Method and device for predicting height change of ultra-high voltage transmission conductor

Abstract

The invention discloses a method and a device for predicting height change of an ultra-high voltage transmission conductor, and relates to the technical field of intelligent operation and maintenance of transmission lines. According to the main technical scheme, the method comprises the steps that operation state parameter data and static parameter data corresponding to a current wire are acquired; the operation state parameter data of the current conductor and the inherent parameter data of the line are processed by using the pre-constructed conductor height change prediction model so as to predict and output the actual conductor sag data of the current conductor, and the conductor height change prediction model is based on a mechanism model based on a physical principle, so that the prediction reliability is ensured, and the prediction accuracy is improved. Meanwhile, a data driving method is used for correcting and optimizing the model, the prediction precision is improved, advantage complementation is achieved, and therefore a high-precision conductor sag prediction result is obtained; and real-time height data corresponding to the current lead is determined in combination with the static geographic parameter data of the current lead, and the real-time height data can be but not limited to the height to the ground or the clearance height to a ground obstacle.

Description

Technical Field

[0001] This application relates to the field of intelligent operation and maintenance technology for power transmission lines, and in particular to a method and device for predicting changes in the height of ultra-high voltage transmission conductors. Background Technology

[0002] The height variation of UHV transmission lines is the result of the coupling effects of multiple physical fields, including thermal expansion and contraction, mechanical creep, and load changes. Thermal expansion and contraction, caused by Joule heating from the conductor's load current and changes in ambient temperature, leads to significant linear expansion and contraction, becoming the most significant factor affecting sag (the height difference between the lowest point and the suspension point of the conductor). Mechanical creep is the irreversible plastic elongation of the conductor under long-term tension, resulting in a permanent increase in sag; this is a slow process that requires long-term consideration. Load changes are caused by external loads such as the significant increase in weight due to icing and wind pressure from strong winds, which dynamically alter the conductor's tension and shape. The combined effect of these three factors results in a dynamic and complex variation in the height of UHV transmission lines.

[0003] Currently, using real-time data reflecting changes in conductor height obtained through measurement or acquisition to calculate the actual conductor height presents numerous practical challenges. These challenges include: firstly, in high-voltage, high-electromagnetic-field environments, directly deploying ranging radars, laser ranging sensors, and other sensors on energized conductors for measurement is technically difficult, costly, and inconvenient to maintain; secondly, accurately analyzing conductor height changes requires the simultaneous and precise acquisition of environmental parameters such as conductor current, conductor temperature, wind speed and direction, and icing thickness to construct a complete input set, and the synchronous and accurate acquisition of this data is itself extremely challenging. These two factors combined significantly increase the difficulty of obtaining real-time changes in the height of UHV transmission conductors.

[0004] Therefore, given the dynamic and complex nature of the height changes in UHV transmission lines, providing a method to more accurately measure these changes is an urgent technical problem to be solved. Summary of the Invention

[0005] This application provides a method and apparatus for predicting the height change of UHV transmission lines. The main purpose is to use a conductor height change prediction model to comprehensively consider the thermal and mechanical properties of the conductor and data-driven correction to predict high-quality real-time conductor sag. Since the change in "conductor height" is directly determined by the change in sag, this allows for the acquisition of more accurate conductor height change results. Even when dealing with the dynamic and complex changes in the height of UHV transmission lines, this application provides a better solution for more accurately obtaining the height change of UHV transmission lines.

[0006] To achieve the above objectives, this application mainly provides the following technical solutions: The first aspect of this application provides a method for predicting the height variation of ultra-high voltage transmission lines, the method comprising: Obtain the current operating status parameter data and static parameter data corresponding to the conductor. The operating status parameter data includes at least conductor current data, conductor temperature data, meteorological data, and tension data. The static parameter data includes at least line inherent parameter data and static geographical parameter data. The static geographical parameter data includes at least ground obstacle height data. The operating state parameter data and the line's inherent parameter data are processed using a pre-constructed conductor height change prediction model to output the target sag data of the current conductor. The conductor height change prediction model is a mechanism data fusion model constructed for the current conductor based on the heat balance equation, the state equation, the catenary equation, and a data-driven correction algorithm. The heat balance equation, the state equation, and the catenary equation are used to predict the real-time sag data of the current conductor based on its thermal and mechanical properties. The data-driven correction algorithm is used to correct and optimize the real-time sag data to obtain the target sag data. Based on the target sag data and the static geographic parameter data corresponding to the current traverse, determine the real-time height data corresponding to the current traverse.

[0007] A second aspect of this application provides a device for predicting changes in the height of ultra-high voltage transmission lines, the device comprising: The first acquisition unit is used to acquire the current operating status parameter data and static parameter data corresponding to the conductor. The operating status parameter data includes at least conductor current data, conductor temperature data, meteorological data and tension data. The static parameter data includes at least line inherent parameter data and static geographical parameter data. The static geographical parameter data includes at least ground obstacle height data. The processing unit is used to process the operating state parameter data and the line inherent parameter data using a pre-built conductor height change prediction model, and output the target sag data of the current conductor. The conductor height change prediction model is a mechanism data fusion model constructed for the current conductor based on the heat balance equation, the state equation, the catenary equation, and the data-driven correction algorithm. The heat balance equation, the state equation, and the catenary equation are used to predict the real-time sag data of the current conductor based on the thermal and mechanical properties of the conductor, and the data-driven correction algorithm is used to correct and optimize the real-time sag data to obtain the target sag data. The determining unit is used to determine the real-time height data corresponding to the current traverse based on the target sag data and the static geographic parameter data corresponding to the current traverse.

[0008] A third aspect of this application provides a computer-readable storage medium storing a computer program that, when executed by a processor, implements the method for predicting changes in the height of ultra-high voltage transmission lines as described above.

[0009] A fourth aspect of this application provides an electronic device, the device including at least one processor, and at least one memory and bus connected to the processor; The processor and the memory communicate with each other via the bus. The processor is used to call program instructions in the memory to execute the method for predicting changes in the height of ultra-high voltage transmission lines as described above.

[0010] By employing the above-described technical solution, the technical solution provided in this application has at least the following advantages: This application provides a method and apparatus for predicting the height change of ultra-high voltage transmission lines. The method involves acquiring operational status parameter data and static parameter data corresponding to the current transmission line. The operational status parameter data includes at least conductor current data, conductor temperature data, meteorological data, and tension data. The static parameter data includes at least line-specific parameter data and static geographic parameter data. A pre-constructed transmission line height change prediction model is used to process the operational status parameter data and line-specific parameter data to predict the actual sag data of the current transmission line. Since the transmission line height change prediction model is based on a mechanistic model based on physical principles, the reliability of the prediction is ensured. Simultaneously, a data-driven method is used to correct and optimize the model, improving prediction accuracy and achieving complementary advantages, thereby obtaining a high-quality prediction result for the current transmission line sag. Furthermore, the static geographic parameter data of the current transmission line is combined to determine the real-time height data corresponding to the current transmission line. This real-time height data may be, but is not limited to, the height above the ground or the clearance height above ground obstacles.

[0011] Compared to existing technologies, which rely heavily on direct measurement and environmental parameters, making it difficult to obtain accurate real-time data reflecting changes in conductor height and thus hindering the acquisition of high-precision UHV transmission conductor height changes, this application uses a conductor height change prediction model that integrates the conductor's thermal and mechanical properties, data-driven correction, and other aspects to predict high-quality real-time conductor sag. Since the change in "conductor height" is directly determined by the change in sag, this allows for the acquisition of high-precision conductor height change results. Even when dealing with the dynamic and complex changes in UHV transmission conductor height, this provides a better solution for more accurately obtaining UHV transmission conductor height changes.

[0012] The above description is only an overview of the technical solution of this application. In order to better understand the technical means of this application and to implement it in accordance with the contents of the specification, and to make the above and other objects, features and advantages of this application more obvious and understandable, specific embodiments of this application are given below. Attached Figure Description

[0013] Various other advantages and benefits will become apparent to those skilled in the art upon reading the following detailed description of preferred embodiments. The accompanying drawings are for illustrative purposes only and are not intended to limit the scope of this application. Furthermore, the same reference numerals denote the same parts throughout the drawings. In the drawings: Figure 1 A flowchart illustrating a method for predicting changes in the height of ultra-high voltage transmission lines, provided as an embodiment of this application; Figure 2 A block diagram illustrating the composition of a device for predicting changes in the height of ultra-high voltage transmission lines, provided in an embodiment of this application; Figure 3 A flowchart illustrating the correction and optimization using a data-driven correction algorithm provided in this application embodiment; Figure 4 A block diagram of another device for predicting the height change of ultra-high voltage transmission lines provided in this application embodiment. Detailed Implementation

[0014] Exemplary embodiments of the present application will now be described in more detail with reference to the accompanying drawings. While exemplary embodiments of the present application are shown in the drawings, it should be understood that the present application may be implemented in various forms and should not be limited to the embodiments set forth herein. Rather, these embodiments are provided so that this application will be thorough and complete, and will fully convey the scope of the present application to those skilled in the art.

