Method and system for intelligently identifying icing microtopography of power transmission line based on icing simulation

Abstract

The invention relates to the technical field of micro-terrain intelligent identification, in particular to a power transmission line icing micro-terrain intelligent identification method and system based on icing simulation. Comprising the following steps: S1, determining position data of a power transmission line, and obtaining remote sensing data covering the power transmission line according to the position data; s2, establishing an icing micro-topography database, and performing micro-topography classification on the icing micro-topography database; the defects in the prior art are effectively overcome through multi-source data fusion and intelligent modeling, a multi-source fusion scheme of SAR data, multispectral data, DEM data and meteorological data is adopted, comprehensive extraction of three-dimensional features of terrains, airflow and water vapor is achieved in combination with accurate position data of a power transmission line management end, and the accuracy of the three-dimensional features of the terrains, the airflow and the water vapor is improved. The problems of single feature dimension and incomplete data coverage in the prior art are solved, richer basic data support is provided for microtopography recognition, and the relevance between the features and the icing mechanism is remarkably improved.

Description

Intelligent Recognition Method and System for Icing Micro-topography of Transmission Lines Based on Icing Simulation Technical Field

[0001] This invention relates to the field of intelligent micro-topography recognition technology, and more specifically, to a method and system for intelligent recognition of icing micro-topography of transmission lines based on icing simulation. Background Technology

[0002] As the core hub of the power system, the safe and stable operation of transmission lines is directly related to the reliability of energy supply. Icing disasters are one of the main risks threatening the safety of transmission lines. Micro-topography can easily form high-intensity icing by changing local airflow, water vapor distribution and temperature conditions, causing serious faults such as line tripping and tower collapse.

[0003] Existing technologies have significant shortcomings and are difficult to meet the actual needs of accurate identification. Geological surveys are limited by terrain accessibility. Transmission lines often pass through remote areas such as high mountains and canyons, and some areas are inaccessible by roads. In addition, the low temperature and rain and snow during the icing period further increase the difficulty and danger of surveys, resulting in extremely low survey efficiency (only 3-5 points can be verified per day) and limited coverage. It is impossible to achieve comprehensive monitoring of thousands of kilometers of line corridors. Traditional remote sensing identification methods rely on a single data source, have insufficient feature extraction dimensions, focus only on terrain morphology parameters, and ignore key influencing factors of icing such as airflow dynamics and water vapor supply. Moreover, micro-topography classification lacks a unified quantitative standard and relies on subjective experience to set thresholds, resulting in low identification accuracy and high rates of missed and false identifications. Therefore, an intelligent identification method and system for icing micro-topography of transmission lines based on icing simulation is proposed. Summary of the Invention

[0004] The purpose of this invention is to provide a method and system for intelligent identification of icing micro-topography of transmission lines based on icing simulation, so as to solve the problems mentioned in the background art.

[0005] To address the aforementioned technical problems, one objective of this invention is to provide an intelligent identification method for icing micro-topography of transmission lines based on icing simulation, comprising the following steps: S1, determining the location data of the transmission line and acquiring remote sensing data covering the transmission line based on the location data; S2, establishing an icing micro-topography database, classifying the icing micro-topography database into micro-topography categories, and then extracting the topographic features, airflow features, and water vapor features of the terrain to be identified along the transmission line based on the remote sensing data; S3, combining the topographic features, airflow features, and water vapor features extracted in S2 with historical identified terrain from various icing micro-topography databases for priority analysis, obtaining the priority of each historical identified terrain, and combining the historical identified terrain... The priority and corresponding micro-topographical features are used to establish icing simulation models for various icing micro-topographic databases; S4, the topographic features, airflow features and water vapor features are input into the icing simulation model, and the icing simulation model outputs predicted icing features. At the same time, the historical icing features of the terrain to be identified are extracted from the remote sensing data. The predicted icing features are combined with the historical icing features for similarity screening, and the icing micro-topographic database corresponding to the predicted icing features with the highest similarity is selected; S5, the deviation data is obtained based on the predicted icing features and historical icing features. The feature patterns are adjusted based on the deviation data, and the adjusted feature patterns are used to form personalized icing micro-topography, which is then entered into the corresponding icing micro-topographic database.

