Method for monitoring and early warning of power transmission line icing by fusing beidou and polsar

Abstract

This invention discloses a method for monitoring and early warning of icing on power transmission lines that integrates BeiDou and PolSAR. The method includes: synchronous acquisition and preprocessing of multi-source data; inversion of high spatiotemporal resolution atmospheric precipitable water using BeiDou CORS station data, combined with temperature and pressure data to calculate the tropospheric delay phase; simultaneous correction of orbital errors in PolSAR images using precise BeiDou ephemeris data; extraction of a set of characteristic parameters closely related to icing conditions; establishment of an icing thickness inversion model based on phase change and polarization characteristics after fusion correction; construction of an icing growth prediction model based on time-series icing thickness inversion results, combined with near-real-time atmospheric temperature and humidity data inverted from BeiDou and numerical weather prediction data; and initiation of icing early warning based on real-time inverted thickness, predicted thickness, and line design icing thickness. This method addresses the problems of existing technologies that primarily focus on icing condition identification but lack accurate prediction and early warning capabilities for icing growth trends.

Description

Technical Field

[0001] This invention belongs to the field of power transmission line operation and maintenance technology, and in particular relates to a method for monitoring and early warning of icing on power transmission lines that integrates BeiDou and PolSAR. Background Technology

[0002] Icing on transmission lines is one of the serious threats to the safe operation of the power grid. Especially in cold and humid areas in winter, icing on conductors can lead to increased sag and tension, which may cause accidents such as line breakage and tower collapse, resulting in large-scale power outages and economic losses. At present, the monitoring of icing on transmission conductors mainly relies on the following methods: (1) Manual inspection and observation: It depends on the regular on-site inspection or telescope observation by maintenance personnel. It is inefficient, risky, and cannot achieve real-time continuous monitoring. (2) Sensor monitoring: Sensors such as tension, tilt angle, and temperature are installed on the conductors to directly measure the changes in physical parameters caused by icing. For example, Chinese patent CN102269573A discloses an icing monitoring system based on tension sensors. However, this method requires the deployment of equipment on each tower or line segment, which is costly, difficult to maintain, and difficult to cover the entire line. (3) Video monitoring: The status of the conductors is captured in real time by a camera, and the image recognition is used to determine whether icing has occurred. This method is greatly affected by weather and sunlight, fails at night, in rain, snow, or fog, and has limited recognition accuracy. (4) Meteorological model forecast: Based on data such as temperature, humidity, and wind speed from meteorological stations, the thickness of ice accretion is predicted using empirical formulas or numerical models. For example, Chinese patent CN103606067A proposes an ice accretion early warning method based on meteorological parameters. However, this method has low spatial resolution, making it difficult to reflect local micro-meteorological differences caused by terrain, and thus has limited forecast accuracy.

[0003] In recent years, synthetic aperture radar (SAR) technology has demonstrated its advantages in disaster monitoring due to its all-weather, all-day, and high-resolution observation capabilities. In particular, polarimetric SAR can acquire scattering information of ground objects under different polarization states and is sensitive to surface cover such as ice, snow, and water. Meanwhile, the BeiDou Navigation Satellite System not only provides high-precision positioning and time reference, but its ground-based augmentation network can also retrieve meteorological parameters such as atmospheric water vapor, temperature, and pressure, supporting atmospheric correction of SAR data.

[0004] However, there is currently no systematic method in the technology that deeply integrates BeiDou and PolSAR for monitoring icing on power transmission lines. While Chinese patent CN112730990A proposes icing identification based on SAR images, it does not consider the impact of atmospheric delay on SAR phase, nor does it incorporate BeiDou data for joint correction and verification, resulting in insufficient reliability and accuracy of the monitoring results. Furthermore, existing methods primarily focus on identifying the icing state, lacking the ability to accurately predict and provide early warnings of icing growth trends. Summary of the Invention

[0005] The technical problem to be solved by this invention is to provide a method for monitoring and early warning of icing on power transmission lines that integrates BeiDou and PolSAR, so as to solve the problem that the existing technology for monitoring icing on power transmission lines does not consider the influence of atmospheric delay on SAR phase, nor does it introduce BeiDou data for joint correction and verification, resulting in insufficient reliability and accuracy of monitoring results; it focuses more on the identification of icing status and lacks the ability to accurately predict and provide early warning of icing growth trends.

