th monitoring point. The specific formula is as follows: The zenith troposphere delay at the k-th monitoring point Converted to interferometric phase delay The specific formula is as follows: in, The incident angle of the LiDAR radar wave. This is the operating wavelength.

7. The multi-source deformation monitoring data fusion method based on scene awareness according to claim 6, characterized in that, Spatially register the GNSS time-series data, the InSAR raster data, and the LiDAR point cloud data to obtain spatially registered GNSS time-series data, InSAR raster data, and LiDAR point cloud data, including: A local coordinate system was established using the GNSS reference station as the origin to construct the BIM model of the dam body; The LiDAR point cloud data and the InSAR raster data are spatially partitioned using an octree index. Each node of the octree is a substructure of the dam body BIM model, and each substructure represents a different spatial region of the dam body. For InSAR raster data and LiDAR point cloud data at each node, using GNSS coordinates as the reference, coordinate transformation is performed on the InSAR raster point coordinates and LiDAR point cloud coordinates through the first coordinate transformation function and the second coordinate transformation function, respectively. The first spatial registration error between the InSAR raster point coordinates and GNSS coordinates is calculated, as well as the second spatial registration error between the LiDAR point cloud coordinates and GNSS coordinates. Based on the first spatial registration error and the second spatial registration error, the coordinates of the InSAR grid points and the coordinates of the LiDAR point cloud are corrected respectively.

8. The multi-source deformation monitoring data fusion method based on scene awareness according to claim 7, characterized in that, The specific formula for calculating the first spatial registration error between InSAR grid point coordinates and GNSS coordinates is as follows: in, Indicates the coordinates of the GNSS reference station. Indicates the coordinates of InSAR grid points. This represents the first coordinate transformation function. Indicates the first spatial registration error; To minimize the first spatial registration error To find the first coordinate transformation function, we need to determine the objective function. Obtain the minimized first spatial registration error ; The specific formula for calculating the second spatial registration error between LiDAR point cloud coordinates and GNSS coordinates is as follows: in, Represents LiDAR point cloud coordinates. This represents the second coordinate transformation function. Indicates the first spatial registration error; To minimize the second spatial registration error To find the objective function, solve for the second coordinate transformation function. To obtain the minimized second spatial registration error .

9. A method for fusing multi-source deformation monitoring data based on scene awareness according to claim 8, characterized in that, Based on atmospheric delay-corrected InSAR raster data and spatially registered GNSS time-series data, InSAR raster data, and LiDAR point cloud data, the deformation values ​​of the dam body are calculated, referred to as the first deformation value, the second deformation value, and the third deformation value, respectively, including: Unify the GNSS station coordinates, InSAR raster pixel coordinates, and LiDAR point cloud coordinates into a local coordinate system with the GNSS reference station as the origin, and align them with the spatial structure of the dam body BIM model. In the dam body BIM model, a key point of the dam body to be monitored is selected. The coordinates of the key point of the dam body BIM model are synchronously mapped to the GNSS network, InSAR deformation raster map and LiDAR point cloud dataset through the established coordinate transformation relationship to obtain the target monitoring point at the same physical location. Based on the three-dimensional coordinates of the target monitoring point, calculate the linear rate of change of GNSS data in the elevation direction within the specified monitoring period T, and use the linear rate of change in the elevation direction as the first deformation value. Based on the three-dimensional coordinates of the target monitoring point, the vertical linear change rate of the InSAR raster data within the specified monitoring period T is calculated, and the vertical linear change rate is used as the second deformation value. Based on the three-dimensional coordinates of the target monitoring point, the cumulative elevation change of the LiDAR point cloud data within the monitoring period T is calculated, and the cumulative elevation change is used as the third deformation value.

10. A multi-source deformation monitoring data fusion system based on scene awareness, characterized in that, The system includes: The data acquisition module is used to acquire GNSS time-series data, InSAR raster data, LiDAR point cloud data, and ECMWF meteorological data of the dam monitoring area; The extraction module is used to extract a multi-dimensional scene feature vector containing signal quality dimension and environmental geometry dimension based on the GNSS time series data, the InSAR raster data, the LiDAR point cloud data and the ECMWF meteorological data. The first calculation module is used to calculate, based on the multi-dimensional scene feature vector and a preset physical threshold indication function, the first weight, the second weight, and the third weight corresponding to the GNSS time series data, the InSAR raster data, and the LiDAR point cloud data, respectively. The second calculation module calculates the first deformation value, the second deformation value, and the third deformation value of the dam body monitoring point based on the GNSS time series data, the InSAR raster data, and the LiDAR point cloud data, respectively. The fusion module is used to fuse the first deformation value, the second deformation value, and the third deformation value according to the first weight, the second weight, and the third weight to obtain the fused deformation value of the dam body monitoring point.

Images