[0015] The height variation of UHV transmission lines is the result of the coupling effects of multiple physical fields, including thermal expansion and contraction, mechanical creep, and load changes. Thermal expansion and contraction: Changes in conductor load current (Joule heating) and ambient temperature lead to conductor temperature variations, resulting in significant linear expansion and contraction. This is the most significant factor affecting sag (the height difference between the lowest point of the conductor and the suspension point). Mechanical creep: Under long-term tension, the conductor undergoes irreversible plastic elongation, leading to a permanent increase in sag. This is a slow but necessary process measured in years. Load changes: External loads such as icing (significantly increasing weight) and strong winds (generating wind pressure) dynamically alter the conductor's tension and shape. The combined effect of these three factors results in a dynamic and complex variation in the height of UHV transmission lines.

[0016] Currently, obtaining real-time data that accurately reflects changes in conductor height is fundamental to studying conductor height variations; however, this faces numerous challenges in practice. Examples are as follows: Direct measurement is difficult: The working principle involves high voltage and high electromagnetic field environments that interfere with the normal operation of the sensor, affecting the accuracy of the measurement data. Furthermore, deploying sensors on live conductors presents numerous technical challenges, and subsequent maintenance is inconvenient and costly. The inability to directly and accurately obtain relevant data about the conductor, such as its real-time position, makes it difficult to accurately grasp the dynamic changes in the conductor's height.

[0017] Incomplete Environmental Parameters: The working principle is that conductor height changes are affected by various environmental parameters, requiring simultaneous acquisition of data such as conductor current, conductor temperature, wind speed and direction, and icing thickness to comprehensively analyze the causes and patterns of conductor height changes. However, in practice, deploying equipment for some environmental parameters is difficult and costly, or the acquired data is difficult to synchronize and accurately obtain. Therefore, the lack of key environmental parameters results in an incomplete input dataset, leading to biases in the analysis and prediction of conductor height changes and affecting the accuracy of the research results.

[0018] To this end, the inventors discovered through research that if the reliance on direct measurements and environmental parameters is reduced, and instead they are used merely as auxiliary data bases and applied to a conductor height change prediction model, then by utilizing the model's high-precision prediction processing to address conductor height changes caused by different internal and external factors, the height changes of UHV transmission conductors can be obtained more accurately, thus providing a better solution for studying conductor height changes.

[0019] Based on the above considerations, embodiments of this application provide a method for predicting changes in the height of ultra-high voltage transmission lines, such as... Figure 1 As shown, the following specific steps are provided in this embodiment of the invention: 101. Obtain the current operating status parameter data and static parameter data corresponding to the conductor. The operating status parameter data shall include at least conductor current data, conductor temperature data, meteorological data and tension data. The static parameter data shall include at least line inherent parameter data and static geographical parameter data. The static geographical parameter data shall include at least ground obstacle height data.

[0020] The current operating parameters of the conductor, such as conductor current, conductor temperature, meteorological data, and tension, are the dynamic parameters of the conductor and are the key inputs for subsequent models to perform real-time sag calculations. The real-time performance and accuracy of these parameter data directly affect the timeliness and accuracy of subsequent model predictions.

[0021] The current static parameter data for the conductors includes: inherent line parameters and static geographic parameters. Inherent line parameters include span length, wire diameter, suspension point height, conductor type, etc. These are fixed parameters determined during the construction and design of the transmission line. Span length is the horizontal distance between two adjacent towers, wire diameter is the diameter of the conductor, suspension point height is the height at which the conductor is suspended from the tower, and conductor type (such as ACSR, AAAC) represents the material and structural characteristics of the conductor. These parameters were obtained during the initial stage of system construction by consulting line design drawings and engineering documents, entered into the data layer, and stored long-term. During subsequent operation, unless the line is modified, these data remain unchanged.

[0022] The inherent parameters of the line (such as span, suspension point height, and conductor type) are the basic inputs of the mechanistic model. The span and suspension point height determine the geometric boundary of the catenary equation. The elastic modulus and thermal expansion coefficient associated with the conductor type directly affect the calculation accuracy of the heat balance equation and the state equation, ensuring that the sag prediction conforms to the actual physical characteristics of the line.

[0023] In addition, static geographic parameter data must include at least ground obstacle height data. After the real-time sag of the traverse predicted by the subsequent 102 model output, it is necessary to further use the static traverse parameter data to calculate the current real-time height data of the traverse, including the following examples: By combining the line's "suspension point height" (the fixed height of the conductor on the tower) and "ground / obstacle elevation" (static geographic parameters), we can further calculate: the conductor's "height above ground" (the vertical distance from the lowest point of the conductor to the ground), or the conductor's "clearance height" relative to other objects (such as the safe distance from trees and buildings).

[0024] Furthermore, in this embodiment of the application, an "indirect sensing" method is used instead of a direct measurement method to obtain the operating status parameter data corresponding to the current conductor. Examples include the following: (1) Use a data acquisition and monitoring control system (SCADA) to collect the current data of the conductor in real time, and convert the current signal into an electrical signal through a sensor to obtain the current data of the conductor.

[0025] (2) Deploy a dual-spectrum thermal imager (visible light + thermal infrared) to monitor the vegetation below the line and to measure the temperature of the conductor in a non-contact manner. This is the most critical data input.

[0026] Dual-spectrum thermal imagers possess both visible light and thermal infrared imaging capabilities. Visible light imaging can be used to monitor the growth of vegetation along power lines, promptly identifying potential safety hazards caused by excessively tall vegetation. Thermal infrared imaging utilizes the differences in infrared radiation emitted by objects to measure their temperature. Different conductor temperatures result in different infrared energy emissions. The thermal imager receives the infrared radiation signals from the conductors, converts them into temperature data, and achieves non-contact measurement of conductor temperature.

[0027] On the one hand, monitoring the growth of vegetation below ground provides a basis for subsequent tree obstacle risk warning and maintenance work; on the other hand, non-contact measurement of conductor temperature avoids the difficulties and dangers of directly contacting live conductors to measure temperature. The obtained conductor temperature data is a key input parameter for model calculation and is crucial for accurately predicting changes in conductor height.

[0028] (3) Install a micro weather station on the tower to measure local wind speed, sunshine and ambient temperature and other meteorological data within the span.

[0029] The micro-weather station, installed on a steel tower, measures local meteorological parameters such as wind speed, solar radiation intensity, and ambient temperature within a specified span. Wind speed sensors acquire wind speed data by measuring airflow velocity, solar radiation sensors acquire solar radiation data by detecting solar radiation intensity, and temperature sensors directly measure ambient air temperature. These sensors collect data in real time and transmit it to a data center via a data transmission module.

[0030] Obtaining accurate local meteorological parameters within the span avoids the potential representativeness issues associated with using regional meteorological station data. Local meteorological parameters significantly affect conductor temperature, tension, and sag variations. Accurate meteorological data provides the model with inputs that better reflect actual operating conditions, improving the model's accuracy in predicting conductor height changes.

[0031] (4) Deploy tilt sensors on the tower to monitor the angle change of the suspension point and infer the conductor tension.

[0032] Tilt sensors, installed at appropriate locations on the tower (such as near the conductor suspension point), can measure changes in the tower's tilt angle in both vertical and horizontal directions. Changes in conductor tension alter the tension on the tower, consequently causing changes in its tilt angle. By establishing a mathematical model relating the tower's tilt angle to the conductor tension, the magnitude of the conductor tension can be calculated backwards from the tilt angle data measured by the tilt sensor.

[0033] This method avoids the technical difficulties and safety risks of directly deploying tension sensors on conductors in high-voltage, high-electromagnetic-field environments, acquiring conductor tension data through indirect measurement. Conductor tension is a crucial parameter for calculating conductor sag and height changes. This method provides accurate tension data for the model, helping to improve the model's prediction accuracy.