[0006] Technical solution of the present invention:

[0007] A method for monitoring and early warning of icing on power transmission lines integrating BeiDou and PolSAR, the method comprising:

[0008] Step 1: Collect time-series PolSAR image data covering the target transmission line corridor, observation data from the continuously operating reference stations of the BeiDou ground-based augmentation network, vector data of transmission line towers and conductors, digital elevation model data, and meteorological reanalysis data, and perform preprocessing.

[0009] Step 2: Use BeiDou CORS station data to retrieve high spatiotemporal resolution atmospheric precipitable water, and combine it with temperature and pressure data provided by meteorological reanalysis data to calculate the tropospheric delay phase; at the same time, use BeiDou precise ephemeris to assist in correcting the orbital errors of PolSAR images.

[0010] Step 3: Based on the transmission line vector data, extract conductor pixels from the PolSAR intensity image, perform polarization decomposition on each conductor pixel, and extract the set of feature parameters closely related to the icing state.

[0011] Step 4: Establish an ice thickness inversion model based on phase change and polarization characteristics after fusion correction;

[0012] Step 5: Based on the time series ice thickness inversion results, and combined with the near real-time atmospheric temperature and humidity data and numerical weather forecast data retrieved from BeiDou, construct an ice growth prediction model.

[0013] Step 6: Based on the real-time inversion thickness Predicted thickness and the thickness of icing in line design ,when An icing warning will be activated if it is predicted that the threshold will be reached within the next 4 hours.

[0014] The PolSAR imagery data acquisition selected Sentinel-1 data with a short revisit period in the C-band and SAR satellite data with dual polarization (VV+VH) or full polarization capabilities. The data coverage area included the target transmission line corridor, and the time series covered the entire icing-prone season. BeiDou data was obtained from the BeiDou ground-based augmentation network covering the study area, acquiring raw observation data and simultaneously downloading corresponding precise ephemeris and clock bias products. Auxiliary data collection included transmission line geographic information system data, such as tower coordinates, conductor sag parameters and design icing thickness, 30-meter resolution digital elevation model data, and meteorological reanalysis data from medium-term weather forecasts. Data preprocessing methods included: radiometric calibration, multi-look processing, precise registration, and geocoding of PolSAR data to generate intensity maps and coherence maps in the geographic coordinate system; BeiDou data was used to calculate the total zenith tropospheric delay for each CORS station, and the zenith wet delay was separated using the Saastamoinen model and surface meteorological data, subsequently retrieving atmospheric precipitable water volume (PWV) with a time resolution of 5 minutes. Perform coordinate system and format conversion on the DEM and line vectors.

[0015] Atmospheric delay phase calculation includes:

[0016] For each image in a PolSAR interferometer pair, the atmospheric delay phase in the line-of-sight direction Represented as dry delay With wet delay sum:

[0017] Among them, the dry delay phase utilizes the Saastamoinen model and surface air pressure. calculate: ;

[0018] Atmospheric precipitable water retrieved from BeiDou using wet delay phase. With water vapor conversion coefficient The conversion yields: ;

[0019] In the formula: For radar wavelength, For radar incident angle, , where is the refractive constant. , where is the gas constant for dry air. It is the acceleration due to gravity. The water vapor conversion coefficient is calculated from the weighted average temperature. calculate: ,in For the density of water, The constant of water vapor. Here is the refractive constant;

[0020] Interference phase correction methods include:

[0021] PolSAR differential interferometric phase Corrections were performed to obtain the main phases related to icing deformation. : ;

[0022] The remaining orbital error phase after correction using BeiDou-assisted precision orbital data.

