Power line icing prediction method and system

Abstract

The invention relates to the technical field of electric power icing prediction, and discloses an electric power line icing prediction method and system, and the method comprises the steps: obtaining electric power data and meteorological data, and carrying out the data preprocessing; and extracting a first feature of the line icing to obtain a second feature and a third feature of the line icing, constructing a line icing prediction model through a machine learning algorithm, predicting the icing condition of the power line in real time, and outputting a line icing thickness prediction result. According to the method, physical prediction and a machine learning model are fused, multi-dimensional information such as weather, line states and historical data is combined, the icing thickness prediction precision is remarkably improved, real-time monitoring and dynamic early warning can be achieved, minute-level icing thickness updating and risk grading are supported, ice melting or dispatching instructions are triggered in time, and the ice melting efficiency is improved. The method still has relatively high robustness in complex environments such as mountainous areas, high altitudes and extreme weather; the manual inspection frequency can be reduced, and the operation and maintenance cost is reduced.

Description

Technical Field

[0001] The present invention relates to the technical field of power icing prediction, and particularly to a method and system for predicting power line icing. Background Art

[0002] During the long-term outdoor operation of the power grid, transmission lines are extremely vulnerable to various natural disasters. Among them, line icing may not only cause line galloping, insulator flashover, and pole tilt, but also, as the ice accretion increases, it is very likely to cause serious accidents such as wire breakage and tower collapse, resulting in large-scale power outages in the surrounding areas of the icing region and even the collapse of the local power grid, causing huge economic losses. Summary of the Invention

[0004] In view of the above existing problems, the present invention is proposed. Therefore, the present invention provides a method for predicting power line icing to solve the above problems.

[0005] To solve the above technical problems, the present invention provides the following technical solutions: In a first aspect, the present invention provides a method for predicting power line icing, including: Obtaining power data and meteorological data, and performing data preprocessing; Based on the processed power data and meteorological data, extracting the first features of line icing, and performing feature analysis according to the first features to obtain the second features of line icing; Based on the first features and the second features, constructing a third feature related to the icing height according to the icing characteristics; Based on the first features, the second features, and the third features, constructing a line icing prediction model through a machine learning algorithm to predict the icing situation of the power line in real time and output the prediction result of the line icing thickness.

[0006] As a preferred solution of the method for predicting power line icing according to the present invention, wherein: obtaining power data and meteorological data, and performing data preprocessing includes: The power data includes line data and geographical data. The line data includes wire tension, inclination angle, vibration frequency, current load, resistance change, and icing thickness. The geographical data includes altitude, terrain, line orientation, wire material, and diameter; The meteorological data includes temperature, humidity, wind speed, wind direction, precipitation, atmospheric pressure, dew point temperature, and sunshine intensity; The data preprocessing includes data cleaning, time alignment, and data standardization.

[0007] As a preferred solution of the method for predicting power line icing according to the present invention, wherein: obtaining the second features of line icing includes: Define a historical time window, divide the historical time window at a preset sampling interval to determine sample collection points, obtain power data, line data, and meteorological data at the sample collection points, and calculate statistical quantities for the data to extract the first feature of line icing. According to the first feature, analyze and obtain the second feature of line icing, where the second feature includes meteorological features, line features, and geographical features. Among them, the meteorological feature obtains the water content and wet bulb temperature through the meteorological data in the first feature. The line feature obtains the degree of change in conductor tension, vibration energy, and line heat loss through the line data in the first feature, and analyzes the vibration spectrum. The geographical feature corrects the temperature and wind speed data in the first feature through the geographical data in the first feature.

[0008] As a preferred solution of the power line icing prediction method of the present invention, among them: the meteorological feature includes: Calculate the water content in the air through the statistical quantities of air pressure, temperature, and wind speed at the sample collection point, expressed as: ; Among them, represents the air density, represents the average precipitation, represents the speed of supercooled water droplets, represents the maximum wind speed, represents the time period; The wet bulb temperature is expressed as: ; Among them, represents the temperature, represents the relative humidity. If the wet bulb temperature , then the current state of the line is the icing state. If the wet bulb temperature , then the current state of the line is the non-icing state.