[0006] As a further improvement to this technical solution, in step S1, a communication connection is established with the transmission line management terminal to obtain the location data of the transmission line from the transmission line management terminal, and remote sensing data from multiple satellites is also obtained. The remote sensing data covering the transmission line is filtered based on the location data of the transmission line. The location data includes transmission line corridor vector data, tower coordinate data, and line attribute information. The sources of the remote sensing data include SAR data, multispectral data, DEM data, high-resolution data, and meteorological data, and the coverage area is the transmission line corridor and the area on both sides.

[0007] As a further improvement to this technical solution, the steps of S2 are as follows: S2.1, collect historical identified terrain, and establish an icing micro-terrain database based on the historical identified terrain. At the same time, set five types of micro-terrain, and then classify the icing micro-terrain database according to the five types of micro-terrain, so that the icing micro-terrain database is divided into five types of icing micro-terrain databases; S2.2, extract the terrain features, airflow features and water vapor features of the terrain to be identified corresponding to the transmission line based on remote sensing data.

[0008] As a further improvement to this technical solution, in S2.1, the five types of micro-topography include high mountain watershed type, topographic uplift type, mountain pass type, canyon wind tunnel type, and water vapor enhancement type.

[0009] As a further improvement to this technical solution, step S3 is as follows: S3.1 Extracting the terrain features, airflow features, and water vapor features of historically identified terrains from various icing micro-terrain databases. Simultaneously, combining the terrain features, airflow features, and water vapor features extracted in S2.2, priority analysis is performed on the historically identified terrains to obtain the priority of each historically identified terrain. The higher the similarity between the terrain features, airflow features, and water vapor features, the higher the priority. S3.2 Setting specific feature rules for micro-terrains in various icing micro-terrain databases, and combining historically identified terrains and their corresponding priorities, an icing simulation model is established so that each type of icing micro-terrain database has a corresponding icing simulation model.

[0010] As a further improvement to this technical solution, in S3.2, the icing simulation model adopts the WRF-LightGBM hybrid model framework. The WRF model is used to forecast meteorological elements, and the output meteorological data is used as the input features of the LightGBM algorithm. Parameter calibration is performed for different micro-topographic types, as follows: for high mountain watershed types, the focus is on calibrating altitude and wind speed amplification coefficient; for topographic lifting types, the focus is on calibrating slope and aspect; for mountain pass types, the focus is on calibrating topographic contraction; for canyon wind tunnel types, the focus is on calibrating width-to-depth ratio and wind direction matching degree; and for water vapor enhancement types, the focus is on calibrating minimum distance to water bodies and relative humidity.

[0011] As a further improvement to this technical solution, step S4 is as follows: S4.1, input the terrain features, airflow features, and water vapor features of the terrain to be identified into the icing simulation model corresponding to various icing micro-terrain databases, and output the predicted icing features corresponding to the terrain to be identified by the icing simulation model; S4.2, obtain the historical icing features of the terrain to be identified based on remote sensing data, and then perform similarity filtering by combining the predicted icing features with the historical icing features to obtain the similarity between the predicted icing features and the historical icing features, and then select the icing micro-terrain database of the icing simulation model corresponding to the predicted icing feature with the highest similarity.

[0012] As a further improvement to this technical solution, step S5 is as follows: S5.1, combine the predicted icing features of the icing simulation model selected in S4.2 with historical icing features to obtain deviation data, and obtain the deviation data between the predicted icing features and historical icing features; S5.2, make specific adjustments to the feature patterns corresponding to the icing micro-topography based on the deviation data, obtain the feature patterns corresponding to the terrain to be identified, and combine the feature patterns corresponding to the terrain to be identified with terrain features, airflow features and water vapor features to form a personalized icing micro-topography, and enter it into the selected icing micro-topography database.