[0023] The specific methods for BeiDou-assisted calibration include:

[0024] Step 2.1: Generate differential interferograms from time-series PolSAR images, selecting images from dates with no ice or light ice as the main image;

[0025] Step 2.2: For each differential interferogram, use BeiDou PWV data to generate a PWV grid map that spatially matches the SAR image using Kriging interpolation.

[0026] Step 2.3: Calculate the tropospheric delay phase for each grid point. The phase is then subtracted from the original interference phase to obtain the preliminarily corrected phase. ;

[0027] Step 2.4: Refine the SAR satellite orbit using BeiDou orbit data and remove remaining orbital error phases. The final phase related to icing deformation was obtained. ;

[0028] Step 2.5, for Phase unwrapping is performed, converting the deformation along the radar line-of-sight direction. : .

[0029] Methods for extracting feature parameter sets closely related to icing conditions include:

[0030] Step 3.1: On the intensity map after PolSAR geocoding, according to the transmission line vector, set a buffer with a width of 3-5 pixels along the direction of the conductor, and extract all pixels in the buffer as potential conductor targets;

[0031] Step 3.2: For each potential conductor pixel, calculate the polarization coherence matrix or covariance matrix;

[0032] Step 3.3: Perform H / A / α polarization decomposition on each pixel to extract the feature vector. Volume scattering was extracted using the Freeman-Durden three-component decomposition method. Dihedral scattering Surface scattering Power; Result:

[0033] ;

[0034] In the formula: The total scattering power, The average scattering angle, For entropy, It is anisotropic.

[0035] The constructed ice thickness inversion model includes:

[0036] In the formula: The line-of-sight deformation phase caused by conductor icing; For the first One polarization characteristic parameter; These are the regression coefficients of the model; This is the error term.

[0037] The construction of the icing growth prediction model includes: ;

[0038] In the formula: For the future Predicted ice thickness at any given time; Wet-bulb temperature; Ambient temperature; Wind speed; It refers to atmospheric precipitable water. These are the model parameters.

[0039] Methods for predicting icing growth include:

[0040] Step 5.1: For the line segment, extract the time series of average ice thickness. ;

[0041] Step 5.2: Obtain synchronized ambient temperature data from BeiDou data or ERA5 data. relative humidity Wind speed And calculate the wet-bulb temperature. and PWV;

[0042] Step 5.3: Using the prediction model, with the data from the previous N hours as the training window, update the model parameters online using the recursive least squares method. And predict the future ice thickness ;

[0043] Step 5.4: Repeat this process to generate the icing thickness prediction curve for the next 24 hours.

[0044] When some BeiDou CORS station data is missing, meteorological field data provided by ERA5 meteorological reanalysis data is used in conjunction with digital elevation models to simulate atmospheric delay phase through physical models, serving as a supplement or replacement for BeiDou data; the acquisition of PolSAR image data also incorporates L-band ALOS-2 or domestic Gaofen-3 SAR data.

[0045] Line design icing thickness It's not a fixed value; it depends on the real-time wind speed. Dynamic adjustments are made, and the adjustment formula is as follows: ;

[0046] This is the wind load factor related to the conductor type.

[0047] The beneficial effects of this invention are:

[0048] This invention achieves large-scale, high spatiotemporal resolution monitoring: a single imaging operation using spaceborne PolSAR can cover hundreds of kilometers of power transmission corridors, achieving a spatial resolution down to the meter level; atmospheric correction is achieved by fusing BeiDou data, ensuring the reliability of phase information and significantly improving monitoring accuracy. Compared to traditional point sensors, this method realizes full-line monitoring combining line and surface approaches.

[0049] This invention has all-day, all-weather monitoring capabilities: PolSAR is not affected by sunlight or cloud and rain weather, and BeiDou data can be acquired around the clock. The combination of the two enables effective monitoring even under adverse conditions such as night, rain, snow, and fog, overcoming the limitations of optical and video methods.