[0009] As a preferred solution of the power line icing prediction method of the present invention, among them: the line feature includes: Calculate the degree of change in conductor tension through the line data at the sample collection point, expressed as: ; Among them, represents the conductor tension at the current moment, represents the conductor tension at the previous moment, represents the sampling interval; The vibration energy is expressed as: ; Among them, represents the vibration frequency, represents the wire mass, represents the amplitude.

[0010] As a preferred solution of the power line icing prediction method described in the present invention, among them: constructing the third feature related to the icing height includes: By calculating the temperature difference between the wire surface temperature and the environment, the line temperature difference is obtained, expressed as: ; ; Among them, represents the wire surface temperature, represents the average environmental temperature, represents the average current load of the wire, represents the AC resistance of the wire, represents the convective heat transfer coefficient, represents the surface area of the wire; By the icing thickness and the thermal conductivity of ice, the icing thermal resistance is calculated, expressed as: ; Among them, represents the icing thickness, represents the thermal conductivity of ice.

[0011] As a preferred solution of the power line icing prediction method described in the present invention, among them: outputting the prediction result of the line icing thickness includes: By the water content, wind speed, and wire area at the prediction moment, the line icing growth rate is calculated, and based on the current icing thickness, the first icing thickness prediction value at the prediction moment is obtained. The line icing growth rate is expressed as: ; Among them, represents the water content in the air, represents the wind speed, represents the cross-sectional area of the wire, represents the collision efficiency, represents the freezing efficiency; Construct a line icing prediction model through the XGBoost algorithm, use historical data for model training and model evaluation, generate a feature matrix based on the water content, wet bulb temperature, wire tension change degree, vibration energy line heat loss, line temperature difference, icing thermal resistance, and wind speed features, perform feature correlation analysis on the feature matrix and the icing thickness, screen out the features strongly correlated with icing, and use them as model inputs to predict the icing thickness residual correction value; Add the predicted value of the first ice coating thickness to the corrected value of the ice coating thickness residual to obtain the final predicted value of the transmission line ice coating thickness.

[0012] In a second aspect, the present invention provides a power line ice coating prediction system, including: A collection module, configured to obtain power data and meteorological data and perform data preprocessing; A first calculation module, configured to extract the first features of the transmission line ice coating based on the processed power data and meteorological data, and perform feature analysis according to the first features to obtain the second features of the transmission line ice coating; A second calculation module, configured to construct a third feature related to the ice coating height according to the ice coating characteristics based on the first features and the second features; A prediction module, configured to construct a transmission line ice coating prediction model through a machine learning algorithm based on the first features, the second features, and the third features, predict the ice coating condition of the power line in real time, and output the predicted result of the transmission line ice coating thickness.

[0013] In a third aspect, the present invention provides an electronic device, including: A memory and a processor; The memory is used to store computer-executable instructions, and the processor is used to execute the computer-executable instructions. When the computer-executable instructions are executed by the processor, the steps of the power line ice coating prediction method are implemented.

[0014] In a fourth aspect, the present invention provides a computer-readable storage medium, which stores computer-executable instructions. When the computer-executable instructions are executed by a processor, the steps of the power line ice coating prediction method are implemented.

[0015] Compared with the prior art, the beneficial effects of the present invention are as follows: By integrating physical prediction and machine learning models, and combining multi-dimensional information such as meteorology, line status, and historical data, the present invention significantly improves the prediction accuracy of the ice coating thickness; it can monitor and give dynamic early warnings in real time, support minute-level updates of the ice coating thickness and risk grading, trigger de-icing or dispatching instructions in a timely manner, and still has strong robustness in complex environments such as mountainous areas, high altitudes, and extreme weather; it can reduce the frequency of manual inspections and lower the operation and maintenance costs. Description of the Drawings

[0016] In order to more clearly illustrate the technical solutions of the embodiments of the present invention, the drawings required for the description of the embodiments will be briefly introduced below. Obviously, the drawings in the following description are only some embodiments of the present invention. For those of ordinary skill in the art, other drawings can be obtained based on these drawings without creative efforts.

