Three-dimensional deformation monitoring method and system for power transmission line, equipment and storage medium

Abstract

Embodiments of the present application disclose a kind of power transmission line three-dimensional deformation monitoring method and system, equipment and storage medium, method includes: using the preliminary elevation estimate of permanent scatterer point on power transmission line and in the plane coordinate of multi-temporal synthetic aperture radar image, construct the three-dimensional space network of permanent scatterer point;For each arc segment of three-dimensional space network, establish fine phase model, and calculate the coherence coefficient of the observation value of each arc segment differential interferometric phase and the fitting value of differential interferometric phase under different model parameter combinations, to generate the coherence coefficient solution space of each arc segment;Using the preset adaptive threshold detection method, according to the local background statistical characteristics of coherence coefficient in coherence coefficient solution space, dynamic decision threshold is determined to identify high-quality deformation arc segment;According to the model parameters corresponding to high-quality deformation arc segment, the three-dimensional deformation of power transmission line is determined.A kind of high-precision, strong robustness improved time series interferometric synthetic aperture radar deformation monitoring method is realized.

Description

Technical Field

[0001] This invention relates to the field of power transmission line deformation monitoring technology, and in particular to a three-dimensional deformation monitoring method, system, equipment and storage medium for power transmission lines. Background Technology

[0002] Temporal interferometric synthetic aperture radar (TIAR) technology, particularly its permanent scatterer interferometry (PSI), is an effective means of achieving high-precision monitoring of surface deformation by analyzing the phase information of long-term radar image sequences. It has been widely applied in fields such as urban subsidence and landslide monitoring. However, when this technology is directly applied to slender targets with large spatial spans and non-rigid structures, such as power transmission lines, significant technical drawbacks exist.

[0003] First, the deformation of transmission lines (such as foundation settlement and tower tilt) and elevation information are severely coupled in the interference phase of long-term time series. Although existing technologies perform elevation phase compensation, traditional separation strategies based on steady-state assumptions are difficult to effectively decouple, leading to mutual contamination of elevation and deformation results and reducing monitoring accuracy. Second, the scattering characteristics of transmission lines fluctuate drastically due to environmental factors such as wind load and temperature changes. Existing technologies typically use a fixed coherence coefficient threshold to screen high-quality monitoring points, but in cases of signal instability, this method is prone to mistakenly rejecting useful signal points or retaining unstable noise points, affecting the reliability of monitoring results. Even with adaptive detection methods, such as constant false alarm rate detection, the screening effect still needs improvement when faced with the complex signal characteristics of transmission lines. Furthermore, existing methods are mostly based on constructing scattering point networks in a two-dimensional plane, such as constructing a two-dimensional irregular triangular network. This ignores the three-dimensional geometric relationship of transmission towers and conductors in the vertical direction, resulting in insufficient spatial constraints and errors that easily accumulate and propagate in complex terrain. Finally, existing models do not adequately characterize the temperature effect and fail to accurately model the thermal expansion and contraction of conductors and changes in sag caused by temperature variations. This introduces systematic biases and further affects monitoring accuracy. Summary of the Invention

[0004] The main objective of this invention is to provide a method, system, device, and storage medium for three-dimensional deformation monitoring of transmission lines, which can solve the problem of the lack of high-precision and robust three-dimensional deformation monitoring methods for transmission lines in scenarios with strong non-rigidity and strong disturbance in the existing technology.

[0005] To achieve the above objectives, the first aspect of the present invention provides a three-dimensional deformation monitoring method for transmission lines, the method comprising: A preliminary elevation estimate of the permanent scatterer points on the transmission line is determined; the preliminary elevation estimate is obtained by simplified elevation inversion based on small time baseline image pairs, which are image pairs selected from the multi-temporal synthetic aperture radar images of the transmission line with time baselines less than a preset threshold. Using the preliminary elevation estimates of the permanent scatterer points and the planar coordinates of the permanent scatterer points in the multi-temporal synthetic aperture radar image, a three-dimensional spatial network of the permanent scatterer points is constructed. The three-dimensional spatial network is used to reflect the spatial constraint relationship of the permanent scatterer points of the transmission line geometry. Based on the three-dimensional spatial network, a fine phase model is established for each arc segment of the three-dimensional spatial network, and the coherence coefficient of the observed value of the differential interferometric phase of each arc segment and the fitted value of the differential interferometric phase under different combinations of model parameters are calculated to generate the coherence coefficient solution space of each arc segment. The fine phase model is used to express the model parameters that decompose the differential interferometric phase into at least residual elevation components, deformation components and temperature effect deformation components. The observed value of the differential interferometric phase is obtained by interferometric processing based on the multi-temporal synthetic aperture radar image. Using a preset adaptive threshold detection method, a dynamic decision threshold is determined in the coherence coefficient solution space based on the local background statistical characteristics of the coherence coefficient, and high-quality deformed arc segments are identified using the dynamic decision threshold and the coherence coefficient solution space of each arc segment. The three-dimensional deformation of the transmission line is determined based on the model parameters corresponding to the high-quality deformed arc segment.

[0006] To achieve the above objectives, a second aspect of the present invention provides a three-dimensional deformation monitoring system for transmission lines, the system comprising: A preliminary elevation unit is used to determine the preliminary elevation estimate of permanent scatterer points on the transmission line. The preliminary elevation estimate is obtained by simplified elevation inversion based on small time baseline image pairs. The small time baseline image pairs are image pairs selected from the multi-temporal synthetic aperture radar images of the transmission line with time baselines less than a preset threshold. A three-dimensional network unit is used to construct a three-dimensional spatial network of the permanent scatterer points using the preliminary elevation estimate of the permanent scatterer points and the planar coordinates of the permanent scatterer points in the multi-temporal synthetic aperture radar image. The three-dimensional spatial network is used to reflect the spatial constraint relationship of the permanent scatterer points of the transmission line geometry. The parameter estimation unit is used to establish a fine phase model for each arc segment of the three-dimensional spatial network based on the three-dimensional spatial network, and to calculate the coherence coefficient of the observed value of the differential interferometric phase of each arc segment and the fitted value of the differential interferometric phase under different combinations of model parameters, so as to generate the coherence coefficient solution space of each arc segment. The fine phase model is used to express the model parameters that decompose the differential interferometric phase into at least residual elevation components, deformation components and temperature effect deformation components. The observed value of the differential interferometric phase is obtained by interferometric processing based on the multi-temporal synthetic aperture radar image. An adaptive selection unit is used to determine a dynamic decision threshold in the coherence coefficient solution space based on the local background statistical characteristics of the coherence coefficient using a preset adaptive threshold detection method, and to identify high-quality deformed arc segments using the dynamic decision threshold and the coherence coefficient solution space of each arc segment. The result output unit is used to determine the three-dimensional deformation of the transmission line based on the model parameters corresponding to the high-quality deformed arc segment.

[0007] To achieve the above objectives, a third aspect of the present invention provides a computer-readable storage medium storing a computer program, which, when executed by a processor, causes the processor to perform the steps shown in the first aspect and any feasible implementation.

[0008] To achieve the above objectives, a fourth aspect of the present invention provides a computer device including a memory and a processor, the memory storing a computer program, which, when executed by the processor, causes the processor to perform the steps shown in the first aspect and any feasible implementation.