[0013] The second objective of this invention is to provide an intelligent identification system for icing micro-topography of transmission lines based on icing simulation, including the intelligent identification method for icing micro-topography of transmission lines based on icing simulation as described in any one of the above-mentioned methods, comprising a data acquisition unit, a model building unit, an icing prediction unit, and a terrain recording unit; the data acquisition unit is used to determine the location data of the transmission line, acquire remote sensing data covering the transmission line based on the location data, establish an icing micro-topography database, classify the micro-topography of the icing micro-topography database, and then extract the terrain features, airflow features, and water vapor features of the terrain to be identified along the transmission line based on the remote sensing data; the model building unit is used to perform priority analysis by combining the extracted terrain features, airflow features, and water vapor features with historical identified terrains from various icing micro-topography databases to obtain... The system prioritizes historically identified terrain features and, based on these priorities and the specific characteristics of each micro-topography, establishes icing simulation models for various icing micro-topography databases. The icing prediction unit inputs terrain features, airflow features, and water vapor features into the icing simulation model, which then outputs predicted icing features. Simultaneously, it extracts historical icing features from remote sensing data, combines the predicted icing features with historical icing features for similarity filtering, and selects the icing micro-topography database corresponding to the predicted icing features with the highest similarity. The terrain input unit obtains deviation data based on the predicted icing features and historical icing features, performs specific adjustments based on the deviation data, and constructs personalized icing micro-topography based on the adjusted features, which is then entered into the corresponding icing micro-topography database.

[0014] Compared with existing technologies, the beneficial effects of this invention are as follows: 1. The intelligent identification method and system for icing micro-topography of transmission lines based on icing simulation effectively makes up for the shortcomings of existing technologies through multi-source data fusion and intelligent modeling. It adopts a multi-source fusion scheme of SAR data, multispectral data, DEM data and meteorological data, combined with the accurate location data of the transmission line management end, to achieve comprehensive extraction of three-dimensional features of terrain, airflow and water vapor. It solves the problems of single feature dimension and incomplete data coverage in traditional technologies, provides richer basic data support for micro-topography identification, and significantly improves the correlation between features and icing mechanism.

[0015] 2. A method and system for intelligent identification of icing micro-topography of transmission lines based on icing simulation. By constructing a database of five types of micro-topography, a dedicated WRF-LightGBM hybrid model is designed for the dominant icing factors of each type of micro-topography, and differential parameter calibration is performed. This avoids the limitations of a one-size-fits-all general model, greatly improves the identification accuracy of different types of micro-topography, effectively reduces the risk of missed or false identification, and provides a precise basis for the classification of micro-topography types for differentiated anti-icing design.

[0016] 3. An intelligent identification method and system for icing micro-topography of transmission lines based on icing simulation. By introducing historical terrain priority analysis and deviation correction mechanism, and using cosine similarity algorithm to select highly correlated historical samples to participate in model construction, the system dynamically adjusts feature pattern parameters by combining the deviation data between predicted icing features and historical icing features, forming a closed-loop system of data-model-validation-optimization. This not only improves the generalization ability of the model and adapts it to application scenarios in different climate zones such as southern mountainous areas and plateau regions, but also continuously optimizes the recognition performance with the accumulation of historical data, solving the problems of poor generalization and inability to dynamically update traditional technologies. Attached Figure Description

[0017] Figure 1 is a flowchart illustrating the intelligent identification method for icing micro-topography of transmission lines based on icing simulation according to the present invention; Figure 2 is a flowchart illustrating the process of collecting historical identified terrain according to the present invention; Figure 3 is a flowchart illustrating the process of obtaining the priority of each historical identified terrain according to the present invention; Figure 4 is a flowchart illustrating the process of outputting the predicted icing features of the corresponding terrain to be identified from the icing simulation model according to the present invention; Figure 5 is a flowchart illustrating the process of making specific adjustments to the feature patterns corresponding to the icing micro-topography based on the deviation data according to the present invention. Detailed Implementation

[0018] 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.