[0017] Figure 1Schematic diagram of the overall process of the ice coating prediction method for a power line according to an embodiment of the present invention. Detailed implementation manners

[0018] To make the above objects, features, and advantages of the present invention more obvious and understandable, the following will describe in detail the specific implementation manners of the present invention with reference to the accompanying drawings of the specification. Obviously, the described embodiments are part of the embodiments of the present invention, rather than all embodiments. All other embodiments obtained by those of ordinary skill in the art based on the embodiments of the present invention without creative efforts shall fall within the protection scope of the present invention.

[0019] Refer to Figure 1 , an embodiment of the present invention provides a method for predicting ice coating on a power line, including: S101, obtain power data and meteorological data, and perform data preprocessing; S102, based on the processed power data and meteorological data, extract the first features of line ice coating, and perform feature analysis according to the first features to obtain the second features of line ice coating; S103, based on the first features and the second features, construct the third features related to the ice coating height according to the ice coating characteristics; S104, based on the first features, the second features, and the third features, construct a line ice coating prediction model through a machine learning algorithm, predict the ice coating condition of the power line in real time, and output the prediction result of the line ice coating thickness.

[0020] In a preferred implementation manner, obtaining power data and meteorological data and performing data preprocessing includes: The power data includes line data and geographical data. The line data includes wire tension, inclination angle, vibration frequency, current load, resistance change, ice coating thickness, and the geographical data includes altitude, terrain, line direction, wire material, and diameter; The meteorological data includes temperature, humidity, wind speed, wind direction, precipitation, atmospheric pressure, dew point temperature, and sunshine intensity; The data preprocessing includes data cleaning, time alignment, and data standardization.

[0021] Specifically, the power data and meteorological data are mainly obtained through camera images (ice coating visual detection), fiber optic sensing data, micro-meteorological station data, etc. Among them, data cleaning in data preprocessing mainly includes processing missing values (interpolation method, filling with data from neighboring stations), outliers (3σ principle or box plot filtering); time alignment is used to unify the timestamps of different data; standardization is used to perform Min-Max or Z-Score standardization on parameters such as temperature and humidity.

[0022] In a preferred embodiment, based on the processed power data and meteorological data, the first features of line icing are extracted, and feature analysis is performed according to the first features, and the second features of line icing are obtained, including: Define a historical time window, divide the historical time window at a preset sampling interval, determine the sample collection points, obtain the power data, line data, and meteorological data of the sample collection points, and calculate the statistical quantities of the data to extract the first features of line icing; According to the first features, the second features of line icing are analyzed and obtained. The second features include meteorological features, line features, and geographical features, and the sensitive factors for ice accretion growth can be identified; Among them, the meteorological features obtain the water content and wet-bulb temperature through the meteorological data in the first features; The line features obtain the wire tension change degree, vibration energy, and line heat loss through the line data in the first features, and analyze the vibration spectrum; The geographical features correct the temperature and wind speed data in the first features through the geographical data in the first features.

[0023] Specifically, according to the time scale of ice formation (usually several hours to dozens of hours), the length of the historical time window can be determined, and the sliding window size (such as 3 hours, 6 hours) can be selected. Among them, the time window can determine the periodicity by calculating the autocorrelation coefficient (ACF) of the ice thickness, or can be determined by the RMSE of the prediction model under different window lengths (such as 1h / 3h / 6h) through grid search. Then, according to the size of the time window, the sampling interval (which can be 10 / 20 / 30 minutes) is defined, and then the sample collection points within the time window are determined, and the statistical quantities of the time series data within the window are calculated. The types of statistical quantities can be the mean, standard deviation, maximum value, or minimum value within the historical time window. The first features include characteristic data such as wire tension, vibration frequency, current load, resistance, temperature, humidity, wind speed, precipitation, and their statistical quantities.