[0034] 102. The pre-built conductor height change prediction model is used to process the operating status parameter data and the inherent parameter data of the line, and output the target sag data of the current conductor.

[0035] The conductor height change prediction model is a mechanism data fusion model constructed for the current conductor based on the heat balance equation, state equation, catenary equation, and data-driven correction algorithm. The heat balance equation, state equation, and catenary equation are used to predict the real-time sag data of the current conductor based on the thermal and mechanical properties of the conductor, and the data-driven correction algorithm is used to correct and optimize the real-time sag data to obtain the target sag data.

[0036] In this embodiment of the application, the constructed conductor height change prediction model is based on a mechanism model based on physical principles to ensure the reliability of the prediction. At the same time, a data-driven method is used to correct and optimize the model to improve the prediction accuracy and achieve complementary advantages.

[0037] Specifically, the heat balance equation, state equation, and catenary equation corresponding to the current conductor are used as the mechanistic model for the current conductor. This mechanistic model, based on physical principles and relevant industry standards (such as IEEE Std 738), describes the thermal and mechanical properties of the conductor by establishing mathematical equations, thereby calculating key parameters such as the conductor's temperature, sag, and shape, providing a solid physical basis for predicting conductor height changes. The mechanistic model is then used to process the current conductor's operating state parameter data to predict and output real-time conductor sag data. This application's embodiments explain the specific implementation process of "obtaining real-time conductor sag data," including the following: The heat balance equation provided in this application embodiment is as follows: (1) ; in, I It is electric current, and when current flows through a wire, it generates... Joule heat; This refers to the solar heat absorption power, which is the solar radiation energy absorbed by the conductor. and These represent convective and radiative heat dissipation power, respectively, with the conductor dissipating heat to the surrounding environment through convection and radiation. Thermal capacity is the capacity of a conductor, reflecting its ability to store heat. The temperature change rate represents how the conductor temperature changes over time. This equation describes the dynamic balance between heat generation by the current, heat absorption by the environment, and heat dissipation. When heat generation and heat dissipation reach equilibrium, the conductor temperature tends to stabilize; when heat generation exceeds heat dissipation, the conductor temperature rises; conversely, the temperature decreases.

[0038] Formula (1) combines input meteorological parameters (such as wind speed, ambient temperature, and solar radiation intensity) and conductor material properties (such as resistivity and coefficient of thermal expansion) to calculate the core and surface temperatures of the conductor. Under steady-state conditions, it can determine the stable temperature of the conductor under specific current and environmental parameters; under transient conditions (such as sudden load changes), it can consider the delayed effect of heat capacity, accurately calculate the temperature change process of the conductor, and provide key temperature parameters for subsequent calculation of conductor sag.

[0039] Furthermore, equation (1) is used to calculate the real-time temperature data of the conductor under different operating conditions. The specific implementation methods are as follows: Based on the heat balance equation, a numerical iterative algorithm is used for calculation. First, based on known initial conditions (such as initial conductor temperature, current, meteorological parameters, etc.), an assumed conductor temperature value is substituted into the heat balance equation to calculate the heat generation and dissipation power. If the calculated heat generation power and dissipation power are not equal, the assumed temperature value is adjusted, and the calculation is repeated. This iterative process continues until the heat generation power and dissipation power reach equilibrium. The temperature value obtained at this point is the actual conductor temperature. During the calculation process, the influence of meteorological parameters and conductor material properties on various heat outputs is fully considered.

[0040] This implementation method can accurately calculate the temperature of the conductor under different working conditions, providing accurate temperature data for subsequent calculation of conductor stress and sag using the equation of state. It is a key link connecting the thermal and mechanical properties of the conductor and directly affects the accuracy of conductor height change prediction.

[0041] Furthermore, by substituting the real-time temperature data of the current conductor into the state equation for processing, the real-time stress data of the current conductor is output. The specific implementation method includes the following: In this application embodiment, a state equation is provided, as shown in formula (2): ; in, For elastic modulus, The coefficient of thermal expansion is... For temperature T1 The stress at that time; this equation of state is used to describe the relationship between conductor stress and sag under different meteorological conditions. For example, when the temperature changes from... T1 Change to T2 Stress changes The sag can be calculated using the equation of state. Temperature changes cause thermal expansion and contraction of the conductor, leading to changes in conductor length. These changes in length are constrained by tension, resulting in changes in stress. This equation comprehensively considers the thermal expansion and elastic properties of the conductor, establishing a quantitative relationship between temperature and stress changes. Therefore, the sag can be calculated using stress changes.

[0042] This implementation method uses state equations to calculate changes in conductor stress based on variations in meteorological conditions such as temperature. Combined with the catenary equation, it can accurately calculate changes in sag, thereby enabling prediction of conductor height changes under different meteorological conditions. Especially under extreme conditions (such as high current or freezing), it can ensure the reliability of the prediction.

[0043] Furthermore, by substituting the real-time stress data of the current conductor into the catenary equation for processing, the real-time sag data corresponding to the current conductor is output. The specific implementation method includes the following: In this embodiment of the application, a catenary equation is provided, as shown in the following formula (3); ; Among them, The stress at the lowest point of the conductor. The equation represents the comprehensive load, including self-weight, ice weight, and wind pressure, with x representing the horizontal distance. It describes the sag curve of the conductor under the influence of self-weight, tension, and temperature. The conductor can be considered a flexible body, naturally forming a sag curve under loads such as self-weight and tension. By introducing parameters such as the stress at the lowest point of the conductor and the comprehensive load, this equation accurately describes the shape of the conductor's sag curve from a mechanical perspective. A specific functional relationship exists between the horizontal distance x and the vertical height y, reflecting the height variation of the conductor at different locations.

[0044] Based on this equation, and given parameters such as the lowest point stress of the conductor, the comprehensive specific load, and the horizontal distance, this implementation method can calculate the vertical height of the conductor at the corresponding horizontal position, thereby determining the real-time shape and sag of the conductor.

[0045] After using the above-mentioned mechanism model to predict the real-time sag data of the output conductor, the embodiments of this application use a data-driven method to correct and optimize the model, which may include, but is not limited to, the following: S201. Use machine learning algorithms to learn the residuals of the machine model. The residuals include at least the effects of creep, unmeasured micro-wind vibrations, and model simplification error factors on the prediction of conductor height changes.

[0046] S202. Use the residuals of the machine model as machine learning compensation to correct the real-time sag data in order to obtain the target sag data.

[0047] In this embodiment, machine learning algorithms (such as Gradient Boosting and LSTM networks) are used to learn the residuals of the mechanistic model. First, a large amount of historical data is collected, including input parameters (such as current and meteorological parameters) under different operating conditions, the predicted sag of the mechanistic model, and the corresponding measured sag. Then, the difference between the measured sag and the predicted sag of the mechanistic model is calculated; this difference is the residual of the mechanistic model. The residual includes the influence of factors not fully considered by the mechanistic model, such as creep, unmeasured micro-wind vibrations, and model simplification errors. Using the input parameters as features and the residuals as labels, the machine learning model is trained, enabling the model to accurately predict the residuals based on the input parameters.

[0048] In actual prediction, the predicted sag from the mechanistic model is added to the residual predicted by the machine learning model to obtain a more accurate conductor sag prediction result. Through machine learning compensation, errors caused by the mechanistic model neglecting some complex factors and simplifying the processing can be effectively compensated, improving the accuracy of conductor height change prediction. In particular, it has a good correction effect on the influence of slowly changing factors such as conductor creep during long-term operation.

[0049] S203. During the operation of the guide height change prediction model, the guide point cloud data obtained by UAV lidar inspection is used to reverse-calibrate and correct the parameters of the mechanism model in order to obtain the target sag data.

[0050] During operation, the model utilizes high-precision point cloud data (considered "true values") of the traverse line acquired through occasional UAV LiDAR inspections to perform reverse calibration and correction on the parameters of the mechanistic model (such as elastic modulus and creep coefficient). First, high-precision 3D point cloud data of the traverse line is acquired through UAV LiDAR inspections, and the actual sag and shape of the traverse line are obtained through data processing. Then, the actual sag is compared with the predicted sag of the mechanistic model under the same input parameters, and the deviation between the two is calculated. Based on the deviation, optimization algorithms (such as least squares method and gradient descent method) are used to adjust the parameters in the mechanistic model in reverse, minimizing the deviation between the predicted sag and the actual sag. Furthermore, as new point cloud data is continuously acquired, the model parameters are continuously calibrated and optimized to ensure the long-term accuracy of the model.