[0019] As shown in Figures 1-5, one of the objectives of this invention is to provide an intelligent identification method for icing micro-topography of transmission lines based on icing simulation, comprising the following steps: S1, determining the location data of the transmission line, and obtaining remote sensing data covering the transmission line based on the location data; obtaining precise location data and multi-source remote sensing data covering the transmission line to provide a data foundation for subsequent feature extraction and model construction; in S1, by establishing a communication connection with the transmission line management terminal, obtaining the location data of the transmission line from the transmission line management terminal, and simultaneously obtaining multi-source satellite remote sensing data, filtering the remote sensing data covering the transmission line based on the location data of the transmission line; establishing an encrypted communication connection with the transmission line management terminal through the TCP / IP protocol to obtain data access permissions; wherein, the location data includes transmission line corridor vector data, tower coordinate data, and line attribute information; obtaining multi-source satellite remote sensing data through a satellite data interface (such as the ESA Copernicus Data Center API), the sources of which include SAR data, multispectral data, DEM data, high-resolution data, and meteorological data, covering the transmission line corridor and the area on both sides.

[0020] Based on the vector data of the transmission line corridor, a spatial overlay analysis method was used to select remote sensing data covering the transmission line corridor and a 5km radius on both sides; S2, an icing micro-topography database was established, and the micro-topography of the icing micro-topography database was classified. Then, the topographic features, airflow features, and water vapor features of the terrain to be identified along the transmission line were extracted based on the remote sensing data; a classified icing micro-topography database was established, and the topographic-airflow-water vapor three-dimensional features of the terrain to be identified were extracted to provide data carriers and feature dimensions for intelligent analysis.

[0021] The steps in S2 are as follows: S2.1, collect historical identified terrain data and establish an icing micro-topography database based on the historical identified terrain data. At the same time, set five types of micro-topography, and then classify the icing micro-topography database according to the five types of micro-topography, so that the icing micro-topography database is divided into five types of icing micro-topography databases; collect historical icing micro-topography survey data, icing fault records and corresponding remote sensing image data along the transmission line, screen valid samples (remove records with vague terrain descriptions and missing icing data), and establish a structured database. The fields include historical terrain ID, geographical coordinates, terrain description, icing characteristic parameters (icing thickness, type, duration), and corresponding remote sensing data index. Then, identify high mountain watershed type, terrain uplift type, pass type, canyon wind corridor type, and water vapor enhancement type as the core classification types, and define the basic morphological description of each type of micro-topography; In S2.1, the five types of micro-topography include high mountain watershed type, terrain uplift type, pass type, canyon wind corridor type, and water vapor enhancement type.

[0022] Based on the morphological characteristics and icing mechanisms of historically identified terrain, they are categorized into five micro-topographic sub-databases, forming sub-databases for high mountain watersheds, terrain uplift, mountain passes, canyons and wind tunnels, and water vapor enhancement. S2.2: Based on remote sensing data, terrain features corresponding to the terrain to be identified for power transmission lines are extracted. Using DEM data, a neighborhood analysis algorithm is employed to extract elevation, slope, aspect, relative height, and terrain roughness. Ridge density is extracted using a ridge / valley detection algorithm. For airflow characteristics, wind speed amplification coefficients are retrieved based on SAR data. Terrain contraction (ratio of cross-sectional area between narrow and open sections) and canyon width-to-depth ratio (ratio of average width to maximum depth) are calculated using terrain vector analysis, and wind direction matching is calculated in conjunction with prevailing wind direction data. For water vapor characteristics, water bodies are identified using the NDWI index based on multispectral data. The minimum distance from the terrain to the water body and the area of ​​the surrounding water body are calculated. Water vapor convergence intensity and cloud / fog coverage frequency are retrieved based on meteorological data.