[0024] It should be noted that ice formation is an accumulation process that depends on continuous meteorological conditions (such as continuous low temperature and high humidity). By calculating the statistical quantities within the time window, the overall trend of short-term meteorological changes is captured, providing context information in the time dimension for the model. Through sliding window statistics and comprehensive statistical quantity calculation, the original data is transformed into intermediate features reflecting time, space, and physical laws.

[0025] In a preferred embodiment, the meteorological features include: By calculating the statistical quantities of the air pressure, temperature, and wind speed at the sample collection points, the water content in the air can be calculated, which can quantify the concentration of freezable water droplets in the air and directly affect the ice accretion rate, expressed as: ; Among them, represents the air density, represents the average precipitation, represents the supercooled water droplet velocity, which is set according to meteorological observations, represents the maximum wind speed, represents the time period, which can be taken as 1 hour to match the precipitation unit; The wet-bulb temperature reflects the actual cooling capacity of the air and is expressed as: ; Among them, represents the temperature, represents the relative humidity. If the wet-bulb temperature , then the current state of the line is the icing state. If the wet-bulb temperature , then the current state of the line is the non-icing state.

[0026] It should be noted that a single meteorological parameter (such as temperature) is difficult to directly reflect the icing potential. It is necessary to integrate multiple parameters to construct a comprehensive index. For example, the water content combines precipitation, wind speed, and temperature to directly quantify the concentration of freezeable water droplets in the air.

[0027] In a preferred embodiment, the line characteristics include: The line data passing through the sample collection point is used to calculate the change degree of the conductor tension, which reflects the change rate. A rapid increase in tension may indicate ice accumulation and is expressed as: ; Among them, represents the conductor tension at the current moment, represents the conductor tension at the previous moment, represents the sampling interval; The vibration energy is used to evaluate the severity of the conductor vibration and is expressed as: ; Among them, represents the vibration frequency, represents the conductor mass, represents the amplitude.

[0028] It should be noted that the dynamic changes in line sensor data (such as conductor tension, vibration frequency) directly reflect the physical effects of ice accumulation. Ice accumulation increases the weight of the conductor, resulting in an increase in tension and a decrease in vibration frequency.

[0029] In an alternative embodiment, the heat loss can be calculated by multiplying the current load, the resistance change, and the time period to obtain the heat loss over a certain period of time, which reflects the energy loss of the line. The vibration spectrum analysis mainly performs a Fourier transform on the wire vibration signal to extract the main frequency component. The main frequency represents the frequency with the highest energy in the spectrum, and the frequency band energy ratio represents the proportion of the energy in the low-frequency band (0 - 5 Hz) to the total energy. Icing may cause an increase in low-frequency vibration. The increased ice adds mass to the wire, resulting in a decrease in the main vibration frequency and an increase in the proportion of low-frequency energy.

[0030] In an alternative embodiment, the geographical features can be used to correct the temperature by altitude, taking into account the influence of altitude on temperature. Generally, for every 100-meter increase in altitude, the temperature drops by approximately 0.6 °C. The temperature can be corrected based on this relationship, that is, the temperature difference is calculated based on the height difference between the current location and the altitude, and the collected temperature is corrected; the wind speed is corrected according to the terrain. For different terrains, such as valleys and ridges, the wind speed will be affected, and the wind speed is corrected based on the terrain characteristics and meteorological principles.

[0031] In a preferred embodiment, constructing the third feature related to the icing height includes: The third feature includes the line temperature difference and the icing thermal resistance, constructing high-value features; By calculating the temperature difference between the wire surface temperature and the environment, the line temperature difference is obtained, expressed as: ; ; where represents the wire surface temperature, represents the average environmental temperature, represents the average current load of the wire, represents the AC resistance of the wire, represents the convective heat transfer coefficient, with a value range of 5 - 25 W / (m²·K), represents the surface area of the wire; The icing thermal resistance is calculated through the icing thickness and the thermal conductivity of ice. The greater the thermal resistance, the more difficult it is for the wire to dissipate heat, and the slower the ice growth. The icing thermal resistance is expressed as: ; where represents the icing thickness, represents the thermal conductivity of ice.