[0009] The embodiments of the present invention have the following beneficial effects: This invention provides a three-dimensional deformation monitoring method for transmission lines. The method includes: determining preliminary elevation estimates of permanent scatterer points on the transmission line; the preliminary elevation estimates are obtained by simplified elevation inversion based on small time baseline image pairs, wherein the small time baseline image pairs are selected from multi-temporal synthetic aperture radar (MSAR) images of the transmission line with time baselines less than a preset threshold; constructing a three-dimensional spatial network of the permanent scatterer points using the preliminary elevation estimates and the planar coordinates of the permanent scatterer points in the MSAR images, wherein the three-dimensional spatial network reflects the spatial constraints of permanent scatterer points on the geometric structure of the transmission line; and establishing fine phases for each arc segment of the three-dimensional spatial network based on the three-dimensional spatial network. The model is used to calculate the coherence coefficients of the observed differential interference phase of each arc segment and the fitted values ​​of the differential interference phase under different combinations of model parameters, so as to generate the coherence coefficient solution space of each arc segment. The fine phase model is used to express the model parameters that decompose the differential interference phase into at least residual elevation components, deformation components and temperature effect deformation components. The observed values ​​of the differential interference phase are obtained by interferometric processing of the multi-temporal synthetic aperture radar image. Using a preset adaptive threshold detection method, a dynamic decision threshold is determined in the coherence coefficient solution space according to the local background statistical characteristics of the coherence coefficient. The dynamic decision threshold and the coherence coefficient solution space of each arc segment are used to identify high-quality deformation arc segments. The three-dimensional deformation of the transmission line is determined according to the model parameters corresponding to the high-quality deformation arc segments.

[0010] The above scheme achieves several advantages. First, by using short-time baseline imagery to obtain preliminary elevation information, the coupling effect between elevation and slow deformation is effectively decoupled, reducing systematic bias at the source and significantly improving the accuracy of subsequent deformation calculations. Second, the constructed three-dimensional network model reflects the true spatial geometry of the transmission line, achieving stable scatterer connections across towers and spans, enhancing spatial constraints, effectively suppressing error propagation in complex terrain, and improving the overall robustness of the monitoring network. Furthermore, the adaptive threshold detection method dynamically adjusts the decision threshold based on local noise levels, avoiding the shortcomings of traditional fixed threshold methods that are prone to false detections and missed detections under low signal-to-noise ratio conditions, significantly improving the reliability of high-quality arc segment identification and the credibility of deformation parameter inversion. Finally, by explicitly introducing a temperature effect component into the phase model, non-rigid deformations such as conductor sag can be more accurately characterized, improving the model's adaptability to seasonal environmental changes and making the monitoring results more consistent with physical reality. This results in a high-precision, robust, and compatible improved temporal interferometric synthetic aperture radar deformation monitoring method. Attached Figure Description

[0011] To more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0012] in: Figure 1 This is a flowchart of a three-dimensional deformation monitoring method for transmission lines according to an embodiment of the present invention; Figure 2 This is a structural block diagram of a three-dimensional deformation monitoring system for a power transmission line according to an embodiment of the present invention; Figure 3 This is a structural block diagram of a computer device in an embodiment of the present invention. Detailed Implementation

[0013] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0014] It is understood that the typical scenario in which this application embodiment is applied is as follows: In this scenario, a radar satellite operates in a predetermined orbit, periodically transmitting radar beams to the ground and receiving echo signals from ground objects. Its monitoring target is a power transmission line traversing complex mountainous terrain, which mainly consists of a series of transmission towers and conductors suspended between the towers. It is understood that at different heights of the transmission towers (such as the tower base, tower body, and tower arm) and on the conductors, there are points whose radar scattering characteristics remain stable over a long period. These points are commonly referred to in the art as permanent scatterers, forming the basis for high-precision deformation measurement.

[0015] The three-dimensional deformation monitoring method for transmission lines provided in this application mainly includes three core stages: preliminary elevation estimation, three-dimensional network construction and parameter refinement, and adaptive quality control. The specific steps of the above three stages will be described in detail below.

[0016] Please see Figure 1 , Figure 1This is a flowchart illustrating a three-dimensional deformation monitoring method for transmission lines according to an embodiment of the present invention. This method can be applied to either a terminal or a server. The terminal can be a desktop terminal or a mobile terminal; a mobile terminal can be at least one of a mobile phone, tablet computer, or laptop computer. The server can be a standalone server or a server cluster composed of multiple servers. This embodiment uses a server as an example for illustration. Figure 1 The method shown includes the following steps: 101. Determine the preliminary elevation estimate of the permanent scatterer points on the transmission line; It should be noted that the preliminary elevation estimate h1 is obtained by simplified elevation inversion based on small time baseline image pairs. The small time baseline image pairs are selected from the multi-temporal synthetic aperture radar images of the transmission line, with time baselines less than a preset threshold. That is, based on the multi-temporal synthetic aperture radar images, small time baseline image pairs with time baselines less than a preset threshold are selected to generate interferometric pairs. Based on the interferometric pairs with small time baselines, the preliminary elevation estimate of the permanent scatterers on the transmission line is obtained by solving a preset simplified differential phase model, whereby the small time baseline makes the deformation component within the two imaging interval negligible.

[0017] The purpose of preliminary elevation estimation is to obtain the initial elevation values ​​of each permanent scatterer on the transmission line with minimal interference from deformation information, thereby laying the foundation for accurate separation of elevation and deformation parameters in the future.

[0018] For linearly extending structures such as transmission lines, the deformation process is typically slow and continuous. When the time baseline is too long, effects such as tower tilt, conductor sag, and foundation settlement will superimpose in the phase, increasing the uncertainty of parameter solutions. To reduce such accumulated errors and improve the elevation inversion accuracy of permanent scattering points, a shorter time baseline range must be used, ensuring that the interval between two imaging operations is short enough that the deformation components of the transmission line within that time period can be ignored or treated as constant.

[0019] Therefore, this invention employs a small time baseline strategy. The core idea of ​​this method is to limit the time baseline to a small range to effectively reduce the deformation components caused by environmental or structural disturbances between adjacent images, minimizing their impact on subsequent elevation calculations. Through this strategy, the coupling interference of deformation on elevation estimation is significantly suppressed, thereby ensuring the accuracy and stability of the elevation inversion results for transmission line areas.

[0020] After generating multiple interferometric pairs, the flat-earth phase component needs to be removed first to extract the pure differential phase. Subsequently, the SAR image time series is normalized for amplitude, and a double-threshold method for amplitude dispersion is used to identify permanent scatterer (PS) points. The identified PS points are connected using a two-dimensional Delaunay triangulation network to establish a differential phase model of the permanent scatterer, the expression of which is as follows: ; in: The radar wavelength; The slant range at which the radar reaches the target (transmission tower or conductor scatterer); Angle of incidence; The vertical baseline; Elevation of the permanent scatterer; For linear deformations (such as the displacement rate of the top of a tower or the suspension rate of a conductor). The coefficient of thermal expansion (reflects the change in conductor sag caused by temperature); This refers to the change in temperature. This is the atmospheric phase delay term; This is the residual phase term.

[0021] In the network, the two permanent scatterer points at the ends of each connecting edge are considered as a pair of adjacent PS points, used to calculate the phase difference between adjacent points. The differential phase between adjacent permanent scatterer points can be expressed as: ; When constructing a three-dimensional permanent scatterer network, it is necessary to first determine the prior information of elevation. Since this invention employs a small time baseline strategy, the minute changes in the transmission lines (such as conductor vibration or instantaneous wind-induced displacement) within the imaging period of each interferometer pair can be ignored, thus simplifying the parameter estimation process to solving for the terrain elevation corresponding to the permanent scatterer points. At this point, the differential phase (i.e., the simplified differential phase model) on each Delaunay network arc segment can be expressed as: ; Subsequently, an exhaustive search is performed in the defined solution space of elevation parameters to determine the optimal parameter values. By combining the least squares adjustment method with outlier removal techniques, a preliminary elevation estimate for each permanent scatterer point can be obtained. This method ensures high accuracy and stability in the estimation of permanent scatterer elevations, providing reliable basic data support for the monitoring of deformations such as tower structures, conductor sag, and foundation settlement in complex transmission line areas.