[0050] This invention improves the accuracy of icing inversion and prediction: by fusing two types of information, namely SAR phase variation (sensitive to conductor deformation caused by icing) and polarization scattering characteristics (sensitive to icing type and thickness), and combining them with precise meteorological parameters provided by BeiDou, an icing inversion and prediction model with clear physical meaning is constructed.

[0051] This invention achieves intelligent early warning and decision support: it establishes a hierarchical early warning mechanism based on real-time monitoring and trend prediction, and the early warning threshold can be flexibly adjusted according to line grade and regional characteristics. The system automatically generates visual reports, providing scientific and intuitive basis for emergency decisions such as de-icing and ice melting.

[0052] This invention significantly reduces construction and maintenance costs: it eliminates the need for large-scale deployment and maintenance of ground sensors along vast power transmission corridors, relying primarily on satellite remote sensing data and existing BeiDou infrastructure. It is highly economical and scalable, and suitable for disaster prevention and mitigation systems of power grid companies at all levels.

[0053] This addresses the shortcomings of existing technologies for monitoring icing on transmission lines, which fail to consider the impact of atmospheric delay on SAR phase and do not incorporate BeiDou data for joint correction and verification, resulting in insufficient reliability and accuracy of monitoring results. Furthermore, these technologies tend to focus more on identifying icing conditions and lack the ability to accurately predict and provide early warnings of icing growth trends. Attached Figure Description

[0054] Figure 1 This is a schematic diagram of the process of the present invention;

[0055] Figure 2 This is a schematic diagram illustrating the principle of PolSAR atmospheric and orbital error correction based on BeiDou data. Detailed Implementation

[0056] A method for monitoring and early warning of icing on power transmission lines that integrates BeiDou and PolSAR includes:

[0057] Step S1: Synchronous Acquisition and Preprocessing of Multi-Source Data

[0058] Acquire temporal PolSAR imagery data covering the target transmission line corridor, observation data from continuously operating reference stations of the BeiDou ground-based augmentation network, vector data of transmission line towers and conductors, digital elevation model data, and meteorological reanalysis data.

[0059] Step S2: PolSAR atmospheric and orbital error correction based on BeiDou data:

[0060] High spatiotemporal resolution atmospheric precipitable water was retrieved using BeiDou CORS station data, and combined with temperature and pressure data provided by meteorological reanalysis, the tropospheric delay phase was calculated. Simultaneously, precise BeiDou ephemeris data was used to assist in correcting orbital errors in PolSAR imagery. The correction model is as follows:

[0061] S2.1 Atmospheric Delay Phase Calculation:

[0062] For each image in a PolSAR interferometer pair, the atmospheric delay phase along the line of sight is... This can be expressed as a dry delay. With wet delay sum:

[0063] Among them, the dry delay phase utilizes the Saastamoinen model and surface air pressure. calculate:

[0064] Atmospheric precipitable water retrieved from BeiDou using wet delay phase. With water vapor conversion coefficient The conversion yields:

[0065] In the formula:

[0066] The radar wavelength (m) is given.

[0067] The radar incident angle (in radians).

[0068] , where is the refractive constant.

[0069] , where is the gas constant for dry air.

[0070] The acceleration due to gravity (m / s²)

[0071] The water vapor conversion coefficient is calculated from the weighted average temperature. calculate: ,in For the density of water, The constant of water vapor. is the refractive constant.

[0072] S2.2 Interferometric Phase Correction: Correction of PolSAR Differential Interferometric Phase Corrections were performed to obtain the main phases related to icing deformation. :

[0073]

[0074] in, The remaining orbital error phase after correction using BeiDou-assisted precision orbital data.

[0075] Step S3: Target extraction and polarization scattering feature interpretation:

[0076] Based on transmission line vector data, conductor pixels are accurately extracted from PolSAR intensity images. Polarization decomposition is performed on each conductor pixel to extract a set of feature parameters closely related to icing conditions. :

[0077] In the formula: This represents the total scattered power. The average scattering angle reflects the dominant scattering mechanism; Entropy characterizes the randomness of scattering; It is anisotropic; These represent the volume scattering, dihedral scattering, and surface scattering power obtained from the Cloude-Pottier decomposition, respectively.