[0032] It should be noted that icing is the result of thermodynamic equilibrium. The heat dissipation of the wire and the heat absorption of ice formation need to satisfy the law of conservation of energy. By constructing features (such as wire surface temperature, thermal resistance) to directly quantify the physical conditions for icing formation, combined with thermodynamics and the icing model, high-value physical features are constructed to enhance the interpretability and prediction accuracy of the model.

[0033] In a preferred embodiment, the predicted results of the ice thickness on the output line include: Calculate the line ice accretion growth rate based on the water content, wind speed, and wire area at the prediction moment, and obtain the first predicted value of the ice thickness at the prediction moment according to the current ice thickness. The line ice accretion growth rate is expressed as: ; Wherein, represents the water content in the air, represents the wind speed, which determines the water droplet transportation rate, represents the cross-sectional area of the wire, which affects the water droplet collision area, represents the collision efficiency, with a value range of [0.6, 1], indicating the probability of supercooled water droplets colliding with the wire, represents the freezing efficiency, with a value range of [0.5, 1], indicating the proportion of actually frozen droplets among the colliding droplets; Construct a line ice accretion prediction model through the XGBoost algorithm, use historical data for model training (70% for the training set, 20% for the validation set, and 10% for the test set) and model evaluation, optimize the model parameters, generate a feature matrix based on the water content, wet-bulb temperature, wire tension change degree, vibration energy, line heat loss, line temperature difference, ice accretion thermal resistance, and wind speed characteristics, conduct feature correlation analysis between the feature matrix and the ice thickness, screen out the features strongly correlated with ice accretion, and use them as model inputs to predict the ice thickness residual correction value; Add the first predicted value of the ice thickness and the ice thickness residual correction value to obtain the final predicted value of the ice thickness on the line.

[0034] Specifically, first, the increase in ice mass per unit time can be obtained from the ice accretion growth rate of the line, and the change in ice accretion mass over time can be directly predicted. Combining with the ice density, it can be converted into the growth of ice accretion thickness. Secondly, XGBoost (eXtreme Gradient Boosting), an efficient gradient boosting decision tree algorithm, is suitable for processing structured data and can automatically capture non-linear relationships and complex patterns. The specific parameter design includes learning rate, tree depth, subsample ratio, column sampling ratio, and regularization term to reduce model complexity and prevent overfitting. Among them, the learning rate can be tuned through grid search. The ice accretion prediction model is trained with historical power data, meteorological data, actual ice accretion thickness, ice accretion residual values, etc., and the model performance is evaluated through cross-validation. The parameters and weights are adjusted to continuously optimize the model. Feature matrix and ice accretion thickness are subjected to feature correlation analysis, and features strongly correlated with ice accretion are selected as model inputs, that is, a data matrix containing the above features is generated (each column is a feature, and each row is a sample at a time point). Dimensionality reduction methods such as principal component analysis (PCA) can be used to analyze the linear relationship between multiple features, or partial least squares (PLS) can be used to analyze the potential relationship between multiple features and ice accretion thickness to find the feature combination and potential factors that have the greatest impact on ice accretion thickness. Covariance analysis can also be used to calculate the Pearson correlation coefficient and select features greater than the threshold, and key features are screened through association analysis to reduce noise and improve model efficiency.

[0035] It should be noted that the ice accretion thickness prediction model in this embodiment constructs a multi-dimensional feature machine learning model by integrating meteorological data, line status, and physical laws, and finally realizes accurate prediction, and is applicable to ice accretion prediction under extreme meteorological conditions.

[0036] The present invention significantly improves the prediction accuracy of ice accretion thickness by integrating physical prediction and machine learning models, combining multi-dimensional information such as meteorology, line status, and historical data. The prediction error (such as RMSE) is reduced by 30% - 50%. It can monitor in real time and give dynamic early warnings, support minute-level ice accretion thickness updates and risk grading, and trigger de-icing or dispatching instructions in a timely manner. Moreover, it still has strong robustness in complex environments such as mountainous areas, high altitudes, and extreme weather. It can reduce the frequency of manual inspections and lower the operation and maintenance costs.