[0051] It can promptly correct deviations in the mechanistic model parameters caused by long-term operation of the conductor and environmental changes, ensuring the predictive accuracy of the mechanistic model during long-term operation. Simultaneously, by utilizing high-precision point cloud data as "true values," it provides a reliable basis for model calibration, enabling the model to better adapt to changes in the actual operating state of the conductor, further improving the reliability and accuracy of conductor height change prediction.

[0052] It should be noted that S201-S202 above are machine learning compensation methods provided, and S203 is an adaptive calibration method provided. In the embodiments of this application, any one of these methods, or a combination of the two methods, can be used to correct and optimize the mechanism model in order to obtain higher quality real-time conductor sag.

[0053] 103. Based on the target sag data and static geographic parameter data corresponding to the current traverse, determine the real-time height data corresponding to the current traverse.

[0054] Conductor sag refers to the vertical distance between the lowest point of the suspension curve and the suspension point of a conductor under the action of its own weight, tension, and external loads (such as icing and wind pressure). The change in "conductor height" is directly determined by the change in sag: when the sag increases, the lowest point of the conductor will decrease, which means that the "height above the ground" (or the clearance height above other objects) of the conductor will decrease; when the sag decreases, the lowest point of the conductor will rise, and the "height above the ground" will increase.

[0055] In this application embodiment, the static parameter data includes at least the line's inherent parameter data and static geographic parameter data. The static geographic parameter data includes at least the ground obstacle height data. For example, by combining the line's "suspension point height" (the fixed height of the conductor on the tower) and "ground / obstacle elevation" (static geographic parameter), the following can be further calculated: the conductor's "height to the ground" (the vertical distance from the lowest point of the conductor to the ground), or the conductor's "clearance height" relative to other objects (such as the safe distance to trees and buildings).

[0056] In some modified embodiments, for using data-driven methods to correct and optimize the model, the machine learning compensation methods provided in S201-202 in this application are explained in detail as follows: Based on the IEEE Std 738 standard, the conductor temperature is calculated using the heat balance equation (Formula (1)). Combined with the catenary equation, the theoretical sag value under current meteorological conditions is predicted. Parameter calibration and residual correction: Adjusting the parameters of the state equation obtained from the fitting of the actual sag. Compared with theoretical value In contrast, the elastic modulus E and the coefficient of thermal expansion in the equation of state are optimized using the Levenberg-Marquardt algorithm. For example, when At that time, increase To reflect the true effect of thermal expansion of the conductor.

[0057] This enables data-driven residual learning by constructing a Long Short-Term Memory (LSTM) network model to learn historical residual sequences. The model maps sag predictions to meteorological parameters (wind speed, humidity). For example, in high temperature and high humidity environments, the model automatically corrects sag predictions to compensate for nonlinear factors not considered in the mechanistic model.

[0058] In the embodiments of this application, a Long Short-Term Memory (LSTM) network model is constructed using historical residual sequences. The training data includes corresponding meteorological parameters (wind speed, humidity, etc.). LSTM networks have the ability to remember long-term dependencies and learn how residuals change with meteorological parameters and other factors. Through training, the LSTM model can accurately predict residuals under current operating conditions based on current meteorological parameters. For example, in high-temperature and high-humidity environments, the model can automatically identify the influence pattern of this environment on residuals, correct sag predictions, and compensate for nonlinear factors not considered by the mechanistic model. Furthermore, it compensates for the shortcomings of the mechanistic model by learning the variation patterns of residuals, performing a secondary correction on the mechanistic model's prediction results, making the final sag prediction more accurate. It has particularly good prediction and correction effects for nonlinear residual changes caused by meteorological parameters and other factors, improving the overall accuracy of conductor height change prediction.

[0059] In some modified embodiments, for the use of data-driven methods to correct and optimize the model, the adaptive calibration method provided in S203 in this application embodiment is explained in detail as follows: (1) High-precision point cloud data acquisition, which achieves multi-source device collaboration and spatiotemporal synchronization in data acquisition. Specifically, the implementation process includes the following: For data acquisition, multi-source device collaboration is achieved: a drone equipped with a LiDAR (Light Detection and Ranging) is used for periodic scanning. The LiDAR emits a laser beam, and the time it takes for the laser beam to travel from emission to reflection back to the sensor is measured. Combined with the drone's position and attitude information, the three-dimensional coordinates of each point on the conductor surface are calculated. Simultaneously, BeiDou positioning terminals are deployed at key locations such as towers. These terminals receive signals from multiple BeiDou satellites to determine their own high-precision coordinates (accuracy down to the millimeter level), thereby obtaining the high-precision coordinates of the towers and conductors.

[0060] The drone equipped with lidar can quickly and over a wide area acquire three-dimensional point cloud data of the conductor, covering the entire conductor situation; the high-precision coordinates provided by the Beidou positioning terminal provide an accurate spatial reference for the point cloud data, ensuring that the acquired conductor coordinate data has high accuracy and providing a reliable data foundation for subsequent model calibration.

[0061] To achieve spatiotemporal synchronization in data acquisition: Utilizing BeiDou short message communication technology, timestamps are added to the coordinate data of control points (such as BeiDou positioning terminals on towers) and the point cloud data collected by UAV lidar. Then, the data is aligned based on the timestamps to ensure that the spatial location information corresponding to the control point coordinates and the point cloud data is consistent at the same point in time, eliminating spatial location deviations caused by asynchronous data acquisition.

[0062] Achieving spatiotemporal consistency in data acquisition ensures accurate temporal and spatial correspondence between point cloud data and control point coordinate data, avoiding inaccurate spatial positioning of point cloud data due to time differences, and providing accurate spatiotemporal reference for subsequent point cloud data processing and model calibration.

[0063] (2) Data preprocessing of high-precision point cloud data. Noise filtering and point cloud segmentation are used in data preprocessing. Specifically, the implementation process includes the following: Noise filtering is employed in data preprocessing using a combination of statistical outlier removal (SOR) and voxel grid filtering. SOR calculates the average k-neighbor distance for each point, sets a standard deviation threshold (typically 3 times the standard deviation), and removes points with distances greater than this threshold. These outliers are primarily noise points caused by measurement errors, environmental interference, and other factors. Voxel grid filtering divides the point cloud space into a voxel grid of a certain size (e.g., 10cm × 10cm × 10cm). For each voxel, the centroid of all points within it is calculated, and the centroid replaces all points within that voxel, thereby reducing data redundancy and improving the efficiency of subsequent data processing.

[0064] Noise points in point cloud data are removed to improve the quality of the point cloud data and reduce the impact of noise on subsequent conductor shape reconstruction and model fitting. At the same time, voxel mesh filtering reduces the amount of data, lowers the computational cost of subsequent data processing, and improves the data processing speed, laying the foundation for fast and accurate model calibration.

[0065] For data preprocessing, point cloud segmentation is employed: Based on a weakly supervised learning framework, a small number of guide points are first manually labeled to generate pseudo-labels. Then, the point cloud data with pseudo-labels is input into a random forest model for training. The random forest model learns the differences in features (such as spatial location, reflection intensity, and normal vectors) between guide points and background points (e.g., poles, vegetation) to build a classification model. After training, the model can automatically identify guide regions in the point cloud data, segment the guide point cloud from background point clouds such as poles and vegetation, and crop out the target guide point cloud.

[0066] By separating the traverse point cloud from the background point cloud, the interference of background factors on the traverse shape analysis and model calibration is eliminated, and pure traverse point cloud data is extracted. This provides accurate data for subsequent traverse 3D morphology reconstruction and model fitting, ensuring the accuracy of model calibration.

[0067] (2) Three-dimensional morphological reconstruction and model fitting of the conductor, the specific explanation of which includes the following: In fitting the catenary equation, span identification is achieved: based on the tower point cloud data, the positions of adjacent towers are determined through methods such as cluster analysis, thereby determining the coordinates of adjacent suspension points. Then, using ArcGIS Pro's ExtractPowerLinesFromPointCloud tool, based on a preset span range (e.g., 50-500 meters) and combined with the spatial distribution characteristics of the traverse point cloud, the traverse point cloud is divided into different segments according to different spans, with each segment corresponding to a traverse span.