[0022] In one feasible implementation, determining the preliminary elevation estimate of the permanent scatterer on the transmission line includes: acquiring multi-temporal synthetic aperture radar (SAR) images of the transmission line; selecting small-time-baseline image pairs from the SAR images whose time baselines are less than a preset baseline threshold, generating an interferometric pair, wherein the preset baseline threshold is used to make the deformation component within the two imaging intervals negligible; and obtaining the preliminary elevation estimate of the permanent scatterer point by performing differential interferometry processing and phase unwrapping on the interferometric pair.

[0023] Specifically, step 101 is achieved through the following steps S1 to S3: Step S1: Acquire image data. In step S1, it is necessary to acquire N time-series synthetic aperture radar (SAR) image data covering the target transmission line area. To ensure the temporal resolution and accuracy of deformation monitoring, the number of images N usually needs to be greater than 20, and the time span should preferably cover at least one complete year in order to capture seasonal deformation characteristics. It is understood that the image data can come from various radar satellite platforms, such as the European Space Agency's Sentinel-1 satellite (providing C-band data), Germany's TerraSAR-X satellite (providing X-band data), or Canada's RADARSAT series satellites. After acquiring the data, a series of preprocessing steps are required, including orbit refinement and image registration. The image with the best quality is selected as the main image, and the remaining N-1 auxiliary images are accurately registered to the geometric space of the main image to ensure that the same ground feature is located at the same pixel position in images from different time phases.

[0024] Step S2: Identify Permanent Scattering Points. On the registered time-series image stack, identify pixels with stable backscattering characteristics over a long time series; these are called permanent scattering points. Typically, permanent scattering points correspond to rigid structures on power transmission lines, such as the tower body, base, and crossarms of transmission towers, as well as certain fixed components on conductors (such as suspension clamps, vibration dampers, etc.) or stable man-made structures on the ground. A specific identification method is the amplitude dispersion method, which calculates the ratio of the standard deviation of the amplitude of each pixel over the entire time series to its mean amplitude. When this ratio is less than a preset threshold (e.g., 0.25), the pixel's scattering characteristics are considered stable, and it is considered a candidate permanent scattering point. To further improve the reliability of the selection, an amplitude threshold can be added as a constraint, requiring the average amplitude of candidate points to be higher than a certain threshold to eliminate stable dark targets with low signal-to-noise ratios. Through this step, a set of permanent scattering points with planar coordinates (e.g., x, y coordinates in a radar coordinate system or geographic coordinate system) can be obtained.

[0025] Step S3: Obtain Preliminary Elevation. Step S3 aims to assign a preliminary elevation value to each identified permanent scatterer point, thereby effectively separating elevation error from deformation signals in the subsequent phase model. As an optional implementation, this step is achieved by screening interferometric pairs with short time baselines. Specifically, from all possible interferometric pairs composed of N images, interferometric pairs with time baselines (i.e., the acquisition time interval between the main and auxiliary images) less than a preset baseline threshold are selected. This threshold can be set according to the expected maximum deformation rate of the monitored area; for example, for general foundation settlement, it can be set to 24 days or 36 days. The basic assumption is that within such a short time interval, the cumulative amount of linear deformation caused by foundation settlement, tower tilt, etc., and seasonal temperature deformation is extremely small and can be ignored. Therefore, the differential interferometric phase of these short time baseline interferometric pairs is mainly contributed by the topographic phase and the atmospheric delay phase. By performing differential interferometry (i.e., removing the reference ellipsoid phase from the interferogram and removing most of the terrain phase using an external coarse digital elevation model) and phase unwrapping (e.g., using minimum cost flow or branching methods) on the selected interferometric pairs, a system of equations can be established regarding the residual elevation of each permanent scatterer point. Solving this system of equations yields the residual elevation of each permanent scatterer point relative to the reference digital elevation model. Adding this residual elevation to the elevation of the reference digital elevation model provides its preliminary elevation information.

[0026] For these candidate points, a simplified differential phase model is established. For the arc segment connecting any two candidate points i and j, its phase on the 1st... Differential interferometric phase on the baseline interferometer pair between hours , can be represented as: ; in, It is the difference in elevation between two points Related topographic phase components, It is atmospheric delay phase difference. This refers to residual phase terms such as system noise and orbital errors. The terrain phase component can be further expressed as: ; In the formula, The operating wavelength of the radar system. For the first The vertical baseline of each interference pair The slant distance to the target. Let be the incident angle of the radar wave. Since the deformation component is assumed to be negligible, the model does not include a deformation term. By combining the phase observations on all small time-baseline interferometric pairs, an overdetermined system of equations can be constructed and solved using the least squares method or weighted least squares method, thus obtaining the elevation estimate of each permanent scatterer candidate point relative to a reference point, i.e., the preliminary elevation.

[0027] The second stage, namely the construction of the three-dimensional network and parameter refinement, will then be implemented. Using the preliminary elevation data obtained in the previous step, an observation network that accurately reflects the geometry of the transmission line will be constructed. Based on this network, a refined model will be established to solve for multiple parameters, including residual elevation, deformation rate, and temperature effects. See steps 102 and 103 below for details.

[0028] 102. Using the preliminary elevation estimate of the permanent scatterer points and the planar coordinates of the permanent scatterer points in the multi-temporal synthetic aperture radar image, construct a three-dimensional spatial network of the permanent scatterer points; The three-dimensional spatial network is used to reflect the spatial constraint relationship of permanent scatterers in the geometric structure of the transmission line; the construction of the three-dimensional spatial network of the permanent scatterer points includes: using the three-dimensional coordinates of the permanent scatterer points to construct a three-dimensional Delaunay triangulation network to generate a set of three-dimensional arc segments connecting each permanent scatterer point, wherein the three-dimensional coordinates include preliminary elevation estimates and planar coordinates.

[0029] After completing the initial elevation estimation, a 3D permanent scatterer network (3-D PS Network) is constructed. This process integrates the spatial location information of each permanent scatterer point in the SAR image, and its expression is as follows: ; in: Represents a 3D Delaunay triangulation network; For Delaunay triangulation functions; The planar coordinates of the permanent scatterer point in the SAR image; These are the preliminary elevation values ​​obtained through hourly time baseline estimation.

[0030] A three-dimensional permanent scatterer network is constructed by incorporating the elevation information of permanent scatterer points. Compared with the traditional two-dimensional network, the three-dimensional network introduces the spatial geometric relationship between transmission towers and conductors, making the connections between permanent scatterer points much closer. This feature is mainly reflected in two aspects: (1) The number of connecting arc segments has increased significantly, and more permanent scattering points at different elevation levels (such as the top of the tower, the middle section of the conductor and the anchor point) are effectively connected through spatial information; (2) The connection methods are more diversified. The interconnection between permanent scattering points can span adjacent towers and different span conductor areas, thus forming a more complex and denser arc network.

[0031] These arcs span both horizontal and vertical dimensions, effectively enhancing the spatial integrity and resolution of transmission line deformation monitoring. By introducing three-dimensional spatial information, the constructed permanent scatterer network not only suppresses the error propagation effect caused by terrain undulations and reduces the average residual phase standard deviation, but also significantly improves the overall monitoring accuracy and system robustness, providing reliable spatial structural support for transmission line deformation analysis under complex terrain.