[0023] S3. Prioritize the topographic features, airflow features, and water vapor features extracted in S2 by combining them with historically identified terrain features from various icing micro-topographic databases. Obtain the priority of each historically identified terrain. Combine the priority of the historically identified terrain with the corresponding micro-topographic-specific feature patterns to establish icing simulation models for various icing micro-topographic databases. Focus on highly correlated historical data through priority filtering to construct micro-topographic-specific icing simulation models, improving prediction accuracy. The steps in S3 are as follows: S3.1. Extract the topographic features, airflow features, and water vapor features of historically identified terrain features from various icing micro-topographic databases. Simultaneously, combine the topographic features, airflow features, and water vapor features extracted in S2.2 to perform priority analysis on the historically identified terrain, obtaining the priority of each historically identified terrain. The steps are as follows: From the five types of icing micro-topographic sub-databases (high mountain watershed type, topographic lifting type, etc.), retrieve the topographic features (slope, aspect, etc.) and airflow features of each historically identified terrain. Features (such as wind speed amplification factor) and water vapor features (such as NDWI index) are used to form a historical feature matrix. Then, the min-max normalization method is used to normalize the historical features and the terrain features to be identified extracted in S2.2 to eliminate dimensional differences and map the feature values ​​to the [0,1] interval. Based on the normalized feature matrix, the cosine similarity algorithm is used to calculate the similarity between the terrain features to be identified and each historical identified terrain feature. The historical identified terrains are sorted from high to low according to the similarity value. The higher the similarity, the higher the priority. The top 30% of high-priority historical samples are selected for subsequent model construction. The higher the similarity between terrain features, airflow features, and water vapor features, the higher the priority. In S3.2, in various icing micro-terrain databases, specific feature rules for micro-terrain are set, and combined with historical identified terrains and their corresponding priorities, an icing simulation model is established so that each type of icing micro-terrain database has a corresponding icing simulation model.

[0024] In S3.2, the icing simulation model adopts a WRF-LightGBM hybrid model framework. The WRF model is used for meteorological element forecasting, and the output meteorological data serves as the input features for the LightGBM algorithm. Initial meteorological field data is input, with a grid resolution of 1km × 1km and an integration time step of 60s set. The model outputs forecast data for meteorological elements such as temperature, humidity, and wind speed for the next 1-7 days. The WRF output meteorological data and the three-dimensional features of high-priority historical samples are used as input, and icing thickness / rate is used as output to construct a regression prediction model. The model was tested, and parameters were calibrated for different micro-topographic types, as follows: For high mountain watershed types, the focus was on calibrating altitude (error ≤ 50m) and wind speed amplification factor (error ≤ 0.1); for topographic lifting types, the focus was on calibrating slope (error ≤ 2°) and aspect (error ≤ 10°); for mountain pass types, the focus was on calibrating topographic contraction (error ≤ 0.05); for canyon wind tunnel types, the focus was on calibrating width-to-depth ratio (error ≤ 0.1) and wind direction matching degree (error ≤ 5°); for water vapor enhancement types, the focus was on calibrating minimum distance to water bodies (error ≤ 200m) and relative humidity (error ≤ 3%), using the following formulas: ;in, Let i be the standardized feature value of the terrain to be identified. The i-th standardized feature value of the historical topography. For the total number of features, The similarity score ranges from 0 to 1, with values ​​closer to 1 indicating higher similarity. ;in, These are the calibrated parameter values. Predict parameter values ​​for the model. These are actual measured parameter values. (This is a calibration coefficient, ranging from 0.9 to 1.1, adjusted according to the micro-topography type). ;in, Mean square error, For high-priority historical sample numbers, Let be the actual icing feature value of the k-th sample. S4. Input the terrain features, airflow features, and water vapor features into the icing simulation model. The icing simulation model outputs the predicted icing features. At the same time, extract the historical icing features of the terrain to be identified based on remote sensing data. Combine the predicted icing features with the historical icing features for similarity screening and select the icing micro-topography database corresponding to the predicted icing features with the highest similarity. By combining the predicted icing features output by the icing simulation model with the historical icing features for similarity matching, determine the micro-topography category to which the terrain to be identified belongs. The steps of S4 are as follows: S4.1. Input the terrain features, airflow features, and water vapor features of the terrain to be identified into the corresponding icing micro-topography databases of various types. The icing simulation model outputs predicted icing features for the terrain to be identified. The steps are as follows: The three-dimensional features (terrain features, airflow features, and water vapor features) of the terrain to be identified extracted in S2.2 are input into the icing simulation models (high mountain watershed type model, terrain lifting type model, etc.) corresponding to the five types of icing micro-terrain databases. The WRF-LightGBM hybrid calculation process of each icing simulation model is started, and the predicted icing features (including icing thickness, icing rate, icing duration, and icing type) of the terrain to be identified corresponding to each type of model are output. The predicted icing features are classified according to the micro-terrain type to form a "micro-terrain type - predicted icing feature" correspondence table.