[0068] Accurately identifying conductor segments of different spans provides a basis for subsequent model fitting of conductors in each span, ensuring that the conductor model of each span can accurately reflect the actual shape of the conductor within that span, and avoiding model fitting errors caused by span confusion.

[0069] In fitting the catenary equation, curve fitting is achieved: the least squares method is applied to fit the catenary equation (formula (3)) for each traverse point cloud. First, based on the coordinate data of the traverse point cloud, the correspondence between the horizontal distance x and the vertical height y is determined. Then, assuming a set of and The initial value is substituted into the catenary equation to calculate the corresponding theoretical y-value, which is then compared with the y-value of the actual traverse point cloud to calculate the error. The error is then calculated by adjusting... and The optimal fitting parameters are those that minimize the sum of squared errors between the theoretical and actual values. During the fitting process, a point tolerance (e.g., 80cm) is set to control the fitting accuracy. To mitigate the impact of wind, wind correction parameters are used to eliminate wind deviation and optimize the fitting results. For line segments with spans exceeding 60 meters, an improved wind correction algorithm is employed. The actual offset of the conductor is calculated based on the point cloud distribution, and the catenary parameters are adjusted to simulate the theoretical shape under windless conditions, eliminating the interference of gusts on sag measurements.

[0070] By fitting the catenary equation, the accurate morphological parameters of each span of conductor can be obtained. and This method determines the actual sag and shape of the conductor, providing accurate actual data for subsequent comparison with the sag predicted by the mechanism model and for model parameter calibration. It also eliminates the interference of external factors such as wind on sag measurement, ensuring the accuracy of actual sag data.

[0071] In summary, the core technical means of this application's embodiment, which "processes the operating state parameter data and line inherent parameter data using a pre-constructed conductor height change prediction model to output the target sag data of the current conductor," is to complete the accurate calculation of sag through a mechanism model chain of "thermal balance equation → state equation → catenary equation," combined with data-driven correction. Specifically, this corresponds to "theoretical sag calculation" and "precise sag calibration." Theoretical sag calculation: Based on the IEEE Std 738 standard, input current (SCADA real-time data), ambient temperature / wind speed (micro-weather station data), conductor material parameters (static data), calculate conductor temperature through thermal balance equation, then substitute into state equation to calculate stress change, and finally obtain "theoretical sag" using catenary equation. Precise sag calibration: The measured sag is fitted using UAV LiDAR point cloud data, and the residual between the theoretical sag and the measured sag is compensated using an LSTM model. At the same time, parameters such as the elastic modulus are calibrated in reverse to ensure the accuracy of the calculation.

[0072] In this embodiment of the application, the influence of temperature on the calculated sag change is fully considered when using the "conductor height change prediction model" to predict sag. The specific explanation is as follows: This application takes the "influence of temperature on sag" as the core mechanism and achieves full-link quantitative calculation through a mathematical model. For example, the "temperature source" and "quantification of sag change" used in this application embodiment are explained as follows: Temperature source: It not only includes ambient temperature, but also integrates "current Joule heating" and "sunlight absorption" through the heat balance equation to accurately calculate the "actual temperature of the conductor" (instead of using only ambient temperature).

[0073] Quantification of sag variation: directly linking temperature change through the equation of state. The model uses the catenary equation to convert the stress change into the sag change. For example, if the temperature increases by 5°C, the model can directly calculate how many centimeters the sag increases, thus achieving a quantitative measurement of the "temperature → sag change".

[0074] Furthermore, the embodiments of this application also fully consider the influence of wind force on the measured sag change, and the specific explanation includes the following: This application embodiment decomposes the "influence of wind force on sag" into "static load (wind pressure)" and "dynamic disturbance (gusts)," which are calculated using models and algorithms respectively. Taking the "static wind pressure influence" and "dynamic gust correction" used in this application embodiment as examples, the explanation is as follows: Static wind pressure influence: In the "comprehensive specific load" of the catenary equation ( The document explicitly includes "wind pressure load" (the greater the wind force, the higher the load). The larger the value, the more significant the lateral shift of the conductor sag (directly quantifying the impact of wind force on the sag shape).

[0075] Dynamic gust correction: For line segments with a span of >60 meters, this application's embodiment designs a "dynamic wind offset correction algorithm". By using UAV data points to calculate the actual offset of the conductor, the catenary parameters are adjusted to simulate the "theoretical sag in a windless state", eliminating the interference of gusts on sag calculation and ensuring the accuracy of the calculation of "wind force change → sag change".

[0076] In addition, the embodiments of this application also take into account the impact of cable aging on sag calculation. However, the embodiments of this application do not adopt a "direct calculation" method, but rather an "indirect compensation" method. The specific explanation includes the following: This application does not use the method of "directly calculating the degree of aging (such as aging coefficient, remaining life)", but indirectly covers the "sag change caused by aging" through "data-driven residual compensation". The specific differences are as follows: (1) The definition of “cable aging” in this application embodiment: It is essentially the long-term accumulation of “mechanical creep”, in which the core manifestation of “cable aging” is “mechanical creep”, where the conductor undergoes irreversible plastic elongation under long-term tension, resulting in a permanent increase in sag (this is the main effect of cable aging on sag). However, this application embodiment does not define “aging coefficient” or “aging degree index” separately, but classifies “sag deviation caused by creep” as “error term that cannot be covered by the mechanism model”.

[0077] (2) Processing method of this application embodiment: Indirect compensation is achieved through "residual learning + adaptive calibration" rather than direct calculation. Since "cable aging (creep)" is a slow and nonlinear process, it cannot be directly quantified and calculated through mechanism models (such as thermal balance, catenary equation). This application embodiment adopts an "indirect compensation" strategy. Taking "residual learning" and "long-term trend calibration" used in this application embodiment as examples, the explanation is as follows: Residual Learning: Using an LSTM network to learn the residuals of "measured sag (UAV LiDAR) - theoretical sag (mechanistic model)" This residual includes errors such as "creep (aging) and unmeasured micro-wind vibrations," which is equivalent to indirectly "capturing" the effect of aging on sag through data.

[0078] Long-term trend calibration: The calibrated parameters are regressed and analyzed with historical point cloud data every quarter to identify long-term trends such as "conductor stress relaxation" (e.g., "elastic modulus decreases by 0.5% per year" in the example of this application). "Stress relaxation" is essentially the result of cable aging. The sag deviation caused by aging is indirectly compensated by adjusting the model parameters.

[0079] The conductor sag calculation method implemented in this application embodiment can calculate the sag changes caused by factors such as temperature, wind force, and cable aging, thereby improving the accuracy of conductor sag calculation.

[0080] Furthermore, based on the data-driven method for model calibration and optimization (including the calibration and optimization methods provided in S201-202 and S203 respectively), this application embodiment further provides an implementation process for dynamic updating and closed-loop calibration, specifically including the following: (1) Real-time data triggering, based on an event-driven mechanism: During the operation of the conductor, parameters such as temperature and wind speed are monitored in real time. When a sudden change in conductor temperature (e.g., ±5℃), wind speed exceeding 15m / s, or new point cloud data updates are detected, the calibration process is triggered. For example, the point cloud data is analyzed in real time by the edge computing unit to calculate the offset of the conductor coordinates. When the offset exceeds 5cm, it is considered that the conductor state has changed significantly, and the parameter optimization program is automatically started to calibrate the model parameters.

[0081] It can promptly detect significant changes in the conductor's operating status, trigger the model calibration process, ensure that the model can adapt to changes in the conductor's status in a timely manner, avoid a decrease in model prediction accuracy due to changes in the conductor's status, and guarantee the real-time performance and accuracy of model predictions.

[0082] (2) Model iteration and feedback, based on online parameter update: The latest elastic modulus E and thermal expansion coefficient obtained through parameter calibration are used to achieve the model iteration and feedback. The parameters are synchronized to the mechanistic model, updating the model's computational parameters. Simultaneously, the coefficients of the LSTM residual correction model are updated based on the new residual data. During the parameter update process, an adaptive Kalman filter algorithm is employed to dynamically adjust the process noise covariance matrix Q and the observation noise covariance matrix R based on the information covariance (i.e., the covariance of the deviation between predicted and actual values). This allows the filtering process to better track the real-time state changes of the conductor, ensuring the accuracy and stability of the model parameter updates.

[0083] Real-time online updates of model parameters ensure the model can always make predictions based on the latest conductor conditions and environmental conditions, guaranteeing the model's long-term accuracy and adaptability. The application of adaptive Kalman filtering further improves the accuracy and stability of parameter updates and reduces the impact of noise on parameter updates.