[0078] Step S4: Inversion of icing thickness by fusing phase and polarization features:

[0079] Establish icing thickness with phase change and polarization characteristics after fusion correction Inversion model:

[0080] in: The line-of-sight deformation phase caused by conductor icing obtained in step S2 (converted to linear displacement, unit: mm); For the first step S3, the extracted One polarization characteristic parameter; These are the regression coefficients of the model; This represents the error term. The model coefficients are obtained by training on historical icing event samples (including measured icing thickness, synchronous PolSAR and BeiDou data), and can be obtained using multiple linear regression or machine learning algorithms.

[0081] Step S5: Modeling and Predicting Icing Growth Trends

[0082] Ice thickness inversion results based on time series A prediction model for icing growth is constructed by combining near-real-time atmospheric temperature and humidity data retrieved from BeiDou and numerical weather prediction data. An improved autoregressive model considering environmental driving forces is adopted:

[0083] In the formula: For the future Predicted ice thickness at any given time; The wet-bulb temperature is (°C). Ambient temperature (°C); Wind speed (m / s); Atmospheric precipitable water (mm); These are the model parameters, learned from historical data.

[0084] Step S6: Early Warning and Information Dissemination:

[0085] Based on real-time inversion thickness and predicted thickness Combined with the line design, icing thickness ,when An icing warning will be activated if the threshold is predicted to be reached within the next 4 hours. The warning information, along with an icing spatial distribution map and a predicted trend curve, will be automatically pushed to operations and maintenance personnel through a dedicated platform.

[0086] The technical solution of the present invention will be further illustrated by the following examples:

[0087] like Figure 1 As shown, the transmission line icing monitoring and early warning method integrating BeiDou and PolSAR described in this invention is implemented according to the following steps:

[0088] Step S101: Synchronous acquisition of multi-source data:

[0089] PolSAR data: Select SAR satellite data with short revisit periods (e.g., 6 days for Sentinel-1) and dual polarization (VV+VH) or full polarization capabilities. The data coverage area should include the target transmission line corridor, and the time series should cover the entire icing-prone season.

[0090] BeiDou data: Raw observation data is obtained from the BeiDou ground-based augmentation network (CORS) covering the study area, and corresponding precise ephemeris and clock error products are downloaded at the same time.

[0091] Supporting data: Collect data from the geographic information system of transmission lines (including tower coordinates, conductor sag parameters, and design icing thickness), 30-meter resolution digital elevation model data, and meteorological reanalysis data from medium-term weather forecasts.

[0092] Step S102: Data preprocessing:

[0093] PolSAR data: Perform radiometric calibration, multi-view processing, precise registration, and geocoding to generate intensity maps and coherence maps in geographic coordinate systems.

[0094] BeiDou data: The total zenith tropospheric delay of each CORS station is calculated, and the zenith wet delay is separated using the Saastamoinen model and ground meteorological data. Then, the atmospheric precipitable water volume (PWV) is retrieved with a time resolution of up to 5 minutes.

[0095] Auxiliary data: Coordinate system and format conversion of DEM and line vectors.

[0096] Step S103: PolSAR Interferometric Processing and BeiDou-Assisted Correction: 1. Generate differential interferograms for the time-series PolSAR images, selecting images from dates with no or light icing as the main image. 2. For each differential interferogram, using the BeiDou PWV data obtained in step S102, generate a PWV grid map spatially matched to the SAR image through Kriging interpolation. 3. Calculate the tropospheric delay phase for each grid point according to the formula. The phase is then subtracted from the original interference phase to obtain the preliminarily corrected phase. 4. Refine the SAR satellite orbit using BeiDou precise orbit data, estimate and remove remaining orbital error phases. The final phase related to icing deformation was obtained. 5. Regarding Phase unwrapping is performed to convert it into a deformation along the radar line-of-sight direction. (Unit: m):