[0025] S4.2. Obtain historical icing features of the terrain to be identified based on remote sensing data. Then, combine the predicted icing features with the historical icing features for similarity screening to obtain the similarity between the predicted and historical icing features. Finally, select the icing micro-topography database of the icing simulation model corresponding to the predicted icing feature with the highest similarity. The steps are as follows: Based on the geographical coordinates of the terrain to be identified, retrieve the historical icing monitoring data and corresponding remote sensing-derived icing features of the area from the icing micro-topography database as the comparison benchmark. Use the min-max normalization method to unify the dimensions of the predicted and historical icing features, mapping the feature values ​​to the [0,1] interval. Calculate the similarity between each set of predicted and historical icing features using the Euclidean distance algorithm. Then, sort the results according to the similarity (the smaller the Euclidean distance, the higher the similarity). Select the icing simulation model corresponding to the predicted icing feature with the highest similarity. The micro-topography database to which it belongs is the matching database for the terrain to be identified. The formula is as follows: ;in, The distance is Euclidean (the smaller the distance, the higher the similarity). The number of icing characteristics, Let be the standardized value of the i-th predicted icing feature. is the standardized value of the i-th historical icing feature.

[0026] S5. Obtain deviation data based on predicted icing features and historical icing features. Adjust the feature patterns using the deviation data. Based on the adjusted feature patterns, create personalized icing micro-topography and input it into the corresponding icing micro-topography database. Optimize feature patterns through deviation correction to generate personalized icing micro-topography and update the database, achieving self-iteration of the method. The steps of S5 are as follows: S5.1. Obtain deviation data by combining the predicted icing features of the icing simulation model selected in S4.2 with historical icing features, obtaining the deviation data between the predicted and historical icing features. Align the predicted icing features output by the highest similarity icing simulation model selected in S4.2 with the corresponding historical icing features according to dimensions such as "icing thickness, icing rate, and icing duration." Then calculate the absolute and relative deviations between the predicted and historical icing features under each dimension to form a deviation dataset. S5.2. Adjust the feature patterns corresponding to the icing micro-topography based on the deviation data, obtaining the feature patterns corresponding to the terrain to be identified. Based on the feature patterns corresponding to the terrain to be identified, combine terrain features, airflow features, and water vapor features to create personalized... The following steps are taken to identify and input the icing micro-topography into the selected icing micro-topography database: For the selected icing micro-topography database (such as the mountain pass type sub-database), extract its preset micro-topography-specific feature patterns (such as terrain contraction ≥ 0.6, wind speed amplification factor ≥ 1.5). Then, based on the effective deviation data, use the deviation weighted correction method to adjust the core parameter thresholds of the feature patterns (e.g., if the predicted deviation of icing thickness is -5mm, adjust the wind speed amplification factor threshold from 1.5 to 1.55). Integrate the adjusted feature patterns with the three-dimensional features of the terrain to be identified (terrain, airflow, water vapor features) to generate a personalized icing micro-topography description that includes micro-topography type, core feature parameters, icing risk level, and anti-icing suggestions. Then, input the personalized icing micro-topography information into the selected icing micro-topography database, synchronously update the historical sample database and feature pattern parameter table, and associate it with the corresponding remote sensing data index.

[0027] The second objective of this invention is to provide an intelligent identification system for icing micro-topography of transmission lines based on icing simulation, including any of the above-mentioned intelligent identification methods for icing micro-topography of transmission lines based on icing simulation, comprising a data acquisition unit, a model building unit, an icing prediction unit, and a terrain recording unit; the data acquisition unit is used to determine the location data of the transmission line, acquire remote sensing data covering the transmission line based on the location data, establish an icing micro-topography database, classify the icing micro-topography database into micro-topography, and then extract the terrain features, airflow features, and water vapor features of the terrain to be identified on the transmission line based on the remote sensing data; the model building unit is used to perform priority analysis by combining the extracted terrain features, airflow features, and water vapor features with the historical identified terrain of various icing micro-topography databases, and obtain the topographic features, airflow features, and water vapor features of each icing micro-topography database. The system prioritizes historically identified terrain features and, based on these priorities and the specific characteristics of each micro-topography, establishes icing simulation models for various icing micro-topography databases. The icing prediction unit inputs terrain, airflow, and water vapor features into the icing simulation model, which then outputs predicted icing features. Simultaneously, it extracts historical icing features from remote sensing data, combines the predicted and historical icing features for similarity filtering, and selects the icing micro-topography database corresponding to the most similar predicted icing features. The terrain entry unit obtains deviation data based on the predicted and historical icing features, performs specific adjustments based on the deviation data, and constructs personalized icing micro-topography based on the adjusted features, which is then entered into the corresponding icing micro-topography database.