[0032] Specifically, step 102 is to construct a three-dimensional spatial network. This is based on the planar coordinates (x, y) of the permanent scatterer points obtained in step S2 and the preliminary elevation information. A three-dimensional network model (i.e., a three-dimensional spatial network) is constructed to represent the true spatial adjacency relationships between permanent scatterer points. The planar coordinates of each permanent scatterer can be obtained from the row and column numbers and geocoding information of radar imagery. It should be noted that in a two-dimensional network, all permanent scatterer points are projected onto the same plane and connected by two-dimensional connecting arcs. This connection method ignores elevation differences and cannot accurately reflect the spatial structure of transmission towers and conductors. However, in this embodiment, the three-dimensional coordinates (x, y, ...) of each permanent scatterer point are used... A three-dimensional Delaunay triangulation network is constructed. This network generates a series of three-dimensional connecting arcs that link adjacent permanent scattering body points, forming a set of non-overlapping three-dimensional tetrahedra. It is understood that the three-dimensional Delaunay triangulation has the "empty circumsphere" property, meaning that the circumsphere of each tetrahedron does not contain any other permanent scattering body points, ensuring that the network connects spatially nearest points. The three-dimensional network can establish connections between permanent scattering body points located at different elevations, such as connecting points on the base, body, and top of the same transmission tower, or establishing span-span connections between points on a conductor and points on adjacent transmission towers. This three-dimensional network model provides stronger spatial constraints for subsequent parameter calculations, effectively suppressing the accumulation and propagation of errors in complex terrain.

[0033] 103. Based on the three-dimensional spatial network, establish a fine phase model for each arc segment of the three-dimensional spatial network, and calculate the coherence coefficient of the observed value of the differential interference phase of each arc segment and the fitted value of the differential interference phase under different model parameter combinations, so as to generate the coherence coefficient solution space of each arc segment. The fine phase model is used to express the model parameters that decompose the differential interferometric phase into at least residual elevation components, deformation components, and temperature effect deformation components. The observed values ​​of the differential interferometric phase are obtained by interferometric processing of the multi-temporal synthetic aperture radar image. It should be noted that after the 3D network construction was completed, a refined inversion of elevation and deformation parameters was immediately performed. Using time-series differential interferometry, the differential phase was first calculated, and then the phase component caused by the initial elevation was subtracted from the total phase of each permanent scatterer (PS) point according to a formula. After this correction step, each PS point obtained a new phase value, which mainly reflects the residual deformation signal. The new permanent scatterer phase... It can be represented as: ; Since the deformation of transmission lines typically ranges from centimeters to decimeters, after eliminating the initial elevation phase, the elevation of each permanent scatterer point needs further refinement. Therefore, a deformation phase model (i.e., a refined phase model) incorporating residual elevation, linear deformation, and thermal expansion coefficient is established for each network arc segment. Its expression is as follows: ; Based on the differential phase model established by the formula, parameters are estimated for each connecting arc segment in the network, with a focus on optimizing key parameters such as elevation, linear settlement rate, and conductor thermal expansion coefficient. This refined modeling process lays a solid foundation for subsequent deformation analysis of transmission tower tilt, foundation settlement, and conductor sag changes. Finally, the elevation of each permanent scattering body point is given by the sum of the two solutions: .

[0034] It is evident that after completing the spatial network construction and phase difference calculation, the solution for the deformation parameters requires the selection of an appropriate estimation method, which directly affects the accuracy and stability of the final results. This application employs an adaptive selection of the optimal model coherence coefficient based on constant false alarm rate (CFAR). Its core mathematical form is the calculation formula for the coherence coefficient, defined as follows: ; in, This represents the phase difference between the observed value and the fitted value; This represents the number of interferometric images in the time series.

[0035] In the framework of permanent scatterer interferometry, to accurately evaluate and screen high-quality arc segments in transmission line deformation monitoring, this invention transforms the arc segment quality assessment problem into a peak detection problem in a three-dimensional solution space. By analyzing the distribution characteristics of the coherence coefficient, intelligent identification and retention of stable scatterer arc segments are achieved.

[0036] The core idea of ​​this method is to detect the peak position of the coherence coefficient in the three-dimensional solution space and retain those tower or conductor arc segments that show high stability and consistency in the time series, thereby significantly improving the accuracy and robustness of deformation parameter estimation.

[0037] Specifically, a refined phase model is established and the coherence coefficient is calculated in step 103. Based on the constructed three-dimensional network model (the aforementioned three-dimensional spatial network), a refined differential interferometric phase model is established for each three-dimensional connecting arc segment (connecting points i and j) in the network. For the interferogram formed by the k-th image, the differential phase observations on arc segment (i, j) are... It can be represented as the sum of multiple components: ; in: It can also be described as at any given time as Differential phase on the image (relative to the main image); This is the residual elevation component, caused by the inaccuracy of the initial elevation information, and it is perpendicular to the spatial baseline. and residual elevation difference Proportional: Here, R is the radar wavelength, and R is the slant range. It is the radar incident angle; As a deformation component, in this embodiment, we mainly consider the linear deformation along the radar line of sight, which is related to the time baseline. and the difference in linear deformation rate Proportional: ; This is the temperature-effect deformation component, used to characterize the thermal expansion and contraction and sag changes of transmission lines (especially conductors) caused by temperature variations. This component introduces the coefficient of thermal expansion. Temperature difference at the time of acquisition of each image To model: Among them, temperature data T k Information can be obtained from weather stations near the monitoring area; and These represent the atmospheric delay phase and the noise phase, respectively, and are treated as random error terms in the subsequent calculation of the coherence coefficient. Accordingly, the modeled phase can be expressed as: ; In the formula, Residual elevation components; For deformation components; This represents the temperature-dependent deformation component. Let be the differential interference phase of the interferogram formed by the k-th image pair of arc segment (i, j); The radar wavelength; The slant range of the radar to the permanent scatterer; Angle of incidence; Let be the vertical baseline of the k-th image pair; Let (i, j) be the residual elevation difference of arc segment (i, j). Let (i, j) be the difference in linear deformation rate of arc segment (i, j). The time baseline is the time interval between the kth image and the main image; The difference in the coefficients of thermal expansion between arc segments (i, j); The temperature difference between the k-th image pairs; The combination of model parameters to be solved includes , , .

[0038] Subsequently, in response to ( , , These three unknown parameters are searched within a predefined three-dimensional solution space. For example, The search range can be set to [-50 meters, +50 meters]. The search range can be set to [-30 mm / year, +30 mm / year]. The search range can be set to [-0.5 mm / °C, +0.5 mm / °C]. After discretizing the solution space into a series of grid points, for each grid point representing a set of parameters ( , , ), calculate its time series coherence coefficient. In the three-dimensional solution space ( , , In ), coherence coefficient The calculation formula is: ; In the formula, j is the imaginary unit, and N is the number of images. By traversing and calculating all parameter combinations in the solution space, a three-dimensional coherence coefficient cube, i.e., the coherence coefficient solution space, can be generated for each arc segment.

[0039] The calculation of the coherence coefficient includes: for each arc segment in the network, the system traverses every grid point in its three-dimensional parameter solution space, i.e., every parameter combination ( , , Substituting this parameter combination into the refined phase model, the predicted phase time series is calculated. (That is, the fitted value under this set of model parameters). Then, the model-predicted phase is compared with the actual observed interferometric phase. Compare and calculate the time coherence coefficient value corresponding to this parameter combination. By traversing and calculating all parameter combinations in the solution space, a three-dimensional coherence coefficient cube, i.e., the coherence coefficient solution space, can be generated for each arc segment. The coherence coefficient value... The value of is between 0 and 1. The closer the value is to 1, the better the model's predicted phase matches the actual observed phase. After traversing all parameter combinations, a three-dimensional coherence coefficient matrix is ​​obtained, where the value of each element corresponds to the quality of a parameter combination. The coherence coefficient solution space includes the coherence coefficient matrix.