[0084] (3) Historical Data Fusion: Regression analysis is performed quarterly on the calibrated model parameters and historical point cloud data. First, the calibration parameters and corresponding point cloud data (actual sag) at different time points in the past quarter are collected. Then, through regression analysis, the relationship between model parameters and factors such as time and conductor operating status is established to identify long-term trends (such as conductor stress relaxation). For example, by comparing point cloud data and calibration parameters over three years, it is found that the elastic modulus decreases by 0.5% per year. Based on this long-term trend, the calculation method of the elastic modulus in the prediction model is adjusted to incorporate the time factor into the calculation of the elastic modulus.

[0085] By integrating historical data, it is possible to identify the changing trends of conductors during long-term operation, such as stress relaxation and material aging. By incorporating these long-term trend factors into the model, the model parameters and prediction algorithms can be further optimized, enabling the model to more accurately predict long-term changes in conductor height and improving the model's long-term prediction accuracy and reliability.

[0086] (4) Calibration effect verification and error control, using accuracy verification and dynamic error compensation; among which accuracy verification includes multi-source data comparison and statistical analysis; dynamic error compensation, such as extreme working condition correction; Provide multi-source data comparison: Use a ground-based laser rangefinder (accuracy ±2mm) to conduct on-site measurements of conductor sag and obtain the measured value of conductor sag. Compare this measured value with the sag result obtained by point cloud fitting and calculate the deviation between the two. For example, in a 1000kV line, measure the deviation between the point cloud-fitted sag and the ground-based laser rangefinder's measured value, and control the deviation within ±3cm to meet the requirements of DL / T 5092-1999 "Technical Specification for Design of 110~500kV Overhead Transmission Lines".

[0087] By comparing data from multiple sources, the accuracy of the fitted sag data in the point cloud was verified, ensuring the reliability of the point cloud data as the "true value". Simultaneously, it also provides direct verification of the model calibration effect; if the deviation between the fitted sag and the measured value is small, it indicates that the data used for model calibration is accurate and the calibration effect is good.

[0088] Provide statistical analysis: Collect a large amount of sample data (e.g., 200 samples), each sample including the model-predicted sag value and the corresponding actual sag value (e.g., point cloud fitted sag or ground-measured sag). Calculate the prediction error for each sample, and then calculate the root mean square error (RMSE) and mean absolute error (MAE) based on the error data. RMSE reflects the overall distribution of prediction errors by calculating the square root of the average of the sum of squares of all errors; MAE reflects the average level of prediction errors by calculating the average of the absolute values ​​of all errors. Perform statistical analysis regularly (e.g., monthly) to evaluate changes in model performance. For example, through the analysis of 200 samples, the RMSE stabilized at 4.2 cm, and the MAE was 3.1 cm, representing reductions of 35% and 40% respectively compared to the traditional mechanistic model.

[0089] Statistical analysis allows for the quantitative evaluation of the model's predictive accuracy and performance stability. Indicators such as RMSE and MAE directly reflect the magnitude of the model's prediction error, and comparisons with traditional models highlight the advantages of this model. Regular evaluations can promptly identify declining performance trends, providing a basis for further model optimization and ensuring the model maintains consistently high predictive accuracy.

[0090] In dynamic error compensation, extreme condition corrections are provided: For extreme events such as icing and wildfires, historical conductor operation data (e.g., temperature, sag, meteorological parameters) and corresponding model prediction error data from these events are collected to establish a historical case library. This library is used to train a decision tree model. The model learns the relationship between meteorological parameters, conductor state parameters, and sag deviation patterns under extreme conditions to establish classification rules. When extreme conditions are detected (e.g., wildfires near the line causing abnormally high local temperatures), the decision tree model identifies the corresponding sag deviation pattern based on current operating parameters and automatically applies preset correction coefficients to adjust the model's predicted sag value. For example, when an abnormally high local temperature is detected, the model automatically applies the correction coefficients, increasing the predicted sag value by 15% to cope with increased thermal expansion.

[0091] Under extreme operating conditions, it can quickly and accurately correct the predicted sag of the model, make up for the large errors that may occur in the mechanism model under extreme conditions, ensure the accuracy of the prediction of conductor height change under extreme weather conditions, provide strong protection for the safe operation of the power grid, and avoid safety accidents caused by inaccurate prediction under extreme operating conditions.

[0092] In a specific application example, to implement the method for predicting changes in the height of ultra-high voltage transmission lines as described above (including 101-103 and S201-203), this application provides a system architecture. This system architecture includes a data layer, a model layer, and an application layer, which are explained in detail below: (1) The data layer is the foundation of the system and is responsible for collecting and storing various types of data related to changes in conductor height, including: real-time data: conductor current (SCADA), conductor temperature (thermal imaging), meteorological data (micro weather station), tension (tilt sensor); periodic data: high-precision conductor point cloud (UAV LiDAR); static data: line parameters (span, wire diameter, hanging point height, conductor type).

[0093] (2) The model layer is the core of the system. Based on the data provided by the data layer, it predicts the changes in conductor height through physical calculations and data-driven correction, and provides services to the application layer through APIs, including: Physics engine: calculates and predicts sag based on heat and state equations; AI correction module: compensates for physical model errors based on historical data; Adaptive calibration module: optimizes model parameters by using UAV data. (3) The application layer provides diversified application services for power grid operation and maintenance personnel and managers based on the prediction results provided by the model layer, so as to realize effective monitoring and management of conductor height changes and ensure the safe operation of the power grid, including real-time sag / distance to ground display; future trend prediction: combined with weather forecast to predict sag changes in the next 24-72 hours; safety early warning: linked with tree growth model to output accurate risk work orders; auxiliary decision: recommend dynamic capacity expansion limit or de-icing strategy.

[0094] Specifically, this application provides two application examples: future trend prediction and decision support.

[0095] Example 1: In the "Future Trend Prediction" example, the implementation steps include: obtaining the target sag data of the current conductor output by the conductor height change prediction model; obtaining the current conductor ground distance safety threshold and the preset maximum allowable sag data corresponding to the line design.

[0096] The application layer acquires weather forecast data (such as temperature, wind speed, and precipitation) for the next 24-72 hours and passes it as input parameters to the model layer. The model layer's physics engine combines the weather forecast data, static data, and current conductor operation data (such as current tension and temperature) to calculate the predicted conductor sag at different future time points. The application layer then organizes and analyzes this future sag prediction data to generate sag trend charts (such as line graphs and heat maps) to display the changing trend of conductor sag over a future period.

[0097] Predicting conductor height changes in advance allows maintenance personnel to anticipate potential abnormal sag situations (such as increased sag due to high temperatures). Based on the predictions, proactive measures can be taken (such as adjusting line loads and arranging vegetation trimming), shifting from reactive responses to proactive prevention and improving the foresight and effectiveness of power grid maintenance.

[0098] Example 2: Based on the "Future Trend Prediction" example, the "Assisted Decision" example is implemented to recommend dynamic capacity expansion limits. The implementation steps include: under the current meteorological conditions, combined with the target sag data of the current conductor, the safety threshold and the preset maximum allowable sag data, to make assisted decisions and recommend dynamic capacity expansion limits.

[0099] Dynamic capacity expansion refers to temporarily increasing the current carrying capacity of a conductor under specific meteorological conditions (such as low temperature and strong winds, which are conducive to conductor heat dissipation). Based on the conductor sag changes predicted by the model layer under different current loads, combined with the conductor-to-ground distance safety threshold and the maximum allowable sag in the line design, the application layer determines the maximum current increment the conductor can withstand under the current meteorological conditions, i.e., the dynamic capacity expansion limit. Simultaneously, considering factors such as the line's equipment capacity and protection settings, the capacity expansion limit is verified and adjusted, ultimately recommending a reasonable dynamic capacity expansion limit value to the dispatchers.

[0100] While ensuring the safe operation of the lines, we can fully tap their transmission potential, increase the power grid's transmission capacity, alleviate the power supply shortage during peak hours, eliminate the need for new lines, reduce power grid construction costs, and improve the economic efficiency of power grid operation.

[0101] Example 3: Building upon the "Future Trend Prediction" example, the "Decision Support" example recommends de-icing strategies and provides the following implementation steps: obtaining the target sag data of the current conductor from the output of the conductor height change prediction model; when icing is detected on the current conductor or icing weather is predicted, calculating the current icing thickness of the current conductor based on meteorological and tension data in the current conductor's operational status parameters and combined with the target sag data; and making decision support based on the current icing thickness to recommend a de-icing strategy.