[0028] The foregoing has shown and described the basic principles, main features, and advantages of the present invention. Those skilled in the art should understand that the present invention is not limited to the above embodiments. The embodiments and descriptions in the specification are merely preferred examples and are not intended to limit the invention. Various changes and modifications can be made to the invention without departing from its spirit and scope, and all such changes and modifications fall within the scope of the present invention as claimed. The scope of protection of the present invention is defined by the appended claims and their equivalents.

Claims

1. A method for intelligent identification of icing micro-topography of transmission lines based on icing simulation, characterized in that: The process includes the following steps: S1. Determine the location data of the transmission line and obtain remote sensing data covering the transmission line based on the location data; S2. Establish an icing micro-topography database, classify the icing micro-topography database into micro-topography categories, and then extract the topographic features, airflow features, and water vapor features of the terrain to be identified along the transmission line based on the remote sensing data; S3. Combine the topographic features, airflow features, and water vapor features extracted in S2 with the historical identified terrains in various icing micro-topography databases for priority analysis, obtain the priority of each historical identified terrain, and combine the priority of the historical identified terrains with the corresponding micro-topography-specific feature patterns to classify various types of icing micro-topography. S4. Input the topographic features, airflow features, and water vapor features into the icing simulation model. The icing simulation model outputs predicted icing features. At the same time, extract the historical icing features of the terrain to be identified based on remote sensing data. Combine the predicted icing features with the historical icing features for similarity screening. Select the icing micro-topography database corresponding to the predicted icing features with the highest similarity. S5. Obtain deviation data based on the predicted icing features and historical icing features. Make specific adjustments to the features based on the deviation data. Based on the adjusted features, form personalized icing micro-topography and enter it into the corresponding icing micro-topography database.

2. The intelligent identification method for icing micro-topography of transmission lines based on icing simulation according to claim 1, characterized in that: In step S1, a communication connection is established with the transmission line management terminal to obtain the location data of the transmission line from the terminal, and remote sensing data from multiple satellites is also obtained. The remote sensing data covering the transmission line is filtered based on the location data. The location data includes transmission line corridor vector data, tower coordinate data, and line attribute information. The sources of the remote sensing data include SAR data, multispectral data, DEM data, high-resolution data, and meteorological data, and the coverage area is the transmission line corridor and the area on both sides.

3. The intelligent identification method for icing micro-topography of transmission lines based on icing simulation according to claim 1, characterized in that: The steps of S2 are as follows: S2.1: Collect historical identified terrain and establish an icing micro-terrain database based on the historical identified terrain. At the same time, set five types of micro-terrain and then classify the icing micro-terrain database according to the five types of micro-terrain, so that the icing micro-terrain database is divided into five types of icing micro-terrain databases; S2.2: Extract the terrain features, airflow features and water vapor features of the terrain to be identified corresponding to the transmission line based on remote sensing data.

4. The intelligent identification method for icing micro-topography of transmission lines based on icing simulation according to claim 3, characterized in that: In S2.1, the five types of micro-topography include high mountain watershed type, topographic uplift type, mountain pass type, canyon wind tunnel type, and water vapor enhancement type.