[0037] The above is a schematic solution of a method for predicting ice accretion on a power line in this embodiment. It should be noted that the technical solution of the power line ice accretion prediction system belongs to the same concept as the technical solution of the above power line ice accretion prediction method. For the details not described in detail in the technical solution of the power line ice accretion prediction system in this embodiment, reference can be made to the description of the technical solution of the above power line ice accretion prediction method.

[0038] The power line ice accretion prediction system in this embodiment includes: A collection module, configured to obtain power data and meteorological data, and perform data preprocessing; A first calculation module, configured to extract a first feature of line icing based on the processed power data and meteorological data, and perform feature analysis according to the first feature to obtain a second feature of line icing; A second calculation module, configured to perform feature correlation analysis based on the first feature and the second feature to construct a third feature strongly related to icing; A prediction module, configured to construct a line icing prediction model through a machine learning algorithm based on the first feature, the second feature, and the third feature, and predict the icing condition of the power line in real time, and output a prediction result of the line icing thickness.

[0039] This embodiment further provides an electronic device, applicable to the situation of power line icing prediction, including: A memory and a processor; the memory is used to store computer-executable instructions, and the processor is used to execute the computer-executable instructions to implement the method for realizing power line icing prediction as proposed in the above embodiment.

[0040] This embodiment further provides a storage medium, on which a computer program is stored, and when the program is executed by a processor, it implements the method for realizing power line icing prediction as proposed in the above embodiment.

[0041] The storage medium proposed in this embodiment and the method for realizing power line icing prediction proposed in the above embodiment belong to the same inventive concept. The technical details not described in detail in this embodiment can be referred to the above embodiment, and this embodiment has the same beneficial effects as the above embodiment.

[0042] Through the above description of the embodiments, those skilled in the art can clearly understand that the present invention can be implemented by means of software and necessary general hardware, and of course, it can also be implemented by hardware, but in many cases, the former is a better implementation method. Based on such an understanding, the technical solution of the present invention, in essence, or the part that contributes to the prior art, can be embodied in the form of a software product, and this computer software product can be stored in a computer-readable storage medium, such as a floppy disk, a read-only memory (ROM), a random access memory (RAM), a flash memory (FLASH), a hard disk, or an optical disc of a computer, etc., including several instructions for causing a computer device (which can be a personal computer, a server, or a network device, etc.) to execute the methods of the various embodiments of the present invention.

[0043] It should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention rather than to limit it. Although the present invention has been described in detail with reference to the preferred embodiments, those of ordinary skill in the art should understand that the technical solutions of the present invention can be modified or equivalently replaced without departing from the spirit and scope of the technical solutions of the present invention, and all of them should be covered by the scope of the claims of the present invention.

Claims

1. A method for predicting icing of power lines, characterized in that: include: Obtain power data and meteorological data, and perform data preprocessing; Extracting a first feature of line icing based on the processed power data and meteorological data, and performing feature analysis based on the first feature to obtain a second feature of line icing; Based on the first feature and the second feature, and according to the ice coating characteristics, constructing a third feature related to the ice coating height; Based on the first feature, the second feature and the third feature, a line icing prediction model is constructed through a machine learning algorithm to predict the icing condition of the power lines in real time and output the prediction results of the line icing thickness.

2. The power line icing prediction method according to claim 1, characterized in that: Obtaining power data and meteorological data and performing data preprocessing include: The power data includes line data and geographic data. The line data includes conductor tension, inclination, vibration frequency, current load, resistance change, and ice thickness. The geographic data includes altitude, terrain, line direction, conductor material, and diameter. The meteorological data include temperature, humidity, wind speed, wind direction, precipitation, atmospheric pressure, dew point temperature, and sunshine intensity; The data preprocessing includes data cleaning, time alignment and data standardization.