[0102] When conductor icing is detected or icing weather is predicted, the application layer establishes an icing strategy evaluation model based on the current ice thickness (calculated from meteorological data and conductor tension and sag changes), conductor temperature, and future weather conditions calculated by the model layer, combined with the icing efficiency, energy consumption, and applicability of different icing methods (such as DC icing and AC icing). This model evaluates the feasibility, safety, and economy of different icing methods, such as calculating the icing time required, energy consumption, and impact on normal power supply to the line. Based on the evaluation results, the optimal icing strategy is recommended to maintenance personnel (e.g., recommending AC icing for thinner ice and DC icing for thicker ice).

[0103] To provide scientific decision support for conductor de-icing work, ensure that the de-icing work is carried out efficiently, safely and economically, remove ice from conductors in a timely manner, avoid excessive conductor tension and abnormal sag caused by ice, and even major safety accidents such as tower collapse and line breakage, and ensure the safe and stable operation of the power grid in icy weather.

[0104] Furthermore, as a response to the above Figure 1 , Figure 2 The present application provides a device for predicting changes in the height of ultra-high voltage (UHV) transmission lines, as illustrated in the method described. This device embodiment corresponds to the aforementioned method embodiment. For ease of reading, the details of the aforementioned method embodiment will not be repeated here, but it should be understood that the device in this embodiment can implement all the contents of the aforementioned method embodiment. This device is applicable even to situations where the height of UHV transmission lines exhibits dynamic and complex changes, providing a better solution for more accurately obtaining changes in the height of UHV transmission lines. Specifically, as shown... Figure 3 As shown, the device includes: The first acquisition unit 31 is used to acquire the current operating status parameter data and static parameter data corresponding to the conductor. The operating status parameter data includes at least conductor current data, conductor temperature data, meteorological data and tension data. The static parameter data includes at least line inherent parameter data and static geographical parameter data. The static geographical parameter data includes at least ground obstacle height data. Processing unit 32 is used to process the operating state parameter data and line inherent parameter data using a pre-constructed conductor height change prediction model, and output the target sag data of the current conductor; the conductor height change prediction model is a mechanism data fusion model constructed for the current conductor based on the heat balance equation, the state equation, the catenary equation and the data-driven correction algorithm; wherein, the heat balance equation, the state equation and the catenary equation are used to predict the real-time sag data of the current conductor based on the thermal and mechanical properties of the conductor, and the data-driven correction algorithm is used to correct and optimize the real-time sag data to obtain the target sag data; The determining unit 33 is used to determine the real-time height data corresponding to the current traverse based on the target sag data and the static geographic parameter data corresponding to the current traverse.

[0105] In some modified implementations, such as Figure 4 As shown, the processing unit 32 includes: The first determining module 321 is used to take the operating status parameter data and the line inherent parameter data of the current conductor as the initial condition data of the current conductor. The second determining module 322 is used to determine the heat generation and heat dissipation power of the current conductor by substituting the initial condition data into the heat balance equation. The execution module 323 is used to repeatedly adjust the initial condition data by iteratively adjusting the heat generation and heat dissipation power of the current conductor if the heat generation and heat dissipation power of the current conductor are not balanced, so as to obtain the real-time temperature data of the current conductor. The first acquisition module 324 is used to process the real-time temperature data of the current conductor by substituting it into the state equation and output the real-time stress data of the current conductor. The second acquisition module 325 is used to process the real-time stress data of the current conductor by substituting it into the catenary equation, and output the real-time sag data corresponding to the current conductor. The correction module 326 is used to obtain the target sag data predicted by the conductor height change prediction model from the real-time sag data corresponding to the current conductor using a data-driven correction algorithm.

[0106] In some modified implementations, such as Figure 4 As shown, the correction module 326 is specifically used for: The heat balance equation, the state equation, and the catenary equation corresponding to the current conductor are used as the mechanism model corresponding to the current conductor. Machine learning algorithms are used to learn the residuals of the machine model, which at least include the effects of creep, unmeasured micro-wind vibrations, and model simplification error factors on the prediction of conductor height changes. The residuals of the machine model are used as machine learning compensation to correct the real-time sag data to obtain the target sag data; and / or, During the operation of the conductor height change prediction model, the conductor point cloud data obtained by UAV lidar inspection is used to reverse-calibrate and correct the parameters of the mechanism model in order to obtain the target sag data.

[0107] In some modified implementations, such as Figure 4 As shown, the device further includes: The first acquisition unit 31 is also used to acquire weather forecast data for a future preset time range and add it to the operating status parameter data of the current conductor; The processing unit 32 is also used to process the current operating status parameter data of the conductor using the conductor height change prediction model, and output the target sag data at each preset future time point in the preset time range; The generation unit 34 is used to generate a trend chart of conductor sag changes corresponding to the preset time range based on each of the target sag data as the conductor sag prediction.

[0108] In some modified implementations, such as Figure 4 As shown, the device further includes: The second acquisition unit 35 is used to acquire the target sag data of the current conductor output by the conductor height change prediction model; The third acquisition unit 36 ​​is used to acquire the current conductor-to-ground distance safety threshold and the preset maximum allowable sag data corresponding to the line design. The first recommendation unit 37 is used to make auxiliary decisions and recommend dynamic capacity expansion limits based on the target sag data of the current conductor, the safety threshold and the preset maximum allowable sag data under the current meteorological conditions.

[0109] In some modified implementations, such as Figure 4 As shown, the device further includes: The second acquisition unit 35 is also used to acquire the target sag data of the current conductor output by the conductor height change prediction model; The calculation unit 38 is used to calculate the current icing thickness of the current conductor based on the meteorological data and tension data in the current conductor's operating status parameter data and the target sag data when icing is detected or icing weather is predicted. The second recommendation unit 39 is used to make auxiliary decisions based on the current ice thickness and recommend ice melting strategies.

[0110] In summary, the device for predicting changes in the height of ultra-high voltage transmission lines includes a processor and a memory. The first acquisition unit, processing unit, and determination unit mentioned above are all stored as program units in the memory, and the processor executes the program units stored in the memory to achieve the corresponding functions.

[0111] The processor contains a kernel that retrieves the corresponding program units from memory. One or more kernels can be configured, and by adjusting kernel parameters, a "direct sensing" method is used instead of direct measurement to acquire dynamic parameter data of the current conductor. Then, a conductor height change prediction model is used, integrating the conductor's thermal and mechanical properties with data-driven correction, to predict high-quality real-time conductor sag. Since the change in "conductor height" is directly determined by the change in sag, this allows for further high-precision results of conductor height changes. Even when dealing with the dynamic and complex changes in the height of UHV transmission conductors, this provides a better solution for more accurately obtaining the height changes of UHV transmission conductors.

[0112] This application provides a computer-readable storage medium storing a computer program that, when executed by a processor, implements the method for predicting changes in the height of ultra-high voltage transmission lines as described above.

[0113] This application provides an electronic device, which includes at least one processor, at least one memory and a bus connected to the processor; wherein the processor and the memory communicate with each other through the bus; the processor is used to call program instructions in the memory to execute the method for predicting changes in the height of ultra-high voltage transmission lines as described above.

[0114] This application also provides a computer program product that, when executed on a data processing device, is adapted to perform the steps of a method for initializing the prediction of changes in the height of ultra-high voltage transmission lines.

[0115] This application is described with reference to flowchart illustrations and / or block diagrams of methods, apparatus (systems), and computer program products according to embodiments of this application. It will be understood that each block of the flowchart illustrations and / or block diagrams, and combinations of blocks in the flowchart illustrations and / or block diagrams, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, special-purpose computer, embedded processor, or other programmable data processing apparatus to produce a machine, such that the instructions, which execute via the processor of the computer or other programmable data processing apparatus, generate instructions for implementing the flowchart... Figure 1 One or more processes and / or boxes Figure 1 A device that provides the functions specified in one or more boxes.

[0116] In a typical configuration, the device includes one or more processors (CPUs), memory, and a bus. The device may also include input / output interfaces, network interfaces, etc.

[0117] Memory may include non-persistent memory in computer-readable media, such as random access memory (RAM) and / or non-volatile memory, like read-only memory (ROM) or flash RAM, and memory includes at least one memory chip. Memory is an example of computer-readable media.