[0097] Step S104: Transmission Line Target Identification and Feature Extraction: 1. On the geocoded intensity map of PolSAR, based on the transmission line vector, set a buffer zone with a width of 3-5 pixels along the transmission line direction, and extract all pixels within the buffer zone as potential transmission line targets. 2. For each potential transmission line pixel, calculate its polarization coherence matrix or covariance matrix. 3. Perform H / A / α polarization decomposition on each pixel and extract feature vectors. Simultaneously, volume scattering was extracted using Freeman-Durden three-component decomposition. Dihedral scattering Surface scattering Power. 4. Because icing significantly alters the scattering mechanism of the conductor, therefore , and It is a key feature for identifying icing.

[0098] Step S105: Ice Thickness Inversion: 1. Collect a sample set of historical icing events. Each sample contains: the point... Polarization eigenvectors And the measured ice thickness obtained through manual inspection or tension sensors. 2. With and As the independent variable, Using the variable as the dependent variable, a random forest regression algorithm is used to train an inversion model for ice thickness. Random forests can handle nonlinear relationships and are insensitive to feature collinearity. 3. The trained model is applied to each conductor cell of the entire transmission line to output a spatial distribution map of ice thickness. .

[0099] Step S106: Icing Growth Prediction: 1. For each key line segment, extract its average icing thickness time series. 2. Obtain synchronized ambient temperature data from BeiDou data or ERA5 data. relative humidity Wind speed And calculate the wet-bulb temperature. And PWV. 3. Using the prediction model described in step S5, with the data from the previous N hours as the training window, update the model parameters online using the recursive least squares method. And predict the future ice thickness 4. Repeat this process to generate a prediction curve for the icing thickness over the next 24 hours.

[0100] Step S107: Warning Generation and Issuance: 1. The system monitors each line partition in real time. and 2. Automatically determine and issue warnings based on preset thresholds. 3. Once a warning is triggered, the system automatically highlights the warning segment on the platform and generates a standardized warning report containing "time, location, current thickness, predicted thickness, and recommended measures." 4. Send the warning information to the relevant regional operations and maintenance personnel in real time via SMS, app push notifications, email, etc.

[0101] In another preferred embodiment, to address the potential impact of extreme weather conditions on BeiDou signals and enhance system robustness, the following strategies are adopted: Step S201 Data Assimilation: When some BeiDou CORS station data is missing, meteorological field data provided by ERA5 meteorological reanalysis data is used, combined with a digital elevation model, and atmospheric delay phase is simulated using physical models (such as MM5, WRF) as a supplement or substitute for BeiDou data. Step S202 Multi-Source SAR Data Fusion: In addition to using C-band Sentinel-1 data, L-band ALOS-2 or domestic Gaofen-3 SAR data are also introduced. The L-band has stronger penetration, is more sensitive to icing layers, and is less affected by atmospheric conditions. By fusing multi-band SAR information, the thickness and density of icing can be more comprehensively retrieved. Step S203 Dynamic Adjustment of Warning Threshold: The warning threshold... It's not a fixed value, but rather depends on the real-time wind speed. Dynamic adjustments should be made because wind significantly increases the mechanical load on icy conductors. The adjustment formula is:

[0102] in This is a wind load factor related to the conductor type. This allows for early warnings to be issued during strong winds.

[0103] Summary of working principle and process:

[0104] This invention acquires high-resolution backscattering and phase information of transmission lines using PolSAR, where phase changes accurately reflect conductor deformation caused by icing. Utilizing high-precision, high-temporal-resolution atmospheric parameters provided by the BeiDou system, precise atmospheric delay correction is applied to the SAR phase, extracting the pure icing deformation signal. Simultaneously, the polarization characteristics of PolSAR are sensitive to physical properties of icing, such as dielectric properties and surface roughness. By fusing the corrected phase and polarization characteristics, a more reliable icing thickness inversion model is constructed. Based on this, combined with real-time meteorological data and numerical forecasts provided by BeiDou, a dynamic prediction model for icing growth is established, achieving a leap from "monitoring the current situation" to "predicting the future." Finally, the system automatically triggers early warnings based on real-time inversion results and predicted trends, providing intelligent decision support for the safe operation of the power grid. The entire process is automated, achieving integrated "sky-ground" intelligent perception and early warning of icing conditions on a large scale of transmission lines.