3. The power line icing prediction method according to claim 2, characterized in that: The second feature of line icing includes: Define a historical time window, divide the historical time window according to a preset sampling interval, determine a sample collection point, obtain power data, line data, and meteorological data at the sample collection point, calculate statistics on the data, and extract a first feature of line icing; According to the first feature, analyzing and obtaining a second feature of line icing, the second feature including meteorological features, line features and geographical features; Wherein, the meteorological feature obtains water content and wet-bulb temperature through the meteorological data in the first feature; The line feature obtains the conductor tension variation, vibration energy and line heat loss through the line data in the first feature, and analyzes the vibration spectrum; The geographic feature modifies the temperature and wind speed data in the first feature by using the geographic data in the first feature.

4. The power line icing prediction method according to claim 3, characterized in that: The meteorological characteristics include: The water content in the air is calculated by the statistics of air pressure, temperature and wind speed at the sample collection point, which is expressed as: ； in, represents the air density, represents the mean precipitation amount, represents the supercooled water droplet velocity, Indicates the maximum wind speed. Indicates a time period; The wet bulb temperature is expressed as: ； in, Indicates temperature, Indicates relative humidity. If the wet-bulb temperature , the line is currently in an icing state. If the wet-bulb temperature , the current state of the line is not frozen.

5. The method for predicting icing of power lines according to claim 3, characterized in that: The line features include: The wire tension variation is calculated through the line data of the sample collection point, which is expressed as: ； in, Indicates the wire tension at the current moment, represents the wire tension at the previous moment, represents the sampling interval; The vibration energy is expressed as: ； in, represents the vibration frequency, Indicates the quality of the wire. Indicates amplitude.

6. The method for predicting icing of power lines according to claim 3, characterized in that: Constructing the third feature related to ice cover height includes: By calculating the difference between the conductor surface temperature and the ambient temperature, the line temperature difference is obtained, which is expressed as: ； ； in, Indicates the conductor surface temperature, Indicates the average ambient temperature, Indicates the average current load of the conductor, is the AC resistance of the conductor, represents the convective heat transfer coefficient, Represents the surface area of ​​the conductor; The ice thermal resistance is calculated by the ice thickness and the thermal conductivity of ice, which is expressed as: ； in, Indicates the ice thickness, represents the thermal conductivity of ice.

7. The method for predicting icing of power lines according to claim 6, characterized in that: The output line ice thickness prediction results include: The line ice growth rate is calculated by the water content, wind speed and conductor area at the prediction time, and the first ice thickness prediction value at the prediction time is obtained according to the ice thickness at the current time. The line ice growth rate is expressed as: ； in, Indicates the water content in the air. Indicates wind speed, Indicates the cross-sectional area of ​​the conductor. represents the collision efficiency, represents the freezing efficiency; A line icing prediction model is constructed by using the XGBoost algorithm, and historical data is used for model training and model evaluation. A feature matrix is ​​generated according to the water content, wet bulb temperature, conductor tension change, vibration energy, line heat loss, line temperature difference, icing thermal resistance and wind speed characteristics. The feature matrix is ​​subjected to feature correlation analysis with icing thickness, and features strongly correlated with icing are screened out and used as model input to predict the residual correction value of icing thickness. The first ice thickness prediction value is added to the ice thickness residual correction value to obtain a final line ice thickness prediction value.

8. A power line icing prediction system, using the method according to any one of claims 1 to 7, characterized in that: include, The collection module is used to obtain power data and meteorological data and perform data preprocessing; A first calculation module is used to extract a first feature of line icing based on the processed power data and meteorological data, and perform feature analysis based on the first feature to obtain a second feature of line icing; A second calculation module, configured to construct a third feature related to ice cover height based on the first feature and the second feature and according to ice cover characteristics; The prediction module is used to build a line icing prediction model based on the first feature, the second feature and the third feature through a machine learning algorithm, predict the icing condition of the power line in real time, and output the prediction result of the line icing thickness.

9. An electronic device, characterized in that: include: Memory and processor; The memory is used to store computer-executable instructions, and the processor is used to execute the computer-executable instructions. When the computer-executable instructions are executed by the processor, the steps of the method described in any one of claims 1 to 7 are implemented.

10. A computer-readable storage medium, characterized in that: It stores computer executable instructions, which, when executed by a processor, implement the steps of the method described in any one of claims 1 to 7.

Images