[0118] Computer-readable media includes both permanent and non-permanent, removable and non-removable media that can store information using any method or technology. Information can be computer-readable instructions, data structures, modules of programs, or other data. Examples of computer storage media include, but are not limited to, phase-change memory (PRAM), static random access memory (SRAM), dynamic random access memory (DRAM), other types of random access memory (RAM), read-only memory (ROM), electrically erasable programmable read-only memory (EEPROM), flash memory or other memory technologies, CD-ROM, digital versatile optical disc (DVD) or other optical storage, magnetic tape, magnetic magnetic disk storage or other magnetic storage devices, or any other non-transferable medium that can be used to store information accessible by a computing device. As defined herein, computer-readable media does not include transient computer-readable media, such as modulated data signals and carrier waves.

[0119] It should also be noted that the terms "comprising," "including," or any other variations thereof are intended to cover non-exclusive inclusion, such that a process, method, article, or apparatus that comprises a list of elements includes not only those elements but also other elements not expressly listed, or elements inherent to such process, method, article, or apparatus. Unless otherwise specified, an element defined by the phrase "comprising one..." does not exclude the presence of other identical elements in the process, method, article, or apparatus that includes that element.

[0120] Those skilled in the art will understand that embodiments of this application can be provided as methods, systems, or computer program products. Therefore, this application can take the form of a completely hardware embodiment, a completely software embodiment, or an embodiment combining software and hardware aspects. Furthermore, this application can take the form of a computer program product embodied on one or more computer-usable storage media (including, but not limited to, disk storage, CD-ROM, optical storage, etc.) containing computer-usable program code.

[0121] The above are merely embodiments of this application and are not intended to limit the scope of this application. Various modifications and variations can be made to this application by those skilled in the art. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of this application should be included within the scope of the claims of this application.

Claims

1. A method for predicting the height variation of ultra-high voltage transmission lines, characterized in that, The method includes: Obtain the current operating status parameter data and static parameter data corresponding to the conductor. The operating status parameter data includes at least conductor current data, conductor temperature data, meteorological data, and tension data. The static parameter data includes at least line inherent parameter data and static geographical parameter data. The static geographical parameter data includes at least ground obstacle height data. The operating state parameter data and the line's inherent parameter data are processed using a pre-constructed conductor height change prediction model to output the target sag data of the current conductor. The conductor height change prediction model is a mechanism data fusion model constructed for the current conductor based on the heat balance equation, the state equation, the catenary equation, and a data-driven correction algorithm. The heat balance equation, the state equation, and the catenary equation are used to predict the real-time sag data of the current conductor based on its thermal and mechanical properties. The data-driven correction algorithm is used to correct and optimize the real-time sag data to obtain the target sag data. Based on the target sag data and the static geographic parameter data corresponding to the current traverse, determine the real-time height data corresponding to the current traverse.

2. The method according to claim 1, characterized in that, The process of using a pre-built conductor height change prediction model to process the operating status parameter data and the inherent line parameter data, and outputting the current conductor's target sag data, includes: The current conductor's operating status parameter data and the line's inherent parameter data are used as the current conductor's initial condition data. By substituting the initial condition data into the heat balance equation, the heat generation and heat dissipation power of the current conductor are determined; If the heat generation and heat dissipation power of the current conductor are not in balance, the initial condition data is adjusted iteratively until the heat generation and heat dissipation power of the current conductor are in balance using the heat balance equation, so as to obtain the real-time temperature data of the current conductor. By substituting the real-time temperature data of the current conductor into the state equation for processing, the real-time stress data of the current conductor is output. By substituting the real-time stress data of the current conductor into the catenary equation for processing, the real-time sag data corresponding to the current conductor is output. The target sag data is obtained by using a data-driven correction algorithm on the real-time sag data corresponding to the current conductor and the output of the conductor height change prediction model.

3. The method according to claim 2, characterized in that, The step of using a data-driven correction algorithm to obtain the target sag data predicted by the conductor height change prediction model from the real-time sag data corresponding to the current conductor includes: The heat balance equation, the state equation, and the catenary equation corresponding to the current conductor are used as the mechanism model corresponding to the current conductor. Machine learning algorithms are used to learn the residuals of the machine model, which at least include the effects of creep, unmeasured micro-wind vibrations, and model simplification error factors on the prediction of conductor height changes. The residuals of the machine model are used as machine learning compensation to correct the real-time sag data to obtain the target sag data; and / or, During the operation of the conductor height change prediction model, the conductor point cloud data obtained by UAV lidar inspection is used to reverse-calibrate and correct the parameters of the mechanism model in order to obtain the target sag data.

4. The method according to any one of claims 1 to 3, characterized in that, The method further includes: Obtain weather forecast data for a future preset time range and add it to the current conductor's operating status parameter data; The current operating status parameter data of the conductor is processed using the conductor height change prediction model, and the target sag data at each preset future time point within the preset time range is output. Based on each target sag data as the predicted sag of the conductor, a trend chart of conductor sag change corresponding to the preset time range is generated.

5. The method according to any one of claims 1 to 3, characterized in that, The method further includes: Obtain the target sag data of the current conductor output by the conductor height change prediction model; Obtain the current conductor-to-ground distance safety threshold and the preset maximum allowable sag data corresponding to the line design; Under current meteorological conditions, a dynamic capacity expansion limit is recommended by combining the target sag data of the current conductor, the safety threshold, and the preset maximum allowable sag data for auxiliary decision-making.

6. The method according to any one of claims 1 to 3, characterized in that, The method further includes: Obtain the target sag data of the current conductor output by the conductor height change prediction model; When icing is detected on the current conductor or icing weather is predicted, the current icing thickness of the current conductor is calculated based on the meteorological data and tension data in the current conductor's operating status parameter data, as well as the target sag data. Based on the current ice thickness, auxiliary decisions are made, and ice-melting strategies are recommended.

7. A device for predicting changes in the height of ultra-high voltage transmission lines, characterized in that, The device includes: The first acquisition unit is used to acquire the current operating status parameter data and static parameter data corresponding to the conductor. The operating status parameter data includes at least conductor current data, conductor temperature data, meteorological data and tension data. The static parameter data includes at least line inherent parameter data and static geographical parameter data. The static geographical parameter data includes at least ground obstacle height data. The processing unit is used to process the operating state parameter data and the line inherent parameter data using a pre-built conductor height change prediction model, and output the target sag data of the current conductor. The conductor height change prediction model is a mechanism data fusion model constructed for the current conductor based on the heat balance equation, the state equation, the catenary equation, and the data-driven correction algorithm. The heat balance equation, the state equation, and the catenary equation are used to predict the real-time sag data of the current conductor based on the thermal and mechanical properties of the conductor, and the data-driven correction algorithm is used to correct and optimize the real-time sag data to obtain the target sag data. The determining unit is used to determine the real-time height data corresponding to the current traverse based on the target sag data and the static geographic parameter data corresponding to the current traverse.

8. The apparatus according to claim 7, characterized in that, The processing unit includes: The first determining module is used to take the operating state parameter data of the current conductor as the initial condition data of the current conductor; The second determining module is used to determine the heat generation and heat dissipation power of the current conductor by substituting the initial condition data into the heat balance equation. The execution module is used to iteratively adjust the initial condition data if the heat generation and heat dissipation power of the current conductor is not balanced, until the heat generation and heat dissipation power of the current conductor is balanced by the heat balance equation, so as to obtain the real-time temperature data of the current conductor. The first acquisition module is used to process the real-time temperature data of the current conductor by substituting it into the state equation and output the real-time stress data of the current conductor. The second acquisition module is used to process the real-time stress data of the current conductor by substituting it into the catenary equation, and output the real-time sag data corresponding to the current conductor. The correction module is used to obtain the target sag data predicted by the conductor height change prediction model from the real-time sag data corresponding to the current conductor using a data-driven correction algorithm.

9. A computer-readable storage medium, characterized in that, The computer-readable storage medium stores a computer program that, when executed by a processor, implements the method for predicting changes in the height of ultra-high voltage transmission lines as described in any one of claims 1-6.

10. An electronic device, characterized in that, The device includes at least one processor, and at least one memory and bus connected to the processor; The processor and the memory communicate with each other via the bus. The processor is used to call program instructions in the memory to execute the method for predicting the height change of ultra-high voltage transmission lines as described in any one of claims 1-6.

Images