[0040] The next stage is the third stage, which is adaptive quality control. Its purpose is to accurately identify the optimal combination of parameters representing the real physical process from the noisy solution space and to select high-quality deformation segments. See steps 104 and 105 below for details.

[0041] 104. Using a preset adaptive threshold detection method, a dynamic decision threshold is determined in the coherence coefficient solution space based on the local background statistical characteristics of the coherence coefficient, and the dynamic decision threshold and the coherence coefficient solution space of each arc segment are used to identify high-quality deformed arc segments. As described above, this application provides an adaptive selection scheme for the optimal model coherence coefficient based on constant false alarm rate (CFAR). In step 104, a preset adaptive threshold detection method is used to determine a dynamic decision threshold in the coherence coefficient solution space based on the local background statistical characteristics of the coherence coefficients. Then, the dynamic decision threshold and the coherence coefficient solution space of each arc segment are used to identify high-quality deformed arc segments. This adaptive threshold detection method can be a constant false alarm rate (CFAR) processing method.

[0042] Among them, the optimal coherence coefficient adaptive selection method based on constant false alarm rate (CFAR) is used to dynamically detect and adjust the threshold of the entire solution space. In the PSI deformation inversion process, the solution space contains all possible parameter combinations, such as key variables like linear settlement rate, elevation error, and conductor thermal expansion coefficient.

[0043] By introducing a CFAR detection mechanism into the solution space, the detection threshold can be automatically adjusted under varying noise and signal conditions to maintain a constant false alarm rate. This allows for adaptive identification of highly coherent and reliable deformation segments without increasing the risk of false detection. This method effectively enhances the stability and reliability of multi-temporal deformation parameter estimation for transmission lines, providing crucial technical support for high-precision, all-weather transmission line safety monitoring.

[0044] The time-domain coherence coefficient distribution is an important indicator for evaluating the stability and coherence of phase information of permanent scatterer points in transmission line deformation monitoring. It exhibits representative distribution characteristics in the solution space. High-quality deformation arcs typically show a clear main peak in the distribution, while unstable arcs affected by factors such as electromagnetic interference, wind load vibration, or thermal expansion and contraction correspond to lower side peaks.

[0045] To effectively screen for high-quality deformable segments, constant false alarm rate (CFAR) detection is first applied at each sampling point in the solution space. By setting an appropriate detection threshold, sidelobes with low temporal coherence coefficients can be removed, thereby reducing the impact of noise and environmental disturbances on the segment quality assessment. After this processing, a refined subspace rich in high-confidence scatterer connectivity is obtained.

[0046] The three-dimensional CFAR detection process can be briefly summarized as the following steps (1), (2), (3) and (4): (1) Define three-dimensional reference units: Protective units and reference units are established around the detection unit (Cell Under Test, CUT), and these units are distributed in different directions in three-dimensional space. The number and range of reference units are set according to the spatial layout of the transmission line and the imaging geometry to cover the multi-height scattering areas such as conductors and towers.

[0047] (2) Estimating background noise: The background noise is estimated by calculating the average coherence coefficient of all reference cells. This value is used as the benchmark for subsequent detection. The calculation formula is as follows: ; in, Total number of reference units; For the first The coherence coefficient values ​​of each reference unit.

[0048] (3) Set the detection threshold: The background noise estimate is proportional to the threshold factor. Multiply them to determine the detection threshold of the current detection unit: ; This threshold is used to distinguish highly coherent targets from noise signals.

[0049] (4) Target detection: Compare the coherence coefficient of the detection unit with the set threshold: If the coherence coefficient is greater than the threshold, the arc segment corresponding to the unit is determined to be a high-quality deformed arc segment and retained; if the coherence coefficient is lower than the threshold, it is removed.

[0050] The CFAR-based adaptive detection strategy ensures that only towers and conductor segments with excellent coherence and the highest stability are selected for subsequent analysis. By suppressing noise and side-peak effects caused by terrain undulations, climatic conditions, and electromagnetic interference, this method significantly improves the overall accuracy and reliability of transmission line time-series deformation monitoring.

[0051] In high signal-to-noise ratio (SNR) scenarios, significant peaks in the solution space are easily identified, with prominent main peak characteristics that contrast sharply with surrounding sampling points. This main peak corresponds to the optimal time-domain coherence coefficient, thus the obtained results are consistent with the optimal solution of the traditional "maximum time-domain coherence coefficient method," enabling accurate identification of tower and conductor deformation. However, under low SNR conditions, high-quality sampling points are often masked by low-quality points, resulting in significant noise interference. If the maximum time-domain coherence coefficient method is still used in this situation, low-quality points affected by wind loads or thermal expansion effects may be mistakenly identified as high-quality points, leading to errors in deformation parameter estimation.

[0052] To address this issue, this invention introduces a three-dimensional CFAR detection mechanism. This mechanism dynamically adjusts the detection threshold based on the background noise level and performs signal identification within the three-dimensional solution space, effectively distinguishing between real signal points and spurious noise points, thereby narrowing the search range and eliminating invalid points. After this process, the system can more accurately extract the true optimal solution from the refined solution space.

[0053] In this invention, a targeted search is performed on the main peak of the time-domain coherence coefficient within the refined solution space, and high-quality deformation arc segments are extracted by identifying local maxima. To further improve the stability of the screening results, an additional judgment condition is set: if no other peak with a similar amplitude to the main peak exists in the remaining solution space, the arc segment is considered to have high temporal stability, and its deformation parameter estimation is reliable. This method effectively reduces the risk of misjudgment and incorrect rejection due to external interference, uneven terrain scattering, or meteorological changes.

[0054] Unlike traditional fixed threshold methods, the adaptive strategy proposed in this invention can dynamically adjust parameters based on data characteristics under different scenarios. By combining CFAR detection and peak identification technology, adaptive evaluation and optimized selection of deformation arc segments of transmission lines can be achieved. (1) CFAR detection adaptively sets the detection threshold based on temporal coherence characteristics, filters out low coherence sidelobes and retains potential high-quality arc segments; (2) Identify the main peak with a high time domain coherence coefficient in the refined solution space, and perform secondary verification and robustness confirmation on the high-quality arc segment.

[0055] This comprehensive strategy significantly improves the accuracy and reliability of transmission line deformation segment analysis, providing strong technical support and application potential for line safety monitoring under conditions such as complex terrain, wind vibration, and temperature changes.

[0056] In one feasible implementation, the solution space of the coherence coefficients for each arc segment contains a large number of discrete units, and each discrete unit corresponds to a set of model parameters. , , Given the coherence coefficients under the model parameters, the step of determining a dynamic decision threshold in the coherence coefficient solution space based on the local background statistical characteristics of the coherence coefficients using a preset adaptive threshold detection method includes the following steps A01 to A03: A01. For each unit to be detected in the coherence coefficient solution space, determine the reference unit corresponding to the unit to be detected. The reference unit is a discrete unit set within a preset range of the unit to be detected. The unit to be detected is any discrete unit. A02. Calculate the average value of the coherence coefficients in all reference cells to obtain the background noise estimate of the cell to be detected. The local background statistical characteristics include the background noise estimate. A03. Multiply the estimated background noise value by a preset scaling factor to determine the dynamic decision threshold of the unit to be detected.