Claims

1. A method for monitoring and early warning of icing on power transmission lines integrating BeiDou and PolSAR, characterized in that: The method includes: Step 1: Collect time-series PolSAR image data covering the target transmission line corridor, observation data from the continuously operating reference stations of the BeiDou ground-based augmentation network, vector data of transmission line towers and conductors, digital elevation model data, and meteorological reanalysis data, and perform preprocessing. Step 2: Use BeiDou CORS station data to retrieve high spatiotemporal resolution atmospheric precipitable water, and combine it with temperature and pressure data provided by meteorological reanalysis data to calculate the tropospheric delay phase; at the same time, use BeiDou precise ephemeris to assist in correcting the orbital errors of PolSAR images. Step 3: Based on the transmission line vector data, extract conductor pixels from the PolSAR intensity image, perform polarization decomposition on each conductor pixel, and extract the set of feature parameters closely related to the icing state. Step 4: Establish an ice thickness inversion model based on phase change and polarization characteristics after fusion correction; Step 5: Based on the time series ice thickness inversion results, and combined with the near real-time atmospheric temperature and humidity data and numerical weather forecast data retrieved from BeiDou, construct an ice growth prediction model. Step 6: Based on the real-time inversion thickness Predicted thickness and the thickness of icing in line design ,when An icing warning will be activated if it is predicted that the threshold will be reached within the next 4 hours.

2. The method for monitoring and early warning of icing on power transmission lines integrating BeiDou and PolSAR as described in claim 1, characterized in that: The PolSAR imagery data acquisition selected Sentinel-1 data with a short revisit period in the C-band and SAR satellite data with dual polarization (VV+VH) or full polarization capabilities. The data coverage area included the target transmission line corridor, and the time series covered the entire icing-prone season. The BeiDou data was obtained from the BeiDou ground-based augmentation network covering the study area, and the corresponding precise ephemeris and clock bias products were downloaded. The auxiliary data collection included data from the transmission line geographic information system, including tower coordinates, conductor sag parameters and design icing thickness, 30-meter resolution digital elevation model data, and meteorological reanalysis data from medium-term weather forecasts. Data preprocessing methods include: radiometric calibration, multi-view processing, precise registration, and geocoding of PolSAR data to generate intensity maps and coherence maps in geographic coordinate systems; The BeiDou data is used to calculate the total zenith tropospheric delay of each CORS station, and the zenith wet delay is separated using the Saastamoinen model and ground meteorological data. Then, the atmospheric precipitable water volume (PWV) is obtained by inversion with a time resolution of 5 minutes. The DEM and line vector are converted to coordinate system 1 and format.

3. The method for monitoring and early warning of icing on power transmission lines integrating BeiDou and PolSAR as described in claim 1, characterized in that: Atmospheric delay phase calculation includes: For each image in a PolSAR interferometer pair, the atmospheric delay phase in the line-of-sight direction Represented as dry delay With wet delay sum: Among them, the dry delay phase utilizes the Saastamoinen model and surface air pressure. calculate: ; Atmospheric precipitable water retrieved from BeiDou using wet delay phase. With water vapor conversion coefficient The conversion yields: ； In the formula: For radar wavelength, For radar incident angle, , where is the refractive constant. , where is the gas constant for dry air. It is the acceleration due to gravity. The water vapor conversion coefficient is calculated from the weighted average temperature. calculate: ,in For the density of water, The constant of water vapor. Here is the refractive constant; Interference phase correction methods include: PolSAR differential interferometric phase Corrections were performed to obtain the main phases related to icing deformation. : ; The remaining orbital error phase after correction using BeiDou-assisted precision orbital data.