5. The intelligent identification method for icing micro-topography of transmission lines based on icing simulation according to claim 1, characterized in that: The steps in S3 are as follows: S3.1 Extract the terrain features, airflow features, and water vapor features of historically identified terrains from various icing micro-terrain databases. At the same time, combine the terrain features, airflow features, and water vapor features extracted in S2.2 to perform priority analysis on the historically identified terrains and obtain the priority of each historically identified terrain. The higher the similarity between the terrain features, airflow features, and water vapor features, the higher the priority. S3.2 Set the characteristic rules specific to each micro-terrain in various icing micro-terrain databases, and combine the historically identified terrains and their corresponding priorities to establish an icing simulation model, so that each type of icing micro-terrain database has a corresponding icing simulation model.

6. The intelligent identification method for icing micro-topography of transmission lines based on icing simulation according to claim 5, characterized in that: In S3.2, the icing simulation model adopts the WRF-LightGBM hybrid model framework. The WRF model is used to forecast meteorological elements, and the output meteorological data is used as the input features of the LightGBM algorithm. Parameter calibration is performed for different micro-topographic types, as follows: for high mountain watershed types, the focus is on calibrating altitude and wind speed amplification coefficient; for topographic lifting types, the focus is on calibrating slope and aspect; for mountain pass types, the focus is on calibrating topographic contraction; for canyon wind tunnel types, the focus is on calibrating width-to-depth ratio and wind direction matching; and for water vapor enhancement types, the focus is on calibrating minimum distance to water bodies and relative humidity.

7. The intelligent identification method for icing micro-topography of transmission lines based on icing simulation according to claim 1, characterized in that: The steps of S4 are as follows: S4.1, input the terrain features, airflow features and water vapor features of the terrain to be identified into the icing simulation model corresponding to various icing micro-terrain databases, and output the predicted icing features of the terrain to be identified by the icing simulation model. S4.

2. Obtain historical icing features of the terrain to be identified based on remote sensing data. Then, combine the predicted icing features with the historical icing features for similarity screening to obtain the similarity between the predicted icing features and the historical icing features. Finally, select the icing micro-terrain database of the icing simulation model corresponding to the predicted icing feature with the highest similarity.

8. The intelligent identification method for icing micro-topography of transmission lines based on icing simulation according to claim 1, characterized in that: The steps in S5 are as follows: S5.1, Combine the predicted icing features of the icing simulation model selected in S4.2 with historical icing features to obtain deviation data, and obtain the deviation data between the predicted icing features and historical icing features; S5.2, Based on the deviation data, make specific adjustments to the feature patterns corresponding to the icing micro-topography, obtain the feature patterns corresponding to the terrain to be identified, and combine the feature patterns corresponding to the terrain to be identified with terrain features, airflow features and water vapor features to form a personalized icing micro-topography, and enter it into the selected icing micro-topography database.

9. An intelligent identification system for icing micro-topography of transmission lines based on icing simulation, used to implement the intelligent identification method for icing micro-topography of transmission lines based on icing simulation as described in any one of claims 1-8, characterized in that: The system includes a data acquisition unit, a model building unit, an icing prediction unit, and a terrain recording unit. The data acquisition unit determines the location data of the transmission line, acquires remote sensing data covering the transmission line based on the location data, establishes an icing micro-topography database, classifies the icing micro-topography database into micro-topography types, and then extracts the terrain features, airflow features, and water vapor features of the terrain to be identified along the transmission line based on the remote sensing data. The model building unit performs priority analysis by combining the extracted terrain features, airflow features, and water vapor features with historical identified terrains from various icing micro-topography databases, obtains the priority of each historical identified terrain, and combines the priority of the historical identified terrains with the corresponding micro-topography-specific features. The system establishes icing simulation models for various icing micro-topographic databases. The icing prediction unit inputs topographic features, airflow features, and water vapor features into the icing simulation model, which outputs predicted icing features. Simultaneously, it extracts historical icing features of the terrain to be identified based on remote sensing data, combines the predicted icing features with historical icing features for similarity filtering, and selects the icing micro-topographic database corresponding to the predicted icing features with the highest similarity. The terrain input unit obtains deviation data based on the predicted icing features and historical icing features, performs custom adjustments based on the deviation data, and forms personalized icing micro-topographic features based on the adjusted features, which are then entered into the corresponding icing micro-topographic database.

Images