[0057] It should be noted that adaptive threshold detection is used to identify high-quality arc segments. Traditional fixed threshold methods (e.g., >0.7) Difficult to adapt to complex noise backgrounds. This embodiment employs an adaptive threshold detection method, namely a cell average constant false alarm rate detector, to process the coherence coefficient solution space generated in step 103. The principle of adaptive threshold detection is as follows: for each cell to be detected in the coherence coefficient solution space, a "protection cell" and a "reference cell" are defined around it. The protection cell is a small region adjacent to the cell to be detected, and this region is set to prevent the energy of the target signal itself from leaking into the background estimation; the reference cell is the outermost region surrounding the protection cell. The detection method first calculates the average coherence coefficient values ​​of all cells within the reference cell, and uses this average value as an estimate of the local background noise level at the current location of the cell to be detected (i.e., the background noise estimate). Then, the local background noise estimate is... Multiplying by a preset scaling factor K (e.g., K=1.5) yields a dynamic detection threshold S (also known as a dynamic decision threshold). Finally, the coherence coefficient value of the unit to be detected is compared with this dynamic detection threshold S. If the coherence coefficient value of the unit to be detected is higher than this threshold, the unit is considered a signal significantly different from background noise, i.e., the main peak of the coherence coefficient, and its corresponding arc segment is identified as a high-quality deformation arc segment; otherwise, the unit is considered to belong to background noise. By scanning the entire coherence coefficient solution space point by point, all high-quality deformation arc segments can be screened out.

[0058] Furthermore, the method of identifying high-quality deformed arc segments using dynamic decision thresholds and the coherence coefficient solution space of each arc segment includes: if the coherence coefficient of the unit to be detected is greater than the dynamic decision threshold, then the arc segment is retained and determined to be a high-quality deformed arc segment; if the coherence coefficient of the unit to be detected is not greater than the dynamic decision threshold, then the arc segment is discarded.

[0059] 105. Determine the three-dimensional deformation of the transmission line based on the model parameters corresponding to the high-quality deformed arc segment.

[0060] In one feasible implementation, determining the three-dimensional deformation of the transmission line based on the model parameters corresponding to the high-quality deformation arc segment includes: for each high-quality deformation arc segment, determining the maximum coherence coefficient from the detection units whose coherence coefficient is greater than the dynamic decision threshold; and using the model parameters corresponding to the maximum coherence coefficient of each high-quality deformation arc segment to determine the three-dimensional deformation of the transmission line.

[0061] Understandably, performing the above detection process on all elements in the entire three-dimensional coherence coefficient matrix will yield a set of high-quality solutions that pass the detection. Among these solutions, the parameter combination corresponding to the solution with the largest coherence coefficient value is selected. , , The residual elevation, deformation rate, and thermal expansion coefficient of the arc segment are used as the final residual elevation, deformation rate, and thermal expansion coefficient. Arc segments that pass this step are considered high-quality deformed arc segments and are retained for subsequent overall network adjustment and deformation analysis, while arc segments that fail the test are discarded.

[0062] For example, step 105, determining the three-dimensional deformation, can specifically be as follows: For each arc segment identified as having high-quality deformation, extract the model parameters corresponding to the point with the largest coherence coefficient value in the coherence coefficient solution space (i.e., the main peak of the coherence coefficient). , , These parameters represent the optimal relative residual elevation, relative linear deformation rate, and relative thermal expansion coefficient between two permanent scatterer points on the arc segment. By adjusting these relative quantities across all high-quality arc segments in the entire 3D network (e.g., least-squares adjustment using a stable reference point as a baseline), the absolute linear deformation rate v and absolute thermal expansion coefficient k of each permanent scatterer point relative to that reference point can be obtained. T At the same time, the calculated residual elevation will be... Adding the preliminary elevation obtained in step 101 to the final elevation yields a more accurate result. Finally, the linear deformation rate v of each permanent scatterer point is output as its subsidence or uplift deformation information. Combined with its precise three-dimensional coordinates, three-dimensional deformation monitoring of the transmission line can be achieved.

[0063] In summary, through the above three stages of processing, this embodiment can effectively decouple elevation and deformation signals, enhance spatial constraints using a three-dimensional network, and adaptively filter out reliable deformation information. Finally, it outputs the precise three-dimensional coordinates of each permanent scatterer in the transmission line corridor, the millimeter-level linear deformation rate, and the temperature-related deformation parameters, thereby achieving high-precision and high-robustness monitoring of the health status of the transmission line.

[0064] This invention provides a three-dimensional deformation monitoring method for transmission lines. First, preliminary elevation information is obtained by using short-time baseline imagery, effectively decoupling the coupling effect between elevation and slow deformation, reducing systematic bias at the source, and significantly improving the accuracy of subsequent deformation calculations. Second, the constructed three-dimensional network model reflects the true spatial geometry of the transmission line, realizing stable scattering body connections across towers and spans, enhancing spatial constraints, effectively suppressing error propagation in complex terrain, and improving the overall robustness of the monitoring network. Furthermore, an adaptive threshold detection method is employed, which dynamically adjusts the decision threshold based on local noise levels, avoiding the shortcomings of traditional fixed threshold methods that are prone to false detections and missed detections under low signal-to-noise ratio conditions, significantly improving the reliability of high-quality arc segment identification and the credibility of deformation parameter inversion. Finally, by explicitly introducing a temperature effect component into the phase model, non-rigid deformations such as conductor sag can be more accurately characterized, improving the model's adaptability to seasonal environmental changes and making the monitoring results more consistent with physical reality. This led to the development of an improved time-series interferometric synthetic aperture radar deformation monitoring method that is highly accurate, robust, and compatible with existing interferometric measurement procedures.

[0065] In another embodiment, the main difference lies in the specific method of constructing the three-dimensional network model. The three-dimensional network unit is constructed using three-dimensional Delaunay triangulation. However, in some cases, such as when the distribution of permanent scatterer points is extremely uneven, Delaunay triangulation may produce some slender tetrahedrons with poor geometry, thereby affecting the stability of the network.

[0066] In this embodiment, the "construction of a three-dimensional network" adopts a strategy based on the K-nearest neighbor algorithm. The specific process is as follows: After obtaining the three-dimensional coordinates (x, y, h) of all permanent scatterer points, for each permanent scatterer point Pi, the distance d between it and all other permanent scatterer points Pi in three-dimensional Euclidean space is calculated. ij Then, for all distances d ij Sort the points and find the K nearest neighbors to Pi, where K is a preset integer, such as 8 or 10. Finally, connect Pi to these K nearest neighbors to form K three-dimensional connecting arcs. Repeat this process for all permanent scatterer points. All generated arcs together constitute the final three-dimensional network model (i.e., a three-dimensional spatial network).

[0067] The advantage of using the K-nearest neighbor algorithm to construct a network is that it can ensure that each node (permanent scatterer point) in the network has a fixed connectivity (i.e., K), avoiding isolated points in sparse areas of the point cloud or over-connection in dense areas of the point cloud, making the network structure more uniform and robust.

[0068] The other steps in this embodiment are the same as those described above. This embodiment shows that the method for constructing a three-dimensional network model is not limited to Delaunay triangulation; any method that can characterize adjacency relationships in three-dimensional space falls within the protection scope of this application.

[0069] In another embodiment, the main difference lies in the specific implementation of the adaptive threshold detection method. When the adaptive selection uses a cell-average constant false alarm rate detector, this method works well when the background noise follows a Gaussian distribution and is relatively stable. However, if other strong interference signals or noise spikes are mixed into the reference cell, the calculated average value will be too high, resulting in an excessively high dynamic detection threshold (i.e., dynamic decision threshold), which may miss the true, but slightly weaker, main peak of the coherence coefficient.