4. The method for monitoring and early warning of icing on power transmission lines integrating BeiDou and PolSAR as described in claim 3, characterized in that: The specific methods for BeiDou-assisted calibration include: Step 2.1: Generate differential interferograms from time-series PolSAR images, selecting images from dates with no ice or light ice as the main image; Step 2.2: For each differential interferogram, use BeiDou PWV data to generate a PWV grid map that spatially matches the SAR image using Kriging interpolation. Step 2.3: Calculate the tropospheric delay phase for each grid point. The phase is then subtracted from the original interference phase to obtain the preliminarily corrected phase. ; Step 2.4: Refine the SAR satellite orbit using BeiDou orbit data and remove remaining orbital error phases. The final phase related to icing deformation was obtained. ; Step 2.5, for Phase unwrapping is performed, converting the deformation along the radar line-of-sight direction. : .

5. The method for monitoring and early warning of icing on power transmission lines integrating BeiDou and PolSAR as described in claim 1, characterized in that: Methods for extracting feature parameter sets closely related to icing conditions include: Step 3.1: On the intensity map after PolSAR geocoding, according to the transmission line vector, set a buffer with a width of 3-5 pixels along the direction of the conductor, and extract all pixels in the buffer as potential conductor targets; Step 3.2: For each potential conductor pixel, calculate the polarization coherence matrix or covariance matrix; Step 3.3: Perform H / A / α polarization decomposition on each pixel to extract the feature vector. Volume scattering was extracted using the Freeman-Durden three-component decomposition method. Dihedral scattering Surface scattering Power; Result: ； In the formula: The total scattering power, The average scattering angle, For entropy, It is anisotropic.

6. The method for monitoring and early warning of icing on power transmission lines integrating BeiDou and PolSAR as described in claim 1, characterized in that: The constructed ice thickness inversion model includes: In the formula: The line-of-sight deformation phase caused by conductor icing; For the first One polarization characteristic parameter; These are the regression coefficients of the model; This is the error term.

7. The method for monitoring and early warning of icing on power transmission lines integrating BeiDou and PolSAR as described in claim 1, characterized in that: The construction of the icing growth prediction model includes: ; In the formula: For the future Predicted ice thickness at any given time; Wet-bulb temperature; Ambient temperature; Wind speed; It refers to atmospheric precipitable water. These are the model parameters.

8. The method for monitoring and early warning of icing on power transmission lines integrating BeiDou and PolSAR as described in claim 5, characterized in that: Methods for predicting icing growth include: Step 5.1: For the line segment, extract the time series of average ice thickness. ; Step 5.2: Obtain synchronized ambient temperature data from BeiDou data or ERA5 data. relative humidity Wind speed And calculate the wet-bulb temperature. and PWV; Step 5.3: Using the prediction model, with the data from the previous N hours as the training window, update the model parameters online using the recursive least squares method. And predict the future ice thickness ; Step 5.4: Repeat this process to generate the icing thickness prediction curve for the next 24 hours.

9. A method for monitoring and early warning of icing on power transmission lines integrating BeiDou and PolSAR as described in claim 2, characterized in that: When some BeiDou CORS station data is missing, meteorological field data provided by ERA5 meteorological reanalysis data is used in conjunction with digital elevation models to simulate atmospheric delay phase through physical models, serving as a supplement or replacement for BeiDou data; the acquisition of PolSAR image data also incorporates L-band ALOS-2 or domestic Gaofen-3 SAR data.

10. A method for monitoring and early warning of icing on power transmission lines integrating BeiDou and PolSAR as described in claim 1, characterized in that: Line design icing thickness It's not a fixed value; it depends on the real-time wind speed. Dynamic adjustments are made, and the adjustment formula is as follows: ; This is the wind load factor related to the conductor type.

Images