[0070] To address this complex background, this embodiment employs an ordered statistical constant false alarm rate (CFAR) detector in the "adaptive detection" process. The specific procedure is as follows: Similar to the cell-averaged CFAR detector, for each cell to be detected in the coherence coefficient solution space, its surrounding guard cells and reference cells are defined. However, instead of calculating the average of all coherence values ​​within the reference cell, the adaptive cell selection first sorts all M coherence coefficient values ​​within the reference window in ascending or descending order. Then, the value located at a specific percentile (or a specific rank k) after sorting is selected as an estimate of the local background noise. For example, the value located at the 75th percentile (i.e., the 0.75*Mth value) after sorting can be selected as the noise estimate. This approach effectively eliminates the influence of a few extremely high values ​​(interference) within the reference window. This is achieved when noise estimation is obtained. Then, it is multiplied by a scaling factor K to obtain the dynamic detection threshold S, i.e., S = K * Z. The subsequent comparison and decision process is the same as the aforementioned implementation process.

[0071] The use of an ordered statistical constant false alarm rate detector can significantly improve detection performance in noisy backgrounds with non-uniformity or outlier interference points, reduce the false alarm rate, and thus more reliably identify high-quality deformed arc segments. The other steps in this embodiment are exactly the same as described above. This embodiment shows that the specific implementation of the adaptive threshold detection method is not limited to the cell averaging method; other methods based on local background statistical characteristics, such as ordered statistical methods and maximum / minimum selection methods, are also within the scope of protection of this application.

[0072] Compared with existing technologies, the technical solution provided in this application has the following beneficial effects: First, by acquiring preliminary elevation information in advance, the coupling effect between elevation and slow deformation is effectively decoupled, reducing systematic biases at the source and significantly improving the accuracy of deformation calculation. Second, the constructed three-dimensional network model explicitly encodes the real spatial geometric relationship of the transmission line, realizing stable scattering body connections across towers and spans, enhancing spatial constraints, effectively suppressing the propagation of errors in complex terrain, and improving the overall robustness of the monitoring network. Furthermore, the adaptive threshold detection method can dynamically adjust the decision threshold according to the local noise level, avoiding the defects of traditional fixed threshold methods that are prone to false detection and missed detection under low signal-to-noise ratio conditions, and significantly improving the reliability of high-quality arc segment identification and the credibility of deformation parameter inversion. Finally, by explicitly introducing a temperature effect component into the phase model, non-rigid deformations such as conductor sag can be more accurately characterized, improving the model's adaptability to seasonal environmental changes and making the monitoring results more consistent with physical reality.

[0073] Please see Figure 2 , Figure 2 This is a structural block diagram of a three-dimensional deformation monitoring system for transmission lines according to an embodiment of the present invention, as shown below. Figure 2 The system shown includes: Preliminary elevation unit 201 is used to determine the preliminary elevation estimate of permanent scatterers on the transmission line; the preliminary elevation estimate is obtained by simplified elevation inversion based on small time baseline image pairs, the small time baseline image pairs are image pairs selected from the multi-temporal synthetic aperture radar images of the transmission line with time baselines less than a preset threshold. The three-dimensional network unit 202 is used to construct a three-dimensional spatial network of the permanent scatterer points using the preliminary elevation estimate of the permanent scatterer points and the planar coordinates of the permanent scatterer points in the multi-temporal synthetic aperture radar image. The three-dimensional spatial network is used to reflect the spatial constraint relationship of the permanent scatterer points of the transmission line geometry. The parameter estimation unit 203 is used to establish a fine phase model for each arc segment of the three-dimensional spatial network based on the three-dimensional spatial network, and to calculate the coherence coefficient of the observed value of the differential interferometric phase of each arc segment and the fitted value of the differential interferometric phase under different combinations of model parameters, so as to generate the coherence coefficient solution space of each arc segment. The fine phase model is used to express the model parameters that decompose the differential interferometric phase into at least residual elevation components, deformation components and temperature effect deformation components. The observed value of the differential interferometric phase is obtained by interferometric processing based on the multi-temporal synthetic aperture radar image. The adaptive selection unit 204 is used to determine a dynamic decision threshold in the coherence coefficient solution space based on the local background statistical characteristics of the coherence coefficient using a preset adaptive threshold detection method, and to identify high-quality deformed arc segments using the dynamic decision threshold and the coherence coefficient solution space of each arc segment. The result output unit 205 is used to determine the three-dimensional deformation of the transmission line based on the model parameters corresponding to the high-quality deformed arc segment.

[0074] It should be noted that, Figure 2 The function of each unit in the system shown Figure 1 The steps in the method shown are similar, and will not be repeated here to avoid repetition. For details, please refer to [reference needed]. Figure 1 The content of each step in the method shown.

[0075] This invention provides a three-dimensional deformation monitoring system for transmission lines. First, by acquiring preliminary elevation information through short-time baseline imagery, the coupling effect between elevation and slow deformation is effectively decoupled, reducing systematic bias at the source and significantly improving the accuracy of subsequent deformation calculations. Second, the constructed three-dimensional network model reflects the true spatial geometry of the transmission line, achieving stable scattering body connections across towers and spans, enhancing spatial constraints, effectively suppressing error propagation in complex terrain, and improving the overall robustness of the monitoring network. Furthermore, the adoption of an adaptive threshold detection method dynamically adjusts the decision threshold based on local noise levels, avoiding the shortcomings of traditional fixed threshold methods that are prone to false detections and missed detections under low signal-to-noise ratio conditions, significantly improving the reliability of high-quality arc segment identification and the credibility of deformation parameter inversion. Finally, by explicitly introducing a temperature effect component into the phase model, non-rigid deformations such as conductor sag can be more accurately characterized, improving the model's adaptability to seasonal environmental changes and making the monitoring results more consistent with physical reality. This resulted in an improved time-series interferometric synthetic aperture radar deformation monitoring system that is highly accurate, robust, and compatible with existing interferometric measurement procedures.

[0076] Figure 3 An internal structural diagram of a computer device in one embodiment is shown. This computer device can specifically be a terminal or a server. Figure 3 As shown, the computer device includes a processor, memory, and a network interface connected via a system bus. The memory includes a non-volatile storage medium and internal memory. The non-volatile storage medium stores an operating system and may also store a computer program, which, when executed by the processor, causes the processor to perform the aforementioned methods. The internal memory may also store a computer program, which, when executed by the processor, causes the processor to perform the aforementioned methods. Those skilled in the art will understand that… Figure 3 The structure shown is merely a block diagram of a portion of the structure related to the present application and does not constitute a limitation on the computer device to which the present application is applied. Specific computer devices may include more or fewer components than those shown in the figure, or combine certain components, or have different component arrangements.

[0077] In one embodiment, a computer device is provided, including a memory and a processor, the memory storing a computer program that, when executed by the processor, causes the processor to perform actions such as... Figure 1 The steps of the method shown.

[0078] In one embodiment, a computer-readable storage medium is provided storing a computer program that, when executed by a processor, causes the processor to perform the following actions: Figure 1 The steps of the method shown.

[0079] Those skilled in the art will understand that all or part of the processes in the above embodiments can be implemented by a computer program instructing related hardware. The program can be stored in a non-volatile computer-readable storage medium, and when executed, it can include the processes of the embodiments described above. Any references to memory, storage, databases, or other media used in the embodiments provided in this application can include non-volatile and / or volatile memory. Non-volatile memory can include read-only memory (ROM), programmable ROM (PROM), electrically programmable ROM (EPROM), electrically erasable programmable ROM (EEPROM), or flash memory. Volatile memory can include random access memory (RAM) or external cache memory. By way of illustration and not limitation, RAM is available in various forms, such as static RAM (SRAM), dynamic RAM (DRAM), synchronous DRAM (SDRAM), dual data rate SDRAM (DDRSDRAM), enhanced SDRAM (ESDRAM), synchronous link DRAM (SLDRAM), RAMbus direct RAM (RDRAM), direct memory bus dynamic RAM (DRDRAM), and RAMbus dynamic RAM (RDRAM), etc.

[0080] The technical features of the above embodiments can be combined in any way. For the sake of brevity, not all possible combinations of the technical features in the above embodiments are described. However, as long as there is no contradiction in the combination of these technical features, they should be considered to be within the scope of this specification.

[0081] The embodiments described above are merely illustrative of several implementation methods of this application, and while the descriptions are specific and detailed, they should not be construed as limiting the scope of this patent application. It should be noted that those skilled in the art can make various modifications and improvements without departing from the concept of this application, and these all fall within the protection scope of this application. Therefore, the protection scope of this patent application should be determined by the appended claims.

Claims

1. A method for three-dimensional deformation monitoring of transmission lines, characterized in that, The method includes: A preliminary elevation estimate of the permanent scatterer points on the transmission line is determined; the preliminary elevation estimate is obtained by simplified elevation inversion based on small time baseline image pairs, which are image pairs selected from the multi-temporal synthetic aperture radar images of the transmission line with time baselines less than a preset threshold. Using the preliminary elevation estimates of the permanent scatterer points and the planar coordinates of the permanent scatterer points in the multi-temporal synthetic aperture radar image, a three-dimensional spatial network of the permanent scatterer points is constructed. The three-dimensional spatial network is used to reflect the spatial constraint relationship of the permanent scatterer points of the transmission line geometry. Based on the three-dimensional spatial network, a fine phase model is established for each arc segment of the three-dimensional spatial network, and the coherence coefficient of the observed value of the differential interferometric phase of each arc segment and the fitted value of the differential interferometric phase under different combinations of model parameters are calculated to generate the coherence coefficient solution space of each arc segment. The fine phase model is used to express the model parameters that decompose the differential interferometric phase into at least residual elevation components, deformation components and temperature effect deformation components. The observed value of the differential interferometric phase is obtained by interferometric processing based on the multi-temporal synthetic aperture radar image. Using a preset adaptive threshold detection method, a dynamic decision threshold is determined in the coherence coefficient solution space based on the local background statistical characteristics of the coherence coefficient, and high-quality deformed arc segments are identified using the dynamic decision threshold and the coherence coefficient solution space of each arc segment. The three-dimensional deformation of the transmission line is determined based on the model parameters corresponding to the high-quality deformed arc segment.

2. The method according to claim 1, characterized in that, The coherence coefficient solution space of each arc segment contains a large number of discrete units, and each discrete unit corresponds to a set of model parameters and coherence coefficients under the model parameters. The step of determining a dynamic decision threshold in the coherence coefficient solution space based on the local background statistical characteristics of the coherence coefficients using a preset adaptive threshold detection method includes: For each unit to be detected in the coherence coefficient solution space, a reference unit corresponding to the unit to be detected is determined. The reference unit is a discrete unit set within a preset range of the unit to be detected. The unit to be detected is any discrete unit. The average coherence coefficients in all reference cells are calculated to obtain the background noise estimate of the cell to be detected. The local background statistical characteristics include the background noise estimate. The dynamic decision threshold of the unit to be detected is determined by multiplying the estimated background noise value by a preset scaling factor.

3. The method according to claim 2, characterized in that, The method of identifying high-quality deformed arc segments using dynamic decision thresholds and the coherence coefficient solution space of each arc segment includes: If the coherence coefficient of the unit to be detected is greater than the dynamic decision threshold, then the arc segment is retained and determined to be a high-quality deformable arc segment. If no coherence coefficient of the unit to be detected is greater than the dynamic decision threshold, then the arc segment is discarded.

4. The method according to claim 3, characterized in that, The step of determining the three-dimensional deformation of the transmission line based on the model parameters corresponding to the high-quality deformed arc segment includes: For each high-quality deformed arc segment, the maximum coherence coefficient is determined from the detection units whose coherence coefficient is greater than the dynamic decision threshold; The three-dimensional deformation of the transmission line is determined by using the model parameters corresponding to the maximum coherence coefficient of each high-quality deformation arc segment.

5. The method according to claim 1, characterized in that, The determination of the preliminary elevation estimate of the permanent scatterer point on the transmission line includes: Acquire multi-temporal synthetic aperture radar images of the transmission line; From the multi-temporal synthetic aperture radar images, select small temporal baseline image pairs with time baselines less than a preset baseline threshold to generate interferometric pairs, wherein the preset baseline threshold is used to make the deformation component within the two imaging interval negligible; By performing differential interferometry and phase unwrapping on the interference pair, a preliminary elevation estimate of the permanent scatterer point is obtained.

6. The method according to claim 1, characterized in that, The construction of the three-dimensional spatial network of the permanent scatterer points includes: Using the three-dimensional coordinates of the permanent scatterer points, a three-dimensional Delaunay triangulation network is constructed to generate a set of three-dimensional arc segments connecting each permanent scatterer point. The three-dimensional coordinates include preliminary elevation estimates and planar coordinates.

7. The method according to claim 1, characterized in that, For the interferogram formed by the k-th image, the fine phase model on arc segment (i, j) includes: ； In the formula, Residual elevation components; For deformation components; This represents the temperature-dependent deformation component. Let be the differential interference phase of the interferogram formed by the k-th image pair of arc segment (i, j); The radar wavelength; The slant range of the radar to the permanent scatterer; Angle of incidence; Let be the vertical baseline of the k-th image pair; Let (i, j) be the residual elevation difference of arc segment (i, j). Let (i, j) be the difference in linear deformation rate of arc segment (i, j). The time baseline is the time interval between the kth image and the main image; The difference in the coefficients of thermal expansion between arc segments (i, j); The temperature difference between the k-th image pairs; The combination of model parameters to be solved includes , , .

8. A three-dimensional deformation monitoring system for transmission lines, characterized in that, The system includes: A preliminary elevation unit is used to determine the preliminary elevation estimate of permanent scatterer points on the transmission line. The preliminary elevation estimate is obtained by simplified elevation inversion based on small time baseline image pairs. The small time baseline image pairs are image pairs selected from the multi-temporal synthetic aperture radar images of the transmission line with time baselines less than a preset threshold. A three-dimensional network unit is used to construct a three-dimensional spatial network of the permanent scatterer points using the preliminary elevation estimate of the permanent scatterer points and the planar coordinates of the permanent scatterer points in the multi-temporal synthetic aperture radar image. The three-dimensional spatial network is used to reflect the spatial constraint relationship of the permanent scatterer points of the transmission line geometry. The parameter estimation unit is used to establish a fine phase model for each arc segment of the three-dimensional spatial network based on the three-dimensional spatial network, and to calculate the coherence coefficient of the observed value of the differential interferometric phase of each arc segment and the fitted value of the differential interferometric phase under different combinations of model parameters, so as to generate the coherence coefficient solution space of each arc segment. The fine phase model is used to express the model parameters that decompose the differential interferometric phase into at least residual elevation components, deformation components and temperature effect deformation components. The observed value of the differential interferometric phase is obtained by interferometric processing based on the multi-temporal synthetic aperture radar image. An adaptive selection unit is used to determine a dynamic decision threshold in the coherence coefficient solution space based on the local background statistical characteristics of the coherence coefficient using a preset adaptive threshold detection method, and to identify high-quality deformed arc segments using the dynamic decision threshold and the coherence coefficient solution space of each arc segment. The result output unit is used to determine the three-dimensional deformation of the transmission line based on the model parameters corresponding to the high-quality deformed arc segment.

9. A computer-readable storage medium storing a computer program, characterized in that, When the computer program is executed by a processor, it causes the processor to perform the steps of the method as described in any one of claims 1 to 7.

10. A computer device, comprising a memory and a processor, characterized in that, The memory stores a computer program that, when executed by the processor, causes the processor to perform the steps of the method as described in any one of claims 1 to 7.

Images